Preface In three decades or so of widespread use, antibiotics have wrought a revolution in the medical, veterinary, and ...
26 downloads
1093 Views
46MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Preface In three decades or so of widespread use, antibiotics have wrought a revolution in the medical, veterinary, and agricultural sciences; indeed, in all of the biological sciences. Early research on antibiotics was necessarily directed at the production, isolation, characterization, and pharmacology of this important class of natural products. It was soon apparent that microorganisms could develop resistance to antibiotics and that this resistance was due, at least in part, to the possession of enzymes that could, in some fashion, chemically modify the antibiotic. TO keep ahead (or even abreast) of antibiotic resistance, it is necessary either to constantly discover new antibiotics or to develop derivatives that are insensitive to the enzymes that cause inactivation of the natural compound. Both avenues have been tried. It is evident that the search for new antibiotics must eventually reach the point of diminishing returns and that the second approach offers the best hope of extendiag the life of an antibiotic. Research on the biosyntheses of antibiotics as well as on enzymatic means of degradation has received less emphasis than other aspects of antibiotic chemistry, but, as it has become apparent that a knowledge of biosynthetic pathways can assist in isolating intermediate compounds capable of being modified chemically, this area has received more attention. Similarly, degradative enzymes have been found that are capable of providing antibiotic derivatives which can be chemically modified, allowing production of large numbers of semisynthetic antibiotics. Other practical uses of antibiotic enzymology include the use of enzymes as analytical reagents in determining the concentrations of antibiotics in samples and in the effective removal of antibiotics from reaction mixtures by converting them to inactive compounds. Aside from the utilitarian aspects of antibiotic enzymology a principal driving force behind research on these enzymes is the intellectual curiosity as to the raison d'etre of antibiotic synthesis. The production of these secondary metabolites, which serve no evident function in the producing organism, requires large quantities of energy and considerable metabolic machinery. In some instances very complicated pathways involving 20-30 enzymes are required to synthesize an antibiotic. As the pathways are unraveled and the branch points with normal metabolic routes are established, light may be shed on the mechanisms of antibiotic synthesis. Even with the renewed interest in antibiotic enzymology, the extent and scope of research in this area are uneven, and the published results are scattered throughout the scientific literature. The aim of this volume is to collate in one source as much information concerning antibiotic xiii
xiv
PREFACE
enzymology as possible. The work is divided into three sections. The first is concerned with methods used in the study of antibiotics, and covers techniques from culturing the producing organism to various chromatographic methods to sophisticated physical techniques. The second and third sections are devoted to enzymes involved in antibiotic biosynthesis and antibiotic degradation and modification, respectively. In some cases the division between the second and third section is quite arbitrary because it is not always clear whether an enzyme belongs in a biosynthetic or degradative pathway. The coverage of enzymes represents the state of the art of antibiotic enzymology; the range extends from pure enzymes that have been sequenced to enzymes that have been studied only in crude extracts. Many other enzymes that act on antibiotics, antibiotic precursors, or antibiotic derivatives have been detected in extracts or whole cells. Most of these had to be omitted because of a paucity of information. It is evident that only a small part of antibiotic enzymology has reached the stage where it can be consolidated into a treatise of this kind, and it is hoped that this volume will serve as a stimulus for further research on the enzymes involved with this important class of compounds. I am indebted to many people for many ideas and suggestions, but I am especially indebted to investigators in pharmaceutical laboratories throughout the world for their ideas and contributions. JOHN H. HASH
Contributors to Volume X L I I I Article numbers are in parentheses following the names of contributors. Affiliations listed are current. ABBOTT ( 5 5 b ) , Fermentation Cambridge, Massachusetts Products Research, Lilly Research JOHN E. DOWDING (48), Department o] Biochemistry, The University o] WisLaboratories, Eli Lilly and Company, Indianapolis, Indiana consin, Madison, Wisconsin COLETTE DUEZ (53f), Institut de BotaE. P. ABRAHAM (29, 55a), Sir William nique, Service de Microbiologie, UniDunn School o] Pathology, University o] Ox/ord, Ox]ord, England versitg de Liege, Sart Tilman, Liege, ADORJAN ASZALOS (8), The Squibb InstiBelgium tute ]or Medical Research, New E. F. ELSTNER (37), Department o] BioBrunswick, New Jersey, and Princeton logical Organic Chemistry, Albert Ein,~tein Medical Center, Philadelphia, University, Princeton, New Jersey MOHINDER S. BATHALA (17), College o/ Pennsylvania Pharmacy, The Ohio State University, AMEI)EO A. FANTINI (2), Department o] Microbiology, Lederle Laboratories, Columbus, Ohio Pearl River, New York VLADIMIR BETINA (7), Department o] PATRICIA FAWCETT (29, 55a), Sir William Technical Microbiology and BiochemDunn School o] Pathology, University istry, Faculty o] Chemistry, Slovak o] Ox]ord, Ox]ord, England PoIytechnical University, Bratislava, HEINZ G. FLOSS (34), Department o] Czechoslovakia Medicinal Chemistry and PharmacogDONALD B. BORI)ERS (10), Department o] nosy, School o] Pharmacy and PharFermentation and Isolation, Lederle macal Sciencies, Purdue University, Laboratories, Pearl River, New York West La]ayette, Indiana EBERHARD BREUKER (41), Lohman and Company AG, Neu]elder Strasse, JEAN-MARIE FRERE (53f), Institut de Botanique, Service de Microbiologie, 219 Cuxhaven, Germany Universit~ de Liege, Sart Tilman, RICHA~ BaU~NEa (31), Institut ]iir BioLiege, Belgium chemische Technologie und Mikrobiologie, Technischen HochschuIe DAvm FROST (8), The Squibb Institute ]or Medical Research, New Brunswick, Wien, Wien Getreidemarkt, Austria New Jersey JOHN H. COATS (58), Research Laboratories, The Upjohn Company, Kala- DAVID S. FUKUDA (55b), Lilly Research Laboratories, Eli Lilly and Company, mazoo, Michigan Indianapolis, Indiana M. COLE (54a, 54b), Research Division, STEN GATENBECK (30), Division o] Pure Beecham Pharmaceuticals Betchand Applied Biochemistry, The Royal worth, Surrey, England Institute o] Technology, Stockholm, JOHN W. CORCORAN (33), Department o] Biochemistry, Northwestern UniverSwede~ sity School o] Medicine, Chicago, G. M. GAUCHER (40), Department o] Chemistry, The University of Calgary, Illinois Calgary, Alberta, Canada LYMAN C. CRAIG (16), The Rocke]eUer JEAN-MARIE GHUYSEN (53f), Institnt de University, New York, New York Botaniq~te, Service de Microbiologie, JULIAN DAVIES (3), Department o] BioUniversitg de Liege, Sart Tilman, chemistry, The University o/ WisconLihge, Belgium sin, Madison, Wisconsin ARNOLD L. DEMAIN (52), Department o] MICHAEL J. HAAS (48), Department o] Biochemistry, The University o] WisNutrition and Food Science, Massaconsin, Madison, Wisconsin chusetts Institute of Technology,
BERNARD J.
ix
X
CONTRIBUTORS TO VOLUME XLIII
Inc., Nutley, New Jersey Microbiology, Vanderbilt University, LESTER A. MITSCHER (17), College o/ Pharmacy, The Ohio State University, School o/ Medicine, Nashville, Columbus, Ohio Tennessee GERHARD HEINRICH (41), Case Western NORBERT NEUSS (20), Lilly Research Laboratories, Eli Lilly and Company, Reserve School o] Medicine, CleveIndianapolis, Indiana land, Ohio ULFERT HORNEMANN (34), Department CYNTHIA H. O'CALLAGHAN (5), Bacterial Chemotherapy Unit, Glaxo Research o] Medicinal Chemistry and PharmaLimited, Green/ord, Middlesex, cognosy, School o] Pharmacy and England Pharmacal Sciences, Purdue UniverSEAN C. O'CONNOR (14), Lilly Research sity, West La]ayette, Indiana Laboratories, Eli Lilly and Company, EDWARD INAMINE (52), Developmental Indianapolis, Indiana Microbiology Department, Merck Sharp and Dohme Research Labora- HENRY PAULUS (44), Department o/Biological Chemistry, Harvard Medical tories, Rahway, New Jersey School, Boston, Massachusetts KENNETH JOHNSON (53f), Biochemistry Laboratory, National Research Council D. PERLMAN (60, 61), School o] Pharmacy, The University o/ Wisconsin, o] Canada, Ottawa, Canada Madison, Wisconsin FREDERICK KAVANAGH (4), 231 Blue PETER PFAENDER (41), Institut ]~r BioRidge Road, Indianapolis, Indiana logische Chemie und ErntihrungswisL. A. KOMINEK (35), Fermentation Resenscha/t, Universitat Hohenheim, search and Development, The Upjohn Hohenheim, Germany Company, Kalamazoo, Michigan SHINICHI KONDO (11, 12), Institute o] BURTON M. POGELL (36), Department o/ Microbiology, St. Louis University Microbial Chemistry, Shinagawa-ku, School o] Medicine, St. Louis, MisTokyo, Japan souri ZOFIA KUR,.~_~o-BoRoWSKA (42), The Rocke]eller University, New York, JOHN N. PORTER (1), Department o] Microbiology, Lederle Laboratories, New York Pearl River, New York SCREN G. LALAND (43), Department o] Biochemistry, University o] Oslo, Oslo, M. H. RICHMOND (6, 53C, 53d), Department o/ Bacteriology, The Medical Norway School, University o/ Bristol, Bristol, SUNG G. LEE (45), The Rocke]eller UniEngland versity, New York, New York ROBLEY J. LIGHT (39), Department o] HANSPETER RIEDER (41), Institut ]i~r Biologische Chemie und ErntihrungswisChemistry, Florida Stale University, senscha]t, Universittit Hohenheim, Tallahasee, FloT~da Hohenheim, Germany FRITZ LIPMANN (45), The Rocke]elIer JOHN H. ROBERTSON (9), Fermentation University, New York, New York Research and Development, The UpF. LYNEN (38), Max-Planck-Institut ]~r john Company, Kalamazoo, Michigan Biochemie, Munich, Germany GARY G. MARCONI (13), Lilly Research MAX RSHR (31), Institut ]~r Biochemische Technologie und MikrobioIogie, Laboratories, Eli Lilly and Company, Technische Hochschule Wien, Wien, Indianapolis, Indiana Austria It. F. MEYER (35), Fermentation Research and Development, The Upjohn GORDON W. Ross (5, 53e), Glaxo Research Limited, Green]ord, Middlesex, Company, Kalamazoo, Michigan England PHmIP A. MILLER (46, 47), Department o] Microbiology, Ho]]mann-La Roche W. A. SAVIDGE (54a, 54b), Biochemical
JOHN H. HASH (46, 47), Department o/
CONTRIBUTORS TO VOLUME X L I I I
Services Unit, Beecham Pharmaceuticals, Betchworth, Surrey, England W. V. SHAW (57), Department of Biochemistry, University of Leicester, Leicester, England HOWARD SIEGEBMAN (18), Princeton Applied Research Corporation, Princeton, New Jersey MAHAVIR M. SIMLOT (41), Agricultural Experimental Station, University of Udaipur, Udaipur (Rajasthan), India GEORGE SLOMP (19), Fermentation Research and Development, The Upjohn Company, Kalamazoo, Michigan JOHN SOGN (16), The Rockefeller University, New York, New York THEODORE S. SOKOLOSKI (17), College of Pharmacy, The Ohio State University, Columbus, Ohio MARILYN K. SPEEDIE (34), School of Pharmacy, Oregon State University, Corvallis, Oregon BRIAN SPENCER (32), Department of Biochemistry, University of Dublin, Trinity College, Dublin, Ireland R. J. SUHADOLNIK (37, 59), Department of Biological Organic Chemistry, Albert Einstein Medical Center, Philadelphia, Pennsylvania DAVID ~. THATCHER (53a, 53b), Department of Molecular Biology, University
xi
of Edinburgh, Edinburgh, Scotland KivosHi T s w i (9, 15), Control Analyti-
cal Research and Development, The Upjohn Company, Kalamazoo, Michigan W. UEMATSU (59), Department of Biological Organic Chemistry, Albert Einstein Medical Center, Philadelphia, Pennsylvania HAMAO UMEZAWA (11, 12), Institute of Microbial Chemistry, Shinagawa-ku, Tokyo, Japan HUBERT VANDERHAEGHE (54a, 54c), Rega Institute, University of Leuven, Leuven, Belgium L. C. VINING (56), Department of Biology, Dalhonsie University, Halifax, Nova Scotia, Canada GiiTNTEB VOGEL (38, 39), Max-PIanckInstitute ]iir Biologie, Tubingen, Germany JAMES B. WALKER (21, 22, 23, 24, 25, 26, 27, 28, 49, 50, 51), Department of Bio-
chemistry, Rice University, Houston, Texas MAR6ABET S. WALKER (50, 51), Department of Biochemistry, Rice University, Houston, Texas TRINE-LIsE ZIMMER (43), Department of Biochemistry, University of Oslo, OsIo, Norway
METHODS IN ENZYMOLOGY EDITED
BY
Sidney P. Colowick and N a t h a n O. Kaplan DEPARTMENT OF C H E M I S T R Y
VANDERBILT UNIVERSITY
UNIVERSITY OF CALIFORNIA
SCHOOL OF MEDICINE
AT SAN DIEGO
NASHVILLE, T E N N E S S E E
LA JOLLA, CALIFORNIA
I. II. III. IV. V. VI.
Preparation and Assay of Enzymes Preparation and Assay of Enzymes Preparation and Assay of Substrates Special Techniques for the Enzymologist Preparation and Assay of Enzymes Preparation and Assay of Enzymes (Continued) Preparation and Assay of Substrates Special Techniques VII. Cumulative Subject Index
XV
METHODS IN ENZYMOLOGY EDITORS-IN-CHIEF
Sidney P. Colowick
Nathan 0. Kaplan
VOLUMEVIII. Complex Carbohydrates
Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG VOLUME IX. Carbohydrate Metabolism
Edited by WILLIS A. WOOD VOLUME X. Oxidation and Phosphorylation
Edited by RONALDW. ESTABROOKAND MAYNARDE. PULLMAN VOLUME XI. Enzyme Structure
Edited by C. H. W. HIRS VOLUME XII. Nucleic Acids (Parts A and B)
Edited by LAWRENCEGROSSMANAND KIVIE MOLDAVE VOLUME XIII. Citric Acid Cycle
Edited by J. M. LOWENSTEIN VOLUME XIV. Lipids Edited by J. M. LOWENSTEIN VOLUME XV. Steroids and Terpenoids
Edited by RAYMOND B. CLAYTON VOLUME XVI. Fast Reactions
Edited by KENNETH KUSTIN VOLUME XVII. Metabolism of Amino Acids and Amines (Parts A and B) Edited by HERBERTTABORANDCELIAWHITE TABOR VOLUMEXVIII. Vitamins and Coenzymes (Parts A, B, and C)
Edited by DONALDB. MCCORMICKAND LEMUELD. WRIGHT VOLUME XIX. Protcolytic Enzymes
Edited by GERTRUDEE. PERLMANNAND LASZLOLORAND VOLUME XX. Nucleic Acids and Protein Synthesis (Part C)
Edited by KIVIE MOLDAVEAND LAWRENCEGROSSMAN xvii
xviii
METHODS
IN
ENZYMOLOGY
VOLUME XXI. Nucleic Acids (Part D)
Edited by LAWRENCEGROSSMANAND KIVIE MOLDAVE VOLUME XXII. Enzyme Purification and Related Techniques Edited by WILLIAMB. JAKOBY VOLUME XXIII. Photosynthesis (Part A)
Edited by ANTHONY SAN PIETRO VOLUMEXXlV. Photosynthesis and Nitrogen Fixation (Part B) Edited by ANTHONY SAN PIETRO VOLUME XXV. Enzyme Structure (Part B)
Edited by C. H. W. HIRS AND SERGEN. TIMASHEFF VOLUME XXVI. Enzyme Structure (Part C)
Edited by C. H. W. HIRS AND SERGEN. TIMASHEFF VOLUME XXVII. Enzyme Structure (Part D)
Edited by C. H. W. HIRS AND SERGEN. TIMASHEFF VOLUME XXVIII. Complex Carbohydrates (Part B)
Edited by VICTOR GINSBURG VOLUMEXXIX. Nucleic Acids and Protein Synthesis (Part E)
Edited by LAWRENCEGROSSMANAND KIVlE MOLDAVE
VOLUMEXXX.
Nucleic Acids and Protein Synthesis (Part F)
Edited by KIVlE MOLDAVEAND LAWRENCEGROSSMAN VOLUME XXXI, Biomembranes (Part A) Edited by SIDNEY FLEISCHERAND LESTER PACKER VOLUME XXXII. Biomembranes (Part B)
Edited by SIDNEY FLEISCHERAND LESTER PACKER VOLUMEXXXIII. Cumulative Subject Index Volumes I-XXX
Edited by MARTHAG. DENNIS AND EDWARDA. DENNIS VOLUMEXXXIV. Affinity Techniques (Enzyme Purification: Part B) Edited by WILLIAMB. JAKOBYAND MEIR WILCHEK
METHODS IN ENZYMOLOGY
xix
VOLUME XXXV. Lipids (Part B) Edited by JOHN M. LOWENSTEIN VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones)
Edited by BERT W. O'MALLEYAND JOEL G. HARDMAN VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones)
Edited by BERT W. O'MALLEYANDJOEL G. HARDMAN VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides) Edited by JOEL G. HARDMANANDBERT W. O'MALLEY VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues, and Organ Systems) Edited by JOEL G. HARDMANANDBERT W. O'MALLEY VOLUMEXL. Hormone Action (Part E: Nuclear Structure and Function)
Edited by BERT W. O'MALLEYAND JOEL G. HARDMAN VOLUME XLI. Carbohydrate Metabolism (Part B)
Edited by W. A. WOOD VOLUME XLII. Carbohydrate Metabolism (Part C)
Edited by W. A. WooD VOLUME XLIII. Antibiotics
Edited by JOHN H. HASH
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
3
[1] Cultural Conditions for Antibiotic-Producing Microorganisms By
JOHN N. PORTER
I. II. III. IV. V.
Introduction . . . . . . . . . . . . . . . . . . Culture Maintenance . . . . . . . . . . . . . . . . Culture Preservation . . . . . . . . . . . . . . . . . Antibiotic Production in Liquid Culture . . . . . . . . . . Selected Antibiotic Fermentations by Fungi . . . . . . . . . A. Cephalosporin C . . . . . . . . . . . . . . . . B. Griseofulvin . . . . . . . . . . . . . . . . . C. Penicillin . . . . . . . : . . . . . . . . . . VI. Selected Antibiotic Fermentations by Eubacteria . . . . . . . A. Bacitracin . . . . . . . . . . . . . . . . . . B. Polymyxins . . . . . . . . . . . . . . . . . VII. Selected Antibiotic Fermentations by Actinomycetes . . . . . . A. Chloralnphenicol . . . . . . . . . . . . . . . . B. Chlortetracycline . . . . . . . . . . . . . . . . C. Erythromycin . . . . . . . . . . . . . . . . . D. Gentamicin . . . . . . . . . . . . . . . . . E. Neomycin . . . . . . . . . . . . . . . . . . F. Streptomycin . . . . . . . . . . . . . . . . .
3 3 8 II 14 14 15 16 16 16 17 18 18 19 20 21
22 23
I. I n t r o d u c t i o n Of t h e a p p r o x i m a t e l y 3000 k n o w n a n t i b i o t i c s a b o u t 70% are d e r i v e d f r o m a c t i n o m y c e t e s , e s p e c i a l l y species of Streptomyces, 2 0 % f r o m fungi, a n d 10% f r o m e u b a c t e r i a . M o s t of the p r o d u c i n g o r g a n i s m s , b u t b y no m e a n s all, were i s o l a t e d f r o m soils in p r o g r a m s designed to d i s c o v e r new a n t i b i o t i c s . Some of t h e m o r e p r o m i n e n t c u l t u r e collections f r o m which m a n y of these o r g a n i s m s are o b t a i n a b l e a r e t h e following: A T C C - A m e r i c a n T y p e C u l t u r e Collection, R o c k v i l l e , M a r y l a n d ; C B S - - C e n traalbureau voor Schimmelcultures, Baarn, The Netherlands ; CMI--Comm o n w e a l t h M y c o l o g i c a l I n s t i t u t e , K e w , E n g l a n d ; I F O - - I n s t i t u t e for F e r m e n t a t i o n , O s a k a , J a p a n ; N C I B - - N a t i o n a l C o l l e c t i o n of I n d u s t r i a l Bacteria, Aberdeen, Scotland ; NRRL--Northern Utilization Research and D e v e l o p m e n t D i v i s i o n , U.S. D e p a r t m e n t of A g r i c u l t u r e , P e o r i a , Illinois. II. Culture Maintenance
The g r o w t h r e q u i r e m e n t s of m i c r o o r g a n i s m s cover a v e r y wide r a n g e of conditions. H o w e v e r , few a n t i b i o t i c p r o d u c e r s are p a r t i c u l a r l y f a s t i d i ous, a n d m o s t can be r e a d i l y c u l t i v a t e d b y t h e p r o p e r selection of m e d i a a n d o t h e r e n v i r o n m e n t a l factors.
4
METHODS FOR T H E STUDY OF ANTIBIOTICS
[1]
Culture Media. Solid media for maintaining eubacteria commonly contain proteins or peptones as sources of both carbon and nitrogen. Media for growing fungi, on the other hand, more often contain a carbohydrate to supply carbon, and nitrates or ammonium salts to supply nitrogen. Actinomycetes, which are related to the eubacteria but not to fungi, are ordinarily grown on media containing both a carbohydrate and an organic nitrogen source. Since spores have more prolonged viability than vegetative cells, a maintenance medium should encourage sporulation rather than vegetative growth. Czapek's agar, for example, serves as a good medium for species of Penicillium and Aspergillus, fungi which grow sparsely on this medium but sporulate readily. I n the preparation of media containing agar, the media should be heated for 15-30 min at 100 ° in order to melt the agar before dispensing. Sterilization of small volumes of media, such as in test tubes and small flasks, is carried out in an autoclave at 120 ° for 15 min at 15 psi. Large volumes require longer sterilization times. Formulations for some of the most commonly used agar media are given below. Some m a y be obtained commercially and should be prepared as directed on the container. M E D I A FOR MAINTAINING FUNGI 1,2
Potato dextrose agar Potatoes, peeled and sliced Distilled water
300 g 1000 ml
Boil the potatoes or heat momentarily in 500 ml of water in an autoclave at 121°. Filter through cheesecloth. Make up the volume to 1000 ml and add Glucose Agar
20 g 15 g
Heat to boiling, mix, then dispense and autoclave. Czapek's solution agar NaNOa K~HPO4 MgSO4 -7H20 KCI FeSO4 •7H20 Agar Distilled water to volume
(g/liter) 3 1 0.5 0.5 0.01 15
Melt and add sucrose 30 g/liter, pH not adjusted. 1F. A. Weiss, in "Manual of Microbiological Methods," p. 99. McGraw-Hill, New York, 1957. *W. C. Haynes, L. J. Wickerham, and C. W. Hesseltine, Appl. Microbiol. 3, 361 (1955).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
5
Malt extract agar
Malt extract Glucose Peptone Agar Distilled water to volume p H n o t adjusted
(g/liter) 20 20 1 20
MEDIA FOR MAINTAINING EUBACTERIA3
Nutrient agar
(g/liter) 3 5 15
Beef extract Peptone Agar The addition of yeast extract (5 g/liter) is optional. TG Y a#ar
Tryptone Yeast extract Glucose K~HPO4 Agar Adjust to pH 7.0
(g/liter) 5 5 1 1 2O
MEDIA FOR MAINTAINING ACTINOMYCETES4'5
Yeast extract agar
Yeast extract Malt extract Glucose Agar Distilled water to volume Adjust to pH 7.3 with NaOH
(g/liter) 4 10 4 20
3 The American Type Culture Collection. Catalogue of Strains, 10th ed., 1972. T. G. Pridham, P. Anderson, C. Foley, L. A. Lindenfelser, C. W. Hesseltine, and R. G. Benedict, Antibiot. Annu. 1956/1957, p. 947 (1957). 5S. T. Williams and T. Cross, in "Methods in Microbiology" (J. R. Norris and D. W. Ribbons, eds.), Vol. 4, p. 295. Academic Press, New York, 1971.
6
METHODS FOR THE STUDY OF ANTIBIOTICS
[1]
Tomato paste oatmeal agar
(g/liter) Solution 1 Heinz baby oatmeal Contadina fancy tomato paste Add to 500 ml of boiling tap water Solution 2 Agar Tap water
20 20
15 500 ml
Melt agar by steaming at 100° for 15-20 min. Mix the two solutions, steam at 100 ° for 10 rain, dispense, and sterilize. No pH adjustment. Oatmeal agar
Oatmeal Agar
(g/liter) 20 18
Cook or steam the oatmeal in 1 liter of distilled water for 20 min. Filter through cheesecloth. Add water to restore volume to 1 liter. Add 1 ml of trace salts solution. Adjust pH to 7.2 with NaOH and add agar. Pridham and Gottlieb trace salts solution in g/liter: CuSO4.5H20, 0.64; FeSO4.7H~O, 0.11; MnC12- 4H~O, 0.79; ZnS04.7H~O, 0.15. Bennett's agar
Yeast extract Beef extract N-Z Amine A (Casein digest: Sheffield Farms) Glucose Agar Distilled water to volume
(g/liter) 1 1 2 10 15
Trypticase-yeast extract agar 6 (Recommended for thermophilic actinomycetes)
Trypticase Yeast extract Sucrose Dung extract Molasses MgSO4 - 7H20 FeSO4.7H~O Microelement solution Agar Distilled water to volume Adjust pH 7.0-7.2
(g/liter) 5 3 5 5 ml 5 ml 0.5 0.01 1 ml 20
Dung extract: suspend 25% dried sheep manure in tap water, autoclave for 30 min, filter, refrigerate under toluene. M. D. Tendler and P. R. Burkholder, Appl. Microbiol. 9, 394 (1961).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
7
Microelement stock solution (per ml): Fe [as Fe(NH4)2(SO4)2], 1.0 mg; Zn (as ZnSO4), 1.0 mg; ]V[n (as MnSO4), 0.5 mg; Cu (as CuSO~), 0.08 rag; Co (as COSO4), 0.1 rag; B (as H3BO3), 0.1 mg. N - Z amine-starch-glucose medium ~
(Recommended for maintaining Micromonospora)
Glucose Soluble starch Yeast extract N-Z Amine A (Casein digest; Sheffield Farms) CaCO~ Agar Distilled water to volume
(g/liter) 10 20 5 5 1 15
Temperature. Most antibiotic-producing microorganisms are mesophilic; that is, the optimum temperature for growth is somewhere in the range of 23-37 °. A few, particularly those from marine sources, are psychrophilic and have an optimum temperature below 20 °. Antibiotics are also known to be produced by certain thermophilic actinomycetes with optima above 50 °. Agar media on which thermophiles are grown tend to dry out rapidly, and sufficient moisture should be provided during the period of incubation. The majority of fungi grow well at room temperature, ordinarily 20-22 °, and when incubators are used the temperature is usually set at 23-25 °. Some fungi are thermotolerant and grow satisfactorily above 30 °, but most do not. Actinomycetes and eubacteria may also be grown at these temperatures, but actinomycetes in particular are normally incubated at a somewhat higher temperature, e.g., 28 °. While pathogenic bacteria frequently require incubation temperatures of 37 ° , most antibiotic-producing strains are grown at temperatures of 32 ° or lower. pH. Fungi grow well in acid environments, but they are not restricted to them. Most will grow normally within a pH range of 4.5 to 8.0. However, the range for sporulation is generally narrower than for growth. M a n y bacteria and actinomycetes, on the other hand, will not grow at a pH much below 5.5-6.0 and the optimum lies between 6.5 and 7.5. In fact, an alkaline environment tends to encourage actinomycetes. The pH changes during growth depend on a number of factors, the most important of which are the buffering capacity of the medium and the type of metabolites produced. Aeration. Practically all antibiotic producers are highly aerobic and grow on the surface of agar and liquid media. As will be pointed out later, oxygen must be provided in deep fermentations. Light. Most fungi grow and sporulate well in the light but some sporu-
8
METHODS FOR THE STUDY O F ANTIBIOTICS
[1]
late better in the dark. It is a safer procedure to avoid incubating cultures in sunlight, however, because certain actinomycetes in particular will not grow or will not sporulate in the presence of light. III. Culture Preservation
Periodic subculture of strains to fresh media may lead to a loss of the ability to produee antibiotics or to other biosynthetic changes. Frequent transfer may also allow low-producing strains to become predominant and in this case it is necessary to plate out the culture and reisolate the desired colony type. Storing cultures at refrigerator temperatures or under oil will reduce or eliminate the possibility of deterioration. A safer method than refrigeration for preserving viability is simply to place fresh agar slants in a freezer, with the caution that repeated freezing and thawing is detrimental to survival. Culture collections, particularly the larger ones, use the lyophil technique because of reliability and the fact that large numbers of cultures can be stored in a relatively small space. Some nonsporulating cultures, however, may be lost by using this method. In these cases storage at very low temperature under liquid nitrogen is the most dependable procedure. Periodic Trans]er. Mycelial organisms, such as fungi and Streptomyces, are grown on test tube slants composed of an agar medium that favors sporulation. Transfers of spore masses are then made periodically by means of a wire loop to fresh slants of the same medium. With nonsporulating cultures, transfers should be made at frequent intervals from young, actively growing marginal areas where these are evident. In making the transfer, a portion of the agar supporting the mycelium should also be dug out and the material on the loop smeared on the surface of the new slant. Cultures should be incubated at an appropriate temperature until they reach maturity and then stored at about 5 ° . Soil Culture. 4,%s For actinomycetes, a loamy soil at neutral or slightly alkaline pH is preferred. If the soil is acid, add 1 g of CaCO3 per 100 g of soil and if lacking in organic matter add 0.25 g of dried blood or casein at the same rate. Enough water should be added to reach a level of about 60% of the maximum water-holding capacity. Place in tubes or flasks, plug with cotton, and autoclave at 15 psi for 1 hr or, alternatively, four times for 30 min each on alternate days. Inoculate with a ' S. A. Waksman, "The Actinomycetes," Vol. I, pp. 26-27. Williams & Wilkins, Baltimore, Maryland, 1959. s D. I. Fennell, Bot. Rev. 26, 79 (1960).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
9
spore suspension, substrate mycelium or young broth culture and incubate at the optimal temperature for a particular culture, preferably with occasional shaking. Allow soil to desiccate after growth. For fungi, inoculate 5 g of orchard loam soil (20% moisture) with 1 ml of a heavy aqueous suspension of conidia. Dry at room temperature and store in the refrigerator. If growth rather than simply retention of spores is required, then larger inocula or more soil moisture should be used followed by a period of incubation. Mineral Oil. s This method is used with agar slants and has been preferred for some nonsporulating cultures which are not particularly amenable to lyophilization. Autoclave a medicinal grade mineral oil at 121 ° for 2 hr and dry in an oven at 170 ° for 1-2 hr. Cover slants completely with a layer at least 1 cm deep of the sterile oil. None of the agar should remain uncovered. Cultures are then stored at the same temperature at which unsealed cultures are normally stored. Freezing. Grow cultures on a medium conducive to sporulation if possible. When mature, place in a deep freeze unit at approximately --22% Prevent unnecessary thawing since repeated thawing and freezing reduces viability. Liqzdd Nitrogen. The method of storing cultures under liquid nitrogen as described below is essentially that of Hwang2 Nonsporulating filamentous cultures are grown on appropriate agar media in petri dishes. Upon maturity cut agar plugs with a No. 3 cork borer and transfer the plugs to sterile cotton-plugged glass vials by means of a sterile spatula. Add 0.8 ml of sterile 10% glycerol or 10% dimethyl sulfoxide. Remove the cotton plugs and heat-seal the ampules. The ampules may be cooled rapidly by plunging directly into liquid nitrogen or slowly frozen to --35 ° and then rapidly to --196 ° by immersing in liquid nitrogen or by the sudden introduction of cold nitrogen gas into the cooling chamber. Store in a liquid nitrogen refrigerator at --150 ° to --196 °. To retain maximum viability the cultures should be retrieved by thawing for 0.5-1 min in an agitated water bath at 38-40 ° . Sporulating cultures may be preserved by the same method, except that harvested spores rather than agar plugs are suspended in 10-15% glycerol or dimethyl sulfoxide. Lyophilization. The basic steps in this process involve the preparation of a dense suspension of cells in a selected fluid, freezing the suspension in vials or ampules in a dry ice bath, evacuating the vials while frozen until the contents are completly desiccated, and then hermetically sealing the evacuated containers. There are several modifications of the 9S. Hwang, Mycologia 60, 613 (1968).
10
METHODS FOR THE STUDY OF ANTIBIOTICS
[1]
lyophil process and the equipment to carry it out from which the investigator may choose in selecting the method most suitable to his particular problem. With some suggested possible variations the procedure outlined below is that of Raper and Alexander 1° and has been in use at the Northern Regional Research Laboratory, Peoria, Illinois, and in other laboratories for many years. Grow fungi and actinomycetes on appropriate maintenance media in test tubes or petri dishes until abundant spores are produced, normally a matter of 7-10 days. Place approximately 0.25 ml of a sterile suspending liquid in a plugged and sterilized agglutination tube and add spores to make a dense suspension. Nonsporulating strains can be homogenized in the suspending medium before freeze drying. For bacteria, wash down the culture growing on an agar slant with the suspending agent to form a dense suspension of cells or centrifuge about 40 ml of a liquid culture, decant, and take up the pellet in the suspending fluid. Some of the more popular suspending media, to be used only after sterilization, are beef serum, double-strength skim milk, 10% high molecular weight dextran, or a gelatin/sucrose solution (5%/5%). Use a micropipette to dispense about 0.05 ml of suspension into each of four sterile lyophil tubes or ampules. Micropipettes can be drawn from 10-mm glass Pyrex tubing. Lyophil tubes may be made of 4-inch lengths of 6-ram Pyrex glass tubing, sealed at one end and lightly fire-polished at the other, plugged with cotton, and sterilized. After adding a suspension to a lyophil tube replace the cotton plug, burn off the excess cotton, and push the cotton about 0.5 inch into the tube. Attach lyophil tubes to a manifold by means of rubber sleeves. An important and frequently overlooked step at this point is "degassing." Enough vacuum should now be used to bubble off most of the air present in the material before freezing. Turn off the vacuum and lower the manifold into a bath of dry ice in either Methyl Cellosolve or 95% ethanol. When the tubes are completely frozen, raise them above the surface of the bath, where a temperature of about --10 ° is maintained, and begin evacuation by means of a vacuum pump. Best results are obtained with a vacuum between 200 and 500 ~m of mercury although some investigators using other apparatus recommend 70 to 100 ~m. The time of drying may be from less than 2 to more than 18 hr, depending on the nature of the organism and the preference of the investigator. By some methods water vapor removed from frozen preparations is taken up in a column of anhydrous calcium sulfate. 1oK. B. Raper and D. F. Alexander, MycoIogia 37, 499 (1945).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
ll
When the pellets are dry continue evacuation at room temperature for 0.5 hour. Then seal off the tubes under vacuum with a gas-oxygen torch. Refrigerated storage of lyophilized preparations is preferable to storage at room temperature.
IV. Antibiotic Production in Liquid Culture If only small quantities of metabotites are desired, it is customary to grow cultures in selected liquid media in flasks placed on a reciprocating or rotary shaking apparatus. Occasional use is also made of stationary cultivation on liquid media and on solid substrates, such as agar, bran, corn meal. When larger quantities of metabolites are needed it is necessary to scale up the fermentation into stirred, aerated bottles or tanks. In this case the conditions which have first been established in flasks may have to be modified to produce maximum results in larger volumes of media. The procedures and culture media described in this section are for the most part those pertaining to antibiotic production in shake flasks. In general, microorganisms can be grown in the same media minus agar that are employed for culture maintenance. Other environmental conditions, such as temperature, pH, also apply. However, these conditions, especially media, do not necessarily promote a desirable level of antibiotic activity. It may be necessary, therefore, to undertake a systematic study of a number of environmental factors before such a level is achieved when dealing with an organism for which appropriate conditions have not been established. If a strain selection and improvement program is undertaken, it should be kept in mind that selected strains sometimes require conditions somewhat different from those which gave superior results with the original strain. Inoculum. Inoculation of a very limited number of fermentation flasks can be accomplished directly from agar slants to fermentation media. Add several milliliters of sterile tap water or saline solution (0.8% NaC1) to the slant and vigorously scrape the surface with a wire loop to make a suspension of spores and vegetative cells. At this point the use of a vortex test tube mixer is sometimes employed in order to enhance the uniformity of the cell suspension, but splashing of the contents onto the plug or cap should be avoided. Using a pipette with a large tip orifice deliver the suspension into liquid media at the rate of about 3% v/v. It is not advisable to initiate a fermentation by digging out a small piece of agar and adding this to the fermentation flask. If more than two or three fermentation flasks are to be used it will
12
METHODS FOR T H E STUDY OF ANTIBIOTICS
[1]
be necessary to employ an intervening liquid inoculum step to provide enough volume. Practical medium/flask relationships are 30 ml of medium in a 250-ml Erlenmeyer flask or 100 ml in a 500-ml flask. Inoculate these flasks as described above and place on a shaker at a suitable temperature: normally 25 ° for fungi, 28-37 ° for eubacteria, and 28 ° for actinomycetes. Incubate long enough to provide a young actively growing inoculum for the fermentation step: in most cases overnight for eubacteria and 36-72 hours for fungi and actinomycetes. Levels of inoculum used for antibiotic fermentations vary, but 3-5% v / v is a useful range. Fermentation, The word fermentation is now commonly employed in relation to antibiotic production under the modern broad definition of the term, namely, the production of metabolites by microorganisms on culture media. The most important environmental conditions to be considered in carrying Out a successful antibiotic fermentation are medium, aeration, pH, and temperature. Each culture passes through the stages of growth, product formation, and senescence and the goal of fermentation improvement is to increase and extend the period of biosynthesis-that is, to achieve and maintain a high level of the desired active principle. Practically all antibiotic producers are highly aerobic organisms. Aeration must be provided, and this is done by placing flasks on a rotary or reciprocating shaking machine for the duration of the fermentation. In addition to the speed of shaking, increased aeration can be obtained by using flasks with indentations or by reducing the volume in each flask. Neomycin is an example of an antibiotic produced in higher yield under conditions of increased aeration. In deep fermentations, air is supplied both by vigorous stirring and by the introduction of air at the bottom of the fermentor. Foaming in tank fermentations then frequently becomes a problem and can be controlled by the addition of one of several agents, a popular example of which is 3% octadecanol in lard oil. However, this may be toxic to some organisms, causing depressed yields, and another type of agent should be used. Most antibiotic fermentations start at a pH close to neutrality, drop to about 6.0, and then rise to 8.0 or above as the carbohydrate becomes exhausted. Maximum antibiotic production often occurs during the earlier stages of this rise in pH. The inclusion of phosphates and CaCO3 in culture media exerts a buffering and stabilizing effect on the fermentation. Occasionally, sterile NaOH or H2S04 is added during the course of the fermentation to maintain a particular pH range. In griseofulvin production, glucose syrup is added at intervals to keep the pH from rising much above 7.0, the optimum for production of the antibiotic. The media used in antibiotic production originate from laboratory
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
13
studies whose purpose is to elicit maximum yields from a particular strain of organism. Such studies may also have the additional purpose of obtaining significant yields of one particular component from among two or more components which a culture is capable of making. A wide variety of substrates from which to choose is available. Carbon sources may include such carbohydrates as the various sugars, starch, or glycerol and a number of plant oils. Commonly used nitrogen sources are amino acids, casein and milk products, animal and plant meals, corn steep liquor, meat extract, and various peptones. Growth-promoting substances, such as distillers' solubles or yeast extract, may be added. Sodium chloride, di- and monopotassium phosphate, magnesium sulfate, sodium nitrate and other salts may be included for supplementary or buffering purposes. However, many of the natural products contain enough material to satisfy any requirements for trace elements. Some representative media for specific antibiotic fermentations are given at the end of this account. In addition to these, there are general purpose media which have been found empirically useful in screening programs to discover new antibiotics because diverse species of microorganisms are capable of producing acceptable antibiotic levels on them. These media can frequently be used for both inoculum and fermentation purposes, and some are suitable for all three major groups of antibioticproducing microorganisms. Warren et al. 1~ listed a number of general purpose media, the following four examples of which appeared to them to be superior. Medium A-4
Soybean meal Glucose NaC1 CaCO~
(g/liter) 10 10 5 1
Medium A-~h
Soybean meal Glucose Curbay BG Glycerol NaCI CaCO3
(g/liter) 15 15 5 2.5 5 1
11It. B. Warren, Jr., J. F. Prokop, and W. E. Grundy, Antibiot. Chemother. (Washington, D.C.) 5, 6 (1955).
14
METHODS FOR THE STUDY OF ANTIBIOTICS
[1]
Medium A-9 (g/liter) 5 10 20
Peptone Glucose Molasses
Medium A-l$ Soybean meal Corn steep liquor Dextrin NaC1 CaCO, K2HPO4
(g/liter) 10 20 10 5 2 2
V. Selected A n t i b i o t i c F e r m e n t a t i o n s b y F u n g i
A. Cephalosporin C Organism. Cephalosporium acremonium grown on C z a p e k - D o x agar with 5 % lactose s u b s t i t u t e d for sucrose.
Media Inoculum medium TM Corn steep solids Ammonium acetate Sucrose Adjust to pH 7.2
(g/liter) 11 4.4 20
Complex medium TM Meat extract Fish meal Corn steep solids Ammonium acetate Sucrose Glucose n~Methionine Rotary shaker, 220 rpm, 28°
(g/liter) 10 10 2.5 2 36 9 0.5
12A. L. Demain and J. F. Newkirk, Appl. Microbiol. 10, 321 (1962).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
15
Defined medium 13 Sucrose Glucose DL-Methionine Oleic acid (NH4)~SO4 KH~PO4 K2HPO4 Na~S04 MgS04" 7H20 Fe(NH4) ~(S04)2.6H20 CaC12 MnSO~ ZnSO4.7H20 CuSO4 •5H20 Distilled water to volume
(g/liter) 36 27 5 1.5 7.5 15 21 O. 75 0.18 0.15 0.06 0.03 0.03 0. 0075
Autoclave glucose separately; adjust initial pH to 7.4; optimum yields are obtained when the flask capacity is 5 times the volume of the medium.
B. G r i s e o f u l v i n Organism.
Penicillium
spp.,
especially
P.
patulum
Bainier-Thom
gr o wn on C z a p e k - D o x agar.
Media Corn steep-lactose medium 14 Corn steep liquor Lactose KC1 KH2PO, CaCO3
(g/liter) (to give 0.15% N) 70 1 4 8
Inoculate with 1-10% of a well-grown vegetative inoculum; incubate on a shaker at 25 ° for 10-i2 days.
Corn steep-glucose medium (for tank fermentation)14 Corn steep liquor CaCO~ KH~PO4 KCI Antifoam agent
(g/liter) (to give 0. 175% N) 4 4 1.5 0.05
Adjust the pH to 5.5 before sterilization inoculate with 10% vegetative inoculum. Glucose syrup is added after the ninth hour at a rate of 0.75 liter per 400 liters of medium per hour to maintain the pH as close to 7.0 as possible. 13A. L. Demain, J. F. Newkirk, and D. Hendlin, J. Bacteriol. 85, 339 (1963). " A . Rhodes, in "Progress in Industrial Microbiology" (D. J. D. Hockenhull, ed.), Vol. IV, p. 165. Gordon & Breach, New York, 1964.
16
METHODS FOR THE STUDY OF ANTIBIOTICS
[1]
C. P e n i c i l l i n Organism. Penicillium chrysogenum. Media Complex medium 15 Lactose Corn steep solids KH2PO4 MgSO4 •7H20 NaNOs ZnS04 •7H~O CaCOs Dow-Corning silicone antifoam
(g/liter) 20 20 0.5 0.25 3 0.04 2.75 2 drops
Adjust to initial pH of 4.7; incubate on a shaker at 22-24 °.
Defined medium 18 Lactose Glucose Starch Acetic acid Citric acid Phenylacetic acid (NH4) 2SO4 Ethylamine Distilled water to volume
(g/liter) 30 10 15 2.5 10 0.5 5 3
VI. Selected Antibiotic Fermentations by Eubacteria
A. B a c i t r a c i n Organism. Bacillus subtilis m a i n t a i n e d on n u t r i e n t agar. Media Complex medium 17 Defatted soya flour Sucrose (NH4)~SO4 CaCO3
(g/liter) 70 12 2 2
lS B. W. Churchill and J. F. Stauffer, in "Biogenesis of Antibiotic Substances" (Z. Vanek and Z. Host£lek, eds.), p. 43. Academic Press, New York, 1965. 1~D. J. D. Hockenhull, in "Biochemistry of Industrial Microorganisms" (C. Rainbow and A. H. Rose, eds.), p. 227. Academic Press, New York, 1963. " J . Ziffer, U.S. Patent 2,813,061 (1957).
[ll
ANTIBIOTIC-PRODUCING MICROORGANISMS
17
Inoculate with 0.5% by volume of a 20-24 hr culture of an appropriate strain of
B. subtilis; incubate at 28 ° on a rotary shaker, 225 rpm, for 45 hr. Defined medium is (g/liter) L-Glutamic acid Glucose Citric acid K2HPO4 KH2PO4 MgSO4 • 7H~O MnSO4- 4H~O FeSO4.7H20 Distilled water to volume
l0 5 1 0.5 0.5 0.2 0.01 0.01
Adjust to p H 6.8-7.0 with N a O H before sterilization; sterilize glucose separately in solution and add aseptically. Inoculate each flask from an agar slant or with 48-hr submerged growth. Incubate at 28 ° on a rotary shaker, 220 rpm.
B. P o l y m y x i n s
Organism. Bacillus polymyxa m a i n t a i n e d o n n u t r i e n t a g a r . Media Nutrient broth-glucose medium (polymyxins A, B, and E)19 (g/liter) N u t r i e n t broth MnSO4 Glucose (NH4) 2HPO4
100 ml 0.02 30 6
For inoculum grow the culture in nutrient broth for 18-24 hr at 37°; use at 5 % v / v . Incubate flasks as static shallow layers or on a shaker at 22-28 ° for 20-24 hr.
Corn steep-glucose medium (polymyxin D)20 (g/liter) Corn steep liquor solids Glucose CaCO3
20 40 10
Inoculate with 2.5-4% v / v liquid inoculum grown 18 hr at 30 °. Incub'~te flasks on a shaker for 4 days at 30 °. is D. Hendlin, Arch. Biochem. Biophys. 24, 435 (1949). 1~G. C. Ainsworth and C. G. Pope, U.S. Patent 2,695,261 (1954). ,0 R. G. Benedict and F. H. Stodola, U.S. Patent 2,771,397 (1956).
18
METHODS FOR THE STUDY OF ANTIBIOTICS
[1]
Defined medium (polymyxin B) ~1 (g/liter) 5
Glucose
(NH4)~S04
1.5
MgSO4.7H20 NaC1 CaCI~ FeSO4.7H~O ZnSO4 MnSO4.4H20 Potassium phosphate Biotin
0.2 0.1 0.1 0.01 0.01 0. 0075 40 ml of 0.5 M, pH 7.5 0.5 gg
Heat to boiling and filter before autoclaving; add glucose as a 50% solution aseptically after sterilization. Incubate without shaking at 30°. The following starter medium for this fermentation is suggested: (in g/liter) glucose, 10; (NH4)2S04, 20; Bacto yeast extract, 5; K~HP04, 2.6; MgSO4.7H~O, 0.5; NaC1, 0.05; FeSO4.7H~O, 0.01. V I I . Selected A n t i b i o t i c F e r m e n t a t i o n s b y A c t i n o m y c e t e s A. C h l o r a m p h e n i c o l
Organism. S t r e p t o m y c e s venezuelae. Media Complex medium ~2 Glycerol Tryptone NaC1 B.Y. fermentation solubles Distilled water to volume
(g/liter) 10 5 5 5
Defined medium 2~ Glycerol Serine NaC1 Sodium lactate K~HPO4 -3H~O KH2PO4 MgSO4.7H20 Distilled water to volume
(g/liter) 10 5 3 11 2.8 1.4 2
Adjust above media to pH 7.0; inocula can be prepared on the same media; incubate inocula on a shaker for 3 days at 26°. 21H. Paulus and E. Gray, J. Biol. Chem. 239, 865 (1964). = M. Legator and D. Gottlieb, Antibiot. Chemother. (Washington, D.C.) 35 809 (1953).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
19
B. Chlortetracycline Organism. S t r e p t o m y c e s aureo]aciens. Media Inoculum medium 2~ Sucrose Corn steep liquor CaCOa (NH,)2SO4
(g/liter) 30 20 7 2
I n c u b a t e a t 26.5 ° for 24 hr on a r o t a r y shaker, 185 r p m ; use at the rate of 4 % v / v .
Starch-corn steep medium 23 (g/liter) 55 25 9 5 1.5 0.06 0.05 0.1 Add individually to flasks at the rate of 20 m l / l i t e r
Starch Corn steep liquor CaCO3 (N H4) 2SO~ NH~C1 FeSO~- 7H20 MnSO~. 4H20 ZnSO4 •7H~O Lard oil
Defined medium 24 CaCO8 (NH4)~SO4 NH4CI MgCl~_ • 6H~O KC1 H3PO4 FeS04.7H20 ZnSO4.7H20 MnSO4 •4H20 CoC12.6H~O Starch (or sucrose) L-Histidine I,-Methionine Lard oil
(g/liter) 9 5 1.5 2 1.3 0.4 0.06 0.1 0.05 0. 005 55 0.8 0.8 Add individually to flasks at the rate of 20 m l / l i t e r
I n c u b a t e at 26.5 ° on a r o t a r y shaker, 185 rpm. 23j. j. Goodman, M. Matrishin, R. W. Young, and J. R. D. McCormick, J. Bacteriol. 78, 492 (1959). ~*J. R. D. McCormick, N. O. Sjolander, S. Johnson, and A. P. Doerschuk, J. Bacteriol. 77, 475 (1959).
20
METHODS FOR THE STUDY OF ANTIBIOTICS
[1]
C. E r y t h r o m y c i n
Organism. Streptomyces erythreus grown on nutrient agar or on the following sporulation medium at 28-37 ° for 10-14 days (light inhibits sporulation). Media S p o r u l a t i o n m e d i u m 25 E
Starch Glucose Tryptone Betaine Curbay BG K2HPO4 NaC1 CaCl~ Mineral mixture Agar Deionized water to volume
(g/liter) 75 5 5 0.5 2 0.2 10 0.08 2 ml 20
Mineral mixture (g/liter) in deionized water: MgSO,.7H~O, 100; FeSO,.7H20 2; ZnSO4.7H20, 1; CuSO,. 5H20, 0.5; MnSO4.5H,O, 0.4; CoCl~. 6H20, 0.1; conc. HC1, 1 ml l n o c u l u m m e d i u m 25
Glucose Sucrose Bactotryptone Bacto yeast extract
(g/liter) 5 10 5 2.5
Incubate 48-72 hr on a shaker at 30-32 °. Complex m e d i u m ~5
Starch Glucose Soybean meal Corn steep solids Yeast NaC1 CaCO8 CoC12.6H20
(g/liter) 25 25 47 10 5 5 2 Trace
~5W. M. Stark and R. L. Smith, in "Progress in Industrial Microbiology" (D. J. D. Hockenhull, ed.), Vol. III, p. 211. Wiley (Interscience), New York, 1961.
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
21
Defined medium ~ (g/liter) 2.5 5 0.5 0.02 0.05 O. 001 0. 001 3 68.4 7.5 1.8
K2HP04 (sterilized separately) NaC1 MgSO4 • 7H~O FeSO4 - 7H~O ZnSO4.7H20 MnCI2- 4H20 CoCI~. 6H~O CaCO8 Sucrose Glycine L-Tyrosine Deionized water to volume Original pH of 7.8 is left unadjusted.
D. Gentamicin
Organism. M i c r o m o n o s p o r a purpurea m a i n t a i n e d o n a g a r s l a n t s cont a i n i n g 1 % y e a s t e x t r a c t , 1 % glucose, a n d 0 . 3 % C a C 0 3 . Media Inoculum medium ~e Beef extract Tryptone Glucose Soluble starch Yeast extract CaCOs
(g/liter) 3 5 1 24 5 4
Incubate on a rotary shaker, 280 rpm, for 4-5 days at 37°; transfer to the same medium (5% v / v ) and incubate at 28 ° for 72 hr; use at a level of 5% v / v .
Fermentation media 2G,27 Yeast extract Glucose CaCOa
(g/liter) 10 10 1
Soybean meal Glucose CaCO~
(g/liter) 30 40 4
Incubate for 5-6 days on a rotary shaker at 26 °. 2, M. J. Weinstein, G. M. Luedemann, E. M. Oden, and G. H. Wagman, Antimicrob. Ag. Chemother. 1963, 1 (1964). 2, G. M. Luedemann and M. J. Weinstein, U.S. Patent 3,091,572 (1963).
22
METHODS FOR THE S T U D Y
OF ANTIBIOTICS
[1]
E. Neomycin
Organism. S t r e p t o m y c e s /radiae. Media Inoculum medium ~8 Glucose N-Z amine A (casein digest: Sheffield F a r m s ) (NH4) 2HPO4 MgS04.7H~O FeSO4 •7H20 CuSO4 •5H20 CaC03 Deionized water to volume
(g/liter) 10 10 5 0.5 0.05 0.05 10
Sterilize the glucose separately as a 50 % stock solution and add to the medium before inoculation; adjust p H to 7.2-7.3 prior to sterilization, I n c u b a t e at 25--27 ° for 1-2 days on a r o t a r y shaker and use at a level of 3 % v / v .
Complex medium ~8 (g/liter) Soybean meal 25 Glucose 10 Brewer's yeast 5 NaC1 5 2 CaCO3 Deionized water to volume Sterilize glucose separately; m a i n t a i n fermentation p H above 7.2 with sterile K O H . M a x i m u m yields occur in a b o u t 5 days at 25-27 ° and under conditions of high aeration (shake flasks w i t h baffles on a rotary shaker at 350 rpm).
Defined medium ~s (g/liter) Glucose 10 L-Histidine 10 (NH4) ~HPO4 1 MgSO4 •7H20 0.5 FeSO4 • 7H~O 0.05 CuSO4.5H20 0.05 CaCO3 10 Deionized water to volume Other f e r m e n t a t i o n conditions remain as above. 2s O. K. Sebek, Arch. Biochem. Biophys. 57, 71 (1955).
[1]
ANTIBIOTIC-PRODUCING MICROORGANISMS
23
F. Streptomycin Organism. S t r e p t o m y c e s griseus. Media Inoculum medium 2~ (g/liter) 10 l0 10 6
Glucose Enzymic hydrolyzate of casein NaC1 Meat extract Distilled water to volume Incubate on a shaker for 48 hr at 28°. Complex medium 29
(g/liter) 25 40 5 2.5
Glucose Soybean meal Distillers dried solubles NaC1 Adjust to pH 7.3-7.5 before sterilization Defined medium 3° Glucose MgS04.7H20 Sodium citrate Glycine NaCl CaCl~- 6H20 KH2PO4 Fe Mg Cu Zn Mo Distilled water to volume
(g/liter) 20 10 10 5 5 0.5 0.5 (mg/liter) ~} as sulfates 0.1 as sodium molybdate
Adjust to pH 7.3 before sterilization. 2gD. J. D. Hockenhull, K. H. Fantes, M. Herbert, and B. Whitehead, J. Gen. Microbiol. 10, 353 (1954). C. G. C. Chesters and G. N. Rolinson, J. Gen. Microbiol. 5, 559 (1951).
24
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
[2] S t r a i n D e v e l o p m e n t
By
AMEDEO A. FANTINI
I. II. III. IV. V.
Introduction . . . . . . . . . . . . . . . . . . Fermentation Conditions . . . . . . . . . . . . . . . Assay for Antibiotic Yield . . . . . . . . . . . . . . Preparation of Suspension for Mutation Treatment . . . . . . . Mutation Techniques . . . . . . . . . . . . . . . . A. Ultraviolet Light (UV) Irradiation . . . . . . . . . . . B. Ionizing Radiation . . . . . . . . . . . . . . . C. Alkylating Agents . . . . . . . . . . . . . . . . D. Other Mutagens . . . . . . . . . . . . . . . . E. Combined Action . . . . . . . . . . . . . . . . VI. Selection from Mutagen-Treated Populations . . . . . . . . .
24 25 26 26 27 29 31 32 35 36 36
I. I n t r o d u c t i o n Strain development, or strain improvement, is a key factor in the development of an antibiotic as a potential therapeutic agent. Before any satisfactory biological testing and chemical characterization of a new antibiotic entity can be accomplished, adequate amounts of the material obviously must be available. As a rule, however, it soon becomes apparent t h a t the producing culture is synthesizing totally inadequate amounts of the antibiotic; thus, the problem of increasing yields becomes one of primary importance, and a pressing one. This discussion will deal with cultures producing low levels of potentially new antibiotics, and the "how to" aspects of improving these strains. B y strain improvement is meant the obtaining of microbial isolates possessing certain desirable characteristics for solving a specific problem. The improvement is only one of utility for the microbiologist, and the culture itself is not necessarily improved from the standpoints of viability, stability, sporulation, pigmentation, or other factors. Further, some of the mutants obtained m a y bear little resemblance to the original parent culture from which they were derived, occasionally creating a problem for the taxonomist. While increasing yields is usually one of the first tasks to be accomplished, it is not the only problem encountered; others include the elimination of unwanted antibiotic components where mixtures are produced, the elimination of undesirable pigments, the selection of mutants with improved growth characteristics, lessened foaming in fermentors, mutants resistant to phage, or mutants synthesizing a modified antibiotic mole-
[2]
STRAIN DEVELOPMENT
25
cule. Whatever the specific problem, its solution comes under the heading of strain development. Simply stated, strain development involves mutation and selection of mutants. For a successful strain improvement program, however, some additional aspects should be considered before starting with mutation techniques; most are obvious and basic, but to ignore them would lessen the effectiveness of the program. They may be listed as follows: fermentation conditions; assay for antibiotic yields; preparation of sample for mutation treatment; mutation techniques; selection from mutagentreated populations. It may also be mentioned at this point that while strain improvement has been applied to innumerable antibiotic-producing cultures over the years, very little has actually been published on the subject. Partly, this is due to the fact that antibiotics are manufactured by a very competitive pharmaceutical industry, so that much of the technical information and data on yields has been considered confidential information. The unavailability of this information, however, should not be considered a serious loss, as much of it would probably be very specific for a particular organism and product. Thus, for an investigator embarking on a strain development program for a novel antibiotic, it may well be more advantageous not to be too influenced by specific techniques which may be useless for a culture for which they were not developed. What may be considered a classic in this field is the work of Backus and Stauffer1 with Penicillium chrysogenum at the University of Wisconsin. This paper is important not only because it reports an increase in penicillin yields 40 times greater than those produced by the parent culture, but primarily because it presents an overall view of the techniques and approach followed over a 10-year period.
II. Fermentation Conditions
These will not be discussed in any detail here; suffice it to say that with a mutation program a parallel study should be initiated to develop satisfactory fermentation conditions. These include seed inoculum, fermentation medium, pH of medium, temperature of incubation, and aeration. In the early stages of improving yields, the above factors are extremely important, and frequently marked gains are made by a study of fermentation conditions alone. In addition, since improvements by mutation are often small, potenM. P. Backus and J. F. Stauffer, Mycologia 47, 429 (1955).
26
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
tially improved mutants will go undetected unless satisfactory fermentation conditions are available. Further, a continuing study of fermentation conditions should be done as an important part of a strain development program, as new mutants will be found that will perform better under conditions other than those originally developed for the parent culture. III. Assay for Antibiotic Yield A reasonably sensitive assay system must be available at the start, as improvements in yields are frequently reflected by relatively small increases in the diameter of a zone of inhibition against a sensitive organism; thus, small changes may well go undetected even though they may represent a 2-fold or better yield improvement. In any case, the assay procedure should be relatively simple, and above all it should not be a limiting factor for the number of mutagentreated isolates that can be fermentation-tested. IV. Preparation of Suspension for Mutation Treatment For efficiency, a suspension of cells or spores lends itself well to mutagenic treatment, subsequent dilution, and plating. 1. Bacterial suspensions of single cells may be prepared by growing in a liquid medium and then subculturing in fresh medium so as to obtain an actively growing population. 2. Spore suspensions from Streptomyces are prepared by adding 5-10 ml of sterile distilled water to a slant and scraping its surface with a pipette; the addition of 0.05% of a suitable wetting agent to the water will result in a more satisfactory suspension. Mycelial fragments are removed by filtering through Whatman No. 2 filter paper held in a small glass funnel. All manipulations should be done aseptically and with sterile equipment; this recommendation will not be mentioned further. 3. For fungi, essentially the same procedure is employed, but mycelial fragments may be removed more readily by filtering through glass wool held in a cylindrical filter tube, such as Kimax No. 46170. 4. Occasionally a microorganism does not produce spores or a mutant has lost its ability to sporulate. Satisfactory mycelial suspensions may be prepared by scraping the surface of a slant in 10 ml of water as mentioned above, and transferring the entire volume to a relatively narrow sterile tube. The coarser particles are allowed to settle out for a few minutes, and a 1-ml volume can then be pipetted from the supernatant and diluted in 9 ml of water. 5. For more difficult cases, where tough or leathery growth is pro-
[2]
STRAIN DEVELOPMENT
27
duced, the latter is transferred in 3-5 ml of water to a tissue homogenizer. The grinding procedure must be restricted to very few strokes, or viability will be greatly reduced. This method will produce fine suspensions devoid of large particles. 6. For certain asporogenous fungi, particularly those exhibiting a rapid filamentous growth, the hyphal tips may be exposed to a mutagen more advantageously than the entire colony growing on a plate. This may be readily accomplished by placing a sterile square of cellophane (avoid water-proofed cellophane) or other porous membrane 3-4 cm wide on the agar surface of a nutrient medium; the center of the square is inoculated with the fungus, and, after an appropriate period of incubation, the growing front of the colony will grow beyond the edge of the cellophane. When the hyphal tips reach a 2-5-ram length beyond this front, the cellophane may be peeled from the surface of the plate, leaving behind the delicate hyphal tips for exposure to mutagenic treatment. 7. Certain spores produced by microorganisms can be very resistant structures. Suspensions of these are made as described above, but in order to make the DNA more accessible to the mutagen it is preferable to pregerminate these spores in a suitable liquid medium prior to treatment. When a germ tube twice the size of the spore is microscopically visible, the suspension is ready for mutagenic treatment. This germination time should not be overextended. 8. Pregerminating spores (or dividing single cells) prior to or during the application of the mutagen has also been employed by some investigators even when good suspensions of uninucleated spores were available, on the assumption that the replicating DNA would be more sensitive to the mutagenic treatment. Unless a definite advantage, or a specific requirement due to the nature of the mutagen, is obvious, the use of germinating spores of filamentous organisms will introduce difficulties in the eventual selection of mutants. Dividing cells must be employed, however, when purine-pyrimidine analogs are used as mutagens, so that they may be incorporated into the replicating DNA molecule. In another application, a mutagenic treatment has been applied at various time intervals to a synchronized cell population, so as to make single-stranded DNA available to the mutagen in a sequential fashion during the replication cycle. 2 V. Mutation Techniques A microbial culture is composed of a dynamic population of cells, and from the standpoint of searching for improved antibiotic producers 2 F. J. Ryan and S. D. Cetrulo, Biochem. Biophys. Res. Commun. 12, 445 (1963).
28
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
it is an advantage that such populations do not remain homogeneous for many generations. Some gains can thus frequently be made in the early stages of a strain development program by simply plating suitable dilutions of the culture and isolating a few of the resultant colonies for fermentation testing. Unfortunately, many mutations, whether induced or spontaneous, are not "gain" mutations; thus, some improved producers may, with time, show signs of lessened antibiotic productivity. In such cases it will frequently be useful to plate appropriate dilutions and isolate a number of the resulting colonies for recovery of the original improved isolate. In any case, if any significant improvements are to be made, the introduction of mutagenic agents becomes necessary. While the chemical changes brought about by the interaction of some physical and chemical mutagens with chromosomal DNA are known, in practice it is not possible to recommend a specific agent that will prove successful in solving a problem like improving the ability of a culture to synthesize greater quantities of an antibiotic molecule. For one reason, the chemical nature of the DNA molecule is such that too many common denominators exist as targets for a mutagen, and thus "directed" mutation of a specific site presents a difficult problem. And perhaps even more important, as Demain ~ pointed out, we are ignorant of the genetic and enzymic changes brought about in a cell as the result of the rare "successful" mutations. Over the years, a number of agents have been described as mutagenic, and it is outside the scope of this article to discuss the application of most of these agents or their mode of action. A paper by Auerbach 4 presents an interesting background history of the better known mutagens, and Heslot ~ has reviewed the principal groups of mutagens and their mode of action. It is usually not necessary to employ many different mutagens in order to conduct a successful strain development program. Actually some agents that have found wide application in the elucidation of the nature of the gene have not been particularly useful in strain development work. Further, most studies of mutagens have been based on the induction of reverse mutations; effectiveness in such mutation induction may not have a parallel in inducing the type of mutant sought in strain development. A. L. Demain, in "Fermentation TechnologyToday" (Proc. 4th Int. Ferm. Syrup.) (G. Terui, ed.), p. 239. Soc. Fermentation Technology,Japan, 1972. 4C. Auerbach, in "Chemical Mutagens; Principles and Methods for Their Detection" (A. Hollaender, ed.), Vol. 3, p. 1. Plenum, New York, 1973. s H. Heslot, Proc. Symp. Radiat. Radloisotop. Ind. Microorganisms, p. 13. Int. Atomic Energy Agency,Vienna, 1971.
[2]
STRAIN DEVELOPMENT
29
Thus, since we do not know what kind of change in the genetic material will bring about the desired effect, selection of a mutagenic treatment should be based on practical considerations: (a) the procedure should not be excessively hazardous (many mutagens are carcinogenic), (b) the techniques should be relatively simple, and (c) equipment and chemicals should be readily available. A magic mutagen for strain development has not yet been found, and what has been effective for culture A may very well be much less so for culture B. Thus it is important to be flexible, and a different mutagen should be applied when few or no improved mutants are obtained. The application of a few mutagenic techniques will now be described.
A. Ultraviolet Light (UV) Irradiation This is probably one of the most frequently employed methods of mutation induction, particularly at the start of a strain development program. An ordinary germicidal lamp equipped with a mercury vapor tube (or tubes) rated to emit radiation of 2537 A wavelength is commonly used. Since the absorption spectrum for DNA (2600 A) is close to this value, this has been suggested as the cause of the mutagenie effect of UV. Absorption of UV by DNA appears to cause bonding of adjacent thymine bases (dimerization) on the same DNA polynucleotide chain. The resulting distortion of the DNA may then cause additional crosslinking of thymine on opposite strands and eventual mutation as the result of base pair substitutions. Since an enzyme requiring visible light as an energy source has been found to repair (by splitting pyrimidine dimers) the lethal and mutagenic effects of UV irradiation (photoreaetivation), all manipulations should be conducted under a limited light source such as a 25-W yellow or red bulb. Suitable spore or cell suspensions are prepared as described earlier; a volume of 10 ml is adequate. In order to determine the effect of irradiation on survival, a count must be made before UV treatment. While some investigators find it convenient to do cell counts by means of a hemaeytometer or Petroff-Hausser counting chamber, determination of the viable count by plating of suitable dilutions is to be preferred. In addition, isolation of a few colonies from this untreated population and eventual testing for their antibiotic-producing ability may uncover a naturally occurring improved mutant. For a viable count then, 1.0 ml of the suspension is diluted serially 1"10 (10-1 dilution) in saline or water with Tween 80 (0.05%) to 10-8; more or fewer dilutions may be required depending on the density of
30
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
the suspension. A 0.1-ml volume from dilution tubes 10.4 to 10-s is then spread on the surface of plates containing an appropriate agar medium; at least 2 plates should be used per dilution. The remainder of the suspension (about 9 ml) is then placed in a petri plate in preparation for UV irradiation. To establish percentage survivals for different exposure times, the distance from UV lamp to the surface of the suspension is maintained constant; a good starting point would be 20 cm. It is important to standardize one's conditions of treatment right at the start, as changes may then be more accurately made based on the results obtained. Further, changes in sensitivity or resistance to UV irradiation may be more readily detected. A range of exposures of 15, 30, 45, and 60 sec for Streptomyces should make possible the selection of an exposure time that will result in a 95-99% kill. UV exposure times should be controlled by opening and closing the petri plate cover, not by turning the lamp on and off. The UV rays will not significantly penetrate the glass or plastic covers for the short working periods here involved. During the exposure times, the suspension should be agitated by gently rotating the plate. At time 0, the cover is removed and then replaced at 15 sec (or other preselected time); a 1.0-ml volume is then pipetted from the plate and into a tube containing 9.0 ml of water and labeled 15 sec (10-1). The operation is repeated for an additional 15 sec, and again a 1.0-ml volume is diluted and labeled 30 sec (10 -1) and so on as outlined. Additional dilutions and subsequent platings should then be carried out; obviously, the greater the UV exposure the fewer the dilutions required to obtain well-spaced colonies following incubation. A reasonable initial plating scheme for a member of the streptomyceres may follow the pattern shown in Table I. All plates should then be incubated at a suitable temperature and in reasonable darkness so as to avoid possible photoreactivation. After colonies develop (3-6 mm in diameter), counts should be made for all exposure times wherever possible. The percentage survival for each exposure period may then be calculated. While a 1-5% rate of survival has been found more useful for obtaining mutants than greater survival values, it should be kept in mind that such data are based on reverse mutation studies and may not be particularly significant for the "gain" mutations sought in strain development. It may thus be worthwhile to occasionally isolate colonies from UV treatments allowing greater or lesser survival rates. In any case, subsequent UV treatments utilizing the survival data
[2]
STRAIN
DEVELOPMENT
31
TABLE I U V EXPOSURE AND INITIAL PLATING SCHEME FOR
Streptomyces
Dilutions UV exposure (sec)
10 -x
10 - z
0 15 30 45
60
x
10 - a
10 -4
10-a
10-~
10-7
10-s
X x
x x
X x x
x
Xa x x x
X
x x x
x
x
x
a x Represents dilution and plating to be done; at least two or more plates should be used per dilution.
obtained, can be designed to more specifically produce a desired percentage of survivors, which with proper dilutions will give counts of 10-60 colonies per plate. At this point, the number of plates employed should be increased so that a reasonably large number of colonies will be available for isolation. It should not be assumed that a high kill rate is a general prerequisite for high mutation efficiency; while it may be true for some mutagens, it is definitely not the case for others.
B. Ionizing Radiation X-Rays, y-rays, a-particles, and fast neutrons are effective forms of radiation for mutation induction. The term ionization refers to their common property of dissipating energy in matter by causing ionization of atoms.
X-Rays are produced by commercially available machines, van de Graaff generators, linear accelerators, and synchrotrons. -/-Rays are given off by radioactive elements such as cobalt-60, and have wavelengths corresponding to very short X-rays. Fast neutrons are usually obtained from a cyclotron, atomic pile, or indirectly from a van de Graaff generator. As mentioned earlier, the selection of a mutagenic treatment must be based on practical considerations. It is obvious that most of the above equipment for generation of ionizing radiation is not readily available from a local scientific supply house; arrangements can usually be made, however, with a university or other institute for irradiation of selected samples. The unit of measurement for X- and -/-rays is the roentgen (r) ; for neutrons, the rad is employed as the unit of measure.
32
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
In contrast to the selective absorption of UV by DNA, ionizing radiation may act directly on chemical bonds, or indirectly by producing free radicals from organic molecules or water, which in turn affect the DNA molecule. The entire mechanism, however, is not completely understood. While X-rays have been shown to produce point mutations, they are capable of producing macrolesions as well (deletions, inversions, translocations) .6 X-Rays have been employed rather extensively in strain development work, and their application seems to have earned them the reputation for inducing useful mutations where other agents have failed. Preparation of a cell or spore suspension for X-ray treatment is the same as described earlier for UV; a volume of about 5 ml is usually placed in a special vial to fit the generating equipment employed. A 2-ml sample of the untreated suspension should be set aside to determine viable counts and calculation of percentage survival for the irradiated portion of the suspension. After irradiation, the suspension is diluted and plated on an appropriate medium. Dilutions of 10-1 to 10-4 are usually satisfactory for the purpose, and a number of plates (10 or more) should be used for each dilution so that recovery of a substantial number of colonies may be anticipated. In the case of a first trial with X-irradiation, plating of the treated suspension may be done on a limited number of plates (2 per dilution) and dilutions extended from 10-1 to 10-e, while the remainder of the irradiated suspension is maintained under refrigeration. As colonies become evident on the plates, a dilution may be selected which will give counts of 10-80 colonies per plate; a final plating may then be done based on this information. Thus, if 10-3 appears optimal, platings should be done at 10-2, 10-8, and 10-4, using several plates. Doses of 100i000-150,000 r have been found satisfactory for Streptomyces species; the percentage survival may be expected to be in the range of 0.1 to 10.
C. Alkylating Agents A number of compounds belongs to this group of agents. They possess one or more alkyl groups capable of being transferred to other molecules. Some of these compounds are sulfonates, ethyleneimines, certain lactones, epoxides, nitrogen mustards, and nitroso compounds. While some alkylate F. J. DeSerres, H. V. Mailing, and B. B. Webber, Brookhaven Syrup. Biol. 20,
56 (1967).
[2]
STRAIN DEVELOPMENT
33
ing agents are carcinostatic, some are also carcinogenic, and safety precautions must be applied in pipetting (use a safety bulb) and disposal. It is beyond the scope of this article to present procedures for the application of many of the alkylating agents as mutagens; while a number have been shown to be effective mutagens in various genetic studies, probably two, the nitrogen mustards and the nitroso compounds, have had the widest application in strain development programs. Nitrogen Mustards. Of these, the most commonly employed is methylbis(fl-chloroethyl)amine, commonly referred to as "HN2." Other members of this family have been labeled "HN," " H N I , " and "HN3" from the wartime code for mustard gas of " H " ; the numbers refer to the replacement of the methyl groups of trimethylamine by one or more 2-chloroethyl groups. A convenient procedure, representing a modification of the technique reported by Roegner et al., 7 is described below. Care should be taken that the mutagenic treatment be done in test tubes large enough to conveniently hold a total volume of 10 ml without danger of spilling; small flasks may be more conveniently used for the purpose. The following reagents should be prepared: Methyl-bis(fl-chloroethyl)amine (HN2) at 1% in sterile distilled water (do not sterilize HN2) NaHC03 at 2%, sterile Decontaminating solution, consisting of: NaHC03 7.0 g and glycine 6.0 g per liter of distilled water Portions of the decontaminating solution should be distributed in tubes, 9 ml per tube, and sterilized; it will be used to end the action of the HN2, and for serial dilutions of the treated spore suspension. The remainder should be kept available for use in case of an accidental spill, and for submerging glassware on completion of the procedure. A spore or cell suspension should be prepared in the usual manner (6-7-ml volume), diluted 1:5, and 5.0-ml aliquots distributed into 5 small flasks (or test tubes) labeled 1 to 5 (Table II). Some investigators prefer pregermination of the spore suspension. The 2% N a H C Q solution is added to each flask at the volumes indicated in the table. The HN2 is then added to 4 of the flasks as shown in the table; flask No. 1 is the control. The reaction is usually allowed to proceed for 30 min, preferably with slight agitation (reciprocating shaker). At this time, 1.0-ml volumes are removed from each reaction flask, and diluted 1:10 in decontaminating solution tubes. This will end the reaction; additional serial dilutions and 7 F. R. Roegner, M. A. Stahmann, and J. F. Stauffer, Amer. J. Bot. 41, 1 (1954).
34
METHODS FOR THE STUDY OF ANTIBIOTICS
T A B L E II PROCEDURE FOR N I T R O G E N M U S T A R D
Spore suspension Flask no. (ml) 1 2 3 4 5
5.0 5.0 5.0 5.0 5.0
[2]
( H N 2 ) ~ TREATMENT
2% NaHCO3 (ml)
1% HN2 (ml)
HN2 Conc. (%)
Exposure time (rain)
5.0 4.9 4.75 4.5 4.0
0 0.1 0.25 0.5 1.0
0 0.01 0. 025 0.05 0.1
30 30 30 30 30
• HN2 = methyl-bis(fl-chloroethyl)amine. platings on suitable media are then completed. The 0.05 and 0.1~o H N 2 levels frequently produce a reasonably satisfactory balance between number of survivors and improved mutants in Streptomyces. Exposure time or H N 2 concentration m a y be increased if necessary; killing of spores is usually rather rapid, and an increase in morphological variants is frequently observed. N-Methyl-N'-nitro-N-nitrosoguanidine ( N T G ) . This compound deserves special mention as a very effective mutagen, s,9 Its action has been attributed to its decomposition prSducts (diazoalkanes)~o; this decomposition appears to occur more readily at alkaline pH. Until fairly recently it was believed t h a t N T G was more effective as a mutagen in dividing cells; thus, spore suspensions were usually pregerminated and N T G was then added during a set incubation period. Now a procedure has been described for Streptomyces coelicolor 11 which differs somewhat in detail from those applied to other bacteria and fungi, and the results obtained from reversion studies show no marked differences between germinated or ungerminated spore suspensions. A stock solution of N T G should be newly prepared for each experiment and kept at low temperature to prevent its premature decomposition. The buffer system to be employed (Tris buffer is convenient for this purpose) should be adjusted to p H 9.0 and sterilized. A filtered spore suspension is prepared in the usual manner, centrifuged at 1000 g for 10 rain, and resuspended in 2.5 ml of buffer. A 1.0-ml volume of the suspension is serially diluted 1:10, and plated as a control 8E. A. Adelberg, M. Mandel, and G. C. C. Chen, Biochem. Biophys. Res. Commun. 18, 788 (1965). 9 E . Cerdh-Olmedo and P. C. ttanawalt, Biochlm. Biophys. Acta 142, 450 (1967). 1oE. Cerd~-Olmedo and P. C. Hanawalt, Mol. Gen. Genet. 101, 191 (1968). ,1V. Delic, D. A. Hopwood, and E. J. Friend, Murat. Res. 9, 167 (1970).
[2]
STRAIN DEVELOPMENT
35
for viable counts. Another 1.0 ml volume of the spore suspension is added to 9.0 ml of NTG solution in a small flask; the solution is prepared so as to give a final concentration of 1 mg/ml. This is considered time 0, and the mixture is incubated in a water bath (28-30°), preferably with gentle agitation. At times 0.5, 1.0, 1.5, and 2.0 hr, 2.0-ml volumes of the reaction mixture are withdrawn from the flask, immediately centrifuged in sterile tubes, and resuspended in 2.0 ml of water; this ends the reaction. Since the spores are exposed to NTG during the centrifugation period, this time should be recorded and considered part of the total treatment time. Dilutions are then made from each exposure time, and plating is done on appropriate media. The procedure described should prove effective with most S t r e p t o m y ces species; NTG concentrations as high as 3.0 mg/ml were reported by Delic et al. ~ in the above study. An exploratory run will provide useful data for establishing more suitable NTG concentrations and exposure times for the particular organism under investigation. NTG levels for other microorganisms have usually been lower than those reported above; for E. coli, a 30-min exposure at 0.1 mg/ml pH 6 has been found very effective,8 and for Saccharomyces cerevisiae, a 30-rain exposure at 0.7 mgfml pH 4 has been reported as quite suitable. 12 A potentially useful property of NTG is that it has been reported to act preferentially at certain sites on the genome, inducing closely linked multiple mutations; this was reported for synchronized E. coli cultures. ~3 A comparable approach applied to an antibiotic producing organism may prove helpful in the development of a less random approach to strain development.
D. Other Mutagens A number of other agents has been reported as effective mutagens in various genetic studies; some of the better known are base analogs, hydroxylamine, nitrous acid, and the acridines. These agents have not been applied widely to strain improvement, possibly because they seem to induce primarily mutations of the microlesion type, whereas multiple mutations or macrolesions among survivors appear more likely to have a positive advantage in strain improvement. 14 12F. K. Zimmerman, R. Schwaier, and V. von Laer, Z. Vererbungslehre 97, 68 (1965). 1~N. Guerola, J. L. Ingraham, and E. Cerd£-Olmedo, Nature (London) New Biol. 230, 122 (1971). ~' D. A. Hopwood, in "Actinomycetales: Characteristics and Practical Importance" (G. Sykes and F. A. Skinner, ed.), p. 131. Academic Press, New York, 1973.
36
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
E. Combined Action More efficient mutagenesis has been reported in a few cases by the application of two different agents. Improved erythromycin 15 and chlortetracycline TM producers were obtained by combined treatment with ethyleneimine plus UV irradiation. A synergistic effect has also been reported when UV-treated bacterial cells are grown in the presence of caffeine. T M It has been suggested that caffeine interferes with the excision repair of pyrimidine dimers from the UV-irradiated DNA. The procedure is essentially as described for UV treatment, but the irradiated suspension is plated on a medium supplemented with 500 ~g/ml of caffeine. The nonirradiated control suspension should also be plated on caffeine-supplemented plates in order to determine a nontoxic concentration. The 500 ~g/ml level appears to be effective for E. coli and Streptomyces. More recently a procedure that significantly improves the percentage of mutants among survivors was described by Townsend et al. TM for Streptomyces. In this case, near UV light in the presence of the photosensitizing chemical, 8-methoxypsoralen was employed. A longwave UV lamp (emitting at >3000 A), commonly referred to as "black light," is set up as previously described for UV treatments. The 8-methoxypsoralen is dissolved in ethanol at 1 mg/ml, and then diluted 1:10 in a water suspension of spores; exposure times of 10-60 min appear to be adequate. The suspension is then irradiated for periods of 0, 5, 10, and 20 min; dilutions and plating are done as mentioned earlier. The authors report that this procedure may be a useful alternative to NTG treatment, as it appears to be less specific as to site of action (at least in E. coli) and less hazardous to use. VI. Selection from Mutagen-Treated Populations After a microbial population has been treated with a mutagen, a number of the survivors will have to be tested for the desired improvement. Two questions will immediately arise: (a) how many colonies should be isolated for testing, and (b) which ones should be isolated. Statistical approaches have been suggested in answer to (a) which might indicate the probability of obtaining an improved mutant from 1~S. I. Alikhanian and V. G. Zhdanov, Dold. Acad. Naulc. SSSR Set. Biol. 125, 1353 (1959). 1, S. I. Alikhanian and N. B. Romanova, Antibiotiki (Moscow) 10, 11'13 (1965). 1~D. M. Shankel, J. Bacteriol. 84, 410 (1962). 18K. Shimada and Y. Takagi, Biochim. Biophys. Acta 145, 763 (1967). ~*M. E. Townsend, H. M. Wright, and D. A. I-Iopwood, J. Appl. Bacteriol. 34, 799 (1971).
[2]
STRAIN DEVELOPMENT
37
a certain number of colonies isolated and tested. However, since the induced improvements are usually rather small and the experimental error and the variability of testing and assaying can be of considerable magnitude, such approaches have not been practical. The answer to (a) then, will be dictated by the practical limiting factors of personnel, incubator and shaker space, and time. The answer to (b), which colonies should be isolated in the search for mutants with the desired characteristics among the many nonmutated ones, or among those which were changed negatively, is a perplexing one, and frequently only a random laborious approach will succeed. The techniques described below have generally been successful, but the technique of choice is the one that will result in the isolation of the greater number of gain mutations over the total number of isolates tested over a period of time. 1. The simplest and most laborious approach is the random isolation of a set number of colonies from a mutagen-treated population. Frequently, especially at the start of a strain improvement program, no other choice may be apparent. 2. Selection of colonies can be based on obvious morphological differences. This approach at least gives the investigator the impression that he is being selective, and a pattern will eventually develop from the testing of these colonies that will at least suggest that certain morphological types are poorer antibiotic producers than others and can thus be avoided. 3. Selection of improved antibiotic-producing colonies has been made by the application of various "overlay techniques." Essentially the procedure consists of flooding the plates containing the treated colonies with a thin layer (4 ml) of soft nutrient agar (0.7%) seeded with a sensitive organism. After incubation, inhibition zones of different diameter will suggest which colonies are better producers, or at least indicate which are obviously poorer. Colony diameter of the producing organism must be considered in evaluating zone sizes. A modification of this technique by Dulaney and Dulaney 2° makes use of a cellophane sheet between the colonies and the seeded agar layer. In a somewhat different approach, single colonies are grown on agar plugs 21 (or a plug of agar is cut around a colony:2), which are then placed on a medium seeded with a sensitive organism; the resulting zones of inhibition are then used as an index of the antibiotic-producing potential of each colony. The application of the various overlay techniques has been found very :~ E. L. Dulaney and D. D. Dulaney, Trans. N.Y. Acad. Sci. 29, 782 (1967). 21T. Ichikawa, M. Date, T. Ishikura, and A. Osaki, Folia Microbiol. (Prague) 16, 218 (1971). 22R. C. Pittenger and E. McCoy, J. Bacteriol. 65, 56 (1953).
38
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
advantageous by some investigators in selecting improved antibiotic producers; others have reported poor correlation when isolates selected by this method were eventually tested in shaker flasks. In any case, the following suggestions will help avoid pitfalls: (a) If the antibiotic is not bactericidal, the colonies to be overlayered should first be replica-plated 2~ to another set of plates before flooding with the test organism. Failure to do so will result in isolates contaminated with the test organism. Alternatively, the cellophane technique should be employed. (b) Frequently the resulting zones of inhibition are too large; obviously a less sensitive organism should be used, or a plating medium less favorable to antibiotic elaboration should be employed, or the thickness of the seeded layer should be increased. 4. Selection of nutritional mutants (auxotrophs). While mutation to the auxotrophic state has frequently been observed to have a negative effect on antibiotic yields, cases of improved productivity have also been reported? 4 This approach is worth investigating, as auxotrophs may be readily recognized and isolated. One such procedure makes use of the replica-plating technique, and consists of replicating colonies resulting from a mutation treatment to a defined medium on which the parent culture is capable of growing. Comparison of the master with the replica plates following incubation will indicate which colonies are auxotrophic mutants by their inability to grow on the defined medium. Various procedures have been described for increasing the percent of auxotrophs among survivors of a treated population. Only two will be mentioned here: Filtration enrichment technique for fungi,2~ and the penicillin technique for bacteria. 26,~7 Basically, both approaches result in the selective elimination of many of the prototrophic survivors with consequent improvement in the ease of recovery of auxotrophs. 5. Selection of prototrophs derived from auxotrophs (revertants). These may be readily selected by, plating a mutagen-treated auxotroph (spore or cell suspension) on a defined medium; only revertants will be able to grow. Some revertants may be due to partial reverse mutation; others to suppression of the original mutation by mutation at a different locus. The reversion approach has been applied with some success in increasing yields of chlortetracycline~° and actinomycin28 A discussion of J. Lederberg and E. M. Lederberg, J. Baclerlol. 63, 399 (1952). E. L. Dulaney, Ann. N.Y. Acad. Sci. 60, 155 (1954). ~ N. Fries, Nature (London) 159, 199 (1947). B. D. Davis, J. Amer. Chem. Soc. 70, 4267 (1948). 2,L. Gorini and H. Kaufman, Science 131, 604 (1960). 2~M. Polsinelli, A. Albertini, G. Cassani, and O. Ciferri, J. Gen. Microbiol. 39, 239 (1965).
[2]
STRAIN DEVELOPMENT
39
the possible biosynthetic pathways and relative value of these approaches has been written by Demain. 29 6. Selection of mutants resistant to toxic analogs of precursors can be an effective method for obtaining isolates capable of synthesizing increased levels of antibiotic. The procedure has been applied in the selection of improved vitamin producers ~° and consists of exposing a microbial population to a gradient concentration of an antimetabolite that is inhibitory to the growth of the cells; survivors usually consist of some mutants capable of producing increased yields of the metabolite. A similar approach was applied by Elander e t al. 31 for the isolation of improved producers of the antibiotic pyrrolnitrin, n-Tryptophan is a direct precursor for this antibiotic but impractical to use in large-scale fermentations; an effort was made to isolate analog-resistant mutants no longer subject to feedback inhibition of the end product. Gradient plates were employed in conjunction with the following D-tryptophan analogs: 5-fluorotryptophan, 6-fluorotryptophan, and 5-methyltryptophan. By isolating colonies resistant to these analogs, a mutant was eventually selected that produced 2-3 times more antibiotic than the parent culture, and in addition D-tryptophan was no longer stimulatory to pyrrolnitrin yields. This approach has particular potential for selection of mutants where something is known of the biosynthesis of the antibiotic. 7. Selection of mutants producing only one antibiotic entity where a mixture is commonly produced by the parent culture. Biosynthesis of antibiotic mixtures is frequently encountered; the mixtures may be made up of related chemical structures, or they may be totally different. The problem is to produce the desired antibiotic component exclusively, or at least to preferentially synthesize more of the desired one and less of the others. Essentially two approaches are employed to achieve the desired result: mutation, and/or manipulation of fermentation conditions. Mutation was proved to be successful in obtaining isolates capable of preferentially synthesizing various components of the nebramycin antibiotic group2 2 Similarly, the investigation of various fermentation conditions (carbon and nitrogen sources, temperature, pH, aeration, carbon to nitrogen ratios) has been useful in favoring production of a desired antibiotic entity. Further, certain components in a mixture may be synthesized relatively early or late during the fermentation period; thus, '~ A. L. Demain, Advan. Appl. Microbiol. 16, 177 (1973). 30G. H. Scherr and M. E. Rafelson, J. Appl. Bacteriol. 25, 187 (1962). ~1R. P. Elander, J. A. Mabe, R. L. Hamill, and M. Gorman, Folia Microbiol. (Prague) 16, 156 (1971). W. M. Stark, N. G. Knox, and R. M. Wilgus, Folia Microbiol. (Prague) 16, 205 (1971).
40
METHODS FOR THE STUDY OF ANTIBIOTICS
[2]
by judicious monitoring of time or pH conditions, improved yields of a specific component may be obtained by harvesting at appropriate times. The prevailing pH values during the fermentation period can also be very important in affecting the ratio of components in a mixture as their stability may be quite different. Thus, various approaches are available for the solution of this rather challenging problem. A complicating difficulty is that the presence or absence of specific antibiotic components will have to be confirmed by chromatography or some other analytical procedure. Since this could be a limiting factor in evaluating numerous isolates, it is usually well worth searching for a differential assay as a preliminary screen. Differences in the chemical structure of the antibiotics in a mixture are occasionally reflected by differences in the sensitivity of various test organisms. These differences may be absolute, such as complete resistance or sensitivity, or only partial, but they can prove useful in detecting changes in the composition of mixtures. 8. In the search for mutants with desired biosynthetic capabilities, the investigator should be aware of the possibility that a genetic modification of the antibiotic-producing culture may well result in the production of a modified antibiotic. In some cases the new molecule may well have improved therapeutic value~3; more frequently, it will be found to be less desirable than the antibiotic produced by the parent culture. In concluding, it must be stated that strain development is frequently a laborious and time-consuming challenge, and every avenue should be explored which may lead toward a more selective isolation of "gain" mutations. The utilization of more than one mutagen is necessary as no single chemical or physical agent can be said to be more effective than another in inducing productive strains. Further, a change in mutagenic agents should be made not only when its effectiveness seems lessened, but also at random. It is important to build a strain improvement program around more than just one improved key mutant, so that if a problem develops mutation work can continue with another improved isolate. Finally, it may be well to repeat that strain improvement has been a key factor for the manyfold antibiotic yield increases which have made possible the development of antibiotics as significant therapeutic agents and economic products. As a possible criticism, it may be said that the application of the empirical approach in strain development has been so successful over the u j . R. D. McCormick, N. O. Sjolander, U. Hirsch, E. Jensen, and A. P. Doerschuk, J. Amer. Chem. Soc. 79, 4561 (1957).
[3]
METHODS FOR STUDY OF ANTIBIOTIC RESISTANCE PLASMIDS
41
years that industrial laboratories seem to have been reluctant to investigate the potential benefits to be derived from the application of more sophisticated genetic studies.
[3] Genetic Methods for the Study of Antibiotic Resistance Plasmids B y JULIAN DAVIES I. Introduction . . . . . . . . . . . . . . . . II. Genetic Transfer of Resistance Plasmids . . . . . . . . A. Conjugation . . . . . . . . . . . . . . . B. Transduction . . . . . . . . . . . . . . . C. Transformation . . . . . . . . . . . . . . III. Enhanced Segregation of Resistance Plasmids by Curing Agents A. Acridine Orange, Quinacrine, and Ethidium Bromide Curing B. Sodium Dodecyl Sulfate . . . . . . . . . . . . C. Mitomycin . . . . . . . . . . . . . . . . IV. Compatibility Properties of R Plasmids . . . . . . . . A. Determination of fi Character . . . . . . . . . . B. Testing for Compatibility of R Factors . . . . . . . .
. . . . .
. . . . .
. . . . . . . . . .
41 43 43 45 48 49 51 51 52 52 54 54
I. I n t r o d u c t i o n T h e extensive use of c h e m o t h e r a p e u t i c agents has p r o v i d e d a n e n v i r o n m e n t conducive to the selection of b a c t e r i a l s t r a i n s I r e s i s t a n t to these agents. Since t h e i r discovery in J a p a n in the 1950's, the occurrence of b a c t e r i a r e s i s t a n t to a n t i m i c r o b i a l agents in clinical s i t u a t i o n s has been reported t h r o u g h o u t the world, a n d it is n o w clear t h a t the presence of such r e s i s t a n t b a c t e r i a can, u n d e r m a n y circumstances, prove to be a serious i m p e d i m e n t to n o r m a l a n t i b a c t e r i a l t h e r a p y Y T h e m e c h a n i s m s of r e s i s t a n c e to m a n y of the c o m m o n l y used a n t i m i c r o b i a l agents have 1Unfortunately the bacterial strains and bacteriophage stocks referred to in this article "are not available from any central source. Many E. coli strains may be obtained from the Coli Genetic Stock Center, Yale University (Curator, Dr. Barbara Bachman), and S. typhimurium can be obtained from the Salmonella Stock Center, University of Alberta, Calgary (Curator, Dr. Kenneth Sanderson). The authors of referenced manuscripts are the best source of strains which have been described in publication. 2S. Mitsuhashi, "Transferable Drug Resistance Factor R." University Park Press, Baltimore, Maryland, 1971. "The Problems of Drug-Resistant Pathogenic Bacteria" (E. L. Dulaney and A. L Laskin, eds.), Ann. N.Y. Acad. Sci. Vol. 182, 1971.
[3]
METHODS FOR STUDY OF ANTIBIOTIC RESISTANCE PLASMIDS
41
years that industrial laboratories seem to have been reluctant to investigate the potential benefits to be derived from the application of more sophisticated genetic studies.
[3] Genetic Methods for the Study of Antibiotic Resistance Plasmids B y JULIAN DAVIES I. Introduction . . . . . . . . . . . . . . . . II. Genetic Transfer of Resistance Plasmids . . . . . . . . A. Conjugation . . . . . . . . . . . . . . . B. Transduction . . . . . . . . . . . . . . . C. Transformation . . . . . . . . . . . . . . III. Enhanced Segregation of Resistance Plasmids by Curing Agents A. Acridine Orange, Quinacrine, and Ethidium Bromide Curing B. Sodium Dodecyl Sulfate . . . . . . . . . . . . C. Mitomycin . . . . . . . . . . . . . . . . IV. Compatibility Properties of R Plasmids . . . . . . . . A. Determination of fi Character . . . . . . . . . . B. Testing for Compatibility of R Factors . . . . . . . .
. . . . .
. . . . .
. . . . . . . . . .
41 43 43 45 48 49 51 51 52 52 54 54
I. I n t r o d u c t i o n T h e extensive use of c h e m o t h e r a p e u t i c agents has p r o v i d e d a n e n v i r o n m e n t conducive to the selection of b a c t e r i a l s t r a i n s I r e s i s t a n t to these agents. Since t h e i r discovery in J a p a n in the 1950's, the occurrence of b a c t e r i a r e s i s t a n t to a n t i m i c r o b i a l agents in clinical s i t u a t i o n s has been reported t h r o u g h o u t the world, a n d it is n o w clear t h a t the presence of such r e s i s t a n t b a c t e r i a can, u n d e r m a n y circumstances, prove to be a serious i m p e d i m e n t to n o r m a l a n t i b a c t e r i a l t h e r a p y Y T h e m e c h a n i s m s of r e s i s t a n c e to m a n y of the c o m m o n l y used a n t i m i c r o b i a l agents have 1Unfortunately the bacterial strains and bacteriophage stocks referred to in this article "are not available from any central source. Many E. coli strains may be obtained from the Coli Genetic Stock Center, Yale University (Curator, Dr. Barbara Bachman), and S. typhimurium can be obtained from the Salmonella Stock Center, University of Alberta, Calgary (Curator, Dr. Kenneth Sanderson). The authors of referenced manuscripts are the best source of strains which have been described in publication. 2S. Mitsuhashi, "Transferable Drug Resistance Factor R." University Park Press, Baltimore, Maryland, 1971. "The Problems of Drug-Resistant Pathogenic Bacteria" (E. L. Dulaney and A. L Laskin, eds.), Ann. N.Y. Acad. Sci. Vol. 182, 1971.
42
METHODS FOR T H E S T U D Y OF ANTIBIOTICS
[3]
TABLE I RESISTANCE CHARACTERS ASSOCIATED WITH RESISTANCE PLASMIDS a
~-Lactam antibiotics (penicillin, cephalosporins) Streptomycin Spectinomycin Gentamiein Tobramycin Neomycin Kanamycin Lividomycin Butirosin Chloramphenicol
Tetracycline Sulfonamides Trimethoprim Erythromycin Lincomycin Fusidic acid Bacteriophages (restriction) Metals (cadmium, nickel, cobalt, bismuth, lead, antimony, mercury, arsenate, arsenite) Ultraviolet irradiation
a This list includes characters on conjugative plasmids (gram-negative organisms) and nonconjugative plasmids (gram-positive bacteria). been characterized, and there are only few antibiotics which are refract o r y to the resistance mechanisms of these strains. Covalent changes in antibiotics are the most common forms of resistance mechanism, but alterations in permeability and other mechanisms are possible. The discovery of this form of resistance, in which genes coding for resistance mechanisms to several antibiotics (see Table I) are linked on an extrachromosomal genetic element or plasmid which replicates autonomously and in many cases codes for its own conjugal transfer, required the application of a number of genetic techniques which are summarized here. In addition, a variety of physical techniques, 4,5 such as dye-buoyant density centifugation, 6 equilibrium density centrifugation, 6 nitrocellulose fractionation, ~ and heteroduplex mapping, s have been applied to the study of resistance plasmids. The term "resistance plasmid" is used to describe extrachromosomal elements in general; these include those plasmids capable of transfer by conjugation (conjugative plasmids, also known as R factors) and those not capable of self-transmission (nonconjugative plasmids). 5,9,1° The latter are found in many gram-positive bacteria and some gram-negative bacteria; the former are found exclusively in gram-negative organisms (see Table II). D. Freifelder, this series, Vol. 21 [6]. R. C. Clowes, Bacteriol. Rev. 36~ 361 (1972). e W. Szybalski, this series, Vol. 12B [124]. Tj. A. Boezi and R. L. Armstrong, this series, Vol. 12A [90b]. 8R. W. Davis, M. Simon, and N. Davidson, this series, Vol. 21 [31]. 9j. Davies and R. Rownd, Science 176, 758 (1972). 1oG. G. Meynell, "Bacterial Plasmids." M.I.T. Press, Cambridge, Massachusetts, 1973.
[3]
METHODS
F O R S T U D Y OF A N T I B I O T I C
RESISTANCE
PLASMIDS
43
TABLE II BACTERIAL SPECIES CAPABLE OF HARBORING
Acinetobacter calcoaceticus A eromonas spp. A lcaligenes faecalis Arizona Citrobacter spp. Enterobacter spp. Erwinia spp. Escherichia coli KlebsieUa pneumoniae Neisseria perflava Proteus spp.
R FACTORS a
Providence spp. Providencia stuartii Pseudomonas spp. Rhizobium japonicum Rhodopseudomonas spheroides RhodospiriUum rubrum Salmonella spp. Serratia marcescens Shigella spp. Vibrio cholerae Yersinia spp.
" Not all R factors have a host range that includes all these strains; in addition many resistance plasmids are not self-transmissible in these species. Among gram-positive strains of bacteria, Staphylococcus aureus is known to harbor nontransmissible resistance plasmids. There have been extensive studies of the structure of plasmids both from a molecular and genetic standpoint. I t is clear t h a t most plasmids consist of covalently closed circular D N A duplexes with molecular weights in a range from about 20 X 106 to 70-80 X 106 (larger composite structures can occur under certain circumstances). Plasmids studied in Escherichia coli, Proteus mirabilis, and Salmonella t y p h i m u r i u m have been found to consist of two separate parts which can dissociate in the bacterium. These two components are the resistance transfer factor, with genes for maintenance, replication, and conjugal transfer, and the resistance determinants, which are the genes coding for the various resistance characters of the particular R factor.
II. Genetic Transfer of Resistance Plasmids
I t is now possible to transfer resistance plasmids between gram-negative species by conjugation, transduction, or transformation. In grampositive species transduction is the most common technique although transformation has also been reported.
A. Conjugation 11 T h e demonstration of transfer and subsequent autonomous maintenance of resistance characters in a recipient strain provides the most con11D. Freifelder,this series,Vol. 21 [34].
44
METHODS FOR THE STUDY OF ANTIBIOTICS
IS]
vincing evidence for the presence of con]ugativc resistance plasmids (R factors) in strains such as E. coli. 1~ The recipient strain generally carries a genetic marker which can be used for counterseleetion against the donor strain. The commonly used markers are nalidixic acid resistance (nalr), sodium azide resistance (aziR), or rifamycin resistance (rif R) although other counterselective markers such as auxotrophy can be used. 1~ Cultures of donor and recipient cells are grown to saturation in a rich medium [for example, Penassay Broth (Antibiotic Medium No. 3)] and 0.1 ml of each is added to the same medium (5 ml) and allowed to stand at 37 ° for 1-2 h r - - t h e mixture is then incubated with shaking for 3-4 hr and samples (0.1 ml or, a sterile loopful) are streaked onto solid medium containing the counterselecting agent (depending on the strain, nalidixic acid 20 ~g/ml, sodium azide 200 ~g/ml, or rifampicin 50 ~g/ml) and one or more of the drugs for which resistance is coded by the plasmid, at appropriate concentrations (since there are often host effects on the levels of expression of resistance, it is advisable to use a low concentration of antibiotic (20-30 ~g/ml) in selecting for transfer. Often the frequency of transfer is low, and this is particularly true during intergeneric crosses, e.g., E. coli X P. aeruginosa, and even in P. aeruginosa X P. aeruginosa. 14 In such cases, methods involving more constant contact between the p u t a t i v e mating pairs is advisable, and thins can be accomplished by carrying out the coniugation on the surface of a nitrocellulose filter or in a thin layer in a medicine bottle. In crosses involving P. aeruginosa and E. coli, since P. aeruginosa is a strict aerobe, anaerobic conditions (anaerobic jar) can be used to counterselect against the P. aeruginosa strain although this is not as convenient as counterselection with an antimicrobial agent, such as rifampicin. I t is important to realize that failure to observe transfer does not mean that a resistance plasmid is not present and even less so that the resistance genes in question are chromosomal. I t is often necessary to examine a large number of bacterial isolates for the presence of con]ugative plasmids, and under these circumstances it is unwieldy to set up many tube crosses in liquid medium, as described above. I t is convenient to use "plate" crosses when m a n y colonies are to be screened. Using a sterile toothpick, single colonies of the desired recipi12H. Watanabe, Bacteriol. Rev. 27, 87 (1963). 13Suitable resistant mutants may be isolated by direct selection; a variety of resistance strains may be obtained from the Coli Genetic Stock Center, Yale University. ~' L. E. B~an, S. D. Semaka, H. M. Van den Elzen, J. E. Kinnear, and R. L. S. Whitehouse, Antimicrob. Ag. Chemother. 3, 625 (1973).
[3]
METHODS FOR STUDY OF ANTIBIOTIC R E S I S T A N C E PLASMIDS
45
ent strain (a nalidixie acid resistant strain, for example) are gently stabbed into the surface of a rich agar plate (containing no selective agents). The colonies should be well separated, about 50 or so can be put on a standard sized petri dish. Colonies of putative R + strains are then stabbed into the recipient colonies, in turn, using sterile toothpicks. The colonies m a y be numbered on the back of the plate with a marker. Thus, each (master) agar plate will contain 50 or more mixed colonies of different R ÷ bacteria with recipient bacteria. After overnight incubation at 37 °, the master plate is replicated (using velvet pads) onto suitable agar plates containing the necessary nutrients and inhibitors that select only the recipients which have received resistance plasmids.
B. T r a n s d u c t i o n ~
Transduction of resistance plasmids was observed soon after they were discovered. Bacteriophage P1 kc has been used with E. coli, and this phage transduces all the resistances plus the resistance transfer factor of a conjugative resistance plasmid. On the other hand, bacteriophage P22, which is used with S. t y p h i m u r i u m is not capable of transducing the entire resistance plasmid and transduction by P22 often leads to loss of transferability and loss of one or more of the resistance characters. This difference is presumably due to the fact that P1 (73 X 106 daltons DNA) is much bigger than P22 (26 X 106 daltons DNA) and is thus capable of incorporating more D N A for transduction. 1~,17 In addition, transduction of resistance plasmids has been reported in Shigella flexneri Is and P r o t e u s mirabilis ~9 among gram-negative species, and transduction is the most convenient way of transferring resistance plasmids between strains of S t a p h y l o c o c c u s aureus. 2°-2~- Stable transduction into a recombination-deficient recipient provides a simple test for the existence of resistance characters on an autonomously replicating unit. Transductional studies have also proved to be useful in genetic mapping studies in both S a l m o n e l l a t y p h i m u r i u m (P22) and S t a p h y l o coccus aureus. Deletions are generated during transduction because of 1, L. Caro and C. M. Berg this series, Vol. 21 [35]. 1~T. Watanabe and T. Fukasawa, Y. Bacteriol. 82, 202 (1961). 1, Mitsuhashi,2 pp. 51-54. 18R. Nakaya, A. Nakarnura, and T. Murata, Biochem. Biophys. Res. Commun. 3, 654 (1960). 1~R. Nakaya and R. Rownd, ]. Bacteriol. 106, 773 (1971). 20R. P. Novick, Virology 33, 155 (1967); K. Smith and R. P. Novick, Y. Bacteriol. 112, 761 (1972). 21R. P. Novick, Bacteriol. Rev. 33, 210 (1969). ~2R. P. Novick and D. Bouanchaud, Ann. N.Y. Acad. Sci. 182, 279 (1971).
46
METHODS FOR THE STUDY OF ANTIBIOTICS
[3]
the limited capacity of the transducing bacteriophages for DNA; these deletions can then be used in the deletion-mapping of point mutations.
Transduction with P1 in E. coli Preparation o] Phage Stock. Two drops of a saturated culture of the donor bacterium (containing the resistance plasmid) and one drop of a suspension of bacteriophage P1 kc (or a clear mutant) (10 s PFU/ml) are added to 2 ml of soft (0.6% agar) nutrient-calcium agar (16 g of nutrient broth (Difco), 8 g of sodium chloride, and agar in 1000 ml of H~O; sterile calcium chloride added to 1 mM after autoclaving) at 45 °. The combination is mixed by agitation and immediately poured onto the top of a fresh (wet and warm, usually prepared the same day) agar (1% agar) plate of nutrient-calcium agar. The petri plate is then incubated at 37 ° for 8-12 hr (5 plates per phage stock are prepared). One milliliter of rich broth is added to each plate and the top agar and medium is scraped into a 50-ml centrifuge tube with a spatula or a thin glass rod which has been bent to form a scraper. Two drops of chloroform are added and the mixture is agitated on a Vortex mixer for 30 sec. The tube is then centrifuged in a table top centrifuge at room temperature (approximately 5000 rpm for 10 min) and the liquid layer of phage stock carefully poured off. This stock is then titered against a suitable indicator strain; many workers prefer Shigella, but a strain such as E. coli C600 is usually satisfactory. The titer of the stock should be 5 X 109/ml or greater, the above procedure is then repeated with a dilution of this phage stock, so that it is grown through the donor strain twice. Conditions ]or Transduction. For transduction of resistance characters, the recipient strain is grown in 5 ml of rich broth, centrifuged, and resuspended in 2 ml of MC solution (1 mM CaCl~, 10 mM MgCl~, pH 7.0), and to 0.9 ml of this suspension is added 0.1 ml of the P1 phage suspension (the multiplicity of infection should be 5:1 for P1 kc, if the clear-plaque mutants P1 vir or P1 clr are used, the multiplicity of infection should be 0.1-0.2 phage per bacterium). The bacteria/phage mixture is allowed to stand at room temperature for 20 min and then centrifuged, resuspended in rich broth (10 ml), and incubated with shaking at 37 °. At appropriate time intervals, 0.1-ml samples can be spread onto rich agar or minimal agar containing nutrients necessary for growth of the recipient strain and also the required selective antibiotic. The required linkage analyses may be performed on clones appearing after incubation by replicating. Propagation in liquid medium after transduction is necessary for phenotypic expression of the resistance characters. As an alterna-
[3]
METHODS FOR STUDY OF ANTIBIOTIC RESISTANCE PLASMIDS
47
tive to growth in liquid medium, the mixture, after transduction, may be spread onto the surface of rich agar plates (approximately 108 cells) and incubated at 37 ° for 4-6 hr. At this time the agar is rimmed with a sterile spatula and gently lifted up from one side of the plate. An appropriate concentration of antibiotic is then pipetted under the agar, the agar is replaced, and the petri plate is allowed to stand in a refrigerator overnight, to allow diffusion of the antibiotic through the agar. Incubation at 37 ° is then continued and the resistant clones picked and scored for necessary auxotrophic and resistance markers.
Transduction with Phage P22 in S. typhimurium Preparation o] Phage Stock. A saturated culture of the donor strain in rich broth is prepared by overnight incubation and 0.I ml of this culture is diluted into 1 ml of rich broth and incubated at 37 ° for 2 hr. To this 10 ml of culture (approximately l0 s cells/ml) is added 107 particles of phage P22 and vigorous incubation is continued for 5-8 hr until the broth begins to clear slightly. Centrifuge the cultures at low speed, remove the supernatant, and centrifuge the supernatant in a Spinco Model L ultracentrifuge at 30,000 g for 1 hr (40 rotor). Remove and discard the supernatant and resuspend the phage pellet in 2 ml of T2 buffer (T2 buffer contains 10 ml of 0.1 M MgS04, 10 ml of 0.01 M CaC1._,, 1.0 ml of a 1% solution of gelatin, 3 g of Na:HP04, 1.5 g KH~P04, 4 g of NaC1, 5 g of K2S04 and distilled water to 1000 ml; when the solution is autoclaved it becomes cloudy, but this clears on cooling). To ensure that the phage suspension is completely sterile, 0.5 ml of chloroform is added and the mixture is well shaken. The chloroform can be removed by bubbling sterile air through the suspension. The phage suspension is stored in a refrigerator; the titer should be approximately 1011 PFU/ml. Conditions ]or Transduction. For transduction, the recipient strain of S. typhimurium is grown overnight in rich broth and 1 ml is placed in a sterile tube and incubated at 37°; to this tube is added 0.1 ml of P22 phage stock, and incubation is continued for 10 rain to allow for phage adsorption. Phenotypic expression of resistance characters is then carried out by incubation in broth or on agar plates as described for P1 transduction in E. coli; following this period of growth, the bacteria are plated on media containing appropriate antibiotics to select for transductants which have received the resistance plasmid. The colonies are subsequently scored for other phenotypes of the plasmid; remember that in P22 transduction the transfer genes are not necessarily cotransduced with the resistance genes. Scoring is most conveniently carried out by
48
METHODS FOR THE STUDY OF ANTIBIOTICS
[3]
transferring colonies to a grid on a "master" plate (containing no antibiotic), incubating at 37% and replica plating onto antibiotic plates. C. Transformation
The discovery of transformation in E. coli23,24has provided an important tool for the study of R plasmid structure and function; it seems likely that further work will provide the technical details whereby other bacterial strains and their plasmids may be studied in similar fashion. Preparation of R PMsmid DNA. It is advisable to use R plasmid DNA in the covalently closed circular form, which can be obtained by ethidium bromide-cesium chloride equilibrium centrifugation or nitrocellulose chromatography. Both these methods provide convenient ways of obtaining active transforming DNA; such DNA preparations have been adequately described in a recent review by Clowes2 The DNA is kept in TEN buffer (20 mM Tris, 1 mM EDTA, 20 mM NaC1, pH 8.0). Conditions ]or Trans]ormation in E. coli. The calcium chloride pretreatment, as described below, is satisfactory for converting most E. coli strains into efficient recipients for transformation. However, certain substrains and mutants lose viability when exposed to calcium chloride. It is important to perform adequate controls and modify the treatment (exposure to calcium chloride, stage of growth, length of 42 ° heat pulse) where necessary, so as to maintain viability of the recipient. The recipient E. coli strain is grown in 10 ml of phosphate-buffered minimal medium to a cell density of 2-3 X 108/ml and chilled quickly by plunging the culture tube in ice water. The cells are collected by centrifugation (5000 rpm X 10 rain) and washed in 5 ml of 10 mM NaC1. The cells are then resuspended in 5 ml of cold 30 mM CaC12 and allowed to stand in ice for 20 min, pelleted by centrifugation and resuspended in 0.5 ml of 30 mM CaC12. To 0.2 ml of cells in CaCi~ is added 0.1 ml of DNA solution (in T E N buffer adjusted to 30 mM CaC12 ; various concentrations should be tested, but 1 ~g is suitable for most experiments), and the mixture is allowed to stand in ice for 1 hr, placed in a 42 ° water bath for 2 min, and chilled to room temperature. The transformants can be detected by one of the following methods: (a) Samples of the transformation mixture are plated directly onto rich medium containing the appropriate selective antibiotics and incubated at 37 °. (b) Samples of the transformation mixture are diluted into rich broth containing no antibiotic, incubated at 37 ° for at least 1 hr, and then plated on solid medium containing antibiS. N. Cohen, A. C. Y. Chang, and L. Hsu, Proc. Nat. Acad. Sci. U.S. 69, 2110 (1972). S. D. Cosloy and M. Oishi, Proc. Nat. Acad. Sci. U~. 70, 84 (1973).
[3]
METHODS FOR STUDY OF ANTIBIOTIC RESISTANCE PLASMIDS 1
I
I
1
1
I
I
1
|
49
12
4O 0 ~c
8
~, 6 E ~4 I--
0.40 0.80
I
1.20
t.60 2.00 2 . 4 0
A m o u n t of DNA ( p . g / m l )
FIG. 1. Effect of concentration of R factor DNA on transformation frequency. Various concentrations of covalently closed R6 DNA were assayed for their ability to transform CaCI2-treated Escherichia coli to kanamycin resistance. Transformation frequency was determined after a 120-min incubation in antibiotic-free medium to allow complete expression of kanamycin resistance. [From S. N. Cohen, A. C. Y. Chang, and L. Hsu, Proc. Nat. Acad. Sci. U.S. 69, 2110 (1972).]
otics. (c) Samples of the transformation mixture are spread onto rich agar medium lacking antibiotic, incubated at 37 ° and subsequently replica-plated onto medium containing antibiotic. In general, method (b) is the most convenient since it allows ample time for phenotypic expression of the antibiotic resistance characters. Method (a) is unsatisfactory for certain antibiotic resistances such as streptomycin, which have a long phenotypic lag. Linkage relationships and other properties may be examined by replica plating. Figures 1 and 2 show a typical DNA requirement for transformation, and the kinetics of appearance of resistance phenotypes after transformation. Transformation of resistance plasmids by similar methods has also been reported for S. aureus. T M
III. Enhanced Segregation of Resistance Plasmids by Curing Agents Evidence for the presence of a resistance plasmid in a bacterial strain can be obtained by demonstrating the loss of the resistance characters 24aL. Rudin, J. E. SjSstrSm, M. Lindberg, and L. Philipson, J. Bacteriol. 118, 155
(1974).
50
METHODS FOR THE STUDY OF ANTIBIOTICS I
I
I
I
|
I
I 20
I 40
J 60
I 80
I t00
I ~20
[3]
I0-6
~:
40 - ~
c ~r
~0 -~
h
10-9 0
(rain)
Fla. 2. Kinetics of expression of kanamycin resistance in transformed Escherichia coli. After incubation of CaC12-treated cells with R6 D N A (0.6 ~g/ml), bacteria were diluted 10-fold into antibiotic-free L broth. At the times shown, 0.1-ml samples of the bacterial culture were spread onto nutrient agar plates containing kanamycin and incubated overnight at 37 ° for determination of number of transformants. An identical sample was diluted appropriately, and plated on antibiotic-free nutrient agar to determine the total number of viable cells. Transformation frequency is expressed in terms of the number of kanamycin-resistant bacteria relative to the total number of viable cells. [From S. N. Cohen, A. C. Y. Chang, and L. Hsu, Proc. Nat. Acad. Sci. U.S. 69, 2110 (1972).]
following loss of the plasmid. In some strains (notably Salmonella typhimurium) and with certain plasmids a relatively high spontaneous segregation of the resistance plasmid ( > 5 % ) is common. ~,5,l° For this reason it is recommended that all stock cultures of resistant strains be mainrained on agar containing a low concentration of one or more of the antibioties to which resistance is determined by the plasmid. However, in most bacterial strains carrying resistance plasmids, spontaneous segregation of the plasmid is not detectable and "curing" agents must be used to enhance this. These curing agents consist of compounds such as acridine orange, quinacrine, ethidium bromide, mitomycin, or
[3]
METHODS FOR STUDY OF ANTIBIOTIC RESISTANCE PLASMIDS
51
sodium dodecyl sulfate. 2,5,1° A variety of other methods are also effective. Some typical protocols are described below. Again, it should be emphasized that failure to remove resistance characters by curing agents is not strong evidence for the lack of a plasmid, neither is it evidence for chromosomal resistance genes.
A. Acridine Orange, Quinacrine, and Ethidium Bromide Curing 25-27 An overnight culture of the resistant strain is diluted into five different 5-ml cultures of Penassay broth or nutrient broth, previously adjusted to pH 7.5, to give a cell concentration of 10~ to 105 per milliliter. Increasing concentrations of the curing agent added to the five tubes cover the range from 10 to 200 ~g/ml. The cultures are then incubated overnight at 37 ° and observed for growth. The cells from the culture tube that contains the highest concentration of curing agent permitting visible growth (usually 50-75 ~g/ml) are diluted and plated onto rich agar plates and grown up to single clones. These clones are tested for the absence of resistance characters by replica-plating onto plates containing antibiotics. It is essential that several of the resistance characters be examined since, in certain cases, resistance plasmids show a tendency to segregate single resistance characters without loss of the plasmid.
B. Sodium Dodecyl Sulfate 2s Sterile Penassay broth containing 10% sodium dodecyl sulfate is prepared and is inoculated with a low dilution of cells of the resistant strain (to 103 to 104 cells/ml). The culture is then incubated for 48-72 hr at 37°; at this time the culture medium should be (apparently) free of bacteria and very viscous. Most of the cells have lysed at this stage, but the survivors are usually enriched for bacteria which have lost the plasmid. It appears that cell surface characters associated with plasmids (pill, etc.) make these cells more sensitive to lysis by the detergent than the corresponding R- strain. Recovery is low, and this cannot be used as a quantitative measure of induced plasmid loss. 2: T. Watanabe and T. Fukasawa, J. Bacteriol. 81, 679 (1961). 28D. Bouanchaud, M. R. Scavizzi, and Y. A. Chabbert, J. Gen. Microbiol. 54, 417 (1969). 27F. E. Hahn and J. Ciak, Ann. N . Y . Acad. Sci. 182, 295 (1971). :s M. Tomoeda, M. Inuzuka, N. Kubo, and S. Nakamura, J. Bacteriol. 95, 1078 (1968).
52
METHODS FOR THE STUDY OF ANTIBIOTICS
[3]
C. M i t o m y c i n 29
This drug has been used as a means of curing plasmids from strains of P. aeruginosa. An overnight culture of the resistant strain is diluted to 105 cells/ml in tubes of rich medium containing successively higher concentrations of mitomycin C (a maximum of 100/~g/ml is usually sufficient). The tubes are incubated overnight with shaking and organisms from the tube containing the highest concentration of mitomycin which permits growth are plated onto rich agar. Colonies are then picked onto rich agar plates and replica-plated to score for antibiotic resistance characters. A relatively high frequency of curing induced by any of these agents is good evidence for the presence of a resistance plasmid, although lack of curing cannot be taken as evidence for lack of a plasmid, or the existence of chromosomal resistance genes since many resistance plasmids are quite refractory to curing. Demonstration of concomitant loss of resistance phenotype and plasmid DNA (by methods such as dye-buoyant density analysis) provides strong support for plasmid-mediated resistance. In those instances where it is simply desired to obtain the host bacterium free of resistance plasmid, any of the above methods can be combined with B-lactam antibiotic screening (penicillin selection) for the antibiotic strains. Thus if a tetracycline-sensitive strain is required, curing treatment followed by two cycles of treatment with a fl-lactam antibiotic in the presence of tetracycline can be used to enrich for rare tetracycline-sensitive plasmid-negative strains, or mutants. The survivors of this treatment should be screened against other antibiotics for which resistance is determined by the R plasmid to determine whether some or all of the resistance genes have been lost or mutated. The choice of /3-1actam antibiotic is important; for most E. coli or S. typhimurium strains ampicillin selection is sufficient, although cefotoxin or another/3lactamase resistant penicillin or cephalosporin is convenient when the strain is a/3-1actamase producer. For P. aeruginosa carbenicillin selection can be used, but resistance to this drug is common in this organism. Penicillin G or methicillin (for/3-1actamase producers) are used for screening of S. aureus strains. IV. Compatibility Properties of R Pla'smids The most common way of characterizing R plasmids is on the basis of their property to coexist or not to coexist in the same bacterium. The original observation of incompatibility was by Watanabe, 12 who found ~oj. G. Rheinwald, A. M. Chakrabarty, and I. C. Gunsalus, Proc. Nat. Acad. Sci. U.S. 70, 855 (1973).
[3]
METHODSFOR STUDY OF ANTIBIOTIC RESISTANCE PLASMIDS
53
TABLE I I I R FACTOF~COMPATIBILITYGROUPSa Representative plasmid RA1 R57b R386b Rldrdl6
ColB.K98 R124 Flac
R811 R27 R144 R483 R621a JR66a R391 R387 R446b N3T R724 RP4 R751 Rtsl R388 RCK
Compatibility group
Resistances b
A
T,Su
C
A,C,Gk, Su
Ft FII FII Fiv Fv G H I I I I J K M N O P P T W X
T K T A T T,K S,Tp T S,K K S,C S,T T S,T,C,Su A,T,K Tp K Su,Tp A,S
See J. N. Coatzee, N. Datta, and R. W. Hedges, J. Gen. Microbiol. 72, 543 (1972); Y. A. Chabbert, M. R. Scavizzi, J. L. Witchitz, G. R. Gerband, and D. H. Bouanchaud, J. Bacteriol. 112, 666 (1972); N. D. F. Grindley, J. N. Grindley, and E. S. Anderson, Mol. Gen. Genet. 119, 287 (1972) ; R. W. Hedges and N. l)atta, J. Gen. Microbiol. 77, 19 (1973). bT, tetracycline; Su, sulfonamide; A, ampicillin; C, chloramphenicol; GK, gentamicin/kanamycin; K, kanamycin; Tp, trimethoprim, S, streptomycin that the R plasmids which he was studying could be divided into two classes, one that would allow a sex factor to coexist in the same host (fi-), and the other which led to the expulsion of the sex factor (fi÷), presumably by segregation during replication and division. The molecular basis for this expulsion is still not understood, but by examining the abilities of various R plasmids to coexist in E . coli it has been possible to assign more than 20 different compatibility types~°; a similar phenomenon has been described in S t a p h y l o c o c c u s aureus. '~1 The prototype for studies of this kind is the work of Coatzee et al., 3° and Table I I I shows the currently known compatibility types among a group of plasmids. ~J. N. Coatzee, N. Datta, and R. W. Hedges, J. Gen. Microbiol. 72, 543 (1972).
54
METHODS FOR THE STUDY OF ANTIBIOTICS
[3]
A. D e t e r m i n a t i o n o f fi C h a r a c t e r 3°
The R factors to be tested are transferred by conjugation or transduetion to an E. coli Hfr donor such as Hfr Hayes or Hfr Cavalli. To facilitate transfer it is convenient to use a nalidixic acid- or rifampicinresistant mutant of the recipient strain. Conjugation is carried out as described previously. A culture of the R÷/Hfr strain is then grown in rich medium and 2-3 drops spread onto a rich agar plate containing CaC12 (1 raM). Samples of a stock lysate of the male-specific phages MS2, f2, or R17 are dropped (spotted) onto the bacterial lawn with a Pasteur pipette or spread on the surface of the lawn with a sterile loop or paper strip. Visible lysis after overnight incubation at 37 ° indicates that the plasmid is fi-; no lysis indicates that the plasmid is fi+ and is repressing the synthesis of sex pili necessary for attachment and infection by the male-specific bacteriophages.
B. Testing for Compatibility of R Factors 3°-~3 This is determined by testing the ability of two different R factors of a particular group to coexist stably in the same host. For this to be done, the two R plasmids must differ in resistance characters in order that a combination of resistance characters from the two may be selected. For ease of selection, it is convenient to have one of the R plasmids present in a standard nalidixic acid, or rifampicin-resistant recipient to allow selection against growth of the donor host harboring the entering R plasmid. The frequency of transfer is compared against that into the same recipient which does not carry an R plasmid. Recipient colonies are purified on nonselective medium and tested, by replica plating, for the presence of characters of each plasmid to give a measure of compatibility. In this way it can be easily seen whether the presence of one R plasmid excludes (or is incompatible) with another, and they can be assigned to incompatibility groups (Table III). If on the basis of resistance phenotypes, two R plasmids apparently coexist in the same host it is important to determine whether they remain as independent R plasmids, since recombination could have occurred. To test for this, the host carrying the two R factors is submitted to a series of conjugations with a suitable recipient (as previously de~1y. A. Chabbert, M. R. Scavizzi, J. L. Witchitz, G. R. Gerband, and D. H. Bouanchaud, J. Bacteriol. 112, 666 (1972). s2N. D. F. Grindley, J. N. Grindley, and E. S. Anderson, Mol. Gen. Genet, 119, 287 (1972). 3~R. W. Hedges and N. Datta, J. Gen. Microbiol. 77, 19 (1973).
[4]
ANTIBIOTIC ASSAYS
55
scribed). Each resistance character of the two coexisting plasmids is selected in turn and the transcipients are analyzed separately for the resistance characters of both plasmids, by replica plating. In this w a y it can be seen whether the characters of the two plasmids remained intact in the host, or any assortment occurred. In like manner incompatibility groups have been assigned for resistance plasmids in P. aeruginosa. 14 In S. aureus the stable coexistence of R plasmids in the same host can be tested by transduction. 21
[4] A n t i b i o t i c Principles
Assays--
and Precautions
B y FREDERICK KAVANAGH I. Introduction . . . . . . . . . . . . . . . . . . II. General Considerations . . . . . . . . . . . . . . . A. Organisms Used in Assays . . . . . . . . . . . . . B. Preliminary Screening Methods . . . . . . . . . . . . C. Quantitative Aspects of Relatively Pure Samples . . . . . . . III. Kinds of Assays . . . . . . . . . . . . . . . . . A. General Comments . . . . . . . . . . . . . . . B. Diffusion Methods . . . . . . . . . . . . . . . C. Dilution Methods . . . . . . . . . . . . . . . . IV. Assays of Diverse Biological Samples . . . . . . . . . . . V. Relative Inhibitory Coefficient of Antibiotics . . . . . . . . .
55 57 57 58 58 59 59 60 62 65 66
I. I n t r o d u c t i o n Anyone who works with antibiotics needs to understand the principles of the several kinds of assays, their applicabilities, and their deficiencies. In laboratories t h a t have assay groups, the researcher should always discuss the reasons for the assays and a n y anticipated difficulties with this group. Otherwise the preparations will be processed in the routine assay. The numbers obtained in such assays m a y have little intrinsic meaning, but neither the researcher nor assay group will realize it. Without a common understanding of principles, the two groups cannot hold meaningful discussions of their common problems. Analytical microbiology is a p a r t of the general subject of analysis, not simply a minor branch of bacteriology. The bacteria are employed as reagents and usually are the most dependable p a r t of the methods.
[4]
ANTIBIOTIC ASSAYS
55
scribed). Each resistance character of the two coexisting plasmids is selected in turn and the transcipients are analyzed separately for the resistance characters of both plasmids, by replica plating. In this w a y it can be seen whether the characters of the two plasmids remained intact in the host, or any assortment occurred. In like manner incompatibility groups have been assigned for resistance plasmids in P. aeruginosa. 14 In S. aureus the stable coexistence of R plasmids in the same host can be tested by transduction. 21
[4] A n t i b i o t i c Principles
Assays--
and Precautions
B y FREDERICK KAVANAGH I. Introduction . . . . . . . . . . . . . . . . . . II. General Considerations . . . . . . . . . . . . . . . A. Organisms Used in Assays . . . . . . . . . . . . . B. Preliminary Screening Methods . . . . . . . . . . . . C. Quantitative Aspects of Relatively Pure Samples . . . . . . . III. Kinds of Assays . . . . . . . . . . . . . . . . . A. General Comments . . . . . . . . . . . . . . . B. Diffusion Methods . . . . . . . . . . . . . . . C. Dilution Methods . . . . . . . . . . . . . . . . IV. Assays of Diverse Biological Samples . . . . . . . . . . . V. Relative Inhibitory Coefficient of Antibiotics . . . . . . . . .
55 57 57 58 58 59 59 60 62 65 66
I. I n t r o d u c t i o n Anyone who works with antibiotics needs to understand the principles of the several kinds of assays, their applicabilities, and their deficiencies. In laboratories t h a t have assay groups, the researcher should always discuss the reasons for the assays and a n y anticipated difficulties with this group. Otherwise the preparations will be processed in the routine assay. The numbers obtained in such assays m a y have little intrinsic meaning, but neither the researcher nor assay group will realize it. Without a common understanding of principles, the two groups cannot hold meaningful discussions of their common problems. Analytical microbiology is a p a r t of the general subject of analysis, not simply a minor branch of bacteriology. The bacteria are employed as reagents and usually are the most dependable p a r t of the methods.
56
METHODS FOR THE STUDY OF ANTIBIOTICS
[4]
Basic knowledge of bacteriology, especially bacterial physiology, is necessary to the proper practice of microbiological assaying. Most of the causes of the low quality of assays come from lack of both analytical knowledge and of analytical sense by practitioners of the art. Microbiological assaying should be considered to be an exacting profession. No one would expect to become an analytical chemist or an expert at synthesizing new antibiotics by reading a brief chapter in a book; he would expect to spend years learning the trade. The purpose of this brief article is to give principles of the several kinds of assays for antibiotics, to discuss problems of application, to indicate reasons for assaying, and to consider the general topic of relation between structure and activity. Details of assay methods will not be given because to do so would fill the volume and would duplicate other publications. The general references, where details may be found, are the two volumes of "Analytical Microbiology," edited by F. Kavanagh. 1,2 Microbiological assays are performed for two reasons. The first is that antimicrobial activity is the only property the rather diverse group of compounds known as antibiotics have in common, and the appropriate method for measuring their quantity is by means of that activity. The second reason is that the alternatives to microbiological methods, namely, chemical methods, may take too much time to develop and may lack specificity for only those compounds with antimicrobial activity. Antibiotic assays are relative assays in which responses to a sample are compared to responses to a standard. This relativity places certain constraints upon the system if the assays are to be valid. The most important constraint is the requirement of qualitative identity of standard and sample solutions and of their treatments. To meet these requirements of validity, standard and sample must have the same chemical compositions, must be dissolved in identical solvents (usually aqueous solutions that have the same pH and buffer capacity), and the solutions must be processed in the assay in an indistinguishable manner. Usually the first requirement, that of chemical identity, is not established. The chemical steps are under the control of the analyst. The final requirement, that of identity of processing, is provided by the AUTOTURB ® System2 ,4 Close approximation to these requirements can be achieved by any analyst without extensive mechanical aids and automation if he is skilled in the art and practice of microbiological assaying. 1 F. Kavanagh, ed., "Analytical Microbiology." Academic Press, New York, 1963. 2 F. Kavanagh, ed. "Analytical Microbiology," Vol. II. Academic Press, New York, 1972. s N. Kuzel and F. Kavanagh, J. Pharm. Sci. 60, 764 (197,1). 4 N. Kuzel and F. Kavanagh, J. Pharm. Sci. 60, 767 (1971).
[4]
ANTIBIOTIC ASSAYS
57
II. General Considerations A. Organisms Used in Assays There are several reasons for measuring the susceptibility of different microbial species to antibiotics. One is to obtain data to guide clinical applications. Another is an aid in selecting appropriate assay organisms. Certain bacteria are employed in diffusion assays and others in tube methods. The most popular test organisms for diffusion assays are Staphylococcus aureus, Sarcina lutea, and Bacillus subtilis. Occasionally Esch-
erichia coli, Klebsiella pneumoniae, Photobacterium fished, Salmonella gallinarium, and Pseudomonas sp. are used for specific antibiotics. Food and Drug Administration '~ uses 15 species and strains of bacteria and fungi to assay 56 different antibiotics by the plate method. They use four bacteria, S. aureus, K. pneumoniae, E. coli, and Streptococcus ]aecalis, and one yeast, Saccharomyces cerevisae, to assay 23 antibiotics by turbidimetric methods. Staphylococcus aureus is the most often used organism in tube assays. Streptococcus faecalis is employed when the pH of the assay medium is low as, for example, in the assay for monensin ~ in animal feeds. Salmonella gallinarium is sensitive enough to preservatives, such as thimerosal, 7"that it can be used in a turbidimetric assay for preservatives in biological preparations as well as for antibiotics. Escherichia coli, K. pneumoniae, and P. aeruginosa are used to assay antibiotics with special activity against gram-negative organisms. The test organism must be susceptible to the drug to be measured. It must grow reasonably fast in practical assay media. It should not be virulent to man for reasons of safety. It should be easy to maintain. Its susceptibility should not change with successive subcultivation. Old laboratory strains are much more likely to satisfy these criteria than freshly isolated strains. Most test organisms are old laboratory strains and have been in ~lse for 30 years or more. The population of cells produced in the inoculum broth should be of uniform susceptibility. In certain types of assays, e.g., overnight serial dilution, resistant forms could grow out and give confusing answers. The first example of this was a tube assay for streptothricin by means of S. aureus (Heatly strain, ATCC 9144). A strain of Klebsiella pneumoniae, ATCC 9997, did not have the resistant forms and was suitable 5B. Arret, D. P. Johnson, and A. Kirshbaum, Y. Pharm. Sci. 60, 1689 (1971). 6 F. Kavanagh and M. Willis, J, Ass. Offic. Anal. Chem. SS, 114 (1972). 7F. Kavanagh, in "Analytical Microbiology" (F. Kavanagh, ed)., Vol. 2, p. 343. Academic Press, New York, 1972.
58
METHODS FOR THE STUDY OF ANTIBIOTICS
[4]
for streptothricin and streptomycin assays. This experience with streptomycin assay indicates the necessity for ascertaining the suitability of test organisms. More details and method of testing the organism are given by Kavanagh. 8,9 If a preparation of a single component antibiotic, for example, penicillin G, is to be assayed in terms of a single component standard of the same chemical structure, any susceptible test organism could be used under any suitable condition. When both standard and sample are mixtures, as is usual, the organism and operating conditions must be specified if others are to obtain comparable answers. Interpretation of answers in terms of weight of active drug must be done with caution when standard and samples are mixtures.
B. Preliminary Screening Methods The usual way of detecting antibacterial activity in preparations of unknown activity is by means of a simple diffusion test. Concentrations of about 1, 10, and 100 ~g/ml in water, buffer, or dilute alcohol are prepared. Filter paper disks saturated with the solutions are placed on agar plates seeded with B. subtilis, S. lutea, S. aureus, or E. coli. The plates are then incubated at 33-35 ° overnight to permit growth of the test bacteria. Clear zones of inhibition will be found around certain of the disks if the compound has antibacterial activity. A zone around only the 100 ~g/ml disk on S. lutea plates usually indicates a very weak antibiotic. If zones at the l~g/ml concentrations occur on the S. lutea, S. aureus, and B. subtilis, but not on the E. coli, plates at 100 ~g/ml, then the compound has good activity against gram-positive bacteria but little against gram-negative bacteria. If activity is mainly on the E. coli plates, the compound should be tested against other gram-negative bacteria. Activities in this crude screening test are used as a guide in designing assay methods for the compound. The character of the zone edge as well as sensitivity needs to be considered in selecting a plate method. Usually zone edges are better defined in the B. subtilis assay than in the others. If activity against S. aureus, or E. coli is indicated, then one of them could also be used in a tube method.
C. Quantitative Aspects of Relatively Pure Samples More antibiotic assays are done to measure the quantity of an active antibiotic than for all other purposes. Samples may be pharmaceutical s F. Kavanagh, Bull. Torrey Bot. Club 74, 303 (1947). 9F. Kavanagh, in "Analytical Microbiology" (F. Kavanagh, ed.), Vol. 1, p. 125. Academic Press, New York, 1963.
[41
ANTIBIOTIC ASSAYS
59
dosage forms of high purity and precisely known composition, agriculture products such as premixes and animal feeds in which the active drug may be of high purity, and process samples of lessened purity but known identity. The first two classes of products must meet a label claim of quantity. The third class of samples are of great interest to the chemists who must supervise production of the product. The assays enable them to monitor their operations and to determine yields across purification steps. All samples discussed in this section are of known identity and of high purity so that measurement of quantity is possible. This is a good place to make a point crucial to microbiological assays as well as to many chemical assays in which a sample is measured in terms of a standard. It is that the quantity of sample can be exactly determined only when sample and standard have identical composition. When the composition of sample is unknown or is known to differ from the standard, then the sample is said to have an activity equivalent to a certain quantity of standard when assayed in a particular assay under specified conditions. The exact quantity of active product as a weight of product compound cannot be obtained from such assays. The degree of purity of a preparation can be determined if the active drugs in the samples have the same relative concentrations as the active materials in the standard. The sample under these conditions may be considered to be a diluted standard so long as the inert materials do not interfere with the assay. Preparations obtained during purification of an antibiotic being manufactured usually meet the criteria. The safest practice is to consider the figures obtained on impure samples only as approximate values until the preparations become nearly pure or chromatography indicates no unexpected active substances.
III. Kinds of Assays A. General Comments The two kinds of assays are those usually referred to as dilution and diffusion. In the former, the antibiotic is presented to the test organism at one concentration. The concentration is uniform throughout the medimn. The medium may be liquid as in tube methods or solidified by agar as in a plate dilution method. Activity may be measured by an end point, as in serial dilution in tubes s,%1° or agar dilution in plates, 1° or as a graded response, as in the diffusion and turbidimetric methods. 1,~ lo H. M. Ericsson and J. (3. Sherris, "Antibiotic Sensitivity Testing." Acta Pathol. Microbiol. Stand., Suppl. 217 (1971).
60
METHODS FOR THE STUDY OF ANTIBIOTICS
[4]
Three kinds of dilution methods are in use; namely, agar dilution, serial dilution in tubes, and photometric assays. Some sort of all-or-none response is obtained with the first two types. The third represents the high precision turbidimetric assays in which graded response is obtained to graded concentrations of drug. B. Diffusion Methods
In diffusion assays, the concentrations of antibiotic vary between the point of application and the point of inhibition. The response is observed as a clear zone of no growth in which the diameter of the zone is approximately proportional to the logarithm of concentrations of antibiotic for small zones. Theory of the method was reviewed by Cooper. 1,2 The two general methods and their applications will be considered in detail in the following sections. Diffusion assays may be performed in tubes or in dishes. The tube method is used in Japan where it is called the superposition method. The test organism is in nutrient agar in a layer about 2 cm deep in small tubes. The sample is added to the tube to a depth of about 1 cm. The tubes are incubated to permit development of the uninhibited bacteria. The length of the zone of no growth is measured. Zone length is proportional to logarithm of concentration of antibiotic. 2 The tube method is a simple method that does not require special equipment. Incubation can be done in a water bath. Most diffusion assays are done in petri dishes or in large plates. The samples may be applied in disks, in cylinders, or in holes in the agar layer. Inoculation of the agar may be uniform or as a thin layer on the surface of the bulk agar for aerobic organisms or next to the glass for microaerophilic organisms. Details of specific methods are given in the two general references. 1,2 When a new substance is to be assayed, the general procedure is to select a test organism that has requisite susceptibility for plate assaying. The assay design can be selected from those given for penicillin or for a substance of the same general class as that of the new substance. All designs are derivatives of those devised for assay of penicillins. Advantages and disadvantages need to be considered before plate methods are applied. The method is deceptively simple but the theory is complex. Details must be carefully observed if precision is to be high. Precautions to be observed to obtain high precision assays are outlined by Kavanagh. 11 The diffusion assay is a physicochemical method in which a bacterium 11F. Kavanaugh, in "Analytical MicrobiOlogy" (F. Kavanaugh, ed.), Vol. 2, p. 31. Academic Press, New York.
[4]
ANTIBIOTIC ASSAYS
61
is used as indicator of concentration of active compound. Diffusion of the drug is determined by diffusion constant, temperature, and to a minor extent by pH, salt concentration, concentration of agar, etc. The diffusion constant is related to molecular weight of the diffusing species. The smaller the molecular weight, the farther the drug diffuses in a given time. Location of the edge of the inhibition zone is determined by arrival of a certain concentration (critical concentration) of the drug before the bacteria have grown to a critical concentration of cells. 1-~ Because zone size is dependent upon several physical and biological factors, size per se is not an indication of concentration of drug or of sensitivity of test organism. Diffusion assays are relative; a standard is required to calibrate the system. Standard and sample must be of the same chemical species for the assays to have meaning. When the dose-response line relating zone diameter to logarithm of concentration of drug covers a relatively short range of concentrations, graphs of the lines on semilog paper are straight. A range of less than 8 times the least concentration showing a zone of inhibition usually will give straight or nearly straight lines. Closely related antibiotics will plot into families of approximately parallel lines. Dissimilar substances may give lines that are not parallel. Thus two factors govern the appearance of dose-response lines. One is the sensitivity of the system which locates the starting point. The other factor is the slope of the line which is determined by the product of the diffusion coefficient and critical time (DTo). Influence of the several factors affecting zone size was considered in detail by Cooper 12,13 and Kavanagh. TM Conclusions from this discussion of diffusion assays are that the method is suitable for obtaining a quick indication of activity, and, if care is exercised, for obtaining a precise measure of a sample in terms of a standard. It cannot be used to obtain an absolute measure of susceptibility of bacteria or of activity of a drug. As ordinarily done, diffusion assays may determine concentration of a sample with errors varying from 1 to 10%. The former errors are for very carefully performed large plate assays. Because of poor control of operating factors, petri dish assays of drugs in nearly pure solution may have errors ranging from 5 to 10%. High accuracy and precision are much more difficult to achieve with diffusion than with turbidimetric methods. 1: K. E. Cooper, in "Analytical Microbiology" (F. Kavanagh, ed.), Vol. 1, p. 1. Academic Press, New York, 1963. 1~K. E. Cooper, in "Analytical Microbiology" (F. Kavanagh, ed.), Vol. 2, p. 13. Academic Press, New York, 1972. 14F. Kavanagh, J. Pharm. Sci. in press.
62
METHODS FOR THE STUDY OF ANTIBIOTICS
[4]
C. Dilution Methods
1. Agar Dilution The active substance is diluted in agar contained in petri dishes. The surface is streaked with several species of bacteria or small areas are inoculated with a large number of strains of bacteria by means of a replicator. The plates are incubated overnight at suitable temperatures and observed for growth of test bacteria. The lowest concentration of drug that shows no growth of a particular test organism is taken as the minimum inhibitory concentration (MIC). This very simple method is suitable for obtaining approximate MIC's for a number of bacteria simultaneously. The error in the determinations comes from both the design and the difficulty in determining the end point, which is not always sharp. Sharpness of end point depends, in part, upon the size of the step in concentration. Usually concentration changes by 2-fold steps. In other words, the concentrations form a geometric series, such as 1, 2, 4, 8, etc., just as in diffusion assays. In a 2-fold dilution test as described, there is an inherent uncertainty of 50% (1 step in the concentration scale) in determining the end point in addition to the subjective one mentioned above. Agar dilution is a widely used test of severely restricted accuracy.
2. Serial Dilution in Tubes Dilution in tubes was used early in the penicillin program to measure penicillin in experimental samples. It had two forms. One was a 2-fold serial dilution method and the other was one in which concentrations formed an arithmetic series of closely spaced values. The latter was considerably more accurate and precise than the serial dilution method. The end point in these tube assays is indicated by the absence of perceptible growth after 16 hr of incubation, by absence of hemolysis of red cells in the short-term test, 15 or by absence of luminescence of photobacteria. TM The last-named test could provide independent information on both suppression of luminescence and of growth of the test photobacterium. Details of the methods and their inherent errors are given by Kavanagh2 '17 One advantage of the tube dilution methods is the ease of performing tests with a large variety of organisms in many media. Results from such tests may be adequate for survey work early in a program. The dilution of the test substance causing inhibition may provide useful information 1~G. Rake and H. Jones, Proc. Soc. Expl. Biol. Med. 54, 189 (1943). ~eG. Rake, C. M. McKee, and H. Jones, Proc. Soc. Exp. Biol. Med. 51, 273 (1942) 1TF. Kavanagh, Bull. Torrey Bot. Club 74, 414 (1947).
[4]
ANTIBIOTIC ASSAYS
63
in the absence of a standard. A serial dilution test may be used for measuring MIC of clinical isolates but not for quantitation of active drugs when accuracy is important. 3. Photometric Methods
Serial dilution methods soon were found to be inadequate for the penicillin program and were supplanted by the method of measuring response of bacteria to graded concentrations of antibiotics. The first successful turbidimetric assay was that of McMahan. TM It was used essentially unchanged for nearly 25 years by laboratories in pharmaceutical companies, official bodies, and regulatory agencies. The method was derived from microbiological vitamin assay methods. It is very simple in principle. Graded quantities of drug are added to test tubes, inoculated broth added, tubes incubated until substantial growth has occurred in the control tubes, growth stopped, and concentration of bacteria in the tubes determined by a photometric method. Potency of samples is obtained by interpolation from a calibration curve prepared from responses of the standard tubes in each test. The assays can be of high accuracy. Both manual and partially automated methods are used. Details of both will be found in the appropriate chapters of the two volumes of "Analytical Microbiology. ''1,2 A turbidimetric method can be designed for nearly any antibiotic or antibacterial substance. A wide variety of gram-positive and gram-negative bacteria, yeasts, protozoa, and algae may be used as the test organism. Quite often the test organism is selected from the dominant group inhibited by the drug. The requirement is that growth of the test organism be inhibited by the drug and that the generation time be less than 60 rain, otherwise incubation time becomes excessive. The assay medium must support growth of the test organism and not interfere to any great extent with the action of the drug. The test organism should not be appreciably pathogenic. It must grow uniformly suspended, and neither clump, form strings, nor form a surface pellicle. In the range of concentrations of drug of interest in assaying, most. drugs interact with the test organism to reduce the growth rate. Therefore, the assay really is a rate method in which the integrated effect of reduction of growth rate is measured. Since a lag period also is involved, the total incubation time is longer than the log phase growth period. The length of the incubation is unimportant so long as it is sufficient and is exactly the same for standards and samples. Since the assays are growth rate methods and the medium has an upper limit to the maximum 18j. R. McMahan, J, Biol. Chem. 153, 249 (1944).
64
METHODS FOR THE STUDY OF ANTIBIOTICS
[4]
concentration of bacteria possible, unduly prolonging incubation time will eliminate the effect of the drug. Anything affecting growth rate other than drug will cause a bias in the assay. 19 The analyst must always be alert to such interferences. They are very common in feed assays but may be rare in pharmaceutical dosage forms and relatively pure preparations from chemical syntheses. An important interfering substance is a slightly different relative of the principle drug. An example would be deacetyl cephalothin in a sample of cephalothin. A theory of the relation between drug and organism for the simple case described above was given by Kavanagh. 19,2° The following theoretical equation is based on the work of Garrett and Miller 2~ N = No exp (ko - k~C)(t - no)
(1)
where N is the concentration of bacteria as number of organisms per milliliter after an incubation time of t, No is the concentration of bacteria at time zero, ko is the growth rate constant in the absence of antibiotic, k~ is the inhibitory coefficient characteristic of antibiotic and organism, C is the concentration of antibiotic, and Lo is the lag time from inoculation to the beginning of exponential growth. In any particular turbidimetric assay No, ko, ka, t, and Lo are constant for all values of C of both standard and sample; otherwise, the assay is invalid. Therefore, Eq. (1) may be reduced to the form logN = A-BC
(2)
A graph of N against C on semilogarithmic paper would be a straight line of slope B3 ° This equation has been a valuable dose-response line for obtaining potencies of unknowns. It has been applied to assay of penicillins, cephalosporins, and erythromycin in pharmaceutical dosage forms and of tylosin, monensin, hygromycin B, and tetracyclines in animal feeds. The equation does not apply to assays employing Klebsiella pneumoniae (ATCC 10031) for unknown reasons. The value of N in Eq. (2) can be obtained from photometric measurements with a calibrated instrument.19, 22 In preparing calibration lines for assays, absorbance is used in place of N to avoid the errors and inconvenience of converting absorbance into N. The line will be somewhat more curved than the log N vs. C line. 1~F. Kavanagh, in "Analytical Microbiology" (F. Kavanagh, ed.), Vol. 2, p. 44. Academic Press, New York, 1972. ~oF. Kavanagh, Appl. Microbiol. 16, 777 (1968). 21E. R. Garrett and G. I-I. Miller, J. Pharm. Sci. 54, 427 (1965). ~F. Kavanagh, in "Analytical Microbiology" (F. Kavanagh, ed.), Vol. 1, p. 141. Academic Press, New York, 1963.
[4]
ANTIBIOTIC ASSAYS
65
The dose-response line is drawn point-to-point. An important point is that the concentrations of standards must be on a linear scale and not form a geometric series as in diffusion assays.
IV. Assays of Diverse Biological Samples The implicit assumption made to this point ~in the discussion is that the samples are reasonably pure and derived from chemical operations. Such samples usually are available in adequate quantity for any kind of test and do not contain substances of biological origin that interfere with assays. The situation changes drastically when the antibiotics are tested in various biological systems because the samples are small and may contain interfering substances from the system. Usually the most difficult to assay accurately are blood samples. Urine samples may present fewer problems because the quantity is larger and, quite often, the concentration is higher than in blood. Diffusion assays have been the method of choice because the assays can be done with small samples and the diffusion process reduces interferences. Diffusion assays usually are more sensitive than tube methods because the most susceptible test bacteria are suitable for plate methods, but not for tube methods. Usually, tube methods are considered to be too insensitive for assay of blood samples. Unless the sample can be diluted substantially, the proteins and other materials in the plasma or serum may affect a tube assay. A problem more likely to occur in samples of biological origin is the presence of active metabolic products of the drug. An example would be deacetyl cephalothin 23 in blood or urine from animals given cephalothin. The metabolic product has considerably less activity than the parent compound when assayed by S. aureus. The derivative may be as active as the parent in certain plate assay systems. When this is so, the assay is a measure of activity, not of a quantity of a particular substance. Quite often the derivative is less active than the parent and may not be or may be only incompletely measured in the assay, which then becomes a method for the parent compound. However, in a turbidimetrie method, all active compounds will be measured because the test organism is exposed to all. There may be pharmacological reasons to use derivatives of the active drug to increase stability, to improve rate of absorption, to decrease toxicity, or to increase activity. Derivatives of several antibiotics have no activity until hydrolyzed to release the parent compound. When as~ R. J. Simmons, in "Analytical Microbiology" (F. Kavanagh, ed.), Vol. 2, p. 193. Academic Press, New York, 1972.
66
METHODS FOR THE STUDY OF ANTIBIOTICS
[4]
sayed without preliminary hydrolysis, apparent potency of the derivatives will depend on the kind of assay and extent of incubation. V. Relative Inhibitory Coefficient of Antibiotic8 The scientist who works with a series of related antibiotics, such as the penicillins or cephalosporins, needs a meaningful method for comparing antibacterial activities. The activity of each new compound must be compared with the parent or a model compound to learn whether the new molecular change has altered the antibacterial activities. The search is for products with higher activities and activities against a broader range of bacteria than the reference compound. Changes in molecular structure may cause an increase in activity, a decrease in activity, or no change, and it is necessary to know how much the change in structure has changed activity. The more accurately changes can be measured, the more reliably can the decision be made concerning the effect of molecular changes. To 1973, the great advances in preparing semisynthetic penicillins and cephalosporins were all made using a very crude measure of activities. How many superior compounds were missed and how much needless work was done because of inadequate measurements of antibacterial activities can not be estimated. In the work published, 24-26 activities were all reported as MIC (minimum inhibitory concentration) ascertained by either a tube or agar dilution method. The minimum inherent error 8,1° in MIC tube methods is 50% and usually is no less in agar dilution assays. These assays could not distinguish between compounds differing in activity by 20%. MIC is a quick way to ascertain influence of changes in structure upon activity against many organisms. The resulting spectrum of activities can be a valuable guide for selecting candidates for the more precise characterization of activities to be described next. Nonetheless, the assays are a great improvement over pure chance. A much more precise measure of activity of a new compound can be obtained by relating its activity to a parent compound or other reference substance of closely related structure. The factor that characterizes antibacterial activity of a compound in a particular turbidimetric assay is the specific inhibitory coefficient, ka, of Garrett and Miller. 21 The theoretical equation, Eq. (1), applicable to many assay systems relates inoculum size, growth constant (ko), specific inhibitory coefficient (ka), and concentration of antibiotic. 19,~° Because inoeulum size and 24j. p. Hou and J. W. Poole, J. Pharm. Sci. 60, 503 (1971). ~5 M. Misiek, T. A. Pursiano, L. B. Crast, :F. Leitner, and K. E. Price, Antimlcrob. Ag. Chemother. 1, 54 (1972). ~e M. Gorman and C. W. Ryan, in "Cephalosporins and Penicillins" (E. Flynn, ed.), Chapter 12, pp. 532-582. Academic Press, New York, 1972.
[4]
ANTIBIOTIC ASSAYS
67
growth constant vary from assay to assay and are inconvenient to measure, absolute values of ka are not determined. Inoculum size and growth constant of the test organism are the same for all antibiotics in one assay. The necessity for knowing them is obviated by obtaining the ratio of /ca of new compounds to k~ of reference compound. The ratio is called the relative intrinsic activity of the new substance relative to reference substance. The value is, of course, a constant only for a particular assay organism and changes with test bacteria. The value also changes with activities of the reference compound. The relative inhibitory coefficient can be easily measured in modern automated turbidimetric assays. Should absolute values of ka be needed, the method of Garrett and Miller 21 or a photometric equivalent ~9 can be used. A procedure was developed for determining the relative inhibitory coefficient for those antibiotics and test organisms that follow Eq. (1). By referring the inhibitory coefficients of several drugs to that of a reference compound, the influence of unavoidable variables in the test system will be reduced to second-order importance. The following assumptions are made: 1. The compound reduces growth rate of the test bacteria. 2. The compound does not affect lag time in a manner different from the standard. 3. The bacteria in all test solutions are still in the log phase of growth when growth is terminated. 4. The apparent growth rate constant is a linear function of concentration of antibiotic. 5. The apparent growth rate ( k o - k.~C) is constant throughout the incubation period for each concentration of each test substance. 6. Drug is not consumed by the test bacteria. 7. Organisms grown in the presence of the test compounds possess optical properties identical with those grown in the absence of the compounds, Given the above assumptions, the growth of test organisms in the presence of a growth-rate inhibitory compound may be expressed by Eq. 1. In an assay, ko, t - - L o , and No are the same for all tubes, are unknown and are not easily measured. However, the value of k , ~ ( t - L~,) can be obtained from measurements of N for two values of C for each antibiotic as follows:
ka(t -- L0) = [2,3 log (NI/N2)]/(C2 - CI)
(3)
where N1 is the concentration of bacteria at antibiotic concentration C1 and N2 the concentration at C2. Although total incubation time can be measured accurately, t - Lo is unknown because the value of L0 is unknown. The values of (t - L0),
6S
M~.T~ODS ~Oa TI~. STVOY Or ANTIBIOTICS
[4]
and N o , are different in different tests even though the tests are prepared from the same inoculated broth. A way around the obvious difficulties is to refer all activities to a standard compound in the test and to treat each test as a unit. Represent the k~ of the reference substance by t l! t! t k,, and the k, of samples by k~. The ratio k~/k~ should be independent of No, t, L0, and k0 because kT(t -
Lo)/k'~(t - Lo) = kT/k'~ = R
(4)
where R is the relative inhibitory coefficient. R is a function of the activity of the reference substance as well as of the sample. If the same reference compound is used in testing a series of derivatives, the several values of the R may be used to measure the influence of a change in molecular structure upon activity. As a practical matter, medium, buffer, size of inoculum, and condition of inoculum are controlled. Restrictions are placed on the assay systems by choosing concentration of antibiotics so that at least one permits growths between 50 and 70% of uninhibited growth and at least one permits growths between 50 and 30%. In addition, a plot of log N vs C on semilog paper of N obtained for at least three concentrations of drug including the above two concentrations should approximate a straight line. The inhibitory coefficients are computed from activities around the 50% growth point of the inhibition curve and gives no indication of activity near the ends of the curve. The full range of activities are better shown in a log-probability plot. 19,22 In these graphs, bacterial concentration relative to uninhibited growth (C = 0) for a number of concentrations of antibiotic are plotted against corresponding C on log-probability paper. The lines should be slightly curved toward the C axis. A family of compounds should form a series of approximately parallel lines. Only for such families can the relative inhibitory coefficients have meaning. The compounds may be expected to have different MR (median response), which is defined as the concentration of antibiotic permitting 50% growth. A procedure for computing relative inhibitory coefficients from the several median responses can be derived from Eq. (1). Although the two procedures have equal theoretical validity, the one given is preferred because no assumptions must be made about linearity of the line representing uninhibited growth. Many measurements of numerous systems showed inhibited growth following the theoretical growth equation much more accurately than uninhibited growth. Lack of linearity of the uninhibited growth causes the computed median responses to be larger than the true value.
[5]
f~-LACTAMASE ASSAYS
69
Nothing can be learned from the dose-response line about resistance to enzymic attack, animal toxicity, absorption and excretion, and effectiveness in treating an infected animal. The dose-response line is the quickest way to obtain an idea of the quantitative response of the test organisms to graded concentrations of the antibiotic. A high accuracy turbidimetric assay such as that provided by an AUTOTURB ® System should be used to obtain the basic data. A reference standard and five samples at three concentrations would constitute one test in the AUTOTURB ® System. The important parameter in Eq. (3) are the values of N which are obtained from a calibration curve relating transmittance or absorbance of the suspension to concentration of bacteria.19, 22 In a study of one series of commercially important antibiotics using S. aureus as the test organism, R ranged from 0.03 to 1.56. The value of R, obviously, can have nearly any value. If the molecular weights of the compounds vary significantly within a series, C should be in molar and not in weight concentrations of active compounds.
[5] f~-Lactamase Assays B y GORDON W. Ross and CYNTHIA H. O'CALLAGHAN I. I n t r o d u c t i o n . . . . . . . . . . . . . . . . . . II. D e f i n i t i o n of U n i t . . . . . . . . . . . . . . . . . III. Assay Methods . . . . . . . . . . . . . . . . . A. M e t h o d s in W h i c h Penicillin a n d C e p h a l o s p o r i n B r e a k d o w n P r o d u c t s Are M e a s u r e d . . . . . . . . . . . . . . B. M e t h o d s in W h i c h t h e C o n c e n t r a t i o n of R e s i d u a l Penicillin or C e p h a l o s p o r i n Is M e a s u r e d . . . . . . . . . . . . . IV. D i s c u s s i o n of A s s a y M e t h o d s . . . . . . . . . . . . . A. I o d o m e t r i c A s s a y . . . . . . . . . . . . . . . . B. A c i d i m e t r i c A s s a y . . . . . . . . . . . . . . . . C. S p e c t r o p h o t o m e t r i c (UV) A s s a y . . . . . . . . . . . D. H y d r o x y l a m i n e A s s a y . . . . . . . . . . . . . . .
I. H R--CO--N~.H
C--C [
r
H
A penicillin
C
J
CH
oo"
74 80 82 83 84 84 85
Introduction
CHs H ~ . ~ [/CH s
o~C--N
69 74 74
÷ ..o
CH~
R-- CON..H HIS~ . 11CHs C--C C I i J
o ,C
OH
CO0"
A penicilloic acid
(I)
[5]
f~-LACTAMASE ASSAYS
69
Nothing can be learned from the dose-response line about resistance to enzymic attack, animal toxicity, absorption and excretion, and effectiveness in treating an infected animal. The dose-response line is the quickest way to obtain an idea of the quantitative response of the test organisms to graded concentrations of the antibiotic. A high accuracy turbidimetric assay such as that provided by an AUTOTURB ® System should be used to obtain the basic data. A reference standard and five samples at three concentrations would constitute one test in the AUTOTURB ® System. The important parameter in Eq. (3) are the values of N which are obtained from a calibration curve relating transmittance or absorbance of the suspension to concentration of bacteria.19, 22 In a study of one series of commercially important antibiotics using S. aureus as the test organism, R ranged from 0.03 to 1.56. The value of R, obviously, can have nearly any value. If the molecular weights of the compounds vary significantly within a series, C should be in molar and not in weight concentrations of active compounds.
[5] f~-Lactamase Assays B y GORDON W. Ross and CYNTHIA H. O'CALLAGHAN I. I n t r o d u c t i o n . . . . . . . . . . . . . . . . . . II. D e f i n i t i o n of U n i t . . . . . . . . . . . . . . . . . III. Assay Methods . . . . . . . . . . . . . . . . . A. M e t h o d s in W h i c h Penicillin a n d C e p h a l o s p o r i n B r e a k d o w n P r o d u c t s Are M e a s u r e d . . . . . . . . . . . . . . B. M e t h o d s in W h i c h t h e C o n c e n t r a t i o n of R e s i d u a l Penicillin or C e p h a l o s p o r i n Is M e a s u r e d . . . . . . . . . . . . . IV. D i s c u s s i o n of A s s a y M e t h o d s . . . . . . . . . . . . . A. I o d o m e t r i c A s s a y . . . . . . . . . . . . . . . . B. A c i d i m e t r i c A s s a y . . . . . . . . . . . . . . . . C. S p e c t r o p h o t o m e t r i c (UV) A s s a y . . . . . . . . . . . D. H y d r o x y l a m i n e A s s a y . . . . . . . . . . . . . . .
I. H R--CO--N~.H
C--C [
r
H
A penicillin
C
J
CH
oo"
74 80 82 83 84 84 85
Introduction
CHs H ~ . ~ [/CH s
o~C--N
69 74 74
÷ ..o
CH~
R-- CON..H HIS~ . 11CHs C--C C I i J
o ,C
OH
CO0"
A penicilloic acid
(I)
70
METHODS FOR THE STUDY OF ANTIBIOTICS
,I
i
I
CO0A cephalosporin
+
H20
~-
C~ C I
o.
I
[5]
CHi I
oo-
7/ I
]I
(2)
Unstable intermediate
Breakdown products
fl-Lactamases [penicillin (cephalosporin) fl-lactam amidohydrolases, EC 3.5.2.6] are enzymes of bacterial origin which hydrolyze the C - - N bond in the fl-lactam ring of a penicillin or a cephalosporin [Eqs. (1) and (2), respectively]. The term fl-laetamase covers a variety of enzymes from a variety of different organisms, fl-Lactamase activity has been detected in all organisms examined to date when an isoelectric focusing technique has been used. 1 The effect of fl-lactamases was first noted with benzylpenicillin (penicillin G), and this resulted in the term penicillinase being applied to any biological entity that rapidly inactivated benzylpenicillin. 2 With the discovery of the cephalosporins, it became evident that the term penicillinase was too narrow to encompass the many different enzymes capable of catalyzing the hydrolysis of the fl-lactam ring of cephalosporins and penicillins, and the general name fl-lactamase is now used. fl-Lactamases from at least 25 different strains of bacteria have been purified, and properties such as substrate profile, molecular weight, effect of pH and temperature, electrical charge, and susceptibility to inhibition have been studied. M a n y other fl-lactamases have been observed but not studied in detail. In a recent review, a classification of fl-lactamases from gram-negative bacteria into five main classes containing 15 different types of enzyme was proposed2 Some fl-lactamases are highly species specific, such as the Sabath and Abraham enzyme found in all strains of P s e u d o m o n a s aeruginosa, 4 whereas others, being mediated by an R factor such as RTEM,5 m a y be found in several different species. Some enzymes inactivate cephalosporins very much more rapidly than they inactivate penicillins, but even so their activity can usually be expressed in terms of benzylpen!cillin. If the main interest is in cephalosporins, 1M. Matthew, A. M. Harris, M. J. Marshall, and G. W. Ross, J. Gen. Microbiol., in press (1975). 2E. P. Abraham and E. Chain, Nature (London) 146, 837 (1940). s M. H. Richmond and R. B. Sykes, Advan. Microb. Physiol. 9, 31 (1973). 4L. Sabath, M. Jago, and E. P. Abraham, Biochem. J. 96, 739 (1965). N. D a t t a and M. H. Richmond, Biochem. J. 98, 204 (1966).
[5]
~-LACTAMASIg ASSAYS
71
then often the activity is related to that against cephaloridine, rather than benzylpenicillin. In the course of the past 30 years, many penieillins and cephalosporins have been prepared and tested as substrates for many fl-lactamases. Most of the compounds tested have been found to be susceptible to at least one fl-lactamase; recently, a new type of cephalosporin known generically as cephamyeins, has been described as highly resistant to most fllaetamases. As yet, little is known about the behavior of these compounds in the assay methods given here. Whatever the preparation of enzyme, or the purpose of the assay, the concentration of the fl-lactamase is primarily estimated from its effect on a known concentration of benzylpenicillin and/or cephaloridine, although its activity on other penicillins or cephalosporins may be tested in parallel at the same time. Methods for assaying fl-laetamase activity can be divided into two main groups. In one, the concentration of the unchanged substrate remaining is measured; in the other, the assay determines the concentration of breakdown products. In general, benzylpenieillin is comparatively simple to assay by a variety of methods. However, if these methods are extrapolated to other penieillins, care must be taken that the method chosen is still applicable. This is even more important with eephalosporins, and it cannot be assumed that the assay method selected as the most convenient for the estimation of benzylpenieillin will work in the same way, or even work at all for a particular cephalosporin. The fl-laetamase preparation may also affect the assay. A highly purified enzyme is tedious and time consuming to make but usually gives good results with most assays. However, if the preparation is very impure and has, for example, strongly reducing tendencies, then this can seriously interfere with several assay methods. When the fl-laetam ring of a penicillin is hydrolyzed by a fl-laetamase IEq. (1) 1, the corresponding penicilloic acid is produced in stoichiometrie proportions. A penieilloic acid has two acidic groups where the parent compound had only one [Eq. (1) ], and it is also a stable substance which can be readily assayed. The situation with eephalosporins is more complicated owing to the presence of another substituent on the nucleus at position 3. The first product of fl-laetamase attack on a eephalosporin is hypothetically a eephalosporoie acid, analogous to penieilloie acid [Eq. (2)]. However, unlike penieilloie acids, these compounds are usually very unstable and cannot be isolated because they undergo rapid decomposition. Only two exceptions are known to this, one being eephalosporins, where the 3-sub-
72
METHODS H
R--CO--N
F O R T H E S T U D Y OF ANTIBIOTICS
[5]
H
S-.
R--CO--N~_~S.~
0~ ~0
COO-
(a)
/---" NO~
(b)
FIG. 1. Stable products of hydrolysis of #-lactam ring of cephalosporins formed when 3-substituent is (a) a ~/-lactone or (b) 2:4-dinitrostyrene.
O ~ 9 " - - - - N ~ _ CH2OCOCHs COO
L
coo~
L
j
(Hypothetical, transient, intermediates)
coo-
/
Further fragmentation FIG. 2. Production of a second acidic group from cephalothin on fl-lactam hydrolysis. s t i t u e n t is a ,/-lactone 6 (Fig. la) ; the other comprises compounds where the 3-substituent is the highly conjugated s y s t e m of 2 : 4 - d i n i t r o s t y r e n e , 7 (Fig. l b ) . 6 E. P. Abraham and G. G. F. Newton, Biochem. J. 79, 377 (1961). 7 C. H. O'Callaghan, A. Morris, S. M. Kirby, and A. H. Shingler, Antimicrob. Ag. Chemother. 1, 283 (1972).
[5]
/~-LACTAMASE ASSAYS
73
CH~--CO--N..~_~S-~ 0
l_
CH2---N
~oo--
coo-
/}
1
Fia. 3. Decomposition products of eephaloridine formed on rupture of ~-laetam ring.
Cephalothin gives the expected acidic group from the fl-lactam ring but, in addition to this, the acetoxy group at position 3 is simultaneously expelled from the molecule, producing another acidic group 4 (Fig. 2). Thus, unlike the penicillins, cephalothin gives two additional acidic groups. Other cephalosporins with an acetoxy group at position 3, such as cephaloglycin, cephapirin, and cephacetrile, also give extra acid groups on fl-lactamase hydrolysis. This does not apply to all cephalosporins. Cephaloridine, for example (Fig. 3), also loses its 3-substituent simultaneously with rupture of the fl-lactam ringS; this 3-substituent is pyridine--a base, not an acid. Therefore, while the intact compound is a betaine with no net charge, the initial decomposition products have two acidic groups and one basic group. For each different cephalosporin, the overall effect will depend on the relative strengths of the acidic and basic groups. Not all cephalosporins lose their 3-substituent when the fl-lactam ring is hydrolyzed. Cephalexin retains its methyl group at position 3. Compounds such as cefazolin and cephamandole probably lose their 3-substituent, but this reaction may not proceed as rapidly as the decomposition of the fl-lactam ring, and the nature of the decomposition products is not yet firmly established. It is thus evident that while the determination of benzylpenicillin in a fl-lactamase reaction mixture is a.relatively simple matter, the situation with cephalosporins is much more complex. 8 C. H. O'Callaghan, S. M. Kirby, A. Morris, R. E. Waller, and R. E. Duneombe, J. Baeteriol. 110, 988 (1972).
74
METHODS FOR T H E STUDY OF ANTIBIOTICS
[5]
II. Definition of Unit A unit of fl-lactamase activity is commonly defined as the amount of enzyme which hydrolyzes 1 ~mole of substrate per minute at 37 ° and optimum pH. Measurement of enzyme activity at 37 ° rather than the 25 ° or 30 ° often used is preferred for these bacterial enzymes, which will usually be acting in vivo at 37 °. A standard pH of 7.0 has often been used for the same reason. The unit defined by Pollock and Torriani 9 (one unit hydrolyzes 1 ~mole of benzylpenicillin per hour at 30 ° and pH 7.0) has also been frequently used in fl-lactamase studies. Because of the diversity of the fl-lactamases and the many independent groups of workers investigating various aspects of the problem, there has been a notable lack of correlation between methods and conditions used. This has made comparisons between various sets of results ~lmost impossible in many cases. There is a great need to standardize methods and conditions to make comparisons possible. The obvious conditions to use should be those most closely related to the bacterial environment in which the fl-lactamases will do the most damage, i.e., pH 7 and 37%
III. Assay Methods
A. Methods in Which Penicillin and Cephalosporin Breakdown Products Are Measured 1. Iodometric Assay
This method is the most widely used at the present time, and several variations have been reported. Perret's macroiodometric method 1° is relatively straightforward, gives reproducible results, and does not require any specialized or expensive equipment. An excess of iodine, buffered at pH 4, is used to stop the enzyme reaction. The hydrolysis products of the fl-lactam antibiotic react with the iodine and the remaining iodine is measured by titration with sodium thiosulfate solution. Full details of this method are given below. Sargent 11 has described a variation of the Perret method in which the decrease in optical density at 490 nm is used instead of thiosulfate ' M. R. Pollock and A. M. Torriani. C. R. Acad. Sci. Paris 237, 276 (1953). soC. J. Perret, Nature (London) 174, 1012 (1954). 11M. G. Sargent, J. Bacteriol. 95, 1493 (1968)
[5]
/~-LACTAMASE ASSAYS
75
titration as a measure of iodine uptake by the products of enzyme action. Ferrari et al. 12 have described an Autoanalyzer method in which iodine uptake is followed spectrophotometrically at 420 nm. Zyk 13 has suggested that the enzyme reaction should be stopped by addition of iodine-tungstate solution before measuring the decrease in optical density at 620 nm due to the reaction of iodine with the products of fl-lactam hydrolysis. The macroiodometric methods, although easy to use, are relatively insensitive and cannot be used, for example, for studies on reaction kinetics. Several microiodometric techniques have been described in which a 1000-fold increase in sensitivity can be obtained by following spectrophotometrically the rate of decolorization of the starch-iodine complex by the penicilloic acid. Novick" followed the decolorization at 620 nm, but his method gave results that were 40% lower than those obtained by Perret's method. Sykes and Nordstrom 1~ showed that this discrepancy could be eliminated if sufficient time was allowed for a steady state to be reached between formation of penicilloic acid and its reaction with the starch-iodine complex; their method is given in detail below. None of the microiodometric assays has been applied successfully to cephalosporin substrates. Lindstrom and Nordstrom 16 adapted Novick's method for use on an Autoanalyzer. Goodall and Davies 1~ had previously described an automated procedure in which excess iodine remaining after reaction with the penicilloic acid was mixed with a starch-potassium iodide reagent and the resulting starch-iodine complex was measured at 526 nm. Cole et al. ~s described an automated procedure in which uptake of iodine was followed by measuring the reduction in height of the 288 nm absorption peak of the iodine/potassium iodide reagent. Mavroiodometric Determination of fl-Lactamase Activity Reagents
Phosphate buffer, 0.1 M, pH 7.0 Acetate buffer, 2 M, pH 4.0 Hydrolyzed starch, 2% Sodium thiosulfate, 0.0166 N (41.19 g/liter, dilute 1:10 v/v before use) 1,~A. Ferrari, F. M. Russo-Alesi, and J. M. Kelly, Anal. Chem. 31, 1710 (1959). 13N. Zyk, Antimcirob. Ag. Chemother. 2, 356 (1972). 1, R. P. Novick, Biochem. J. 83, 236 (1962). 15R. B. Sykes and K. Nordstrom, Antimicrob. Ag. Chemother. I, 94 (1972). I"E. B. Lindstrom and K. Nordstrom, A~timicrob. Ag. Chemother. 1, 100 (1972). 1~R. R. Goodall and R. Davies, Analyst 86, 326 (1961). ~8M. Cole, S. Elson, and P. D. Fullbrook, Biochem. J. 127, 295 (1972).
76
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
Iodine reagent is 0.0166 N iodine, 60 mM potassium iodide and 2 M acetate buffer, pH 4.0. Dissolve 40.6 g of iodine + 200 g of KI in 1 liter of demineralized water. Dilute 1:20 with 2 M acetate buffer, pH 4.0, just before use. Substrate (5 mM) is prepared daily in phosphate buffer pH 7.0 Procedure. The reaction is carried out in a shaking water bath at the appropriate temperature.
50-ml flasks Addition
Control
Test
Substrate (ml) E n z y m e (ml)
5 --
5 1.0
Allow substrate and enzyme to reach operating temperature before addition of enzyme. Incubation in the shaking water bath is continued for 30 min before t h e reaction is stopped by addition of iodine reagent (10 ml) ; 10 ml of iodine reagent is also added to the control flask, followed by 1 ml of enzyme preparation. Ten minutes later (20 min for cephalosporins) the flasks are removed from the bath and titrated with 0.00166 N thiosulfate, using starch as indicator. If the titration is less than 4.5 ml, the assay is repeated with diluted enzyme. Under these conditions, 1 ml of 0.0166 N iodine is equivalent to 2 ~moles of a penicillin destroyed. The macroiodometric assay of cephalosporins is less reliable than with the penicillins, largely because the breakdown pattern of these molecules is less clear cut; 1 ml of 0.0166 N iodine is equivalent to about 4 t~moles of a cephalosporin destroyed, but the stoichiometry of the reaction varies somewhat with the nature of the cephalosporin. Calculation. C o n t r o l - test = x ml of iodine consumed; .'. x/15 = t~moles of penicillin substrate destroyed per minute, and x/7.5 = t~moles of cephalosporin substrate destroyed per minute. Microiodometric Determination o] fl-Lactamase Activity Reagents
Phosphate buffer, 0.1 M, pH 7.0 Hydrolyzed starch, 0.2% Iodine reagent 0.08 M in 3.2 M potassium iodide (2 g of I2 + 53.29 g of KI in 100 ml)
[5]
~-LACTAMASE ASSAYS
77
Starch iodine solution is prepared by adding 0.15 ml of iodine reagent to 100 ml of starch solution (giving a final iodine concentration of 120 t2Y/). Substrate (0.20 mM) in phosphate buffer, pH 7.0, is prepared daily and kept on ice for the period of the experiment. Procedure. The reaction is followed in a recording spectrophotometer with a heated cell carrier.
Cuvettes
Addition Starch-iodine (ml) Substrate (ml) Phosphate buffer (ml) E n z y m e (ml)
1
2
3
E n z y m e control
Substrate control
Test
1 -l. 9 0.1
1 1 1 --
1 1 0.9 0.1
Allow cuvettes to reach operating temperature in heated cell carrier before initiating reaction by addition of enzyme. Absorbance at 620 nm is then measured at various times. The initial absorbance of the normal assay mixture is 1.20. The reaction becomes linear after 15-20 minutes for enzyme activities up to 0.001 unit per assay mixture. At higher activities the starch-iodine is completely decolorized before the steady state is achieved. Calculation. The amount of iodine used in the assay corresponds to 30 nmoles of penicillin. At this concentration, the optical density (OD) at 620 nm decreases from 1.20 to 0 when a penicillin is completely destroyed by a fl-lactamase..'. ( 5 0 D / m i n / 1 . 2 0 ) X 0.03 X 10 = ~moles destroyed per minute per milliliter of enzyme. 2. Acidimetric Methods (or Alkalimetric Methods) These methods depend fundamentally on measurement of the rate of increase in acidic groups when the fl-lactam ring is ruptured. This rate can be measured directly by continuous titration with alkali in a pH stat, ~,19 or else by the change in color of an indicator which can be continuously recorded on a spectrophotometer. ~° It can be determined indil~j. p. Hou and J. W. Poole, J. Pharm. Sci. 61, 1594 (1972). 2oF..4.. Rubin and D. H. Smith, Antimicrob. Ag. Chemother. 3, 68 (1973).
78
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
rectly by manometric measurement of carbon dioxide continuously released from a sodium bicarbonate solution by the increasing acidity.21 The techniques are relatively simple for penicillins but more difficult to apply to cephalosporins. a. D i r e c t M e t h o d s
Wise and Twigg 22 suggested a method for determining penicillinase activity in which the acid produced by the enzyme action was titrated with 0.01 N NaOH to a constant pH. (They chose pH 7.8, the optimum for their enzyme.) No indicator was used. Selzer and Wright ~3 used a similar method when studying inhibition of penicillinase by penicillinaseresistant penicillins. Sodium hydroxide (0.1 N) was added from a burette and the pH maintained at 7.0. Later workers used a pH star and different conditions. Sabath, Jago, and Abraham 4 hydrolyzed benzylpenicillin and cephalosporin C with a fl-lactamase from P. aeruginosa. They used Radiometer pH-stat equipment to measure uptake of 0.1 N sodium hydroxide solution, and hence the rate of fl-lactam hydrolysis when investigating pH activity curves and Michaelis constants. Hou and Poole 19 have recently used a similar method when investigating a variety of new penicillins. Details of this method are given below. An indicator method was first described by Saz et al., ~4 and the sensitivity of the method has recently been increased by Rubin and Smith 2° by decreasing the buffer concentration in the reaction mixture. They added enzyme solution (10 t~l, capable of hydrolyzing 2-10 nmoles of substrate per minute) to a reaction mixture (1.2 ml) containing substrate (0.2 raM), phenol red indicator (0.0013% w/v), and phosphate buffer (0.4 mM, pH 7.6). The absorption at 558 nm was followed with a recording spectrophotometer for 5 min at the temperature chosen for the assay. The assay should be standardized for each substrate by relating change in absorption at 558 nm to micromoles of substrate hydrolyzed. Imsande 2~ described an indicator method in which bromothymol blue was used instead of phenol red. The change in absorption at 620 nm of a reaction mixture containing substrate (6 mM), bromothymol blue indicator (0.0033% w/v), and cacodylate buffer (3 mM, pH 7.3) was followed for 12 min after addition of enzyme. 21R. J. Henry and R. D. Housewright,J. Biol. Chem. 167, 559 (1947). W. S. Wise and G. H. Twigg,Analyst 75, 106 (1950). ~8G. B. Selzer and W. W. Wright, in "Antimicrobial Agents and Chemotherapy" (G. Hobby, ed.), p. 311. AmericanSociety for Microbiology (1964). ~A. K. Saz, D. L. Lowry, and L. J. Jackson, J. Bacteriol. 82, 298 (1961). 25j. Imsande, J. Bacteriol. 69, 1322 (1965).
[5]
~-LACTAMASE ASSAYS
79
pH Stat Titration Method Reagents Potassium hydroxide solution, 0.04 N Substrate (5 mM) in dilute aqueous potassium chloride solution (ionic strength 0.025). Prepared daily and kept on ice before use Nitrogen gas washed with 10% sulfuric acid, 10% sodium hydroxide solution, and finally water. Procedure. Substrate solution (20 ml, 100 ~moles) is equilibrated to the operating temperature by stirring in a water-jacketed reaction vessel. A slow stream of the nitrogen gas is allowed to flow over the surface of the solution. B-Lactamase (less than 0.5 ml containing sufficient activity to hydrolyze 2-10 ~moles of substrate per minute) is added, together with enough dilute acid or base to set the initial pH. The rate of hydrolysis of 50% of the substrate is followed by the rate of addition of alkali required to maintain a constant pH, and the enzyme activity (~moles per minute) can be obtained from a calibration curve relating enzyme units to the time required for 100% hydrolysis of the 100 ~moles of substrate. b. Indirect Method Manometric measurement of released carbon dioxide was first described by Henry and Housewright21 in 1947 for the estimation of the rate of decomposition of benzylpenicillin by Bacillus cereus. The reaction is carried out in constant-volume Warburg respirometers. The enzyme solution and the sodium bicarbonate solution (total volume 4 ml) are placed in the main body of the vessel. The sidearm of the vessel contains 0.5 ml of penicillin solution, the penicillin being dissolved in the same concentration of bicarbonate solution as is contained in the main body of the vessel. The vessels and manometers are moved to and fro through a distance of 3 cm at a rate of at least 90 cycles per minute (to permit maximum C02 evolution); readings are taken every 2 rain until the reaction system is in equilibrium with the gas mixture. The penicillin solution in the sidearm is tipped into the main flask, and the evolution of CO~ begins. Readings are taken every minute, and often there will be a lag period of up to 4 rain before the rate of CO~ evolution becomes constant. The rate is measured in mm3/min of CO,.. evolved. The amount evolved is plotted against time, and the slope of the straight portions of the line is determined by the method of least squares. Concentrations of penicillin and enzyme must be adjusted so that the evolution of CO,, is fairly brisk. If the enzyme solution is excessively
80
METHODS FOR THE STUDY OF ANTIBIOTICS
[51
dilute, so that only a small amount of penicillin is hydrolyzed, then the amount of COz is too small to measure accurately. Similarly, if the penicillin solution is too dilute, complete decomposition occurs too quickly for the rate to be accurately measured, and the assay has a very low precision. The best results are obtained for a penicillin concentration of about i mM in 0.2% sodium bicarbonate solution. The preparations used must be in solutions of low buffering capacity, otherwise the C02 is not evolved in stoichiometric proportions. Iron, calcium, copper, and zinc ions interfere with the assay at levels as low as 5 t~g/ml for Fe 3÷. This method is not always readily extrapolated to cephalosporins, and, ideally, pilot experiments should be done first with new enzyme preparations to determine the best concentration of the enzyme preparation to use. It is a laborious and time-consuming method of penicillinase assay, and of poor reliability for cephalosporins; this method is now mainly obsolete. B. Methods in Which the Concentration of Residual Penicillin or Cephalosporin Is Measured
I. Spectrophotometric (UV) Assay The method described by O'Callaghan, Muggleton, and Ross 26 depends on the direct measurement of eephalosporin concentration and has proved to be particularly useful for comparison of the fl-lactamase resistanee of new eephalosporins and for/~-lactam inhibition studies. The rate of hydrolysis of the fl-laetam ring is followed by measuring the rate of decrease in optical density of a cephalosporin solution at the wavelengths of maximum absorption associated with the fl-lactam ring. This wavelength is 255 nm for cephaloridine; it must be determined for each analog assayed, together with the decrease in optical density at this wavelength for 100% hydrolysis. In a typical assay with cephaloridine as substrate, the concentration of the enzyme solution is adjusted so that a 10-15 ~1 sample mixed with 3 ml of a 0.1 mM (41.5 ~g/ml) solution of cephaloridine in 0.1 M phosphate buffer, pH 7.0, in a 1 cm euvette and incubated at 37°C will completely hydrolyze the cephaloridine in 5-10 min. The rate of decrease in optical density at 255 nm is followed for the first minute in a recording spectrophotometer (e.g., Unieam SP800). The change in optical density must be measured for each cephalosporin 2~C. H. O'Callaghan, P. W. Muggleton, and G. W. Ross, in "Antimicrobial Agents and Chemotherapy" (G. Hobby, ed.), p. 57. American Society for Microbiology, 1968.
[5]
~-LACTAMASE ASSAYS
81
being investigated. For example, the optical density of 0.1 mM solution of cephaloridine falls from 1.4 to 0.6 for complete hydrolysis; therefore, t~moles cephaloridine hydrolyzed/minute = (A OD/min/0.8)}( [(3 }( 41.5)/415] = A OD/min )< ~ . The range of the assay can be extended to cover other substrate concentrations by the use of optical cells of different light paths. 2. Hydroxylamine Assay This assay is based on that of Boxer and Everett :7 as modified by Hamilton-Miller et al. 2s Hydroxylamine reacts with the intact fl-lactam compound to give a hydroxamic acid which forms a chromogen with ferric ions. Reagents 1. Hydroxylamine hydrochloride (174 g in 500 ml of solution in water) 2. Buffer, pH 11.2 (173 g of sodium hydroxide + 20.6 g of anhydrous sodium acetate in 1 liter of solution) 3. Absolute alcohol + reagent 1 + reagent 2 + demineralized water (4:1 : 1 : 1.5 v/v) ; made freshly each day 4. Ferric ammonium sulfate (200 g) + concentrated sulfuric acid (95 ml) added to demineralized water to give 1 liter of solution 5. Substrate, 1 mM in 10 mM phosphate buffer, pH 7.0 Procedure. Reagent 3 (7.5 ml) is added to dilutions of substrate [1 ml in wide (diameter 1 inch) glass tubes] and mixed thoroughly (Vortex mixer). After 3..5 min (critical for cephalosporins) reagent 4 (2 ml) is added and mixed thoroughly. The mixture is immediately read in a spectrophotometer at 490 nm or in a colorimeter (623 filter) against a blank of reagent 3 (7.5 ml), phosphate buffer (1 ml), and reagent 4 (2 ml). Calibration curves are obtained for each substrate; they are usually straight lines passing through the origin (e.g., cephaloridine from 75 to 1500 ~g/ml). Substrate (1 mM) is incubated with enzyme at 37 ° for 5 min. Samples (1 ml) are assayed for residual substrate as described above. Five replicate samples are usually assayed. 3. Biological Assay Residual cephalosporin or penicillin can be estimated after fl-lactamase attack by biological plate assay against a suitable organism. Princi27G. E. Boxer and P. M. Everett, Anal. Chem. 21,670 (1949). J. M. T. Hamilton-Miller, J. T. Smith, and R. Knox, Nature (London) 208, 235 (1965).
82
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
ples and methods are given in detail in this volume [4]. Choice of the organism will depend on the particular substrate being assayed; for example, cephaloridine is usually assayed against Staphylococcus aureus, Bacillus subtilis, or Sarcina lutea. These organisms are all highly sensitive to ccphaloridine and will assay concentrations down to 0.1 ~g/ml. However, solutions of 12-500 ~g/ml can be assayed using Escherichia coll. This organism has the advantage that it is not sensitive to many other cephalosporins and penicillins, so that it will measure cephaloridine in the presence of a substance, such as cloxacillin, which may be present because it is being used as an enzyme inhibitor. It is very necessary, when using a biological method of estimating residual antibiotic, to stop the reaction immediately the sample for assay has been taken. This can be done by the addition of an iodine solution which is adequate to inactivate the enzyme without interfering with the growth of the assay organism. 29 Different enzymes require different amounts of iodine to inactivate them; the enzyme from Enterobacter cloacae P99 can be inhibited by 10 ~M iodine, but the one from E. coli TEM ÷ is not inactivated until the concentration of iodine is 0.7 mM. Method. Iodine reagent: 17.8 mg of iodine is dissolved in 100 ml of 0.06% potassium iodide solution. The chosen substrate at 1 mM is incubated with the enzyme preparation at 37 ° for 5 min. The reaction mixture is then diluted 100-fold with the iodine reagent (0.7 mM) to stop the enzyme reaction, and residual antibiotic assayed by a suitable plate diffusion assay. The concentration of iodine used does not interfere with the assay of antibiotic activity by either Staphylococcus aureus NCTC 7447 or Bacillus subtilis NCIB 8533. 59 IV. Discussion of Assay Methods The range of the different fl-lactamase assays described above and the large number of variations of these assays appearing in the literature, gives some indication of the widespread interest in fl-lactamases. The methods used to estimate their range and potency are numerous and modifications are constantly being made to the basic techniques. Consequently, it is very difficult to compare the results obtained by different groups of workers, although at least one attempt has been made to correlate the results obtained by running five methods in parallel, using 10 substrates and 3 fl-lactamases. ~9 At present, no single assay would seem to satisfy all requirements although there are two or three with definite advantages. aSG. W. Ross, K. V. Chanter, A. M. Harris, S. M. Kirby, M. J. Marshall, and C. H. O'Callaghan, Anal. Biochem. 54, 9 (1973).
[5]
~-LACTAMASE ASSAYS
83
For those workers requiring a rapid and sensitive assay for studying the fl-lactamase sensitivity and inhibitory activity of a large number of cephalosporin derivatives, the spectrophotometric assay is the method of choice. For similar studies on penicillin analogs, one of the recent microiodometric assays is suitable. If both cephalosporin and penicillin substrates are being compared, then microbiological assay is a very sensitive way of measuring residual substrate but is not well suited to enzyme kinetics. Some acidimetric assays offer rather less sensitive but more precise methods for measuring fl-lactamase hydrolysis of penicillins and cephalosporins, provided that the breakdown products of the cephalosporin analogs are sufficiently well characterized for the method to be properly calibrated. If the assay requirement is for a rapid, reliable but comparatively insensitive method of comparing the effect of fl-lactamase on a range of penicillins and cephalosporins, without the need for expensive equipment, then the macroiodomeric method of Perret is the obvious choice. If a method is used which requires intermittent sampling and reading, then the enzyme activity must be stopped at the time of sampling. In some assay methods, the reagents will themselves inhibit the action of the enzyme, for example, addition of hydroxylamine or iodine. This problem does not occur if a continuous recording device is employed.
A. Iodometric Assay Perret's macro assay is relatively straightforward, and reproducible results can be obtained quickly using inexpensive reagents and apparatus. The mechanism of action of iodine with breakdown products of fl-lactam antibiotics is uncertain and care has to be taken with tile interpretation of results for each new substrate studied. Fortunately, few intact substrates react with iodine, although examples are known. Grove and RandalP ° reported that intact allylmercaptomethylpenicillin reacts with iodine ; 3-dimethyldithioearbamoyloxymethyl-7fl-benzylthioacetan~idoceph-3-em-4-carboxylic acid also takes up iodine before/~-lactam hydrolysis, thus giving a high blank value in the assay, p-Hydroxybenzyl penicillin also gives a high blank value21 The enzyme preparation should be routinely checked for iodine uptake. The enzyme reaction can be stopped after 5 min instead of the usual 30 min if required. Care should be taken that the enzyme reaction is stopped by addition of the iodine reagent and Zyk '3 has followed the suggestion of Cshnyi '~'-' D. C. Grove and W. A. Randall, "Assay Methods of Antibiotics," p. 16. Medical Encyclopaedia Inc., Chicago. (1955). ~1p. H. A. Sneath and J. F. Collins, Biochem. J. 79, 512 (1961). 32V. Cs~nyi, Acta Physiol. Acad. Sci. Hung. 18, 261 (1961).
84
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
to include sodium tungstate in the iodine reagent to ensure inactivation of the enzyme. Perret's method is particularly useful for comparison of the substrate specificities of fl-lactamases, when a wide range of substrates has to be used. Its main disadvantage is lack of sensitivity; the usual substrate concentration is 5 mM. The microiodometric methods are extremely sensitive. Sykes and Nordstrom used a substrate concentration of 0.067 mM in the assay mixture and enzyme activities of below 0.001 ~mole of substrate per minute. At higher enzyme activities, the steady state is not reached before the starch-iodine complex is completely decolorized. Enzyme activities as low as 0.0004 unit per milligram dry weight of bacteria can be readily measured. 1~ The only equipment required is a relatively unsophisticated spectrophotometer. Microiodometric methods are particularly useful in studies of enzyme kinetics or in inhibition studies. A major disadvantage of the micro methods is that they cannot be used with cephalosporin substrafes. The effect of the starch-iodine complex on each new fl-lactamase must also be carefully examined.
B. Acidimetric Assay Although the original methods were "manual" assays, pH star equipment is required to obtain good precision and reproducibility in the direct assay. A substrate concentration of 5 mM is commonly used. Acidimetric assays have been applied principally to penicillin substrates rather than to the cephalosporins, where careful calibration is required owing to the possible production of more than one acidic group or even a basic group after hydrolysis of the fl-lactam ring (see introduction). Farrar and Krause have shown that the number of equivalents of acid released per mole of antibiotic hydrolyzed is 2 for cephalothin, 1.6 for cephaloglycin, and 0.6-1.0 for cephalexin. ~3 The indicator method is a useful alternative, especially for laboratories which have a spectrophotometer but not pH star equipment. This method is of intermediate sensitivity; a substrate concentration as low as 0.2 mM can be used. The manometric method is mainly of historical interest and is now seldom used.
C. Spectrophotometrie (UV) Assay This assay is easy to operate, is rapid, and gives valuable information on the enzyme-substrate interaction in the form of a reaction profile. 3a W. E. Farrer and J. M. Krause,
Inject. Immunity 2, 610 (1970).
[5]
~-LACTAMASE ASSAYS
85
Up to four simultaneous enzyme reactions can be followed in a recording spectrophotometer, and this has been particularly useful for comparison of the fl-lactamase resistance of new cephalosporins and for inhibition studies. The products of fl-lactam hydrolysis may absorb in the UV range so that at any one time the absorption of a mixture of substrate and product(s) is being measured. The effect of UV absorption by the products is minimized by comparing only the initial rates of hydrolysis over the first minute. Reliable quantitative results are obtainable when a callbration curve is constructed relating fall in UV absorption to hydrolysis of the cephalosporin. The assay has been used for substrate concentrations from 0.05 to 1 mM. The main disadvantage of this technique is that it cannot be applied to penicillin substrates because they do not have the conjugated system associated with the fl-lactam ring in cephalosporins which gives rise to the absorption monitored in this method. 26 However, Jansson has reported 34 that the rate of hydrolysis of ampicillin by a B. cereus B-lactamase can be related to a fall in absorption at 244 nm. Spectrophotometric measurement has been used indirectly for assay of the products of B-lactamase hydrolysis of penicillins in the microiodometric and indicator methods. All fl-lactamases tested to date, including both penicillinase and cephalosporinase types, can be assayed by the chromogenic cephalosporin method and this is particularly useful for comparing several enzymesJ
D. Hydroxylamine Assay This assay has been found to be more difficult to operate than the other assay procedures described and it gave higher relative activities against penicillins and some cephalosporins in a study in which five assay methods were compared. 29 Mixing of the reagents is critical and unwanted precipitation can occur after addition of the ferric ammonium sulfate reagent or if the concentration of the phosphate buffer is increased. The chromogen is unstable, and gas bubbles formed on mixing the alcoholic and aqueous solutions can interfere with absorption measurements. When purified enzyme preparations are used, the assay measures residual intact B-lactam ring in the substrate, but, in less pure reaction mixtures, hydroxylamine can react with compounds containing a carbonyl group, or with esters, anhydrides, and amides25 The assay is relatively insensitive and has no apparent advantage. 34j. A. T. Jansson, Biochim. Biophys. Acta 99, 171 (1965). ~sj. :H. Ford, Ind. Eng. Chem. Anal. Ed. 19, 1004 (1947).
86
METHODS FOR THE STUDY OF ANTIBIOTICS
[6] I m m u n o l o g i c a l
Techniques
[6]
for
Studying/~-Lactamases
By M. H. :RICHMOND I. Introduction . . . . . . . . . . . . . . . . . . II. Preparation of Antisera . . . . . . . . . . . . . . . A. Using Purified Enzyme Preparations . . . . . . . . . . B. Using Crude Enzyme Preparations . . . . . . . . . . . C. Isolation of ~-Lactamase-Less Mutants . . . . . . . . . D. Inoculation Program for Crude Preparations . . . . . . . . III. Examination of ~-Lactamases with Specific Antisera . . . . . . . A. Neutralization Analysis . . . . . . . . . . . . . . B. Precipitation Analysis . . . . . . . . . . . . . . .
86 86 86 87 89 89 90 90 98
I. Introduction I n principle, fl-lactamases are no different from other enzymes in respect to their action with antisera, and m a n y of the techniques to be described below are capable of much wider application t h a n has occurred in the past. The reason for describing t h e m here, however, is t h a t they have a particular relevance to fl-lactamases since m a n y of the minor variants of this enzyme t h a t are found in naturally occurring strains have been characterized b y these immunological methods.
I I . P r e p a r a t i o n of Antisera
A. Using Purified Enzyme Preparations For the preparation of specific antisera, and those for reaction with enzymes are no exception, it is preferable to have the antigen as pure as possible. I n this w a y the sera are likely to have their highest titers and to be as free as possible of nonspecific side reactions. In practice, preparation of a pure antigen m a y be more exacting than normal protein purification since small amounts of impurity m a y be disproportionately immunogenic. On the other hand, these considerations do not often invalidate the use of sera for analytical purposes with fl-lactamases. Indeed all the studies on staphylococcal and Escherichia coli B-lactamases described in this section were carried out with enzymes purified by the methods described elsewhere in this volume. 1,2 1This volume [53c]. This volume [53d].
[6]
IMMUNOLOGICAL TECHNIQUES FOR STUDYING ~-LACTAMASES
27
A number of different animal species and dosage regimes have been used to raise fl-lactamase sera. All the most effective, however, use inoculation of rabbits with purified enzyme preparations treated in various ways2 ,4 For antistaphylococcal penicillinase,4 sandy lop rabbits are each injected in alternate thighs at 3-day intervals with a total of four 0.2-ml portions of a solution of purified staphylococcal penicillinase made by suspending about 5 mg of purified enzyme per milliliter of Freund's adjuvant2 After a further 2 weeks, each rabbit is injected intravenously through the ear vein with three successive 0.2-ml samples of an alum-precipitated preparation of purified enzyme (approximately 5 mg/ml) at 2-day intervals. Samples of blood are withdrawn from the rabbit's ear vein 6 days after the end of the second phase of treatment, the serum is separated by conventional techniques, and the antibody titer is determined. In practice, once a serum of suitable titer has been obtained, it is better to kill and exsanguinate the rabbit since this procedure provides maximum quantities of a serum of uniform properties. In a typical example with staphylococcal penicillinase, about 100 ml of a serum, with titer about 104 units per milliliter of undiluted serum, was obtained from a single sandy lop rabbit. After separation of the serum, the material is best stored by dividing into ampules in about 5-ml quantities and freezedrying. Ampules may then be reconstituted for use as required by adding 5 ml of distilled water. If diluted serum is needed, the reconstituted serum should be diluted for use in physiological saline.
B. Using Crude Enzyme Preparations The above method is the one that should be followed when reasonable supplies of purified enzyme are available. It is possible, however, to prepare adequate specific sera by inoculating relatively crude enzyme preparations in Freund's adjuvant, and then absorbing out the unwanted antibodies before use. For this procedure to be successful, it is vital to have a mutant of the producer strain that makes no fl-lactamase. Before going on to describe the actual techniques for preparation of fl-lactamase from crude ])reparations, the purification procedure for crude enzyme and the methods of isolating fl-lactamase-less mutants will be described. Preparations of fl-lactamase from most gram-negative bacteria may be prepared as follows. The method is based on the one described elsewhere in this volume for the purification of the fl-lactamase from Escherichia coli (Rr+~M).2 A culture of the producer strain growing exponena M. R. Pollock, J. Gen. Microbiol. 14, 90 (1956). 4 M. H. Richmond, Biochem. J. 88, 452 (1963). 5 Difco Bacto-adjuvant, Complete (Freund), 0638-60-7.
88
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
tially in 1% CY medium 1 is centrifuged and the bacteria are collected. These organisms are resuspended in 0.1 M Na2HPO4/KH2P04 buffer, pH 7.0, to a density of about 30 mg dry weight of bacteria per milliliter and disrupted in an ultrasonic disintegrator. ~ The broken bacteria are stored at 2 ° until disruption of the whole batch is complete; they are then centrifuged, first at 5000 g for 15 rain at 4 ° (to remove large cell debris), then at 40,000 g for 4 hr at 4 ° (to remove smaller fragments), and last at 105,000 g for 2 hr at 4 ° to sediment ribosomes and minute pieces of comminuted membrane. After centrifugation the supernatant containing the enzyme is dialyzed against 100 volumes of 0.1 M Na~ HPO4/KH2P04 buffer, pH 7.0, for 24 hr at 2 °. During this process a precipitate often forms, and this is removed by centrifugation at 5000 g for 15 min at 4 °. After dialysis and any necessary centrifugation the enzyme is run through a Sephadex column. Adequate preparations are obtained if a Sephadex G-75 column equilibrated against 0.1 M Na~HPO~/KH2P04 buffer, pH 7.0, is used, but cleaner enzyme preparations may be obtained by using either DEAE-Sephadex or CM-Sephadex, as appropriate. The choice of Sephadex depends on the nature of the enzyme, and to decide this it is useful to know something of the ionic properties of the protein. In general all fl-lactamases from gram-negative species with a predominant activity against cephalosporins are positively charged at neutral pH values 6 and therefore should be purified on CM-Sephadex. The remaining fl-lactamases are likely to be more acidic 6 and in this case DEAE-Sephadex may be more appropriate./~lthough a general tendency, however, this correlation between charge andsubstrate profile should not be taken as absolute, and it is always necessary to carry out some preliminary test absorptior~ experiments with the enzyme and various substituted Sephadexes when an enzyme is being examined for the first time. As a guide it is always helpful to know the electrophoretic properties of the enzyme before doing the experiments. When G-75 Sephadex is used, the column (150 X 1.5 cm) is equilibrated against 0.1 M Na2 HPO4/KH2P04 buffer, pH 7.0, and the material is eluted with similar buffer. After elution from the Sephadex column the enzyme preparation is dialyzed against distilled water to remove salts and then stored at 2 ° until required for use. In general it seems better not to freeze dry enzyme preparations destined for raising antisera since the freezing and thawing process is likely to denature some of the enzyme molecules, and this in turn is likely to broaden the specificity of the resulting antiserum. M. H. Richmond and R. B. Sykes, Recent Adv. Microb. Physiol. 9, 36 (1973).
[6]
IMMUNOLOGICALTECHNIQUES FOR STUDYING ~-LACTAMASES
89
C. Isolation of/~-Lactamase-Less Mutants The fl-lactamase producing culture is grown exponentially in any convenient growth medium and then treated with N-methyl-N-nitro-Nnitrosoguanidine (NMG) as described elsewhere in this volume. 1 After treatment and the necessary period of growth following removal of the mutagen, the bacteria are plated as single colony-forming units on 1% CY agar l and incubated overnight at 37 °. In the morning the colonies are flooded on the plate with a solution (5 mg/ml) of the chromogenic cephalosporin 87/312. 7 In this example, the great majority of the colonies will turn red (indicating production of fl-lactamase), but a few (probably not more than 1/50,000) will produce no color. It is therefore necessary to do this experiment on a large scale. About 100 plates with 1000 colonies per plate are commonly employed; but it may be necessary to work on an even larger scale. Any colorless colonies are picked onto fresh agar and purified by restreaking. The isolates are next checked for their production of fl-lactamase, and any producing no detectable enzyme by quantitative methods are retained for use. The process of isolating mutants in this way is tedious, since methods for selection of fl-lactamaseless mutants are not available, but fl-lactamase-less mutants of this type must be available for specific sera to be obtained when crude enzyme preparations are used as immunogens.
D. Inoculation Program for Crude Preparations As with the program for purified enzyme, rabbits are the most convenient animal for raising sera in the conventional laboratory. The larger the rabbit the better, since the final yield of serum will depend to some extent on the blood volume of the animal; and absorbed sera are likely to have lower titers than those prepared against purified antigens. A typical program for raising an antiserum using a crude enzyme preparation is as follows. Samples of crude enzyme, purified as described above, are injected after mixing with equal volumes of Freund's adjuvant into alternate thighs of a sandy lop rabbit. In all, a sequence of six injections at 2-day intervals is used. Two weeks after the last injection, the animal is injected through the ear vein with 0.2 ml of alum-precipitated crude enzyme. Four injections of this material are made at 2-day intervals. About 10-14 days after the end of this sequence of injections, the animal is test bled, and the antiserum titer is determined against the appropriate enzyme (see below). When two test bleeds 2 days apart show the same titer, the animal is killed and exsanguinated, and serum is prepared from TC. H. O'Callaghan, A. Morris, S. M. Kirby, and A. H. Shingler, Antimicrob. Ag. Chemother. 1, 283 (1972).
90
METHODS FOR T H E STUDY OF ANTIBIOTICS
[6]
the whole blood by conventional techniques. Titers of about 5000 units of neutralizing ability are not uncommon by this technique, but it must be stressed that not all anti-fl-lactamase sera are of the neutralizing type (see below). Once an active serum has been obtained by this method it is then treated with crude protein prepared from the fl-lactamase-less mutant strain. Ideally the serum should be treated with material made in an identical manner to the crude enzyme, protein being collected from the Sephadex G-75 column at the point at which the relevant fl-lactamase is known to elute. However, it is equally effective to use some of the total cell protein obtained after ultrasonic disruption of the fl-lactamaseless mutant. However, if material that is the product of ultrasonic disintegration without centrifugation is used, losses in the titer of the serum are likely to be high, but if the material at the stage at which it would normally be loaded onto the Sephadex column is employed, absorption is more effective and losses of specific anti-fl-lactamase activity is minimized. The adsorption technique involves adding the absorbing protein to the serum in a series of 0.2-ml quantities until precipitation ceases. Ideally this should lead to no loss in titer of the antiserum, but in practice some loss seems difficult to avoid, and it is usually necessary to compromise between having an absolutely specific serum and one of such low titer that it is virtually useless. In a typical absorption experiment, 0.2-ml quantities of absorbing protein were added at intervals to 5 ml of undiluted serum. The serum was then incubated at 30 ° for 2 hr before the precipitate was centrifuged off at 5000 g for 15 min. In one experiment about 0.75 ml of absorbing protein had to be added to the 5-ml batch of serum before precipitation ceased. In order to sterilize the serum during this procedure, 0.1% (w/v) sodium thiomersalate is added to the serum. Once absorption is complete, the serum is transferred to ampules and freeze-dried in the normal way. III. Examination of fl-Lactamases with Specific Antisera Once antisera to a B-lactamase have been obtained, they can be used to study the enzymes in a number of ways. These include precipitation analyses both in solution and by gel diffusion. But perhaps the most effective is by neutralization of enzyme activity since this can be studied with as little as 0.1 ~g of enzyme in impure preparations.
A. Neutralization Analysis When antibodies interact with enzymes, the activity of the preparation may be altered. However, interaction with an antibody molecule
[~]
IMMUNOLOGICAL TECHNIQUES FOR STUDYING ~-LACTAMASES
91
need not affect activity, and it is impossible to predict the manner in which an antiserum preparation raised against either a crude or a purified preparation of fl-lactamase will act. Indeed, antisera obtained with samples of the same enzyme preparation in different rabbits may v a r y greatly in their effect, even when the immunization program is identical and carried out in parallel. It follows, therefore, that when antisera are being raised it is important to obtain as large a quantity as possible so that a large series of tests may be made with a serum of uniform properties. Provided an antiserum affects enzyme activity, this interaction m a y be used to provide much information about the nature of a fl-lactamase. Most sera reduce the activity of B-lactamases, 3,~ and this neutralizing effect will be discussed first. However, stimulatory sera, ~ or even sera that stimulate and then inhibit at higher concentration are known, 9 but their use will be discussed later. The most common way to use a neutralizing serum is to study the enzyme/antiserum interaction as a titration of a constant amount of antigen. Increasing amounts of antiserum are added to a series of vessels each containing a fixed amount of antigen. After a period for the antigen/antibody reaction to occur (usually about 10 rain at room temperature) the residual enzyme activity is assayed by a convenient method (usually the iodometric method of Ferret TM as modified by N o v i c k ' ) . With E. coli fl-lactamase, where a serum with titer about 104 units/ml is available, the usual experimental approach is to put about 50-75 units of enzyme in each of a series of conical flasks and immediately add increasing amounts of serum, leaving the first without addition to act as a measure of the untreated enzyme activity. The reaction is allowed to continue for 10 rain. Then 10 ml of a prewarmed solution of benzyl penicillin (1%, w/v, in 0.1 M Na~ H P O J K H 2 P O 4 buffer, pH 5.9) is added. After 6 rain further incubation, 10 ml of standard iodine solution in 2 M sodium acetate buffer, pH 4.2, is added and the excess iodine back-titrated against sodium thiosulfate." Figure 1 shows a typical neutralization curve obtained in this way2 There is a linear decrease in residual enzyme activity with increasing antiserum concentration up to a point (the equivalence point: E--see Fig. 1) at which no further inactivation occurs however much serum is added. The activity remaining after the equivalent amount of serum has been added is known as the "residual activity" and probably represents the activity of the enzyme/antibody complex. The slope of the neutralization curve G. W. Jack and M. H. Richmond, J. Gen. Microbiol. 61, 43 (1970). 9 M. R. Pollock, Immunology 7, 707 (1964). ~oC. J. Perret, Nature (London) 174, 1012 (1954). See also this volume [5]. ~ R. P. Novick, Biochem. J. 83, 229 (1962).
92
METHODS FOR THE STUDY OF ANTIBIOTICS
[6]
F
100~-,,~
'~: .I- \ . 6o-
\o
\o
--
'~EO--~O
20 I 0.02
I 0,04
I 0,06
I 0.08
Antiserum
I 0.1
I 0.12.
: ml
FIG. 1. Neutralization of type I I I a ~-lactamase by anti-type I I I a fl-lactamase serum [G. W. Jack and M. I-I. Richmond, J. Gen. Microbiol. 61, 43 (1970)]. E, equivalence point.
gives the titer of the serum; and if the turnover number of the fl-lactamase is known, this titer m a y be expressed as the number of enzyme molecules neutralized per milliliter of antiserum. In practice 10 min is normally allowed for the enzyme/antibody complex to forIn at room temperature, but a study of the time course of enzyme/antiserum interaction shows that the reaction occurs very rapidly (Fig. 2). Neutralization with more than the equivalent amount of serum is more than 90% complete in 2 min at room temperature. Figure 1 illustrates a typical neutralization curve obtained with an anti-fl-lactamase serum. However, all sera do not neutralize. Figure 3
~
c t00 ]uu .2
6{ E
E E
•
O
/
°~° 2C I 2
I 4
I 6
I 8
I 10
minutes
FIG. 2. Time course of reaction between type I l i a /~-lactamase and a specific antiserum.
[6]
IMMUNOLOGICAL TECHNIQUES FOR STUDYING /~-LACTAMASES
93
241 20(
//
ID
/* /* 40,;I 0.05
i 0.1 Antiserum
| 0.15
I 0-2
: ml
FIG. 3. Stimulation of staphylococcal ~-lactamase activity by a specific antiserum [M. H. Richmond, Biochem. J. 88, 452 (1963)]. E, equivalence point.
shows the effect of addition of increasing amounts of anti-staphylococcal serum to staphylococcal fl-lactamase type A. In this case this particular batch of serum stimulates the activity of the enzyme to a maximum about 4.5 times higher than that of the enzyme without serum. ~ Once again the response is linear with increasing amounts of antiserum, and once more there is an equivalence point (E) above which further addition of serum produces no change in activity. As with neutralization effects, stimulation by antiserum can be used to characterize an enzyme. The slope of the curve (Fig. 3) gives a measure of the titer of the serum, and the extent of stimulation is constant for a given type of fl-lactamase. At present it is uncertain how combination with an antiserum increases the activity of an enzyme. Intuitively one feels that the process must hold the enzyme in a more favorable conformation, but convincing experimental evidence for this sort of mechanism is lacking. Nevertheless one thing seems certain: The binding of stimulatory sera cannot block the active center of the enzyme. Certain sera show even more complex effects with fl-lactamases than the examples shown in Figs. 1 and 3. Pollock has studied the interaction of specific serum with purified fl-lactamase from Bacillus licheni]ormis2 In this example (Fig. 4) addition of antiserum causes stimulation and then inhibition. Such a curve has two equivalence points (El and E.~) but is difficult to use for analytical purposes, although the shape of the curve has great value as a qualitative means of identifying an enzyme type. This complex curve is certainly due to the interaction of the enzyme with an antiserum containing both inhibitory and stimutatory compo-
94
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
3001 /\,E1 .. 200 >
0
i ,°0,
\,
o
E2
I
I
O.2
o.1 ml
of
I
O.3
antiserum
FIG. 4. Interaction of fl-lactamase from Bacillus licheni]ormis 6346 with specific antiserum [M. R. Pollock, Immunology 7, 707 (1964)]. E,, stimulatory equivalence point; E~, neutralizing equivalence point.
nents. Partial absorption of the serum with enzyme removes the stimulatory component and gives a residual serum which is purely neutralizing2
I. Characterization o] Variants in Naturally Occurring Strains Since the effect of antisera on fl-lactamase activity can be studied with such small amounts of enzyme and since contaminating protein does not interfere significantly with the interaction, tests in solution are an ideal way of looking for molecular variants of fl-lactamase whether of natural or laboratory (i.e., mutational) origin. An example of the detection of natural variants comes from studies that have been made on staphylococcal fl-lactamase. 12,13 The reaction of anti-staphylococcal fllactamase serum with the fl-lactamase found in most clinical isolates of Staphylococcus aureus gives a stimulatory curve already discussed (Fig. 1). However, analysis of a large number of clinical isolates shows that four types of response are to be found (Fig. 5). The majority show the standard stimulation by a factor of about 4.5-fold---known as A-type response--while a substantial minority show no stimulation at all. Indeed in this group (C type) there is no evidence that the serum (which was raised against purified A-type enzyme) interacts with the fl-lactamase at all until a precipitation analysis is carried out. Under these conditions an enzyme/antiserum precipitate is obtained, but this has the expected activity of the amount of enzyme combined in the precipitate. Two other ~2M. It. Richmond, Biochem. J. 94, 584 (1968). ~3V. T. Rosdahl, J. Gen. Microbiol. 77, 229 (1973).
[{)]
IMMUNOLOGICAL TECHNIQUES FOR STUDYING ~-LACTAMASES
95
20o
"~160 •;
120
o / / -
• ~
• ~
• ~ OA
o
o
o
I
I
/+
i40, ~-~ I
0.1 0.2 0.3
I
Od
Co
I
0.5
ml of antiserum
Fro. 5. Response of the four variants of staphylococcal f~-lactamase after interaction with a serum raised against type A [M. H. Richmond, Biochem. J. 94, 584 (1968)]. A, B, C, D: fl-lactamase types based on their response to serum.
types of response are also found. In the first (type B), the antiserum stimulates, but the titer is about one-fifth of that expressed with A-type fl-lactamase. In fact, further analysis shows that this response is due to expected antiserum binding to a fl-laetamase molecule with about onefifth the turnover number of A-type fl-lactamase. The fourth type of response (type D) gives a lower maximum degree of stimulation (about 1.5-fold at most) with a lower apparent titer. 13 In this case, the turnover number of the fl-lactamase variant is the same as that of A-type enzyme, and therefore the serum interacts with the enzyme with lower affinity. Subsequent experiments have shown that types A, B, and C fl-lactamase from S. a u r e u s (and probably type D as well) are minor variants of a common structure probably with one or a few amino differences in their primary sequence. Similar experiments have been done with fl-lactamases from other bacterial species, and in general these experiments have led to the view that these enzymes tend to be species specific, but that within a given species a relatively large number of minor variants are to be found. 6''4 The cross reaction found between the minor variants within a given species, but the absence of cross reaction between the fl-lactamases from different species suggests that perhaps interaction between enzyme and antiserum is particularly sensitive to changes in the primary sequence of the enzyme protein, and that a change in only one or two residues can be tolerated before the binding of the serum becomes suSstantially impaired. Nor is such a view incompatible with the normal high specificity of antibody molecules. ~ N. Cirri and M. R. Pollock, Advan. Enzymol. 28, 237 (1966).
96
METHODS FOR THE STUDY OF ANTIBIOTICS
[5]
2. The Use o] Antisera to Study Laboratory-Induced Mutations The main objective here, since one starts with the information that any changes in reaction are due to changes in amino acid sequence, is to try to discover whether the mutation has affected only the turnover number of the enzyme, or whether there has also been some alteration in affinity for antibodies. A typical example is the examination of the pen P2 mutant of staphylococcal fl-lactamase. 1~ Bacterial cultures synthesizing this mutant enzyme express about 5% of the normal wild-type level of activity. Comparison of the serum interaction of this mutant enzyme with the wild type shows that the total stimulation obtained is the same (i.e., about 4.5-fold), but that the antiserum titer with the mutant enzyme is about ½oth of that found with the wild type. These results suggest strongly, therefore, that the mutational change in P2 has resulted in a protein with about 5% the turnover number of the wild-type enzyme but with no alteration in the affinity of interaction between serum and enzyme. This result is only an indication and must be confirmed by purification of the enzyme; and indeed in the case of mutant P2 the turnover of the purified mutant enzyme is found to be about 5% of the unmutated parental fl-lactamase. As yet this approach to the study of fl-lactamases is largely unexploited since no systematic analysis of the effect of primary sequence changes on enzyme activity and reaction with antiserum has yet been made. Once this approach is developed, however, the use of sera in an appropriate way will give much useful preliminary information about the properties of mutant enzymes, even if the final situation can be confirmed only by the painstaking isolation and purification of each mutant.
3. Use o] Antisera to Analyze One Component in a fl-Lactamase Mixture The one group of strains where the general rule that fl-lactamases tend to be species specific breaks down in that group of gram-negative bacteria susceptible to R factor infection. In this group it is possible to find bacterial isolates that express two distinct types of fl-lactamase,8 and it can therefore be valuable to have methods for determining the relative amounts of the two enzymes expressed by such strains. One typical example is the use of anti-type I I I a fl-lactamase serum to analyze the enzyme produced by a strain of E. coli which can be shown by electrophoresis of a crude enzyme preparation to make two enzymes, one a cephalosporinase of type Ib and the other an R factor-mediated enzyme (type IIIa). The type Ib enzyme does not react with the anti-type IIIa serum ~ See footnote 4, and also unpublished experiments b y M. It. Richmond.
[6]
IMMUNOLOGICAL TECHNIQUES FOR STUDYING ~-LACTAMASES
97
whereas the t y p e I I I a e n z y m e - - b e i n g the enzyme against which the serum was raised in the first p l a c e - - g i v e s a neutralization reaction with a residual activity of about 35% (see Fig. 1). About 100 units of the enzyme mixture was titrated with anti-type I I I a serum as a constant antigen titration. The serum partially neutralized the activity of the mixture, and the residual activity was about 67% of the initial activity (Fig. 6). Since (a) only the type I I I a enzyme reacts with the antiserum, and (b) the presence of type I b enzyme does not interfere with the reaction, it is possible to calculate the relative amounts of the two enzymes in the original mixture using the fact that the residual activity obtained with type I I I a enzyme and this antiserum is 35% of the initial activity. I n this particular example, in fact, the calculation shows that the original mixture was made up of about 50 units of type Ib enzyme and 50 units of t y p e I I I a . This approach to analyzing the composition of mixtures of fl-lactamases is very useful, but it is subject to two limitations. First, the antiserum must be specific to one of the two component enzymes; and second, the amount of the two components must be approximately equal. Clearly the total activity of small amounts of type I I I a enzyme in the presence of large amounts of a nonreactive fl-lactamase will show relatively little decrease of activity in the presence of antiserum, and any subsequent calculations must be subject to large errors. 120
\ 8( .>
E
'esidual activity
40
0.1 ml
0.2
I
of a n t i s e r u m
FIG. 6. Neutralization of a mixture of type Ib and type IIIa fl-lactamase with anti-type IIIa serum. E, equivalence point. Type Ib enzyme does not react with anti-type IIIa serum. Moreover this serum neutralizes type IIIa fl-lactamase to 35% residual activity. The actual neutralization observed in Fig. 6 is 35 units out of a total of 103 present; and this corresponds to the neutralization of 54 units of type IIIa fl-lactamase. The type Ib component therefore amounts to 49 units, by difference.
98
METHODS FOR THE STUDY OF ANTIBIOTICS
[6]
B. Precipitation Analysis ~6
Even if antisera to fl-lactamases do not affect enzyme activity, they will precipitate enzyme protein provided the concentration of the reactants is high enough. However, since most enzymes have such high specific enzyme activities it is unusual to have sufficient enzyme protein in a test system to produce an effective precipitate. It follows, therefore, that quantitative precipitation analysis of fl-lactamases is usually confined to rather special circumstances; and it is normally necessary to label the enzyme protein with radioactivity in order to follow the process accurately. Quantitative precipitation analysis has been used effectively to discover whether fl-lactamase I induction in Bacillus cereus was due to the biosynthesis of enzyme de novo, or whether conversion of an inactive precursor was involved.16 In these experiments a culture of Bacillus cereus growing exponentially was induced by adding 6 gg of benzyl penicillin per milliliter, and at the same time a source of 14C-labeled amino acids was added to the culture. Incubation was continued, and samples of the culture were withdrawn at intervals for precipitation analysis. The amounts of fl-lactamase I present in the growth medium in this experiment (the enzyme is an extracellular protein in B. cereus) were far too low for precipitation to occur without the addition of carrier fllactamase. Accordingly, each sample from the culture supernatant was treated with enough purified nonradioactive B. cereus fl-lactamase I to bring the total activity of the sample to 9000 units, followed by a 50% excess of specific antiserum. The precise experimental sequence was carried out as follows.16 Samples of whole bacterial culture (4.0 or 5.0 ml) were pipetted directly into a centrifuge tube containing a few crystals of oxine (8-hydroxyquinoline) and sufficient solid sodium chloride to bring the final concentration to molar strength. After shaking to dissolve the solids, the tubes were cooled rapidly and the bacteria centrifuged down. The penicillinase activity in the supernatant was assayed and enough carried fl-lactamase added to 3.0 ml of the total sample to bring the total penicillinase activity to 9000 units. This amount of enzyme is equivalent to about 3 mg of pure enzyme. The carrier supplemented samples were then dialyzed overnight at 2 ° against 1 mM KH2P04/K~HPO~ buffer, pH 7.0. After overnight dialysis, the carrier-supplemented enzyme preparations were transferred to centrifuge tubes and 0.15 ml of a suspension of "fine-mesh" glass powder (100 mg of glass per milliliter) added to each. The preparation was left at room temperature for 3 min, then the glass powder was centrifuged off and washed once with 5 ml of dis~6M. R. Pollock and M. Kramer, Biochem. J. 70, 665 (1958).
[6]
IMMUNOLOGICAL TECHNIQUES FOR STUDYING ~-LACTAMASES
99
tilled water. The enzyme was then eluted from the glass with two successive washes (2.5 ml and 1.0 ml, respectively) of alkaline 1 M NaC1 at pH 8.5. These two washings were combined and filtered through an Oxoid membrane filter to sterilize them and made up to 5.0 ml with distilled water; their fl-lactamase content was assayed. Enough carrier fl-lactamase to return the total activity of the preparation to 9000 units was then added to 4.0 ml of the eluate. Anti-exo-penicillinase serum was then added to exceed the equivalence point by 50%, and the mixture was incubated overnight at 35% The next morning the precipitate was centrifuged down after cooling to 2 °, resuspended in 2 ml of distilled water, and centrifuged down once more. This precipitate was then prepared for radioactivity measurements in the conventional way. If this procedure is followed meticulously, at least 90% of the fl-lactamase activity is precipitated as an antigen/antibody complex; and the application of these precipitation techniques allowed Pollock and Kramer TM to show conclusively that B-lactamase induction was not due to the activation of an inactive precursor, but was the result of de novo fl-lactamase synthesis. Gel Precipitation and Other Qualitative Procedures The main objective in this article has been to describe the preparation of specific anti-fl-lactamase sera and to show how they may be used for the quantitative study of penicillinases and cephalosporinases. Once the sera have been obtained, however, they may be used for the full range of gel precipitation and immune-electrophoresis procedures available for other enzymes and proteins. These have frequently been reviewed, and the reader is referred to them. 17 One special point must be borne in mind when applying these techniques to fl-lactamases. Several of the enzymes, but notably all the B-lactamase variations synthesized by plasmid carrying strains of Staphylococcus aureus are relatively basic proteins and adsorb strongly to acidic polymers such as agar. Gel diffusion experiments involving these enzymes should therefore be carried out in the presence of 1 M NaC1. Without this no diffusion of enzyme occurs and precipitation, if it occurs at all, is either on the edge of the well or even in the enzyme well itself. Unfortunately, it is impossible to add such high salt concentrates to electrophoretic systems, and so far immune electrophoresis of staphylococcal fl-lactamase has been unsuccessful. These enzymes do not adsorb to starch gels, but the precipitation bands are difficult to locate in this medium unless they are visualized by staining the gels for enzyme activity. The best method of doing this is to treat starch gels after electro1, C. A. Williams and M. W. Chase, Eds., "Methods in Immunology and Immunochemistry," Vol. 3, pp. 118 and 234. Academic Press, New York, 1971.
100
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
phoresis with the reagent used to detect fl-lactamase by colonies growing on starch agar. 1 This method is less effective than might be expected, however, since the reaction spreads rapidly and the precipitation bands appear clearly only for a short time.
[7] Paper Chromatography of Antibiotics By VLADIMIRBETINA I. II. III. IV.
V.
VI.
VII.
VIII.
IX.
X. XI. XII.
Introduction . . . . . . . . . . . . . . . . . . General T e c h n i q u e s . . . . . . . . . . . . . . . . Bioautography . . . . . . . . . . . . . . . . . Paper C h r o m a t o g r a p h y of fl-Lactam Antibiotics . . . . . . . A. Penicillins a n d 6-Aminopenicillanic Acid . . . . . . . . . B. Cephalosporin C F a m i l y , 7-ACA, a n d C e p h a m y c i n s . . . . . C a r b o h y d r a t e Antibiotics . . . . . . . . . . . . . . A. Aminoglycosidic Antibiotics . . . . . . . . . . . . B. Streptothricin Group . . . . . . . . . . . . . . C. Oligosaccharides w i t h C h r o m o p h o r e . . . . . . . . . . D. E v e r n i n o m i c i n a n d L i n c o m y c i n Group . . . . . . . . . E. Other Sugar D e r i v a t i v e s . . . . . . . . . . . . . Macrocyclic Lactone Antibiotics . . . . . . . . . . . . A. N o n p o l y e n e Glycosidic Macrolides . . . . . . . . . . B. Polyene Macrolides . . . . . . . . . . . . . . . C. Other Macrocyclic Antibiotics . . . . . . . . . . . . Quinone Antibiotics . . . . . . . . . . . . . . . A. Tetracyclines . . . . . . . . . . . . . . . B. A n t h r a c y c l i n e s a n d A n t h r a c y c l i n o n e s . . . . . . . . . Amino Acid a n d Peptide Antibiotics . . . . . . . . . . A. Diketopiperazine D e r i v a t i v e s . . . . . . . . . . . B. H o m o p e p t i d e s . . . . . . . . . . . . . . . C. H e t e r o m e r Peptides . . . . . . . . . . . . . . D. Sideromycins . . . . . . . . . . . . . . . . E. Bleomycin Group . . . . . . . . . . . . . . F. A c t i n o m y c i n s . . . . . . . . . . . . . . . G. E c h i n o m y c i n - T y p e Antibiotics . . . . . . . . . . N i t r o g e n - C o n t a i n i n g Heterocyclic Antibiotics . . . . . . . A. M i t o m y c i n Group . . . . . . . . . . . . . . B. P y r i m i d i n e Nucleosides . . . . . . . . . . . . . C. P u r i n e Nucleosides . . . . . . . . . . . . . . O x y g e n - C o n t a i n i n g Heterocyclic Antibiotics . . . . . . . . Alicyclic Antibiotics . . . . . . . . . . . . . . . A r o m a t i c Antibiotics . . . . . . . . . . . . . . A. Chloramphenicol a n d I t s Derivatives . . . . . . . . . B. Novobiocin . . . . . . . . . . . . . . . .
101 102 105 110 110 116 119 119 123 126 128 128 129 129 134 136 137 137 141 144 144 145 146 146 148 150 150 152 152 152 153 154 156 159 159 160
100
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
phoresis with the reagent used to detect fl-lactamase by colonies growing on starch agar. 1 This method is less effective than might be expected, however, since the reaction spreads rapidly and the precipitation bands appear clearly only for a short time.
[7] Paper Chromatography of Antibiotics By VLADIMIRBETINA I. II. III. IV.
V.
VI.
VII.
VIII.
IX.
X. XI. XII.
Introduction . . . . . . . . . . . . . . . . . . General T e c h n i q u e s . . . . . . . . . . . . . . . . Bioautography . . . . . . . . . . . . . . . . . Paper C h r o m a t o g r a p h y of fl-Lactam Antibiotics . . . . . . . A. Penicillins a n d 6-Aminopenicillanic Acid . . . . . . . . . B. Cephalosporin C F a m i l y , 7-ACA, a n d C e p h a m y c i n s . . . . . C a r b o h y d r a t e Antibiotics . . . . . . . . . . . . . . A. Aminoglycosidic Antibiotics . . . . . . . . . . . . B. Streptothricin Group . . . . . . . . . . . . . . C. Oligosaccharides w i t h C h r o m o p h o r e . . . . . . . . . . D. E v e r n i n o m i c i n a n d L i n c o m y c i n Group . . . . . . . . . E. Other Sugar D e r i v a t i v e s . . . . . . . . . . . . . Macrocyclic Lactone Antibiotics . . . . . . . . . . . . A. N o n p o l y e n e Glycosidic Macrolides . . . . . . . . . . B. Polyene Macrolides . . . . . . . . . . . . . . . C. Other Macrocyclic Antibiotics . . . . . . . . . . . . Quinone Antibiotics . . . . . . . . . . . . . . . A. Tetracyclines . . . . . . . . . . . . . . . B. A n t h r a c y c l i n e s a n d A n t h r a c y c l i n o n e s . . . . . . . . . Amino Acid a n d Peptide Antibiotics . . . . . . . . . . A. Diketopiperazine D e r i v a t i v e s . . . . . . . . . . . B. H o m o p e p t i d e s . . . . . . . . . . . . . . . C. H e t e r o m e r Peptides . . . . . . . . . . . . . . D. Sideromycins . . . . . . . . . . . . . . . . E. Bleomycin Group . . . . . . . . . . . . . . F. A c t i n o m y c i n s . . . . . . . . . . . . . . . G. E c h i n o m y c i n - T y p e Antibiotics . . . . . . . . . . N i t r o g e n - C o n t a i n i n g Heterocyclic Antibiotics . . . . . . . A. M i t o m y c i n Group . . . . . . . . . . . . . . B. P y r i m i d i n e Nucleosides . . . . . . . . . . . . . C. P u r i n e Nucleosides . . . . . . . . . . . . . . O x y g e n - C o n t a i n i n g Heterocyclic Antibiotics . . . . . . . . Alicyclic Antibiotics . . . . . . . . . . . . . . . A r o m a t i c Antibiotics . . . . . . . . . . . . . . A. Chloramphenicol a n d I t s Derivatives . . . . . . . . . B. Novobiocin . . . . . . . . . . . . . . . .
101 102 105 110 110 116 119 119 123 126 128 128 129 129 134 136 137 137 141 144 144 145 146 146 148 150 150 152 152 152 153 154 156 159 159 160
[7]
PAPER
CHROMATOGRAPHY
OF ANTIBIOTICS
C. Griseofulvin . . . . . . . . . . . . . . . . . D. Miscellaneous A r o m a t i c Antibiotics . . . . . . . . . . X I I I . Classification a n d S y s t e m a t i c Analysis of Antibiotics b y M e a n s of Paper C h r o m a t o g r a p h y . . . . . . . . . . . . . . . A. Salting-out C h r o m a t o g r a m s . . . . . . . . . . . . B. p H C h r o m a t o g r a m s . . . . . . . . . . . . . . . C. S u m m a r i z e d C h r o m a t o g r a m s ( C h r o m a t o g r a p h i c Spectra) . . . .
101 161 162 162 162 164 168
I. Introduction
Paper chromatography, introduced in this field in 1946 by Goodall and Levi I still remains one of the principal tools used for separation, characterization, and identification of antibiotics. About two thousand antibiotics hitherto known belong to very different groups of organic compounds. It is therefore ahnost impossible to develop universal solvent systems for their paper chromatographic studies. It is also difficult to use universal chemical detection methods such as can be done for other groups of chemically related natural or synthetic compounds, e.g., sugars, amino acids, steroids, sulfonamides. On the other hand, the biological activity of antibiotics can be exploited for their detection on paper chromatograms by means of bioautography. Paper chromatography has become an indispensable tool in the study of antibiotics especially for the following purposes: (1) characterization and identification in the search for new antibiotics; (2) separation of mixtures of antibiotics; (3) quantitative analysis; (4) studies of biogenesis and biosynthesis; (5) selection of producing strains; (6) development of isolation procedures for unknown substances; (7) control of isolation and purification; (8) preparative chromatography; (9) control of pharmaceutical preparations; (10) control of feed additives; (11) structural studies; (12) studies of enzymic degradation or transformations of antibiotics; (13) studies of movement of antibiotics in animals and plants; (14) production of antibiotics as one criterion in systematics and taxonomy of microorganisms; (15) systematic analysis. In this chapter, principles of paper chromatography of antibiotics are described. They include all steps of analysis beginning with the preparation of samples and papers, development techniques, detection methods, and means of documentation. Paper chromatography of the best known groups of antibiotics is given in separate sections. Finally, a section is devoted to paper chromatographic classification and systematic analysis of antibiotics. General theoretical aspects of paper chromatography are not dealt with except for those that directly emerged from studies of antibiotics. R. R. Goodall and A. A. Levi, Nature (London) 158, 675 (1946).
102
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
For the theory of paper chromatography, specialized monographs should be consulted. ~-~ I t was impossible to pack all available data on paper chromatography of antibiotics into such a short chapter, and a selection of the most important applications was necessary. Of the older reviews concerning various aspects of paper chromatography of antibiotics, that by the author of the present chapter could be helpful. ~ More recently, two books on paper chromatography were published. One of them, by Blinov and Khokhlov, 7 is in Russian; the other, in English, is by Wagman and Weinstein. 8 The latter was characterized by its authors as an "antibioticist's vade mecum" and, besides paper chromatography, includes applications of thin-layer and gas-liquid chromatography together with electrophoresis and countercurrent distribution. A chapter dealing with paper and thin-layer chromatography of antibiotics is included in a recent book on pharmaceutical applications of paper and thin-layer chromatography. 9 II. G e n e r a l T e c h n i q u e s
Preparation of Samples Samples of antibiotics may be of varioqs purity, and sometime special procedures are needed before their application to chromatograms. Pure substances are dissolved in appropriate solvents. From laboratory or industrial fermentations, filtrates of the cultivation media containing antibiotics can be used directly. In some cases, antibiotics are present in the biomass of the producing microorganisms and must be extracted with an organic solvent (methanol, acetone, chloroform, etc.) and concentrated. In screening programs it is often desirable to have crude eoncenJ. W. Copius-Peereboom, "Comprehensive Analytical Chemistry," Vol. IIC, Chapter 1. Elsevier, Amsterdam, t971. s G. J. Giddings, "Dynamics of Chromatography, Part I: Principles and Theory". Dekker, New York, 1965. I. M. I-Iais and K. Macek, Eds., "Paper Chromatography," 3rd ed. Academic Press, New York, 1964. E. Heftmann, Ed., "Chromatography," 2nd ed. Van Nostrand-Reinhold, Princeton, New Jersey, 1967. 6V. Betina, in "Chromatographic Reviews" (M. Lederer, ed.), Vol. 7, p. 121. Elsevier, Amsterdam, 1965. 7 N. O. Blinov and A. S. Khokhlov, "Paper Chromatography of Antibiotics" (in Russion). Izd. Nauka, Moscow, 1970. 8G. H. Wagman and M. J. Weinstein, "Chromatography of Antibiotics." Elsevier, Amsterdam, 1973. V. Betina, in "Pharmaceutical Applications of Thin-Layer and Paper Chromatography" (K. Macek, ed.), p. 503. Elsevier, Amsterdam, .1972.
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
103
trates of antibiotics for chromatographic characterization. Crude concentrates of antibiotics produced by fungi are prepared as follows.TM 1. To 50 ml of a filtrate obtained from a cultivation of a producing strain, 50 ml of acetone are added for precipitation of proteins and other substances that might interfere in chromatographic analysis of the antibiotics in the filtrate. The mixture is warmed to 50 ° for 10 min and then cooled to room temperature. After precipitation and filtration, the filtrate is evaporated in vacuo to dryness. The residue is dissolved in 5 ml of 80% aq. acetone, filtered, and used for chromatography. Such concentrates may be stored in a refrigerator. 2. The corresponding mycelium is extracted twice with ethyl acetate in a mixer, the combined extracts are filtered and evaporated in vacuo to dryness. The residue is dissolved in 80% aq. acetone and used for chromatography of antibiotics from the mycelium. Solutions of antibiotics are applied to chromatograms in the form of spots or a streak (1.5-3 cm long).
Form and Development o] Chromatograms Chromatograms may be prepared in three general forms: as sheets, as strips, or as circular papers. Unless special requirements are stated, standard papers are used, for example, Whatman No. 1 or Schleicher and Schuell 2043b. For bioautographic detection, narrow strips of chromatographic papers are recommended. In our laboratory, standard-strips 1 X 35 cm are used, and ascending development is carried out in narrow glass cylinders. The lower end of the chromatogram is immersed to a depth of 1 cm into the solvent system in the cylinder, and the upper end of the strip is fixed with a glass stopper. Besides the ascending development, descending, horizontal, or radial development of chromatograms is used. In descending, ascending, or horizontal development on paper sheets one-dimensional or two-dimensional development may be applied. Electrophoresis in one direction and chromatographic development in the other (electrochromatography) can also be combined,
Solvent Systems As mentioned above, the main difficulty in studies of antibiotics by paper chromatography lies in their chemical diversity. This means that specific solvent systems and detection methods must be worked out for different chemical groups. In general, systems with an aqueous stationary phase and with a stationary hydrophilic organic solvent, respectively, can be used for antibiotics. In the former type of system, water is held 1oZ. Bar~th, V. Betina, and P. Nemec, J. Antibiot. 17, 144 (1964).
104
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
stationary on the paper, and the solvent, which is usually immiscible with water, passes through. In the latter type, paper is impregnated with a polar nonvolatile organic solvent and another organic solvent, immiscible with the stationary phase, is used as the mobile phase. The Zaffaroni's systems belonging to this category are given in the section on nonpolyene glycosidic macrolide antibiotics. Of the nine basic systems recommended by Macek, 11 the six listed below could be tested in paper chromatography of antibiotics which were not yet analyzed chromatographically. Systems for hydrophilic substances: A: Isopropanol-ammonia-water (9 : 1 : 2) B: n-Butanol-acetic acid-water (4: 1:5), the upper phase being used for development and both phases being used to saturate the atmosphere in the tank Systems for slightly hydrophilic substances: C: Formamide/chloroform. In formamide systems, usually a 40% ethanolic formamide solution is used for impregnation of the paper. The formamide may contain 5% ammonium formate or 0.5% phosphoric acid. Chloroform is used as mobile phase which does not need to be saturated with formamide. D: Formamide/benzene-chloroform. The proportion of the two components in the mobile phase is selected from 1 : 9 to 9:1 in accordance with the character of antibiotics to be analyzed. E: Formamide/benzene F: Formamide/benzene-cyclohexane in proportions from 9:1 to 1:9 Detection Methods
Usually, the chromatogram is dried prior to detection in order to remove the residues of the solvent systems from the paper. T h i s is very important when bioautography is used as the detection method, since some solvents m a y be harmful to test microorganisms. For colorless antibiotics, physical, chemical, radiochemical, or bioautographic detections are used to locate their positions on chromatograms. T h e y are reported directly in the sections dealing with individual antibiotics and their groups. Bioautography is described in a separate section. The position of the chromatographed substance is usually expressed as the relative Rs value, i.e., the ratio of the distance of the center of the spot A from the start, and the distance of the solvent front, F, from the start: R~ = A / F . The so-called hRs values (hRr = R~ X 100) are often used for greater clarity, particularly in tables. In certain cases, mainly when the R~ values cannot be calculated, relative R x values can be estimated. R x value is the distance A of a given spot from the start divided by the distance B of substance X from the start: R x = A / B . 1~K. Macek, in "Pharmaceutical Applications of Thin-Layer and Paper Chromatography" (K. Macek, ed.), p. 16. Elsevier, Amsterdam, 1972.
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
105
Finally, R~ values can be calculated according to Eq. (1). RM
=
log[(1/Rs) -- 1]
(1)
m . Bioautography The antimierobial activity of different antibiotics varies; they may have antiviral, antibacterial, antifungal, antiprotozoal, or antialgal properties. In addition, some antibiotics possess cytotoxic, antineoplastic, and other biological effects. Bioautography is based on the biological activities of antibiotics. The possibilities of the bioautographic detection of antibiotics possessing various biological activities are summarized in Table I. In a typical procedure of bioautographic detection of antibiotics on paper chromatograms, chromatographic sheets or strips are placed on the surface of large nutrient agar plates inoculated with microorganisms that are sensitive to the antibiotics being analyzed. After about 15-30 rain the sheets are removed, while the narrow strips may be left on the seeded surface. In both instances antibiotics diffuse from their positions on the chromatograms into the agar layer and inhibit the growth of tile test organisms. The plates are then incubated at an appropriate temperature until a thin film of the growing microorganisms is visible on the agar surface. Clear zones of inhibition are seen in the areas where antibiotics are present. With bacteria and some fungi the incubation period is about 16-24 hr. TABLE I POSSIBILITIES OF BIOAUTOGRAPHY OF ANTIBIOTICS IN PAPER CHROMATOGRAPHYa
Biological activity Antibacterial Antifungal Antiprotozoal Antialgal Antiphage Phage-inducing Antiviral Cytotoxic
Test systems
Effects
Bacteriophages plus host bacteria Mixtures of lysogenic plus indicator bacteria Virus-infected animal cells Monolayer animal cell cultures
Growth inhibition b Growth inhibition b Growth inhibition b Growth inhibition h Absence of plaques Halo of lysis on indicator bacteria Absence of cytopathic action Dead cells
Bacteria Fungi Protozoa
Algae
" Adapted from V. Betina, J. Chromatogr. 7@, 41 (1973). Manifested by zones of inhibition on agar plates and by inhibition of growth in test tubes with liquid media, respectively.
106
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
Bioautography is carried out by similar techniques in most laboratories. Wagman and Weinstein8 recommend Pyrex baking dishes approximately 8.5 inches wide by 13.5 inches long by 1.75 inches deep (21.6 X 34.3 X 4.4 cm) with stainless steel covers. In our laboratory, flat dishes are prepared of glass plates 40 X 27 cm with removable aluminum frames 37 X 24 cm and 2 cm high. The frames are sterilized separately, placed on the plates previously adjusted to horizontal position, and adhered to them by pipetting a part of melted agarized medium for the bottom layer (about 65 °) around their inner edges. Typically, a 300-ml portion of a base medium and a 100-ml portion of seed medium are used. The seed medium is poured on top of the solidified base (bottom) layer. When air bubbles occasionally form on the surface of the agar, a Bunsen burner flame can be passed rapidly over the agar to break them. The plates are then allowed to harden. Metallic or glass covers with filter paper on their inner sides are used to cover the plates during incubation. The filter paper absorbs vapors and prevents their condensation on the cover. A large variety of microorganisms or animal cell cultures may be used to detect various antibiotics under test. A number of useful media for such tests can be found in special books. 1~,13 Selected procedures given below include bioautographic detections of antibacterial, antifungal, phage-inducing, cytotoxic, and antiviral antibiotics. Bioautography o] Antibacterial Antibiotics
When commercial media for bioautography are not available, the following cultivation medium can be usedg: peptone, 10 g; meat extract, 1000 ml; NaC1, 3 g; Na~HP04, 2 g, pH 7.2; agar, 20 g for base layer, and 10 g for seed layer, respectively. The medium is autoclaved at 120° for 20 min. The base medium is cooled to about 65 ° and poured onto an appropriate framed glass plate. The seed medium is cooled to about 45 °, seeded with a suspension of the test organism, mixed well (avoid formation of air bubbles), and poured onto the solidified base layer. Typically, to 100 ml of a seed medium is added 1.0 ml of the working inoculum. Usdin et al. 14 added into the agar medium 2,3,5-triphenyltetrazolium chloride (TTC) which is reduced to a formazan of a red coloration. Zones of inhibition are then better visible on a red background. Other tetrazolium salts may be used. TTC and also 2,6-dichlorophenolindophenol ~2F. Kavanagh, Ed., "Analytical Microbiology." Academic Press, New York, 1963. 13D. C. Grove and W. A. Randall, "Assay Methods of Antibiotics. A Laboratory Manual." Med. Encycl., New York, 1955. 14E. Usdin, G. D. Shockman, and G. Toennies, Appl. Microbiol. 2, 29 (1954)
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
107
are recommended for shortening the incubation time from 16 to 5 hr and for a better visualization of the zones of inhibition as follows. 1'~ The seeded plates with the chromatographic strips are incubated for 4 hr and then sprayed with an aqueous 0.05% solution of T T C and 0.5% solution of 2,6-dichlorophenolindophenol, respectively. After an additional incubation for half an hour, inhibition zones begin to be visible. With TTC, noncolored zones appear on a red background; and with the other redox indicator, blue zones appear on a bleached background. For documentation purposes, test agar plates with inhibition zones on the chromatographic strips can be photographed in polarized light which gives the zones more contrast. 1~ Agar plates can also be photographed by a direct contact with photographic papers. 1~ Color reprints on filter paper can be prepared from the test plates colored by special techniques. TM The direct bioautography of antibacterial substances on paper chromatograms has been described) 9 The developed chromatographic strips are dried in the air, carefully immersed into a soluble agar medium seeded with sensitive bacteria, and then incubated in a moist atmosphere at a suitable temperature. The growing bacterial culture on the strips are then colored using an available bacteriological staining procedure. This technique was shown to be more sensitive than the bioautographic detection on agar plates.
Bioautography o] Anti]ungal Antibiotics With yeasts as test organisms, Sabouraud's glucose medium (glucose, 40 g; peptone, 10 g; distilled water, 1000 ml; pH 5.7) is recommended2 Add 1.5% of agar to base medium and 1.0% of agar to seed medium. The media are autoclaved and handled as are those for bioautography of antibacterial antibiotics. The incubation temperature is 28% Antibiotics active on filamentous fungi can be detected by similar procedures when the seed agar is inoculated with conidia of a sensitive organism, and the plates are incubated at its optimal temperature. Direct bioautography on paper chromatograms is also possible by applying the following procedure, originally used for detecting synthetic fungicides. 2°,2' Developed chromatograms on Schleicher and Sehuell ~5V. Betina and L. Pil~tov£, Oesk. Mikrobiol. (Prague) 3, 202 (1958). ~6N. A. Drake, J. Amer. Chem. Soc. 72, 3803 (1950). 1~Anonymous, Chem. Eng. News 32, 3940 (1954). 1~j. Stephens and A. Grainger, J. Pharm. Pharmacol. 7, 702 (1955). 1, G. Csob£n and G. Szab6, Kiserl. Orvostud. 4, 387 (1952). 2o H. C. Weltzien, Naturwissenscha]ten 45, 288 (1958). .-1 H. M. Dekhuizen, Meded. Landbouwhogesch. Opzoekingsstat. Slaat Gent ~6, 1542 (1961).
108
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
2043b or Whatman No. 3 M N paper are dried in air and sprayed with a conidial suspension of Stemphylium consortiale or Glomerella cingulata in a medium prepared as follows: sucrose, 50 g; NAN03, 5 g; KH2P04 1.25 g; MgS04.7 H20, 1.25 g; distilled water, 500 ml; pH 4.6; 2-2.5 X 105 conidia are used per milliliter of the medium. The chromatograms are then incubated on glass plates in a moist atmosphere for 2-3 days at 25-27% Inhibitory zones indicate the positions of the fungistatic substances tested.
Bioautography o] Phage-Inducing Antibiotics A procedure for the detection of substances inducing X bacteriophage of Escherichia coli K12 was described by Heinemann et al. 22 The top agar layer is inoculated with a suspension of the latter lysogenic bacterium along with a nonlysogenic culture of E. coli C600. After incubation, a halo of lysis on the colonies of E. coli C600 indicates the position of the inducers studied on paper chromatograms.
Bioautography for Cytotoxic Antibiotics 23 Earle's L cells and Eagle's KB cells are grown in suspension culture using Waymouth's medium MB 752/1 supplemented with 10% (v/v) calf serum; 0.3 g/liter of 4000 cps methyl cellulose (Methocel, Dow Chemical Co.) ; 1 g/liter nonionic polymer (Pluronics, Wyandotte Chemical Co.) ; and 1 g/liter anionic surfactant (Darvan No. 2, R. T. Vanderbilt Co., Inc.) in 250-ml conical flasks on a rotating shaker as described by Perlman et al. 24 The L-1210 cells are grown in this medium without shaking as the cells remain in suspension under these conditions when incubated at 37% The KB cells are grown in the shaken culture for 3 days at 37 ° and then harvested by centrifugation. The cells of each culture are then suspended in fresh medium so that the cell count is 4 X 106 cells/ml. Of this cell suspension, 50 ml are added to a flask containing 37 ml of calf serum, 1 g of glucose, and 3 g of melted agar in 120 ml of water; the resulting suspension is poured into a 3-qt Pyrex baking dish and allowed to solidify. The chromatograms are placed on the agar surface for 40 rain and then removed. The dishes are loosely covered with an alumihum cover and incubated at 37 ° for 18-20 hr. The agar surface is flooded with 0.05% solution of 2,6-dichlorophenolindophenol and allowed to stain for 5 min. After the dye has been poured off, the plates are placed in a 37 ° incubator for 40-60 rain. Under these conditions the viable cells 32B. Heinemann, A. J. Howard, and Z. J. Hollister, Appl. Microbiol. 15, 723 (1967). 23D. Perlman, W. L. Lummis, and H. J. Geiersbach, J. Pharm. Sci. 58, 633 (1969). 54D. Perlman, J. B. Semar, G. W. Krakower, and P. A. Diass, Cancer. Chemother. Rep. 51, 255 (1967).
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
109
reduce the dye whereas the dead cells do not, so that the zones of toxicity can be located. Siminoff and H u r s k y ~ used a monolayer of human H e L a cells covered with an agar medium. After incubation with paper chromatograms on the plates, the cells were fixed and stained. According to another method -~G a suspension of tumor cells is simply brought into contact with a chromatogram, and later the color reaction with dehydrogenases is applied using 2,6-dichlorophenolindophenol in order to locate the zones of inhibition.
Bioautography o] Antiviral Antibiotics An adaptation of the Dulbecco virus-plaque technique 27 was described for this purpose as follows. 2s Paper chromatograms are sterilized by means of ethylene oxide and after drying are placed for 5 min or agar overlays of virus-infected cultures. After removal of papers, baking dishes are sealed with Saran Wrap (Dow Chemical Co.) and incubated for 4 days at 36% The cell layer is then stained with a second agar overlay containing indonitrotetrazolium chloride, and within a few hours plaques can be readily observed. When very sensitive tests are needed, the authors recommend applying the chromatograms very soon after virus infection of the cells. When a more accurate determination of the area of antiviral effects is required, but the sensitivity is not important, the application of the chromatograms can be delayed.
Special Cases o] Bioautography For microorganisms which do not grow well on agar media, chromatographic strips can be cut into short pieces; the pieces are placed into a series of test tubes containing a liquid cultivation medium inoculated with sensitive microorganisms. The absence of growth in some test tubes after an incubation period indicates the positions of antibiotics on the chromatograms. This technique was used for antileptospiraF 9 and antiprotozoaP ° antibiotics. When unknown antibiotics in samples from the cultivation media or from the mycelia of the producing microorganisms are studied, it is necessary to know the antibiotic activity of samples against different microorganisms. In such cases, representatives of all sensitive organisms are used for bioautographic detection. Very often the analyzed sample con25p. Siminoff and V. S. Hursky, Cancer Res. 20, 618 (1960). 56A. Oda and T. Yamamoto, dap. J. Exp. Med. 29, 87 (1959). 57R. Dulbecco, Proc. Nat. Acad. Sci. U.S. 38, 747 (1952). ~s E. C. Herrmann and J. P. Rosselet, Proc. Soc. Exp. Biol. Med. 104, 304 (1960). ~ P. Nemec, V. Betina, and J. Durkovsk:~, Naturwissenscha]ten 47, 235 (1960). 30j. Balan, L. Ebringer, P. Nemec, 8. Kov£~, and J. Dobias, J. Antibiot. 15, 157 (1963).
110
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
rains several antibiotics with different Ry values and with different antimicrobial activities. 1° In the so-called correlative assays, 31 it was demonstrated, by means of paper chromatography and bioautography, that the same component in a certain filtrate from a microbial fermentation inhibited vaccinia virus in vitro and also a yeast, Saccharomyces pastorianus. Other bioautographic techniques may be found elsewhere. 6-9,32
IV. Paper Chromatography of ~-Lactam Antibiotics This group includes penicillins, cephalosporins, and cephamycins. Penicillins and cephalosporins are fungal metabolites, and cephamycins are a new family of antibiotics produced by actinomycetes. Penicillins are N-acyl derivatives of 6-aminopenicillanic acid (6-APA). Cephalosporin C family and cephamycins are derivatives of 7-aminoeephalosporanic acid (7-ACA). Cephalosporin P series, of steroid nature, is included in the section dealing with alicyclic antibiotics. Cephamycins have been known since 1971 (for references, see Stapley et al2 s) and structures of cephamycins A, B, and C have been elucidated. The methyl group in the acetyl residue of cephalosporin C is substituted by other radicals in the cephamycins.
A. Penicillins and 6-Aminopenicillanic Acid Paper chromatography of natural, biosynthetic, and semisynthetie penicillins and of 6-APA is described here. Sampling. Culture filtrates can be applied directly on chromatograms. Penicillins may also be extracted, after chilling the filtrate and adjusting to pH 2, with butyl acetate and back extracted from the organic layer with 0.5% aq. sodium bicarbonate. 34 References to procedures for penicillins in pharmaceutical preparations may be found elsewhere. 9 Solvent Systems. Of many systems described in the literature, those most used have been selected as follows. A: Ether satd. with water on paper buffered with phosphate buffer, pH 5.8 or 6.2, development at 4-5 °1'~5 B: Ethyl acetate satd. with water on paper buffered with McIlwaine's citrate phosphate buffer pH 6.0 s6 ~IL. J. Ha~ka and C. G. Smith, Antimicrob. Ag. Chemother. 1961, p. 677 (1962). s2V. Betina, J. Chromagogr. 78, 41 (1973). 'SE. O. Stapley, M. Jackson, S. Hernandez, S. B. Zimmerman, S. A. Currie, S. Mochales, J. M. Mata, H. B. Woodruff, and D. Hendlin, Antimicrob. Ag. Chemother. 2, 122 (1972). M. Cole, Appl. Microbiol. 14, 88 (1966). M. L. Karnovsky and M. J. Johnson, Anal. Chem. 21, 1125 (1949). V. Betina, Chem. Zvesti (Bratislava) 18, 209 (1964).
[7l
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
111
C: Amyl acetate satd. with water on paper buffered with procaine-citrate buffer pH 5.337 D: n-Butanol-acetic acid-water (12:3 : 5), upper layer 3s E: n-Butanol-ethanol-water (4:1:5), upper layer 38 F: n-Butanol-pyridine-water (1 : 1 : 1) 38 G: 2-Butanol-formic acid-water (75: 15 : 10) 39 H: n-Propanol-water (6: 4) 39 I: n-Propanol-ethanol-water (5: 2 : 3) 39 J : n-Butanol-n-propanol-water (25: 50: 25) 39 K: 27% KH2PO4 pH 6.0 on paper impregnated with 2% liquid paraffin in ether, air dried, and saturated with vapors of butyl acetate 4° L: Both phases of isopropy! ether-isopropanol-water (7:3:10) on paper buffered with phthalate buffer pH 4 or 541 T h e systems A, B, C, and K are convenient for natural and biosynthetic penicillins. With system B, the dependence of the R~ values of natural penicillins on p H of the stationary phase was studied, ~7 and Sshaped curves were obtained showing optimal separation at p H 6 (Fig. 1). The systems D to J and L were used for separation of 6-APA from penicillins, 39 for seminisynthetic penicillins, ~° and for degradation products of penicillins and 6-APA. 42 Detection. For bioautography of natural and biosynthetic penicillins, Bacillus subtilis, Sarcina lutea, or Staphylococcus aureus are recommended. For penicillin N and semisynthetic penicillins active on gramnegative bacteria, an avirulent strain of Salmonella t y p h i m a y be used. 6-APA can be converted to benzylpenicillin as follows. ~8 The chromatogram is sprayed with 5% aqueous N a H C 0 3 , then with 2 - 5 % phenylacetyl chloride in acetone and again with the N a H C Q solution. After drying for a short time in air the chromatographic strip is placed on a plate seeded with S. aureus or B. subtilis. Of chemical detection methods for fl-lactam compounds the following m a y be used. 1. A mixture composed of l0 m M iodine in 3 m M K I (10 ml), 1 M phosphate buffer p H 7.0 (1 ml), and 2% (w/v) sodium starch glycolate in water (9 ml) is used to spray the chromatograms. Compounds with potential sulfhydryl groups (but not the penicillins with a Closed fll a c t a m ring) appear as pale spots on a deep blue background. The paper is then sprayed with a mixture of equal volumes of the starch-iodine solution and a solution of purified penicillinase (e.g., from Bacillus cereus 1000 units/ml). Additional pale spots appear at the site of penicillins ~7H. G. Macmorine, Appl. Microbiol. 5, 386 (1957). s8M. Cole and G. N. Rolinson, Proc. Roy. Soc. Ser. B 154, 490 (1961). 39M. RShr, Microchim. Acta 4, 705 (1965). 4oT. Watanabe, S. Endo, and Y. Iida, J. Antibiot. 15, 112 (1962). 41H. Hellberg, J. Ass. O~c. Anal. Chem. 51, 552 (1968). 42p. H. A. Sneath and J. F. Collins, Biochem. J. 79, 512 (1961).
112
[71
METHODS FOR THE STUDY OF ANTIBIOTICS Rf 1.0
0.8
0.6
0.4
0.2
I
I
1
I
3
~.
5
1
I
I
I
I
6
7
8
9
10
pH
FIG. 1. Separation of natural penicillins. (A) Dependence of RI values on pH of the stationary phase. Solvent system: ethyl acetate saturated with water. Stationary phases: McIlvaine's citrate-phosphate buffers and phosphate buffers, respectively. (B) Separation of penicillins from the same sample using ethyl acetate as mobile phase and McIlvaine's buffer, pH 6.0. Traveling distance of the solvent front, 30 cm. Descending development indicated by the arrow. Bioautography with Bacillus subtilis. Modified from V. Betina, Chem. Zvesti (Bratislava) 18, 209 (1964). whose fl-lactam ring is opened by the penicillinase. Sensitivity with benzylpenicillin is 1 ttg/cm -~. Methicillin is revealed less readily owing to its relative stability to penicillinase, p-Hydroxybenzylpenicillin absorbs iodine before treatment with penicillinase owing to iodination of the phenolic side chain. The method cannot be applied to cephalosporin C . 42
[7]
P A P E R CHROMATOGRAPHY OF ANTIBIOTICS
113
TABLE II PAPER CHROMATOGRAPHIC DATA OF NATURAL AND BIOSYNTHETIC PENICILLINS
RGa in systems b
Penicillins F FH2 G K V X
A at pH 5.6¢ B at pH 6.0d 1.62 2.18 1.00 3.10 -0.17
2.42 2.92 1.00 3.26 -0.11
Rs X 100 in system K b'~ 22 -48 14 36 59
RpenicilliaG. b See solvent systems for penicillins on p. 110. c From S. Yamatodani, in "Papierchromatographie in der Botanik" (H. Linskens, ed.), p. 181. Springer, Berlin, 1955. d Estimated from drawings in V. Betina, Chem. Zvesti (Bratislava) 18, 209 (1964). e From T. Watanabe, S. Endo, and Y. Iida, J. Antibiot. 15, 112 (1962).
a RG
=
2. In a variation of the previous method, the paper is sprayed with 0.5 N N a O H and,dried for 1-15 min (in order to open the fl-lactam ring) before spraying with a reagent composed of a mixture of 1% aqueous starch, glacial acetic acid, and 0.1 N iodine in 4% K I (50:3:1). Decolorization of the iodine reagent by the hydrolyzed penicillins and cephalosporin C yields maximum contrast after 5-10 rain. 4a 3. The chromatogram is immersed into a solution of A g N Q in acetone (1 ml of satd. aqueous AgNO~ added dropwise into 100 ml of acetone, and the precipitate dissolved by adding water). After drying in air, the paper is immersed into a solution of 2.5 ml of 50% N a O H in methanol until there is maximal development of the color reaction. The chromatogram is washed briefly in water and, to decolorize the background, immersed into 6 N NH~OH and washed carefully in tap water. 39 A p p l i c a t i o n s . Mobilities of natural and biosynthetic penicillins in different solvent systems obtained independently by three authors are compared in Table II. 6-APA was separated from natural penicillins 4~ in system E. Mobilities of penicillins in the same system were~: penicillin K, FH~ ~ F, G ~ penicillin 4 ~ penicillin 3 ~ penicillin 2 ~ penicil4~R. Thomas, Nature (London) 191, 1161 (1961). " F. R. Batchelor, F. P. Doyle, J. H. C. Nayler, and G. N. Rolinson, Nature (London) 183, 257 (1959). 4~A. Ballio, E. B. Chain, and F. D. Di Accadia, Nat~re (London) 183, 180 (1959).
114
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
lin 1. RShr, 39 using 9 solvent systems, including G, H, I, and J, characterized phenoxymethyl penicillin, phenoxymethylpenicilloic acid obtained from the former by degradation with penicillinase, 6-APA, and penicilloic acid obtained from 6-APA by degradation with penicillinase. The following R~ values were observed in system F3S: methylpenicillin, 0.58; penicillin X, 0.3; 6-APA, 0.48; penicillin, 0.75. In system E, a-aminobenzylpenicillin had a lower mobility than had a-azidobenzylpenicillin. 46 Hellberg 41 compared 8 penicillins using system L. At pH 5.0, their RI values were as follows: ampicillin, 0.0; benzylpenicillin, 0.23; cloxacillin, 0.56; methicillin, 0.04 oxacillin, 0.50; phenoxymethylpenicillin, 0.33; phenoxyethylpenicillin, 0.55; and phenoxypropylpenicillin, 0.80. At pH 4.0, the separation was less successful. Cole 47 studied a series of reaction mixtures of 6-aminopenicillanic acid with carboxylic acids, amides, and N-acyl derivatives of glycine using system A at pH 6.2 and system E. The chromatograms were detected with Bacillus subtilis. Cole 4s compared several esters of benzylpenicillin and other compounds using paper chromatography in systems D, E, and F. In studying effects of exogenous penicillin V on penicillin biosynthesis, 49 penicillins and 6-APA were identified by applying samples and standards to Whatman No. 1 paper, which underwent an overnight development in system E. After phenylacetylation and drying, chromatograms were placed on agar plates inoculated with B. subtilis, then removed after 15 min contact; the plates were incubated overnight. Duplicate chromatograms were prepared in those studies in which 35S-labeling was used. The RI values of radioactive peaks were compared with those of bioactive zones. The R~ values of 6-APA, penicilloic acid, and penicillin V were 0.19, 0.27, and 0.63, respectively. Penicilloic acid and other biologically inactive penicillins were detected by spraying chromatograms with iodine 43 as described above. Paper chromatography helped to demonstrate that removal of the side chain of penicillins is a minor transformation route in man) ° The resulting 6-APA was detected in urine samples by developing chromatograms in system E and detecting bioautographically with B. subtilis after phenylacetylation (RI = 0.13). Older data concerning paper chromatography of penicillins (biosyn,6E. Hansson, L. Magni, and S. Wahlqvist, Antimicrob. Ag. Chemother. 1967, p. 568 (1968). 4, M. Cole, Biochem. J. 115, 747 (1969). 48M. Cole, Biochem. J. 115, 733 (1969). ~9E. L. Gordee and L. E. Day, Antimicrob. Ag. Chemother. 1, 315 (1972). 50M. Cole, M. D. Kenig, and V. A. Hewitt, Antimicrob. Ag. Chemother. 3, 463 (1973).
[7]
P A P E R CHROMATOGRAPHY OF ANTIBIOTICS
115
thesis by various fungi and their nmtants, commercial preparations, discovery of 6-APA, semisynthetic penicillins, degradation products, etc.) m a y be found elsewhere2 ,s Methods for quantitative analysis of penicillins and 6-APA are described below. 1. ~ A T U R A L AND BIOSYNTHETIC PENICILLINS. Karnovsky and Johnson :'~ used dry filter paper strips buffered with 20% pH 6.2 phosphate buffer. Samples containing 1-2 units of penicillins were applied. After development in system A, the positions and concentrations of the individual components were determined either by measurement of the zones of inhibition produced when the entire strips were laid on a large agar plate seeded with B. subtilis, or by cutting the strips into small uniform squares and measuring the circular inhibition zones produced by the individual squares. A convenient measure of the amount of each penicillin was given by the area under the relevant portion of a standard curve. The bestknown natural penicillins migrated with increasing mobilities as follows: G, F, FH2, and K. 2. LABELEDPENICILLINS. Using the buffer and solvent as in (1), Smith and Allison 51 described a quantitative method for the detection and evaluation of labeled penicillins. The developed chromatographic strips are left in contact with an X - r a y film for several days, and the film is then developed. Two alternative procedures can be then used. (1) Using the radioautograph as a guide, the corresponding strip is cut into sections each containing the whole of one penicillin species. Each section is then extracted by boiling for a few minutes with a very dilute phosphate buffer. An aliquot of each extract is evaporated down on a planchette, and the radioactivity is measured. From the total counts for each penicillin species, the proportions can be readily calculated. (2) Alternatively, again using the radioautographs as a guide, the corresponding paper strips are cut into squares, or, if necessary, rectangles, in such a way as to have each penicillin on a separate square. The squares are then fitted into planchettes and measured radiometrically. 3. 6-AMINOPENICILLANICACID. 6-APA can be estimated quantitatively according to Erickson and Bennett. '~ Penicillins are separated from 6-APA by development, in system E, of chromatograms, which are then air dried and dipped in a solution of 4% phenylacetyl chloride in methyl isobutyl ketone to convert 6-APA to benzylpenicillin. The chromatograms are dried under a hood and then bioautographed with S. lutea. Levels of 6-APA in the samples are determined by plotting zone areas against a standard curve established by treating known amounts of 6-APA in Smith and D. Allison, Analyst 77, 29 (1952). ~2R. C. Erickson and R. E. Bennett, Appl. Microbiol. 13, 738 (1965). 5~ E. L.
116
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
t h e s a m e m a n n e r as t h e samples. T h e e v i d e n c e for t h e presence of 6 - A P A in a r e a c t i o n m i x t u r e is b a s e d on c o m p a r i s o n s w i t h a u t h e n t i c 6 - A P A in r e s p e c t to b o t h c h r o m a t o g r a p h i c m o b i l i t y a n d a n t i b a c t e r i a l p r o p e r t i e s , b o t h before a n d a f t e r p h e n y l a c e t y l a t i o n a n d i n a c t i v a t i o n b y penicillinase.
B. C e p h a l o s p o r i n C F a m i l y , 7 - A C A , a n d C e p h a m y c i n s S o l v e n t S y s t e m s . Selected s o l v e n t s y s t e m s for these c o m p o u n d s are c o m p i l e d below. M o s t of t h e m are used for t h e c e p h a l o s p o r i n C f a m i l y . S y s t e m s K a n d L s e p a r a t e c e p h a l o s p o r i n C, p e n i c i l l i n N, a n d c e p h a l o s p o r i n P. S y s t e m s A a n d S are h e l p f u l in s e p a r a t i n g 7 - A C A a n d its d e r i v atives. S y s t e m s O, P, a n d R are r e c o m m e n d e d for c e p h a m y c i n s . S o l v e n t s y s t e m s for c e p h a l o s p o r i n s , 7 - A C A a n d c e p h a m y c i n s :
A: B: C: D: E: F: G: H: I: J: K: L: M: N: O: P:
n-Butanol-acetic acid-water (4:1 : 5), upper layer 5~ As A, but (5:1:4) 54 As A, but (4:1:2)55 As A, but (3:1:1) 58 As A, but (4:1:4) 57 n-Butanol-ethanol-water (4:1 : 5) 58 n-Butanol satd. with 5 N ammonia ~9 2-Butanol satd. with water 59 2-Butanone satd. with water 59 Methyl ethyl ketone satd. with water e° Methanol-n-propanol-water (6: 2: 1) on paper buffered with 0.75 M phosphate pH 4.0 °1 n-Propanol-0.013 M Na citrate pH 5.5 (7:3) 6~ n-Propanol-water (7:3) 6~ Isopropanol-pyridine-water (65: 5: 30) 63 n-Propanol-pyridine-acetic acid-water (15: 10: 3: 12) 6a n-Propanol-pyridine-acetic acid-CHsCN-water (45: 30: 9: 40: 3 6) 64
~3B. Loder, G. G. F. Newton, and E. P. Abraham, Biochem. J. 79, 377 (1961). 54K. Kariyone, H. Harada, M. Kurita, and T. Takano, J. Antibiot. 23, 131 (1970). 6~j. Kozatani, M. Okui, T. Matsubara, and N. Nishida, J. Antibiot. 25, 86 (1972). 56M. M. ttoehn and C. T. Pugh, Appl. Microbiol. 16, 1132 (1968). 57j. D' A. Jeffery, E. P. Abraham, and G. G. F. Newton, Biochem. J. 81, 591 (1961). ~B. Loder, G. G. F. Newton, and E. P. Abraham, Biochem. J. 79, 408 (1961). 59H. R. Sullivan and R. E. McMahon, Biochem. J. 102, 976 (1967). R. P. Miller, Antibiol. Chemother. 13, 689 (1962). ~ J. L. Ott, C. W. Godzeski, D. Pavey, J. D. Farran, and D. R. ttorton, Appl. Microbiol. 10, 515 (1962). 52A. L. Demain, R. B. Walton, J. F. Newkirk, and I. M. Miller, Nature (London) 199, 909 (1963). ~D. R. Brannon, D. S. Kukuda, J. A. Mabe, F. M. Huber, and J. G. Whitney, Antimicrob. Ag. Chemother. 1, 237 (1972). J. G. Whitney, D. R. Brannon, J. A. Mabe, and K. J. Wicker, Antimicrob. Ag. Chemother. 1, 247 (1972).
[~]
PAPER CHROMATOGRAPHY OF~ ANTIBIOTICS
117
R: I s o p r o p a n o l - w a t e r (70:30), descending development approximately 3.5 hr on 3 M M paper~6 S: E t h y l acetate satd. with aq. N a acetate buffer (0.1 M with respect to N a *) p H 5.2 on paper buffered with same buffer 53
Detections. Bioautography with Bacillus subtilis ATCC 6633 was used for cefazolin, cephalotin, deacetyl cephalotin, cephalosporin C and its deacetyl, lactone, and side-chain derivatives. Sarcina lutea PCI 1001 was used for cephaloglycin and deacetyl cephaloglycin; Staphylococcus aureus (Oxford strain NCTC 6571) for cephalosporin C,. cephalosporin Co, deacetyl cephalosporin C, cephalosporin CA, 7-ACA, and related compounds converted to N-phenylacetyl derivatives; Salmonella typhi for cephalosporin C and its derivatives; Pseudomonas solanaceum Lilly X185 for A16886A, A16886B, cephalosporin C, and deacetyl cephalosporin C; Vibrio percolans MB-1272 for cephamycins A, B, and C. 7-ACA and related compounds must be converted to N-phenylacetyl derivatives by spraying dried chromatograms with pyridine in 50% acetone (v/v) until barely damp. The ehromatograms are then lightly spayed with 2% (w/v) phenylacetyl chloride in acetone, and again with the pyridine solution until a spot of bromocresol green placed on the paper immediately turns blue (pH 5.0). The paper is dried in air for 3-5 min and bioautographed with S. aureus2 G Detection with ninhydrin results in purple spots of cephalosporin CA derivatives. Cephalosporin C is detectable in UV light at 230-400 nm giving dark light-absorbing spots, r'6 Applications. Paper chromatography was one of the important tools in separation and identification of cephalosporins produced by various fungi (for references, see Betina6). It was also used in studies of enzymic degradations of cephalosporins 59,~2,6~,~ A16886A and A16886B, fllactam antibiotics from Streptomyces clavuliger were separated with system p.~4,~5 Cephalosporin C, deacetyl cephalosporin C, and A16886A were separated from each other using system O. 63 Cephamycins were characterized by their mobilities in system R 33 and A16886B also in system p.63 Paper chromatographic data of cephalosporins, their derivatives, and cephamycins are given in Table III. A quantitative paper chromatographic analysis of cephalosporin C and related compounds was described by Miller2 ° An antibiotic prepara6~D. R. Brannon, J. A. Mabe, R. Ellis, J. G. Whitney, and R. Nagarajan. Antimicrob. Ag. Chemother. 1, 242 (1972). C. W. Hale, G. G. F. Newton, and E. P. Abraham, Biochem. J. 79, 403 (1961).
~' C. O'Callaghan and P. W. Muggleton, Biochem. J. 89, 304 (1963). N. Nishida, Y. Yokota, M. Okui, Y. Mine, and T. Matsubara, J. Antibiot. 21, 165 (1968).
118
METHODS FOR THE STVDV OF ANTIBIOTICS
[7]
TABLE III PAPER CHROMATOGRAPHIC DATA Of CEPHALOSPORINS AND CEPHAMYCINS
Rs X 100 in systems a Compounds 7-Aminocephalosporanic acid Cefazolin Cephalosporin C Cephalosporin CA (pyridine) Cephalosporin CA (pyridine) nucleus Cephalosporin Cc Cephalosporin Cc nucleus Cephamycin A Cephamycin B Cephamycin C Deacetyl cephalosporin C N-Phenylacetyl-7-ACAt N-Phenylacetylcephalosporin C N-Phenylacetylcephalosporin Cc
Ab
Cc
14 . -75 04 -O0 . 07 . 09 -26 . . . . . . . . . . --40 . 13 . 23 .
E~
Md
.
. -77 . . 98 .
. -78 . .
. . 85
.
. . . . 57
. . .
. . .
60 . . .
Re
-56
-64 73 44 --
a See solvent systems for cephalosporins, etc., on p. 116. b From B. Loder, G. G. F. Newton, and E. P. Abraham, Biochem. J. 79, 408 (1961). c From J. Kozatani, M. Okui, T. Matsubara, and N. Nishida, J. Antibiot. 25, 86 (1972). From J. D' A. Jeffery, E. P. Abraham, and G. G. F. Newton, Biochem. J. 81,591 (1961). e From E. O. Stapley, M. Jackson, S. Hernandez, S. B. Zimmerman, S. A. Curie, S. Mochales, J. M. Mata, H. B. Woodruff, and D. Hendlin, Antimicrob. Ag. Chemother. 2, 122 (1972). 1 7-ACA = 7-aminocephalosporanic acid. t i o n is dissolved in w a t e r or a m y l acetate, a n d a v o l u m e e s t i m a t e d to c o n t a i n a p p r o x i m a t e l y 0.2 ~g (0.5 ~g for the d e a c t y l a t e d c o m p o u n d s ) is a p p l i e d to each strip of W h a t m a n No. 1 paper. T h e strips are developed in s y s t e m J for 3 hr a t 23 ° a n d detected with B. subtilus. F o r the p a r e n t compounds, the i n c r e m e n t s are 0.05, 0.10, a n d 0.20 ~g. T h e m a x i m u m w i d t h of the zones of i n h i b i t i o n are measured. T h e average m a x i m u m d i a m e t e r s for each of the three s t a n d a r d samples are p l o t t e d a g a i n s t the a m o u n t of a n t i b i o t i c on s e m i l o g a r i t h m i c paper. K n o w i n g the r e l a t i v e specific activities for the v a r i o u s compounds, it is possible to d e t e r m i n e the a m o u n t of p a r e n t c o m p o u n d (cephalosporin C ) , d e a c e t y l c o m p o u n d , a n d l a c t o n e s i m u l t a n e o u s l y in a n u n k n o w n s a m p l e from the one s t a n d a r d curve for the p a r e n t compound. R e a c t i o n m i x t u r e s c o n t a i n i n g 7 - A C A a n d c e p h a l o s p o r i n C are a n a l y z e d for 7-ACA in the same m a n n e r as described a b o v e for 6 - A P A after p h e n y l a c e t y l a t i o n . 52
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
ll9
V. Carbohydrate Antibiotics I n a c c o r d a n c e with B ~ r d y ' s classification of a n t i b i o t i c s p 9 this group includes a n t i m i c r o b i a l agents c o n t a i n i n g sugar as the sole basic constit u e n t , as well as those where the s u g a r m o i e t y c o n s t i t u t e s the m a i n skelet o n of the molecule. T h e r e f o r e p a p e r c h r o m a t o g r a p h y of aminoglycosides, other glycosides (e.g., s t r e p t o t h r i c i n s , v a n c o m y c i n ) a n d v a r i o u s sugar d e r i v a t i v e s (e.g., l i n c o m y c i n ) is described here.
A. A m i n o g l y c o s i d i c A n t i b i o t i c s Solvent Systems A: B: C: D: E: F: G:
n-Butanol with 2% (w/v) p-toluenesulfonic acid TM 5 % Ammonium chloride in water TM n-Butanol-acetic acid-water (2: l: 1) 71 2 % p-toluenesulfonic acid monohydrate in n-butanol satd. with water 72 n-Butanol-methanol-water-p-toluenesulfonicacid (10 ml:40 ml : 20 ml:l g)7~ Methanol-3% NaC1 in water (2: 1) TM 80% aq. methanol-piperidine (100: 10.5) adjusted to pH 9.0-9.5 with acetic acid75 H: Methanol-water-glacial acetic acid (80:15:5) TM I: 50% Acetone, samples applied in 5% NaC1TM J: n-Butanol-water-piperidine (84:16:2) TM K: Methylethyl ketone-t-butanol-methanol-6.5 N ammonia (l 6: 3:1 : 6)77 L: Methanol-ethanol-conc. HCl-water (50: 25: 6: 19) TM M: n-Propanol-glacial acetic acid-water (9: l: 10) 79 N : n-Butanol-pyridine-water (6: 4 : 3) 80 O: Methanol-water (4: l) plus 3% NaC1 on paper buffered with 0.95 M Na2S()4 plus 0.5 M NariS0481 P: n-Propanol-pyridine-acetic acid-water ( 15: 10: 3 : 12) 8, Q: n-Propanol-water-acetic acid (50: 40:4)81 R: 80% aq. phenol 8~ 0~j. B~rdy, In]orm. Bull. Int. Cent. In]orm. Antibiot. 10, 1 (1973). 7oA. C. Sinclair and A. F. Winfield, Antimicrob. Ag. Chemother. 1961, p. 503 (1962). ~l D. J. Mason, A. Dietz, and R. M. Smith, Antibiot. Chemother. I1, 118 (1961). ~ J. N. Pereira, J. Biochem. Microbiol. Tech. Eng. 3, 79 (1961). ~'~N. Ishida, J. Miyazaki, S. Okamoto, and K. Omachi, J. Antibiol. 6, 1 (1953). 74A. Saito and C. P. Schaffner, Proc. Int. Congr. Biochem. 3rd, 1955, p. 98 (1955). 75T. Miyaki, H. Tsukiura, M. Wakae, and H. Kawaguchi, J. Antibiot. 15, 15 (1962). ~ M. K. Majumdar and S. K. Majumdar, Anal. Chem. 39, 215 (1967). ~TM. K. Majumdar and S. K. Majumdar, Appl. Microbiol. 17, 763 (1969). 74H. Umezawa, S. Takasawa, M Okanishi, and R. Utahara, J. Antibiot. 21, 81 (1968). n K. L. Rinehart, Jr., A. D. Argoudelis, W. A. Goss, A. Sohler, and C. P. Schaffner, J. Amer. Chem. Soc. 82, 3938 (1960). 8oW. T. Shier, K. L. Rinehardt, Jr., and D. Gottlieb, Biochemistry 63, 198 (1969). ~' M. J. Weinstein, G. M. Luedemann, E. M. Oden, and G. H. Wagman, Antimicrob. Ag. Chemother. 1963, 1 (1964).
120
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
S: Chloroform-methanol-17 % ammonia (2 : 1 : 1, lower layer) in atmosphere satd. with vapors of the top layer 82 T: n-Butanol-pyridine-water-acetic acid (6:4:3:1) 83 U: Methanol-water-acetic acid-25% aq. NaC1 (600:212:75:20.8) on paper buffered at pH 384 V: Methanol-5% aq. NaC1 (2:1) on paper buffered at pH 384 W: Ethanol-water-acetic acid-25 % sq. NaC1 (250: 500 : 38: 7.5) s4 Systems A and B are used for actinospectacin, C - F for streptomycin (plus system L), kanamycins, and neomycins. Bluensomyein (glebomycin) gives good results in G-I. Systems J, K, and M are also used for neomycins, N for hybrimycin. Gentamicins are analyzed in systems 0 to S, the latter together with T are used for lividomycin. Butirosins m a y be characterized in T to W. Detections. Staphylococcus aureus, Bacillus subtilis, and Escherichia coli are mostly used for bioautography. Chemical detections given below are also recommended. 1. Streptomycin and other antibiotics containing guanidine groups (primycin, viomycin) are detected according to Szilhgyi and Szab6 s5 as follows. Pull chromatograms through a cold 0.01% a-naphthol solution in methanol containing 5% NaOH. D r y in air for 2--3 min. Pull through a cold 0.5% N-bromosuccinimide solution in 10% ethanol; pull through 40% urea solution for stabilization of the color and then dry in air. 2. Ninhydrin reagent for neomycins. ~7 Spray air-dried chromatograms with a ninhydrin solution (0.25 g of ninhydrin in 10 ml of methanol plus 47 ml of butanol plus 3 ml of water plus 50 ml of pyridine) heat at 80-90 ° for 30 min. 3. Starch-iodine reagent for N-acetylneomycins. 7G Dissolve 1 g of starch and 0.25 g of potassium iodide by heating in 3.5 ml of water and then add quickly 0.5 ml of this solution to 50 ml of pyridine. Use freshly prepared reagent. After transferring the sprayed chromatogram to a water-saturated atmosphere, bluish pink spots are obtained. 4. Ninhydrin reagent for gentamicins. 8: Spray with 0.25% ninhydrin in pyridine-acetone (1:1) and heat at 105 ° for several minutes. Purple or blue spots appear against a white background. Application~. Paper chromatographic data of streptamine derivatives are given in Table IV, and relative mobilities of streptomycins in Table 82G. M. Wagman, J. A. Marquez, and M. J. Weinstein, J. Chromatogr. 34, 210 (1968). S. Umezawa, I. Watanabe, T. Tsuchiya, H. Umezawa, and H. Hamada, J. Antibiot. 25, 617 (1972). H. W. Dion, P. W. K. Woo, N. E. Willmer, D. L. Kern, J. Onaga, and S. A. Fusari, Antimicrob. Ag. Chemother 2, 84 (1972). 8~I. Szilhgyi and I. SzabS, Arzneim.-Forsch. 8, 333 (1955).
[7]
PAPER
CHROMATOGRAPHY
121
OF ANTIBIOTICS
tIittll~ 0~
iitiitt~
c~
llrlltl~ ©
Illl[ll~
r~
z
© o
~111t111 o
~IIIIlll o C)
~illllil
~;~o
o
~'~
122
M E T H O D S F O R T H E S T U D Y OF A N T I B I O T I C S
[7]
TABLE V RSTR a VALUES FOR STREPTOMYCINS
RSTR in systems b Antibiotics Streptomycin Dihydrostreptomycin Hydroxystreptomycin Mannosidostreptomycin Mannosidodihydrostreptomycin
D° 1.00 0.62 0.64 0.37 --
Ed 1.00 0.70 -0.40 0.25
a RSTR = Rstreptomycin-
bSee solvent systems for aminoglycosides on p. 119. c From J. N. Pereira, J. Biochem. Microbiol. Technol. Eng. 3, 79 (1961). d From K. Kavanagh, E. Grinnan, E. Allanson, and D. Tunin, Appl. Microbiol. 8. 160 (1960).
V. RI values of gentamicin, kanamycins, neomycins, and their derivatives and of other 2-deoxystreptamine containing antibiotics are compiled in Table VI. Relative mobilities of neomycins and their N-acetyl derivatives m a y be found in Table VII. Relative mobilities of neamine and paromamine as Rp~,......... ine in system T were neamine, 0.47; paromamine, 1.00. 86 Gentamicins are characterized in Table V I I I . Lividomycin is inactivated by P s e u d o m o n a s aeruginosa by the formation of a phosphorylated product. The active and inactive lividomycin in system S had the R~ values 0.77 and 0.82, respectively, s7 I n system T, R l i v i d o m y c i n B of 5"-deoxylividomycin B sa was reported to be 1.1 and Rbutirosin B of 3',4',-dideoxybutirosin B s8 was 1.73. Butirosin complex was differentiated from most aminoglycosidic antibiotics in systems U, V, and W by Rj values of 0.21, 0.5, and 0.71, respectively, s4 Quantitative analysis of neomycins as free bases in samples from fermentations have been described. 77 Paper chromatographic methods for quantitative analysis of gentamicins are also known, s9-91 See also the Addendum. s~K. Tatsuta, E. Kitazawa, and S. Umezawa, Bull. Chem. Soc. Jap. 40, 2371 (1967). 87F. :Kobayashi, M. Yamaguchi, and S. Mitsuhashi, Antimicrob. Ag. Chemother. 1, 17 (1972). 86D. Ikeda, T. Tsuchiya, and S. Umezawa, J. Antibiot. 26, 307 (1973). soG. H. Wagman, E. M. Oden, and M. J. Weinstein, Appl. Mie~iol. 16, 624 (1968). N. Kantor and G. Selzer, J. Pharm. Sci. 57, 2170 (1968). 9, G. H. Wagman, J. V. Bailey, and M. M. Miller, J. Pharm. Sci. 57, 1319 (1968).
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
123
B. Streptothricin G r o u p S o l v e n t S y s t e m s . T h e systems for this f a m i l y of N - g l y c o s i d i c antibiotics are collected below. Systems A and B are used for separation of streptothricins, geomycin, pleocidin, mycothricins, polymycins, p h y t o baeteriomycin, r a c e m o m y c i n s , griseomycin, nourseomycins, virothricins, etc. (see references in Blinov and KhokhlovT). R a d i a l development is m o s t l y used. A: B: C: D: E: F: G: H: I:
n-Propanol-pyridine-acetic acid-water (15: 10:3 : 12) 92 n-Butanol-pyridine--acetic acid-water (15 : 10:3 : 12) 93 n-Butanol-pyridine-acetic acid-water-t-butanol ( 15: 10: 3 : 12 : 4) 94 75% aq. ethanol on paper impregnated with 0.95 M Na sulfate 95 n-Butanol satd. with water plus 2% p-toluenesulfonic acid 96 80% aq. methanol-piperidine (10: 1) adjusted to pH 9.3 with acetic acid 97 75% Phenol 9s n-Butanol-methanol-ammonium hydroxide-water (10: 4 : 3 : 3) 99 Ethanol-water (3: 1) plus 2% NaC11°°
Detection. T h e detection of antibiotics of this group is possible with S. aureus, B. s~btilis, or E. coli and, chemically, with ninhydrin. Applications. T h e following R I values of streptothricins a n a l y z e d in form of hydrochlorides (and s u l f a t e s - - v a l u e s in parentheses) in system A were obtained: ~°~ streptothricin A, 0.20 (0.17); streptothricin B, 0.14 (0.20); streptothricin C, 0.29 (0.27) ; streptothricin D, 0.35 (0.32); streptothricin E, 0.40 (0.36); streptothricin F, 0.48 (0.43). I n system C, r a c e m o m y c i n s were characterized by Rj values: 1°2 A, 0.33; B, 0.18; C, 0.24; and D, 0.11. R~ . . . . . idin ~ values obtained in system I ~°° are: lemacidin B1, 0.34; lemacidin B2, 0.61; lemacidin B:~, 1.00. P r o d u c t s of h y d r o l y sis of streptothricins, L-fl-lysine, streptolidine, N-guan-streptolidylgulosamine and its h y d r o l y z a t e , were separated in s y s t e m C. '~4 9~M. I. Horowitz and C. P. Schaffner, Anal. Chem. 30, 1616 (1958). ~ A. S. Khokhlov and P. D. Reshetov, J. Chromatogr. 14, 495 (1964). g4A. S. Khokhlov and K. I. Shutova, J. Antibiot. 25, 501 (1972). ,s L. M. Larson, H. Sternberg, and W. H. Peterson, J. Amer. Chem. Soc. 75, 2036 (1953). Y. Saburi, J. Antibiot. Ser. B 6, 402 (1953). ~7K. Akasaki. H. Abe, A. Seino, and S. Shirato, J. Antibiot. 21, 98 (1968). ~Y. Kasukabe, Y. Yamauchi, C. Nagatsu, H. Abe, K. Akasaki, and S. Shirato, .1. Antibiot. 22, 112 (1969). Y. Kono, S. Makino, S. Takeuchi, and H. Yonehara. J. Antibiot. 22, 583 (1969). ~ooE. Gaeumann and F. Benz, U.S. Patent 3,689,816 ; May 14, 1963. ~01A. W. Johnson and J. W. Westley, J. Chem. Soc. London p. 1642 (1962). ~0~H. Taniyama, Y. Sawada, and T. Kitagawa, J. Chromatogr. 56, 360 (1971).
124
METHODS FOR T H E STUDY OF ANTIBIOTICS
irlJlrr~l~l ~r
[7]
I
II
I
f J J f r f J~
~f
I
II
I
I r I [ I I i~
~l
I
II
1
I
r l
I
II
I
I
II
I
I
II
[
I
I I
I
I
C
iI
I~ ~ - ~
ii
III
II
I~ ~
I
I I
09
z
•, $
¢q
O
I
~o
[I
I
II
~llll ~1111
o
I
~lr
IIIll
~5
IIIII ©
[
Ill
IIIIIIII
[7]
PAPER
CHROMATOGRAPHY
¢D
¢D
~v
O
~o
OF ANTIBIOTICS
125
126
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
T A B L E VII RELATIVE MOBILITIES OF NEOMYCINS AND /-AcETYLNEOMYCINS
Compounds
RNeamine in system K a'~
Neamine base Neomycin B Neomycin C N-Acetylneamine N-Acetylneomycin B N-Acetylneomycin C
1.00 0.54 0.30 ----
RN-acetylneamine in system Jb,c ---1.00 0.69 0.35
From M. K. M a j u m d a r and S. K. Majumdar, Appl. Microbiol. 17, 763 (1969). a F r o m M. K. M a j u m d a r and S. K. Majumdar, Anal. Chem. 39, 215 (1967). See solvent systems for aminoglycosides on p. 119.
TABLE VIII RELATIVE MOBILITIES OF GENTAMICINS
Compounds Gentamicin Gentamicin Gentamicin Gentamicin Gentamicin
Rneamine Rgentamicin Bl i n s y s t e m K ~'¢ in system S b'¢
A1 A B X B1
Gentamicin Ca sulfate Gentamicin C1~ sulfate Gentamicin C2 sulfate
-----2.25 3.75 3.07
0.36 0.41 0.52 0.55 1.00 ----
a F r o m M. K. M a j u m d a r and S. K. Majumdar, Appl. Microbiol. 17, 763 (1969). b From G. H. Wagman, J. A. Marquez, J. V. Bailey, D. Cooper, M. J. Weinstein, R. Tkach, and P. Daniels, J. Chromatogr. 70, 171 (1972). See solvent systems for aminoglycosides on p. 119.
C. O l i g o s a c c h a r i d e s w i t h C h r o m o p h o r e
Solvent systems. T h e s y s t e m s u s e d f o r t h i s g r o u p a r e a s f o l l o w s : A: n - P r o p a n o l - w a t e r (40:60) 1°8 B: Methyl ethyl k e t o n e - n - b u t a n o l - w a t e r (30: 5: 65) 1°3 C: M e t h a n o l - w a t e r (80: 20) plus 1.5% of NaC1, on paper buffered with a solution ~o3V. Betina, J. Chromatogr. 15, 379 (1964),
[7l
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
D: E: F: G: H: I: J: K: L: M: N:
127
containing 0.95 M Na~S04 plus 50 mM NariS04, equilibration 3 hr, development 16 h TM n-Butanol-pyridine-n-propanol-acetic ucid-water (20: 10 : 5: 3: 32) ~os 90% phenol-m-cresol-pyridine-acetic acid-water (25: 25:1 : 1 : 25) TM Diethyl ether-ethyl acetate (170: 30) ~0e Isopentenyl acetate-acetone (19 : 1) 10e Butyl acetate-pyridine-water (1 : 5 : 10) 107 Benzene-acetic acid-water (20: 25: 5) ~o8 Benzene-butanol-water (18: 2: 20) los Chloroform-carbon tetrachloride (both satd. with water)-methanol (5 : 4 : 1) ~os Diisoamyl ether satd. with water-butanol (20: 10) 1°8 n-Butanol-dichloroethane-formamide-water (5:45:16:34), lower layer on paper impregnated with upper layer (see Blinov and Khokhlov, 7 p. 126) Methanol-benzene (4:6) 1°9
Detection. U V l i g h t a n d b i o a u t o g r a p h y w i t h S. aureus, B. subtilis, or Corynebacterium xerosis a r e used for a n t i b i o t i c s of this group. Applications. B o t h r i s t o c e t i n a n d v a n c o m y c i n were s e p a r a t e d into two c o m p o n e n t s in s y s t e m s A - C . 1°3,1°4 I n s y s t e m D , r i s t o c e t i n B m o v e s f a s t e r t h a n r i s t o c e t i n A. T h e s a m e s y s t e m can be used for t h e c o m p a r i s o n of ristocetins, r i s t o m y c i n s , a n d v a n c o m y c i n . R i s t o m y c i n is a m i x t u r e of 4 components, ristomycin III and IV being identical with ristocetin A and B, r e s p e c t i v e l y . 1°5 A c t i n o i d i n can be s e p a r a t e d into 6 b i o l o g i c a l l y a c t i v e c o m p o n e n t s w h e n s y s t e m D is used. ~1° C h r o m o m y c i n s a r e c h a r a c t e r i z e d in F, G, a n d f u r t h e r s y s t e m s b y M i z u n o , 1°6 who also s t u d i e d c h r o m o m y c i n A3 d e r i v a t i v e s using s y s t e m H. 1°7 " S u m m a r i z e d c h r o m a t o g r a m " of a b u r a m y c i n in 8 s o l v e n t s y s t e m s has been p u b l i s h e d . ~°3 C h r o m o m y c i n A3, a u r e o l i c acid, a n d m i t h r a m y c i n can be c h a r a c t e r i z e d c h r o m a t o g r a p h i c a l l y in four s y s t e m s , iH O l i v o m y c i n is s e p a r a t e d into four m a i n c o m p o n e n t s b y c i r c u l a r c h r o m a t o g r a p h y in s y s t e m M (see B l i n o v a n d K h o k h l o v , : p. 126). lo~M. P. Kunstman, L. A. Mitscher, J. N. Porter, A. J. Shay, and M. A. Darken, Antimicrob. Ag. Chemother. 1968, 242 (1969). 105M. G. l~razhnikova, N. N. Lomakina, M. F. Lavrova, N. V. Tolstykh, M. S. Yurina, and L. M. Klyuyeva, Antibiotiki 8, 392 (1963). lo~K. Mizuno, J. Antibiot. Ser. B 13, 329 (1960). lot K. Mizuno, J. Antibiot. 16, 22 (1963). 10, E. V. Kruglyak, V. N. Borisova, and M. G. Brazhnikova, Antibiotiki 8, 1064 (1963). ,o~A. Aszalos, R. S. Robison, F. E. Pansy, and B. Berk, U.S. Patent 3,551,561; December 29, 1970. 1,0 E. Borowski, H. Chmara, and E. Jereczek-Morawska, Chemotherapia 12, 12 (1967). ~"D. M. Schuurmans, D. T. Duncan, and B. H. Olson, Cancer Res. 24(1), 83 (1964).
128
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
D. Everninomicin and L i n c o m y c i n G r o u p Solvent Systems. The systems used for these antibiotics are as follows. A: Benzene--petroleum ether (bp 30-60°)-acetone (20: 5: 10), descending development, 2 hr m B: Toluene--n-butanol-water-petroleum ether (bp 30-60 °) (20:1.5:7:1.5), descending, 4.5113 C: Petroleum ether (60-90°)-methanol-ethyl acetate-water (5: 8: 5: 2), 2 hr 113 D: Benzene-chloroform (93:7) satd. with formamide on paper dipped, prior to use, in 25% methanolic formamide, blotted, and air dried for 5 rain to remove methanol114 E: Chloroform-benzene (7: 3) on paper satd. with formamide as stationary phase l~s F: n-Butanol-water (84: 16) 116 G: n-Butanol-acetic acid-water (2:1 : 1) 11~ H: 5% Citric acid in water made to pH 7.0-7.5 with ammonia 11s I: 2-Propanol-2 N HC1 (65:35) 11e Systems A - C are for everninomicins and curamicin, D for halomicins, F for avilamicin, exfoliatin, and curamicin. Systems E - I are for the lincomycin group. However, thin-layer chromatography is mostly used to characterize members of ~he latter group. Paper chromatographic data of the everninomicin group are presented in Table IX. In bioautography, S. aureus and B. subtilis (avilamicin, exfoliatin) are used. With lincomycin, NaIO4-KMnO~ reagent m a y be applied: ~6 Spray the developed chromatograms with a mixture of 2% aqueous N a I 0 4 - 1 % K M n 0 4 in 2% aqueous Na.,C03 (2:1). Yellow spots appear on a purple background. E. Other Sugar Derivatives Diumyeins can be separated with n-propanol-n-butanol-0.5 N ammonium hydroxide ( 2 : 3 : 4 ) ; resulting Rf values for diumycin A and B are 0.32 and 0.45, respectively. 117 More recently, the same system differentiated 4 components with the following Rf values: 11s diumycin A, 0.32; diumycin A', 0.35 ; diumycin B, 0.45; diumycin B', 0.52. m M. J. Weinstein, G. M. Luedemann, E. M. Oden and G. H. Wagman, Antimicrob. Ag. Chemother. 1964, 24 (1965). m G. M. Luedemann and M. J. Weinstein, U.S. Patent 3,499,078; March 3, 1970. m M. J. Weinstein, G. M. Luedemann, E. M. Oden, and G. H. Wagman, Antimicrob. Ag. Chemother. 1967, 435 (1968). ~5 F. Buzzetti, F. Eisenberg, H. N. Grant, W. Keller-Schierlein, W. Voser, and H. Z~hner, Experientia 24, 320 (1968). ~ R. Thomas, G. J. Ikeda, and H. Harpootlian, J. Pharm. Sci. 56, 862 (1967). m E. Meyers, D. S. Slusarchyk, J. L. Bouchard, and F. L. Weisenborn, J. Antibiot. 22, 490 (1969). 1~8W. A. Slusarchyk, J. L. Bouchard-Ewing, and F. L. Weisenborn, J. Antibiot. 26, 39 (1973).
[7l
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
129
TABLE IX PAPER CHROMATOGRAPHIC DATA OF EVERNINOMICIN-TYPE ANTIBIOTICS RI X 100 in systems a Antibiotics Everninomlcln Everninomlcm Everninomicin Everninomicin Everninomicin Everninomlcm HaloInicin A c Halomicin B c Halomicin C c Halomicin D c Avilamicin d Exfoliatin ~ Curamicin ~
i b B
b
Cb Db Eb Fb
A
B
C
0 0
00
0 0
- -
- -
14 28 64 74 74
19 42 61 71 71
08 26 60 82 82
------
------
- -
- -
- -
00
- -
- -
- -
- -
35
- -
- -
- -
- -
60
- -
75
-70 58 60
-. . .
-. . .
. . .
-. . .
D
E
a See solvent systems for everninomicin and lincomycin group on p. 128. b F r o m G. M. L u e d e m a n n and M. J. Weinstein, U.S. P a t e n t 3,499,078; M a r c h 3, 1970. Cited by W a g m a n a n d Weinstein. e c M. J. Weinstein, G. M. Luedemann, E. M. Oden, and G. H. Wagman, Antimicrob. Ag. Chemother. 1967, 435 (1968). F. Buzzetti, F. Eisenberg, H. N. Grant, W. Keller-Schierlein, W. Voser, and H. Z~hner, Experientia 24, 320 (1968). e G. H. W a g m a n and M. J. Weinstein, " C h r o m a t o g r a p h y of Antibiotics," p. 63. Elsevier, Amsterdam, 1973.
VI. Macrocyclic
A. Nonpolyene
Lactone
Antibiotics
Glycosidic Macrolides
T h i s f a m i l y c o m p r i s e s t e n s of a n t i b a c t e r i a l a n t i b i o t i c s f r o m a c t i n o m y c e t e s w h i c h h a v e b e e n i n t e n s i v e l y s t u d i e d b y m e a n s of p a p e r chromatography. Filtrates from fermentations are applied directly onto chromatograms while pure samples are dissolved in methanol or ethanol. Bioautography w i t h B. subtilis o r o t h e r g r a m - p o s i t i v e b a c t e r i a is t h e common detection method. Solvent Systems. T h e s y s t e m s f o r s e p a r a t i o n , c h a r a c t e r i z a t i o n , a n d systematic analysis of this family are listed below. Of these, A-H are m o d i f i c a t i o n s of Z a f f a r o n i ' s s y s t e m s . S i n c e a n t i b i o t i c s o f t h i s f a m i l y a r e m o s t l y of b a s i c c h a r a c t e r , s y s t e m s T - W r e p r e s e n t m o d i f i c a t i o n s o f "pH-
130
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
c h r o m a t o g r a p h y ' ' H g - ~ which helps to s e p a r a t e m i x t u r e s of acidic, s6 basic, or a m p h o t e r i c a n t i b i o t i c s b y selecting o p t i m a l p H v a l u e s of the s t a t i o n a r y p h a s e (see Section X I I I , B on p H c h r o m a t o g r a m s ) . Systems m a r k e d * were on p a p e r p r e t r e a t e d with f o r m a m i d e ; **, on p a p e r p r e t r e a t e d with 50% f o r m a m i d e i n m e t h a n o l , s o l v e n t s y s t e m s satd. with f o r m a m i d e . A: B: C: D: E: F: G: H: I: J: K: L: M: N: O: P: Q: R: S: T: U: V: W: X: Y: Z:
Benzene-cyclohexane (1 : 1)*'122 Benzene*'1~2 Benzene-chloroform (3 : 1)*'122 Benzene-chloroform (1: 1) *'122 Benzene-cyclohexane (1 : 1)**.123 Benzene-cyclohexane (2:1)**.123 Benzene--chloroform (3: 1) **'1~3 Benzene-chloroform (1: 1) **'128 5% Aq. ammonium chloride TM Benzene--methanol (4: 1) 124 1 M Phosphate buffer, pH 7.0 TM 0.05 N (0.175%) ammonia satd. with methyl isobutyl ketone TM 1% Aq. ammonia TM 1% Ammonia satd. with methyl isobutyl ketone TM Methanol-acetone-water (19: 6: 75) 125 80% Methanol-3% NaC1 (1:1) on paper buffered with 0.95 M Na sulfate plus 0.5 M Na bisulfate, descending12~ Propanol-pyridine-acetic acid-water (6: 4:1 : 3) ascending128 Butanol-acetic acid-water (4: 1 : 5), ascending~28 80% Phenol, ascending12~ Benzene on paper impregnated with citrate buffer of pH 4.61~7 Butyl acetate on paper impregnated with citrate buffer of pH 4.0 ~7 Isopropyl ether-methylisobutyl ketone-2 % aq. ammonium carbonate (2: 1 : 2) ~2s Methanol-benzene (20:70) buffered with diisopropylamine-acetic acid ~22 Methanol-benzene (20: 70) buffered with pyridine-acetic acid l~g Methanol-dichloroethane (10: 80) buffered with pyridine--acetic acid 122 Methanol-benzene (20: 70) buffered with pyridine-oxalic acid 129
11,V. Betina, Nature (London) 182, 796 (1958). 1~ V. Betina and P. Nemec, Nature (London) 187, 1111 (1960). 12~M. IglSy, Magyar KJm. Lap]a 18, 622 (1963). 1~ A. Zaffaroni, R. B. Burton, and E. H. Keutmann, Science 111, 6 (1950). l~aT. M. Lees, J. De Muria, and W. H. Boegemann, J. Chromatogr. 5, 126 (1961). 1~4H. Koshiyama, M. Okanishi, T. Ohmori, T. Miyaki, H. Tsukiura, M. Matsuzaki, and tt. Kawaguchi, J. Antibiot. 16, 59 (1963). 1~5C. W. Pettinga, W. M. Starke, and F. R. Van Abeele, J. Amer. Chem. Soc. 76, 569 (1954). 1~ M. J. Weinstein, G. M. Luedemann, G. It. Wagman, and J. A. Marquez, U.S Patent 3,632,750; January 4, 1972. 1~ T. Osato, K. Yagashita, and H. Umezawa, J. Antibiot. 8, 161 (1955). 1~ H. A. Whaley, E. L. Patterson, A. C. Dornbush, E. J. Backus, and N. Bohonos, Antimicrob. Ag. Chemother. 1963, p. 45 (1964). ~2~M. IglSy, A. Mizsei, and I. I-Iorv~th, J. Chromatogr. 20, 295 (1965).
[7]
PAPER CHROMATOGRAPHY OF ANTIBIOTICS
131
Five buffer mixtures for systems W to Z are prepared as shown in accompanying tabulation:
Moles per 100 ml solvent Mixtures
Acid
Base
1 2 3 4 5
0.018 0. 018 0.018 0.009 Nil
Nil 0. 009 0.018 0~018 0.018
A p p l i c a t i o n s . P a p e r chromatographic data of 26 macrotides are compiled in T a b l e X. In system O, samples from erythromycin fermentation show two bioactive spots, one of them belonging to erythromycin B which can be isolated by the use of cellulose column chromatography. 12~ A further spot, detected in the course of the isolation of erythromycin B, helped to discover erythromycin C. 13° System 0 helped in finding two components in a commercial preparation of erythromycin (Ilotycin). T M Ochab et al. 132 described a technique for quantitative analysis of erythromycin in the presence of tetracycline in drugs. IglSy et al. ~29 used their modification of p H chromatography in order to compare erythromycin, magnamycin, oleandomycin, picromycin, and methymycin. On paper chromatography, spiramycins proved to be identical with foromacidins A, B, and C, but foromacidin D differed from the three spiramycins. 133 System E was applied in quantitative analysis of triacetyl oleandomycin. T M P a p e r c h r o m a t o g r a p h y helped to separate m e t h y m y c i n from neom e t h y m y c i n ~'~ and to prove the identity of a m a r o m y c i n and albomycetin with pikromycin. ~36 N i d d a m y c i n was compared with tylosin, carbomycin
,~0p. F. Wiley, R. Gale, C. W. Pettinga, and K. Gerzon, J. Amer. Chem. Soc.79, 6074 (1957) "~ V. Betina, pH chromatography of Antibiotics (in Slovak), p. 106. Thesis, Slovak Acad. Sci., Bratislava, 1960. 13~S. Ochab, D. Malysz, and B. Borowiecka, Chem. Anal. (Warsaw) 8, 597 (1963). I~'~R. Corbaz, L. Ettlinger, E. G~iumann, W. Keller-Schierlein, F. Kradolfer, E. Kyburz, L. Neipp, V. Prelog, A. Wettstein, and H. Z~ihner, Helv. Chim. Acta 39, 3O4 (1956). 1~ It. J. Pazdera, personal communication, 1962. 1~ D. Perlman and .E. O'Brien, Antibiot. Chemother. 4, 894 (1954). "eR. Hfitter, W. Keller-Schierlein, and H. Z~ihner, Arch. Mikrobiol. 39, 158 (1961).
132
METHODS
©
FOR
IIl[I
Z
F~
THE
STUDY
OF
[71
ANTIBIOTICS
I
I
I
II
I
r
I
I/
I
J
I
i
I
I
I
I
i
I
F~
0
I
I
I[[I [I
I
I
[I
I
I
11
I
I
~o
0
r~ e~ 0 ©
[7l
PAPER
~o~
~'~
~
,~
CHROMATOGRAPHY
,l~
~ .~.
~
~
~
o~
~
~ •
. ~
o
.~
~
.~.
~
OF
ANTIBIOTICS
133
134
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
A, and carbomycin B. 13~ Blinov et al. 1"8 found that some macrolides and other antibiotics gave double spots on chromatograms even when they were not mixtures of two compounds. Acumycin can be differentiated from 9 macrolides in systems A to C and also by thin-layer chromatography. 139 In structural studies of angolamycin, 149 amino sugars were determined chromatographically in hydrolyzates. Mycaminose was identified by paper chromatography in hydrolyzates of acumycin while a neutral deoxysugar was characterized by thin-layer chromatography. 130 Similar spots are detected in hydrolyzates of miyamycin and erythromycin, m Leucomycins A and B, related to spiramycins, can also be studied by means of paper chromatography. ~4= Aminosugars with similar RI values can be detected in hydrolyzates of leucomycins A~, A2, B1, Ba, B4, and .143
B. Polyene Macrolides Polyene antibiotics are applied in methanolic solutions onto chromatograms which can be then detected bioautographically with Saccharom y c e s cerevisiae, Candida a~bicans, or other fungi and by their fluorescence in UV light. S o l v e n t S y s t e m s . The systems used in studies of polyenes are as follows. A: B: C: D: E: F:
Butanol-acetic acid-water (4:1 : 5) 144 5% Dimethylformamide in methanoP** Butanol-ethanol-water (5: 1: 4) 14s Propanol-water (7: 3) 145 As D, but 8:2145 n-Butanol satd. with water 146
137G. Huber, K. H. Wallhgusser, L. Fries, A. Steigler, and H. L. Weidenmiiller, Arzneim.-Forsch. 12, 1,191 (1962). 138N. 0. Blinov, E. F. Oparysheva, I. N. Trudnikova, T. M. Rozanova, and A. S. Khokhlov, Antibioti£i 6, 660 (1961). is, H. Bickel, E. Giiumann, R. Hiitter, W. Sackmann, E. Vischer, W. Voser, A. Weftstein, and H. Z~hner, Helv. Chim. Acta 45, 1396 (1962). 140R. Corbaz, L. Ettlinger, E. G~umann, W. Keller-Schierlein, L. Neipp, V. Prelog, P. Reusser, and H. Z~hner, Helv. Chim. Acta 38, 202 (1955). 141H. Schmitz, M. Misiek, B. Heinemann, J. Lein, and I. R. Hooper, Antibiot. Chemother. 7, 37 (1957). 1,, T. Hata, F. Koga, and It. Kanamori, J. Antibiot. 6, 110 (1953). 1,, T. Watanabe, Bull. Soc. Chem. Soc. Jap. 34, 15 (1961). I~A. Aszalos, R. S. Robison, P. Lemanski, and B. Berk, J. Antibioti. O.l, 611 (1968). i~ A. P. Struyk, I. ttoette, G. Drost, J. M. Waisviz, T. van Eck, and J. C. Hoogerheide, Antibiot. Annu. 1957/1958, p. 878 (1958). I~K. Dornberg, R. Fugner, G. Bradler, and H. Thrum, J. Antibiot. 24. 172 (1971).
[7]
PAPER CHROMATOGRAPHY
OF ANTIBIOTICS
135
G: 20% Ammonium chloride 14~ H: As G, but 3% 14~ I: 75% Aq. phenol 14e J: 50% Aq. acetone ~4~ K : n-Butanol-methanol-water-methyl orange (40 ml : 10 ml : 20 ml : 1.5 g) ~4~ L: n-Butanol-methanol-water (40: 10: 20) 14~ M : Benzene-methanol (80: 20) 14~ N: Distilled water ~4e O: n-Butanol-pyridine-water (60: 40: 30) ~4~ P: Dimethyl formamide-water (10 : 90) ~46 Q: As P, but 50:5014~ R: As C, but 5:1:5147 S: t-Butanol-water (4: 1) 14s T: n-Butanol-ethyl acetate (1: 1) satd. with water ~4s U: Methy! isobutyl ketone-n-butanol-water (50:15 : 3) ~49 V: Methyl isopropyl ketone-water (satd. ca 1.6%) 149 W: Methyl isopropyl ketone-n-butanol-water (50: 10: 3.5) ~49 X : Methyl isopropyl ketone-methanol-water (50:1 : 0.4) ~0 Y: Ethyl acetate-pyridine-water (4: 3 : 2) 1~0 Z: As O, but 6:4:51~° : Methanol-25 % ammonium hydroxide-water (20:1 : 4) ~0 f~: As A, but 20:1:25 TM 7: As Y, but 6:2.5:7 TM : Chloroform-tetrahydrofuran-formamide (50: 50: 5) ~2 ~: n-Butanol satd. with 0.2% aq. acetic acid ~53 T h e a b o v e s y s t e m s are used for p a p e r c h r o m a t o g r a p h y of different groups of p o l y e n e s as follows. T r i e n e s : s y s t e m s A a n d B for trienine. T e t r a e n e s : s y s t e m A - - f o r t e r r a m y c i n , t e t r i n A, a n d t e t r i n B ; C - - f o r pimaricin; D--for pimaricin and tetramyein; F--for endomycins, flavacid, n y s t a t i n , p i m a r i c i n , p o l y f u n g i n s , r i m o c i d i n , t e r r a m y c i n , a n d unamycin; J--for nystatin, tetramycin, and unamycin; R--for amphot e r i c i n A, a n t i m y c o i n A, c h r o m i n , a n d n y s t a t i n ; S - - f o r e n d o m y e i n s and f l a v a c i d ; T - - f o r e n d o m y c i n s ; s y s t e m s G, H, I, K, L, M , N, O, P, a n d Q for t e t r a m y c i n . P e n t a e n e s : s y s t e m L - - f o r e u r o c i d i n ; E a n d F - - f o r m o l d i e i n B ; syst e m s U, V, W , a n d X - - - f o r f u n g i c h r o m i n . 14~C. P. Schaffner, I. D. Steinman, R. S. Safferman, and H. Lechevalier, Antibiot. Antra. 1957/1958, p. 869 (1958). 14SL. C. Vining and W. A. Taber, Can. J. Chem. 35, 1461 (1957). '*~A. C. Cope, R. K. Bly, E. P. Burrows, O. J. Ceder, E. Ciganek, B. T. Gillis, R. F. Porter, and H. E. Johnson, J. Amer. Chem. Soc. 84, 2170 (1962). 1~,R. Bosshardt and H. Bickel, Experientia 24, 422 (1968). 1~'G. R. Deshpande and N. Narasimhachari, Hindustan Antibiot. Bull. 9, 76 (1966). 1~2R. Schlegel and H. Thrum, J. Antibiot. 24, 360 (1971). 15~H. Lechevalier, R. F. Acker, C. T. Corcke, C. M. Haensler, and S. A. Waksman, Mycologia 45, 155 (1953).
136
METHODS FOR THE STUDY OF ANTIBIOTICS
[7]
H e p t a e n e s : system E - - f o r amphotericin B and h a m y c i n ; F and J - - f o r h a m y c i n ; L - - f o r aureofungin and h a m y c i n ; Y - - f o r candicidin; Z - - f o r candicidin, h a m y c i n , and p e r i m y c i n ; a - - f o r amphotericin B, ascosin, azacolutin, h a m y c i n , and levorin; f l - - f o r ascosin, candidin, and c h a m p a m y c i n ; e - - f o r candicidins. Octaenes: systems A, O, and Z for ochramycin. Others: systems L, Q, and Z for flavofungin and fungichromin.
C. O t h e r M a c r o c y c l i c A n t i b i o t i c s This section includes aglycosidic n o n p o l y e n e macrolides (e.g., cyanein), r i f a m y c i n s and related compounds, antimycins, azalomycins, etc. S o l v e n t S y s t e m s . Systems used for these antibiotics are as follows. A: Phosphate buffer, pH 7.3, containing 0.1% Na ascorbate satd. with n-amy! alcohol-n-butanol (9: 1) TM B: n-Butanol satd. with phosphate buffer, pH 7.3, containing 0.1% Na ascorbate TM C : Phosphate buffer, pH 8.6, containing 0.1% Na ascorbate satd. with n-butanol TM D: Water containing 3 % ammonium chloride plus 1% ascorbic acid 15~ E: M/15 Phosphate buffer pH 8.6 with or without addition of Na ascorbate on paper impregnated with 2-octyl alcohol 15e F: Cyclohexane-chloroform-water (1 : 8: 2) on Schleicher and Schuell No. 589 (Blue Ribbon Special) impregnated with 0.2 M phosphate buffer at pH 4.1 and air dried. Equilibrate for 2 hr in atmosphere of both phases and develop in organic phase 15~ G: n-Hexane-benzene-acetone-water (30: 10: 18: 32) ~ss H: As G, but (1:3:1:3) 1~9 I: 0.075 N ammonium hydroxide satd. with methyl isobutyl ketone 16° J: Water-ethanol-acetone (7: 2: 1) 16' K: Ethanol-n-hexane (1 : 2) le2 L: Water-acetone (75:25) 183 M: 16% Aq. n-propano1182 u4 p. Sensi, C. Coronelli, and B. J. R. Nicolaus, J. Chromatogr. 5, 519 (1961). ~5 p. Sensi, A. M. Greco, and R. Ballotta, Antibiot. Annu. 1959/1960, p. 262 (1960). 1~S. Sferruzza and R. Rangone, Farmaco, Ed. Prat. 19, 486 (1964). u7 p. Siminoff, R. M. Smith, W. T. Sokolski, and G. M. Savage, Amer. Rev. Tuberc. 75, 576 (1957). 158T. Kishi, H. Yamada, M. Muroi, and K. Mizuno, 163rd Meet. Jap. Antibiot. Res. Ass., September, 1968. 15oT. Kishi, H. Yamada, M. Muroi, S. Harada, M. Asai, T. Hasegawa, and K. Mizuno, J. Antibiot. 25, 11 (1972). ~ C. DeBoer, P. A. Meulman, R. J. Wnuk, and D. H. Peterson, J. Antibiot. 23, 442 (1970). 161D. Kluepfel, S. N. Seghal, and C. Vezina, J. Antibiot. 23, 75 (1970). ~ Y. Sakagami, A. Ueda, S. Yamabayashi, Y. Tsurumaki, and S. Kumon, J. Antibiot. 22, 521 (1969).
[7]
PAPER CHROMATOGRAPHYOF ANTIBIOTICS
137
N: Water-ethanol-acetic acid (70:24:6), 18-20 hr at 30° on Eaton-Dikeman No. 613183 O: Benzene-ethyl acetate (1: 2) TM P: 40% Aq. methanol 1~5 Q: As M, but 20% 1~5 R: n-Butanol-benzene-5 % ammonium chloride (l : 9: 10) 166 S: n-Butanol satd. with water 167 T: n-Butanol-methanol-water (4: 1:5), bottom layer 167
Detection. Methods v a r y greatly for these antibiotics. Rifamycins, besides their original color, are detected with Sarcina lutea, streptovaricins with Mycobacterium ranae, tolypomycin Y with Staphylococcus aureus, geldanamycin with Tetrahymena pyriformis, 1~° antimycins with Saccharomyces cerevisiae, oligomycins with Glomerella cingulata, ~63 aabomycin A and ikutamycin with Piricularia oryzae, 164,165 azalomyein B with S. lutea or Mycobacterium phlei, ~66 cyanein with Candida pseudotropicalis. Applications. Systems A through E are recommended for rifamycins, F for streptovaricins, G and H for tolypomycins, I for geldanamycin, J for antimycins, K and L for hondamycins, N for oligomycins, O for aabomycin, P and Q for folimycin and ikutamycin, R for azalomycin B, S, and T for cyanein (brefeldin A). Chromatographic data of rifamycins and other large macrocyclic antibiotics are summarized in Table XI, and those of other macrolactones in Table X I I . Quantitative paper chromatographic analysis of streptovaricins has been described. 1~8
V I I . Quinone Antibiotics
A. Tetracyclines Pure substances are applied on chromatograms in methanol. Samples from fermentations must be acidified in order to liberate tetracyclines from the mycelium. One such procedure has been described. 169 P h a r m a ceutical aqueous suspensions are acidified with HC1 and diluted with le,~M. H. Larson and W. H. Peterson, Appl. Microbiol. 8, 182 (1960). 1~4German "Offenlegungsschrift" 1,961,746 (1970). Cited by Wagman and Weinstein.~ 1,5y. Sakagami, A. Ueda, and S. Yamabayashi, J. Antibiot. 20, 299 (1967). '~ M. Arai. J. Antibiot. 13, 51 (1960). 1~V. Betina, P. Nemec, J. Dobias, and Z. Bar/~th, Folia Microbiol. (Prague) 8, 353 (1962). 1~W. T. Sokolski, N. J. Eilers, and P. Siminoff, Antibiot. Annu. 1957/1958, 119 (1958). 169j. Vondr£Skov/~and O. ~trauchovh, J. Chromatogr. 32, 780 (1968).
138
METHODS
FOR
THE
STUDY
OF
[71
ANTIBIOTICS
0 0~ ¢0
0,1
v
. e~
.
Z
r.,.l
........
Z
z o
e,
~'~'~~ .~.~.~ ~ oooo=
.< mmmm
~
~b ~,l ~
~
~-~
,
~ ~ l l l
0
6
0 0
~-~.~=..~'~ ~o
.~.~.~.~.$.B.~.~
Tr*
(i)
(2}
(I) path will be experienced, so *L -- cR and nL ---- na and no optical activity accompanies this transition. However, unsymmetrical conformation or substitution (II) will lead to a definite preference, and optical activity
5 ~ 1
R23 2
(3)(CHAIR)
S,
2
hY
nL f/TR; ~L# ER
6 (5)
(II)
R (3)(TWIST)
[17]
SPECTROPOLARIMETRY OF ANTIBIOTICS
351
results. The carbon bearing the R group (R =/= H) in formulas (3) and (4) (II) has four different substituents and is a classical asymmetric center. The dependence of chiroptical phenomena on molecular conformation and molecular asymmetry can be readily assimilated from such considerations. This grossly oversimplified exposition documents another important rule. The ORD and CD curve of a given substance encodes the same information regarding molecular asymmetry and the methods are, therefore , complimentary. Specific examples to be given shortly will help the reader comprehend the practical and strategic considerations which may lead the investigator to choose one or the other of these methods as most appropriate in a given case. Table I lists the four classes of chromophores recognized in chiroptical studies and some examples of each type encountered in the study of antibiotics. Crabb5 (pp. 185-191) ~ has an excellent tabular summary of the Cotton effects generated by a wide variety of chromophores and the rules for curve rationalization. Table II lists the families of antibiotics which have been studied spectropolarimetrically and the purposes of the particular study. The literaTABLE I TYPES OF CHROMOPHORES O1~" INTEREST IN CHIROPTICAL STUDIES OF ANTIBIOTICS
Type Symmetrical chromophore in an unsymmetrical molecular environment Inherently disymmetric chromophore due to lack of symmetry of the chromophore, or molecular twisting Homoconjugated. Chromophores interact through space rather than through the molecular framework. Exciton coupled
Extrinsic. Optical activity induced in a symmetrical molecule by binding to an optically active biopolymer (I)NA, RNA, protein). A special subcase exists when the bound molecule also has optical activity in its own right
Examples Ketone n -~ ~r* (erythromycin, fusidic acid, Actidione) benzene ring (chloramphenicol) a,B-Unsaturated ketones (methymycin) conjugated dienes (Leucomycin A3) conjugated ene-one system (tetracyclines) Peptides in helical arrays or otherwise conformationally fixed at angles ~180 ° (tyrocidin S, bacitracin, stendomycin). Other chromophores of similar energy content fixed at angles ~ 180 ° relative to one another (chromomycin, anhydrotetracyclincs, chelocardin) (Actinomycin -I- I)NA, bleomycin + I)NA, etc.)
352
METHODS FOR THE STUDY OF ANTIBIOTICS
[17]
TABLE II CttIROPTICAL STUDIES OF ANTIBIOTICS
Antibiotic a n d / o r family Actinomycin Chlorampheaicol Fusidic acid Glutarimide family ~3-Lactam antibiotics
Macrolide family Peptide family
Tetracycline family Miscellaneous types Azoxyketones Streptothricins Rhodomycins Antimycins Chromomycins Bluensomycin Griseofulvin X-537A I n t e r a c t i o n of antibiotics with biopolymers Bleomycin types Actinomycins Kanchanomycin Netropsins Sulfonamides
Purpose of s t u d y Conformational analysis and solvent interactions a-d Conformational analysis, °-g assignment of absolute configuration,/' g a n d interaction with chelating ionsg Absolute configuration h-i Absolute configuration k-° Investigation of the electronics of the chromophore ~-t a n d q u a n t i t a t i v e analysis ~ Conformational analysis r Conformational analysis, " - ' ' ~' structure d e t e r m i n a t i o n , " ~'' b, solvent effects . . . . . ' Conformational analysis, d'-~' solvent effects,X, ,'. a'-k'. ~,-o,, ~, interaction with chelating ions, k', g" "-~' solvent effects, d'' ,,,g,-k,, ,,,-o,, ~, a n d absolute configuration-structure analysis °''*'' ~' Conformational analysis, ~ ' - ' ' ' ~"' ~ structure studies, ~ ion binding, ~'-'',¢'. ~a q u a n t i t a t i v e analysis ~* Absolute configuration, dd' *o solvent effects, dd' °° conformational analysis d~' ** N a t u r e of the chromophore H Assignment of absolute stereochemistryga Assignment of absolute stereochemistry hh Assignment of absolute stereocheinistry,"' Ji n a t u r e of thc chromophore~, ii Assignment of absolute stereochemistry, kk ion binding kk Structural analysis kk I o n binding u
W i t h D N A =m W i t h D N A . . . . . . pp DNAqq DNA~r, u Plasma albumin ~u
H. Ziffer, K. Yamaoka, a n d A. B. Mauger, Biochemistry 7, 996 (1968). b F, Ascoli, P. DeSantis, M. Lener, and M. Savino, Biopolymers 11, 1173 (1972). D. M. Crothers, S. L. Sabol, D. I. Ratner, and W. Mueller, Biochemistry 7, 1817 (1968). d F. Ascoli, P. DeSantis, a n d M. Savino, Nature (London) 227, 1237 (1970). • L. A. Mitscher, F. Kautz, a n d J. Lapidus, Can. J. Chem. 47, 1957 (1969). I L. A. Mitscher, P. W. Howison, J. B. Lapidus, and T. D. Sokoloski, J. Med. Chem. 16, 93 (1972). o L. A. Mitscher, P. W. Howison, a n d T. D. Sokoloski, J. Med. Chem. 16, 98 (1972). h R. Bucort, M. Legrand, M. Vignau, J. Tessier, a n d V. Delaroff, C. R. Acad. Sci. 257, 2679 (1963).
[17]
SPECTROPOLARIMETRY OF ANTIBIOTICS
353
R. Bucort and M. Legrand, C. R. Acad. Sci. 258, 3491 (1964). i W. O. Gotfredsen, W. V. Daehne, S. Vangedal, A. Marquet, I). Arigoni, and A. Malera, Tetrahedron 9.1, 3505 (1965). k F. Johnson, L. G. Duquette, and H. E. Hennis, J. Org. Chem. 33, 904 (1968). l T. Okuda, M. Suzuki, and Y. Egawa, Chem. Pharm. Bull. 8, 335 (1960). '~ T. Okuda and M. Suzuki, Chem. Pharm. Bull. 9, 1014 (1961). n M. Suzuki, Y. Egawa, and T. Okuda, Chem. Pharm. Bull. 11, 582 (1963). ° T. Okuda, Chem. Pharm. Bull. 7, 259 (1959). 7, [. Z. Siemion, J. Lisowski, B. Tyran, and J. Morawiec, Bull. Acad. Po/. Sci., Ser. Sci. Chim. 20, 549 (1972). q R. Nagarajan and D. 0. Spry, J. Amcr. Chem. Soc. 93, 2310 (1971). S. Kukolja, P. V. Demarco, N. D. Jones, M. O. Chaney, and J. W. Paschal, J. Amer. Chem. Soc. 94, 7592 (1972). L. Neelakanthan and D. W. Urry, Abstr. 158th Meet. Amer. Chem. Soc. 1969, B10L176. t F. Snatzke, Bull. Acad. Pol. Sci. Set. Sci. Chim. 21, 167 (1973). " C. E. Rasmussen and T. Higuchi, J. Pharm. Sci. 60, 1608 (1971). •' T. J. Perun, R. S. Egan, P. H. Jones, J. R. Martin, L. A. Mitscher, and B. J. Slater, Antimicrob. Ag. Chemother. 116 (1969). w S. Omura, A. Nakagawa, N. Yagisawa, Y. Suzuki, and T. Hata, Tetrahedron, 28, 2839 (1972). R. S. Egan, T. J. Perun, J. R. Martin, and L. A. Mitscher, Tetrahedron 29, 2525 (1973). L. A. Mitscher, B. J. Slater, T. J. Perun, P. H. Jones, and J. R. Martin, Tetrahedron Lett. 4505 (1969). J. R. Martin, T. J. Perun, and R. S. Egan, Tetrahedron 28, 2937 (1972). "' C. Djerassi and O. Halpern, Tetrahedron 3, 255 (1958). b, p. Kurath, J. R. Martin, J. Tadanier, A. W. Goldstein, R. S. Egan, and ]). A. Dunnigan, Helv. Chim. Acta 56, 1557 (1973). ~' C. Djerassi and J. A. Zderic, J. Amer. Chem. Soc. 78, 6390 (1957). d' l). W. Urry and A. Ruiter, Biochem. Biophys. Res. Commun. 38, 800 (1970) ~" S. Laiken, M. Printz, and L. C. Craig, J. Biol. Chem. 244, 4454 (1969). /' P. M. Bayley, Biochem. J. 125, 90 (1971). ~' B. E. Isbell, C. Rice-Evans, and G. It. Beaven, F E B S Lclt., 192 (1972). h' D. W. Urry, Biochemistry 11, 487 (1972). i, D. W. Urry, A. L. Ruiter, B. C. Starcher, and T. A. tiinners, Antimicrob. Ag. Chemother., p. 87 (1968). i' F. Quadrifoglio and D. W. Urry, Biochem. Biophys. Rcs. Commun. 29, 785 (1967). k, T. Funk, F. Eggers, and E. Grell, Proc. Ear. Biophys. Congr., 1st, Vol. 3, 37 (1971). i, M. A. Ruttenberg, T. P. King, and L. C. Craig, Biochemistry 5, 2857 (1966). '~' M. A. Ruttenberg, T. P. King, and L. C. Craig, J. Amer. Chem. Soc. 87, 4196 (1965). "' M. Bodarlszky and A. Bodaaszky, Nature (London) 220, 73 (1968). o' Yu. A. Orchinuikov, V. T. Ivanov, B. F. Bystrov, A. I. Miroshnikov, E. N Shepel, N. D. Abdullaev, E. F. Efremov, and L. B. Senyavina, Biochem. Biophys. Res. Commun. 39, 217 (1970). ~" D. Balasubramaniau, J. Amer. Chem. Soc. 89, 5445 (1967).
354
METHODS FOR THE STUDY OF ANTIBIOTICS
[17]
q' V. T. Ivanov, I. A. Laine, N. D. Abdullaev, V. Z. Pletnev, G. M. Lipkind, S. F. Arkhipova, L. B. Senyavina, E. N. Meshcheryakova, and E. M. Popov, Khim. Prir. Soedin. 7, 221 (1971). "' K. A. Zykalova, G. N. Tishcheako, G. A. Kogan, aad V. T. Ivanov, Isv. Akad. Nauk. SSSR, Ser. Khim., p. 1547 (1970). ~' K. Vogler, R. O. Struder, P. Lanz, W. Lergier, E. Boehri, and B. Fust, Helv. Chim. Acta 46, 2823 (1963). t, N. W. Cornell and D. G. Guiney, Jr., Biochem. Biophys. Res. Commun. 40, 530 (1970). -' N. A. Podduknaya and N. Ya. Krasnobrizhii, Zh. Obshch. Khim. 41, 46 (1971). •' A. Bodanzsky, M. Bodanzsky, K. L. Perlman, and D. Perlman, J. Antibiot. 25, 325 (1972). ~" U. Ludescher and R. Schwyzer, Helv. Chim. Acta. 54, 1637 (1971). ~' L. A. Mitscher, B. Slater-Eng, and T. Sokoloski, Antimicrob. Ag. Chemother. 2, 66 (1972). • ' L. A. Mitscher, A. C. Bonacci, B. J. Slater, A. K. Hacker, and T. D. Sokoloski, Antimicrob. Ag. Chemother., p. 111 (1969). y' A. H. Caswell and J. D. Hutchison, Biochem. Biophys. Res. Commun. 43, 625 (1971). •' L. A. Mitscher, A. C. Bonacci, and T. D. Sokoloski, Antimicrob. Ag. Chemother., p. 78 (1968). ~a L. A. Mitscher, A. C. Bonacci, and T. D. Sokoloski, Tetrahedron Lett. 1968, 5361 (1968). b~L. A. Mitscher, W. Rosenbrook, Jr., W. W. Andres, R. S. Egan, J. Schenk, and J. V. Juvarkar, Antimicrob. Ag. Chemother., p. 38 (1970); J. Amer. Chem. Soc. 92, 6070 (1970). cc R. F. Miller, T. D. Sokoloski, L. A. Mitscher, A. C. Bonacci, and B.-A. Hoener, J. Pharm. Sci. 62, 1143 (1973). ~4 W. J. McGahren and M. P. Kunstman, J. Amer. Chem. Soc. 92, 1587 (1970). " W. J. McGahren and M. P. Kunstman, J. Org. Chem. 37, 902 (1972). ss H. Taniyama and Y. Sawada, Chem. Pharm. Bull. 20, 596 (1972). ga H. Brockmann, Jr. and M. Legrand, Tetrahedron 19, 395 (1963). hh M. Kinoshita, S. Aburaki and S. Umezawa, J. Antibiot. 25, 373 (1972). ii N. Harada and K. Nakanishi, Accounts Chem. Res. 5, 257 (1972). iJ N. Harada, K. Nakanishi, and S. Tatsuoka, J. Amer. Chem. Soc. 91, 5896 (1969). k~ C. B. Barlow and L. Anderson, J. Antibiot. 25, 281 (1972). kk, A. Brossi, M. Brumann, and F. Burdhardt, Helv. Chim. Acta. 45, 1292 (1962). zl H. Degani, H. L. Friedman, G. Navon, and E. M. Kosower, Chem. Commun., p. 431 (1973). m,~ W. C. Krueger, L. M. Pschigoda, and F. Reusser, J. Antibiot. 26, 424 (1973). ~ R. B. Homer, Arch. Biochem. Biophys. 129, 405 (1969). oo y. Courtois, W. Guschlbauer, and P. Fromageot, Eur. J. Biochem. 6, 106 (1968). ~" K. Yamaoka and H. Ziffer, Biochemistry 7, 1001 (1968). qq P. A. Friedman, T.-K. Li, and I. H. Goldberg, Biochemistry 8, 1545 (1969). ~ C. Zimmer, G. Luck, H. Thrum, and C. Pitra, Eur. J. Biochem. 26, 81 (1972). ~ C. Zimmer, K. E. Rinert, M. Thrum, U. Waehnert, and G. Loeber, J. Mol. Biol. 56, 329 (1971). u C. Zimmer and G. Luck, F E B S Lett. 10, 339 (1970). ~ G. C. Wood and S. Stewart, J. Pharm. Pharmacol. 2S, Suppl., 248s (1971).
[17]
SPECTROPOLARIMETRY OF ANTIBIOTICS
355
ture is widely scattered, and the authors apologize for inevitable accidental errors of omission.
I I . I m p o r t a n t E q u a t i o n s for D a t a R e d u c t i o n a n d Curve Comparison
ORD M o l e c u l a r r o t a t i o n = [~]~ = [a]x X M / 1 0 0 = (a X 100 X M)/l X ~:' A m p l i t u d e = a --- ([~]1 - [~b]2)/100
CD Specific ellipticity = [~I,]x = (~I, X 104)/l × c' Molecular ellipticity = [O]x = ([xI,] X M ) / 1 0 0 = 3300 [A~] = (3300 X O D X M)/c × 1 A n i s o t r o p y f a c t o r = g = Ae/e
Comparison o] ORD-CD Data a = 40.28Ae = 0.0122 [~]
Meaning o] Terms ~I, = ellipticity in degrees = scale setting x degrees ellipticity for full scale M = g r a m molecular w e i g h t of sample c = c o n c e n t r a t i o n in g / m l c' = c o n c e n t r a t i o n in g / l i t e r l = light p a t h in c m = degrees r o t a t i o n [a]× = specific r o t a t i o n at w a v e l e n g t h ~ in n m [O]t - [~]~ = the difference b e t w e e n the a m p l i t u d e s of the molecular r o t a t i o n s of the p e a k a n d t r o u g h of a g i v e n C o t t o n effect [~]x = eL - ~R O D = optical d e n s i t y difference III. Some Specific Examples
Involving
Antibiotics
T h e spectra described in this section were obtained using a J a s c o ins t r u m e n t and the r e m a r k s a p p l y specifically to d a t a obtained in this way. T h e principles a p p l y also to the C a r y instrument, but one anticipates
356
METHODS FOR THE STUDY OF ANTIBIOTICS
[17]
that there would be some small differences in operating procedure and strategy.
A. Oleandomycin (5)
Favorable Ae/~; expected response o] instrument; representation o] data; ORD/CD ]actorable. The ORD and CD spectra are illustrated CH
Ou /O'~'CH2
CH31~.I."CH3
N(CH3)2
L c.,. CH'30 ~ . , 0 " ~ CH3
0 "I
~""Z--"CH3
~--~OCH 3 u " ""~'-0 CH3 H (6)
HO CH3HN..-.~,,,,,'~ ~
-v VHo.~OH
CH5
CH3 (7)
for this substance in methanol as solvent (Fig. 2). There are two chromophoric groups in the molecule, the a-epoxyketone function absorbing at 302 nm (n--> ~*) and the lactone n--) ~r* transition at 215 nm. The Ac/c ratio of both chromophores is favorable for measurement because of their relatively high local asymmetry and relatively low total absorptivity. At both high (700 nm) and low (200 nm) wavelengths the Xe lamp output is weak and the readings are noisy. In the 700 nm region, the molecule does not absorb light, so the low lamp output simply results in greater uncertainty in absolute intensity than would otherwise be tolerated. This can be overcome substantially by increasing instrument response, by increasing the cell pathlength or the concentration of the solution. These
SPECTROPOLARIMETRY OF ANTIBIOTICS
[17]
357
H IN
/S~cH3
CH5 HN ~
/ ~
CH3
C"
/S~
.CH3
0 CH5 (10)
(9)
H~. .O...H N----.// I
H/. '~.(O
HO "" O-,,J ^ ~CH~, -3 ~ L + ".H
H
O .O O ~H"" ~H"'"
devices cannot be used to improve the readings in the 200-nm region because the sample is also absorbing light fairly strongly (±,/c decreases). It is customary to record the wavelength of the last reading which the investigator believes to be reproducible and reliable, and this point is often indicated by an asterisk after the intensity in the experimental section of the paper. For purposes of data presentation, the following form is convenient: CD spectrum of oleandomycin (c = 0.11, MeOH): [0]3440, [0]3025685, [0]~50- 190. (c = 0.86, MeOH): [0]~40- 500, [01212- 10860, [0121o - 8000*. It is clear from this that it was necessary to employ two different concentrations to get the data.
358
METHODS FOR THE STUDY OF ANTIBIOTICS
HOCH2 H+NHCOCHCI2 HO~H
CH20H 0,2CHCOHHN~~I NO?_ (12)
c
H
[17]
Q (13)
~
C
H
O,~,H..O~,H..O
3
0
0
(14) O R D spectrum of oleandomycin (c = 0.11, M e O H ) : [ ¢ ] 7 0 0 - 470, [¢]589 - 470, [¢]~00 -- 700, [¢]400 -- 2185, [¢]850 - 2655, [¢]3~o - 4690, [¢]29a 0, [¢]27s q- 1875, [¢]264 0, [@]2~5 -- 3590, [¢]220 - 3120".
It is interesting to note t h a t at the scale used in measuring this spectrum, the instrument pen, which makes a 1 mm-wide line, introduces a potential error as large as ± 1 5 6 [@] units, when the blank is also considered, for this is the magnitude indicated for each millimeter of pen deflection even when no noise is present. For a peak of [¢] 350 - - 4690, this error is about 3%, and hence not too serious, but at the sodium D line uncertainty represents about 30% of the reading itself. Thus, readings of [aid using such a spectropolarimeter should be done with much more
70 --
0
x
[]
/
-2
z~-4 '
=o Z I"-o
R 1 = --C(CHs) s or C(CDs) 3 R~ = --C(CHs) s o r C(CDs) s or - - C F 2 - - C F 2 - C F 3 Rs = - - H o r - - D M = Eu, P r , Yb, o r other r a r e e a r t h
(m)
402
METHODS FOR THE STUDY OF ANTIBIOTICS
[19[
sample it complexes with electron-rich functions on the molecule (CO--NH > - - N H > --OH > C-~-O > - - 0 - - > --CN), and by a pseudocontact mechanism it changes the shielding (AS) of nearby protons in an amount depending on their distance (R) and angle of deviation from the coordination axis (8): A S = K [ ( 3 cos 2 0 - - 1 ) / R 3 ] , where K is a scalar constant for the particular experiment including the binding ability of the reagent, its concentration, etc. The shifts in the spectrum can be paramagnetic (downfield) or diamagnetic (upfield) depending on 0 and the metal of the reagent. The desired effect is to increase the separation of the signals in order to make them less overlapped and more firstorder. The method works best in nonpolar solvents and with substrates that have a dominant binding site and therefore does not always work
R = -- C ( C H s ) s o r -- C F 2 - C F z - - C F 3
(iv) on antibiotics. If a chiral shift reagent (IV) is used enantiomers shift differently and can therefore be distinguished because the complex is diasterotopic. ~5 Computer programs are available 6~ which test the molecular geometry of proposed structures by computing the agreement factor between the observed and calculated shifts. This procedure is very useful when choosing from more than one proposed structure. A subtle change in the spectrum brought on by the nuclear Overhauser effects can sometimes be used to confirm the close proximity of protons across space. The test depends on internuclear relaxation, not on spin coupling. A weak radiation is applied to a given proton at its resonance frequency which partially saturates that proton's resonance signal changing its Boltzmann distribution. If a second proton is nearby across space and therefore depends on the perturbed proton for its relaxation, it too will experience a disruption of its Boltzmann distribution with the result that its resonance intensity will grow as much as 50%. The effect falls off rapidly with distance (1/R 6) and is a specific test for hydrogens in an eclipsed 5- or 6-position. R. E. Davis, M. R. Willcott, III, R. E. Lenkinski, W. yon E. Doering, and L. Birladeanu, J. Amer. Chem. Soc. 95, 6846 (1973).
[19]
P M R SPECTROSCOPY OF ANTIBIOTICS
403
TABLE III SHIFTS RESULTING FROM ACETYLATION OF GELDAN:kMYCIN
Proton
Shift, A~
Proton
Shift, A~
A X M B C D E Y2 F G Ha I Ja
+0.38 +0.09 -0.14 -0.04 -0.06 --0.67 --0.13 +0.19 +1.31 +0.04 --0.07 ? 0.01
K3 L N O
+0.05 -0.16 +0.02 -0.11
P3
--0.02
Q
T Ra $3 U Va Wa
?
--0.14 Absent -0.01 -0.26 -0.19 --0.01
The use of higher Larmor frequency has already been noted, and the N M R of other nuclei, such as 13C or 15N, may provide the needed information to solve the P M R spectrum. In the case of geldanamycin acetate (II) the first spectra were observed at 60 MHz. Comparison of the 60 and 100 MHz spectra helped identify the multiplets.
B. Changing the Sample For antibiotics a very useful reaction is acetylation or carbamation of the sample. The alcohols and amines can be characterized from the resulting shifts noting that secondary carbino166 or carbamine 32 hydrogens shift downfield much more than primary ones. The acetyl methyls can usually be counted, and the polarity may be diminished enough to allow additional observations in a less polar solvent. The carbamation reaction can be carried out in the N M R cell by the addition of trichloroacetyl isocyanate reagentY In the geldanamycin example, the acetate (II) was prepared with acetic anhydride and pyridine and the shifts recorded in Table I I I helped identify the molecular structure. Other reactions such as ketone derivatization, hydrogenation, or borohydride reduction have been used to help interpret spectra. In the geldanamycin case the methanolysis product (V) was prepared from (I) with C. R. Narayanan and M. R. Sarma, Tetrahedron Lett. 1553 (1968). 8, V. W. Goodlett, Anal. Chem. 37, 431 (1965).
404
METHODS FOR THE STUDY OF ANTIBIOTICS
[20]
CHsO~H o
~/ HaC/~ C
"VO
~'~NH~ CHaO"
O CH3 CHa C ~
H
3
0
OH
~ OCONH2
(V) p o t a s s i u m c a r b o n a t e in refluxing m e t h a n o l - c h l o r o f o r m ( 1 : 1 ) . T h e P M R s p e c t r u m of (V) h a d the I, Q, a n d T m u l t i p l e t s u n c o v e r e d , a n d a large upfield shift of the a r o m a t i c p r o t o n signal (M) showed t h a t the a m i d e m u s t h a v e been a t t a c h e d ortho to it.
[20] The Use of 13C Labeling in the Study of Antibiotic Biosynthesis By NORBERT NEUSS I. II. III. IV.
Introduction . . . . . . . . . . . . . . . Instrumental Requirements . . . . . . . . . . . Satellite Method . . . . . . . . . . . . . . Experimental Conditions . . . . . . . . . . . . A. Preliminary 14C Experiment . . . . . . . . . . B. Natural Abundance CMR Spectrum . . . . . . . C. Selection of an Appropriate 13C-Enriched Precursor . . D. Consideration of Nuclear Overhauser Effect (NOE) . . E. Conditions of Labeling . . . . . . . . . . . V. CMR in Biosynthetic Studies of Antibiotics . . . . . . VI. Biosynthesis of fl-Lactam Antibiotics . . . . . . . . A. Synthesis of Model Compounds and 1'C Precursors . . B. Fermentation, Labeling, and Isolation . . . . . . . C. Recording of the Spectra . . . . . . . . . . . D. Assignments of Chemical Shifts in the CMR Spectra . . E. Determination of Incorporation Levels of 13C Precursors F. Discussion and Conclusions . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
404 405 406 406 407 407 407 407 407 408 410 411 411 414 415 417 418
I. I n t r o d u c t i o n T h e i n t r o d u c t i o n of a n t i b i o t i c s into the t h e r a p y of b a c t e r i a l i n f e c t i o n s p r o m p t e d chemical i n v e s t i g a t i o n s in m a n y l a b o r a t o r i e s t h r o u g h o u t the
404
METHODS FOR THE STUDY OF ANTIBIOTICS
[20]
CHsO~H o
~/ HaC/~ C
"VO
~'~NH~ CHaO"
O CH3 CHa C ~
H
3
0
OH
~ OCONH2
(V) p o t a s s i u m c a r b o n a t e in refluxing m e t h a n o l - c h l o r o f o r m ( 1 : 1 ) . T h e P M R s p e c t r u m of (V) h a d the I, Q, a n d T m u l t i p l e t s u n c o v e r e d , a n d a large upfield shift of the a r o m a t i c p r o t o n signal (M) showed t h a t the a m i d e m u s t h a v e been a t t a c h e d ortho to it.
[20] The Use of 13C Labeling in the Study of Antibiotic Biosynthesis By NORBERT NEUSS I. II. III. IV.
Introduction . . . . . . . . . . . . . . . Instrumental Requirements . . . . . . . . . . . Satellite Method . . . . . . . . . . . . . . Experimental Conditions . . . . . . . . . . . . A. Preliminary 14C Experiment . . . . . . . . . . B. Natural Abundance CMR Spectrum . . . . . . . C. Selection of an Appropriate 13C-Enriched Precursor . . D. Consideration of Nuclear Overhauser Effect (NOE) . . E. Conditions of Labeling . . . . . . . . . . . V. CMR in Biosynthetic Studies of Antibiotics . . . . . . VI. Biosynthesis of fl-Lactam Antibiotics . . . . . . . . A. Synthesis of Model Compounds and 1'C Precursors . . B. Fermentation, Labeling, and Isolation . . . . . . . C. Recording of the Spectra . . . . . . . . . . . D. Assignments of Chemical Shifts in the CMR Spectra . . E. Determination of Incorporation Levels of 13C Precursors F. Discussion and Conclusions . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
404 405 406 406 407 407 407 407 407 408 410 411 411 414 415 417 418
I. I n t r o d u c t i o n T h e i n t r o d u c t i o n of a n t i b i o t i c s into the t h e r a p y of b a c t e r i a l i n f e c t i o n s p r o m p t e d chemical i n v e s t i g a t i o n s in m a n y l a b o r a t o r i e s t h r o u g h o u t the
[20]
13C LABELING IN ANTIBIOTIC SYNTHESIS
405
world concerned with elucidation of biosynthetic pathways of this important class of natural products. These studies consisted of addition of simple 14C-labeled substrates such as acetate, amino acids, shikimic acid, during the fermentation. After the completion of the feeding experiment, the metabolite was isolated and the labeling pattern was established by suitable chemical operations. The latter procedure can often be time consuming and, in the majority of cases, involves a bond-specific, multistep degradation process on limited amounts of material followed by laborious purification in order to achieve a constant level of radioactivity. In addition, some degradative procedures cannot be applied to complex molecules. Many of these problems can be avoided by the use of 13C-labeled precursors followed by examination of 13C NMR spectrum of the isolated 13C-enriched compound, and a comparison of that spectrum with a natural abundance CMR spectrum of an unprecursed metabolite. The ~'~C NMR or CMR spectroscopy is a powerful technique which has become extremely useful in recent years in structure elucidation and other facets of organic chemical research. Three excellent reviews on this subject have been published recently, 1-3 and therefore we shall only briefly discuss some of the general aspects of this technique.
II. Instrumental Requirements Unlike carbon-14, carbon-13 is a stable, magnetic isotope of carbon; therefore, no precautions are required in handling 13C-containing materials. Its nucleus is similar to that of hydrogen in that it has a spin of 1/,/2 and can be subjected to resonance in a strong magnetic field. Hence, a basic knowledge of proton spectroscopy can be transferred to CMR. The major difficulty with recording CMR spectra is the low natural abundance of 13C (only 1.1%) and a smaller magnetic moment than that of the protons. These two properties result in a considerably reduced sensitivity of detection of resonance (5700-fold). In other words, if a 10-rag sample were the lower limit of sample size for proton spectroscopy using a very widely available 60 MHz spectrometer, a 57-g sample would be required for carbon spectroscopy. Fortunately, a number of developments have helped solve this problem. These include: 1. The new type high field spectrometer of high stability which permits the use of a larger size sample (2 ml) in a 15 mm tube in conjunction 1F. A. Anet and G. C. Levy, Science 180, 141 (1973). 2j. B. Grutzner, Lloydia 35, 375 (1972). 3j. B. Stothers, Appl. Spectrosc. 26, 1 (1972).
406
METHODS FOR THE STUDY OF ANTIBIOTICS
[20]
with a computer capable of adding weak N M R signals with a net result of increase in intensity relative to noise signal. 2. The introduction of proton noise decoupling which results in the appearance of carbon-13 frequencies as sharp singlets rather than multiplets due to ~3C-1H spin-spin coupling. 3. The use of the pulsed Fourier transform which permits approximately a 10-fold increase in sensitivity by averaging, for instance, 100 spectra in the time necessary for the recording of a conventional spectrum. III. Satellite Method
One more aspect of 13C-labeling must be considered. In some instances, particularly when the 13C-label is contained in a carbon-bearing hydrogen atom, the spin-spin coupling between IsC nuclei and directly bonded protons produces "satellite" signals on either side of the main proton resonances in the proton magnetic resonance spectrum. These coupling constants range from 125 to 250 Hz depending on the carbon hybridization and substituents, and can be detected with a 100 MHz spectrometer available in many laboratories. In case of substantial enrichment in 13C and availability of 30-50-mg samples of a small molecular weight substance (e.g., l~C-labeled valine) a spectrum is of a quality permitting a detailed interpretation of coupling constants and degree of incorporation. To illustrate this type of spectrum we refer the reader to characteristic bands in the N M R spectrum of (2RS, 3R)-[4-13C]valine (Fig. 4). In cases of a small degree of incorporation, a C-1024 time averaging computer for multiscan averaging is required in conjunction with a 100 MHz spectrometer. Thus, during the study of biosynthesis of griseofulvin, 4 fusaric acid, 5 and piericidin A 6 the satellite signals from enriched sites in these two antibiotics were observed and used to establish their biosynthetic origin with enrichment as low as 2% and sample size ranging from 25 to 100 mg. IV. Experimental Conditions
Certain conditions have to be fulfilled for a valid experiment in the elucidation of biosynthetic pathway using a 13C precursor and CMR spectroscopy of the enriched metabolite. 4 M. Tanabe and G. Detre, J. Amer. Chem. Soc. 88, 4515 (1966). D. Desaty, A. G. McInnes, and L. C. Vining, Can. J. Biochem. 46, 1293 (1968). eM. Tanabe and H. Seto, J. Org. Chem. 3,5, 2087 (1970).
[20]
13C
LABELING
IN
ANTIBIOTIC
SYNTHESIS
407
A. Preliminary 14C Experiment It is absolutely necessary to check in a preliminary carbon-14 study the degree of incorporation of precursors to ascertain a sufficient level for 13C detection. In most cases 1% incorporation will be satisfactory. This prerequisite is necessitated by the ability to distinguish differences in peak intensity between the labeled compound and its natural ~3C abundance reference. At present this limit is a peak ratio of 1.2:1. In a case of utilization of 90% ~3C-enriched precursor, 0.2% incorporation may be detected.
B. Natural Abundance CMR Spectrum For a complete analysis of a natural abundance CMR spectrum of a metabolite to be studied, it is sometimes necessary to synthesize appropriate model compounds or derivatives in order to eliminate ambiguous assignments.
C. Selection of an Appropriate 1*C-Enriched Precursor Several simple 13C-enriched compounds are commercially available, and others can be synthesized according to the need, using standard procedures described for the synthesis of 14C-enriched compounds.
D. Consideration of Nuclear Overhauser Effect (NOE) The process of proton decoupling results in the increase of certain peaks (NOE). In order to avoid a misinterpretation of such an increase as incorporation of 13C from the precursor, it is advisable to run the natural abundance CMR spectrum of the metabolite and that of the enriched metabolite to be tested, under similar conditions (concentration, instrument setting, solvents, etc.); great care has to be taken to establish as accurately as possible the extent of NOE, generated by proton decoupling or Fourier transform technique. There are also some ways to minimize NOE.
E. Conditions of Labeling Finally, it should be obvious to anyone engaged in the technology of fermentation that conditions used in producing the 13C-labeled metabolite should closely mimic those used in the preparation of metabolite in
408
METHODS FOR THE STUDY OF ANTIBIOTICS
[20]
the regular fermentation run as well as 1~C feeding experiment preceding the 13C experiment. V. C M R in Biosynthetic Studies of Antibiotics The great value of CMR in the study of biosynthetic pathways was recognized early in the development of the technique ~ however, the problem of sensitivity delayed the use of the method until 1970, when biosynthetie formation of radicinin, 8 a microbial metabolite from Stenphylum radicinum was elucidated using the solvent peak as an internal lock. In this important modification, it is possible to switch from noise decoupling to single-frequency or continuous-wave (CW) decoupling when the decoupler is tuned to the exact frequency for irradiation of the solvent protons. Under such conditions most 13C-signals arising from CH3, CH~, or CH groups with proton shifts slightly removed from the solvent proton shifts, give characteristic close-spaced multiplets, thus permitting appropriate assignments. Until the introduction of this modification, biosynthetic investigations of antibiotics relied solely upon the presence of 13C-satellite in proton spectra to study the locus of 13C-label (see above). The summary of biosynthetic studies using the continuous wave (CW) as well as Fourier transform technique is shown in Table I. In all these studies, feeding of the enriched 13C-labeled precursor during the fermentation was followed by the isolation of the metabolite and comparison of its CMR spectrum with that of the natural abundance CMR of the metabolite. Since these experiments were published either as communications to the editor or abstracts of papers given at meetings, the details of experimental conditions and particularly spectra themselves, are not readily available. These considerations, as well as the uniqueness of the principle tested, prompted us to use as an example of a thorough biosynthetic study using CMR spectroscopy, the recent elueidation of some steps in the biosynthetic formation of fl-lactam antibiotics, cephalosporin C and penicillin V2 -12 7j. B. Stothers, Quart. Rev. 19, 144 (1965). 8 M. Tanabe, tI. Seto, and L. F. Johnson, J. Amer. Chem. Soc. 92, 2157 (1970). N. Neuss, C. tt. Nash, P. A. Lemke, and J. B. Grutzner, J. Amer. Chem. Soc. 93, 2337 (1971). ~oN. Neuss, C. tt. Nash, P. A. Lemke, and J. B. Grutzner, Proc. Roy. Soc. Ser. B. 179, 335 (1971). ~ N. Neuss, C. H. Nash, J. E. Baldwin, P. A. Lamke, and J. B. Grutzner, J. Amer. Chem. Soc. 95, 3797 (1973). ~2j. E. Baldwin, J. LSliger, W. Rastetter, N. Neuss, L. L. Huckstep, and N. DeLaHiguera, J. Amer. Chem. Soc. 95, 3796 (,1973).
[20]
13C LABELING IN ANTIBIOTIC SYNTHESIS
409
TABLE I SUMMARY OF BIOSYNTHETIC LABELING STUDIES OF ANTIBIOTICS BY C M R (UNTIL NOVEMnER 1973)
Compound
Precursor
Asperlin Cephalosporin C
[2-13C]Acetate ~ [1-13C]- and [2-1~C]Acetate ~ DL-[1-1~C]- and [2-1~C]d (2RS,3R)-[4-~SC]valine ~ (2S,3S)-[4-13C] Valine I [1-~3C]Acetate b [1-1aC]Propionate, [4-13C]methionineg [1-1~C]Glucosamine h [1-13C]Acetate~ (2S,3S)-[4- ~3C]Valine / (2RS,3R)-[4-~3C]Valine ~ oI~Tryptophan-alanine-[3- ~3C]j [1-1~C], [2-13C], and [3-1aC]Propionate [I-13C]- and [2-13C]Acetatek [1-1~C] and [2-1aC]Acetate~ [1-~3C]Propionate "~ [1-13C]Propionate and [1-~3C]butyraie ~
Cephalosporin C Chlorothricin Geldanamycin Neomycin Nybomycin Penicillin N Penicillin V Pyrrolnitrin Rifamycin Showdomycin Streptovaricin X-537A
M. Tanabe, T. Hamasaki, l). Thomas, and L. F. Johnson, J. Amer. Chem. Soc. 93, 273 (1971). b D. J. Hook, C. J. Chang, H. G. Floss, E. W. IIagaman, and E. Wenkert, Ab,~lr. Pap. Amer. Chem. Soc., 166 M E D I , p. 68 (1973). c N. Neuss, C. H. Nash, P. A. Lemke, and J. B. Grutzner J. Amer. Chem. Soc. 93, 2337 (1971). N. Neuss, C. H. Nash, P. A. Lemke, and J. B. Grutzner, Proc. Roy. Soc. Ser. B 179, 335 (1971). N. Neuss, C. H. Nash, J. E. Baldwin, P. A. Lemke, and J. B. Grutzner, J . . - i ~ e r . Chem. Soc. 95, 3797 (1973). / H. Khtender, C. H. Bradley, C. J. Sih, P. Fawcett, and E. P. Abraham, J. Amer. Chem. Soc. 95, 6149 (1973). g R. l). Johnson, A. Haber, and K. L. Rinehart, Jr., Abstr. Pap. Amer. Chem. Sot., 166 Orgn, p. 124 (1973). h S. T. Truitt, M. Taniguchi, J. M. Malik, R. M. Stroshane, and K. L. Rinehart, Jr., Abstr. Pap. Amer. Chem. Soc. 166 M E D I , p. 69 (1973). W. M. J. Knoll, R. J. Huxtable, and K. L. Rinehart, Jr., J. Amer. Chem. Soc. 95, 2703 (1973). i L. L. Martin, C. J. Chang, and H. G. Floss, J. Amer. Chem. Soc. 94, 8942 (1972). k R. J. White, E. Martinelli, G. G. Gallo, C. Lancini, and P. Beynon, Na/,r~ (London) 243, 273 (1973). z R. J. Suhadolnik and E. F. Elstner, in It. G. Floss, LIoydia 35, 399 (1972). " B. Milavetz, K. Kakinuma, and K. L. Rinehart, Jr., J. Amer. Chem. Soc. 95, 5793 (1973). '~ J. W. Westley, D. L. Pruess, and R. G. Pitcher, Chem. Commun. 161 (1972).
410
METHODS FOR T H E STUDY OF ANTIBIOTICS
[20]
H2N\ /SH e CH-(CH2L-CONHCN-CHzCHICH3)2 OOC/ 'J CO-NH-CH-COOH ~ -(~-AM INOADIPOYL)CYSTEINYLVALINE
o / CH-ICH2I~CONH o
ooc
0//
~
c.u o.
c.,
Q,.;,~N,C/..~ CH3 J COOH
N'~CH 3 COOH PENAM
O~
" ~ xCH2OAC COOH CEPHEM
FIG. 1. Proposed formation of penam and cephem antibiotics. VI. Biosynthesis of •-Lactam Antibiotics In spite of extensive effort over the past two decades, the detailed biosynthetic pathway to the fl-lactam antibiotics, penicillin, and cephalosporin, is still unknown. TM The evident relationship of these antibiotics, known as the penam and 3-cephem derivatives, to the amino acids L-valine and L-cysteine has been established by appropriate 14C incorporation experiments. 14-17 It has been postulated that both these systems could be derived from a common intermediate, an a,fl-dehydrovaline derivative of the tripeptide, 8-(a-aminoadipyl)cysteinylvaline, which after ring closure gives rise to the cephem and penam ring systems (Fig. 1). We decided to use valine with a chiral isotopic '3C-label at the 4-position as a probe for the stereochemical fate of this isopropyl group during conversion to the fl-lactam products. The synthesis of such a precursor required a considerable amount of effort. Therefore, a number of experiments had to be undertaken to reconfirm ~4C experiments and establish appropriate labeling and fermentation conditions suitable for the final experiment with chiral valine. B y way of example, details of these experiments are given below since they are typical for the study of biosynthesis using l~C-labeled precursors and C M R spectroscopy. ,sp. A. Lemke and D. R. Brannon, in "Cephalosporins and Penicillins" (E. H. Flynn, ed.), p. 370. Academic Press, New York, 1972. 14H. R. V. Arnstein and M. E. Clubb, Biochem. J. 65, 618 (1957). i, It. R. V. Arnstein and P. T. Grant, Biochem. J. 57, 353, 360 (1954). 1~S. C. Warren, G. G. F. Newton, and E. P. Abraham, Biochem. J. 163, 902 (1967). 1, E. P. Abraham, G. G. F. Newton, and S. C. Warren, I. A. M. Symp. Appl. Microbiol., Tokyo, 6, 79 (1964).
[20]
laC LABELING IN ANTIBIOTIC SYNTHESIS
411
A. Synthesis of Model Compounds and 13C Precursors The first prerequisite for these studies was fulfilled by the data available from the earlier investigations 14-17 with radioactive precursors, indicating sufficient degree of incorporation for CMR spectroscopy. The localization of the 13C-label was possible only after complete assignment of all 13C-frequencies of 16-carbon atoms in cephalosporin C (I). In order to accomplish this task it was necessary to examine CMR spectra of appropriate model substances. These included: DL-a-aminoadipic acid ethyl amide (II), eephalexin (liD, 3 methyl-7-(2-phenoxyacetamido)-3eephem (IV), and 7-aminocephalosporanic acid iV) (Fig. 2). From precursors to be used in this study [1J3C]- and [2-13C]sodium acetate were commercially available, and contained 62-68% of 13C. DL-[1-13C]-, and ~L-[2-13C]valine were prepared using KI~CN and [2-~'~C]glycine (both containing 62-63% of ~C, respectively), according to procedures described for the synthesis of analogous l~C-labeled compounds. (2RS, 3R)-[4-~'~C]Valine was prepared by reductive opening of optically pure (+)-trans-([1-~C]methyl-2-cyelopropane carboxylic acid ethyl ester giving rise to (3S)-[4-1~C]3-methylbutyric acid ethyl ester. The free acid was brolninated and subjected to aminolysis. The final purification by ion exchange chromatography gave pure (thin-layer chromatography, mass spectrometry, and amino acid analysis) (2RS, 3R)-[4"C]valine (Fig. 3). The estimated isotopic purity was 92% as estimated from the ratios of methyl frequencies containing no ~3C, but two ~CH:~. Chemical shifts and coupling constants in the NMR spectrum (Fig. 4) were as follows: NMR (D._,O)8 1.42 m (J['3C-'H] = 126 and JI'H'H] = 6.9 Hz), 8 2.72 m (J[~C-'H] ~ 1 0 Hz), and ~ 4.04 m (J ["~C-'H] -- 4.2 and J [H-'H] = 4.2 Hz). In addition, the 1~C and ~C isopropyl multiplets showed an isomeric difference of 0.05 ppm arising from the DL center. The estimated optical purity of the chiral center, based on the starting material and its optically active transformation products, was 100%)'-' For comparison the NMR spectrum of commercially available DLvaline is shown in Fig..5.
B. Fermentation, Labeling, and Isolation Submerged cultures of Cephalosporium acremonium, a superior antibiotic-producing mutant, M8650-3 is were grown at 25 ° on a rotary shaker (250 rpm) in a complex medium. 19 Cephalosporin C was labeled with is D. W. D e n n e n and D. D. Carver, Can. J. Microbiol. 15, 175 (1969). 29p. A. Lemke and C. H. Nash, Can. J. Microbiol. 17, 255 (1971).
412
METHODS FOR THE STUDY OF ANTIBIOTICS
0//
[20]
" ~ "~R3 R2
o,, R1 = -C-CH2-CH2-CH2-CH 10 11
12
13 14~00~,.,
Cephalosporin C Na-salt
(z)
15 0 u R2 = COONa; R 3 = CH2OC-CH3 16 17 18 19
o II C-CH2-CH2-CH2-CH C H 3 _ C H 2 _ N H / l O 11 2 1'
2'
12
13 141COOe 15
at- 4)streptidine. Streptomycin has an aldehyde group rather than a hydroxymethyl group at position 3',,. Bluensomycin is N-methyl-a-L-glucosamine(1---> 2)a-L-dihydrostreptose(1--->4)bluensidine. Mild acid hydrolysis of these compounds gives streptidine or bluensidine, respectively, plus the disaccharide, dihydrostreptobiosamine (DSBA). Bluensidine has also been called glebidine. None of the components of these antibiotics has so far been found elsewhere in nature.
430
ANTIBIOTIC BIOSYNTHESIS
[21]
. fP
+~ 0 0
0
-x~_ 0
~ 0
o~
~, 0
"
~-_~~ T
II ~
+~ fill
.,~
~-~
~
~-~-~---~ ~ II ~
0
,.,.,
.
0-0 ~l~l
0
~ o m
Z
~- - ~ o ~ 0
"~ "
*~
I
o--.,~o
I
I
.o
.~
[21]
BIOSYNTHESIS OF GUANIDINATED INOSITOL
431
The following group of articles is concerned with enzymes involved in biosynthesis by Streptomyces strains of the guanidinated inositol moieties of the streptomycin family of antibiotics, including dihydrostreptomycin (I) and bluensomycin (II), whose structures are given in Fig. 1.1 Studies in vivo on the biosynthesis of streptomycin by intact mycelia 2 have provided important guideposts for enzymic studies with cell-free extracts. ~,4 Results of the latter enzymic studies are summarized in Fig. 2. Two analogous sequences of 5 enzymic reactions each, operating in series, appear to be involved in biosynthesis of the streptidine moiety of dihydrostreptomycin (I) from myo-inositol (V). Each sequence converts a hydroxyl group to a guanidino group and involves, in order, a dehydrogenation (C and H ) , transamination (D and I), phosphorylation (E and J), transamidination (F and K), and dephosphorylation (G and N). For certain pairs of corresponding reactions in the two sequences, it is known that different enzymes are involved, although some have overlapping substrate specificities. Only one of the above two sequences appears to be operative in strains that synthesize bluensomycin (II). 4 Descriptions of the individual enzymic reactions in the following articles will follow the suggested biosynthetic sequence, starting with reaction C. Emphasis will be placed in these articles on the preparation and detection of various intermediates and their analogs and on detailed descriptions of several alternative assays for each enzyme. Radiochemical assays involving trace amounts of substrates and products have dominated these early enzymic studies, but it is anticipated that assays can subsequently be developed that utilize, e.g., ninhydrin and Sakaguchi reactions or dansylation of intermediates containing amino, guanidino, a n d / o r keto groups. 1Published and previously unpublished work reported in these articles was supported by the Robert A. Welch Foundation and the National Institutes of Health. 2 A. L. Demain and E. Inamine, Bacteriol. Rev. 34, 1 (1970). 3j. B. Walker, Lloydia 34, 363 (1971). 4j. B. Walker, J. Biol. Chem. 249, 2397 (1974).
FIG. 2. Current concept of enzymic steps involved in biosynthesis of the streptidine moiety of dihydrostreptomycin (I) by Streptomyces humidus ATCC 12760 and the bluensidine moiety of bluensomycin (II) by S. glebozus ATCC 14607, starting from glucose-6-P (III). Extracts of S. glebos~s catalyze reactions C, D, E, F, G, and K. The step at which the carbamoyl group is added has not been established. Enzyme G dephosphorylates substrate molecules which have escaped carbamoylation. Enzymes H and I are apparently missing in S. glebosus. Abbreviations: DSBA, dihydrostreptobiosamine; KGAM, a-ketoglutaramate; Orn, ornithine; Pyr, pyruvate; NDP-sugar, nucleosidediphosphate sugar.
432
ANTIBIOTIC BIOSYNTHESIS
[21]
MOBILITIES OF VARIOUS BIOSYNTHETIC INTERMEDIATES AND ANALOGS Relative H V E mobility d
R/'
HC1 elution from Dowex 50 c o l u m n ~
1D-1-Amino-1-deoxy-scyllo-inositol-6-P 1-Amino- 1-deoxy-scyllo-inositol-4-P
0.04
0.5 N
0.0
O. 10
O. 5 N
- O. 3
d- O. 9
1D-1,3-Diamino-l,3-dideoxy-scyUo-
0.13
1.0 N
--0.5
+0.9
0.08
2.0 N
-0.5
d-l.0
0.17 0.21 0.20
0.5 N 1.0 N 2.0 N
- 0.3 --0.9
+0.2 -+0.1
0.43
2.5 N
--
--
0.33
2.0 N
- 0.9
+0.2
1 D- 1,3-Diguanidino- 1,3-dideoxy-scylloinositol-6-P
0.40
2.0 N
- 0.8
- 0.5
1D-1,3-Diamino-l,2,3-trideoxy-scyllo-
0.25
1.0 N
--
--
0.39 0.59 0.83
0.5 N 1.0 N 2.5 N
-- 1 . 2 - 1.1 - 1.8
-- 0 . 2 - 1.0 - 1.0
0.83
2.5 N
- 1.8
-- 1 . 0
0.92 0.71 Streaks
5.0 N 1.0 N 1.0 N
- 2.0 -- 1.1
-- 1.7 ---
Streaks
H~O
- 0.3
--
Compounda
p H 3.6 p H 10.4 + I. 0
inositol-2-P
1D-1,3-Diamino-l,3-dideoxy-scylloinositol-6-P
1-Guanidino-l-deoxy-scyllo-inositol-4-P 2-Guanidino-2-deoxy-neo-inositol-5-P 1D-1-Amino-3-guanidino-l,3-dideoxy-
scyllo-inosi tol-6- P 1D- 1-Amino-3-guanidino-1,2,3-trideoxy-
scyllo-inositol-6-P 1D-1-Guanidino-3-amino-l,3-dideoxy-
scyllo-inositol-6-P
inositol-6-P
Aminodeoxy-scyllo-inositol Guanidinodeoxy-scyUo-inositol 1D-1-Amino-3-guanidino-1,3-dideoxy-
scyllo-inositol 1D- 1-Guanidino-3-amino-l,3-dideoxy-
scyllo-inositol 1,3-Diguanidino-l,3-dideoxy-scylla-inositol 2-Guanidino-2-deoxy-neo-inositol 1D- 1-Guanidino- 1-deoxy-3-keto-scylloinositol
Keto-scyllo-inositol
Cyclitol n o m e n c l a t u r e according to I U P A C t e n t a t i v e rules [Eur. J. Biochem. 5, 1 (1968)], except t h a t a keto s u b s t i t u e n t is so designated r a t h e r t h a n n a m i n g c o m p o u n d as a derivative of cyclohexanone. b Ascending paper c h r o m a t o g r a p h y , 80% p h e n o l - 2 0 % H20, N H 4 O H a t m o s p h e r e . T h i s v o l u m e [24]. c M i n i m u m concentration of HC1 required to elute c o m p o u n d adsorbed on a column containing Dowex 50 (H +) X 8 cation exchange resin, 200-400 mesh, in the stepwise sequence: H20, 0.5 N HC1, 1.0 N HC1, 2.0 N HC1, 2.5 N HC1, 5.0 N HC1. This v o l u m e [23]. Mobilities relative to distance traveled b y a pieric acid m a r k e r d u r i n g highvoltage paper electrophoresis at t h e indicated p H . T h i s v o l u m e [22]. A m i n u s sign indicates m i g r a t i o n toward t h e n e g a t i v e electrode, a n d a positive sign indicates m i g r a t i o n toward t h e positive electrode. Mobilities v a r y s o m e w h a t w i t h point of application a n d t e m p e r a t u r e .
myo-INOSITOL:NAD + 2-OXIDOREDUCTASE
[22]
433
Mobilities of certain of the biosynthetic intermediates and their analogs are summarized in the table ; these properties can be utilized for both separation and identification. The enzymes themselves emerge from a Sephadex G-100 column in the following approximate order: streptomycin-6-P phosphatase (N), glutamine: keto-scyllo-inositol aminotransferase (D), alanine: 1D-l-guanidino-l-deoxy-3-keto-scyUo-inositol aminotransferase (I), inosamine-P amidinotransferase (F, K), streptomycin 6-kinase (0), guanidinodeoxy-scyllo-inositol-4-P phosphatase (G), and inosamine kinase(s) (E, J). Except for enzymes N and O, most of the enzymes appear to be sulfhydryl enzymes. Although commercial strains of Streptomyces selected by sequential mutation procedures for high productivity might well have higher levels of certain of these biosynthetic enzymes, all the procedures to be described employ the more generally available Streptomyces strains from the American Type Culture Collection (ATCC), 12301 Parklawn Drive, Rockville, Maryland 20852.
[22] m y o - I n o s i t o l : N A D + 2 - O x i d o r e d u c t a s e By
JAMES B . WALKER
OH
OH
+ tt
H
NAD+ ~
0
OH m yo - Inositol
+
NADH
+
(1)
tt
OH Keto- s c y l l o - ino sitol (scyllo-inosose ; myo-inosose-2)
When Streptomyces hygroscopicus ]orma glebosus ATCC 14607 (S. glebosus) is grown with myo-inositol as a major energy source, myoinositol 2-dehydrogenase activity 1 is induced, and activity can be readily measured in crude extracts spectrophotometrically at 340 nm by conventional procedures, in either direction. 2 However, most streptomycin-producing strains cannot utilize myo-inositol as a major energy source, and in such strains myo-inositol 2-dehydrogenase activity has not yet been i E C 1.1.1.18. F o r the e n z y m e f r o m A e r o b a c t e r 2 j . B. Walker, u n p u b l i s h e d data.
a e r o g e n e s , see this series, Vol. 5 [36].
myo-INOSITOL:NAD + 2-OXIDOREDUCTASE
[22]
433
Mobilities of certain of the biosynthetic intermediates and their analogs are summarized in the table ; these properties can be utilized for both separation and identification. The enzymes themselves emerge from a Sephadex G-100 column in the following approximate order: streptomycin-6-P phosphatase (N), glutamine: keto-scyllo-inositol aminotransferase (D), alanine: 1D-l-guanidino-l-deoxy-3-keto-scyUo-inositol aminotransferase (I), inosamine-P amidinotransferase (F, K), streptomycin 6-kinase (0), guanidinodeoxy-scyllo-inositol-4-P phosphatase (G), and inosamine kinase(s) (E, J). Except for enzymes N and O, most of the enzymes appear to be sulfhydryl enzymes. Although commercial strains of Streptomyces selected by sequential mutation procedures for high productivity might well have higher levels of certain of these biosynthetic enzymes, all the procedures to be described employ the more generally available Streptomyces strains from the American Type Culture Collection (ATCC), 12301 Parklawn Drive, Rockville, Maryland 20852.
[22] m y o - I n o s i t o l : N A D + 2 - O x i d o r e d u c t a s e By
JAMES B . WALKER
OH
OH
+ tt
H
NAD+ ~
0
OH m yo - Inositol
+
NADH
+
(1)
tt
OH Keto- s c y l l o - ino sitol (scyllo-inosose ; myo-inosose-2)
When Streptomyces hygroscopicus ]orma glebosus ATCC 14607 (S. glebosus) is grown with myo-inositol as a major energy source, myoinositol 2-dehydrogenase activity 1 is induced, and activity can be readily measured in crude extracts spectrophotometrically at 340 nm by conventional procedures, in either direction. 2 However, most streptomycin-producing strains cannot utilize myo-inositol as a major energy source, and in such strains myo-inositol 2-dehydrogenase activity has not yet been i E C 1.1.1.18. F o r the e n z y m e f r o m A e r o b a c t e r 2 j . B. Walker, u n p u b l i s h e d data.
a e r o g e n e s , see this series, Vol. 5 [36].
434
ANTIBIOTIC BIOSYNTHESIS
[22]
detected in crude extracts by a spectrophotometric assay at 340 nm, even when assayed in the equilibrium-favored reverse direction. Such an assay might be feasible with more purified preparations, when such preparations become available. Since it is not yet known whether the same gene codes for both the biosynthetic and catabolic myo-inositol 2-dehydrogenase activities, two procedures will be described below for detection of biosynthetic myo-inositol dehydrogenase activity in extracts of mycelia grown under conditions in which the myo-inositol catabolic pathway does not appear to be induced. Both assay methods employ radioactive substrates and involve transamination of the keto-scyllo-inositol formed in Eq. (1) to give aminodeoxy-scyllo-inositol (coupling of reactions C and D). s Components of the incubation mixtures are usually separated by high voltage paper electrophoresis at pH 3.6 and counted. Alternatively, separations can be performed with very small Dowex 50 (H ÷) columns, as described for larger columns.4
Assay Methods
Method 15 myo_[U_14C]inosito1 + NAD+ c_~keto_scyllo_[U_14C]inositoI + NADH -{- H+ (2) D
Keto-scyllo-[U-~4C]inositol -}- L-glutamine---> aminodeoxyoscyllo-[U-14C]inositol -~ a-ketoglutaramate (3) Principle. This method is based upon conversion of labeled myo-inositol, having no net charge, to labeled aminodeoxy-scyllo-inositol (pK~ 7.6), having one positive charge at pH 3.6, in a coupled enzyme assay. Reagents Extract of Streptomyces hygroscopicus ]orma glebosus ATCC 14607
(S. glebosus) myo-[U-14C]Inositol, ca. 200 Ci/mole, from New England Nuclear (myo-[U-3H]inositol would also be satisfactory, but not myo- [2-SH ] inositol) NAD ÷, 3 mM Potassium phosphate buffer, 100 mM, containing 13 mM EDTA and 8 mM pyridoxal-P, pH 7.4. L-Glutamine, 40 mM ' This volume [21]. 4This volume [23]. ~J. B. Walker, ]. Biol. Chem. 249, 2397 (1974).
[22]
myo-INOSITOL:NAD + 2-OXIDOREDUCTASE
435
Procedure. The complete incubation mixture contains: myo-[U14C]inositol, 5 ~1 (e.g., 68,000 cpm) ; phosphate buffer, EDTA, and pyridoxal-P, 5 ~1; NAD ÷, 5 ~1, glutamine, 5 ~1, and dialyzed S. glebosus extract, 10 ~1. After incubation in a stoppered 13 X 100 mm test tube at 35 ° for 145 min, 10-~1 aliquots are spotted at the 30-cm mark and the components separated by high-voltage paper electrophoresis at pH 3.6 (see below). Nondialyzed extracts can also be employed; NAD and amino donor requirements are more pronounced with dialyzed extracts. '~ Growth of Mycelia. Well sporulated slants of S. glebosus, as well as most other strains of Streptomyces employed in these studies, are obtained by growth at room temperature on 1% maltose-0.5% Tryptone-0.2% yeast extract-l.5% agar-tap water. After incubation for 4-10 days at room temperature, a 1-cm square of surface growth and conidia is transferred with a stiff iron wire loop to a 2-liter Erlenmeyer flask containing 500 ml of autoclaved medium composed of 0.05% glucose-l% Soytone-l% Tryptone-0.2% yeast extract-0.03% K2HPQ-tap water. To assure dispersal of eonidia and mycelial fragments, the agar-square inoculum is often rubbed vigorously with the loop against the inside of the flask below the surface of the medium. After growing on a New Brunswick Psychrotherm rotary shaker at 240 rpm for 2 or 3 days at 26 °, or on a similar shaker at room temperatures of 24-28 °, mycelia are harvested by suction filtration through Whatman No. 1 qualitative filter paper (12.5 em diameter) on a Biichner funnel, pressed dry between paper toweling for 5-10 min under a 5-1b weight, rolled together into a cylindrical pad, wrapped with Saran wrap, weighed, frozen, and then stored in capped jars in a freezer for subsequent use. Many of the idiophase enzymes from various strains to be described in subsequent sections remain active in frozen pads for more than a year. Higher activities of most of these enzymes are usually obtained when large inocula from fresh, well sporulated slants are employed, using slants which had previously been inoculated heavily with fresh conidia. Rapid initial growth from many disseminated loci (e.g., viable conidia~ favors subsequent high specific activity of these idiophase enzymes. Since idiophase enzyme '~ctivities usually reach a peak after the rapid growth phase, G it is recommended that flasks inoculated at the same time from a given slant be harvested after both 2 days and 3 days of growth to assure getting at least one pad with optimal activity. Yields of 6-15 g of frozen mycelia are obtained per flask, depending upon conditions. In the case of S. glebosus, addition of 1% myo-inositol to the above growth medium causes induction later in the growth phase of high levels of a number of enzymes 6j. B. Walker and V. S. Hnilica, Biochim. Biophys. Acta 89, 473 (1964).
436
ANTIBIOTIC BIOSYNTHESIS
involved- in catabolism 2-dehydrogenase2
of
myo-inositol,
[22]
including
myo-inositol
Preparation o] Extracts. Extracts of frozen mycelial pads are prepared by sonication with a Branson sonifier of batches of 6 g of mycelia dispersed in 12 ml of deionized water contained in a wide-mouthed glass tube, 25-ml capacity, surrounded by an ice-water bath. Sonication is performed i n 1-min segments separated by recooling periods, during which time the suspension is stirred with a thermometer until the temperature is lowered to 5-10% The total time of sonica~ion optimal for a given enzyme should be determined empirically; 3-6 min is satisfactory for most purposes. Sonicates are then centrifuged at 30,000 g for 20 min at 2 °, and the supernatant solutions are stored frozen. Dialyzed S. glebosus extracts for use in studies for which reaction D activity must be retained are prepared by adding to 4 ml of the above supernatant solution 0.3 ml of a solution containing 0.7 M potassium phosphate, 0.3 M EDTA, and 0.04 M pyridoxal-P, pH 7.6. The resulting solution is pipetted into cellophane dialysis tubing, immersed in 4 liters of cold deionized water, dialyzed for 16 hr at 4 °, and stored frozen. High-Voltage Paper Electrophoresis Technique. Separations are performed on buffer-saturated Whatman No. 1 filter paper with a Savant horizontal plate apparatus (46-cm plate size). A refrigerated solution of ethylene glycol and water is circulated through the hollow metal supporting plate during the approximately 90-min run. One plastic sheet (Savant) insulates the filter paper sheet from the grounded aluminum cooling plate, and another plastic sheet is placed on top of the filter paper; a heavy square of thick plate glass rests on the top plastic sheet during a run. Each electrode vessel contains 1 liter of buffer. The filter paper sheet is 57 cm long and has previously been ruled with a hard graphite pencil so as to have 2 to 6 parallel and separated tracks, each 3.7 cm wide. Each track on that 46-cm portion of the paper which will rest on the cooling plate is marked off in 46 numbered 1-cm segments, with the No. 1 segment nearest the negative electrode. Unruled track space (2.5 cm wide) is provided at each side of the paper for the reference compound. The unruled 5.5-cm portions at each end of the filter paper serve as wicks when immersed, right after sample application, in the buffer vessels containing the electrodes. Before a run is started, the filter paper sheet is dipped in buffer contained in a wide pan, quickly placed in position on top of the lower plastic sheet, blotted as dry as possible with paper toweling and aligned in final position before samples are applied. At the start of a run a small drop of a saturated solution of picric acid is applied as a mobility reference marker at the 10-cm segment level on each edge of the paper. After the 10-td sample aliquots and the picric
[22]
myO-INOSITOL : N A D + 2-OXIDOREDUCTASE
437
acid markers are spotted, the paper ends are immersed in buffer, and the upper plastic sheet and then glass plate are placed on top and the cooling water circulation is started. The power supply normally is set at 1500 V. The run is usually terminated after the yellow pierie acid spots have migrated from 11 to 18 era. A razor blade is used to sever tile wet filter paper next to the buffer vessels, and the sheet is quickly hung by one (anode) end on a horizontal glass rod, employing stainless steel clamps, Excess buffer is blotted from the extreme lower (eathodet end of the hanging paper sheet. The sheet is dried at room temperature in a hood. Strips are cut from the dried paper, spliced end-to-end with Scotch transparent tape, with suitable lengths of filter paper strips (1.5-em wide) as spacers, and run through a Nuclear Chicago radioactive strip counter. Quantitative data are obtained by cutting numbered l-era segments from tile strips, placing the two halves of each segment flat on the bottom of a scintillation bottle containing 5 ml of diluted Liquifluor solution in toluene, and counting with a liquid scintillation system. Two buffers have been employed in our electrol)horetic separations: (a) an acid buffer in which amino and guanidino groups each have a single positive charge, and i)hosphate ester groups have a single negative charge; and (b) a basic buffer in which guanidino groups have one positive charge, amino groups have no charge, and phosphate ester groups have two negative charges. These buffers are prepared as follows, using glass-distilled water. (a) Acid buffer (pH ca. 3.6): 26 g of ammonium formate and 17.4 ml of 88% formic acid are dissolved in 4 liters of water. (b) Basic buffer (pH ca. 10.4): 72.6 g of glycine and 32.8 g of sodium hydroxide pellets arc dissolved separately and made up to 4 liters with water. Buffers are stored in plastic bottles. Each batch can t)e used for many runs before replacement; after each run the buffer solutions are returned to the bottle. The approximate relative mobilities of certain of the biosynthetic intermediates are given elsewhere. 3 Mobilities vary somewhat with the point of sample application.
Method II 5 myo-Inositol + NA1) + ~ keto-scgllo-inositol + N A I ) t t + H + •
4
(4)
D
Keto-sc?tllo-inositol + l-ammo- l-deoxy-scyllo-[ l-1 C]inositol --~ aminodeoxy-scyllo-inositol + l-keto-scyllo-[l-l~C]inositol
(5)
Principle. In this method a high concentration of nonlabeled myoinositol is employed to help overcome the unfavorable equilibrium of reaction C. The nonlabeled keto-scyllo-inositol formed in Eq. (4) serves
438
ANTIBIOTIC BIOSYNTHESIS
[22]
as acceptor of an amino group from labeled aminodeoxy-scyllo-inositol in an exchange reaction catalyzed by transaminase D. Labeled aminodeoxy-scyUo-inositol with one positive charge at pH 3.6 is converted to labeled keto-scyllo-inositol with no net charge.
Reagents Extract of S. glebosus ATCC 14607 1-Amino-l.deoxy-scyllo- [ 1-14C] inositol, 3.5 Ci/mole, from California Bionuelear Corp., Sun Valley, (scy!lo-inosamine, Cat. 72673). Labeled compound can also be prepared as described in Method I and isolated by ion-exchange chromatographyA Potassium phosphate buffer, 100 mM, containing 13 mM EDTA, and 8 mM pyridoxal-P, pH 7.4 myo-Inositol, 110 mM NAD ÷, 3 mM
Procedure. The complete incubation mixture contains: 1-amino-1deoxy-scyUo-[1-1~C]inositol, 5 ~l (e.g., 68,000 cpm); phosphate buffer, EDTA, and pyridoxal-P, 5 ~l; myo-inositol, 5 ~1; NAD ÷, 5 ~l; and nondialyzed or dialyzed extract of S. glebosus (see Method I), 10 ~l. After incubation in are spotted at voltage paper formed in Eq. the extract
a stoppered 13 X 100 mm test tube at 35 °, 10-/A aliquots the 30-cm segment and the components separated by highelectrophoresis at pH 3.6. The labeled keto-scyllo-inositol (5) is transformed to negatively charged metabolites when is prepared from mycelia previously induced with
myo-inositol. Properties myo-Inositol 2-dehydrogenase activity has only recently been detected in cell-free extracts of Streptomyces, 5 and extensive studies have not yet been made. Biological Distribution. myo-Inositol 2-dehydrogenase activity has also been detected in cell-free extracts of S. bikiniensis ATCC 11062, a streptomycin-producing strain, employing Method II. In addition to the occurrence of this enzyme in streptomycin producers, it is anticipated that strains of Streptomyces which can utilize myo-inositol as a carbon source would also have myo-inositol 2-dehydrogenase activity induction by substrate. It is not yet known whether the inducible dehydrogenase of S. glebosus is identical with the myo-inositol dehydrogenase involved in biosynthesis of bluensomycin or streptomycin.
KETO-scyllo-INOSITOL AMINOTRANSFERASE
[23l
439
Purification. Purification has not yet been performed, so it is not known whether enzymes C and D are physically associated. The assays of Methods I and II are adequate for detection and preliminary substrate specificity studies, but they would not be useful for purification purposes unless a preparation of purified transaminase D is available. If the inducible and biosynthetic myo-inositol 2-dehydrogenase of S. glebosus are identical, a spectrophotometric assay based on either the forward or reverse reactions of Eq. (1) would be the method of choice for purification studies. Specificity. D-chiro-Inositol and L-chiro-inositol are not effective substrates. In crude extracts N A D P + has less than 10% of the activity of N A D t In the coupled assay of Method I, aminodeoxy-scyllo-inositol is a better amino donor than L-glutamine, streptamine (1,3-diamino-l,3dideoxy-scyllo-inositol) is as effective as a-glutamine, and L-alanine has only slight activity at high concentrations. The foregoing amino donor specificities, of course, are a property of transaminase D, not the 2-dehydrogenase. I t appears that enzymes C and H are different enzymes?
[23] L-Glutamine:
Keto-scyllo-inositol A m i n o t r a n s f e r a s e ~ By JAMES B. WALKER HeC--CH 2
tt
O II
OH
O
I C--NH Ctt 2 2 I O + CH~ I HC,--NH:
@
H OH
Keto-s c y l l o inositol
o/~C" O-
L -
Glutamine
0 II
OH
~ N
tt2
D
+ HO
OH
Aminodeoxy-
s c y l l o -inositol ( s c y l l o -inosamine)
C--NH2 I CHz CH2
(1)
E
o/~C'~o-
a-Ketoglutaramate
This reaction is unusual among reactions involving glutamine in that the a-NH._, group, rather than the amide group, is transferred, the first
1EC 2.6.1.50.
KETO-scyllo-INOSITOL AMINOTRANSFERASE
[23l
439
Purification. Purification has not yet been performed, so it is not known whether enzymes C and D are physically associated. The assays of Methods I and II are adequate for detection and preliminary substrate specificity studies, but they would not be useful for purification purposes unless a preparation of purified transaminase D is available. If the inducible and biosynthetic myo-inositol 2-dehydrogenase of S. glebosus are identical, a spectrophotometric assay based on either the forward or reverse reactions of Eq. (1) would be the method of choice for purification studies. Specificity. D-chiro-Inositol and L-chiro-inositol are not effective substrates. In crude extracts N A D P + has less than 10% of the activity of N A D t In the coupled assay of Method I, aminodeoxy-scyllo-inositol is a better amino donor than L-glutamine, streptamine (1,3-diamino-l,3dideoxy-scyllo-inositol) is as effective as a-glutamine, and L-alanine has only slight activity at high concentrations. The foregoing amino donor specificities, of course, are a property of transaminase D, not the 2-dehydrogenase. I t appears that enzymes C and H are different enzymes?
[23] L-Glutamine:
Keto-scyllo-inositol A m i n o t r a n s f e r a s e ~ By JAMES B. WALKER HeC--CH 2
tt
O II
OH
O
I C--NH Ctt 2 2 I O + CH~ I HC,--NH:
@
H OH
Keto-s c y l l o inositol
o/~C" O-
L -
Glutamine
0 II
OH
~ N
tt2
D
+ HO
OH
Aminodeoxy-
s c y l l o -inositol ( s c y l l o -inosamine)
C--NH2 I CHz CH2
(1)
E
o/~C'~o-
a-Ketoglutaramate
This reaction is unusual among reactions involving glutamine in that the a-NH._, group, rather than the amide group, is transferred, the first
1EC 2.6.1.50.
440
ANTIBIOTIC BIOSYNTHESIS
[23]
such reaction observed in a prokaryoteY Meister and co-workers had earlier discovered an L-glutamine:a-ketoacid aminotransferase in mammalian liver; they found that such transaminations are made relatively irreversible by either spontaneous cyclization of a-ketoglutaramate or deamidation by an amidase to form a-ketoglutarate. ~
Assay M e t h o d
Principle. Transaminations involving pyridoxal-P as coenzyme are the sum of two half-reactions. One of the half-reactions of Eq. (1) is Eq. (2), which can be assayed as indicated in Eq. (3). The amino group of labeled aminodeoxy-scyllo-inositol is transferred to keto-scyllo-inositol, giving labeled keto-scyllo-inositol. Reactants and products are usually separated by high voltage paper electrophoresis. Alternatively sepaD
Keto-scyllo-inositol + enzyme-pyridoxamine-P aminodeoxy-scyllo-inositol + enzyme-pyridoxal-P (2) D Keto-scyllo-inositol + aminodeoxy-scyllo-[14C]inositol --~ aminodeoxy-scyllo-inositol + keto-scyllo-[14C]inositol (3) rations can be performed with small Dowex 50 (H ÷) columns. This assay can be used for both crude and purified preparations; purified preparations must contain pyridoxal-P at all times to retain activity.
Reagents Extract of Streptomyces hygroscopicus ]orma glebosus ATCC 146074 1-Amino-l-deoxy-scyllo- [1-14C]inositol, 3..5 Ci/mole, from California Bionuclear Corp., Sun Valley, Calif. (scyllo-inosamine[1-1~C], Cat. No. 72673). Labeled compound can also be prepared as described below. Potassium phosphate buffer, 100 raM, containing 13 mM EDTA and 8 mM pyridoxal-P, pH 7.4 Keto-scyllo-inositol, 20 m M (freshly prepared solution), from Sigma (myo-inosose-2, Cat. No: I 5375). This compound can also be prepared from myo-inositol by either catalytic oxidation or fermentation with Acetobacter suboxydans ATCC 621. ~ J. B. Walker and M. S. Walker, Biochemistry 8, 763 (1969). s See this series, Vol. XVIIA, addendum to article [136]. This volume [22]. T. Posternak, Biochem. Prep. 2, 57 (19527
[23]
KETO-seyllo-INOSITOL AMINOTRANSFERASE
441
Procedure. The complete incubation mixture contains: 1-amino-1deoxy-scyllo-[14C]inositol, 5 tL1 (e.g., 68,000 cpm); phosphate buffer, EDTA, and pyridoxal-P, 5 t~l; keto-scyllo-inositol, 5 ~l; and nondialyzed or dialyzed extract of S. glebosus, l0 t~l. After incubation in a stoppered 13 X 100 mm test tube at 35 °, a 10-~1 aliquot is spotted at the 30-cm segment and the components separated by high-voltage paper electrophoresis at pH 3.6, as described earlier. ~ The labeled amino donor has one positive charge (pK,, = 7.6) at pH 3.6, whereas the labeled product is uncharged (see Table I of Walker 6 for mobil[ties). The extract should be prepared from cells not induced with myo-inositol, to prevent catabolism of keto-scyllo-inositol. Preparation o] Labeled Aminodeoxy-scyllo-inositol. Any of several methods can be employed. For example, 1-amino-l-deoxy-scyllo-[114C]inositol can be chemically synthesized from glucose and ['4C]nitromethane, as the penultimate intermediate in an established procedure for synthesis of myo-[2-14C]inositol. 7 In our early work, labeled aminodeoxy-scyllo-inositol and its 4-phosphate, along with other ring-labeled precursors of the streptidine moiety of streptomycin, were prepared by feeding a pulse of myo-[14C]inositol to a culture of S. griseus ATCC 12475 and extracting the mycelia. ~ However, the most efficient and dependable conversion of myo-[l~C]inositol to aminodeoxy-scyllo[14C]inositol employs the coupled enzymic procedure described in Method I of an earlier article, ~ with reagent volumes suitably scaled-up for preparative purposes. Labeled substrate and labeled product can be readily separated by ion-exchange column chromatography (see below). Unconverted myo-[~4C]inositol can be isolated and used again to increase the ultimate yield. When extracts of certain streptomycin-producing strains are used as the source of enzymes C and D, the desirability of adding excess myo-inositol 2-dehydrogenase (Sigma Cat. No. I 5503) should be explored in a preliminary microassay.
Isolation o] Various Labeled Biosynthetic Intermediates by Ion-Exchange Column Chromatography. Samples are deproteinized, if necessary, with the minimal required amount of trichloroacetic acid and centrifuged to remove precipitated macromolecules. The precipitate is washed with water and centrifuged again. The combined supernatant solutions are applied to the top of a glass column, 1 cm X 22 cm, containing Dowex 50W (H +) X8 cation exchange resin (Bio-Rad analytical grade), 200-400 mesh. (The resin is used as purchased without treatment, except that the column is washed with 500 ml of glass-distilled water shortly before This volume [21]. G. I. Drummond,J. N. Aronson, and L. Anderson,J. Org. Chem. 26, 1601 (1961). J. B. Walker and M. S. Walker, Biochemistry 6, 3821 (1967).
442
ANTIBIOTIC BIOSYNTHESIS
[23]
use.) The column is washed with water, and successive stepwise elutions are performed, employing in sequence 0.5 N, 1.0 N, 2.0 N, 2.5 N, and 5.0 N concentrations of HC1. Approximately 80-120 ml total volume of water or eluent are employed at each step. Column fractions are 3-4 ml each (timed collection, 5-6 rain/tube) ; 5 ~l from each tube are spotted on a separate filter paper square. The paper is dried with a hot air blower, placed flat on the bottom of a scintillation bottle containing 5 ml of Liquifluor, and the sample is counted. Column fractions containing the desired labeled component are pooled in a ceramic evaporating dish, which is placed in a large vacuum desiccator adjacent to two petri plates containing pellets of CaCl~ and NaOH, respectively; the desiccator is carefully evacuated with a vacuum pump. The next day the dry fractions are taken up in a small volume of water with the aid of Pasteur pipettes and stored frozen. Most intermediates, except keto derivatives, are relatively stable for years in the frozen state. The elution behavior of various intermediates, using this protocol, are summarized in the table2 Tris buffer should not be used in an incubation when the compound to be prepared would be eluted in the 1.0 N HC1 fraction. Properties
Biological Distribution. L-Glutamine:keto-scyllo-inositol aminotransferase has also been found in S. bikiniensis ATCC 11062, S. griseus ATCC 12475, and S. ornatus ATCC 23265. In streptomycin producing strains, L-glutamine:keto-scyllo-inositol aminotransferase can be distinguished from L-alanine: 1D-l-guanidino-l-deoxy-3-keto-scyUo-inositol aminotransferase 9 by separation of the two enzymes on a Sephadex G-100 column, as well as by differential heat inactivation of the latter enzyme2 Specificity. 1° From stability studies (see below), pyridoxal-P appears to be a required cofactor. L-Glutamine appears to be the physiological amino donor. Very high concentrations of L-alanine and L-glutamate have slight donor activity. Aminodeoxy-scyllo-inositol and 1D-4-amino-4deoxy-myo-inositol are excellent amino donors. Streptamine (1,3-diamino-l,3-dideoxy-scyllo-inositol) and 2-deoxystreptamine are also good amino donors; it is not known which one of their two amino groups undergoes transamination. Relatively inactive amino donors include: L-asparagine, L-aspartate, D-alanine, D-glutamine, L-isoglutamine, L-glutamic-7-hydroxamate, glycine, 2-amino-2-deoxy-neo-inositol, and 2-amino-2-deoxy-myo-inositol. Among the amino acceptors, 1D-4-keto-myo-inositol is an excellent substrate. This compound can be prepared as the racemic mixture [myo9This volume [27]. loj. B. Walker, Lloydia 34, 363 (1971).
[23]
KETO-scyllo-INOSlTOL AMINOTRANSFERASE
443
inosose-4(6)] by heating myo-inositol with fuming nitric acid in a platinum crucible. 5 Pyruvate and ~-ketoglutarate at high concentrations can also serve as amino acceptors. This latter property is of practical interest, since labeled keto-scyllo-inositol can be prepared from labeled aminodeoxy-scyllo-inositol by transamination with pyruvate, followed by successive passage of the reaction mixture through Dowex 50 (H +) and Dowex 1 (C1-) columns. Labeled keto-scgllo-inositol appears in the water washes from both columns. It should be noted that keto-inositols are very unstable in alkaline solutions. Other Assays. Several additional assay procedures can be envisioned. For example, either of the two coupled enzyme assays previously described ~ could be modified by adding an excess of myo-inositol 2-dehydrogenase (Sigma Cat. No. I 5503) in order to make the transamination step rate-limiting. Alternatively, keto-scyllo-inositol (Sigma Cat. No. I 5375) could be transaminated with uniformly labeled L-['~C]glutamine, and the labeled a-ketoglutaramate formed could be measured by its decarboxylation in the presence of eerie sulfate2 On occasion the aminodeoxy-scyllo-inositol formed in Eq. (1) has been measured in a multistep assay by its conversion to 1-[~4C]guanidino-l-deoxy-scgllo-inositot 4-phosphate. Stability. After mycelial extracts of S. bikiniensis have been dialyzed overnight at 4 ° in the absence of pyridoxal-P, the aminotransferase activity of Eq. (1) cannot be detected, even when pyridoxal-P is included in the incubation mixture. Aminotransferase activity can be retained if pyridoxal-P is present during dialysis." Presence of pyruvate during dialysis spares much of the pyridoxal-P requirement. Loss of activity during dialysis is not prevented by the presence of pyridoxamine-P, pyridoxal, or L-alanine. Presumably these results reflect a need to keep this cofactor in the aldehyde form for optimal binding through Schiff base formation with a lysine residue at the active site. Furthermore the apoforms of certain pyridoxal-P dependent enzymes have recently been shown to be susceptible to degradation by specific proteolytic enzymes in other organisms, and this might be applicable in Streptomyces, especially since these bacteria have very active proteolytic enzymes. S. bikiniensis glutamine: keto-scyllo-inositol aminotransferase is relatively resistant to heating at 55 ° for 5 rain, in contrast to L-alanine: 1D1-guanidino-l-deoxy-3-keto-scyllo-inositol aminotransferase," which is inactivated by this treatment2 °
444
ANTIBIOTIC BIOSYNTHESIS
[24]
[24] A T P : I n o s a m i n c P h o s p h o t r a n s f e r a s c ( s )
By
1
JAMES B . WALKER
OH
OH
+ MgATP
~ MgADP +
H
4
i
(i)
z Osp OH
OH
Aminodeoxyscyllo- inositol ( s cyllo- inosamine)
'
7~+ NH~
NH,+
I C:NH +
NH~
I__,
+
(2) HO ~
H OH
OPO~-
1 D- 1-Guanidino- 3-amino1, 3- dideoxy-s cyllo - inositol
(N-amidinostreptamine) OH
OH
+ HO
2-
~
MgATP
~ MgADP
+
6
s
(3)
2-OsP
Deoxystreptamine
Equation (1) is a reaction involved in biosynthesis of both the bluensidine moiety of bluensomycin and the streptidine moiety of streptomycinY-4 Equation (2) is a reaction involved in biosynthesis of the streptidine moiety of streptomycin 5 (for biosynthetic pathway see an earlier 1EC 2.7.1.65. ~J. B. Walker and M. S. Walker, Biochemistry 6, 3821 (1967). 3j. B. Walker, L l o y d i a 34, 363 (1971), 4j. B. Walker, J. Biol. C h e m . 249, 2397 (1974). 5j. B. Walker and M. S. Walker, Biochem. Biophys. Res. C o m m u n . 26, 278 (1967).
[24]
ATP :INOSAMINE PHOSPHOTRANSFERASE(S)
445
articleG). In Eq. (3) 2-deoxystreptamine or streptamine serves as an analog of a physiological phosphate acceptor. 2 Of all the enzymatic activities known to be involved in the biosynthesis or enzymic modification of the streptomycin family of antibiotics, the activities of Eqs. (1) through (3) are the most difficult to demonstrate consistently in cell-free extracts. Optimal conditions for growing and extracting mycelia to give reproducible, active inosamine kinase preparations have not yet been established.
Assay Methods Method I E
Aminodeoxy-scyllo-[~4C]inositol + MgATP ---> 1-amino-l-deoxy-scyllo-[14C]inosit.ol-4-P + MgA1)P (4) J
lD-l-[~4C]Guanidino-3-amino-l,3-dideoxy-scyllo-inositol + MgATP --~ 11)-1-[14C]guanidino-3-amino-1,3-dideoxy-scyllo-inositol-6-P + MgAI)P (5) Principle. Labeled phosphorylated product is separated from labeled substrate by paper chromatography or high-voltage paper electrophoresis and counted. L-Ornithine is included in the incubation mixture to inhibit subsequent transamidination of the phosphorylated inosamine product. Reagents Extract of Streptomyces bikiniensis ATCC 11062 1-Amino-l-deoxy-scyllo- [ 1-1'~C]inositol, 3.5 Ci/mole, from California Bionuclear Corp., Sun Valley, Calif. (scyllo-inosamine[1-14C], Cat. No. 72673). Labeled compound can also be prepared as described in an earlier article. 7 1D-l- [ 14C] Guanidino-3-amino-l,3-dideoxy-scyllo-inositol (N-amidinostreptamine), prepared as described ~ Tris C1, 0.5 M, containing 40 mM MgCl~, pH 7.4 L-Ornithine, 80 mM 2-Mercaptoethanol, 0.3 M ATP, 36 mM, pH 6.8-7.0
Procedure. The complete incubation mixture contains: either aminodeoxy-scyllo- [ 1~C] inositol or 1D- 1- [1~C] guanidino-3-amino- 1,3-dideoxyThis volume [21]. 7This volume [23]. 8This volume [27].
446
ANTIBIOTIC BIOSYNTHESIS
[24]
scyllo-inositol, 5 td; Tris-Mg buffer, 5/zl; ornithine, 5 ~1; mercaptoethanol, 1 tL1; ATP, 5/zl; and dialyzed extract of S. bikiniensis, 10 t~l. After incubation in a stoppered 13 X 100 mm test tube at 35 °, 10 ~l are spotted (a) at the 22-cm segment and the components separated by high-voltage paper electrophoresis at pH 3.6, 9 or (b) on a paper chromatogram and developed with ammoniacal phenol (see below). Mobilities are given in the table of an earlier article2 Growth of Mycelia. The procedures for maintaining slants and growing and harvesting mycelia of S. bikiniensis are similar to those described for S. glebosus2 However, most work involving S. bihiniensis has utilized mycelia grown for 2-3 days on a medium of 2% peptone-0.2% yeast extract-tap water. Preparation of Extracts. Extracts of S. bikiniensis, as well as other streptomycin-producing strains, can be prepared by sonication, as described2 However, another extraction procedure, which requires no special equipment and can be readily scaled up to process large quantities of mycelia, can be employed for extraction of many enzymes, including aminotransferases, kinases, amidinotransferase, and phosphatases. This latter method utilizes hydrolysis of a portion of the peptidoglycan cell wall with lysozyme plus EDTA. Most strains of Streptomyces form a DNA gel which prevents separation by centrifugation during extraction with lysozyme; this gel can be dissolved by adding a small amount of deoxyribonuclease and Mg 2÷. For some reason lysozyme extracts of S. bikiniensis do not form this gel. In this procedure, frozen 2- to 3-day mycelial pads of S. bikiniesis are shaved with scissors into a beaker and extracted at room temperature for 1 hr with 3 volumes of 100 mM potassium phosphate containing 13 mM EDTA and 1 mg/ml crystalline egg-white lysozyme, pH 7.4. The mixture, immersed in a water bath, is stirred either continuously or intermittently during cell-wall digestion. The resulting suspension is centrifuged for 30 min at 30,000 g and 2 °. Except for the case of aminotransferases, 7,8 the supernatant solution is dialyzed overnight at 4 ° against 4 liters of 3 mM Tris C1, pH 7.4, to which 0.1 ml of mercaptoethanol has been added. The dialysis tubing had previously been treated with EDTA and stored in deionized water at 4 ° . The dialyzed solution is stored in the freezer, where it is stable for several months. Paper Chromatography. The RI values listed in the table of an earlier article 6 were obtained by ascending paper chromatography with Whatman No. 1 filter paper, developed with a solvent of 80% phenol-20% H_~O, NH40H atmosphere. Paper is cut in sheets 23 cm wide and 27 cm long, This volume [22].
[24]
ATP :INOSAMINE PHOSPHOTRANSFERASE(S)
447
and ruled in pencil as follows: (a) vertically, 5 lines, 5 cm apart, with margins of 1.5 cm at each side; (b) horizontally, 25 lines, 1 cm in width, starting 2 cm from the bottom. Samples are spotted with a 5-td micropipette, which is rinsed 3 times with water between samples, in the center of each of the bottom 1-cm segments. Samples are dried with a portable hair drier, and additional aliquots are applied as desired in 5-td increments. The paper is rolled into a cylinder, edges not touching, and stapled at two places, 7 cm from each end. Separation is performed in a cylindrical glass chromatography jar, 6 inches X 12 inches (e.g., Corning 431428) with ground top lightly greased and covered with a weighted round glass plate during a separation. To conserve developing solvent, an inverted shallow petri dish cover is placed on the bottom of the jar, and a petri dish cover (100 mm }( 20 mm) containing 80% phenol-20% H20 to a depth of less than 1 cm is placed on top of the inverted dish. After 0.8 ml of concentrated NH40H is added by pipette along the inside wall of the jar, the bottom end of the paper cylinder is placed in the phenol, and the cover is placed on top of the jar. The solvent usually reaches the top of the paper within 16 hr at room temperature. The paper is then removed from the jar, and its lower end is placed on paper toweling for 10 rain. The paper is then clamped, lower end up, with a stainless steel clamp to a horizontal glass rod suppogted by a ring stand, and dried in a hood. Drying takes several hours at room temperature, or less if portable hair dryers are used to blow heated air on the suspended sheet. The horizontal 1-cm segments along each path are cut into 3 pieces, placed flat on the bottom of a bottle containing 5 ml of Liquifluor scintillation fluid in toluene, and counted with a liquid scintillation system.
Method II Inosamine -~- MgATP E MgADP + inosamine-P (6) F Inosamine-P + L-[guanidino-l~C]arginine--~ L-ornithine + N-[14C]amidinoinosamine-P (7)
Principle. This assay involves coupling of ATP:inosamine phosphotransferase with L-arginine:inosamine-P amidinotransferase. TM The inosamine substrate of Eqs. (1), (2), or (3) is not labeled; label is introduced into the final product from commercially available L-[guanidino14C] arginine. Labeled product is separated by paper chromatography and counted. Some dephosphorylation of the amidinated product can occur if phosphatases are present. 10This volume [25].
448
ANTIBIOTIC BIOSYNTHESIS
[24]
Reagents Dialyzed extract of S. bik~iniensis ATCC 11062 (see Method I) Tris C1, 0.5 M, containing 40 mM MgCI~, pH 7.4 ATP, 36 mM, pH 6.8-7.0 2-Mercaptoethanol, 0.3 M L- [Guanidino-~4C] arginine, 10-40 Ci/mole, 33 t~Ci/ml Inosamine derivative, 5 mM (aminodeoxy-scyllo-inositol, monoamidinated streptamine, streptamine, or 2-deoxystreptamine)
Procedure. The complete incubation mixture contains: inosamine derivative, 5 t~l; [guanidino-~4C]arginine, 5 ~l; Tris-Mg, 5 ~l; mercaptoethanol, 1 ~l; ATP, 5 ~l; and dialyzed extract of S. bikiniensis, 10 ~l. After incubation in a stoppered 13 X 100 mm test tube at 35 °, 10-~l aliquots are spotted on a paper chromatogram, and the components are separated with ammoniacal phenol (Method I) and subsequently counted. Mobilities of the expected products ~ are given in Table I of an earlier article2 For purification purposes, an excess of purified inosamine-P amidinotransferaseTM would have to be added if this assay is employed. Preparation o] Aminodeoxy-scyllo-inositol. Any one of several procedures can be employed. For example, aminodeoxy-scyllo-inositol can be prepared from bluensomycin by mild acid hydrolysis to give bluensidine (1D-l-O-carbamoyl-3-deoxy-3-guanidino-scyllo-inosi~ol). Alkaline hydrolysis of bluensidine gives aminodeoxy-scyllo-inositol. 11 Alternatively, the hydrolysis could be performed in reverse order. Alkaline hydrolysis of bluensomycin gives deamidinodecarbamoylbluensomycin, which would give aminodeoxy-scyllo-inositol on subsequent acid hydrolysis. If bluensomycin is not available, aminodeoxy-scyllo-inositol can be chemically synthesized either from myo-inositol by way of an azido de- , rivative or from keto-scyllo-inositol (myo-inosose-2, Sigma Cat. No. I 5375) by reduction of its oxime with sodium amalgam22 Preparation o] Streptamine and Streptidine ]rom Dihydrostreptomyc/n. Mild acid hydrolysis of dihydrostreptomycin gives streptidine 13,~4 (1,3-diguanidino-l,3-dideoxy-scyllo-inositol), which can be converted to streptamine (1,3-diamino-l,3-dideoxy-scyUo-inositol) by alkaline by1~B. Bannister and A. D. Argoudelis,J. Amer. Chem. Soc. 85, .119 (1963). 12L. Anderson and H. A. Lardy, J. Amer. Chem. Soc. 72, 3141 (1950). ~' R. L. Peck, R. P. Graber, A. Walti, E. W. Peel, C. E. Hoffhine,Jr., and K. Folkers, J. Amer. Chem. Soc. 68, 29 (1946). ~ M. S. Walker and J. B. Walker, J. Biol. Chem. 241, 1262 (1966}.
[24]
A T P : INOSAMINE PHOSPHOTRANSFERASE(S)
449
drolysis. 1~,15 A covered beaker containing 107 g of dihydrostreptomycin sulfate dissolved in 360 ml of 1 N H~S04 is placed in a 37 ° water bath. The reaction mixture is incubated for 2 days, and the crystals of streptidine sulfate are collected by suction filtration with a sintered glass filter, washed twice with a small amount of cold water, washed with acetone, and dried. Yield: 47 g of crude streptidine sulfate. Several successive recrystallizations are perfo¢med by dissolving, e.g., 108 g of crude strew tidine sulfate in 1600 nfl of boiling water slightly acidified with H2S04, placing the solution in the refrigerator overnight, and collecting, washing, and drying the crystals as before. Streptamine sulfate is prepared by refluxing for 24-48 hr a mixture of 25 g of streptidine sulfate and 1250 ml of a saturated aqueous solution of barium hydroxide. Then 1 N H..SO~ is slowly added with stirring to the hot mixture until the pH stabilizes at 5-6 (about 415 ml required). The mixture is filtered while hot, and the filtrate is cooled. To the chilled filtrate 0.4 volume of acetone is slowly added, and the suspension is filtered under suction. The precipitate is washed three times with a total of 30 ml of cold water, washed with acetone, and dried. Yield: 11.5 g of streptamine sulfate. Preparation o] Monoamidinated Streptamines. To a covered beaker is added 8.1 g of streptamine sulfate and 5.2 g of barium hydroxide in 125 ml of hot water. The mixture is digested on a steam bath for 5 min, filtered by suction, and the barimn sulfate precipitate is washed with hot water. The filtrate plus washings are evaporated by boiling to a volume of 25-50 ml, and the beaker is placed in a hood. Powdered S-methylisothiouronium sulfate is added, with stirring, in portions over several days; 2 g are added initially, with a total of 2 g more added over several days. After 6 days the mixture is neutralized with 2 N H..,SO~. Fractional precipitation with increasing amounts of acetone give several precipitates, which are then assayed for guanidino compounds by the Sakaguchi reaction. A late-precipitating acetone fraction usually contains monoamidihated streptamines relatively free from streptamine and streptidinc. Preparation o] 2-Deoxystreptamine ]rom Kanamycin. ~,~6 To 5 g of kanamycin sulfate (Schwarz/Mann) in 10 ml of hot water is slowly added, with stirring, 2.2 g of barium hydroxide. After 15 min, the mixture is centrifuged. A saturated aqueous solution of barium hydroxide is slowly added, with stirring, to the supernatant solution plus washing until no more barium sulfate precipitate is formed. After centrifugation, the 1~R. L. Peck, C. E. Hoffhine, Jr., E. W. Peel, R. P. Graber, F. W. Holly, R. Mozingo, and K. Folkers, J. Amer. Chem. Soc. 68, 776 (1946). 1~M. J. Cron, D. L. Johnson, F. M. Palermiti, Y. Perron, H. D. Taylor, D. F. Whitehead, and I. R. Hooper, J. Amer. Chem. Soc. 80, 752 (1958).
450
ANTIBIOTIC BIOSYNTHESIS
[24]
supernatant solution is refiuxed for 75 min with an equal volume (ca. 25 ml) of concentrated HC1. The hydrolyzate is decolori~ed with acidwashed charcoal and filtered; the filtrate is concentrated in a vacuum desiccator over NaOH pellets. Ethanol is added slowly to the concentrated solution until turbidity appears. On chilling, an oil is formed which is converted to a :precipitate by the addition of methanol. The precipitate is washed twice with cold 75% methanol, then acetone, and dried. Yield: 560 mg of deoxystreptamine dihydrochloride.
Properties
Purification. Inosamine kinase activities, concentrated by treatment of an extract with Mn 2÷ and precipitation with (NH4)2S0~, can be separated from most other enzymes involved in streptidine biosynthesis by (a) Sephadex G-100 column chromatography, 17 (b) batch treatment with DEAE-cellulose, is or (c) DEAE-cellulose column chromatography, as described. 1° A mercaptan should be present at all stages of purification. Biological Distribution. Inosamine kinase activities are believed to occur in all strains of Streptomyces which synthesize the streptomycin family of antibiotics, s° However, as stated earlier, optimal conditions of growth and enzyme extraction to obtain consistently active preparations have not yet been determined. Specificity. There are two separate specificity problems. One concerns whether more than one enzyme is responsible for catalyzing Eqs. (1), (2), and (3) ; if more than one enzyme is involved, each enzyme presumably has its own characteristic substrate specificity. At present it can only be stated that streptamine is also an excellent phosphate acceptor. 2 Guanidinodeoxy-scyllo-inositol and streptidine ~8 are not phosphate acceptors, nor is the enantiomer of the acceptor of Eq. (2). At higher concentrations it appears that a number of isomeric inosamines can accept a phosphate group and subsequently be transamidinated when crude extracts are the source of enzymes. These findings must, however, be checked with more purified enzyme preparations, because of the probable presence of epimerases in crude extracts. In crude extracts ATP and dATP can serve as phosphate donors, whereas CTP, UTP, dTTP, and G T P are inactive. 2,~ Either Mg 2÷ or Mn 2÷ is required as cofactor. Assignment of the phosphate group para to the amino group which is subsequently to be transamidinated is not based on direct evidence. The possibility that the phosphate group is initially esterified at another position 1~A. L. Miller and J. B. Walker, J. Bacteriol. 99, 401 (1969). 18j. B. Walker and M. S. Walker, Biochim. Biophys. Acta 148, 335 (1967).
[25]
L-ARGININE :INOSAMINE-P AMIDINOTRANSFERASE(S)
451
and transferred to the para position by a phosphomutase has not been ruled out, but is considered unlikely?" Alternate Assay. With purified enzyme preparations, a radiochemical assay can be developed which utilizes [V-32P]ATP as the phosphate donor. Small columns containing Dowex 1 (C1-) resin can be used; labeled product is not adsorbed on such columns, whereas labeled substrate and inorganic phosphate are adsorbed. This rapid assay is also useful for determining acceptor specificity, since candidate substrates are not usually available in labeled form. Inhibition. Sulfhydryl reagents such as p-chloromercuribenzoate and formamidine disulfide inhibit these kinase activities. Purified enzyme preparations are stabilized by the addition of serum albumin and mercaptans. 1~M. S. Walker and J. B. Walker, J. Biol. Chem. 246, 7034 (1971).
[2 5] L - A r g i n i n e : I n o s a m i n e - P A m i d i n o t r a n s f e r a s e (s)
By JAMES B. WALKER H
L -
NH~ F, K _
Arginine +
L-Ornithine +
2-OsP OH R = (a) - - O H ;
(c)
- 0 - C(=
--- ~3
(1) OH
(d) --
(b) --NH-- C(=NH+)NI-I, ;
60~ R z-Osp
NH~
N H 2 ; and their corresponding 2-deoxy derivatives
O)NIi~;
L-Canavanine -[- L-ornithine ~- L-canaline + L-arginine
(2)
L-Arginine + N H ~ O H - * L-ornithine -k H 2 N - - C ( = N H 2 + ) N H - - O H
(3)
Prior to the discovery of this enzyme in extracts of mature mycelia of certain Streptomyces, 2,3 the only known amidinotransferases were those involved in the biosynthesis of creatine and certain other phosphagens in higher animals. 4,5 Rapid and convenient colorimetric assays have been developed for amidinotransferases which are based on reactions re1See also this series, Vol. 17A [146]. 2 j. B. Walker, J. Biol. Chem. 231, 1 (1958). s M. S. Walker and J. B. Walker, J. Biol. Chem. 241, 1262 (1966). 4 j. B. Walker, in "The Enzymes" 3rd ed. (P. Boyer, ed.), Vol. 9, p. 497. Academic Press, New York, 1973. E. Grazi and F. Conconi, this series, Vol. 17A [145].
[25]
L-ARGININE :INOSAMINE-P AMIDINOTRANSFERASE(S)
451
and transferred to the para position by a phosphomutase has not been ruled out, but is considered unlikely?" Alternate Assay. With purified enzyme preparations, a radiochemical assay can be developed which utilizes [V-32P]ATP as the phosphate donor. Small columns containing Dowex 1 (C1-) resin can be used; labeled product is not adsorbed on such columns, whereas labeled substrate and inorganic phosphate are adsorbed. This rapid assay is also useful for determining acceptor specificity, since candidate substrates are not usually available in labeled form. Inhibition. Sulfhydryl reagents such as p-chloromercuribenzoate and formamidine disulfide inhibit these kinase activities. Purified enzyme preparations are stabilized by the addition of serum albumin and mercaptans. 1~M. S. Walker and J. B. Walker, J. Biol. Chem. 246, 7034 (1971).
[2 5] L - A r g i n i n e : I n o s a m i n e - P A m i d i n o t r a n s f e r a s e (s)
By JAMES B. WALKER H
L -
NH~ F, K _
Arginine +
L-Ornithine +
2-OsP OH R = (a) - - O H ;
(c)
- 0 - C(=
--- ~3
(1) OH
(d) --
(b) --NH-- C(=NH+)NI-I, ;
60~ R z-Osp
NH~
N H 2 ; and their corresponding 2-deoxy derivatives
O)NIi~;
L-Canavanine -[- L-ornithine ~- L-canaline + L-arginine
(2)
L-Arginine + N H ~ O H - * L-ornithine -k H 2 N - - C ( = N H 2 + ) N H - - O H
(3)
Prior to the discovery of this enzyme in extracts of mature mycelia of certain Streptomyces, 2,3 the only known amidinotransferases were those involved in the biosynthesis of creatine and certain other phosphagens in higher animals. 4,5 Rapid and convenient colorimetric assays have been developed for amidinotransferases which are based on reactions re1See also this series, Vol. 17A [146]. 2 j. B. Walker, J. Biol. Chem. 231, 1 (1958). s M. S. Walker and J. B. Walker, J. Biol. Chem. 241, 1262 (1966). 4 j. B. Walker, in "The Enzymes" 3rd ed. (P. Boyer, ed.), Vol. 9, p. 497. Academic Press, New York, 1973. E. Grazi and F. Conconi, this series, Vol. 17A [145].
452
ANTIBIOTIC BIOSYNTHESIS
[25]
lated to the arginine:ornithine exchange reaction 6 catalyzed by these enzymes. 2 Either one of the assays indicated in Eqs. (2) 2'7 and (3) 2,8 can be utilized (a) for detecting strains of Streptomyces capable of synthesizing the streptomycin family of antibiotics, and (b) for studying the mechanisms of controls which govern differentiation of these strains to the antibiotic-synthesizing state after a phase of rapid vegetative growth. In this article assays will be described based on each of the above equations. It is not yet known whether one enzyme9 or two different enzymes1° catalyze reactions F and K in the reaction scheme shown in an earlier article2 ~
Assay Methods Method 18,9 Principle. This assay is based on Eq. (1). L-[Guanidino-l*C]arginine donates its labeled amidino group to a natural or synthetic inosamine-P derivative, and the labeled product is separated by paper chromatography and counted. 1~ This method, unlike the other methods to be described, is specific for inosamine-P amidinotransferase. Reagents Dialyzed extract of mature mycelia of Streptomyces bikiniensis ATCC 11062 or other strain of Streptomyces which synthesizes one of the streptomycin family of antibiotics L- [Guanidino-14C] arginine, 12-30 Ci/mole Tris C1,0.5 M, containing 10 mM EDTA, pH 7.4 2-Mercaptoethanol, 0.3 M Solution containing chemically phosphorylated inosamine derivative or hot water extract of mature mycelia of a producing strain grown in presence of 0.5 % myo-inositol
Procedure. The complete incubation mixture contains: labeled arginine (33 gCi/ml), 5 gl; Tris-EDTA, 5 t,1; mercaptoethanol, 1 t*l; phosphorylated inosamine derivative, 10 gl; and dialyzed extract of S. bikiniensis or S. glebosus ATCC 14607, 10 ~1. After incubation in a stoppered ej. B. Walker, J. Biol. Chem. 221, 771 (1956). J. B. Walker, Biochim. Biophys. Acta 73, 241 (1963). s j. B. Walker, d. Biol. Chem. '235, 2357 (1960). ~J. B. Walker, J. Biol. Chem. 249, 2397 (1974). lo L. C. Pla, Biochim. Biophys. Acta 242, 541 (1971). 11This volume [21]. ~2This volume [24].
[25l
L-ARGINI NE :INOSAMINE-P AMIDINOTRANSFERASE (S)
453
13 X 100 mm test tube at 35 °, a 10-t,1 aliquot is spotted on a paper chromatogram, and the components are separated with ammoniacal phenol and counted. 12 Mobilities of expected products are given in Table I of an earlier article. 11 Excessive amounts of inorganic phosphate markedly lower the Rr values. The presence of EDTA inhibits action of phospharases on the reactants, and in the case of chemically phosphorylated acceptors chelates inhibitory Ba 2÷. Streptomyces Extracts. Mycelia are grown and harvested as described. 13 Dialyzed extracts can be prepared from sonicates TM or from lysozyme extracts22 Nondialyzed supernatant solut!ons from sonicates can be utilized if desired as the source of both amidinotransferase and phosphorylated acceptors2 In such cases the addition of carbamoyl-P to the incubation mixture often markedly improves the yield of labeled product. 9 Preparation o] Natural Acceptors. Mycelia of a strain capable of synthesizing streptomycin or bluensomycin are grown (with 0.5 to 1.0% myo-inositol included in the growth medium), harvested, and stored as described. 1~ Freshly harvested, or frozen, mycelia are added to an equal weight of hot water contained in a centrifuge tube immersed in a boiling water bath. The mixture is stirred for 7 to 10 rain with a glass stirring rod, then cooled and centrifuged. The supernatant solution is stored in the freezer. The hot-water extracts with the highest levels of acceptors are determined by assay with an active dialyzed amidinotransferase preparation (Method I).
Preparation of Chemically Phosphorylated Inosamine Derivatives. 3 The nonspecific procedure of Plimmer and Burch TM is employed to produce mixtures of phosphorylated isomers of a given inosamine derivative. The crude reaction mixtures after removal of most phosphate as the barium salt, are used as sources of amidino acceptors, without further purification. Amidinotransferase is believed to react with only one phosphorylated isomer in each case. Inosamine derivatives which can be phosphorylatcd to give amidino acceptors include aminodeoxy-scyllo-inosamine, streptamine, 2-deoxystreptamine, 2-amino-2-deoxy-neo-inositol, streptidine, and 2-deoxystreptidine. Commercially available antibiotics can also be directly utilized. For example, phosphorylation of dihydrostreptomycin gives a mixture which includes the correct streptidine-P isomer, and phosphorylation of kanamycin gives a mixture which includes the correct 2-deoxystreptamine isomer; in both instances hydrolysis of glycosidic bonds accompanies phosphorylation. Amidinotransferase plus a-[guani13This volume [22]. 1~R. H. A. Plimmer and W. J. N. Burch, Biochem.J. 31,398 (1937).
454
ANTIBIOTIC BIOSYNTHESIS
[25]
dino-14C]arginine react with the correct isomer in a mixture, and product labeled in the guanidino group can then be separated on a Dowex-50 (H ÷) column 15 (see Table I of an earlier article11). In a typical preparations, 2-5 g of the inosamine derivative are added to a 50-ml round-bottom flask containing 10 ml of concentrated H3P04. Then 5 g of powdered P205 are quickly added, with minimal exposure to the atmosphere, and a loosely stoppered reflux condenser is immediately joined to the flask. The mixture is heated on a steam bath for 6 hr and left at room temperature overnight. A hot saturated solution of Ba (OH)2 is then slowly added to the mixture, with thorough stirring, to give a slightly alkaline pH. The completely neutralized mixture is filtered, and the precipitate is washed several times with hot water. The original filtrate and each successive wash are separately frozen for subsequent assay for amidino acceptor activity. NazS04 is added to filtrates, if necessary, to precipitate any Ba 2+ remaining in solution. These preparations should be assayed in the presence of EDTA.
Method IP Principle. This assay is based on Eq. (2). Canavanine is a naturally occurring analog of arginine, so this reaction is analogous to the arginine: ornithine exchange transamidination catalyzed by both animal glycine amidinotransferase 6 and bacterial inosamine-P amidinotransferase. ~ The arginine formed is assayed by any of several modifications of the Sakaguchi reaction. If desired, canaline can be removed as it is formed by formation of a ketoxime with acetone to make this readily reversible reaction unidirectional/ It should be noted that streptomycin, bluensomycin, and their guanidinated precursors also react with the Sakaguchi reagent, so dialyzed extracts or purified enzyme preparations must be used. Reagents Dialyzed extract of mature mycelia of S. bikiniensis ATCC 11062 Potassium phosphate, 1 M, pH 7.5 L-Canavanine sulfate, adjusted to pH 7.5 with KOH, 0.25 M L-Ornithine.HC1, 0.1 M Acetone, 20% (v/v) in water Trichloroacetic acid, 30% (w/v) in water Sodium hydroxide, 10% (w/v) in water ~-Naphthol, 1 mg/ml, in 95% ethanol ~ This volume [23].
[25]
L-ARGININE :INOSAMINE-P AMIDINOTRANSFERASE(S)
455
Urea, 20% (w/v) in water Clorox, 50% (v/v) in water
Procedure. The complete incubation mixture contains, in a final volume of 1.0 ml: canavanine, 0.2 ml; phosphate, 0.1 ml; ornithine, 0.2 ml; acetone solution, 0.1 ml; and dialyzed extract, After incubation in a stoppered 13 X 100 mm test tube at 35% the mixture is treated with 0.3 ml of trichloroacetie acid solution. After 10 rain the mixture is centrifuged, and an aliquot of the supernatant solution is added to a 20 X 150 mm test tube and made up to 5.0 ml with water; 1 ml of NaOH solution and 1 ml of a-naphthol solution are added and mixed thoroughly. After 5 min, the Clorox solution is added, and the tube is shaken immediately after addition; 60 sec later 2 ml of urea solution are added with shaking. Absorbancy at 540 nm is measured against a control consisting of the complete mixture minus ornithine. Method l i p Principle. This assay, the standard method used in our laboratory, is based on Eq. (3) and serves equally well as an assay for animal glycine amidinotransferase,s Hydroxylamine reacts with the enzyme-amidine ("active urea") intermediate 2,4 to give hydroxyguanidine, which is measured as its colored complex with pentacyanoaminoferrate. Crude extracts can be assayed provided that arginase activity is low, since ornithine strongly inhibits hydroxyguanidine formationJ Hydroxylamine reacts with esters and certain other acyl derivatives often present in crude extracts to form hydroxamates, which give a similar color with pentacyanoaminoferrate. Excessive amounts of EDTA or mercaptans can interfere with the color development. Reagents Extract of Streptomyces bikiniensis ATCC 1106212 L-Arginine.HC1, 1 M L-Ornithine.HC1, 0.6 M NH.,OH • HC1, neutralized with cold 2 M KOH, 2 M, stored frozen Potassium phosphate buffers, 1 M, pH 7.4 and pH 7.0, respectively Acetone Na3[Fe(CN)~NH~], 1% in water, aged at least 1 day (Fisher Cat. No. S-659) Hydroxyguanidine hemisulfate • H.20, 300 t~g/ml (Eastman Cat. No. 9241). Synthesis of this compound was described previously in this series. 1 Trichloroacetic acid, 30% (w/v) in water
456
ANTIBIOTIC BIOSYNTHESIS
[25]
Procedure. The complete incubation mixture contains, in a final volume of 1.0 ml: arginine, 0.1 ml; NH~OH, 0.3 ml; phosphate buffer, pH 7.4, 0.1 ml; and enzyme solution. For a blank, 0.1 ml of ornithine is substituted for arginine. For a standard, 0.5 ml of hydroxyguanidine is substituted for the enzyme solution. After incubation in a stoppered 13 X 100 mm test tube at 37 °, the reaction is stopped with 0.4 ml of trichloroacetic acid, and 1.0 ml of water is added. After 10 min the mixture is centrifuged in the same test tube, and a 1.5-ml aliquot of the supernatant solution is pipetted into another 13 X 100 mm test tube, followed by addition of 0.5 ml of water, 2.0 ml phosphate buffer, pH 7.0, 0.3 ml acetone, and 0.3 ml pentacyanoaminoferrate. The mixture is rapidly stirred, and after 10 min the absorbance is read at 480 nm. The color does not form if excess NH20H is not removed as the ketoxime of acetone. Definition of Unit and Specific Activity. A unit of amidinotransferase is defined as the amount that catalyzes the formation of 1 t~mole of hydroxyguanidine per hour in the assay described in Method III. Specific activity (units/mg of protein) is based on measurement of protein by the Lowry method? 6 Purification Procedure Growth and Extraction of Mycelia. Mycelia of S. bikiniensis ATCC 11062 are grown at 24-28 ° for 2 days on a medium of 2% peptone-0.2% yeast extrac~tap H20 and harvested as described. 1~ Frozen mycelial pads are extracted with lysozyme-EDTA as described. 12 The maximal amount of MnCl~ 'solution which can be employed without precipitating amidinotransferase is determined with an aliquot of the supernatant solution. Treatment with Mn 2÷ and (NH4)2SO~. To 70 ml of the supernatant solution is added, dropwise with stirring, 7 ml of 10% MnCl~ • 4H~O in 0.1 M phosphate buffer, pH 7.4. After 20 min the mixture is centrifuged. To 70 ml of the supernatant solution is added 70 mg of L-arginine.HCl and 7 tzl of mercaptoethanol; 17 g of powdered (NH~)2SO4 is next added in small portions, with stirring, over a 35-min period, with the pH maintained near 7.0 by cautious addition of dilute NH4OH. After 20 min the precipitate is removed by centrifugation. To 76 ml of the supernatant solution is added 15.5 g of powdered (NH4)2SO~, following the same procedure, but this time saving the precipitate. The precipitate is dissolved in 3 ml of 0.1 M phosphate buffer, pH 7.4, containing EDTA, 5 mg/ml, and dialyzed overnight against 4 liters of 1 mM phosphate ~60. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem. 193, 265 (1951).
[25]
L-ARGI NINE :INOSAMINE-P AMIDINOTRANSFERASE (S)
457
PURIFICATION OF INOSAMINE-P AMIDINOTRANSFERASF
Step
Total activity (units)
Total protein (rag)
Lysozyme extract Mil s+ s u p e r n a t a n t solution (NH4)~SO4 fraction, 4 0 - 7 0 % DEAE-cellulose, 0.3 M NaC1 eluate
672 490 367 280
840 350 204 28
Specific activity Recovery (units/mg) (%) 0.8 1.4 1.8 10.0
100 73 55 42
buffer, pH 7.4, containing 200 mg of EDTA and 0.1 ml of mercaptoetltanol. DEAE-CelIulose Chomatography. DEAE-eellulose (medium mesh, 92 mEq/g) is washed twice with 1 N NaOH and then water. The adsorbent is treated with 1 N HC1, stirred under suction until bubbles stop evolving, and washed with water. Fines are decanted at each step. A 58 X 2-cm column is packed and equilibrated with 10 mM potassium phosphate buffer, pH 7.0, containing EDTA, 0.5 mg/ml, and mercaptoethanol, 0.1 ~l/ml. The dialyzed (NH4)2SO~ fraction, 5.5 ml, containing 204 mg protein, is added to the column, followed by 70 ml of equilibration buffer; 6-ml fractions are collected. The protein is eluted with one column volume of 0.3 M NaC1 in the same buffer. Amidinotransferase is eluted with 0.3 M NaC1, separating it from ATP:inosamine phosphotransferasp, 12 which is eluted with 0.2 M NaC1, and from ATP:streptomycin 6-phosphotransferase, ~7 which is eluted with 0.6 M NaC1 in buffer. The purification procedure is summarized in the table.
Properties Biological Distribution. Amidinotransferase specific activity increases 50-fold after the phase of rapid growth in strains of Streptomyces which produce the streptomycin family of antibiotics, 18 including S. grise'us strains ATCC 12475, 10137, 11429, 11984, and 27001; S. humidus ATCC 12760; S. griseocarneus ATCC 12628; S. galbus ATCC 14077; S. ornatus ATCC 23265; and a bluensomycin-producing strain, S. hygroscopicus forma glebosus ATCC 14607. 9 Specificity and Alternative Assays. In addition to the specificities shown in Eqs. (1), (2), and (3), inosamine-P amidinotransferase can readily amidinate 2-amino-2-deoxy-neo-inositol-5-P, which surprisingly differs in configuration from the normal acceptors of Eq. (1) in positions ~TThis volume [49]. lsj. B. Walker and V. S. Hnilica, Biochim. Biophys. Acta 89, 473 (1964).
458
ANTIBIOTIC BIOSYNTHESIS
[25]
1, 2, 4, and 5. Evidently the relationship between positions 3 and 6 are most important for acceptor activity. Inosamine-P amidinotransferase, like glycine amidinotransferase, also amidinates the ornithine analogs glycylglycine and 1,4-diaminobutylphosphonic acid. 19 Inosamine derivatives must be phosphorylated to serve as substrates, and the phosphate must be esterified with the hydroxyl group para to the amino group to be transamidinated~°; derivatives phosphorylated in the ortho position are not active. 21 When L-[guanidino-~4C]arginine and amidinotransferase are incubated with either chemically phosphorylated streptidine, or with streptidine-P synthesized within mycelia following the addition of 1 mg/ml of streptidine sulfate to a mature culture of almost any strain of Streptomyces 3-10 hr prior to harvest, label is introduced by an exchange reaction into the 3-guanidino moiety of streptidine-6-P2 However, derivatives of streptidine-6-P in which position 4 is substituted with a bulky group, as in streptomycin-6-P, are not substrates for amidinotransferase. Since most amidino donors can couple with most amidino acceptors, a variety of additional assays can be employedA9 However, canavanine:NH20H transamidination is obscured by a nonenzymic reaction between these two compounds. When arginine is the donor, the ornithine formed can be measured by the method of Chinard. 5 Inhibitors. L-0rnithine is the most potent competitive inhibitor. Product inhibition by ornithine particularly interferes with kinetic studies, except in the case of Method II. If necessary, ornithine can be removed by enzymic conversion to citrulline in the presence of added carbamoyl-P and ornithine carbamoyltransferase. Strong inhibition occurs with the sulfhydryl reagents, formamidine disulfide. 2HC1 (freshly prepared aqueous solution, kept slightly acid at all times), cystine, cystamine, and p-chloromercuribenzoate,is
1,j. B. Walker, Lloydia 34, 363 (1971). =oM. S. Walker and J. B. Walker, J. Biol. Chem. 046, 7034 (1971). ~lj. B. Walker and M. Skorvaga,J. Biol. Chem. 248, 2441 (1973).
[26]
GUANIDINODEOXY-8eyIIo-INOSITOL-4-P PHOSPHOHYDROLASE
459
[26] 1-Guanidino-l-deoxy-scyllo-inositol-4-P Phosphohydrolase B y JAMES B. WALKER NH
OH
2
NH
~=NH +
2
L
H + H20
Mg2+ G ~ HPO~- +
2-OsPO ~
I~H
(1)
H OH
OH
This enzyme participates in the biosynthesis of streptomycin 1-3 but is apparently not required for biosynthesis of bluensomycin, although it is present in bluensomycin producers. 4 Transamidination products '~' substituted at position 5 with a guanidino group (streptidine-P) or an O-carbamoyl group (bluensidine-P) are not good substrates.
Assay Method Principle. 1- [ ~4C] Guanidino- 1-deoxy-s cgllo-inositol-4-P is incubated with an enzyme preparation, and the components of the reaction mixture are separated by paper chromatography. This assay can be used for both crude and purified preparations. Reagents Extract of mature mycelia of Streptomyces bikiniensis ATCC 11062, or other strain which synthesizes one of the streptomycin family of antibiotics
1- [~4C ] Guanidino-l-deoxy-scyllo-inositol-4-P Tris C1, 0.5 M, containing 40 m M MgCl~, pH 7.4 2-Mercaptoethanol, 0.3 M
Procedure. The complete incubation mixture contains: 1-[14C]guanidino-l-deoxy-scyllo-inositol-4-P, 10 t~l; Tris C1-Mg, 5 ~l; mercaptoethanol, 1 ~l; and enzyme preparation, 10 ~l. After incubation in a stoppered 13 }( 100 mm test tube at 35 °, a 10-t~l aliquot is spotted, and the compo1j. B. Walker and M. S. Walker, Biochem. Biophys. Res. Commun. 26, 278 (1967). 2 M. S. Walker and J. B. Walker, J. Biol. Chem. 246, 7034 (1971). 3j. B. Walker, Lloydia 34, 363 (1971). 4j. B. Walker, J. Biol. Chem. 249, 2397 (1974). This volume [251.
460
ANTIBIOTIC BIOSYNTHESIS
[25]
nents are separated by paper chromatography with ammoniacal phenol and counted2 For mobilities, see the table of a previous articleY Preparation o] 1-[14C]Guanidino-1-deoxy-scyllo-inositol-4-P. The simplest procedure is to grow mycelia of S. bikiniensis ATCC 11062, S. griseus ATCC 12475, or other streptomycin producing strain as described 6 with 0.5% myo-inositol added to the growth medium. (Do not use S. glebosus ATCC 14607.) Frozen mycelial pads are sonicated as described, 8 and 10-~l aliquots of the nondialyzcd supernatant solutions from various pads are incubated with L-[guanidino-~4C]arginine as described with a dialyzed extract in Method I of a previous article2 However, here the supernatant solutions are utilized as a source of both inosamine-P amidinotransferase and amidino acceptors. Generally two peaks of radioactive products will appear on ammoniacal phenol paper chromatograms2 The peak at an Ry of 0.17 is 1-[~4C]guanidino-l-deoxy-scyllo-inositol-4-P. The broad peak centering around R~ 0.40 is streptidine-P. It is advisable to assay a number of different mycelial pads, so that the best preparation can be utilized for a suitably scaled-up incubation. The labeled products from a scaled-up incubation are separated on a Dowex 50 (H ÷) column2 1-[~C]Guanidino-l-deoxy-scyllo-inositol-4-P is eluted with 0.5 N HC1, and labeled streptidine-P is eluted with 2.0 N HC1 (Table I). 7 Alternatively, chemically phosphorylated aminodeoxy-scyllo-inositol can be employed as the amid no acceptvr ~ with a dialyzed extract of a streptomycin or bluensomycin producer as the source of amidinotransferase, and the labeled product is isolated by the same procedure.
Separation o] 1-Guanidino-l-deoxy-scyllo-inositol-4-P Phosphatase ]rom Streptomycin-6-P PhosphataseY Both of these phosphatase activities can be readily demonstrated in dialyzed lysozyme extracts ~ or sonicates s of frozen mycel~al pads of streptomycin producing strains. The following procedure is typical, but not necessarily optimal. All operations are carried out at 4 ° or less. Frozen pads of mature mycelia of S. bikiniensis ATCC 11062, 18.7 g, are sonicated with an equal weight of water in a beaker for a total of 15 rain in 30-sec bursts, and centrifuged 20 min at 30,000 g. To 20 ml of supernatant solution is slowly added, with stirring, 6.7 ml of 10% MnCl_~ • 4H~O in 0.1 M Tris C1, pH 7.4. After 20 rain the mixture is centrifuged. Powdered (NH4)~S04 is added to the supernatant solution to bring it to 20% of saturation. After 15 min the mixture is centrifuged. The supernatant solution is slowly brought to 65% saturation with powdered (NH~):SO~ while the pH is adjusted to neutralThis volume [24]. 'This volume [21]. 8This volume [22]. 9This volume [23].
[26]
GUANImNODEOXY-scyllo-INOSITOL-4-PPHOSPHOHYDROLASE
461
ity with 1 M NH4OH. After 15 rain the mixture is centrifuged, and the precipitate is taken up in 5 ml of 0.1 M potassium phosphate buffer, pH 7.4, containing 10 mM EDTA, and dialyzed for 2 days against 4 liters of 1 mM phosphate buffer, pH 7.4, containing 0.1 mM EDTA and 0.1 ml of mercaptoethanol, with one change in external medium. A 0.5-ml aliquot of the dialyzed preparation, containing 17.5 mg of protein (Lowry), is diluted with 1 ml of 0.1 M Tris C1, pH 7.4, and applied to a Sephadex G-200 column, 3 cm )< 67 cm, previously equilibrated with 10 mM Tris C1, pH 7.4, plus mercaptoethanol (10 t~l/10 ml). The column is eluted at the rate of 1 drop per 37 sec with the equilibrating buffer; 2.5-ml fractions are collected, and 1.5 mg of bovine serum albumin is added to odd-numbered tubes, starting with tube 61, to preserve activity of guanidino-deoxy-scyllo-inositol-P phosphatase. The fractions are stored frozen. Streptomycin-6-P phosphatase activity 1° is almost completely excluded and is eluted between tubes 45 and 60 (peak at 53), with the bulk of the protein. Guanidinodeoxy-scyllo-inositol-P phosphatase activity is eluted in a broad peak centering on tube 111.
Properties
Biological Distribution. This enzyme apparently occurs in all strains of Streptomyces which produce the streptomycin family of antibiotics, 5 including bluensomycin producers. 4 It is not yet known whether the similar enzyme detected in S. hygroscopicus ]orma glebosus ATCC 14607 differs in its molecular properties from the corresponding enzyme in streptomycin-producing strains. Specificity. 1-Guanidino-l-deoxy-scyllo-inositol-4-P phosphatase acts equally well on 2-[14C]guanidino-2-deoxy-neo-inositol-5-P, ~- formed by transamidination of chemically phosphorylated 2-amino-2-deoxyneo-inositol derived from hygromycin. 1D-[l'C]Guanidino-5-amino-l,5dideoxy-scyllo-inositol-4-P, formed by transamidination of phosphorylated streptamine, can also serve as a substrate. 2 1D-1,5-diguanidino1,5-dideoxy-scyllo-inositol-4-P (streptidine-P) and 1D-l-guanidino-ldeoxy-5-O-carbamoyl-scyllo-inositol-4-P (bluensidine-P) ~ are not readily dephosphorylated. This phosphatase has an absolute and relatively specific requirement for Mg 2+. Mercaptans stimulate activity of purified preparations. Inhibitors. Sulfhydryl reagents, such as p-chloromercuribenzoate, cystamine, and freshly prepared aqueous solutions of formamidine disulfide • 2 HC1, strongly inhibit activity. lOThis volume [281.
462
[27]
ANTIBIOTIC BIOSYNTHESIS
[27] L-Alanine:lD-1-Guanidino-l-deoxy-3-ketoscyllo-inositol Aminotransferase By JAMES B. WALKER
o __A
I•H2
L-Alanine +
NH~
NI__A ?=Nil: ,
It
Pyruvate +
(1)
tt Oil
Ott 1 D-Guanidino-3amino- 1, 3-dideoxy-
scyllo- inositol
This aminotransferase 1,2 participates in the biosynthesis of streptomycin but is absent from strains which synthesize bluensomycin 3 (structure in previous article4). This aminotransferase is somewhat less specific with respect to amino donors than is L-glutamine:keto-scyllo-inositol aminotransferase, '~ but the two enzymes appear to be closely related, possibly being derived from a common ancestral gcne. :*
Assay Method
Principle. The assay is carried out in the reverse direction of Eq. (1). Amino donor labeled in the guanidino group, with two positive charges at p H 3.6, is converted to labeled product with one positive charge. Labeled substrate and labeled product are separated by high-voltage paper electrophoresis at p H 3.6 and counted. 6 This assay can be used with both crude and purified preparations; purified preparations must contain pyridoxal-P at all times to retain activity. Reagents Extract of Streptomyces bikiniensis A T C C 11062, or another strain which synthesizes streptomycin 1 D - l - [ 14C ] Guanidino-3-amino-l,3-dideoxy-scyUo-inositol 1j. B. Walker and M. S. Walker, Biochem. Biophys. Res. Commun. 26, 278 (1967) 2j. B. Walker and M. S. Walker, Biochemistry 8, 763 (1969). 3j. B. Walker, J. Biol. Chem. 249, 2397 (1974). 4This volume [21]. 5 This volume [23]. 6 This volume [22].
[27]
GUANIDINODEOXY-3-KETO-8cyllo-INOSITOLAMINOTRANSFERASE 463 Potassium phosphate, 0.2 M, containing 13 mM EDTA, pH 7.4 Sodium pyruvate, 0.15 M, freshly prepared
Procedure. The complete incubation mixture contains: 1D-I-I~4CI guanidino-3-amino-l,3-dideoxy-scyllo-inositol, 5 ~l; phosphate- EDTA, 5 ~1; pyruvate, 5 td; and extract of S. bikiniensis, 10 ~l. After incubation in a stoppered 13 X 100 mm test tube at 35 °, a 10-~l aliquot is spotted at the 34-cm segment, and components are separated by high-voltage paper eleetrophoresis at pH 3.6 and counted2 For mobilities, see the table of a previous article. ~
Preparation of ID-1-[14C] Guanidi~m-3.amino- l ,3-dideoxy-scyllo-inositol. 1-['4C]Guanidino-l-deoxy-scyllo-inositol-4-P, prepared as described, ~ is successively dephosphorylated, dehydrogenated, and transamihated (cf. Fig. 2 of a previous article *) during a single incubation with a nondialyzed supernatant solution of a sonicate of S. bikiniensis, prepared as described2 The complete incubation mixture contains: 1-[~C]guanidino-l-deoxy-scyllo-inositol-4-P, 0.72 ml (5.3 X 10'~ cpm); 0.5 M Tris C1, pH 8.8, 0.3 ml; 0.2 M L-alanine, 0.2 ml; and supernatant solution from sonieate of S. bikiniensis, 0.72 ml. The mixture is incubated 3 hr at 35 ° in a 100-ml beaker, with occasional swirling to assure sufficient aeration. (The dehydrogenation step will not proceed with this volume of solution in a small test tube where the depth of solution limits availability of oxygen.) The mixture is treated with 0.2 ml of 30% trichloroacetic acid and centrifuged. Labeled components in the supernatant solution plus washing are separated on a column containing Dowex 50 (H+I as described ~ (see the table~). Unchanged labeled substrate, as well as labeled intermediates, can be reclaimed during the same column isolation procedure. Occasionally the incubation mixture has been fortified with pyridoxal-P; this is essential for purified preparations, but can often be omitted when crude sonicates are the source of enzyme. More often, the labeled substrate is preincubated with Escherichia coli alkaline phosphatase (Sigma type III) to increase the final yield.
Preparation o] 1D-I- [~4C] Guanidino:-l-deoxy-3-keto-scyllo-inositol. This compound can be prepared by suitably scaling up the quantities used in the assay procedure, and subsequently separating the labeled components on a Dowex 50 (H +) column (the table~). Unreacted substrate can be reclaimed. Tris buffer should not be used, because it is eluted by 1.0 N HC1 along with the desired product. Separation o] the Two Inosamine Transaminases. 2 L-Glutamine:ketoscyllo-inositol aminotransferase5 can be prepared free from the aminotransferase activity described in this article by (a) use of S. glebosus This Yolume [26].
464
ANTIBIOTIC BIOSYNTHESIS
[27]
ATCC 14607 as the enzyme source a (b) treatment of an extract of S. bikiniensis at 55 ° for 5 min, 8 or (c) separation of both aminotransferases on a Sephadex G-100 column. For separation of the two aminotransferases, either sonicates or lysozyme extracts of S. bikiniensis can serve as starting material. All procedures are performed at 4 ° or less. To 30 ml of a lysozyme extract 9 of S. bikiniensis is added 0.3 ml of neutralized pyridoxal-P (20 mg/ml), and then 2 ml of 10% MnCl~ • 4H20 is added slowly with stirring. After 20 rain the suspension is centrifuged. Powdered (NH4)..SO,, 12 g, is slowly added with stirring to 27 ml of the supernatant solution; the solution is kept neutral by the careful addition of 2 M NH4OH. After 20 min the suspension is centrifuged, and the precipitate is taken up in 8 ml of 0.1 M potassium phosphate buffer, pH 7.4, containing 10 mM EDTA and 1 mg pyridoxal-P per milliliter. The solution is dialyzed for several hours; 1 ml of this preparation, containing 19.6 mg of protein (Lowry), is applied to a Sephadex G-100 (40-120 ~m) column, 2 X 54 cm, which had previously been equilibrated with 1 mM potassium phosphate containing per milliliter 0.1 mg of EDTA and 0.1 mg of pyridoxal-P, pH 7.4. Fractions of 2.4 ml (40 drops) are collected, 1 drop per second. Protein is first detected in tube 19. L-Glutamine:ketoscyllo-inositol aminotransferase activity is eluted first in a peak centering on tube 25, whereas L-alanine:lD-l-guanidino-l-deoxy-3-keto-scylloinositol aminotransferase activity is ¢luted in a peak centering on tube 28; there is thus some overlap, but tubes can be selected which have essentially only one of tho two activities. -° Properties
Biological Distribution. L-Alanine: 1D-l-guanidino-l-deoxy-3-ketoscyllo-inositol aminotransferase occurs in S. bikiniensis ATCC 11062, S. griseus ATCC 12475, S. ornatus ATCC 23265, and presumably other strains which synthesize streptomycin. This enzyme appears to be absent from S. hygroscopicus ]brma glebosus ATCC 14607, which synthesizes bluensomycin2 Specificity and Alternate Assays. From enzyme stability studies pyridoxal-P appears to be a required cofactor. At a concentration of 2 mM, L-alanine is the most active amino donor among the amino acids tested, followed by L-glutamate, and then L-glutamine. The corresponding Damino acids are inactive. When L-glutamine is the amino donor the reaction should be relatively irreversible because of cyclization or enzymic deamidation of the a-ketoglutaramate formed? There is evidence that s j. B. Walker, Lloydia 34, 363 (1971). gThis volume [24].
[28]
STREPTOMYCIN-6-P PHOSPHOHYDROLASE
465
at high concentrations keto-scyllo-inositol can serve as an amino acceptor, and a large number of other compounds can serve as amino donors. This enzyme catalyzes a very active exchange transamination between 1D-1-gu anidino-3-amino- 1,3.dideoxy.scyllo-inositol and 1D-l- [ 14C] guanidino-l-deoxy-3-keto-scyllo-inositol, ~,~ and this reaction can serve as an alternate assay. 1L-l-[14C]Guanidino-3-amino-l,3-dideoxy-scyllo-inositol, prepared by enzymic dephosphorylation of the transmidination product with streptamine-P as amidino acceptor, 1° cannot serve as amino donor in the reverse direction of Eq. (1). Stability. This aminotransferase is unstable to dialysis in the absence of pyridoxal-P and cannot be significantly reactivated by the subsequent addition of pyridoxal-P. Pyruvate spares the pyridoxal-P requirement during dialysis. Pyridoxamine-P, pyridoxal, and L-alanine do not protect this enzyme during dialysis. The enzyme from S. bikiniensis is inactivated by heating at 55 ° for 5 min, even in crude extracts. lOThis volume [25].
[28]
Streptomycin-6-P Phosphohydrolase
By JAMES B. WALKER I~N--C--~NH+ I
NH H
NH
O
[\ ,
HsC
.N/ l
[
~\ CH.OH
NH~ I
/I
I
Ix, O H
/[
Dihydrost repto my cin~ 6- P
I Dihydrostreptomycin
+
HPO~-
This enzymic reaction is believed to be the final step in streptomycin biosynthesis (Fig. 2 of a previous article1). Evidence that this reaction is the final step is derived primarily from the observations that streptomycin-6-P accumulates in the culture medium of (a) a particular strain of Streptomyces gri, eus (HUT 6037) grown in the presence of 5% glucose, 2 and (b) all streptomycin producers tested, following addition of This volume [21]. : R. Nomi, O. Nimi, T. Miyazaki, A. Matsuo, ami H. Kiyohara, Agr. Biol. Chem. 31, 973 (1967).
[28]
STREPTOMYCIN-6-P PHOSPHOHYDROLASE
465
at high concentrations keto-scyllo-inositol can serve as an amino acceptor, and a large number of other compounds can serve as amino donors. This enzyme catalyzes a very active exchange transamination between 1D-1-gu anidino-3-amino- 1,3.dideoxy.scyllo-inositol and 1D-l- [ 14C] guanidino-l-deoxy-3-keto-scyllo-inositol, ~,~ and this reaction can serve as an alternate assay. 1L-l-[14C]Guanidino-3-amino-l,3-dideoxy-scyllo-inositol, prepared by enzymic dephosphorylation of the transmidination product with streptamine-P as amidino acceptor, 1° cannot serve as amino donor in the reverse direction of Eq. (1). Stability. This aminotransferase is unstable to dialysis in the absence of pyridoxal-P and cannot be significantly reactivated by the subsequent addition of pyridoxal-P. Pyruvate spares the pyridoxal-P requirement during dialysis. Pyridoxamine-P, pyridoxal, and L-alanine do not protect this enzyme during dialysis. The enzyme from S. bikiniensis is inactivated by heating at 55 ° for 5 min, even in crude extracts. lOThis volume [25].
[28]
Streptomycin-6-P Phosphohydrolase
By JAMES B. WALKER I~N--C--~NH+ I
NH H
NH
O
[\ ,
HsC
.N/ l
[
~\ CH.OH
NH~ I
/I
I
Ix, O H
/[
Dihydrost repto my cin~ 6- P
I Dihydrostreptomycin
+
HPO~-
This enzymic reaction is believed to be the final step in streptomycin biosynthesis (Fig. 2 of a previous article1). Evidence that this reaction is the final step is derived primarily from the observations that streptomycin-6-P accumulates in the culture medium of (a) a particular strain of Streptomyces gri, eus (HUT 6037) grown in the presence of 5% glucose, 2 and (b) all streptomycin producers tested, following addition of This volume [21]. : R. Nomi, O. Nimi, T. Miyazaki, A. Matsuo, ami H. Kiyohara, Agr. Biol. Chem. 31, 973 (1967).
ANTIBIOTIC BIOSYNTHESIS
466
[28]
15 mM inorganic phosphate to the medium. 3 It is believed that accumulation of streptomycin-6-P in both instances results from inhibition of Eq. (1): inhibition in the case of HUT 6037 strain by the low pH of the medium, and inhibition in streptomycin producers in general by product inhibition by the added inorganic phosphate. Streptomycin-6-P phosphatase has marked phosphotransferase activity and can be employed to transfer phosphate groups to hydroxyls adjacent to nitrogenous groups of certain aminoglycoside antibiotics and other amino alcohols. 4-6
Assay Methods Method I Principle. Dephosphorylation of [3'a-3H] dihydrostreptomycin-6-P is followed by paper chromatography. This assay can be used for both crude extracts and partially purified preparations. This assay should be employed at intervals, during purification procedures which utilize less specific assays, to be certain that the desired phosphatase is being isolated. Reagents
Extract of mature mycelia of Streptomyces bikiniensis ATCC 110627,s or other strain which synthesizes streptomycin (EDTA not present; ~ 1 mM inorganic phosphate) [3'a-3H]Dihydrostreptomycin-6-P (see below) Tris-C1, 0.5 M, containing 40 mM MgC12, pH 9.0 Procedure. 4 The complete incubation mixture contains: [3H]dihydrostreptomycin-6-P, 5 t~l (e.g., 40,000 cpm); Tris-Mg, 5 ~l; and enzyme preparation, 20 td. After incubation in a stoppered 13 X 100 mm test tube at 35 °, a 10-~l aliquot is spotted, and the components are separated by paper chromatography with ammoniacal phenol and counted, s For mobilities, see the table of a previous article. 1 Since ~H-labeled compounds cannot be counted accurately on strips at R r values greater than 0.9 because of quenching by colored decomposition products of phenol, activity is measured as decrease in counts per minute remaining in phosphorylated substrate. Similar assays can be employed with 1D-1,3114C]diguanidino-l,3-dideoxy-scyllo-inositol-6-P (streptidine-P) or 1D-l-amino-3- [1~C]guanidino-l,3-dideoxy-scyllo-inositol-6-P as substrate; in the latter cases appearance of labeled product can be readily followed. A. L. Miller and J. B. Walker, J. Bacteriol. 104, 8 (1970). M. S. Walker and J. B. Walker,J. Biol. Chem. 246, 7034 (1971). J. B. Walker and M. Skorvaga, Y. Biol. Chem. 248, 2441 (1973). ' J. B. Walker and M. Skorvaga,,/. Biol. Chem. 248, 2435 (1973). This volume [22]. SThis volume [24].
[281
STREPTOMYCIN-6-P
PHOSPHOHYDROLASE
467
Preparation of [3'~-3H]Dihydrostreptomycin-6-P. [3'a-~H]Dihydrostreptomycin (700--3000 Ci/mole) is obtained from Amersham-Searle. The complete incubation mixture contains: [3H]dihydrostreptomyein, 0.5 ml (6 X 106 cpm) ; 0.5 M Tris-C1, pH 9.0, containing 40 mM MgCl~, 0.5 ml; 36 mM ATP, pH 7, 0.5 ml; and dialyzed lysozyme extract of S. bikiniensis ATCC 11062,8 1.0 ml. After incubation in a stoppered 13 X 100 mm test tube at 35 ° for 90 min, the reaction is stopped by heating at 100 ° for 4 min and the solution is centrifuged. The supernatant solution is added to a column (1 X 25 cm) containing Bio-Rex-70 (NHg) carboxylic acid resin, 100-200 mesh. Fractions of 3 ml are collected, at a flow rate of ca. 0.6 ml/min. The column is successively eluted stepwise with a total of 60 ml of each of the following concentrations of ammonium formate: 0.1 M, 0.3 M, 0.8 M, and 2.0 M. Labeled dihydrostreptomycin-6-P is eluted in the 0.8 M fraction and unchanged dihydrostreptomycin in the 2.0 M fraction (cf. the table, in a previous article1). Tubes containing the radioactive product are combined in a porcelain evaporating dish and evaporated to dryness in a vacuum desiccator over CaCI~. Residual ammonium formate is removed in a vacuum over a shallow layer of concentrated H._.SQ in petri dishes, to which glass wool has been added to increase the surface area. The labeled compound is taken up in a small amount of water and stored frozen. Pretreatment of Bio-Rex-70 Resin. Bio-Rex-70 (Na +) is obtained from Bio-Rad. The resin, 30 g, is suspended in 800 ml of glass-distilled water, and fines are decanted. The resin is washed two times with 1 N NaOH, then water to pH 7. The resin is next washed with 1 N HC1 then water to pH 7. The resin is then washed once with methanol, washed four times with water, and suspended in 500 ml of 1 M NH40H. The mixture is stirred slowly for 4 hr and then washed with glass-distilled water to pH 7-9 and stored. Method H N
p - N i t r o p h e n y l - P + 2 - d e o x y s t r e p t a m i n e --+ p-nitrophenol + 2-deoxystreptamine-6-P
H2N--~4C --NH
NH2 L- [Guanidino - ~4C ] arginine +
HO~
(2)
[H2 opo~-
~. L-Ornithine
+
~OH
z ~ ~H2
~o~/
(3) opo~-
1 D- 1- Amino- 3guanidino-1, 2, 3trideoxy- s c y l l o inositol-6-P
468
ANTIBIOTIC BIOSYNTHESIS
[28]
Principle. Streptomycin-6-P phosphatase (N) catalyzes transfer of phosphate from dihydrostreptomycin-6-P or p-nitrophenyl-P to amino alcohols such as Tris, streptamine, or 2-deoxystreptamine~; esterification occurs adjacent to the basic group. In the first step of this 2-step enzymic assay [Eq. (2)], a solution containing streptomycin-6-P phosphatase catalyzes transfer of phosphate from p-nitrophenyl-P to streptamine or 2-deoxystreptamine. In the second step [Eq. (3)], EDTA is added to inhibit further phosphatase action, and the streptamine-P isomer shown above is transamidinated in the presence of L-[guanidino-~4C]arginine and a dialyzed extract containing inosamine-P amidinotransferase2 The labeled product is separated by paper chromatography and counted. Reagents Preparation from mature mycelia of streptomycin producing strain of Streptomyces containing streptomycin-6-P phosphatasel°; no EDTA and ( 1 mM Tris or inorganic phosphate Dialyzed extract of mature mycelia of S. glebosus ATCC 14607 ~ or S. bikiniensis ATCC 11062 s p-Nitrophenyl-P, disodium, 6 mM EDTA, 0.3 M, pH 8 Streptamine 2HC1, 120 mM, or 2-deoxystreptamine.2 HC1 85 mM, s adjusted to pH 8 L- [Guanidino-~4C] arginine (12-25 Ci/mole), 33 ~Ci/ml
Procedure. In the first step, the complete incubation mixture contains: p-nitrophenyl-P, 5 ~l; streptamine or 2-deoxystreptamine, 5 ~l; and streptomycin-6-P phosphatase preparation, ~° 10 t~l. (In later stages of purification, Mg 2÷ should be added.) After incubation in a stoppered 13 X 100 mm test tube at 35 ° for 45 min, the following are added: EDTA, 1 ~l; [l*C]arginine, 5/~l; and dialyzed extract of S. glebosus (or S. bikiniensis), l0 ~l. After further incubation at 35 ° for 45 min, a 10-~l aliquot is spotted, and the labeled components are separated on an ammoniacal phenol paper chromatogram and counted, s Mobilities of the expected products are given in Table I of a previous article. 1 Method III Principle. Streptomycin-6-P phosphatase catalyzes transfer of phosphate from p-nitrophenyl-P to both water and Tris. ~ The p-nitrophenol formed is measured at 400 nm. This rapid, but nonspecific, assay can gThis volume [25]. 1oThis volume [26].
[28]
STREPTOMYCIN-6-P PHOSPHOHYDROLASE
469
be utilized with partially purified enzyme preparations during isolation procedures. However, Method I or II should be utilized at intervals to confirm that the phosphatase being isolated is indeed streptomycin-6-P phosphatase.
Reagents Partially purified streptomycin-6-P phosphatase preparation1°; no EDTA and ( 1 mM inorganic phosphate p-Nitrophenyl-P, disodium, 10 mM Tris C1, 0.5 M, containing 10 mM MgCI~, pH 8
Procedure. The complete incubation mixture contains: p-nitrophenyl-P, 0.5 ml; Tris-Mg, 1.0 ml; and enzyme preparation plus water, 2.5 ml. After incubation at 35% the absorbance is measured at 400 nm.
Purification The procedure for separation of streptomycin-6-P phosphohydrolase activity from 1-guanidino-l-deoxy-scyllo-inositol-4-P phosphohydrolase is described in a previous article. 1° The substrate specificity of this preparation has been determined, utilizing yarious mono- and diphosphorylated streptomycin derivatives. 4-6 Nimi et al. n have described a more extensive purification procedure for what appears to be a similar enzyme from their strain of S. griseus, H U T 6037, utilizing DEAE Sephadex A-50 chromatography. The substrafe specificity of their phosphatase preparation has not been reported.
Properties Biological Distribution. Streptomycin-6-P phosphatase appears to occur primarily in mature mycelia of streptomycin producing strains of Streptomyces, such as S. bikiniensis ATCC 11062, S. griseus ATCC 12475, S. griseocarneus ATCC 12628, and S. galbu.~ ATCC 14077. Little or no activity has been detected in S. kanamyceticus ATCC 12853, S. griseus ATCC 10971, and S. ]radiae. Activity appears to be readily released into the medium in Nomi's H U T 6037 strain. 11 Specificity. In the characterization of a given phosphatase, it is essential that the spectrum of substrate specificity be determined. Even nonspecific alkaline phosphatase from Escherichia coli (Sigma type III) will dephosphorylate streptomycin-6-P, given enough enzyme and time. Streptomycin-6-P phosphatase does not transfer phosphate esterified at the 110. Nimi, H. Kiyohara, T. Mizoguchi, Y. Ohata, and R. Nomi, Agr. Biol. Ct~em. 34, 1150 (1970).
470
ANTIBIOTIC BIOSYNTHESIS
[28]
primary 3'a-hydroxymethyl position of dihydrostreptomycin, but it does transfer phosphate esterified at secondary hydroxyl groups adjacent to basic nitrogenous groups, such as occur in streptidine-6-P, streptomycin-3"-P, Tris-P, ethanolamine-P, and numerous aminocyclitol-P derivatives. ~,5 The rate of formation of p-nitrophenol from p-nitrophenyl-P is enhanced in the presence of amino alcohols which can serve as phosphate acceptors. This behavior suggests formation of a phosphoryl enzyme, with dephosphorylation the rate-limiting step. The relatively high phosphotransferase activity of this enzyme perhaps could be employed in molecular modification of antibiotics when it is desired to protect or modify a group adjacent to an amino group. Dialysis of the S. bikiniensis enzyme against EDTA decreases activity; activity can be restored by the addition of Mg ~÷. Alternative Assays. Nimi et al. 11 employ a bioassay for streptomycin released following incubation of enzyme with streptomycin-6-P for 24 hr. Other assays might include action of the enzyme on (a) [3H]dihydrostreptomycin-3.'a,6-diP 12 to give [3H]dihydrostreptomycin-3'a-P which can be separated by paper chromatography (Table I of a previous article1); or (b) dihydrostreptomycin-6-3~P to give 32Pi, which can be detected in the water wash of a small Pasteur pipette column containing Dowex 50 (H ÷) resin, whereas unreacted substrate is retained on the column. Inhibitors. Streptomycin-6-P phosphatase is not inhibited by sulfhydryl reagents, unlike 1-guanidino-l-deoxy-scyllo-inositol-4-P phosphatase, but it is inhibited by EDTA and inorganic phosphate. Certain amino alcohols inhibit formation of inorganic phosphate by competing with water for phosphate transferred from the presumed phosphoryl-enzyme intermediate. 5 Moderate (e.g., 0.03 M) concentrations of inorganic phosphate included in industrial streptomycin fermentation media might well increase the overall yield of antibiotic by decreasing autoinhibition, since streptomycin-6-P has no biological activity.:. 3 Alternatively, mutants lacking streptomycin-6-P phosphatase could be selected and employed to increase overall yield. Active antibiotic could then be generated by incubation with nonspecific alkaline phosphatase or the enzyme described here.
1, This volume [51].
[29]
~- (a-AMINOADIPYL)CYSTEINYLYALINE SYNTHETASE
471
[29] 6-(a-Aminoadipyl)cysteinylvaline Synthetase B y PATRIClA FAWCETT and E. P. ABRAHAM
-O2C. CH(NH3)(CH~)3CONH. CH(CH2SH)CO2- + HaN. CH(CH(CH3)2)CO~- --~ +
-02C. CH (NH3) (CH2)3CONHCH (CH~SH)CON HCH(CH(CH~)~)CO C A tripeptide, g-(a-aminoadipyl)cysteinylvaline, was shown by Arnstein and Morris l to be present in small amounts in the mycelium of Penicillium chrysogenum. A mixture of related peptides was obtained by Loder and Abraham 2 from the mycelium of Cephalosporium acremonium. The major component of the mixture was shown to be g-(L-a-aminoadipyl)-L-cysteinyl-D-valine. 2 Two minor components appeared to be tetrapeptides, one differing from the tripeptide in containing a glycine residue and one also in containing a fl-hydroxyvaline residue in place of valine. 2 g-(a-Aminoadipyl)cysteinylvaline synthetase, which catalyzes the formation of this tripeptide from g-(L-a-aminoadipyl)-L-cysteine and L-valine, has been found in a broken-cell system from C. acremonium2 No synthesis of the tripeptide from L-a-aminoadipic acid and a-cysteinyl-L-valine could be shown to occur. 3
Assay Method Principle. The enzyme is assayed by measurement of the incorporation of 14C from labeled valine into g-(a-aminoadipyl)cysteinylvaline in the presence of ~-(L-a-aminoadipyl)-L-cysteine. The synthesis of g-(L-aaminoadipyl)-L-cysteine has been described by Loder and Abraham. 4 The labeled product is isolated by paper electrophoresis and chromatography, after oxidation to the sulfonic acid form, and identified by comparison with an authent!c sample of the tripeptide. ~-(L-a-aminoadipyl)-L-cysteinyl-D-valine has been synthesized by Usher. ~ The disulfide of the corresponding all-L tripeptide has been synthesized by Rudinger?
1H. R. V. Arnstein and D. Morris. Biochem. J. 76, 357 (1960). 2p. Bronwen Loder and E. P. Abraham. Biochem. J. 123, 471 (1971). 3p. Bronwen Loder and E. P. Abraham. Biochem. J. 123, 477 (1971). 4p. Bronwen Loder, E. P. Abraham, and G. G. F. Newton, Biochem. J. 112, 389 (1969). J. J. Usher, unpublished experiments, 1973. "J. Rudinger, Czech. Chem. Commun. 27, 2246 (1962); personal communication (1968).
472
ANTIBIOTIC BIOSYNTHESIS
[29]
Reagents n-Valine, 0.2 M solution in water DL-[l-14C]Valine, 1.0 ~Ci/~l, 0.03 ~mole/~l $- (L-a-Aminoadipyl) -L-cysteine 4 Adenosine 5'-triphosphate disodium salt, 9 mg Phosphoenolpyruvate tricyclohexylammonium salt, 12 mg Pyruvate kinase, crystalline suspension, 10 mg/ml Enzyme, crude preparation as described below, 3 ml
Procedure. Adenosine triphosphate (9 mg) and phosphoenolpyruvate (12 mg) are weighed into the same tube, and the enzyme preparation (3 ml) is added. The pH of the solution is adjusted to 7.0 with 1 M NaOH, and pyruvate kinase (75 t~l of suspension) is added. Samples (1 ml) of this mixture are added to tubes containing $-(L-~-aminoadipyl)L-cysteine (1 mg) in 175 ~1 of water and 5 ~1 of 1 M NaHCOs. Samples of the solution of n-valine (20 t~l) and DL-[1-14C]valine (5 td) arc added to each tube and the tubes incubated at 27.5 ° for 1 hr. After incubation the mixtures are centrifuged at 20,000 g for 30 min and the supernatants removed and freeze-dried. The residues are extracted with 70% (v/v) ethanol (each 2 X 200 td) and the extracts diluted with water (2 ml) and freeze-dried. The desalted extracts are treated with a 1.5% solution of performic acid in 98% formic acid (0.2 ml) at 0 ° and the mixtures kept at 0 ° for 6 hr. After dilution with water (2 ml) the solutions are freeze-dried. Each residue is dissolved in a solution (25 ~l) of synthetic $-(L-aaminoadipyl)-L-cysteinyl-D-valine (100 ~g). Samples (5 ~l) of the resulting solutions arc spotted onto Whatman No. 1 paper and subjected to electrophoresis (70 v/cm) for 2.5 hr at pH 1.8 in 20% (v/v) acetic acid containing 2% (v/v) formic acid and then to chromatography in the second dimension in butan-l-ol:acetic acid:water (4:1:4 by volume). Under these conditions the labeled tripeptide migrates' slightly less far than glutathione sulfonic acid toward the anode and shows an R value relative to that of glutathione sulfonic acid of about 2.75. The position of the labeled tripeptide on the paper is located by radioautography 7 and the radioactive spot is counted. 7 The paper is then sprayed with ninhydrin and the identity of the labeled product with ~-(a-aminoadipyl)cysteinylvaline is confirmed by the coincidence of the radioactive and ninhydrin positive spots. B. Smith, S. C. Warren, G. G. F. Newton, and E. P. Abraham, Biochem. d. 103, 877 (1967).
[29]
~- (ct-AMINO&DIPYL)CYSTEINYLVALINE SYNTHETASE
473
Preparation Cephalosporium acremonium C91 is grown in shake-flasks in a chemically defined medium as described by Smith et alJ The mycelium is harvested by filtration 72 hr after inoculation and washed on the filter with water. It is then resuspended in 0.1 M Tris • HC1 buffer, pH 7.0, containing 0.1 M KC1, 20 mM MgSO4, and 1 mM cysteine (1 g damp-dry mycelium/3 ml), and the suspension is homogenized in a Potter-Elveh]em homogenizer with a Teflon pestle cooled in ice. The homogenate is diluted with an equal volume of buffer and a sample (20 ml) of the resulting mixture subjected to ultrasonic treatment (60 W, 20 Hz) in a jacketed glass vessel cooled at 0 ° and with a titanium probe, 1 cm in diameter, ending about 2 mm below the surface. Half of the treated suspension is centrifuged for 30 min at 20,000 g, and the supernatant is removed. The particulate fraction is washed by resuspension in the buffer (20 ml) and recentrifugation at 20,000 g. It is finally resuspended in buffer to give a volume equal to that of the suspension from which it has been obtained and used immediately after preparation. 8-(a-Aminoadipyl)cysteinylvaline synthetase is present in both the complete mixture obtained after ultrasonic treatment and in the particulate fraction, but very little is found in the supernatant fraction. The crude enzyme preparation from 1 g damp-dry mycelium synthesises about 15 nmoles of tripeptide in 1 hr under the conditions described.
Specificity $-(~-Aminoadipyl)cysteinylvaline synthetase differs from glutathione synthetase, which is found mainly in the supernatant fraction obtained by centrifugation after ultrasonic treatment of the mycelium. It catalyzes the synthesis of tripeptide from $-(L-a-aminoadipyl)-L-cysteine and L-valine, but nosynthesis has been observed from ~-(D-a-aminoadipyl)-scysteine and L-valine, from ~-(L-~-aminoadipyl)-L-cysteine and D-valine, or from L-~-aminoadipic acid and L-cysteinyl-L-valine?. ~
P. Bronwen Loder, unpublished experiments, 1971.
474
ANTIBIOTIC BIOSYNTHESIS
[30]
[30] A c y l C o A : 5 - A m i n o p e n i c i l l a n i c A c i d A c y l t r a n s f e r a s e
By STEN GATENBECK Acyl CoA + 6-aminopenicillanic acid (6-APA) --* penicillin + CoA Assay Method ~
Principle. The method is based on the determination of radioactive benzylpenicillin formed with 14C-labeled phenylacetyl CoA and 6-APA as substrates. Reagents 6-APA, O.O4M Dithiothreitol, 0.25 M Phenylacetyl-l-l~C CoA, 7 mM. The synthesis of this reagent is performed by reacting the mixed anhydride of the labeled acid and carbonic ester with CoA. Buffer solution, pH 8.4 (0.2 M Tris, 50 mM phosphate, 0.2 M NaC1, 1 mM EDTA) Benzylpenicillin, 3 mM Penicillinase
Procedure. Incubations are carried out at 30 ° in tubes containing the following mixture: Enzyme, 0.1 ml; 6-APA, 0.025 ml; dithiothreitol, 0.01 ml; phenylacetyl-l-l~C CoA, 0.025 ml (200,000 dpm); buffer, 0.1 ml. After 40 min 0.01 ml af carrier benzylpenicillin is added to each tube, and the total content is spotted on a paper chromatogram strip after adjusting the pH of the solution to 5.5. The paper strips, preimpregnated with citrate buffer, pH 5.7, are developed in ether saturated with water. The radioactive benzylpenicillin is localized on the paper strip by using a radiochromatogram scanner. The amount of benzylpenicillin formed is determined by cutting out the radioactive area from the paper chromatogram and by dipping the paper into a solution of PPO and POPOP in toluene-methanol (1 : 1) and measuring the radioactivity in a scintillation spectrometer. Definition of Unit and Specific Activity. One unit is defined as the amount of enzyme that catalyzes the formation of 1 nmole of penicillin in the above test. Specific activity is the number of units of activity per milligram of protein determined by the biuret method. 1S. Gatenbeck and U. Brunsberg, Acta Chem. Scand. 9.2, 1059 (1968).
[30]
ACYL CoA:AMINOPENICILLANIC ACID ACYLTRANSFERASE
475
Other Assay Methods. 2 Methods based on the following exchange reaction have been used for determining the enzyme activity. Acyl APA + *APA;~ acyl *APA + APA
Enzyme Purification Growth of Cells and Extract Preparation. Penicillium chrysogenum ATCC 12,687 is grown on a rotary shaker (1" stroke, 250 rpm) at 28 ° in 500-ml conical flasks each containing 150 ml of substrate of the following composition (all in grams per liter): KH2P04, 3.0; Na~SO4, 0.5; MgS04 • 7H20, 0.25; ZnS04 • 7H20, 0.02; M n S Q • H20, 0.02; Fe (NH4)_~(SO~) 2 • 6H.~O, 0.1 ; CuSO~ • 5H20, 0.005 ; CaC12 • 2H20, 0.05; yeast extract, 1.0; NH4-acetate, 3.5; NH4-1actate, 6.0; glucose, 10; lactose, 30; distilled water. The mycelia from 12 flasks are harvested and washed with buffer (0.2 M Tris, 50 mM phosphate, 0.2 M NaC1, 1 mM EDTA, pH 7.8) after 96 hr of growth. The cells are ground with sand at 4 ° for 10 min, and the cell debris is removed by centrifugation for 20 min at 30,000 g. Ammonium Sul]ate Precipitation. Solid ammonium sulfate is slowly added with stirring to the supernatant solution from the preceding step until a 0.30 saturated solution is obtained. The precipitate is collected by centrifugation (30,000 g for 20 min) and discarded. Additional solid ammonium sulfate is added to the supernatant solution until 0.50 saturation is reached. After centrifugation, the supernatant solution is discarded, and the precipitate is dissolved in 30 inl of 20 mM phosphate buffer, pH 7.8; the solution is passed through a column of Sephadex G-25, coarse. Hydroxyapatite Gel Adsorption and Elution. The eluate from the preceding step is treated with hydroxyapatite (7.0 g). The gel is washed with 20 ml of 20 mM phosphate buffer, pH 7.8, followed by 10 ml of 40 mM phosphate buffer, pH 7.8. The enzyme is solubilized with 25 ml of 70 mM phosphate buffer, pH 7.8. The volume of the enzyme solution is then reduced to 4 ml by ultrafiltration. Sephadex G-200 Column. The enzyme concentrate is placed on a sephadex G-200 column (3 X 70 cm) which has been equilibrated with a buffer of 0.2 M Tris, 50 mM phosphate, 0.2 M NaC1, 1 mM EDTA, pH 7.8. The enzyme is eluted at about 190 ml with the same buffer. The yields and the specific activities of the various fractions obtained during the purification procedure are outlined in the table. 2D. L. Pruess and M. J. Johnson, J. Bacteriol. 94, 1502 (1967).
476
ANTIBIOTIC BIOSYNTHESIS
[31]
PURIFICATION OF 6-AMINOPENICILLANIC ACID ACYLTRANSFERASE FROM
Penicillium chrysogenum
Step Cell extract (NH4)~SO4 fractionation Hydroxyapatite Sephadex G-200
Protein (mg)
Units
Specific activity (units/mg protein)
2640 144 14.4 0.7
1584 2016 252 64
0.6 14.0 17.5 91.0
Yield (%) 100 127 16 4
Properties Specificity. Fractionated enzyme preparations utilize a number of acyl-CoA derivatives although at different rates, e.g., phenoxyacetylCoA, p-methoxyphenylacetyl-CoA, octanoyl-CoA. The enzyme does not show any penicillin acylase activity. Ejlect of pH. In Tris buffer the optimum enzyme activity is approximately pH 8.5. The activity is very low at pH 7. Activators. The enzyme activity is stimulated by the presence of thiol compounds. Comments. An enzyme that catalyzes acyl group exchange between various penicillins has been described 2 under the name penicillin acyltransferase. The likely Uni Uni Uni Uni Ping Pong mechanism for the 6-APA acyltransferase reaction suggests that the two enzymes are identical.
[31] Phenacyl :Coenzyme A Ligase By
RICHARD BRUNNER a n d MAX ROHR
Phenacyl:coenzyme A ligase catalyzes the synthesis of phenacetylCoA and phenoxyacetyl-CoA according to Eq. (1). RCOOH 4- CoA ~- ATP ~- CoA-SCOR -b AMP + PP
(1)
where R is the phenacetyl or phenoxyacetyl radical, known as the acyl side chains of benzylpenicillin and phenoxymethylpenicillin. The enzyme was found to be formed in a penicillin-producing strain of Penicillium chrysogenum in the course of penicillin production experi-
476
ANTIBIOTIC BIOSYNTHESIS
[31]
PURIFICATION OF 6-AMINOPENICILLANIC ACID ACYLTRANSFERASE FROM
Penicillium chrysogenum
Step Cell extract (NH4)~SO4 fractionation Hydroxyapatite Sephadex G-200
Protein (mg)
Units
Specific activity (units/mg protein)
2640 144 14.4 0.7
1584 2016 252 64
0.6 14.0 17.5 91.0
Yield (%) 100 127 16 4
Properties Specificity. Fractionated enzyme preparations utilize a number of acyl-CoA derivatives although at different rates, e.g., phenoxyacetylCoA, p-methoxyphenylacetyl-CoA, octanoyl-CoA. The enzyme does not show any penicillin acylase activity. Ejlect of pH. In Tris buffer the optimum enzyme activity is approximately pH 8.5. The activity is very low at pH 7. Activators. The enzyme activity is stimulated by the presence of thiol compounds. Comments. An enzyme that catalyzes acyl group exchange between various penicillins has been described 2 under the name penicillin acyltransferase. The likely Uni Uni Uni Uni Ping Pong mechanism for the 6-APA acyltransferase reaction suggests that the two enzymes are identical.
[31] Phenacyl :Coenzyme A Ligase By
RICHARD BRUNNER a n d MAX ROHR
Phenacyl:coenzyme A ligase catalyzes the synthesis of phenacetylCoA and phenoxyacetyl-CoA according to Eq. (1). RCOOH 4- CoA ~- ATP ~- CoA-SCOR -b AMP + PP
(1)
where R is the phenacetyl or phenoxyacetyl radical, known as the acyl side chains of benzylpenicillin and phenoxymethylpenicillin. The enzyme was found to be formed in a penicillin-producing strain of Penicillium chrysogenum in the course of penicillin production experi-
[31]
PHENACYL :COENZYME A LIGASE
477
merits on a laboratory scale? It appeared at 2 days when penicillin production was beginning and increased markedly at 3 and 4 days, i.e., just prior to and during the stage of most rapid penicillin fermentation. The finding l-3 that mycelial extracts of Penicilliura chrysogenum contain a 6-aminopenicillanic acid acyltransferase, which catalyzes the formation of penicillins from 6-aminopenicillanic acid and the CoA derivatives of phenylacetic or phenoxyacetie acid, led to the conclusion that the final step in penicillin biosynthesis might involve the direct N-acylation of 6-aminopenicillanic acid by the CoA-activated side-chain precursors.
Assay Method
Principle. The method employed involves the measurement of the rate of formation of acylhydroxamate in the presence of excess ATP and phenylacetate, catalytic amounts of CoA, and neutral hydroxylamine as described elsewhere? Reagents Potassum phosphate buffer, 1 M, pH 7.0 CoA, 1 mM, pH 6.5 ATP, 0.1 M, pH 7.5 Potassium fluoride, 1 M Magnesium chloride, 0.2 M Reduced glutathione, 0.2 M, pH 4.5 Potassium phenylacetate, 0.2 M, pH 6.0 Hydroxylamine solution, 2 M, pH 6.5, freshly prepared by mixing equal volumes of 4 M hydroxylamine hydrochloride and 4 M potassium hydroxide. Ferric chloride reagent containing 0.37 M ferric chloride, 20 mM trichloroaeetic acid, and 0.66 M hydrochloric acid.
Procedure. To a small centrifuge tube are added 0.1 ml of potassium phosphate buffer, 0.1 ml of CoA, 0.1 ml of ATP, 0.05 ml of potassium fluoride, 0.05 ml of magnesium chloride, 0.05 ml of glutathione, 0.1 ml of potassium phenylacetate, 0.1 ml of hydroxylamine, and enough distilled water to bring the final volume to 1.0 ml after the addition of the enzyme solution. After 5 min of temperature equilibration in a water bath at 37 ° , the enzyme is added and incubation is carried out for periods ' R . Brunner, M. RShr, and M. Zinner, Hoppe-Seyler's Z. Physiol. Chem. 349, 95 (1968). 2 B. Spencer, Biochem. Biophys. Res. Commun. 31, 170 (1968). 3 S. Gatenbeck and U. Brunsberg, Acta Chem. Scc,nd. 22, 1059 (1968). 4 p. Berg, this series, Vol. V [62].
478
ANTIBIOTIC BIOSYNTHESIS
[31]
of 20 or 40 min. The reaction is stopped by the addition of 2 ml of the ferric chloride reagent, and after centrifugation the optical density of the solution is measured at 540 nm in a spectrophotometer. The extinction coefficient of phenylacetylhydroxamate under these conditions is 0.90 X 106 cm 2 mole-1. One unit of enzyme activity is defined as the catalytic activity leading to the formation of 1 t,mole of acylhydroxamate in 1 min. Comment. Omission of CoA in parallel tubes as recommended in the assay of acetyl:coenzyme A ligase 4 frequently does not show any significant difference in absorption. Apparently, phenylacetyl-AMP resulting from the reaction of phenylacetate and ATP can readily react with hydroxylamine, but the possibility cannot be ruled out that catalytic amounts of CoA remain in the enzyme preparation at the low degree of purity achieved. In order to ascertain the participation of CoA in the overall coupled reaction, other procedures such as that of Grunert and Phillips, ~ which gave positive results in the authors' laboratory, may be applied; for assays in the course of enzyme purification the hydroxamate procedure appears to be more advantageous.
Production of Enzyme Enzyme-containing mycelium may be obtained from a penicillin plant (harvested at the third or fourth day of fermentation), or produced in the laboratory as follows: Cultures of a penicillin-producing strain of Penicillium chrysogenum are maintained on Sabouraud-sucrose agar slants. Strains yielding 1000-2000 units of penicillin per milliliter (1 unit = 0.6 t'g sodium benzylpenicillin) are sufficient. Spore suspensions are prepared as follows: barley grains are rinsed with tap water and subsequently steeped in a solution containing: lactose, 30 g/liter; KH2PO~, 3.0 g/liter; K~HP04, 3.0 g/liter; MgS04" 7 H20, 0.1 g/liter; corn steep liquor, equivalent of 0.85 g nitrogen per liter. After 30 min the liquid is decanted and layers of grains about 2 cm deep are placed in Erlenmeyer flasks with cotton plugs and sterilized for 20 min at 120 ° on 3 successive days. One milliliter of a spore suspension obtained from an agar slant is used to inoculate each flask. After 1 week of incubation at 25 °, sterile saline is added aseptically, and the spore suspension obtained by shaking the flasks is transferred aseptically to a sterile flask. It should contain 107-10s spores per milliliter and may be kept in the cold for several months. The nutrient medium contains (in grams per liter): glucose, 10.0; lactose, 30.0; starch, 10.0; dextrine, 5.0; yeast extract, 1.0; citric acid, R. R. Grunert and P. H. Phillips, Arch. Biochem. 30, 217 (1951).
[31]
PHENACYL:COENZYME A LIGASE
479
4.0; lactic acid, 3.0; ammonium acetate, 3.5; ammonia solution (25%), 3.0; KH,,PO,, 1.0; M g S Q " 7 H20, 0.5; FeSO4 • 7 H20, 0.05; Z n S Q . 7 H._,O, 0.01; CUSP4.5 H~O, 0.01; MnSO4" 7 H20, 0.01; Co(NO~)2" 6 H,_,O, 0.005; CaCl.., • 12 H.,O, 0.05; NaC1, 1.0; phenylacetic or phenoxyacetic acid, 2.0. The pH is adjusted to 5.8 with 2 N NaOH and 150 ml of medium are placed in l-liter culture flasks and sterilized at 120 ° for 20 min. After inoculation with 0.5 ml of spore suspension propagation is performed at 25 ° on a rotary shaker (e.g., Model V, New Brunswick Scientific Co.) at 280 rotations per minute. Maximum enzyme production is usually attained after 4 days. The mycelium is filtered with suction and washed three times with cold 0.5% potassium chloride. It may t)e stored in the frozen state; alternatively, freeze-dried mycelium may t)e prepared according to common procedures. Such preparations are very stable, and extraction is less diffficult. Drying with acetone results in considerable losses of enzyme activity. To obtain a crude enzyme extract, the following procedures may be used: Fresh or frozen mycelium is mixed with an equal amount of Celite (Hyflo Super Cel, Johns Manville Co., Baltimore) and ground to a homogeneous paste for 15 min in a cooled mortar. The paste is extracted at 0 ° with 2-3 times the mycelial weight of 10 mM K2HPO~ with stirring, and centrifuged for 30 min at 0 ° at 30,000 g. Freeze-dried mycelimn is ground to a fine powder, mixed with 8 times its weight of glass heads (average diameter, 500 ~m), suspended in 7 times the mycelial weight of 50 mM Tris • HC1 buffer pH 9.0, and treated in a vibrating disintegrator '; for 15 min under cooling at 3-5 °. The suspension is separated from the glass beads by low speed centrifugation for about 5 min, the glass beads are washed with 3 times the mycelial weight of 50 mM Tris . HC1 buffer pH 9.0, and the combined supernatants are centrifuged at 30,000 g for 20 rain at 0% In each case the extract should yield approximately 10-15 mg of protein per milliliter. The use of buffer solutions of higher pH, as indicated in the second procedure, has proved especially efficient in the case of mycelia from industrial fermentations, which frequently give lower protein yields. If not used immediately, the extract should be kept frozen. Partial Purification
Since several of the common procedures of enzyme purification (e.g., adsorption on calcium phosphates, precipitation with organic solvents, 6A "VIBROGEN" cell mill (E. Biihler, Tiibingen, Germany), operated at 4000 rpm, was used by the authors.
480
ANTIBIOTIC BIOSYNTHESIS
v
O
¢9
~0
©
~9 Z
O Z O
y, ~9
¢D
O v
O
L~ tt~
[31]
[31]
PHENACYL:COENZYME A LIGASE
481
absorption by DEAE-cellulose at different pH and ionic strengths) have not been successful, a procedure leading only to a moderate degree of purity has been elaborated. DEAE-cellulose (DE 23, Whatman) is pretreated and equilibrated with 50 mM Tris. HCl-buffer of pH 8.0 according to the instructions given by the manufacturer. The pH of the crude extract is adiusted to 8.0. To each 100 ml of extract about 100 ml of DEAE-cellulose suspension corresponding to 140 g of dry cellulosic material per 100 g of protein are added under stirring. After treatment for at least 15 nfin at 0 ° the suspension is filtered with suction over filter paper and the filtrate eolletted. To each 100 ml of filtrate, 71 g of solid ammonium sulfate are added (95% saturation) at 0 ° under constant stirring; stirring is continued for at least 30 rain. After eentrifugation at 30,000 g for 30 rain at 0 °, the precipitate is dissolved in a sufficient amount of 50 mM Tris • HC1 buffer pH 9.0 containing a few crystals of glutathione. The solution is brought to 40% saturation of ammonium sulfate as described above, centrifuged, and the precipitate discarded. The supernatant solution is brought to 70% saturation of ammonium sulfate as described before, and the precipitate obtained after centrifugation at 30,000 g is dissolved in a minimmn amount of 50 mM Tris • HC1 buffer pH 7.0. An example of the whole procedure is given in the table. Properties
Stabilitg. Whereas crude extracts are rather stable to prolonged storage at --20 ° in a pH range of 7 to 9, partially purified preparations are less stable under these conditions and particularly sensitive to repeated freezing and thawing. Addition of sulfhydryl compounds, e.g., glutathione, gives some protection. Rapid inactivation occurs at temperatures above 40% Specificity. Under conditions of the assay procedure, the enzyme preparation catalyzes the activation of phenylaeetic, phenoxyaeetie, and also acetic acid with similar degrees of activity. No change of the proportions of activity against the respective acids was observed during the purification procedure. Effect of pH. The enzyme preparation is maximally active in a pH range of 6 to 7.
482
ANTIBIOTIC BIOSYNTHESIS
[32]
[32] P h e n y l a c e t y l C o e n z y m e A H y d r o l a s e
By
BBIAN SPENCEa
Phenylaeetyl coenzyme A + H~O--~ phenylaeetie acid + eoenzyme A An enzyme catalyzing this reaction has been recognized in extracts of a member (Wis 51-20 F~) of the "Wisconsin Family" of high penicillinyielding mutants of Penicillium chrysogenum and in a number of related mutants?
Assay M e t h o d
Principle. [1-~4C]Phenylacetyl coenzyme A is incubated at 37 ° with P. chrysogenum extract as well as buffer, EDTA, and a sulfhydryl compound. The reaction is stopped by adding an excess of citrate buffer, pH 5.0, and the mixture is chromatographed on paper. The radioactivity of the spot corresponding to [1-14C]phenylacetic acid is counted by liquid scintillation and quantitated by reference to the radioactivity of standard [ 1-~4C] phenylacetyl CoA.
Reagents [1-14C]Phenylacetyl coenzyme A is prepared by the mixed anhydride method of Stadtman. 2 To 50 ~Ci of [1-14C]phenylacetic acid (39 mCi/mmole, The Radiochemical Centre, Amersham) is added 8.9 mg of phenylacetic acid, and the mixture is dissolved in 400 ~l of ether. After addition of 6,1 ~l of pyridine the mixture is cooled in ice and 7 ~l of ethylchloroformate is added slowly with agitation. During standing for 1 hr in ice, the tube is occasionally twirled by hand, causing the precipitated pyridinium chloride to stick to the sides of the vessel. The clear supernatant is pipetted off and added to a solution of 33.8 mg coenzyme A. (Sigma Chemical Co. Grade 1) in 1 ml of water, which has been adjusted to pH 7.6 with solid potassium bicarbonate. The tube containing the mixture is gassed with nitrogen and shaken for 30 rain. The solution is then adjusted to pH 2.0 with HC1 and extracted 4 times with 1.5 ml ether. The remaining ether is evaporated off by a stream of nitrogen. The solution is adjusted to pH 7.0 with solid lB. Spencer and Chit Maung, Biochem. J. 118, 29P (1970). E.R. Stadtman, this series, Vol. 3, p. 931.
[321
PHENYLACETYL COENZYME A HYDROLASE
483
potassium carbonate and distributed in 100-t~l portions in small tubes for storage at --20 ° . Radioactivity is counted by liquid scintillation and thiol ester content is assayed by the DTNB method of Elhnan 3 scaled down to deal with 40 t~l of diluted (1/9 v/v) sample. The yield is 1 ml of 35 mM [1-14C]phenylacetyl coenzyme A, and 10~1 spotted on paper give 450,000 epm measured as described below. Potassium phosphate buffer, 0.1 M, pH 7.5, containing 1 mM EDTA and 10 mM glutathione or N-acetylcysteine Citrate buffer, pH 5.0, 10% (w/v) Ethanol~n-butanol~28% (w/v) ammonium sulfate (2:1:1 by volume) 4 Whatman No. 1 chromatography paper that has been wetted with potassium phosphate buffer, pH 6.0 (75 g/liter of KH2P04, 25 g/liter of K~HPO~), blotted to remove excess buffer and dried 5 2,5-Diphenyloxazole (PP0), 5 g, and of 2,2-p-phenylenebis(5phenyloxazole) (POPOP), 0.3 g, dissolved in 1 liter of toluene
Assay of the Hydrolase. The enzyme preparation, 20 td, is incubated with l0 t~l of 24 mM [l-14C]phenylacetyl CoA and 10 t~l of 0.1 M potassimn phosphate buffer, pH 7.5 containing 1 mM EDTA and 10 mM glutathione. In controls the enzyme preparation nmst be replaced by the buffer-sulfhydryl mixture in which the enzyme sample is contained and the controls must be incubated along with the enzyme reaction tubes. After incubation for 15 rain, the reaction is stopped by adding 20 ~1 of 10% (v/v) citrate buffer, pH 5.0. Portions (10 ul) of the treated reaction mixtures including controls are chromatographed for 5-10 hr by descending chromatography on treated Whatman No. 1 paper using the ethanol/ n-butanol/28% ammonium sulfate solvent. The spot at RI 0.65, corresponding to ['~CIphenylacetic acid, is cut out, placed in a glass vial with 10-12 ml of the scintillation fluid and counted for sufficient time (usually 2-5 rain) in a Packard Tri-Carb liquid scintillation spectrometer Model 3375 to obtain a count rate with an SE. of less than 1%. The count rate is quantitated by reference to the radioactivity of standard amounts of [1-~C]phenylacetyl CoA which are spotted onto a piece of chromatography paper. The assay is linear with enzyme concentration and with time only over 20 rain. C. Ellman, Arch. Biochem. Biophys. 82, 70 (1959). 4 p. L. Tardrev and M. J. Johnson, J. Bacteriol. 76, 400 (1958). D. L. Pruess and M. J. Johnson, g. Bacteriol. 94, 1502 (1967).
484
ANTIBIOTIC BIOSYNTHESIS
[32]
A unit of phenylacetyl CoA hydrolase is defined in the standard manner as that amount which hydrolyzes 1 ~mole of the substrate per minute.
Production and Purification
Culture on Agar Slopes. Penicillium chrysoqenum Wis 51-20F3 is maintained as a spore suspension in soil and subcultured on tomato juice agar [canned tomato juice adjusted to pH 6 with 1 N NaOH and diluted 1:1 (v/v) with 2.5% (w/v) agar solution (Difco Bacto) ]. Shake Cultures. Inoculations are carried out with fresh spore suspension prepared by washing spores from a tomato-agar slope with 5 ml of sterile deionized water containing 0.1% (v/v) Tween 80. Five milliliters of this spore suspension are used to inoculate 250-ml Erlenmeyer flasks containing 40 ml of the following sterile glucose-lactate salts medium (in g/liter): D-glucose, 40.0; ammonium lactate, 21.0; KH:PO~, 3.0; Na2SO~, 0.74; magnesium acetate, 0.25; ZnCl~, 0.02; CaCO3, 13.0; MnCl~. 4H20, 0.02; FeCl~ • 6H20, 0.02; and CuCl~ 2H~_O,0.005. The glucose/ lactate is sterilized separately from the salts solution. The inoculated flasks are shaken for 2 days at 25 ° and 250 rpm on a gyratory shaker. The preculture (40 ml) is used to inoculate l-liter Erlenmeyer flasks containing 400 ml of sterilized fermentation medium composed of (g/l) : lactose, 50; corn liquor, 50; Na2SO~, 1. Growth is carried out at 28 ° on a gyratory shaker (250 rpm). At 24 hr, 1 ml of lard oil containing 5% v / v Tween 80 is added to prevent foaming. After 3 days growth, the mycelium is harvested by filtering through a double layer of cheesecloth and washed thoroughly with cold water. The mycelium is then pressed between filter papers to remove excess water and is used either immediately or after storage at --20 °. A yield of 15-25 g of "pressed-dry" mycelium per flask is obtained.
Purification
Step 1. Ten grams of press-dried mycelium is suspended in 40 ml of ice-cold 50 mM potassium phosphate buffer, pH 7.6 (containing 5 mM dithiothreitol and 1 mM EDTA), and ground lightly in a cooled mortar for about 2 min. The mixture is then extruded twice in a cooled French press at 5000-7000 psi. The pH of the extract is adjusted to pH 7.6 and mycelial debris is removed by centrifugation. The resulting supernatant, 38-40 ml, has a protein concentration of 15-20 mg/ml.
[32[
PHENYLACETYL COENZYME A HYDROLASE
485
Alternatively, the mycelium can be ground with buffer and sand until a thin paste is given and then centrifuged. The volume of supernatant is less and the protein concentration about half that achieved using the French press. Step 2. Solid ammonium sulfate is added to a supernatant solution to a concentration of 209 g/liter (35% saturation). The precipitate is removed by centrifugation, and further ammonium sulfate is added to a final concentration of 313 g/liter (50% saturation). The precipitate is collected by centrifugation and dissolved in 2 ml of 50 mM phosphate buffer, pH 7.5, containing 1 mM D T T and 1 mM EDTA. The enzyme at this stage is stable for several weeks at --20% Step 3. The 35-50% ammonium sulfate fraction (2.5 ml) is desalted on a Sephadex G-25 (coarse) column which has been equilibrated with 50 mM phosphate buffer pH 7.5 containing 1 mM D T T and 1 mM EDTA, the elution being carried out with the same buffer. A DEAE-cellulose column (0.9 }( 12.5 cm) is equilibrated with 50 mM phosphate buffer pH 7.5 containing 1 mM D T T and 1 mM EDTA. The desalted extract (12 ml) is absorbed onto the colmnn which is then washed with 20 ml of the equilibrating buffer. The enzyme is eluted from the column using the equilibrating buffer containing 0.1 M KC1. After discarding the void volume, the next 6 ml are collected. Step 4. A Sephadex G-100 column (1.2 X 75 cm), Vo 34 ml, flow rate 20-25 ml/hr, is first equilibrated with 50 mM phosphate buffer pH 7.5 containing 1 mM DTT and 1 mM EDTA. The DEAE fraction (6 ml) is applied to the column and eluted with the same buffer. Two-milliliter fractions are collected, and the most active fractions (fraction numbers 27-31) are pooled (10 ml). The pooled fractions from step 4 represent a 130-fold purification from step 1 with recovery of 25% of the activity. The specific activity is 0.15 unit per milligram of protein.
Properties of the Purified Enzyme
Specificity. It has been suggested 1 that phenylacetyl CoA hydrolase is only one of four activities attributable to a single thiol-dependent enzyme whose action involves a Ping-Pong Bi Bi mechanism with an alternate hydrolytic step. The other activities are penicillin acyltransferase, 6-aminopenicillanic acid acyltransferase, and penicillin acylase. The evidence is based on the constant ratio between the activities during various fractionation and purification procedures, inhibition, and pH and temperature inactivation. In the crude extract (step 1) the ratio of phenylacetylCoA hydrolase to the other activities is higher than in the purified ex-
486
ANTIBIOTIC :BIOSYNTHESIS
[32]
tracts, suggesting the presence of nonspecific enzymes at this stage which can hydrolyze phenylacetyl-CoA. The enzyme is approximately 5.0 times as active against phenoxyacetyl-CoA as compared to phenylacetyl-CoA, 6 and this compares to a similar ratio between the acylase activity of the preparation in hydrolyzing penicillins V and G. The purified preparation also hydrolyzes p-nitrophenyl acetate and phenoxyacetyl glycine at about the same rate as the hydrolysis of phenylacety]-CoA. The ratio of these various activities is the same in steps 3 and 4, but there is no other indication to confirm that the same enzyme is responsible. Other Properties. The hydrolase activity shows a pH optimum of 7.6-7.8 and a Km of 3.95 raM. The molecular weight by Sephadex gel filtration 7 is 25,000. The enzyme is sensitive to sulfhydryl inhibitors and is completely inhibited by pretreatment wth 2 mM N-ethylmaleimide, 2 mM dithiobis(2-nitrobenzoic acid), and 2 mM p-mercuribenzoate. The inhibition is reversible by subsequent addition of excess thiol compounds. The enzyme is sensitive to oxidation during the purification procedure, and it is necessary to keep the enzyme in a reduced state by the presence of thiol compounds. The oxidized enzyme fractionates differently on DEAE-cellulose. Free thiol compounds are not involved in the mechanism of action, and when they are removed by passing the enzyme preparation through Sephadex G-25 and thiols are omitted from the assay mixture, some activity is still observed. However, thiol compounds need to be present during the assay for maximum activity. Despite the presence of 1 mM D T T and 1 mM EDTA the purified preparation is not stable, and about 50% of the activity is lost during storage at 0 ° for 24 hr. Comments on the Assay Procedure. The necessity for free thiol during assay introduces complications due to a rapid S -~ S intermolecular acyl migration between phenylacetic acid and the thiol that occurs at pH above 6.5 and which leads to the elimination of free CoA. When the added thiol does not contain NH.. or OH groups proximal to the SH group (e.g., N-acetyl cysteine, thioglycolic acid, glutathione), the acyl group simply equilibrates between CoA and the added thiol, the rate of equilibration increasing with the pH. When proximal NH2 or OH groups are present in the thiol (e.g., DTT, DTE, cysteine, cysteamine, mercaptoethanol), the S-> S acyl migration proceeds to completion owing to subsequent S--> N and S--> O intramolecular acyl transfer. Further reactions can Chit Maung, Ph.D. Thesis, Dublin University, Ireland, 1970. 7p. Andrews, Biochem. J. 91, 22"2(1964).
[331
ERYTHROMYCIN C 0-METHYLTRANSFERASE
487
include elimination of free [14C]phenylacetic acid and the formation of cyclic compounds t h a t contain [1-14C]phenylacetic acid. These reactions not only alter the concentration of substrate during the reaction, but nonenzymieally liberated [1-1~C]phenylacetic acid and other products, which chromatograph at the same R~, can lead to spurious high results. By avoiding those thiols t h a t can carry out intramoleeular S --> 0 and S ~ N aeyl migrations, interference during the assay can be limited, even at its fullest extent, to the equilibrimn position governed by the amounts of phenylacetyl-CoA and thiol used. This interference can be further limited by keeping the ratio, phenylaeetyl CoA:thiol, high and the p H and assay time low. The assay conditions recomn~en(ted take these points into consideration. In the preparation of the enzyme it is convenient to use buffer containing D T T , and the residual amounts of this compound will produce some nonenzymieally liberated [~4C~]phenylacetie acid. I t is therefore necessary for the control to contain the same buffer as t h a t in which the enzyme is dissolved and for the control to be incubated.
[33] S-Adenosylmethionine: Erythromycin C O-Methyltransferase By JOHN W. CORCORAN E r y t h r o m y c i n s A, B, and C, 1 the macrolide antibiotics elaborated by Streptomyces erythreus have the structures shown in Fig. 1. I t has been demonstrated t h a t the methyl groups attached to the C-3"-,3"-0, and N atoms of the sugars of the erythromycins are derived from L-methionine. -°,3 The O-methylation of the L-mycarose moiety of erythromycin C by a partially purified enzyme obtained from extracts of S. erythreus is described here. The reaction catalyzed is shown in Fig. 1Abbreviations : Ea, the lactone of erythromycin A, also called erythronolide A; D, D-desosaminyl group; M, L-mycarosyl group; C, b-cladinosyl group; EaDM, erythromycin C; EaDC, erythromycin A; EDTA, ethylenediaminetetracetic acid, disodium salt; DTT, dithiothreitol; SAM, S-adenosyl-L-methionine; SAH, S-adenosyl-L-homoeysteine; REV, relative elution volume, defined as a ratio of the elution volume (Ve) over the bed volume (Vt). 2 j. W. Corcoran, J. Biol. Chem. 236, PC 27 (1961). 3j. Majer, M. Puza, L. Dole~ilov£, and Z. Vanek, Chem. Ind. (London) 1961, p. 669 (1961).
[331
ERYTHROMYCIN C 0-METHYLTRANSFERASE
487
include elimination of free [14C]phenylacetic acid and the formation of cyclic compounds t h a t contain [1-14C]phenylacetic acid. These reactions not only alter the concentration of substrate during the reaction, but nonenzymieally liberated [1-1~C]phenylacetic acid and other products, which chromatograph at the same R~, can lead to spurious high results. By avoiding those thiols t h a t can carry out intramoleeular S --> 0 and S ~ N aeyl migrations, interference during the assay can be limited, even at its fullest extent, to the equilibrimn position governed by the amounts of phenylacetyl-CoA and thiol used. This interference can be further limited by keeping the ratio, phenylaeetyl CoA:thiol, high and the p H and assay time low. The assay conditions recomn~en(ted take these points into consideration. In the preparation of the enzyme it is convenient to use buffer containing D T T , and the residual amounts of this compound will produce some nonenzymieally liberated [~4C~]phenylacetie acid. I t is therefore necessary for the control to contain the same buffer as t h a t in which the enzyme is dissolved and for the control to be incubated.
[33] S-Adenosylmethionine: Erythromycin C O-Methyltransferase By JOHN W. CORCORAN E r y t h r o m y c i n s A, B, and C, 1 the macrolide antibiotics elaborated by Streptomyces erythreus have the structures shown in Fig. 1. I t has been demonstrated t h a t the methyl groups attached to the C-3"-,3"-0, and N atoms of the sugars of the erythromycins are derived from L-methionine. -°,3 The O-methylation of the L-mycarose moiety of erythromycin C by a partially purified enzyme obtained from extracts of S. erythreus is described here. The reaction catalyzed is shown in Fig. 1Abbreviations : Ea, the lactone of erythromycin A, also called erythronolide A; D, D-desosaminyl group; M, L-mycarosyl group; C, b-cladinosyl group; EaDM, erythromycin C; EaDC, erythromycin A; EDTA, ethylenediaminetetracetic acid, disodium salt; DTT, dithiothreitol; SAM, S-adenosyl-L-methionine; SAH, S-adenosyl-L-homoeysteine; REV, relative elution volume, defined as a ratio of the elution volume (Ve) over the bed volume (Vt). 2 j. W. Corcoran, J. Biol. Chem. 236, PC 27 (1961). 3j. Majer, M. Puza, L. Dole~ilov£, and Z. Vanek, Chem. Ind. (London) 1961, p. 669 (1961).
488
[33]
ANTIBIOTIC BIOSYNTHESIS
R,J, 2
0
~'1
-I,~I g
O
_ HO,-,,,~IV~eN/Me
,,Ho I
~
-'~/-~CH3
~.z/~ORz
~O.~',~OH CH 3
Rz CH5 B H CH3 C OH H Fie. 1. Structure of erythromycins A, B, and C. ERYTHROMYCIN A
o° c . , . . O . o . +
~O-'~CH3
Erythromycin C
R= OH
I -]5 SPECIFIC ~.~-OD CHi3~ 0 OCH~ H TRANSFERASE Eo 3 O SAM
_P-- -Lo
c. '
Erythromycin A
FIG. 2. Conversion of erythromycin C into erythromycin A in the presence of S-adenosyl-L-methionine (SAM). 2 and the enzyme specifically converts erythromyein C into erythromycin A in the presence of S-adenosyl-L-methionine (SAM).45
Assay I n c u b a t i o n Conditions. Characterization of the transmethylase activity is done in a very simple incubation mixture. As employed for studying the kinetic p a r a m e t e r s of the transmethylase and for determiming possible control factors the components of the incubation are as follows (in micromoles per total volume of 1.5 ml) : Buffer (potassium phosphate) 125, p H 7.5 E D T A , 0.1 D T T , 1.0
4 T. S. McAlpine and J. W. Corcoran, Fed. Proc., Fed. Amer. Soc. Exp. Biol. 30, 1168 (1971). 5 T. S. McAlpine, Enzymatic O-methylation of erythromycin C in the biogenesis of erythromycin A. M. S. Thesis, Northwestern University, 1971.
[33]
ERYTHROMYCIN C 0-METHYLTRANSFERASE
489
Erythromycin C, 0.5, except when specified otherwise [~4CH:,]S-adenosyl-L-methionine, 0.5 (50,000 cpm, except when specified otherwise) Enzyme (6-10 mg of protein, microsomal fraction or ammonium sulfate fraction)
Assay Procedures Two different assays of SAM:EaDM transmethylase activity have been used. The first measures the formation of the product erythromycin A, which is detected by partition chromatography as such or by the formation and purification of the crystalline 2-O-benzoyl derivative. The second assay procedure follows the transformation of the L-mycarosyl moiety of erythromycin C into L-cladinose, which is liberated from the product, erythromycin A, by mild acid hydrolysis. Chromatographic separation of the two neutral sugars, L-mycarose coming from the substrate and L-cladinose from the product, is rapid and complete, the only possible radioactive product being L-cladinose.
Assay I. Measurement of the Formation o] Erythomycin A ]ro~t Erythromycin C. The incubation is carried out for 60 min at 33 ° in a water bath without aeration. The reaction is terminated by the addition of acetone (2 volumes). After 30 min, the precipitated protein is removed by centrifugation, and most of the acetone is removed by evaporation. The aqueous residue is then extracted with methylene chloride (2 times, 2 ml each time). The extract is washed with a minimal volume of water, and the organic phase is concentrated to dryness by vaporation under nitrogen. The residue may then be subjected to partition chromatography (see Chromatographic Methods).
Assay II. Measurement o] the Radioactivity Incorporated into the L-Cladinose Moiety o] Erythromycin A. The enzyme assay is the same as that described for Assay I except that the incubation is terminated by the addition of trichloracetic acid (10% to pH 2, approximately 0.25 ml). Nonradioactive erythromycin A is added (2.0 rag), and the mixture is heated at 90 ° for 30 rain. After cooling in an ice bath, the precipitated protein is removed by centrifugation and washed with water (2 times. 1.5 ml each). The combined supernatant fraction and washings are passed through an Amt)erlite MB-3 resin column (1 cm X 14 cm), the material not adsorbed to the resin is collected, and the column is eluted with a small amount of water (8 ml). Nonradioactive L-cladinose (5.0 my, prepared by hydrolysis of erythromyein A) is added to the combined eluate and wash, and water is removed under reduced pressure. The oily residue is dissolved in a minimal amount of ethyl acetate and chromato-
490
ANTIBIOTIC BIOSYNTHESIS
[33]
graphed on a TLC plate to separate L-mycarose from L-cladinose (see Chromatographic Methods).
Erythromycin C (EaDM) and Its Spiroketal (6 --> 9; 12 --->9) Erythromycin C is not commercially available and current production strains of S. erythreus accumulate little, if any. However, a sample of mother liquors from crystallization of erythromycin A supplied by Abbott Laboratories was somewhat enriched in EaDM. Gel filtration using a lipophilic derivative of dextran (Sephadex LH-20, Pharmacia Fine Chemicals) permits very superior fractionation and purification of the EaDM and other erythromycin derivatives as well. A column (2.0 X 120 cm) of Sephadex LH-20 (particle size 25-100 ~m) in a chloroform/hexane mixture (1:1) with a bed volume of 300 ml is prepared, and the mixture containing EaDM dissolved in the same solvent is applied. Fractions (5 ml) are collected at a flow rate of 0.1 ml/cm 2 per minute. The relative elution volumes (Ve/Vt, or REV) of erythromycins B, A, C are 0.8, 0.85, and 0.95 respectively, while the 6-> 9; 12-~ 9 spiroketals of erythromycin A and C eluted at REV 1.0 and 1.10. Thin-layer chromatography is carried out using silica gel plates (F-254, E. Merck, Darmstadt) in a solvent mixture of ethyl acetate, isopropanol, 15% ammonium acetate (pH 9.7), 27:21:24 (upper phase)2 Anisaldehyde and concentrated sulfuric acid in ethanol (1:1:9) are used as a detection reagent (colors develop after heating). RI values relative to erythromycin A (R~ = 1.0) are 1.05 for erythromycin B and 0.87 for erythromycin C. The RI values of the respective 6--> 9; 12--> 9 spiroketals are 1.12 (A) and 0.93 (C).
Erythromycin A Commercial antibiotic (Abbott) is repeatedly recrystallized by dissolving in a minimum amount of ice water and then warming to 37 ° (negative temperature coefficient of solubility) until the crystalline material is homogeneous (TLC, partition paper chromatography, etc., see below). It can also be recrystallized from organic solvents (isopropanol, chloroform, etc.).
S-Adenosyl-L- methionine (SAM) Commercial material, both nonradioactive and radioactive, is used. As usually supplied, it is contaminated with a substantial amount of 6 W. Malczewska-Konecka, Z. Piekarska, and Z. Kowszyk-Gindifer, Chem. Anal. (Warsaw) 14 (5), 1093 (1969).
[33]
ERYTHROMYCIN C 0-METHYLTRANSFERASE
491
S-adenosyl-L-homocysteine. As with many transmethylases, 7-9 the enzyme described here is inhibited by this end product of the enzymic transmethylation reaction. Although the reaction can be monitored with unpurified substrate, the rate is quite dependent on the state of its purity. For studies of the kinetic parameters of the enzyme, the commercial SAM is purified by ion-exchange chromatography. The methods of Schlenk and D e P a l m a TM and Shapiro and Ehringer 11 are suitable.
Purification Gro~vth o] S t r e p t o m y c e s erythreus
A slant culture of S. erythreus (Abbott-CA340) is used to inoculate (dry loop) 50 ml of a sterile soluble vegetative medium (glucose, 5 g; commercial brown sugar, 10 g; tryptone, 5 g; yeast extract, 2.5 g; E D T A , 36 mg; betaine, 1.29 g; sodium propionate, 0.11 g; tap water, 1100 ml; final pH 7.0-7.2 adjusted with K O H ) contained in 500 ml wide-mouth Erlenmeyer flasks. The flasks which are closed with 2 milk filter disks (Filter Fabrics Inc., Goshen, Indiana) secured with rubber bands are shaken at 33 ° for 2.5 to 3 days on a New Brunswick gyratory shaker at 200 rpm. The resultant culture is used (10% v / v ) to inoculate 350 ml of the same medium contained in 2-liter Erlenmeyer flasks, and after growth under the same conditions for 24 hr the second culture is used to inoculate (10%, v / v ) a New Brunswick Microferm fermentor, containing 10 liters of the same medium. The fermentor is stirred at 500 rpm at an aeration rate of 10 liters/min. The growth obtained after 12 hr at 33 ° (75 g net weight) is harvested by continuous flow centrifugation at 27,000 g using a Sorvall RC-2B centrifuge. It is washed twice, each time by suspension in 2 volumes of 10 m M phosphate buffer (pH 7.5) containing 0.1 m M E D T A and 0.1 m M D T T (Buffer A) and stored at --80 ° with 40% glycerol ( v / v ) . Culture conditions for S. erythreus have been described by Kaneda et al. 1'~ and Friedman et al. ~3 A. Y. Akamatsu and J. L. Law, J. Biol. Chem. 245, 709 (1970). 8T. Deguchi and J. Barchas, J. Biol. Chem. 246, 3175 (1971). V. Zappia, C. R. Zydek-Cwick, and F. Schlenk, J. Biol. Chem. 244, 4499 (1969). ~°F. Schlenk and R. E. DePalma, J. Biol. Chem. 229, 1051 (1957). 11S. K. Shapiro and D. J. Ehrimger, Anal. Biochem. 15, 323 (1966). 12T. Kaneda, J. C. Butte, S. B. Taubman, and J. W. Corcoran, J. Biol. Chem. 237, 322 (1962). 1~S. M. Friedman, T. Kaneda, and J. W. Corcoran, J. Biol. Chem. 239, 2387 (1964).
492
ANTIBIOTIC BIOSYNTHESIS
[33]
SAM :EaDM Transmethylase Cell-Free Extract and Microsomal Fraction. Either freshly collected or frozen and thawed cells of S. erythreus suspended in 0.1 M phosphate buffer (pH 7.5) containing 0.1 mM EDTA and 1 mM D T T (Buffer B) are disrupted by a single passage through a French pressure cell at pressures between 8000 and 10,000 psi, and the exudate is centrifuged at 48,000 g for 30 min at 5 °. The resultant supernatant solution may be used as a crude cell-free enzyme preparation. For other experiments this crude extract is centrifuged at 105,000 g for 150 min in a Spinco Model L2-65B preparative ultracentrifuge, and the resultant microsomal pellet was then homogenized gently in a Thomas tissue grinder with one-fourth of the original volume (crude extract) of buffer B. Ammonium Sul]ate Precipitation. All operations are conducted at 5 ° . The crude cell-free extract is gradually brought to 70% of saturation with ammonium sulfate by adding the solid salt (490 mg/ml during 30 min). After 30 rain of equilibration the precipitated protein is collected by centrifugation. The precipitate contains the transmethylase activity, and it can be stored at --15 ° for 1 month without significant loss of activity. For assays, the precipitate is diluted to 25% of the original (crude extract) volume of the crude extract with buffer B and dialyzed for 2 hr with three changes of 100 volumes of phosphate buffer (50 mM, pH 7.5) containing EDTA and D T T (0.1 mM each). The dialyzed enzyme can be kept at 4 ° for several days without significant loss in activity. For studies of enzyme properties and substrate specificity, a further fractionation by ammonium sulfate may be done. In the first step, the crude cell-free extract is brought to 30% of saturation with solid ammonium sulfate (210 mg/ml). After 30 rain, the precipitate is removed by centrifugation and discarded. The supernatant solution is brought to 60% of saturation with ammonium sulfate by adding more solid salt (210 mg/ml). The precipitated protein contains the transmethylase activity.
Chromatographic Methods
Partition Chromatography of Erythromycins A and C. The method is based on the procedure described briefly by Friedman et al. 1" It is a somewhat difficult method, especially when compared to TLC chromatography (above and below), but affords superior resolution of closely related members of the erythromycin family and their derivatives. For analytical work and small-scale preparative separations, the chromatog-
[33]
ERYTHROMYCIN C 0-METHYLTRANSFERASE
493
raphy is done with filter paper as the support for the stationary phase (Whatman 1 or 3 MM, depending on sample amount). The solvent system employed consists of benzene (50 ml), n-heptane (25 ml), acetone (15 ml), isopropanol (10 ml), and potassium phosphate buffer (25 ml, pH 7.0). The paper is impregnated with the stationary (lower) phase of the system and blotted briefly between dry sheets of filter paper. The residue to be analyzed is applied as a solution in acetone, and the chromatogram is then developed by the descending method using the upper phase of the system. Depending on the temperature and relative humidity of the laboratory, the time of development (2-4 hr) and the relative mobilities of the two erythromycins (and other derivatives) vary a great deal. In :::any of the past experiments, erythromycin A had an R r value of 0.6-0.7, that of erythromycin C was much lower, in the range of 0.2-0.3. The product of an enzyme assay is run in a lane separated from the remainder of the chromatogram by a narrow gap cut on either side (scalpel or razor blade), and appropriate standards are put on the same sheet. After removal of the unknown lane(s), the standard samples of erythromycins C and A are detected by spraying with a vanillin (0.5% by weight) perchloric acid (1 M in absolute ethanol) reagent and gentle heating in the absence of air (e.g., in Saran foil) at about 80 ° (grayish blue spots on a light background). The lane to which the residue from the assay system is applied is analyzed with a paper strip counter and the amount of radioactivity with the same migration rate as the reference sample of erythromycin A is a measure of the activity of the SAM:EaDM transmethylase. The chromatographic method described is also adaptable to column separations. In this case the stationary phase is :nixed with silica gel (0.5-1.0 g per gram of gel, giving a friable powder) and packed in a column under the mobile phase of the system. The sample is dissolved either in a m:nimal amount of the mobile phase and added to the top of the column or (if relatively insoluble) it is dissolved in a similar volume of stationary phase. In the latter case the solution is mixed with sufficient silica gel to give a seemingly dry powder, and this is layered on the top of the column. Development with the mobile phase is started and fractions are collected. The relative elution volumes are in the same sequence as the relative migrations of the substances on paper, but the absolute elution volumes are not easily predicted from the R~ values. TLC Chromatography. TLC analysis of the erythromycins is described above. The sugars liberated from the erythromycins by acid hydrolysis also are conveniently separated by TLC chromatography. Plates coated with silica gel are used together with a solvent composed of benzene and acetone (7:3, v/v). When the results of enzyme assays with
494
ANTIBIOTIC BIOSYNTHESIS
[331
radioactive SAM are analyzed, the unknown substances are placed on one side of the plate and reference samples of L-myearose and L-eladinose are placed on the other. After development of the plate, the unknown side is covered with aluminum foil or Saran foil and the standard side is sprayed with the vanillin/perchlorie acid reagent (above). Color development is achieved by gentle drying and heating at 80 °. The sugars afford purplish spots on a light background. The side of the plate containing the radioactive unknowns is divided into zones (ca. 2 X 2 em per lane), and each is scraped into a scintillation vial for measurement of total radioactivity. L-Myearose has an Rf value of 0.25 in this TLC system, and L-eladinose has one of 0.55. Derivatization of Erythromycin A. It is useful, in standardizing the assay for SAM:EaDM transmethylase, to make an independent check of the radiopurity of the erythromyein A produced. It also serves as a control on the functioning of the partition chromatographic separation of the EaDC from possible radioactive contaminants. For this purpose, the erythromyein A produced in an enzyme assay (Assay I) may be diluted with carrier nonradioactive EaDC (17.6 mg, 0.024 mmoles) and a monobenzoyl derivative prepared as described by Kaneda et al. 1~ The combined sample of the carrier erythromyein A and the residue from the assay are dissolved in dry acetone (1 ml) and treated with a solution of benzoic anhydride (8.35 mg, 0.037 mmole) in dry acetone (1 ml). The reaction is conveniently done in a conical centrifuge tube (12 ml) and the solutions are warmed briefly to the boiling point using a fine glass rod to prevent bumping. The mixture is left overnight at 25 ° in the dark, and then the acetone is evaporated under a stream of nitrogen. The oily liquid residue is dissolved in ethanol (95%), and the solution is passed through a small column (1 cm diameter X 5 em in length) of Amberlite IR-45 resin (free-base form) which has been washed with ethanol. The effluent solution plus several washes of the resin (3-5 times the bed volume) contains the etude monobenzoyl derivative of erythromyein A. It is free of excess benzoic anhydride and has been converted to the freebase form of the derivative (product is a benzoate salt prior to the resin treatment). After removal of the ethanol by evaporation under reduced pressure, the 2'-O-benzoyl derivative of erythromyein A is crystallized from minimal amounts of boiling isopropanol (ca. 0.3 ml). Recrystallization is done in the same manner, until a constant specific radioactivity is achieved. The specific radioactivity of the derivative will drop on the first crystallization and rise slightly on successive crystallizations if EaDC is the major radioactive substance isolated from the enzyme assay. Miscellaneous. All reagents not otherwise specified are commercial products used as obtained. Ammonium sulfate is "enzyme grade" as sup-
[33]
E R Y T H R O M Y C ICN 0-METHYLTRANSFERASE
495
plied by Schwarz/Mann. Solvents are analytical reagent grade. The antibiotics tylosin and niddamycin are products of Eli Lilly Co. and Abbott Laboratories, respectively. The sugars, L-mycarose and its anomeric methyl glycosides and L-cladinose, are prepared by the mild acid hydrolysis of tylosin TM and erythromycin A, 15 respectively. The methods used are those described in the literature, and no special purification is done. Mycarosyl erythronolide B, a monoglycoside of L-mycarose and the laetone of erythromyein B (lacking a lactone ring hydroxy group at C12) is a product of Abbott Laboratories. Radioactivity is measured with a liquid scintillation spectrometer (Nuclear Chicago Mark I with use of an external standard and a quench correction) and with a paper strip scanner (Nuclear Chicago Actigraph III). The scintillation system used contains naphthalene (125 g), 2,5diphenyloxazole (7 g), and 1,4-bis-2-(5-phenyloxazolyl)benzene in 1 liter of p-dioxane. Protein is measured by the biuret method; using bovine serum albumin as a standardJ ~
Properties Product o] the Transrnethylase Reaction. The sole radioactive product of the reaction between erythromycin C and S-adenosyl-L-methionine (14CH~) appears to be erythromycin A. This is shown by the apparent homogeniety of the radioactive substance which migrates with the same RI value as EaDC in a partition chromatographic system. This material also cochromatographs with added radioactive EaDC. Further confirmation of the radiopurity of the EaDC is afforded by finding that a monobenzoyl derivative [2'-O-benzoate] retains its radioactivity on repeated crystallization. Reversibility o] the Transmethylase Reaction. Incubation of radioactive erythromycin A, N-[14CH3]EaDC, of very high specific radioactivity obtained from Abbott Laboratories, with the SAM:EaDM transmethylase (mierosomal fraction or ammonium sulfate precipitated protein) yields no detectable erythromycin C (Assay I workup conditions). Thus the transmethylase under these conditions appears to be unable to carry out the transformation of erythromycin A into the C form of the antibiotic. Cellular Location o] the Transmethylase. On comparing the total enzymic activity in a crude cell extract with that in the microsomal frac14R. B. Morin, M. Gorman, R. L. Hamill, and P. V. Demarco, Tetrahedron Lett. 1970, 4737 (1970). ~5F. Wiley, R. Gale, C. W. Pettinga, and K. Gerzon, J. Amer. Chem. Soc. 76, 3121 (1955). '~A. G. Gornall, C. S. Bardawill, and M. M. David, J. Biol. Chem. 177, 751 (1949).
496
ANTIBIOTIC BIOSYNTHESIS
[33]
S-ADENOSYL-L-METHIONINE-DEPENDENT CONVERSION OF ERYTHI~OMYCIN C INTO ERYTHROMYCIN A a
Preparation
Total activity (~moles/hr)
Crude cell-free extract 105,000 g supernatant fraction 105,009 g microsomal fraction
387 67 171
Protein (mg)
275.5 189 75.6
Specific activity (~moles/mg protein/hr) 1.23 0.36 2.97
" Assays conducted as described in Assay Method II. tion, it is apparent that the enzyme is associated with the microsomal material (see the table). The specific activity of this microsomal fraction is approximately 8 times that of the remaining supernatant protein and at least twice that of the original crude extract. Kinetic Properties o] the Transmethylase. The conversion of EaDM to EaDC (Assay II) is dependent on SAM and enzyme. It is linear under the conditions used to about 15 mg of enzyme protein per assay and then falls off slowly up to 25 mg of enzyme protein per incubation. The rate of transmethylation is linear for nearly 60 min and continues at a slightly lower rate for another 60 rain. The effect of pH on the transmethylase activity is not dramatic, and a broad optimum is observed between pH 7.5 and 8.8. The maximum rate is seen at about pH 8.2-8.3. The pK~ of erythromycin C is in the range of pH 8.5-9.0 and the enzyme may use the unprotonated species as its substrate. The temperature dependence of the transmethylase is striking, with a fairly sharp optimum seen at 33-35 °. The strain of S. erythreus used (CA340) as a source of the enzyme grows well between 30 ° and 37 °, but its optimum temperature is also very near 33-34% The amount of EaDM used in the assay is nearly optimal, since the dependency of the enzyme on this substrate is nearly linear to about 0.2 ~mole per assay and falls off rapidly above this value. The amount of EaDM used (0.5 ~mole of EaDM) is seemingly in a nearly saturating range, but changes in conditions may affect this markedly. The products of the enzymic reaction [EaDC and S-adenosyl-L-homocysteine (SAH) ] markedly inhibit the transmethylase activity (43% and 76%, respectively, when 0.5 ~mole of each is added to separate incubation systems). The amount of SAM used to measure enzyme activity is well below the optimal amount, and with a constant amount of EaDM (0.5 ~mole) there is a maximum rate of enzyme activity when 4 ~moles of SAM are present. Increasing the level of SAM above this value produces a reduced rate of transmethylase activity, probably because of end-product inhibition.
[33]
ERYTHROMYCIN C 0-METHYLTRANSFERASE
497
An approximate K,~ for erythromycin C in the SAM:EaDM transmethylase reaction of S. erythreus (CA340) was calculated as 0.30 raM. This value and the total conversion of this substrate to product (ca. 3%) is probably very dependent on the assay conditions, and this K,~ value cannot be more than a rough indication of the enzyme's affinity for erythromycin C. Specificity of the SAM:EaDM Transmethylase. The enzyme shows a very high degree of substrate specificity. Aside from erythrolnycin C, it fails to catalyze the methylation of any L-mycarosyl moiety tested. Substances tested were 3-O-mycarosyloxyerythronolide B, the (6 ~ 9:12--~ 9)-spiroketal of erythromycin C, L-mycarose, the anomeric methyl L-mycarosides, and the antibiotics tylosin and niddamycin) ~ The latter two substances have a mycarosyl moiety present as part of a disaccharide attached to a 16-membered lactone whose structure is not very similar to that in erythromycin. A fourth member of the erythromycin family has been identified in this laboratory. It is called erythromycin D and has the same two sugars as does erythromycin C--namely L-mycarose and n-desosamine. The lactone of erythromyein D is that of erythromycin B and it lacks one tertiary hydroxyl group present in erythromycin C (at C-12). The protein (crude and ammonium sulfate precipitated) used in the study of tho EaDM:SAM transmethylase catalyzes the transformation of erythromycin D (EbDM) into erythromyein B. It remains to be proved that the EaDM:SAM and EbDM:SAM transmethylases are one and the same but it is likely that they are. If so, some variation in lactone structm'e is compatible with enzymic activity (hydrogen vs hydroxy at C-12). Space-filling models of erythromyeins C and A show that the neutral sugar in each is closely aligned (almost parallel) with the basic sugar and that both are roughly perpendicular to the plane of the laetone ring. 's It is somewhat difficult to see how the SAM:EaDM transmethylase can act at all, and it is hard indeed to comprehend the strict substrate specificity that it appears to possess. Significance of the S-Adenosyl-L-methionine :Erythromycin C Transmethylase The best evidence available from studies of physiologically intact S.
erythreus strains indicates that erythromycin C is a precursor of erythro~: G. Huber. K. H. Wallhiiusser, L. Fries, A. Steigler, and H. Weidenmtiller. Arzneim.-Forsch. 12, 1191 (1962). ~ T . J. Perun, in "Drug Action and Drug Resistance in Bacteria" (S. Mitsuhashi, ed.), Vol. I, p. 123. Univ. Park Press, Baltimore, Maryland, 1970.
498
ANTIBIOTIC BIOSYNTHESIS
[34]
mycin A. 10 The properties of the enzyme that produces EaDC from EaDM support this relationship. The potent inhibitory effect of both products, EaDC and SAH, on the methylation reaction also indicates that the activity of the transmethylase may be one control point in the overall biogenesis of erythromycin A. The latter accumulates in very high concentration in the fermentation beer of commercially useful strains of S. erythreus (strain CA340 is a former production strain of Abbott Laboratories), and it may be surmised that some barrier separates the S A M : E a D M transmethylase from this reservoir of inhibitor. The intracellular concentration of SAM is not known, but evidence exists (unpublished) to suggest that methylation steps in general are rate limiting in the biogenesis of the erythromycins. It is possible that the intracellular concentration of SAM and SAH is quite low and that recycling of the SAH to SAM reduces the concentration of this end-product inhibitor. The SAM:SAH ratio could be a major regulatory factor of the final step in erythromycin A formation. Since very little if any erythromycin C accumulates in strains of S. erythreus that are useful for erythromycin A accumulation, the activity of the EaDM:SAM transmethylase clearly is sufficiently great to prevent buildup of this direct precursor of erythromycin A. 19j. R. Martin and A. W. Goldstein, Proc. Int. Congr. Chemother. 6th, 2, 1112 (1970).
[34] S - A d e n o s y l m e t h i o n i n e : I n d o l e p y r u v a t e 3-Methyltransferase
By MARILYN K. SPEEDIE, ULFERT HORNEMANN, and HEINZ G. FLoss O II ~CH2--C--COOH H
+ S-Adenosylmethionine
CHs O F }l ~ C H - C--COOH
~
H
+ 5-S - Adenosylhomocysteine
The enzyme is isolated from an indolmycin-producing strain of Streptomyces griseus and is presumably involved in the biosynthesis of this antibiotic. The reaction catalyzed by this enzyme represents the first spe-
498
ANTIBIOTIC BIOSYNTHESIS
[34]
mycin A. 10 The properties of the enzyme that produces EaDC from EaDM support this relationship. The potent inhibitory effect of both products, EaDC and SAH, on the methylation reaction also indicates that the activity of the transmethylase may be one control point in the overall biogenesis of erythromycin A. The latter accumulates in very high concentration in the fermentation beer of commercially useful strains of S. erythreus (strain CA340 is a former production strain of Abbott Laboratories), and it may be surmised that some barrier separates the S A M : E a D M transmethylase from this reservoir of inhibitor. The intracellular concentration of SAM is not known, but evidence exists (unpublished) to suggest that methylation steps in general are rate limiting in the biogenesis of the erythromycins. It is possible that the intracellular concentration of SAM and SAH is quite low and that recycling of the SAH to SAM reduces the concentration of this end-product inhibitor. The SAM:SAH ratio could be a major regulatory factor of the final step in erythromycin A formation. Since very little if any erythromycin C accumulates in strains of S. erythreus that are useful for erythromycin A accumulation, the activity of the EaDM:SAM transmethylase clearly is sufficiently great to prevent buildup of this direct precursor of erythromycin A. 19j. R. Martin and A. W. Goldstein, Proc. Int. Congr. Chemother. 6th, 2, 1112 (1970).
[34] S - A d e n o s y l m e t h i o n i n e : I n d o l e p y r u v a t e 3-Methyltransferase
By MARILYN K. SPEEDIE, ULFERT HORNEMANN, and HEINZ G. FLoss O II ~CH2--C--COOH H
+ S-Adenosylmethionine
CHs O F }l ~ C H - C--COOH
~
H
+ 5-S - Adenosylhomocysteine
The enzyme is isolated from an indolmycin-producing strain of Streptomyces griseus and is presumably involved in the biosynthesis of this antibiotic. The reaction catalyzed by this enzyme represents the first spe-
[34]
499
INDOLEPYRUVATE 3-METHYLTRANSFERASE
cific step in indohnycin biosynthesis according to the pathway postulated by Hornemann et al2
Assay Method Principle. The enzyme catalyzes the transfer of the methyl group from [methyl-14C]S-adenosylmethionine to the 3-position of the aliphatic side chain of indolepyruvate. The radioactive reaction product can be extracted from the acidified reaction mixture with butyl acetate and then counted in a liquid scintillation spectrometer. Reagents
KH2PO4/Na2HPQ, 10 mM, pH 7.5 Indolepyruvate, 2.5 mM, pH 7.5 [methyl-14C]S-Adenosylmethionine, specific mmole, 2.5 mM
activity
0.20
mCi/
Procedure. Incubations are carried out in 15-ml centrifuge tubes. The reaction mixture contains 0.25 ~mole of indolepyruvate, 0.25 ~mole (0.05 ~Ci) [methyl-14C]S-adenosylmethionine, and enzyme in 10 mM phosphate buffer, pH 7.5, in a total volume of 1.0 ml. The reagents are prepared immediately before addition to the incubation mixture. Incubation is at 30 ° for 1 hr. Boiled enzyme serves as a control. The reaction is terminated by acidification to pH 3 with tartaric acid. One milliliter of distilled water and 4 ml of butyl acetate are added. The mixture is agitated on a Vortex mixer for 30 sec, centrifuged for 5 min, and the top layer (butyl acetate) is withdrawn and placed in a second centrifuge tube. Then 1 ml of water is added to the organic phase, and the mixing and centrifugation are repeated. Two milliliters of the butyl acetate phase are withdrawn and placed in a counting vial. When crude and partially purified enzyme preparations are assayed, S-adenosylmethionine is partially decomposed during the reaction period to form radioactive methanol which interferes with the determination of radioactivity in the product. To overcome this problem, any radioactive methanol is removed by repeated addition of unlabeled methanol and evaporation under nitrogen. With enzyme solutions in the later stages of purification there appears to be no methanol formed, so the butyl acetate phase is simply taken to dryness once under a stream of nitrogen. Ten milliliters of toluene scintillation fluid are added to each vial, and the samples arc counted in a liquid scintillation spectrometer.
1U. Hornemann, L. H. Hurley, M. K. Speedie, and H. G. Floss, J. Amer. Chem. Soc. 93, 3028 (1970).
500
ANTIBIOTIC BIOSYNTHESIS
[34]
Definition o] Activity. One unit of enzyme is defined as that quantity of enzyme which will convert 1 t~mole of substrate in 1 rain as measured by the incorporation of the methyl group of S-adenosylmethionine into 3-methylindolepyruvate under the conditions described. Specific activity is expressed as milliunits per milligram of protein. Protein is estimated by the biuret method 2 for steps 1 and 2 and by absorption at 280 nm for the later steps2 Purification Procedure All steps are performed at 0-4 ° unless otherwise stated. Step 1. Growth o] Cultures and Preparation o] the Crude Cell-Free Extract. A strain of Streptomyces griseus (ATCC 12648) is used as the source of the enzyme. The organism is maintained on slants of Emerson agar at 24 °. For production of cells to be used in enzyme preparations, the culture is first grown on a medium described by Rao 4 which contains dextrose, 1.0 g; K2HPO4, 0.5 g; NaCI, 0.2 g; CaC03, 0.20 g; distiller's solubles, 0.25 g; soybean meal, 0.15 g; and distilled water to 100 ml. A part of the mycelial pad from a slant culture is transferred under sterile conditions to a 500-ml Erlenmeyer flask containing 100 ml of the above medium and is allowed to grow for 4-7 days on a rotary shaker at 180 rpm at 24 °. From this culture 2-ml aliquots are withdrawn and transferred to 500-ml Erlenmeyer flasks containing "Phytone" (Baltimore Biological Laboratories), 2.0 g; yeast extract (Difco), 0.2 g; trace element solution, 5 0.1 ml; FeCI~, 0.1 rag; and distilled water to 100 ml. After 30-42 hr of growth in the latter medium, the mycelium is harvested by vacuum filtration and washed twice with distilled water. The mycelium from each flask is suspended in 15 ml of 10 mM phosphate buffer, pH 7.0. The cells are broken by one passage through a French pressure cell at 15,000-20,000 psi. The resulting suspension is centrifuged at 30,000 g for 20 rain to remove cell debris. Step 2. Ammonium Sul]ate Treatment. Solid ammonium sulfate is added over a period of 15 rain to give 35% saturation. The pH is adjusted to 7.0 with solid Na2HPO4, and the solution is stirred for a further 40 rain. Following 20 min of centrifugation at 30,000 g, ammonium sulfate is added to the supernatant in a similar manner to achieve 55% saturation, keeping the pH at 7.0. The protein that precipitates between 35 A. G. Gornall, C. J. Bardawill, and M. M. David, J. Biol. Chem. 177, 751 (1949); see also this series, Vol. 3 [73]. O. Warburg and W. Christian, Biochem. Z. 310, 384 (1941). K. V. Rao, Antibiot. Chemother. (Washington, D.C.) 10, 312 (1960). E. J, Kirsch and J. D. Korshalla, J. Bacteriol. 87, 247 (1964).
[a41
INDOLEPYRUVATE 3-METHYLTRANSFERASE
501
and 55% saturation contains the enzyme activity and is dissolved in 5-10 ml of 10 mM phosphate buffer, pH 7.0, and dialyzed against two changes of 2 liters of the same buffer for 1.5 hr each. Step 3. Sephadex Chromatography. The dialyzed solution is then applied to a 2.5 X 33 cm column of Sephadex G-150 which has been equilibrated with 10 mM phosphate buffer, pH 7.0, and is eluted with the same phosphate buffer. Fractions of 5 ml each are collected. The C-methyltransferase is recovered in fractions 15 through 22. The total volume of the active fractions is reduced to approximately 4 ml using an Amicon pressure dialysis apparatus. Step 4. DEAE-Sephadex Chromatography. The combined fractions from the previous step are applied to a 1.8 X 23 cm column of DEAESephadex which has been equilibrated with 50 mM phosphate buffer, pH 6.8. A 500-ml gradient of 0-0.4 M NaC1 in 50 mM phosphate buffer, pH 6.8, is used to elute the column. Fractions of 5 ml each are collected and monitored for protein and C-methyltransferase activity. The volume of the combined active fractions is reduced to approximately 5 ml using an Amicon pressure dialysis apparatus. Step 5. Bio-Gel A-5m Chromatography. The enzyme preparation from the preceding step is applied to a 1.0 )< 25 cm column of Bio-Gel A-5m which has been equilibrated with 10 mM phosphate buffer, pH 7.0, and subsequently is eluted with the same buffer. One-milliliter fractions are collected. The enzyme activity is recovered in fractions 9 ~hrough 12. By tim above procedure, the enzyme may be purified approximately ll0-fold with an overall yield of 40-45%. The enzyme can be stored at 2 ° for at least 3 weeks without significant loss of activity. A typical purification is summarized in the table.
Properties
Egect of pH. The enzyme is optimally active between pH 7.5 and 8.5. There appears to be some dependence upon the buffer, since in phosSUMMARY OF ENZYME PURIFICATION
Fraction 1. 2. 3. 4. 5.
Crude extract Ammonium sulfate Sephadex G-150 DEAE-Sephadex Bio-Gel A-5m
Total protein (rag)
Activity (milliunits)
Specific activity (mU/mg)
Yield (%)
358.0 171.0 40.2 2.2 1.5
28.0 22.5 22.0 12.8 12.5
0. 078 0.13 0.55 5.82 8.30
100 80.7 78.3 45.8 44.6
502
ANTIBIOTIC
[35]
BIOSYNTHESIS
phate buffer the enzyme is slightly more active at pH 7.5 than at pH 8.0, whereas in other buffers the optimum pH is slightly higher. The enzyme is irreversibly inactivated at pH 5.5 and below. Specificity. The crude extract is capable of methylating phenylpyruvate and p-hydroxyphenylpyruvate, in addition to indolepyruvate, but this capability is lost upon purification of the indolepyruvate 3-methyltransferase. Kinetic Properties. The Km values for indolepyruvate and S-adenosylmethionine are 4.8 ~M and 13 ~M, respectively. Molecular Weight. The molecular weight of the enzyme is estimated to be 55,000 ± 5000 by Sephadex G-200 gel filtration with reference proteins. Inhibitors. The reaction is strongly inhibited by the thiol reagents p-chloromercuribenzoate (10 ~M) and N-ethylmaleimide (4.0 mM). The Zn ~÷ and Fe 2÷ chelators 1,10-phenanthroline and 2,2P-bipyridine also inhibit the enzyme activity. At 2.0 mM the former compound inhibits 60%, the latter 35%. Ethylcnediaminetetraacetic acid (EDTA) has no effect upon the activity in concentrations up to 5 raM.
[35] N o v o b i o c i c A c i d S y n t h e t a s e By L. A. KOMINEK and H. F. MEYER OH J,~NH2
HOOC. ~
,,CH3 ~CH2CH=C
OH ~-~ ~
0II FCH5 ~NH--C. ~.~ ~CH2CH=C
+ H 0 J'-,~TI"~0 ~ 0 CH3
B ring
r3-Amino-4,7- dihydroxy8-methyl coumorin]
l
CH3
A ring
Novobiocic ocid
~4- Hydroxy-3(3- methy!-2bulenylJbenzoic ocid'l
The antibiotic novobiocin consists of a noviose sugar (C ring), a coumarin moiety (B ring), and a substituted benzoic acid moiety (A ring) linked by a glycosidic and an amide bond as shown in Fig. 1. The enzyme forming an amide bond between the A ring and the B ring to produce novobiocic acid has been demonstrated in cell-free extracts of Streptomyces niveus, z The reaction requires adenosine triphosphate (ATP), which indicates that an activation of the carboxyl group of the A ring is involved. The mode of activation has not been determined. 1L. A. Kominek, Antimicrob Ag. Chemother. 1, 123 (1972).
502
ANTIBIOTIC
[35]
BIOSYNTHESIS
phate buffer the enzyme is slightly more active at pH 7.5 than at pH 8.0, whereas in other buffers the optimum pH is slightly higher. The enzyme is irreversibly inactivated at pH 5.5 and below. Specificity. The crude extract is capable of methylating phenylpyruvate and p-hydroxyphenylpyruvate, in addition to indolepyruvate, but this capability is lost upon purification of the indolepyruvate 3-methyltransferase. Kinetic Properties. The Km values for indolepyruvate and S-adenosylmethionine are 4.8 ~M and 13 ~M, respectively. Molecular Weight. The molecular weight of the enzyme is estimated to be 55,000 ± 5000 by Sephadex G-200 gel filtration with reference proteins. Inhibitors. The reaction is strongly inhibited by the thiol reagents p-chloromercuribenzoate (10 ~M) and N-ethylmaleimide (4.0 mM). The Zn ~÷ and Fe 2÷ chelators 1,10-phenanthroline and 2,2P-bipyridine also inhibit the enzyme activity. At 2.0 mM the former compound inhibits 60%, the latter 35%. Ethylcnediaminetetraacetic acid (EDTA) has no effect upon the activity in concentrations up to 5 raM.
[35] N o v o b i o c i c A c i d S y n t h e t a s e By L. A. KOMINEK and H. F. MEYER OH J,~NH2
HOOC. ~
,,CH3 ~CH2CH=C
OH ~-~ ~
0II FCH5 ~NH--C. ~.~ ~CH2CH=C
+ H 0 J'-,~TI"~0 ~ 0 CH3
B ring
r3-Amino-4,7- dihydroxy8-methyl coumorin]
l
CH3
A ring
Novobiocic ocid
~4- Hydroxy-3(3- methy!-2bulenylJbenzoic ocid'l
The antibiotic novobiocin consists of a noviose sugar (C ring), a coumarin moiety (B ring), and a substituted benzoic acid moiety (A ring) linked by a glycosidic and an amide bond as shown in Fig. 1. The enzyme forming an amide bond between the A ring and the B ring to produce novobiocic acid has been demonstrated in cell-free extracts of Streptomyces niveus, z The reaction requires adenosine triphosphate (ATP), which indicates that an activation of the carboxyl group of the A ring is involved. The mode of activation has not been determined. 1L. A. Kominek, Antimicrob Ag. Chemother. 1, 123 (1972).
[35]
NOVOBIOCIC ACID SYNTHETASE 0 II
O-C -NH2
OH
0
503
CH$ =C~, CH5 o.
CH3
t
CH3
Jl C Ring
Jl B Ring
A Ring
I novobiocic ocid
FIG. 1. Structure of novobiocin.
Assay Method
Principles. Novobiocic acid synthetase can be assayed by measuring the rate of novobiocic acid formation. The quantity of novobiocic acid produced is measurable by spectrophotometric methods after extraction into n-butyl acetate. Reagents Ring A [4-hydroxy-3(3-methyl-2-butenyl)benzoic acid] 10 raM. A solution is prepared by dissolving 2.06 mg/ml of the A ring in 0.1 M Tris buffer, pH 8.0. Ring B (3-amino-4,7-dihydroxy-8-methylcoumarin) 10 mM. A solution is prepared by dissolving 2.44 mg/ml of the B ring hydrochloride in 0.1 M Tris buffer, pH 8.0. The solution is freshly prepared under a nitrogen atmosphere to retard degradation of the B ring. ATP, 50 mM in 0.1 M Tris buffer, pH 8.0 n-Butyl acetate Tris. HC1 buffer, 2.0 M, pH 9.0 Potassium phosphate buffer, 0.5 M, pH 6.5 Enzyme: the cell-free extract in 50 mM Tris buffer is used undiluted. The final reaction mixture should contain 5-20 units per milliliter of enzyme activity.
Procedure. The reaction mixture in both the experimental and blank tubes contains: A ring, 2 ml; B ring, 2 ml; ATP, 1 ml; Tris.HC1 buffer, 1 ml (2.0 M). The reaction is started by the addition of 1-4 ml of enzyme to the experimental tubes. Final volumes of experimental and blank tubes are brought to 10 ml with water. The reaction mixture is preincubated at 30 ° prior to the addition of enzyme. Samples (1 ml) are removed at
504
ANTIBIOTIC BIOSYNTHESIS
[35]
1-min intervals (0-6 min) and placed in a 125-ml glass-stoppered Erlenmeyer flask containing 15 ml of n-butyl acetate and 1 ml of potassium phosphate buffer. The flasks are shaken immediately for 15 rain to stop the reaction and extract novobiocic acid. The butyl acetate is separated from the aqueous phase and its absorbance is determined in a UV spectrophotometer at 360 and 310 nm against an n-butyl acetate blank. The B ring interferes in this assay but is corrected for by the use of a two-component equation: The concentration of novobiocic acid in the presence of B ring can be calculated from the following equation derived from the absorbancies of novobiocie acid (A36o = 42.70; A31o = 25.81) and the B ring (A36o = 3.46; A3,o = 13.05). Novobiocic acid (~g/ml) = (27.89 A36o - - 7.39 A 3 1 o ) X dilution factor
(15)
This assay, based on the novobiocin assay, 2 is sensitive, simple, and rapid but not completely specific. If greater specificity is desired, a quantitative paper chromatographic assay may be used. In this assay 10 ml of the n-butyl acetate extract is taken to dryness and the residue is redissolved in 0.5 ml of ethanol. This sample is spotted on Whatman No. 20 paper in a quantity sufficient to deliver approximately 75 ~g of novobiocic acid to the paper. Prior to application of the sample, the paper is dipped in ethylene glycol containing 2% of 85% lactic acid which is the stationary phase. The mobile phase consists of isopropyl ether saturated with ethylene glycol. The chromatogram is developed by the descending method at 2 8 ° for ca. 16 hr and dried. A standard novobiocic acid solution is also spotted. Novobiocic acid is located with reference to the standard by UV light. These are cut from the paper and eluted with 10 ml of acid methanol (0.002 N H oSO,). The absorbance of the eluate is read at 324 nm, and the concentration of novobiocic acid is calculated from its absorptivity (51.9). Qualitative estimation of novobiocic acid formation can be accomplished with thin-layer chromatography (TLC). The samples and standard are prepared as described previously. They are spotted on silica gel G TLC plates and developed in chloroform:methanol (9:1). Detection is by UV light. U n i t s . One unit of activity is defined as the amount of enzyme catalyzing the formation of 1 ~g of novobiocic acid per minute under the assay conditions described. Specific activity is defined as units per milligram of protein measured. 2R. M. Smith, J. J. Perry, G. C. Prescott, J. L. Johnson, and J. H. Ford, Antibiot. Ann. 1957/1958, 43 (1958).
[35]
NOVOBIOCIC ACID SYNTHETASE 0 II O-C-NH2
OH
0
505
CH
Novobiocin
CH3
CH3
(CH3C0)20/plridine I I
CH3COCl / methanol
0 H O-C-NH 2
0--~i-- CH3
~CH3
+
"%o o
H O O C ~ C H 2 C H = C CH3
x:/ "OAc
CH3
CH3
I
0
tl ~ CH3 NH--C~CH2CH=C ~ CH3 HO'~O~O "¢/~OH CH3
HCl, methanol
Novobiocic
I
acid
NoOC2H5
OH
H O ~
OH
/
/CH3
H2
HOOC.~--~CH2CH =C "~OH ~CH5
CH3 B
ring
A ring
FIG. 2. Hydrolysis scheme.
P r e p a r a t i o n o f S u b s t r a t e s a n d Product
Substrates for the biosynthesis of novobiocin are most conveniently obtained by the hydrolysis of novobioein itself. Two hydrolysis routes devised by J. W. Hinman et al. ~ are summarized in Fig. 2. N o v o b i o c i c Acid. One hundred grams of novobioein are dissolved in 1 liter of methanol; 12.5 ml aeetyl chloride are added, and the solution is refluxed for 80 min (the reaction may be followed by TLC on silica gel with 8:2 benzene:methanol, detection by UV). The solution is then cooled to room temperature and poured into 3 liters of cold water. The yellow precipitate is filtered, washed with 0.5 liter of distilled water, and air dried. The air-dried crude novobiocie acid is reerystallized from a minimum of hot aeetone. The yield is about 51 g (79%); mp 2 1 6 218 ° (d).
Thirty-five grams of this material are purified in a 200 X 50 ml -}- 50 ml eountereurrent distribution machine (CCD). It is dissolved in 500 ml of the upper phase of 1 : 2 : 2:3 water : ethanol : ethyl acetate : eyelohexane, filtered from undissolved material, and filled into the first 10 tubes of the CCD machine. The remaining tubes are filled with 50 ml each of lower phase; upper phase is added automatically. After 500 transfers, the lower phases are assayed by solids determination, the appropriate 3ft. W, Hinman, E. L. Caron, and H. I-Ioeksema, d. Amer. Chem. Sac. 79, 3789 (1957).
506
ANTIBIOTIC BIOSYNTHESIS
[35]
tubes ( ~ N o s . 40-120) are combined and concentrated to 250 ml. The resulting precipitate is recrystallized from a minimum of hot acetone. The yield is 12.9 g (37%), mp 229 ° (d). A second crop of 11.2 g can be obtained by concentrating the mother liquor to 500 ml and adding 100 ml of SkellysoIve B. 4-Hydroxy-3(3-methyl-2-butenyl)benzoic Acid (Ring A). One hundred grams of novobiocin are dissolved in 1 liter of freshly distilled pyridine and refluxed with 216 g acetic anhydride for 4 hr. The reaction solution is cooled to 5 °, 1500 ml of water, and 1030 ml of 12 N hydrochloric acid are added while maintaining the temperature below 20 ° . The resulting precipitate is filtered, washed with 500 ml of water, and dried in a vacuum oven at 50 °. The yield is 114 g of crude mixture. 4-Acetoxy-3(3-methyl-2-butenyl)benzoic acid is extracted from the crude precipitate in a Soxhlet extractor with 2 X 300 ml ether for 1 hr each. The ether extracts are evaporated to form the crude acid; this is recrystallized from 300 ml of ethanol and 600 ml of water. The yield is 74 g crude. A second crystallization from 800 ml of 50% aqueous ethanol yields 31 g of pure acid (76%), mp 120 °. This compound is deacetylated by dissolving 25 g in 1 liter of 1 N sodium hydroxide and 2.5 liters ethanol. After 3 hr the solution is adjusted to pH 2 with 6 N hydrochloric acid and concentrated at reduced pressure to an aqueous concentrate. This solution is extracted with 3 X 1 liter ether; the extracts are combined, dried with sodium sulfate, and evaporated. The residue (20.5 g) is crystallized from 200 ml of hot benzene. The yield is 20 g (80%) ; mp, 104-106 °. The overall yield is 61%. 3-Amino-4,7-dihydroxy-8-methyl Coumarin (Ring B). One hundred grams of novobiocin are hydrolyzed as described under 4-hydroxy-3(3methyl-2-butenyl)benzoic acid. After extracting this compound from the described crude mixture, the remainder (51 g) is recrystallized from 750 ml of ethanol and 200 ml of water to yield 46.5 g (65%) of the sugarcoumarin moiety (rings B-C) ; mp 167-172 °. This compound is hydrolyzed by suspending 20 g in 1 liter of methanol and refluxing with 250 ml 4 N methanolic HC14 (the finely powdered crystals dissolve before the boiling point is reached). After 2-hr refluxing the solution is cooled, concentrated at 35°/vacuum to about 200 ml, and cooled to 5 °. The resulting crystals (8.1 g) are filtered; a second crop (1.3 g) is obtained by concentrating the mother liquor to 60 ml and adding 120 ml of ether. Both crops are combined and recrystallized from 200 ml of hot ethanol and 130 ml of water. The yield is 9.0-9.1 g (90-91%) of the hydrochloride ; mp 90-93 °. 4Dry HCI gas is bubbled into 1 liter of cooled absolute methanol until the weight has increased from 791 g to 1064 g. The acid is titrated with 1 N NaOH; an appropriate amount of absolute methanol is added if the normality is >4.
[35]
NOVOBIOCIC ACID SYNTHETASE
507
The free base may be obtained by recrystallizing 1 g of the hydrochloride from 25 ml 90% aqueous ethanol, yielding about 0.8 g.
Preparation of the Enzyme Cultivation o] the Microorganism. Streptomyces niveus strain BC-345 is maintained on agar slants consisting of glucose, 10 g; brewer's yeast, 10 g; distillers solubles, 5 g; KC1, 4 g; C a CQ , 1 g; agar (Difeo), 20 g; and tap water to 1000 ml. Spores from these slants were suspended in sterile distilled water (10 ml) and used to inoculate a seed medium of the following composition: glucose, 10 g; beef extract, 3 g; NZ Amine B, l0 g; and distilled water to 1000 ml. The inoculated medium was incubated at 28 ° on a reciprocating shaker for 72 hr. The fermentation medium consists of glucose, 30 g; sodium citrate, 6 g; L-proline, 6 g; K2HPO~, 2 g; (NH~)~SO~, 1.5 g; NaC1, 5 g; MgSO4, 1 g; CaCl~, 0.4 g; FeSO4.TH20, 0.2 g; ZnSO4"TH20, 0.1 g; and distilled water to 1000 ml. The pH of this medium is adjusted to 7.2. The glucose is sterilized separately and added aseptically prior to inoculation. The inoculum for the fermentation medium is blended for 3 rain in a sterile, chilled Waring Blendor prior to addition. Blending is essential for rapid growth and enzyme production. The fermentation medium is inoculated with 5% of the blended seed culture and incubated at 28 ° on a rotary shaker. Cell-Free Extract. The cells are harvested from the fermentation medium after 4 days of incubation by centrifugation. The mycelia are washed twice in 50 mM Tris.HC1 buffer, pH 8.0. The mycelia are resuspended in this same buffer to about one-third of the fermentation harvest volume, and broken by sonication with a Bronson sonifier (S-125) run at maximal power output for 5 rain. Cell debris is removed by centrifugation at ~0,000 g for 10 min. All manipulations are carried out at 4 °. Protein concentrations are determined by the method of Lowry et al. 5
Properties Substrate Specificity. In addition to the amide bond formation between the A and B rings the reaction also occurs between the A ring and 3 amino-4-hydroxycoumarin, and between the dihydro A ring [4-hydroxy-3(3 methylbutyl)benzoic acid] and B ring. No reaction occurs between 4-hydroxy benzoate and the B ring or tyrosine and the A ring. 0. H. Lowry, N. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem. 193, 265 (1951).
508
[36]
ANTIBIOTIC BIOSYNTHESIS
pH Optimum. In Tris.HC1 buffer the optimum pH of the reaction is 9.0 with a rapid decline in activity above this pH. For this reason it is best to run the reaction at pH 8.0 when using crude cell-free extracts. Stability. The stability of the enzyme in crude extract is poor. Storage at 4 ° results in about 50% loss of activity in 24 hr. Freezing or lyophilization results in a complete loss of activity. Relationship to Novobiocin Biosynthesis. The formation of novobiocic acid may not be the initial reaction in the coupling of the ring systems of novobiocin. Other possible coupling mechanisms include the formation of a glycosidic bond between the sugar (C ring) and coumarin (B ring) moieties prior to amide bond linkage with the A ring or a suitable intermediate of ring A.
[36] S - A d e n o s y l m e t h i o n i n e : O - D e m e t h y l p u r o m y c i n O-Methyltransferase By BURTON M. POGELL The antibiotic and antitumor agent, puromycin, is produced by Streptomyces alboniger and is a well characterized specific inhibitor of protein synthesis which terminates peptide bond elongation by competing with AA-tRNA. 1 The presumed last step in its biosynthesis 2 is catalyzed by the following enzymic reaction:
HsC-.N/CHs
HsC\N/CHs
H
O
~
NH
~
OH
I H, / = ~ O=C--CH--C~ - ~ ~ O H NH2 O-Demethylpuromycin
NH
OH
O : C - - CH-- C"~ NH,
~
O--CHs
Puromycin
+ S - Adenosyl- L- methionine
S-Adenosyl-L-homocysteine
: S. Pestka, Annu. Rev. Microbiol. 25, 487 (1971). 2M. M. Rao, P. F. Rebello, and B. M. Pogell, J. Biol. Chem. 244, 112 (1969).
508
[36]
ANTIBIOTIC BIOSYNTHESIS
pH Optimum. In Tris.HC1 buffer the optimum pH of the reaction is 9.0 with a rapid decline in activity above this pH. For this reason it is best to run the reaction at pH 8.0 when using crude cell-free extracts. Stability. The stability of the enzyme in crude extract is poor. Storage at 4 ° results in about 50% loss of activity in 24 hr. Freezing or lyophilization results in a complete loss of activity. Relationship to Novobiocin Biosynthesis. The formation of novobiocic acid may not be the initial reaction in the coupling of the ring systems of novobiocin. Other possible coupling mechanisms include the formation of a glycosidic bond between the sugar (C ring) and coumarin (B ring) moieties prior to amide bond linkage with the A ring or a suitable intermediate of ring A.
[36] S - A d e n o s y l m e t h i o n i n e : O - D e m e t h y l p u r o m y c i n O-Methyltransferase By BURTON M. POGELL The antibiotic and antitumor agent, puromycin, is produced by Streptomyces alboniger and is a well characterized specific inhibitor of protein synthesis which terminates peptide bond elongation by competing with AA-tRNA. 1 The presumed last step in its biosynthesis 2 is catalyzed by the following enzymic reaction:
HsC-.N/CHs
HsC\N/CHs
H
O
~
NH
~
OH
I H, / = ~ O=C--CH--C~ - ~ ~ O H NH2 O-Demethylpuromycin
NH
OH
O : C - - CH-- C"~ NH,
~
O--CHs
Puromycin
+ S - Adenosyl- L- methionine
S-Adenosyl-L-homocysteine
: S. Pestka, Annu. Rev. Microbiol. 25, 487 (1971). 2M. M. Rao, P. F. Rebello, and B. M. Pogell, J. Biol. Chem. 244, 112 (1969).
[36]
PUROMYCIN
509
A s s a y M e t h o d s 2,3
The two procedures described give comparable results and are based on determination of radioactivity incorporated into puromycin after separation from radioactive S-adenosyl-L-methionine (SAM). Method A is a more reproducible, simple one-step assay in which radioactive product is directly extracted into scintillation fluid. Preparation o] O-Demethylpuromycin. 2 The aminonucleoside of puromycin is coupled with O,N-dicarbobenzoxy-L-tyrosine in the presence of Woodward's reagent K, followed by hydrogenolysis of the carbobenzoxy group. Add 0.2 ml of triethylamine (0.34 mmole) in dimethylformamide over a l-rain period to a magnetically stirred suspension of 0.153 g of O,Ndicarbobenzoxy-L-tyrosine (0.34 mmole) and 0.086 g of Woodward's reagent K (0.34 mmole) in 1.6 ml of dimethylformamide. Continue stirring until the Woodward's reagent is completely dissolved (25 min). Then add 3 ml of a solution of 0.10 g of puromycin aminonucleoside (0.34 mmole) and 0.025 mmole of triethylamine in dimethylformamide and stir for an additional 18 hr. Remove the solvent in a vacumn, and triturate the sticky residue with 15 ml of water. Collect the granular precipitate (formed after standing for 3 hr at 4 °) by filtration, wash several times with water, and dry in a vacuum overnight (yield, 0.213 g, 86%). Heat this product to boiling in 5 ml of absolute ethanol, cool, and collect the residue by filtration (rap 184-187°). Dissolve the above residue (0.173 g) in 5 ml of Methyl Cellosolve at 60-70 °, add 0.4 g of 10% palladium charcoal, and place the suspension in a water bath kept at 60-70 °. Pass a slow stream of hydrogen through the solution until the evolution of carbon dioxide, as tested with barium hydroxide solution on the exit gases, is complete (1.5 hr). Remove the catalyst by filtration, the solvent in a vacuum, and dry the glassy residue overnight in a vacuum (yield, 0.105 g). The crude O-demethylpuromyein is further purified by column chromatography on silica gel and paper electrophoresis. Place the sample in 0.1 ml of methanol on a column of silica gel (1 X 10 cm) in chloroform and elute with 3 % methanol in chloroform. The bulk of the product, as measured by absorption at 275 nm, appears between 130 and 410 ml of effluent. Concentrate this material to dryness in a vacuum and repeat the chromatography. Removal of final traces of yellow impurity is achieved by preparative paper electrophoresis in pyridine-aeetic acidwater (2: 1:20) (pH 5.2) and the sample is eluted with ethanol. Absorption maxima and approximate molar extinction coefficients: X~t~~°1 275 nm 3 L. Sankaran and B. M. Pogell, Anal. Biochern. 54, 146 (1973).
510
ANTIBIOTIC BIOSYNTHESIS
(e, 15,000); ~0mla~HCI 245 nm (5, 9,700).
268 nm
[~5]
(5, 14,400); X°m~NNa°~ 277 nm (5, 14,300),
Reagents. All solutions for the enzyme assay are prepared in 0.1 M sodium phosphate-1 mM EDTA (pH 7.5) O-Demethylpuromycin, 4 mM S-Adenosyl-L-methionine (SAM), 0.32 mM (2.5 X l0 s cpm/0.1 ml), either [3H]methyl or [14C]methyl Sodium borate, 0.1 M-NaCI, 5 M, final pH 9.0 Enzyme. Dilute with buffer if necessary.
Procedure A2 Incubate 5 tLl of O-demethylpuromycin, 5 ~l of SAM, and enzyme in a final volume of 20 ~l at 38 ° for suitable time intervals in 1-dram vials with snap-on plastic caps (Rochester Scientific, Cat. No. 7475, 15 X 45 mm). Add all reagents at the same place in the angle between the bottom and side of the vials, and mix the contents by gentle rotation. Stop the reactions by adding 0.7 ml of 0.1 M sodium borate-5 M NaC1 (pH 9) and 3 ml of scintillation fluid (Spectrafiuor, NuclearChicago, 1:25 dilution in toluene) to each vial. Stopper and shake the vials vigorously in a horizontal position for 10 min in a mechanical shaker (Precision Scientific, Cat. No. 65855), centrifuge to clarify the layers, and place in regular-sized glass scintillation vials for radioactive counting. Always include a control incubation minus O-demethylpuromycin, particularly with crude preparations. With 14C-SAM as substrate, transfer 2 ml of the scintillation fluid to a new vial for counting after extraction. This results in much lower background values. Definition o] Unit and Specific Activity. A unit of enzyme is defined as the amount ~hat will form 1 nmole of puromycin per minute at pH 7.5 and 38 ° . Specific activity is expressed as units of enzyme per milligram of soluble protein. Protein is determined by the Lowry procedure 4 with bovine serum albumin as standard. Units of enzyme are calculated as follows: (a) ~H: cpm of sample 0.875 × 1.6 × cpm/nmole of 3H-SAS/I X min of incubation or
(b) 14C: 1.5 X cpm of sample 0.875 X 1.13 × cpm/nmole of [14C]SA5/I × min of incubation 4 E. Layne, this series, Vol. 3 [73].
[36]
PUROMYCIN
511
where 0.875 = fraction of puromycin extracted by scintillation fluid under above conditions; specific radioactivity of SAM is determined by counting in 95% ethanol:scintillation fluid (3:10); 1.6 = ratio of ~H sample counted in scintillation fluid compared to sample counted in ethanol:scintillation fluid; 1.13 = ratio of 1~C sample counted in scintillation fluid compared to sample counted in ethanol:scintillation fluid; 1.5 - correction for 14C because only two-thirds of the scintillation fluid is counted. Procedure B f- Incubate mixtures as above in 10 X 75 mm test tubes. Terminate reactions by adding 20 ul of ethanol-acetic acid (9:1) anti remove protein by centrifugation. Carefully remove one-half of each supernatant with a micropipette and spot on a precoated thin-layer cellulose plastic sheet (Brinkmann). Develop the sheets in saturated ammonium sulfate-1 M sodium acetate-isopropyl alcohol (80:18:2) with puromycin as marker. After drying, repeat the development once or twice. This procedure separates the unreacted SAM and its decomposition products from puromycin, which remains at the origin. Cut out the areas corresponding to puromycin (approximately 1.5-cm squares), place each in a scintillation vial, add 3 ml of ethanol and 10 ml of scintillation fluid, and determine the radioactivity in each sample. Enzyme units are calculated from the specific radioactivity of the added SAM assuming that radioactivity incorporated is directly proportional to product formed.
E n z y m e Purification ~,5,6
Our original studies on this enzyme were carried out in extracts from a mutant strain of S. alboniger, ATCC 12462. Higher specific activities can be obtained with the wild-type organism, ATCC 12461, by harvesting cells during mid- or late logarithmic growth (48-72 hr) and shortening the time of sonication. Inoculate cells of S. alboniger strain ATCC 12461 into 500 ml of 6% corn steep liquor-2% corn starch, pH 6.4, in 2-liter shake flasks with added Dow-Corning Antifoam. Grow at 28 ° with constant agitation on a New Brunswick gyratory shaker. Harvest cells by centrifuging at 4 °, wash twice with cold water, and store frozen at --20 ° . All purification procedures are carried out at 0-4 °. Results of a typical purification procedure with pooled extracts of specific activity >0.5 are shown in the table. 5L. Sankaran and B. M. Pogell, Nature (London) New Biol. 245, 257 (1973). "L. Sankaran, S. V. K. Narasimha Murthy, M. Kariya, and B. M. Pogell, unpublished observations.
512
ANTIBIOTIC BIOSYNTHESIS
[35]
PURIFICATION OF O-I~ETHYLTRANSFERASE
Step 1. Crude extract Protamine sulfate s u p e r n a t a n t 2. A m m o n i u m sulfate precipitate (4,5-70%) 3. DEAE-cellulose column I Fractions 66-84 Amicon 12 concentrate DEAE-cellulose column I I Amicon 12 concentrate
4. Sephadex G-200 Amicon 12 concentrate
Protein (mg)
ReSpecific covery activity (%) (units/mg)
Volume (ml)
Units
15.9 17.1 1.3
43.0 40.6 25.2
76.3 67.5 23.6
100 94 59
38 1.04
17.5 12.9
---
41 30
---
0.82
7.5
1.53
17
4.9
1.4
7.1
0.43
17
16.5
0.56 0.60 1.1
Step 1. Preparation of Cell-Free Extract and Removal of Nucleic Acids. Prepare a cell-free extract by sonic disintegration of a 10% (w/v) suspension of cells in 0.1 M sodium phosphate buffer, pH 7.5, for 3-4 min in an ultrasonic disintegrator (Measuring and Scientific Equipment, Ltd., London, 60 W). Cool the cell suspension by circulating ice water during the sonic disruption. Remove particulate matter by centrifuging for 30 min at 32,000 g. Remove nucleic acids from the above supernatant by adding a 2% solution of protamine sulfate (pH 7) (ca. 0.11 ml per milliliter of enzyme solution). Collect the supernatant by centrifugation. Step 2. Salt Fractionation. Gradually add solid ammonium sulfate to 45% saturation to the protamine supernatant (0.277 g/ml). Remove the precipitate by centrifugation and raise the concentration of ammonium sulfate to 70% saturation by further addition of salt (0.171 g/ml). Collect the precipitate by centrifugation. Upon further purification, the methylase is found to be unstable in dilute solution. Inclusion of 10 ~M SAM in all solutions used for column chromatography prevents this inactivation (see section on Properties). Step 3. DEAE-Cellulose Column Chromatography. Dissolve the ammonium sulfate precipitate in 1.3 ml of 5 mM sodium phosphate-0.1 mM EDTA-10 ~ / / SAM, pH 7.5 (PES), and place on a column of DEAEcellulose (Whatman, DE-52, 5 g, 0.9 X 13 cm) preequilibrated with the same buffer. Wash with PES and elute with a concave salt gradient at a flow rate of 16 ml/hr. A Technicon l-liter capacity Autograd is used to form the gradient. The first 2 chambers contain 70 ml each of 0.1 M NaC1 in PES and the third chamber contains 70 ml of 0.6 M NaC1
[36]
PUROMYCIN
A z ~J
513
z bJ CORN STEEP LIQUOR-CORN STARCH
HICKEY-TRESNER MEDIUM
0 o.
4.0
o
0.I0
f
¢n
GROWTH
FZ
>I-> I--
3.0 0
•
~'~ SPECIFIC ACTIVITY
0.4 E
-g
:~ SPECIFIC
~,
2.0 ~
~_ 0 . 5
_z
0.05
w 0.2
~
0 LO 0,.
._1 I.w a.
I00
200
0
0
_L
L 200
0
TIME IN HOURS
TIME IN HOURS 0
A tO0
0
FIG. 1. Changes in O-methyltransferase activity during growth in (a) HickeyTresner and (b) corn steep liquor-corn starch media. Streptomyces alboniger cells (ATCC 12461) were grown at 28° and cell-free extracts prepared as described in the text. Total protein was estimated in samples of whole-cell sonicates as a measure of growth. Enzyme specific activity was determined in supernatants after centrifugation at 32,000 g for 30 rain. in PES. Collect 2-ml fractions and assay for enzyme activity. The methylase is eluted as one symmetrical p e a k in fractions 66-84. Pool this m a t e rial and concentrate in an Amicon 12 ultrafiltration unit with a P M 30 membrane. Wash and reconcentrate twice with 9-ml volumes of PES. R e p e a t the above DEAE-cellulose c h r o m a t o g r a p h y procedure with concentrated enzyme. Pool fractions 58-76 and reconcentrate as above. Step 4. Sephadex G-200 Gel Filtration. Place the concentrated enzyme on a Sephadex G-200 column (2.5 X 40 cm) equilibrated with P E S and elute at a flow rate of 15 m l / h r with PES. Collect 2-ml fractions and assay for enzyme activity. T h e methylase is eluted as one peak in fractions 46-55. Pool this material, concentrate by ultrafiltration as above, and store at --65 ° . In another preparation, where Sephadex gel filtration was run before DEAE-cellulose chromatography, the final specific activity was 26. Properties o f the E n z y m e 2,5,
The enzyme activity increases with growth, reaches a m a x i m u m during the mid- or late logarithmic growth phase, and then declines rapidly in the stationary phase (Fig. 1). I n a medium designed by H i c k e y and Tresner 7 to enhance sporulation in Streptomyces, the change in enzyme R. J. Hickey and H. D. Tresner, J. Bacleriol. 64, 891 (1952).
514
ANTIBIOTIC BIOSYNTHESIS
[35]
specific activity follows a bell-shaped curve with time, again becoming zero after a high degree of sporulation occurs; also no enzyme activity has been found in spore extracts. Much higher specific activities are obtained in corn steep liquor-corn starch medium. Both glucose (1%) and subinhibitory levels of ethidium bromide (5 ~M) or acriflavine (3 ~g/ml) differentially inhibit enzyme formation; glucose similarly inhibits antibiotic formation on agar. Possibly related to the rapid turnover of the methylase activity is a recent observation that the enzyme is very unstable in extracts at 38 ° and pH 7.5. Addition of assay levels of either SAM or O-demethylpuromycin protects enzyme activity under these conditions. The enzymic methylation is very specific for SAM and O-demethylpuromycin. 5-Methyltetrahydrofolate is completely inactive as methyl donor and no methylation of tyrosine, several tyrosine derivatives, catechol, or L-epinephrine is detectable. The Km for SAM is 0.01 mM and for O-demethylpuromycin, 0.2 mM. A broad pH optimum is found over the range of pH 7-9. Formation of product is linear with respect to time and enzyme concentration.
Other Aspects of Puromycin Biosynthesis Dialyzed supernatants from extracts of S. alboniger catalyze the formation of two aminopentose phosphates, 2-amino-2-deoxy-D-ribose 5-phosphate and 2-amino-2-deoxy-D-lyxose 5-phosphate, from ribose 5phosphate and NH4÷.s,9 However, no evidence has yet been found for the formation of any 3-aminopentose in cell-free extracts. No other enzymic steps in the biosynthesis of puromycin have been characterized. Three possible precursors of the antibiotic have been isolated in small quantities from commercial preparations of puromyein. 1° Advantage was taken of the solubility of puromycin in chloroform at pH 11-12, conditions where O-demethylpuromycin and other derivatives with a free tyrosyl hydroxyl group remain in the aqueous layer. The compounds and structures correspond to the respective demethyl derivatives of puromycin: N6,N6,0-tridemethylpuromycin, N6,0-didemethylpuromycin, and O-demethylpuromycin. Radioactive precursor studies have shown that the methyl group of methionine is incorporated into the 8B. M. Pogell, P. F. Rebello, and P. P. Mukherjee, in "International Symposium on Metabolism, Physiology, and Clinical Use of Pentoses and Pentitols" (B. L. ttorecker, K. Lang, and Y. Takagi, eds.), p. 135. Springer-Verlag, Berlin, 1969. P. F. Rebello, B. M. Pogell, and P. P. Mukherjee, Biochim. Biophys. Acta 177, 468 (1969). 2oT. N. Pattabiraman and B. M. Pogell, Biochim. Biophys. Acta 182, 245 (1969).
[37]
GUANOSINETRIPHOSPHATE-8-FORMYLHYDROLASE
515
N~,N6-dimethyladenine moiety as well as the O-methyl group of puromycin. 6,1~ Adenine is also incorporated into N'%N6-dimcthyladenine during growth. 11 Thus, it appears that these compounds, or derivatives such as 5'-phosphate esters, are probable precursors of puromycin. H a r d l y anything is known about why antibiotics are formed, or what, if any, are the normal functions of these secondary metabolites in normal cell metabolism. Several compounds containing NG,N6-dimethyladenine in glycosidic linkage have been isolated from spore-rich preparations of S. alboniger (2-3 nmoles of base per milligram dry weight)? 1 A major portion is linked to part of the spore coat. In addition, free N~;,N%dimethyladenine is excreted into the growth medium. The absence of methylated adenine in hyphae suggests a role for N-methyla~ion during differentiation in S. alboniger. No O-methyl-L-tyrosinc was found in acidhydrolyzed hyphae or spores.
~ M. G. Sarngadharan, B. M. Pogell, and M. Kariya, Biochim. Biophys. Acta 262, 405 (1972).
[37] G u a n o s i n e
Triphosphate-8-formylhydrolase
B y E. F. ELSTNER and R. J. SUHADOLNIK
GTP ~ 4-N[(5'-triphospho)- l'-ribosylamino]-2, ~diamino-6-hydroxypyrimidine + formic acid
Assay Method Principle. I t is now well established t h a t carbon-8 of guanine, guanosine, or guanosine triphosphate (GTP) is lost as formic acid in the biosynthesis of riboflavin, the pteridines, a pteridine cofactor for phenylalanine hydroxylation, the azapteridine ring, and the benzimidazole ring. ~-7 The enzyme, GTP-8-formylhydrolase, that catalyzes ~he conversion of G T P to formic acid and 4-N-[ (5'-triphospho)-l'-ribosylamino]2,5-diamino-6-hydroxypyrimidine by S. rimosus has been described? This enzyme appears to be involved in the biosynthesis of the pyrrolopyrimi-
1E. F. Elstner and R. J. Suhadolnik, J. Biol. Chem. 246, 6973 (1971). A. W. Burg and G. M. Brown, J. Biol. Chem. 243, 2349 (1968). s B. Levenberg and D. K. Kaczmarek, Biochim. Biophys. Acta 117, 272 (1966). J. Cone and G. Guroff, J. Biol. Chem. 246, 979 (1971). 5W. L. Alworth, S. H. Lu, and M. F. Winkler, Biochemistry 10, 1421 (1971). 6 R. A. Harvey and G. W. E. Plaut, J. Biol. Chem. 241, 2120 (1966). 7A. Bacher and F. Lingens, J. Biol. Chem. 245, 4647 (1970).
[37]
GUANOSINETRIPHOSPHATE-8-FORMYLHYDROLASE
515
N~,N6-dimethyladenine moiety as well as the O-methyl group of puromycin. 6,1~ Adenine is also incorporated into N'%N6-dimcthyladenine during growth. 11 Thus, it appears that these compounds, or derivatives such as 5'-phosphate esters, are probable precursors of puromycin. H a r d l y anything is known about why antibiotics are formed, or what, if any, are the normal functions of these secondary metabolites in normal cell metabolism. Several compounds containing NG,N6-dimethyladenine in glycosidic linkage have been isolated from spore-rich preparations of S. alboniger (2-3 nmoles of base per milligram dry weight)? 1 A major portion is linked to part of the spore coat. In addition, free N~;,N%dimethyladenine is excreted into the growth medium. The absence of methylated adenine in hyphae suggests a role for N-methyla~ion during differentiation in S. alboniger. No O-methyl-L-tyrosinc was found in acidhydrolyzed hyphae or spores.
~ M. G. Sarngadharan, B. M. Pogell, and M. Kariya, Biochim. Biophys. Acta 262, 405 (1972).
[37] G u a n o s i n e
Triphosphate-8-formylhydrolase
B y E. F. ELSTNER and R. J. SUHADOLNIK
GTP ~ 4-N[(5'-triphospho)- l'-ribosylamino]-2, ~diamino-6-hydroxypyrimidine + formic acid
Assay Method Principle. I t is now well established t h a t carbon-8 of guanine, guanosine, or guanosine triphosphate (GTP) is lost as formic acid in the biosynthesis of riboflavin, the pteridines, a pteridine cofactor for phenylalanine hydroxylation, the azapteridine ring, and the benzimidazole ring. ~-7 The enzyme, GTP-8-formylhydrolase, that catalyzes ~he conversion of G T P to formic acid and 4-N-[ (5'-triphospho)-l'-ribosylamino]2,5-diamino-6-hydroxypyrimidine by S. rimosus has been described? This enzyme appears to be involved in the biosynthesis of the pyrrolopyrimi-
1E. F. Elstner and R. J. Suhadolnik, J. Biol. Chem. 246, 6973 (1971). A. W. Burg and G. M. Brown, J. Biol. Chem. 243, 2349 (1968). s B. Levenberg and D. K. Kaczmarek, Biochim. Biophys. Acta 117, 272 (1966). J. Cone and G. Guroff, J. Biol. Chem. 246, 979 (1971). 5W. L. Alworth, S. H. Lu, and M. F. Winkler, Biochemistry 10, 1421 (1971). 6 R. A. Harvey and G. W. E. Plaut, J. Biol. Chem. 241, 2120 (1966). 7A. Bacher and F. Lingens, J. Biol. Chem. 245, 4647 (1970).
516
ANTIBIOTIC BIOSYNTHESIS
[37]
dine nucleoside antibiotics toyocamycin, tubercidin, and sangivamyein. The assay method is based on two procedures. The first method involves the release of 14C-labeled formic acid from carbon-8 of the imidazole ring of [8-~4C]GTP. The formic acid is oxidized to carbon dioxide2 The second method involves the addition of the enzyme reaction mixture, following a 2-hr incubation, to a 1-ml column (Norit A-Celite, 1:1 (w/w). The column is washed with 5 ml of 1 N formic acid to elute the [~4C]HCOOH. The recovery by this method is 75% compared with the first procedure.
Reagents [8-14C]GTP, 12.5 nmoles, 80,000 cpm Tris buffer, 25 mM, pH 8.0 Enzyme, 0.4 unit/ml
Procedure. Mix the above three reagents and incubate for 2 hr at 38°; total volume, 1 ml. The formic acid released is determined by one of the two methods described above. If the oxidation of formic acid is used, the 1~C0~ is removed from the reaction mixture by bubbling with nitrogen gas for 10 min at 100 ° into 1 ml of 1 N Hyamine (NCS solubilizer) ; the 14CO~ is determined by liquid scintillation counting of 0.5 ml of the Hyamine solution using Bray's scintillation fluid. 8 Definition o] Unit and Specific Activity. One unit of enzyme is that amount of protein which causes the formation of 1 nmole of formic acid per hour at 38 ° . Specific activity is expressed as units per milligram of protein. Protein is determined by the method of Murphy and Kies." Purification Procedure
Step 1. Preparation o] Cell-Free Extracts o] S. rimosus. The cells are harvested 46-48 hr after inoculation and centrifuged at 5000 g for 5 min, washed with 50 mM Tris (pH 8.0), centrifuged, and lyophilized. The lyophilized cells are stored at --20 °. The lyophilized cells (5 g, dry weight) are suspended in cold Tris buffer (50 mM, pH 8.0) and treated with a French pressure cell (Aminco, pressure cell No. 4-3396) at 16,000 psi. The slurry is centrifuged at 48,000 g for 30 min. The supernatant is referred to as crude extract. The crude extract is dialyzed overnight against 5 mM Tris buffer, pH 8.0. Ammonium sulfate precipitations are performed with the highest purity ammonium sulfate. All fractionations s G. A. Bray, Anal. Biochem. 1, 279 (1960). PJ. B. Murphy and M. W. Kies, Biochim Biophys. Acta 45, 382 (1960).
[37]
GUANOSINETRIPHOSPHATE-8-FORMYLHYDROLASE
517
are performed at 4 °. GTP-8-formylhydrolase activity is maximal 48 hr after inoculation. Step 2. Purification on DE-52 Cellulose. The dialyzed crude extract (70 ml, no EDTA), 1200 mg of protein, is applied on a column of DE-52 cellulose (2 X 30 cm) after equilibration with 10 mM phosphate buffer, pH 7.4. The linear gradient is from 10 mM to 0.3 M phosphate buffer (pH 7.4), 200 ml each. Fractions of 10 ml are collected at 4 ° and a flow rate of 0.5 ml per minute is maintained. Step 3. Ammonium Sul]ate Fractionation. Solid ammonium sulfate is added with stirring to 30% saturation. After 15 min, the solution is centrifuged at 15,000 g, 10 min. Solid ammonium sulfate is added to the supernatant to bring the solution to 60% saturation, stirred 15 additional min, and centrifuged at 15,000 g, 10 min. The pellet is dissolved in 10 ml Tris buffer (50 mM, pH 8.0) and dialyzed against 2 liters Tris buffer (50 mM, pH 8.0) overnight. Assays of the crude undialyzed extract are not reliable because there are interfering compounds that inhibit the enzyme. These inhibitors are removed by dialysis. Step 4. Purification on Sephadex G-200. Two milliliters (26 mg prorein) of the GTP-8-formylhydrolase are added to a Sephadex column (1.5 cm X 100 era). The column is previously equilibrated with 10 mM phosphate buffer (pH 7.4; EDTA, 1 mM). The protein is eluted with the same phosphate-EDTA solution. Two-milliliter fractions are collected at a flow rate of 0.5 ml/min; the peak tube for elution of the enzyme is usually tube 30. A purification is summarized in the table.
Properties Specificity. The purified enzyme will only eliminate the ureido carbon of GTP as [14C]formic acid. The following [8-1~C]purine ribonucleoside or nucleotides are not substrates: guanosine, GMP, GDP, ITP, and ATP. The addition of guanosine, GMP, or GDP (50 fM) to assays with the purified enzyme does not inhibit the release of [l~C]formic acid from [8-'~C] GTP. Activators and Inhibitors. No cofactors are required for enzymic activity. GTP-8-formylhydrolase is a sulfhydryl enzyme. Sulfhydryl reducing agents such as mercaptoethanol (10 raM) and cysteine (1 mM) stimulate enzyme activity (110 and 140%, respectively). With 0.1 M mercaptoethanol, the GTP-8-formylhydrolase activity decreases to 15%. Ascorbate gives a slight stimulation (120%) and EDTA (1 raM) completely inhibits the enzyme. Sulfhydryl inhibitors such as showdomycin (1 mM) and p-chloromerc-
518
ANTIBIOTIC BIOSYNTHESIS
O
~q
©
©
(9 ©
Z ©
O
0
¢0
[37]
[371
GUANOSINE TRIPHOSPHATE~8-FORMYLHYDROL ASE
519
uribenzoate (0.5 raM) cause a 79% and 95% inhibition of the enzyme; Mg 2÷ (10 mM) and Fe '-'÷ (1 raM) inhibit the enzyme 87% and 41% ; ATP (5 raM) and I T P (5 mM) inhibit the enzyme 60 and 83%. ITP is a mixed type inhibitor (Ki 45 ~M). ATP is a competitive inhibitor (K~ 0.11 raM). The 65% inhibition of the enzyme by showdomycin (1 raM) is reversed by cysteine (3 mM). When incubations are done under anaerobic conditions, about twice as much formic acid is released as under aerobic conditions. Anaerobic conditions are performed in Warburg flasks by flushing for 3 rain with nitrogen. The center well contains 1 g of pyrogallol in 0.5 ml of 1 N NaOH. The reaction vessel contains 1 mM EDTA, 0.025 M Tris buffer, pH 8.0, 50 ~M [8-14C]GTP (90,000 cpm) and 50 mM of glucose. The reaction is started by adding glucose oxidase anti GTP-8-formylhydrolase from the two side arms. A control vessel is run in which pyrogallol, glucose oxidase, and nitrogen are omitted. The reaction is stopped by cooling to 0 °, transferring the reaction contents quantitatively to the microdistillation apparatus, and oxidizing the [14C] formic acid to 1'C0~. The enzyme exhibits maximal activity at pH 8.0, 50 mM Tris buffer. Michaelis Constant and Enzyme Properties. Tile Miehaelis constant for GTP is 20 uM. This was determined by plotting the reciprocal of the velocity of the reaction against the reciprocal of the concentration of GTP (Lineweaver-Burk method). Product formation is linear for 3 hr (125 mnoles of GTP). The rate of reaction is directly proportional to protein concentration up to 10 mg of protein. Molecular Weight. The molecular weight of GTP-8-formylhydrolase has been determined by comparing the elution of the enzyme to that of urease. The molecular weight of the enzyme eluted from a Sephadex G-200 cohlmn is 500,000. Dissociation of GTP-8-formylhydrolase. GTP-8-formylhydrolase c.m be dissociated into two proteins. To do so, it is necessary to dialyze the ammonimn sulfate fraction (without EDTA) prior to chromatograptly on DE-52. Dialysis against EDTA prevents dissociation of the enzyme. The dialyzed crude extract, containing the GTP-8-formylhydrolase, is precipitated between 30 and 60% saturation with ammoniuIn sulfate. The precipitate (380 rag) is dissolved in 2 ml of 50 mM Tris buffer, pH 8.0, and dialyzed overnight against 5 mM Tris buffer, pH 8.0 (no EDTA). The protein is applied to a DE-52 column (1 em X 15 cm) after equilibration with 10 mM phosphate buffer, pH 7.4. The linear gradient is from 10 mM to 0.3 M phosphate buffer (pH 7.4), 150 ml each. Fractions of 8 ml are collected at 4 ° at a flow rate of 0.5 ml per minute. Each tube is assayed for GTP-8-formylhydrolase, phosphatase activity, and protein concentration. Two protein peaks are observed that have GTP-8-formyl-
520
ANTIBIOTIC BIOSYNTHESIS
[38]
hydrolase activity (peak enzyme tubes: 4 and 20). To prove that GTP is the substrate, phosphatase activity is measured in each fraction. Eighty percent of the phosphatase activity is lost in the 30-60% ammonium sulfate fraction. Phosphatase activity is assayed by incubating the enzyme with 25 nmoles [fl-~-32p]GTP (1400 cpm). The reaction mixture is added to 1 ml of a Norit A-celite column (1:1, w/w). The columns are washed with 3 volumes of 0.4 M phosphate buffer, pH 7.4. The inorganic 32p is counted. [fl-y-32p]GTP is prepared from GMP and KH232p04 by the method of Tochikura et al. ~° [fl-y-82P]GTP is purified by DE-52-cellular chromatography (30 g, gradient elution, 500 ml each of water and 0.4 M triethylammonium bicarbonate, pH 7.3). The fl_~/.82p labeled GTP shows a single UV spot on Avicel thin-layer chromatography plates, solvent: 6% ammonium chloride. P r o d u c t F o r m a t i o n . The predicted product following ring opening by GTP-8-formylhydrolase should be 4-N-[(5'-triphospho)-l'-ribosylamino]-2,5-diamino-6-hydroxypteridine. Only 0.16% of this ring-opened product exists following elimination of carbon-8 of GTP. The major product is neopterin. Since riboflavin and folic acid both use the purine ring for their biosynthesis, attempts to isolate these two compounds from the culture medium were made. These compounds are not present in the culture medium. lo T. Tochikura, K. Kawaguchi, T. Kano, and K. Ogata, J. Ferment. Technol. 47, 564 (1969).
[38] 6-Methylsalicylic Acid Synthetase B y G. VOGEL and F. LYNEI~
Under certain culture conditions several P e n i c i l l i u m species produce 6-methylsalicylic acid, the parent compound in the reaction sequence leading to the antibiotic patulin. 1,2 6-Methylsalicylic acid belongs to the class of naturally occurring compounds derived from head-to-tail condensation of acetate units, as postulated in the polyacetate or polyketide hypothesis2 ,4 Malonyl-CoA has been recognized as the building block 1j. D. Bu'Lock and A. J. Ryan, Proc. Chem. Soc. 222 (1958). 2j. D. Bu'Lock, D. Hamilton, M. S. Hulme, A. J. Powell, D. Shepherd, and G. N. Smith, Can. J. Mikrobiol. 11, 765 (1965). 3j. N. Collie, J. Chem. Soc. 91, 1806 (1907). ' A. J. Birch and F. W. Donovan, Austr. J. Chem. 6, 360 (1953).
520
ANTIBIOTIC BIOSYNTHESIS
[38]
hydrolase activity (peak enzyme tubes: 4 and 20). To prove that GTP is the substrate, phosphatase activity is measured in each fraction. Eighty percent of the phosphatase activity is lost in the 30-60% ammonium sulfate fraction. Phosphatase activity is assayed by incubating the enzyme with 25 nmoles [fl-~-32p]GTP (1400 cpm). The reaction mixture is added to 1 ml of a Norit A-celite column (1:1, w/w). The columns are washed with 3 volumes of 0.4 M phosphate buffer, pH 7.4. The inorganic 32p is counted. [fl-y-32p]GTP is prepared from GMP and KH232p04 by the method of Tochikura et al. ~° [fl-y-82P]GTP is purified by DE-52-cellular chromatography (30 g, gradient elution, 500 ml each of water and 0.4 M triethylammonium bicarbonate, pH 7.3). The fl_~/.82p labeled GTP shows a single UV spot on Avicel thin-layer chromatography plates, solvent: 6% ammonium chloride. P r o d u c t F o r m a t i o n . The predicted product following ring opening by GTP-8-formylhydrolase should be 4-N-[(5'-triphospho)-l'-ribosylamino]-2,5-diamino-6-hydroxypteridine. Only 0.16% of this ring-opened product exists following elimination of carbon-8 of GTP. The major product is neopterin. Since riboflavin and folic acid both use the purine ring for their biosynthesis, attempts to isolate these two compounds from the culture medium were made. These compounds are not present in the culture medium. lo T. Tochikura, K. Kawaguchi, T. Kano, and K. Ogata, J. Ferment. Technol. 47, 564 (1969).
[38] 6-Methylsalicylic Acid Synthetase B y G. VOGEL and F. LYNEI~
Under certain culture conditions several P e n i c i l l i u m species produce 6-methylsalicylic acid, the parent compound in the reaction sequence leading to the antibiotic patulin. 1,2 6-Methylsalicylic acid belongs to the class of naturally occurring compounds derived from head-to-tail condensation of acetate units, as postulated in the polyacetate or polyketide hypothesis2 ,4 Malonyl-CoA has been recognized as the building block 1j. D. Bu'Lock and A. J. Ryan, Proc. Chem. Soc. 222 (1958). 2j. D. Bu'Lock, D. Hamilton, M. S. Hulme, A. J. Powell, D. Shepherd, and G. N. Smith, Can. J. Mikrobiol. 11, 765 (1965). 3j. N. Collie, J. Chem. Soc. 91, 1806 (1907). ' A. J. Birch and F. W. Donovan, Austr. J. Chem. 6, 360 (1953).
[381
6 - M E T H Y L S A L I C Y L I C ACID S Y N T H E T A S E
521
for 6-methylsalicylic acid biosynthesis, and a reaction mechanism similar to that for fatty acid synthetase has been proposed2 The 6-methylsalicylic acid synthetase is a stable multienzyme complex which catalyzes 6-methylsalicylic acid formation according to the following equation: CH3CO-SCoA + 3HOOCCH2CO-SCoA + NADPH ÷ H+ CH3 [~COOH+ "OH
3CO2 + 4HSCoA * NADP++ H20
Assay Methods The activity of the enzyme complex can be determined by measuring the incorporation of radioactively labeled acety1-CoA or malonyl-CoA into 6-methylsalicylic acid2 ,7 6-Methylsalicylic acid has to be separated by gaschromatography or thin-layer chromatography prior to determination, because fatty acid synthetase is also present in crude extracts, producing radioactively labeled fatty acids. After complete separation of fatty acid synthetase, the rate of NADPH oxidation can be followed spectrophotometrically at 340 nm. 6 A more sensitive assay, specific for 6-methylsalicylic acid synthetase, is the fluorometric determination of enzymatically formed 6-methylsalicyclic acid. Fluorimetric A s s a y Principle. The fluorescence spectra of 6-methylsalicyclic acid and NADPH are shown in Fig. 1. In 0.1 M potassium phosphate, pH 7.6, containing 1.25 mg/ml bovine serum albumin, the maximal excitation wavelength is 308 nm and the maximal emission wavelength is 410 nm. Using an excitation wavelength of 310 nm and measuring the fluorescence intensity at 390 nm, 6-methylsalicylic acid determination is not disturbed by the concomitant change of NADPH fluorescence. Fluorescence intensity of 6-methylsalicylic acid is increased about 30-fold in the presence of bovine serum albumin. Fluorescence has been measured at 90 ° to the exciting beam using a Hitachi MPF-3 fluorescence spectrophotometer. The reaction may be carried out using any temperature-stabilized fluorometer which incorporates a recorder fitted with a zero suppression de-
F. Lynen and M. Tada, Angew. Chem. 73, 513 (1961). ~'P. Dimroth, H. Walter, and F. Lynen, Eur. J. Biochem. 13, 98 (1970). R. J. Light, J. Biol. Chem. 242, 1880 (1967).
522
ANTIBIOTIC BIOSYNTHESIS
Sl I
o) r-
I
I
I
l | l I I I I l
/
c o o
o I1.
!
/
/
I
\\
\\
I I
/
U
[38]
l I l I l I I l t l
I
! !
\
420
460
/
\
\
\
,
l I
I
280
320
360
380 Wavelength,
480
nrn
FIG. 1. Fluorescence spectra (left set, excitation spectra; right set, emission spectra) of 6-methylsalicylic acid ( ) and NADPH ( - - - ) in 0.1 M potassium phosphate, pH 7.6, containing 1.25 mg of bovine serum albumin per milliliter. The excitation spectrum of 6-methylsalicylic acid (c = 0.5 #M) was measured by setting the emission wavelength at 410 nm, and the emission spectrum was measured by setting the excitation wavelength at 310 nm. The excitation spectrum of NADPH (c = 20 #M) was measured by setting the emission wavelength at 455 nm, and the emission spectrum was measured by setting the excitation wavelength at 334 nm. The excitation and emission wavelengths used in the assay of 6-methylsalicyclic acid synthetase are indicated by the arrows. vice, n e c e s s a r y for zeroing o u t e x t r a n e o u s fluorescence i n t e n s i t i e s due to the high p r o t e i n c o n t e n t in the cuvette. Reagents P o t a s s i u m p h o s p h a t e buffer, 1 M , p H 7.6 B o v i n e s e r u m a l b u m i n , 25 m g / m l N A D P H , p H 8, 20 m M A c e t y l - C o A , p H 5-7, 10 m M 8 M a l o n y l - C o A , p H 5-7, 20 m M 9 sPrepared from the corresponding anhydrides by the method of E. J. Simon and D. Shemin, J. Amer. Chem. Soc. 75, 2520 (1953). 9For preparation, see F. Lynen, this series, Vol. 5 [60].
[]8]
6-METHYLSALICYLIC ACID SYNTHETASE
523
Procedure. The reaction mixture (in a 3.0-ml cuvette, d = 1 cm) contains in a total volume of 2 ml, 0.2 ml of 1 M potassium phosphate, pH 7.6, 2.5 mg of bovine serum albumin, 0.2 mM NADPH, 0.1 mM acetyl-CoA, and 0.2-1.0 mU of enzyme. The assay is performed at 25 °. The background fluorescence is compensated for by an electrical offset circuit. After establishment of a stable baseline, 0.02 ml of malonyl-CoA (0.2 mM) are added, and the increase in fluorescence associated with 6methylsalicylic acid formation is followed. The extent of fluorescence change is related to that obtained with a standard solution of 6-methylsalicylic acid. TM Sensitivity of the fluorometer is adjusted so that addition of 2 nmoles of 6-methylsalicylic acid (10 ~l of a 0.2 mM solution) gives a full-scale deflection on the recorder under the usual assay conditions. This assay also works with crude extracts of enzyme provided that the solution is clear. Units. One unit of enzyme is defined as that amount of enzyme catalyzing the formation of 1 ~mole of 6-methylsalicylie acid per minute at 25 ° under the previously described assay conditions. Protein is determined by the biuret method 11 or in solutions of low protein content by the spectrophotometric method at 260 and 280 nm. 11 Organism and Growth Conditions Agar slants of Penieillium patulum N R R L 679 (Central bureau voor Schimmelcultures, Baarn, The Netherlands) are maintained at room temperature for a minimum of 10 days with transfers at 4-6-week intervals. To prepare the medium, 15 g of bacto agar (Difco) are dissolved with heating in 1 liter of H~O, and 20 ml of wort (LSwenbrauerei, Munich) arc added. Stock cultures are kept at 4 ° to prevent strain variation. A separate slant is used to inoculate each 2-liter Fernbach flask containing 650 ml of a modified Czapek-Dox medium, ( N a N Q , 3 g; KH2P04, 2.5 g; K2HPO~, 2.5 g; MgSO~'7H20, 0.5 g; KC1, 0.1 g; yeast extract (Difco), 3 g; glucose, 30 g; and distilled water to 1 liter). Flasks are incubated at 30 ° for 36 hr on a rotary shaker using a 5-cm stroke and 100 rotations per minute. The contents of two of these flasks are then added to 10 liters of the modified Czapek-Dox medium in a 14-liter fermentor vessel (New Brunswick, F-14). Cultures are grown in a Fermentor (New Brunswick, FS314) for 24 hr at 25 ° with an aeration rate of 6 liters per minute and an agitation rate of 200 rpm. Foaming is prevented by addition of 0.5 ml of Niax Polyol (Brentag GmbH, Mfilheim, Germany). Yield from lo For preparation, see this volume [39]. 1~E. Layne, this series, Vol. 3 [73].
524
ANTIBIOTIC BIOSYNTHESIS
[38]
a typical 10-1iter fermentation is 150-200 g of wet cells. The mycelia are collected on a cheesecloth lining a wire basket and washed with 0.9% of sodium chloride solution. The cheesecloth is then wrung out forcibly to remove residual liquid. This is essentially the method used by Dimroth et al2 Fresh cells are used immediately or are lyophilized and stored at --20 ° without loss of enzymic activity over several months, if kept in tightly sealed containers. A shake culture technique, which is of advantage to laboratories lacking a fermentor, is described in the article on 6-methylsalicylic acid decarboxylase?°
Purification Procedure All operations are carried out at 0-4 °. Step 1. Cell Breakage and Extraction of the Enzyme. About 1000 g of fresh cells or 250 g of lyophilized mycelia are suspended in 6-8 liters of 0.2 M potassium phosphate, pH 7.6, containing 6% (w/v) polyethylene glycol 6000 (Serva, pract, grade), 5 mM 2-mercaptoethanol, and 5 mM EDTA. The suspension is cooled to 3-5 °, and then the cells are disrupted by passage once through a Gaulin press (Manton-Gaulin, Everett, Massachusetts) at 600 arm. Unbroken cells and cell fragments are removed by centrifugation at 5000 g for 40 min, and the supernatant fraction is saved. Small-scale extraction is carried out by agitating a suspension of 7.5 g wet weight cells in a cell homogenizer12 for 1 min with 30 g of glass beads (size 31/10, Dragonwerk Wild, Bayreuth) in 0.2 M potassium phosphate, pH 7.6, containing 5 mM 2-mercaptoethanol, and 5 mM EDTA. Step 2. Polyethylene Glycol 6000 Precipitation. The desired concentration of polyethylene glycol is achieved by using a stock solution prepared by dissolving 50% (w/w) of polyethylene glycol 6000 in 50 mM potassium phosphate, pH 7.6. To 1 liter of the supernatant fraction 360 ml of this polyethylene glycol solution are added slowly while the mixture is stirred mechanically. The solution is stirred for a further 20 min and centrifuged for 40 rain at 5000 g. Step 3. Polyethylene Glycol 1500 Fractionation. The supernatant fraction is discarded, and the precipitate is well homogenized using a plastic homogenizer with a Teflon plunger in about 300 ml of 0.2 M potassium phosphate, pH 7.6, containing 8% (w/v) polyethylene glycol 1500 1~M. Merkenschlager, K. Schlossmann, and W. Kurz, Biochem. Z. 329, 332 (1957).
[381
6-METHYLSALICYLIC ACID SYNTHETASE
525
(Serva, pratt, grade), 0.5 M sodium chloride, 5 mM 2-mercaptoethanol, and 5 mM EDTA. The resulting precipitate is removed by centrifugation at 10,000 g for 20 min. To each 100 ml of the green supernatant liquid a 70-ml portion of a solution containing 50% (w/w) polyethylene glycol 1500 dissolved in 50 mM potassium phosphate, pH 7.6, is then added dropwise, while the mixture is stirred gently. Stirring is continued for a further 20 min, and the suspension is centrifuged at 15,000 g for 15 min. Step 4. Ammonium Sul]ate Fractionation. The green precipitate resulting from step 3 is homogenized in about 300 ml of 0.2 M potassium phosphate, pH 7.6, containing 38.4 g ammonium sulfate (22% saturation), 5 mM 2-mercaptoethanol, and 5 mM EDTA. The cloudy liquid is centrifuged at 10,000 g for 20 rain, and the supernatant fraction is taken to 38% saturation by the addition of solid ammonium sulfate. The supernatant fraction, containing a green protein component which shows peroxidase activity, is discarded. The precipitated protein is dissolved in a minimal volume (about 50 ml) of 50 mM potassium phosphate, pH 7.6, containing 5 mM 2-mercaptoethanol (EDTA omitted), and desalted by filtration on a 5 X 25 cm column of Sephadex G-25 previously equilibrated against the same buffer. Step 5. Hydroxyapatiie Chromatography. The desalted, slightly yellow solution is then placed on a 5 X 20 cm column of hydroxyapatite (Bio-Gel HT, Bio-Rad) preequilibrated with the desalting buffer. A layer of about 1 cm of the adsorbant is whirled up carefully so as to mix with the protein solution and is then allowed to settle again. By this procedure a sharp starting zone is obtained. This starting zone is readily observable by examining the fluorescence of flavin components of the solution using a UV-lamp. Up to this stage of the purification procedure, the 6-methylsalieylic acid synthetase and the fatty acid synthetase of this organism are purified together, with similar yields and a comparable purification factor. As can be seen from Fig. 2, they are completely separated by hydroxyapatite chromatography. 6-Methylsalicylic acid synthetase is eluted from the column witll a linear gradient established between 800 ml of 50 mM and 800 ml of 0.22 M potassium phosphate, pH 7.6, containing 5 mM 2-mereaptoethanol. An uniform flow rate of 100 ml per hour is maintained by means of a pump. The eluent is collected, and active fractions arc pooled. If desired, fatty acid synthetase can be isolated by a further stepwise elution with potassium phosphate, pH 7.6, of increasing ionic strength (see Fig. 2). Elution, readily observable by the bright yellow fluorescence of the enzyme, usually occurs during the application of a 0.25 M phosphate buffer to the column.
526
[38]
ANTIBIOTIC BIOSYNTHESIS
.....
.,~t100'
OD280n m
"//~
"/t' 10.05M
mm~
eoo~
:. ".
t ~,
"
t
i
[I:: :
!!
[ t| " i"
" -:"
I
I
, I..'""-
'"I
:" :"'"
,~0
-~
iIIi
1 I
,oo
!
...
:'.
°.-".%
"1
..
|
80 Fractions
I
I
D 0.22M I Potassium phosphate rnolarity
"~
~"."
I I i
I"" I I
I
120 0.25M
I 0.5M
I
FIG. 2. Hydroxyapatite column. Desalted protein from step 4 (1.6 g in a volume of 72 ml, 60 units) was placed on a 5 × 20 cm column as described in the text. 6-Methylsalicylic acid synthetase was eluted in 14 hr (overnight) with a linear gradient of 0.05 M and 0.22 M potassium phosphate (pH 7.6, 5 mM 2-mercaptoethanoD, 1600 ml total volume. The flow rate was maintained at about 110 ml/hr, and 15-ml fractions were collected. Fatty acid synthetase was eluted by a stepwise increase of phosphate concentration to 0.25 M. Peak I and peak II contained green fluorescent proteins, visualized with a UV lamp, whereas fatty acid synthetase is bright yellow fluorescent.
Step 6. Gel Filtration on Sepharose 6B. The solution, containing 6methylsalicylic acid synthetase activity, is t a k e n to 45% saturation with solid a m m o n i u m sulfate. The precipitated protein is brought down by centrifugation at 10,000 g for 30 min and dissolved in a minimal volume (about 25 ml) of 50 m M potassium phosphate, p H 7.6, containing 5 m M 2-mercaptoethanol and 5 m M E D T A . The concentrated enzyme solution is placed on a 5 X 90 cm column of Sepharose 6B (Pharmacia) previously equilibrated with the above buffer. The protein is then eluted from the column at a flow rate of approximately 100 ml per hour, 12-ml fractions being collected. E n z y m e activity appears after an elution volume of about 800 ml in the second of three m a j o r protein peaks. Tubes containing activity are pooled for subsequent fractionation. The leading and trailing edges of the enzyme p e a k are discarded. Step 7. DEAE-CeUulose Chromatography. The pooled fractions from the Sepharose-6B column are loaded onto a 5 X 16 cm DEAE-cellulose
[38]
6-METHYLSALICYLIC ACID SYNTHETASE
527
column (Whatman DE-52) preequilibrated with 50 mM potassium phosphate, pH 7.6, containing 5 mM 2-mercaptoethanol, and 5 mM EDTA. The enzyme is eluted by a linear gradient established between 500 ml of the starting buffer and 500 ml of the same buffer containing 0.2 M sodium chloride, 10-ml fractions being collected. The enzyme appears at the end of the gradient as a broad peak. Fractions containing the major portion of the enzyme are pooled and concentrated by ammonium su~.fate precipitation as described in step 6. The precipitated protein is dissolved in a minimal volume of an appropriate storage buffer (see below). Better recovery of enzymic activity is sometimes observed using an alternative procedure. The DEAE-cellulose column is equilibrated with the same equilibration buffer described above, but containing 5% (w/v) polyethylene glycol 1500. In this case the enzyme is eluted by a linear gradient of 0.1 M to 0.4 M sodium chloride in the starting buffer, 1000 ml total volume. The enzyme again appears as a broad peak at the end of the gradient. Fractions containing the bulk of the activity are pooled and the enzyme is concentrated by centrifuging this solution at 100,000 g for 14 hr. The resulting pellet is homogenized in an appropriate storage buffer (see below). The results of a typical purification procedure starting with 950 g of wet weight cells are summarized in the table and show that there is approximately a ll0-fold increase in specific activity with a 20% recovery of enzymic activity.
PURIFICATION OF 6-METHYLSALICYLIC ACID SYNTHETASE
Step and fraction 1. Centrifuged extract 2. Polyethylene glycol 6000 precipitation 3. Polyethylene glycol 1500 fractionation 4. 22-38% Ammonium sulfate fraetionation 5. Hydroxyapatite eluate 6. Sepharose-6B eluate 7. DEAE-cellulose eluate a
Volume (ml)
Protein (mg)
Total Specific activity activity Re(milli(milli- covery units) units/rag) (%)
8200 320
48,000 10,800
93,000 82,000
280
4,700
75,000
16
80
48
1,600
60,000
37.5
64
320 290 270
410 180 92
38,000 29,000 19,000
92.5 161 206
41 31 20
1.9 7.5
100 88
DEAE-eellulose chromatography without polyethylene glycol 1500 in the equilibrating buffer.
528
ANTIBIOTIC BIOSYNTHESIS
[38]
Properties
Stability and Storage. The 6-methylsalicylic acid synthetase is unstable in solutions of low ionic strength or when exposed to acid or alkaline pH. The purified enzyme can be stored for several weeks at --15 ° in 0.2 M potassium phosphate, pH 7.6, containing 50% glycerol, and 1 mM dithiothreitol, or at 2 ° in 0.2 M potassium phosphate, pH 7.6, containing 5% (w/v) polyethylene glycol 1500, and 1 mM dithiothreitol. Pure or partially purified enzyme preparations dissolved in 0.2 M potassium phosphate, pH 7.6, can be lyophilized and stored at --25 ° for at least one year with only slight loss of activity, when protected from moisture. Homogeneity. The purified enzyme forms a single peak in the ultracentrifuge and migrates as a single protein band upon electrophoresis on cellogel strips (Serva) at various pH values. It forms a single precipitation band in Ouchterlony double-diffusion plates and immunoelectrophoresis against rabbit antisera prepared against purified enzyme preparations. With acrylamide disk electrophoresis, two or sometimes three diffuse bands are observed. Possibly dissociation of the enzyme takes place under these conditions. The enzyme is completely free of fatty acid synthetase. Kinetic Properties. The reaction has an optimal pH of about 7.6. 6 The Km values for acetyl-CoA and malonyl-CoA are both 20 ~M. 6 Molecular Weight. A molecular weight of about 1.2 X 106 has been determined by Dimroth et al., 6 using sucrose density gradient centrifugation with catalase and yeast fatty acid synthetase as references. A molecular weight of about 3.7 X 10~ is reported by Light and Hager, 1~ using gel filtration on a standardized Sephadex G-200 column and sucrose density gradient centrifugation. This great discrepancy has not been explained to date. The preparations used in the latter case have been unstable crude extracts. Perhaps the enzyme complex is dissociated under the conditions employed. Inhibition. The enzyme is inhibited by sulfhydryl reagents such as iodoacetamide and N-ethylmaleimide. 6 The inactivation rate with iodoacetamide is pH independent whereas that with N-ethylmaleimide increases with increasing pH. This indicates the participation of two types of SH-groups in the catalytic process, analogous to the yeast fatty acid synthetase. 14 One of these SH groups is identified as part of a 4'-phosphopantetheine residue of the "acyl carrier protein. ''1~ 4'-Phosphopantetheine is released by alkaline hydrolysis from the enzyme complex, and ~' R. J. Light and L. P. ttager, Arch. Biochem. Biophys. 125, 326 (1968). 14F. Lynen, this series Vol. 14 [3]. is G. GreulI, Doctorate Thesis, University of Munich, Munich, Germany, 1973.
[38]
6-METHYLSALICYLIC ACID SYNTHETASE
529
E-SH 1. CH3CO-SCoA ÷ 2HOOCCH2CO-SCoA ~ " CH3COCH2COCH2CO-SE÷2CO2 OH l /,~ (%J..
Triacetic acid lactone
/ H3C~XO/"~O 2. CH3COCH2COCH2CO~SE+ NADPH + H*-
OH m-HO-Benzyl olcohol
OH m - HO- Benzoldehyde
HOO:
CHO
OH m - H O - Benzoic acid
I
HOO: CH3
OH Toluquinol
CH20H NADP~ NADPH OH
COOH OH
Gentisyl olcohol
OH
Genlisaldehyde
Gentisic acid
NADPH
02
~NADP~
r CHO l" L
~. COOH -CH20H
OH coo.] CHO CHzOH
~o0 ~0 H
J
P r e - palulin
Potulin
FIG. 1. The major pathway for patulin biosynthesis in Penicillium urticae (NRIRL 2159A) is indicated by heavy arrows with branch reactions indicated by light arrows and probable additional reactions by dashed arrows. Metabolites which often accumulate to significant levels in the culture medium are drawn with heavy lines. Square brackets indicate proposed intermediates.
thesis steins from its use as a simple model system in examining the enzymology, regulation and raison d'gtre of an important group of secondary metabolites, the polyketides. 1 These secondary metabolites are produced principally b y the fungi, and to a lesser extent by bacteria and higher plants. Among the fungi, polyketides are produced primarily by the fungi imperfeeti and the aseomyeetes and rarely by the basidiomyeetes. The few polyketides of major commercial interest are the antibioties, griseofulvin and the tetraeyelines, and the animal toxins, the aflatoxins. Only the biosyntheses of patulin and the tetraeyelines have been extensively studied at the enzymologieal level. The major in vivo pathway for patulin biosynthesis has been determined ~ by kinetic pulse labeling techniques, and an up-to-date scheme is given in Fig. 1. With reference to this .figure, it is noteworthy that the mono-oxygenases catalyzing steps 3 and 5 of the pathway are the only enzymes not yet detected in cell-free extracts of P. urticae. The most recent prog4 p. I. Forrester and G. M. Gaucher, Biochemistry 11, 1102 (1972).
542
ANTIBIOTIC BIOSYNTHESIS
[40]
ress has been a demonstration of the dioxygenase mediated ring cleavage of gentisaldehyde2 Only 6-methylsalicyclic acid synthetase,6 6-methylsalicylic acid decarboxylase7 and m-hydroxybenzyl-aleohol dehydrogenases have been purified to any extent. The latter enzyme (3-hydroxybenzylalcohol:NADP ÷ oxidoreductase, EC 1.1.1.97) which catalyzes step 4 of the pathway is the subject of this article. Source of E n z y m e
Vegetative cells of a white variant (NRRL 2159A) of the relatively common soil fungus,~ P. urticae, 9 are the source of this enzyme. On the basis of a preliminary survey of seven fungis this dehydrogenase may be restricted to closely related fungi possessing all or part of the patulin pathway. Five milliliters of a detergent spore suspension are prepared from an agar slant IN 15 cm2; 49 g of Czapek solution agar (Difco) plus 5 g of Bacto-Agar (Difco) per liter], incubated for 7 days at 28 °, and stored at 4 ° for not more than 1 month. Three such spore suspensions are used to inoculate 3 liters of medium (Difco yeast extract, 5.0 g; glucose, 40.0 g; and distilled water to 1 liter) in a 4-liter New Brunswick Microferm fermentor operating at 28 ° and with aeration and agitation rates initially at 0.5 liters/min and 200 rpm, and then increased to 3 liters/min and 500 rpm at 5 hr and to 5 liters/min and 700 rpm at 18 hr. Cells (,~ 5-6 g dry/liter) harvested at 28-34 hr yield approximately 180 mU of enzyme per milligram of dry cells. Similar growth conditionss using medium containing 0.3% yeast extract and salts, yield cells (,-, 4 g dry/ liter) at 20-24 hr containing approximately 58 mU of enzyme per milligram of dry cells. After harvesting by suction filtration, cells are washed twice with distilled water, lyophilized, and stored dry at --15 °. Before storage these lyophilized cells contain the same dehydrogenase activity as fresh cells, whereas after 2 months of storage they retain approximately 75% of their dehydrogenase activity. Shake cultures s in 250-ml conical flasks (26-28°; 180 rpm on rotary shaker) containing 100 ml of an essentially identical medium except for the omission of yeast extract yield cells at 8 days containing an e s t i m a t e d 1.6 mU of 5A. I. Scott and L. C. Beadling,Bio-organ. Chem. 3, 281 (1974). eThis volume [38]. This volume [39]. 8p. I. Forrester and G. M. Gaucher,Biochemistry 11, 1108 (1972). ' Penicillium urtlcae Bainier, Penicillium patulum Bainier, and Penicillium flexuosum Dale are synonyms; Penicillium griseo]ulvum Dierckx is closely related.
[40]
?~t-HYDROXYBENZYL-ALCOHOL DEHYDROGENASE
543
enzyme per milligram of dry cells. These latter authors 5 used fresh cells. Preparation of Cell-Free Extracts Large quantities of cells are best broken using the liquid shear techniques of sonication or liquid pressing. 1° Lyophilized cells (2 g) are suspended in 60 ml of 10 mM TES [N-Tris (hydroxymethyl)methyl-2aminoethanesulfonic acid] buffer (pH 7.6) containing 1 mM (DTT) dithiothreitol and 1 mM MgClz. 11 The cell slurry is sonicated in a 100-ml rosette cell immersed in an ice bath for 7 min at the maximum setting of a Bronson sonicator (model 5125). Alternatively 10 g of wet cells per 60 ml of sonicating buffer can be treated as above. The supernatant from a 30-rain centrifugation at 30,000 g yields the crude cell-free extract. In an alternative procedure ~ a slurry of 30 g of wet cells per 100 ml of 50 mM phosphate buffer (pH 7.6) containing 1 mM DTT is passed through a French press at 10,000-20,000 psi. For small samples more reproducible cell breakage is obtained by the liquid shear technique of ballistic disruption. Lyophilized cells (0.1 g), glass beads (20 g; 0.45-0.50 mm; Braun) and 8 ml TES buffer (identical to sonicating buffer above) are added to a 50-ml Braun flask. The flask is then shaken for 30 sec in a Braun shaker (model MSK) at 4000 rpm. The resulting suspension may be clarified by suction filtration through Whatman No. 4 filter paper and by centrifugation. This latter method generally releases more enzyme than sonication. Despite the addition of various stabilizing additives, s such crude extracts retain only 11% dehydrogenase activity after 24 hr at 3% However, treatment with Polyclar AT (insoluble polyvinylpyrrolidone; GAF Corp., New York, New York) has been shown to stabilize and activate (1.5fold) such extracts. 5 After boiling for 10 min in 10% HC1, Polyclar AT was washed free of chloride with distilled water, equilibrated with buffer, and then separated from excess buffer by decantation. A wet weight of this Polyclar slurry equal to the original wet weight of cells broken is added to the cell extract. The suspension is stirred for 10 min and then clarified by centrifugation. Extracts from cells grown for only 28-34 hr as described above are, however, not activated by Polyclar. lOVarious grinding and homogenization techniques generally liberate less dehydrogenase. Since two freeze-thaw treatments yield cells devoid of dehydrogenase activity, the solid shear technique of freeze pressing with a Hughes press is also unsatisfactory. 11In all these procedures the use of freshly prepared buffer containing dithiothreitol is imperative.
544
ANTIBIOTIC BIOSYNTHESIS
[40]
Dehydrogenase Assay Principle. This kinetic, spectroscopic assay is typical of dehydrogenases and depends upon determining the rate of loss of NADPH during the reduction of m-hydroxybenzaldehyde. Reagents Buffer: 10 mM TES (pH 7.6) containing 1 mM MgC12 Substrate: 2 mM m-Hydroxybenzaldehyde (Sigma; recrystallized from water; mp 106 °) in water Cofactor: NADPH.Na4 (Sigma), 5 mM in a 1% NaHC03 solution
Procedure. An appropriate amount (usually 0.02-0.40 ml) of enzyme solution is added to 0.5 ~mole of NADPH and TES buffer previously equilibrated at 30 °. After a stable baseline is achieved, the reaction is started by adding 0.4 ~mole of substrate to yield a total volume of 3.0 ml in a quartz cuvette in the 30 ° thermostated cell compartment of a spectrophotometer. The loss of NADPH is monitored by the decrease in absorbance at 340 nm. The assay is linear with increasing enzyme concentration up to a A absorbance of 0.25. Units and Calculations. One unit of activity is equal to that amount of enzyme necessary to oxidize 1 ~mole NADPH per minute. Assay results are routinely expressed in milliunits (mU), that is nanomoles of substrate converted per minute. Since both NADPH QpH \ 340 7.~ 6.22 × 103) and substrate \(~pH 340 7.6 1.6 × 103) absorb at 340 nm, the sum of their extinction coefficients is used in converting AOD/min into enzyme units. ~ Other Assays. A similar gentisylalcohol dehydrogenase assay may be carried out using gentisaldehyde ( ~ m 2.7 X 103) as substrate. 5 A more time consumingfixed-time radiochemical assay using [1-~4C]m-hydroxy benzaldehyde as substrate has also been reported2 Purification Despite this enzyme's instability some purification has been possible. The crude cell-free extract is brought ~o 2% (w/v) in streptomycin sulfate (Sigma) and after 10 rain is centrifuged at 10,000 g for 10 min. Solid ammonium sulfate (Mann Ultra Pure) is added to the resulting supernatant until 40% saturation is reached. After centrifugation at 10,000 g for 10 min the resulting precipitate is discarded and the supernatant is brought to 65% saturation in ammonium sulfate. The precipitate obtained by centrifugation at 10,000 g for 10 min is dissolved in 12Given that substrate (S) and cofactor (C) both absorb at 340 nm, AOD3,0= esA[S] + e¢A[C], and since MS] = A[C] then A[S] = AOD34o/¢~+ e¢.
[40]
m-HYDROXYBENZYL-ALCOHOL DEHYDROGENASE
545
TABLE I PARTIAL PURIFICATIONOFTn-HYDROXYBENZYL-ALCOHOL])EHYDROGENASE ~
Purification step Crude supernatant Steptomycin sulfate precipitation Ammonimn sulfate fractionation Sephadex G-200
Volume Total (ml) (milliunits) 275 275
115,500 115,500
12
70,000
42
17,000
Protein concentraiion b (mg/ml) 4.2 13 0.55
Specific activity (milliunits/rag of Yield proleiT~) (%)
Purifi alion (-fold)
100 -
100 100
--
455
6l
4.55
655
15
6.55
Reproduced by permission of the American Chemical Society from Biochemislrg 11, 1108 (1972). b Both the Folin-Lowry method and the 280:260 nm ratio method were used ~o determine protein concentration using bovine serum albumin (Sigma) as ~t standard. a minimum volume of 10 m M T E S buffer (pH 7.6) which is 1 m M in both dithiothreitol and MgCl~. This a m m o n i u m sulfate fraction is applied to a P h a r m a e i a K25/100 column packed with Sephadex G-200 (fine) which has been previously equilibrated with the above T E S buffer which also is 0.02 m M in N A D P +. Elution is carried out using the above equilibration buffer and 5 ml fractions are collected. Fractions containing dehydrogenase activity are pooled. The 6.6-fold purification yields a preparation with a specific activity of 655 mitliunits per milligram of protein as detailed in Table I. As recently reported '~ improved purification of what is probably the same dehydrogenase results in a 46-fold purification and a final specific activity of approximately 1100 milliunits per milligram of protein. Tile enhanced purification results from a Polyclar AT treatment of the crude extract and a diethylaminoethyl (DEAE)-cellulose column chromatography step using a 0.05-0.50 M phosphate huffer (pH 7.6) gradient. This purified preparation is 50% active after 48 hr. Finally, this preparation exhibits m-hydroxybenzyl alcohol and some gentisyl "tlcohol dehydrogenase activity. Hence the existence of one or two dehydrogenases is uncertain.
Properties Despite the fact t h a t a stable, highly purified preparation of this dehydrogenase is not available, some characterization has been possible and
546
ANTIBIOTIC BIOSYNTHESIS
[40]
TABLE II INHIBITION STUDY OF UNPURIFIED m-HYDROXYBENZYLALCOHOL DEHYDROGENASE a
Preassay treatment b Control + Iodoacetic acid + NADPH, then iodoacetic acid + NADP +, then iodoacetic acid + m-Hydroxybenzaldehyde, then iodoacetic acid + Diethylpyrocarbonate + NADP +, then diethylpyrocarbonate + m-Hydroxybenzaldehyde, then diethylpyrocarbonate
Relative activity (%) 100 46 97 89 45 19 65 19
a Reproduced by permission of the American Chemical Society from Biochemistry 11, 1108 (1972). b Ten-milliliter portions of crude supernatant were incubated at 0° with either 1 ~mole of m-hydroxybenzaldehyde, 1 ~mole of NADP +, or 1 ~mole of NADPH. After 5 min equilibration, 10 ~moles of iodoacetic acid (BDH, recrystallized before use) were added to each sample. After a further incubation at 0° of 2 hr the activity of each sample was determined by assaying 0.1-ml aliquots. Similarly, 10-ml portions of crude supernatant, adjusted to pH 6.0 with 1 N HC1, were incubated at 0° with 1 ~mole of NADP + or with 1 ~mole of m-hydroxybenzaldehyde. After 5 min of equilibration, 10 ~moles of diethylpyrocarbonate (Eastman) were added. After a further 3 min at 0° the activity of each sample was determined by assaying 0.1-ml aliquots.
has been carried out using crude enzyme preparations, s This intracellular, soluble dehydrogenase (EC 1.1.1.97) has an approximate molecular weight of 120,000 (gel filtration method) and can be readily detected in crude extracts as a single enzymically active band in polyacrylamide disc gel electrophoresis. As indicated in Table I I the inactivation of the dehydrogenase by the sulfhydryl reagent, iodoacetate, and the histidine reagent, diethylpyrocarbonate, and the protection provided by N A D P H agree with what has been found for other dehydrogenases. The enzyme has a fairly sharp pH optimum at 7.6 and exhibits classic MichaelisMenten plots of initial rate versus substrate concentration (approximate K~ values = 4 to 5 X 10-5 M). The equilibrium constant and standard free-energy change for the reaction have been calculated to be: K'app -- [aldehyde] [ N A D P H ] / [ a l c o h o l ] [ N A D P +] = 0.18 (pH 7.6, TES, 30 °) and AG°1 ---- + 1.04 kcal/mole (pH 7.6, TES, 30°). The reduction of m-hydroxybenzaldehyde, which is opposite to the preferred pathway direction, is clearly favored (i.e., the reverse rate relative to the
[40]
m-HYDROXYBENZYL-ALCOHOL DEHYDROGENASE
547
TABLE I I I SUBSTRATI~ SPECIFICITY OF UNPURIFIED m-HYDROXYBENZYLALCOHOL DEHYDROGENASE a
Relative substrate Relative inhibitor activity activity
m-Hydroxybenzaldehydeg m-Methoxybenzaldehyde~ AcetaldehydeS Benzaldehyde p- Hydroxybenzaldehyde o-Hydroxybenzaldehyde Acetophenone m-Hydroxyacetophenone p-Hydroxyacetophenone p-Hydroxyoctaphenone Gentisaldehydeh
(~o)b
(G~)C
100 100 29 20 10 0 0 0 0 0 0
-0 16 0 37g 0 0 0 5()g 70~ 36
a Reproduced by permission of the American Chemical Society from Biochemistry 11, 1108 (1972). b In each case the cofactor NADPH and standard assay conditions of pH 7.6 and 30 ° were used. c Percent decrease in m-hydroxybenzyl-alcohol dehydrogenase activity upon addition of an equimolar amount of various substrate analogs. d Substitution of NADH for NADPH yielded a relative activity of ~12%, while substitution of m-hydroxybenzyl alcohol and NADP O for substrate and cofactor, respectively, yielded a relative activity of ~18%. Furthermore substitution of NADP® or NAD@ for NADPH yielded no activity. e Substitution of m-methoxybenzyl alcohol and NADP O for substrate and cofactor, respectively, yielded a relative activity of ~ 9 %. ] Substitution of NADH for NADPH yielded a relative activity of 35%. g Concentration of added compound equal to only yl~ the substrate concentration since these compounds absorb strongly at 340 nm. h Substitution of NADH for NADPH yielded a relative activity of ~12 %.
forward rate is favored 6: 1). T h e r e a c t i o n ' s r e v e r s i b i l i t y clearly indicates t h a t the e n z y m e is a dehydrogenase, n o t a reductase. F i n a l l y , a l t h o u g h a crude e n z y m e p r e p a r a t i o n was used, the d e h y d r o g e n a s e ' s s u b s t r a t e specificity is d e l i n e a t e d in T a b l e I I I . T h e e n z y m e is clearly a n alcohol r a t h e r t h a n a n a l d e h y d e d e h y d r o g e n a s e a n d is N A D P + specific. A specificity for a r o m a t i c aldehydes with a meta(3)-hydroxy or m e t h o x y s u b s t i t u e n t is a p p a r e n t a n d a r o m a t i c k e t o n e s are good i n h i b i t o r s r a t h e r t h a n substrates. T h e presence of some of both N A D a n d N A D P e t h a n o l d e h y d r o genase a c t i v i t y is p r o b a b l y due to other enzymes. A l t h o u g h no gentisyl alcohol ( N A D P ) d e h y d r o g e n a s e a c t i v i t y was found in this p r e p a r a t i o n ,
548
ANTIBIOTIC BIOSYNTHESIS
[41]
such activity has recently been reported in a different preparation from P. urticaeP As has been discussed, both kinetic pulse-labeling experiments 4 and the characteristics of this enzyme 8 suggest that this dehydrogenase probably catalyzes the rate-determining step of the patulin biosynthetic pathway. Acknowledgment The development of unpublished procedures in this manuscript by Mr. Jan Groot Wassink is gratefully acknowledged.
[41] Bacitracin Synthetase B y HANSPETER RIEDER, GERHARD HEINRICH, EBERtIARD BREUKER,
MAHAVIR M. SIMLOT, a n d PETER PFAENDER
The bacitracins (Fig. 1) are a group of potent peptide antibiotics that are active against a variety of gram-positive, but only a few gram-negative, microorganisms. 1 The bacitracins and licheniformins ~ are produced by Bacillus licheni]ormis. Bacitracins A and B, whose structures are known, are the most common members of the group3,4; bacitracins D and E, whose structures are only partially known, contain valine, bacitracin C also yields glycineP In aqueous solution bacitracin A is transformed into the antibiotically inactive, yet equally nephrotoxic bacitracin F. 1 (B)
VAL
I I ~ (A)
~oRF-' .PHE
C~H~--C--C~-'# x'/---CO-LEU-D-GLU-I LE-*LYS ~ I I 's-# ,~SN CH 3 NH 2 --C ~N
)
HI S
"~-D.ASI~
Fro. 1. Bacitracins A, B, and F. B and F are similar to A except for the variations shown. 1R. J. Hickey, Progr. Ind. Microbiol. 8, 95 (1964). R. K. Callow and T. S. Work, Biochem. J. 51, 558 (1953). a L. C. Craig, J. R. Weisiger, W. Hausmann, and E. J. Harfenist, J. Biol. Chem. 199, 259 (1952). 4C. Ressler and D. V. Kashelikar, J. Amer. Chem. Soc. 88, 2025 (1966). G. G. F. Newton and E. 19. Abraham, Biochem. J. 53, 597 (1953).
548
ANTIBIOTIC BIOSYNTHESIS
[41]
such activity has recently been reported in a different preparation from P. urticaeP As has been discussed, both kinetic pulse-labeling experiments 4 and the characteristics of this enzyme 8 suggest that this dehydrogenase probably catalyzes the rate-determining step of the patulin biosynthetic pathway. Acknowledgment The development of unpublished procedures in this manuscript by Mr. Jan Groot Wassink is gratefully acknowledged.
[41] Bacitracin Synthetase B y HANSPETER RIEDER, GERHARD HEINRICH, EBERtIARD BREUKER,
MAHAVIR M. SIMLOT, a n d PETER PFAENDER
The bacitracins (Fig. 1) are a group of potent peptide antibiotics that are active against a variety of gram-positive, but only a few gram-negative, microorganisms. 1 The bacitracins and licheniformins ~ are produced by Bacillus licheni]ormis. Bacitracins A and B, whose structures are known, are the most common members of the group3,4; bacitracins D and E, whose structures are only partially known, contain valine, bacitracin C also yields glycineP In aqueous solution bacitracin A is transformed into the antibiotically inactive, yet equally nephrotoxic bacitracin F. 1 (B)
VAL
I I ~ (A)
~oRF-' .PHE
C~H~--C--C~-'# x'/---CO-LEU-D-GLU-I LE-*LYS ~ I I 's-# ,~SN CH 3 NH 2 --C ~N
)
HI S
"~-D.ASI~
Fro. 1. Bacitracins A, B, and F. B and F are similar to A except for the variations shown. 1R. J. Hickey, Progr. Ind. Microbiol. 8, 95 (1964). R. K. Callow and T. S. Work, Biochem. J. 51, 558 (1953). a L. C. Craig, J. R. Weisiger, W. Hausmann, and E. J. Harfenist, J. Biol. Chem. 199, 259 (1952). 4C. Ressler and D. V. Kashelikar, J. Amer. Chem. Soc. 88, 2025 (1966). G. G. F. Newton and E. 19. Abraham, Biochem. J. 53, 597 (1953).
[411
BACITRACIN SYNTHETASE
549
The nonribosomal synthesis of cyclic peptide antibiotics, such as gramicidin S and the tyrocidines, 6 on enzyme templates suggested the present study of baeitracin synthetase. 7-9 The study is concerned with the purification and partial characterization of bacitracin synthetase from a m u t a n t of B. licheniformis, L o h m a n n cells, exhibiting a high rate of bacitracin synthesis. Principle. Bacitracin is synthesized by the enzyme via Eq. (1): 3 lie + Cys + Leu + Glu + Lys + Orn + Phe + l i i s + Asp + Asn
n A T P , M g ~+ p H 7.5
, bacitracin + n PPI
(1I
The exact amount of A T P needed is not known. For bacitracin-synthesizing particles it has been determined to be about 2 A T P for 1 peptide bond. The amount of bacitracin formed can be measured by incorporation of one or more labeled amino acids into bacitraein, by radio thin-layer chromatography, 9 or selectron filters, 9 or by growth inhibition test after incubation, lyophilization of the mixture and application to culture plates planted with Micrococcus flavus according to Hoff. TM The activity of baeitracin synthetase can also be evaluated by ATP-32PPi exchange hie a su re menU. 9,~1 One unit of enzyme is defined as t h a t amount of bacitracin synthetase that incorporates 1 ~mole of one of the amino acids (1 t~mole of isoleucine) into baeitracin per minute at 25 °. Materials
Buffers M E , 2-Mercaptoethanol, Merck, D a r m s t a d t , G e r m a n y Buffer A: 25 m M sodium phosphate, p H 7.5 Buffer B: buffer A + 0.25 m M E D T A , 10 m M MgC1._,, 10 m M M E Buffer C: buffer B + 0.5 m M ATP, 0.4 m M bacitracin amino acids except cysteine 1 m M 6F. Lipmann, W. Gevers, H. Kleinkauf, and R. Roskoski, Jr., Advan. Enzymol. 35, 1 (1971). 7p. Pfaender, Zentralbl. Bakteriol. Parasitenk. Infektionskr. Hyg, I. Abt. Orig. A 230, 319 (1972). 8 p. Pfaender, D. Specht, G. Heinrich, E. Schwarz, E. Kuhnle, and M. M. Simlot, FEBS Lett. 32, 100 (1973). M. M. Simlot, P. Pfaender, and D. Specht, FEBS Lett. 35, 231 (1973). 4oD. A. Hoff, R. E. Bennett, and A. R. Stanley, Science 106, 551 (1947). 11M. M. Simlot and P. Pfaender, FEBS Lett. 35, 201 (1973).
550
ANTIBIOTIC BIOSYNTHESIS
[41]
Enzymes Alcohol dehydrogenase (ADH), EC 1. 1.1.1 Catalase, EC 1.11.1.6 Deoxyribonuclease (DNase), EC 3.1.4.5 Gtyceraldehydephosphate dehydrogenase (GA-3P-DH), EC 1.2.1.12 Lactate dehydrogenase (LDH), EC 1.1.1.27 Leucine aminopeptidase (LAP), EC 3.4.1.1 3-Phosphoglycerate kinase, EC 2. 7.2.3 Ribonuclease A (RNase), EC 2. 7.7.16 Urease, EC 3. 5.1.5 (All from Boehringer, Mannheim, Germany) Lysozyme, EC 3.2.1.17, (from Serva, Heidelberg, Germany) Other proteins Crystalline bovine serum albumin (BSA) Crystalline soybean trypsin inhibitor (SBTI) (Both from Serva, Heidelberg, Germany) Amino acids Asparagine, aspartic acid, glutamic acid, histidine, isoleucine, leucine, lysine, ornithine, phenylalanine; all from Merck, Darmstadt, Germany; cysteine from Schuchardt, Munich, Germany Labeled compounds Algae protein total hydrolysate, 250 ~Ci/5 ml [U-14C]Lysine, 50 ~Ci/0.023 mg Tetrasodium pyrophosphate, 2.48 ~Ci/~mole (All from Buchler, Braunschweig, Germany.) Fermentation media Bolus alba (white clay), Haiger, mine Niederdresseldorf, Germany Corn steep liquor, Kellogg, Osterholz, Scharmbeck, Germany Pharmamedia, Traders Protein Division, Fort Worth, Texas Someel, Holtz & Willemsen,Krefeld, Germany Column chromatography steps DEAE-cellulose, Merck, Darmstadt, Germany Sephadex G-50, Sephadex G-200, Pharmacia, Uppsala, Sweden Concentration step XM-50 Amicon membrane filters, 25 ram, Amicon, Oosterhout, The Netherlands Electrophoresis Sodium dodecyl sulfate (SDS), Merck, Darmstadt, Germany Growth inhibition test Triphenyltetrazolium chloride (TTC) Merck, Darmstadt, Germany Bacitracin, 68 IE per mg, Serva, Heidelberg, Germany
[41]
BACITRACIN SYNTHETASE
551
Liquid scintillation counting PPO, diphenyloxazole, Merck, Darmstadt, Germany
Assay Methods The following methods are described: (1) Micrococcus flavus growth-inhibition test (2) ATP-3'-'PPi exchange measurement (3) Millipore filter test (4) radio thin-layer chromatography of incorporated [14C]lysine 1. Micrococcus flavus Growth Inhibition Test. Fifteen milliliters of media I containing 4.0 g/liter of nutrient broth (Difco), 3.0 g/liter of yeast extract (Difco), 3.0 g/liter of Peptone (Difco), and 1.0 g/liter of glucose (Merck) are inoculated with a loop of M. flavus (ATCC 10240) and incubated at 37 ° for 18 hr. Culture plates are prepared with base agar containing 15 g of agar in 1 liter of media I without glucose, and seed agar, containing 15 g of agar, 4.0 g of Casamino acids (Difco), and 10 ml of M. flavus culture in 1 liter of media I. A 3 mm deep layer of base agar at 50 ° is poured onto sterile polyethylene culture plates with a diameter of 8.2 cm. After 20 rain a 1 mm-deep layer of seed agar at 45 ° is poured onto the base agar. Poured plates can be kept for 1 week at 4 °. Just before the assay, 5 sterile steel cylinders (0.5 X 0.7 cm) are gently placed onto each culture plate, 0.1 ml of bacitracin containing solution are filled into each cylinder and the plates are incubated at 37 ° for 18 hr. After incubation the cylinders are removed and the plates are reincubated at 37 ° with 0.1% TTC for 15 rain. The circular zones of inhibition are then measured against the red background. A diameter of 1.5 cm corresponds to 0.1 unit or 1.74 ~g of bacitracin. 2. ATP-3'2PP~ Exchange Measurement. The composition of the incubation mixture is ATP, 4.5 mM; MgCI~, 3.75 mM; KF, 22.5 mM; 2-mercaptoethanol, 7.5 raM; Na432P~O,, 4.5 mM (specific activity 2.48 mCi/mmole) ; BSA, 0.23 mg/ml; cysteine, 15 mM; the nine amino acids of baeitracin, each 4.0 mM; Tris.HC1 buffer, 109 mM; final pH 7.5. The enzyme solution in buffer B (20 ~l) is mixed in the cold with 40 ~l of incubation mixture and incubated at 0 °, 4 °, or 25 ° for the stated periods (Table I). A control is also incubated which contains water in place of amino acids. After incubation, the mixture is immediately frozen to stop the reaction, thawed, and 10 ~l of a solution of 0.4 M Na,P~O; in 15% perchloric acid is added. After shaking the solution it is centrifuged and 50 ~l of
552
ANTIBIOTIC
E~
o~
[41]
BIOSYNTHESIS
=
'°
l
g ©v
@,1 E~ Z
v
©
~°I, i ~ ~_ ~
i®
o
r~
©
0
[41]
BACITRACIN SYNTHETASE
553
the clear solution is then applied to the charcoal filter disk. Disks are held either on pins or laid flat on a bed of pins and numbered. The filter paper disks are allowed to dry at room temperature. For washing the disks, any of the techniques used in filter paper disk methods can be used. However, we have found the following procedure to be satisfactory: filter paper disks are placed in a large petri dish separated from each other and a cold solution of 40 mM sodium pyrophosphate in 1.5% perchloric acid is added. Disks are submerged in the solution and, with the help of forceps, kept from settling to the bottom and adhering to each other. After about 5 rain, the wash solution is withdrawn and replaced with a fresh cold solution. This is followed by a third wash. Disks are now washed in the same manner with cold distilled water. Finally individual disks are held with forceps, washed in a stream of distilled water and placed on a bed of pins for drying. After partial air drying (30-60 rain), the disks are completely dried for 10 rain at 110% After cooling, the disks are placed under 5 ml of 5% PPO in toluene in scintillation counter vials, and the radioactivity is measured in a liquid scintillation counter tLS 150, Beckman Inc., Fullerton, California) using the channel range of 14C + 32p. 3. Millipore Filter Test. The synthesis of bacitracin is measured by incorporation of radioactive amino acids into baeitracin and its retention hy selectron filters (0.45 t~m, 24 mm diameter, Sehleicher & Sehuell, Dassel, Germany). The incubation mixture of 0.1 ml, pH 7.3, contains 8.0 mM ATP; 10 mM cysteine; 3 mM each of the other 9 amino acids (all L-form) ; 0.16 M Tris.HC1 buffer and 50 nCi [U-~C]protein hydrolyzate. Enzyme solution in buffer B (0.] ml) is added in the cold and incubated 10 rain at 4 ° and 25 ° for the stated periods. Blanks are prepared as above, but are frozen immediately. Blanks and the incubated mixtures are filtered through Selectron filters in the cold, washed with huffer B (2 }( 1 ml}, and once with water (1 ml). The filter disks are dried at 50 ° for 20 rain and counted in 5 ml of 57; PPO in toluene in a Beckman scintillation counter.
4. Radio Thin-Layer Chromatographg of Incorporated [HC ]Lgsi~w. The incubated mixture from the Millipore filter test whieh contains 550,000 cpm of [~4C]lysine instead of 1~C label protein hydrotysate, is ehromatographed on TLC-plates silica gel F~.~ (Merck) using r~-butanol:acetie acid:water (4:1:2, v/v) as the developer. After development, the areas corresponding to haeitraeins A and F (Rr 0.43 and 0.46) are measured and integrated by a windowless thin-layer scanner (Duennsehicht-scanner II, Berthold, Wildbad, Germany) (Table I). The areas corresponding to baeitraeins are scraped off, extracted with methanol (2 X 1 ml) and then with water (2)4 0.5 ml). The pooled ex-
554
ANTIBIOTIC BIOSYNTHESIS
[41]
tract is dried under vacuum and lyophilized with a small amount of water to remove any trace of methanol. The residue is taken up in water, and the antibiotic activity is determined. As the antibiotic activity in these samples is usually low, the above incubation is also directly tested for growth inhibition after extraction with methanol. Other Methods
SDS Disc Gel Electrophoresis. Protein, 0.04 ml or 83 ~g, of fractions 18-21 (DEAE-cellulose run) and 0.04 ml or 160 ~g of protein of fractions 10-11 (Sephadex G-200 run) are each diluted with 0.02 ml of an electrode buffer solution from gel-system No. 612 (20% sucrose, 1% SDS, 0.1% ME). Lucite tubes of 0.5 cm diameter containing 7.5 cm of separation gel and 0.5 cm of spacer gel serve as carriers. Electrophoresis is allowed to proceed over a period of 4 hr with a current of 4 mA. The exposed gels are then developed with Amido black and finally decolorized with 7% CH3COOH (20% ethanol). Marker enzymes are LAP, catalase, and ADH. Disc Gel Electrophoresis. Portions, 0.02 ml of fractions 18-21 (DEAEcellulose run) and fractions 10-11 and 14 (Sephadex G-200 run) are diluted as described except that 1% SDS and 0.1% ME are omitted (Fig. 4). Protein Determination. Protein is determined according to the method of Folin-Ciocalteu.is Sucrose Density Gradient Centri]ugation. Protein, 0.1 ml or 0.4 mg, of fractions 10-11 (Sephadex G-200 run) are carefully layered on top of a sucrose density gradient (5-20% sucrose in buffer C), and centrifuged at 204,000 g and 3 ° for 3 hr. Marker enzymes are ADH, LAP, and urease. Two-drop fractions are collected by puncturing the bottom of the tube. All fractions are assayed for bacitracin synthesizing activity utilizing the ATP-PPi exchange method. Three active bands are found (Table II). Bands of marker enzymes are located by measuring the optical density of all fractions in a Zeiss M4QIIId manual UV spectrophotometer (Zeiss, Oberkochen, Germany) at 280 nm. Preparation o] Inoculum/or Fermentation. (A) Mutant 4 rough cells of Bacillus licheni]ormis are obtained from 3% tryptic soy broth slant cultures no older than 24-28 hr. (B) Tryptic soy broth (30 ml) in a 300-ml Erlenmeyer flask is inoculated with a large loop from a slant culture. The flask is then incubated at 36-37 ° for 12 hr and shaken at a rate of 180 strokes per minute with an amplitude of 3 cm. (C) Two 12H. R. Maurer, in "Disc Electrophoresis," pp. 44-45. de Gruyter, Berlin, 1971. l~j. R. Spies, this series, Vol. 3, pp. 467--468.
[41]
B&CITRACIN SYNTHETASE
555
TABLE II MOLECULAR WEIGHT DETERMINATION
Fraction
Molecular weight
DEAE-cellulose fraction 18-21
148,000
Sephadex G-200 fraction 10-I 1
148,000 SDS disc gel electro74,000 phoresis 431,000 Sucrose density gradient 327,000 centrifugation 64,700
Method SDS disc gel electrophoresis
ATP-a'PP~ exchange (cpm) --5;5 92 1092
2000-ml Erlenmeyer flasks containing 500 ml of 3% tryptic soy broth each, pH 6.0-6.2 after sterilization, are inoculated with 5 ml of culture from part B, incubated and shaken for 10 hr at 36-37 °. Fermentation. Fermentation takes place in a K 7 fermenter with 1000liter capacity. The fermentation media (1000 liters) contains 10 kg of Pharmamedia, 30 kg of Someel, 20 kg of starch, 5 kg of corn steep liquor, 5 kg of Bolus alba, 5 kg of CaCO~, 5 g of FeSO~.7 H.20, and 5 g of M n S Q . H 2 0 . The media is sterilized at 121 ° for 40 min. The pH values of the media before and after sterilization are 6.2 and 6.5-7.0, respectively. During fermentation, the media pH is determined every 2 hr (Fig. 2), and adjusted with NaOH or H2S04 when necessary. Fermentation is started by adding to the 1000 liters of sterile media 80-
7.5
T 70
6.51
0
I0
Hours
20
30
FIG. 2. Fermentation, 1000 liters, of Bacillu~ licheni]ormis, mutant 4, rough; pH profile.
556
ANTIBIOTIC BIOSYNTHESIS
[41]
the two 500-ml portions of inoculum prepared (see C, above) and is allowed to proceed at 34 ° for 33-34 hr. The fermentation mixture is stirred with a "Turbo-stirrer" at 180 rpm and aerated with an air flow of 0.25 liter/liter per minute for the first 4 hr, 0.5 liter/liter per minute for the next 4 hr, and 1.0 liter/liter per minute until fermentation is terminated. An antifoaming mixture of soy-oil/lard is added as necessary. Harvesting and Storage of Cells. Figure 2 shows the pH changes of the fermentation media from which the organism utilized in the present study was harvested. The arrow indicates harvesting time. Eight hours before harvesting the fermenter was slowly cooled and the air flow was reduced gradually to allow proper removal of 100 liters of cell suspension at 26 hr of fermentation. The cell suspension was centrifuged to yield 5 kg of sediment (wet weight, 70% H20). Cells were frozen in 100-g portions. Purification of Bacitracin Synthetase All procedures were carried out at 4 ° unless noted otherwise. Step 1. Cell Lysis. Rough mutant 4 cells (RC4) of B. licheniformis, 100 g wet weight, are thawed in 150 ml of 25 mM phosphate buffer, pH 7.5. After 2 hr of gentle stirring the cell suspension is centrifuged for 40 min at 27,000 g. T h e pellets are resuspended in 200 ml of 25 mM phosphate buffer, pH 7.5, containing 0.75 mM EDTA, 1 mM MgCI_~, and 10 mM mercaptoethanol. To the clay-colored suspension are then added 160 mg of lysozyme, 400 ~g of DNase, and 100 ~g of RNase. The gently stirred mixture is incubated at 30 ° for 20 min. After centrifugation at 27,000 g for 40 rain, the pellets are resuspended and treated once more with lysozyme, DNase, and RNase as described above. The turbid light yellow supernatant from both lysis cycles is combined and, under slow stirring, is brought to 45% saturation with crystalline ammonium sulfate. After standing for 2 hr, a minimal amount of precipitate is removed from the suspension by centrifugation at 27,000 g for 40 min. The supernatant is brought to 80% ammonium sulfate saturation in a similar fashion after removal of a thin surface layer of lipid material. Step 2. Sephadex G-50 Chromatography. The suspension from step 1 (saturated to 80% with ammonium sulfate) is centrifuged at 27,000 g for 40 rain. The sediment, containing approximately 120 mg of protein, is dissolved in 5 ml of buffer B and the resulting 8.5 ml of solution are eluted from a Sephadex G-50 column (2.2 X 32 cm) with buffer C. Fractions of 8 ml are collected using a Uvicord with UltroRac (LKB Produkter AB, Bromma, Sweden). Fractions 5, 6, and 7 show bacitracin synthesizing activity (Fig. 1, Table I).
[41]
BACITRACIN SYNTHETASE
557
Step 3. DEAE-Cellulose Chromatography. Fractions 5, 6, and 7 from step 2 are combined to give 18 ml of solution. This solution is then eluted from a DEAE-cellulose column (2 X 24 cm) with an exponential NaC1 gradient in buffer C. The gradient is achieved by running 300 ml of 0.5 M NaCt in buffer C into a mixing chamber containing a constant volume of buffer C (100 ml). Fractions of 8 ml are collected as in step 2. Fractions 14-]7 show bacitracin-synthesizing activity (Fig. 2, Table I). Step 4. Diafiltration and Concentration. Fractions 14-17 from step 3 (23 ml) are concentrated to 6 ml in a Multi-Micro Ultra,filtration System (Amicon, Oosterhout, The Netherlands) with 30 psi of N._,. Step 5. Sephadex G-200 Chromatography. The concentrate from step 4 (6 ml) is eluted with buffer C from a Sephadex G-200 column (2 X 61 cm) in fractions of 8 ml each. Fractions 8-11, 14 show bacitracin synthesizing activity (Fig. 3, Table I). •
-
,
.
,
A[3° 2.O ~.
c
2 ~ 50 o
i.o
I i
-':
"'" .......
'..''""
iO
B L04
:
S 7
~oI00 "'"I,
"k
,~. , 1001
:, '~
~1008 c
~
-
.02 "6
, ." , . , 10 20 30 8 ml F r a c t i o n s
Fro. 3. Column chromatography of bacitraein synthetase. % Transmittance ( ), milligrams of protein per milliliter (---). (A) Sephadex G-50. (B) DEAE-cellulose; 0-0.5 M NaC1 gradient ( . . . ) . (C) Sephadex G-200.
558
ANTIBIOTIC BIOSYNTHESIS
[41]
0-
5.0
cm I 2 3 FIG. 4. Disc gel electrophoresis of enzyme fractions. 1, DEAE-cellulose chromatography, fractions 18-21; 2, Sephadex G-200 chromatography, fractions 10W 11; 3, fraction 14, treated similarly to fraction 10 T 11 in 2.
The Millipore filter test, which measures the sum of specifically, at low temperature, acid stable (thioesters?) TM, and nonspecifically bound intermediates 'of the biosynthesis of bacitracin also shows that the rate of formation of bacitracin from these intermediates rises at 25 °. The 0 ° and 4 ° values of the growth inhibition test reflect only roughly the rise of specific activity of the enzyme containing material over the course of the three column chromatography steps. The growth inhibition test is necessary, however, because during radio thin-layer chromatography, which measures the sum of bacitracin A and F (Rf bacitracin A: 0.43; R~ bacitracin F: 0.46) at least 90% of bacitracin A is transformed into bacitracin F. Lysis and Ammonium Sul]ate Precipitation. Approximately 120 mg of bacitracin synthetase are recovered from a bulk of 15 g of bacterial protein (0.8%). Properties
ATP-82PPI Exchange, Millipore Filter Test, Radio Thin-Layer Chromatography, and Growth Inhibition Test (Table I). The results of ATP-82PPi exchange measurements of the three column chromatography steps clearly show two facts: (a) the purification factor for the very fast reaction. T M E + ATP + AA ~- AMP-AA-E + PPi
(2)
is of the order of 100 at 4 °. (b) While the rate of formation of s2P-labeled A T P rises throughout the three column chromatography steps at 4 °, this rate reaches its peak in step 2 of the assays at 25% 1, W. Gevers, H. Kleinkauf, and F. Lipmann, Proc. Nat. Acad. Sci. U.S: 60, 269 (1968).
[42]
EDEINE SYNTHETASE
559
A comparison with the amounts of bacitracin found in the radio thinlayer chromatography analyses suggests that the rate of formation of bacitracin from aminoacyl adenylates on the highly organized and purified multienzyme complex of column chromatography, step 3, has increased at 25 ° and has reached the same order of magnitude as the back reaction of reaction (2). SDS Disc Electrophoresis, Disc Electrophoresis, and Sucrose Density Gradier.t CentriJugation. The molecular weights as determined by SDS disc gel electrophoresis with a single sharp band of MW 148,000 for the DEAE-cellulose run (Fig. 3) and of MW 148,000 and 74,000 for the Sephadex G-200 run show clearly that bacitracin synthetase has a molecular weight of 431,000 (Table II, sucrose gradient centrifugation) and is composed of 7 subunits. The Sephadex G-200 fractions produce a more differentiated electrophoretic pattern because of constant association and dissociation of enzyme subunits and because of the comparatively low protein concentration. Accordingly, in Fig. 4, bar 2 (Sephadex G-200, fractions 10-11) and Fig. 4, bar 3 (Sephadex G-200, fraction 14) identical bands at 1.8 and 2.4 cm may be seen. The same bands are visible in Fig. 4, bar 1, in the still impure DEAE-cellulose fractions (see bands below 4.1 cm). In contrast, the band of highest molecular weight is seen only in bars 1 and 2 of Fig. 4. Molecular weight determination with a number of marker proteins (LAP, LDH, ADH, BSA, and SBTI) on the Sephadex G-200 column yielded a molecular weight of ~300,000 for fractions 8-9 (elution within exclusion volume).
Acknowledgments Thanks are due to Deutsche Forschungsgemeinschaftand to Fonds der Chemischen Industrie for grants, and to Lohmann & Co. AG for RC4 cells. The skillful assistance of Miss Annerose Bahnmfiller and of Miss Dorothee Specht is greatly appreciated.
[42] Edeine Synthetase By ZOFIA KURYLO-BoRoWSKA Edeines A and B are linear oligopeptides. 1 They are mixtures of biologically active compounds (edeine A1 and B1) (Fig. 1) and inactive isomers (edeine A~. and B2) in which isoserine is linked to the ~- rather T. P. Hettinger, Z. Kurylo-Borowska,and L. C. Craig, Ann. N.Y. Acad. Sci. 170, 1002 (1970).
[42]
EDEINE SYNTHETASE
559
A comparison with the amounts of bacitracin found in the radio thinlayer chromatography analyses suggests that the rate of formation of bacitracin from aminoacyl adenylates on the highly organized and purified multienzyme complex of column chromatography, step 3, has increased at 25 ° and has reached the same order of magnitude as the back reaction of reaction (2). SDS Disc Electrophoresis, Disc Electrophoresis, and Sucrose Density Gradier.t CentriJugation. The molecular weights as determined by SDS disc gel electrophoresis with a single sharp band of MW 148,000 for the DEAE-cellulose run (Fig. 3) and of MW 148,000 and 74,000 for the Sephadex G-200 run show clearly that bacitracin synthetase has a molecular weight of 431,000 (Table II, sucrose gradient centrifugation) and is composed of 7 subunits. The Sephadex G-200 fractions produce a more differentiated electrophoretic pattern because of constant association and dissociation of enzyme subunits and because of the comparatively low protein concentration. Accordingly, in Fig. 4, bar 2 (Sephadex G-200, fractions 10-11) and Fig. 4, bar 3 (Sephadex G-200, fraction 14) identical bands at 1.8 and 2.4 cm may be seen. The same bands are visible in Fig. 4, bar 1, in the still impure DEAE-cellulose fractions (see bands below 4.1 cm). In contrast, the band of highest molecular weight is seen only in bars 1 and 2 of Fig. 4. Molecular weight determination with a number of marker proteins (LAP, LDH, ADH, BSA, and SBTI) on the Sephadex G-200 column yielded a molecular weight of ~300,000 for fractions 8-9 (elution within exclusion volume).
Acknowledgments Thanks are due to Deutsche Forschungsgemeinschaftand to Fonds der Chemischen Industrie for grants, and to Lohmann & Co. AG for RC4 cells. The skillful assistance of Miss Annerose Bahnmfiller and of Miss Dorothee Specht is greatly appreciated.
[42] Edeine Synthetase By ZOFIA KURYLO-BoRoWSKA Edeines A and B are linear oligopeptides. 1 They are mixtures of biologically active compounds (edeine A1 and B1) (Fig. 1) and inactive isomers (edeine A~. and B2) in which isoserine is linked to the ~- rather T. P. Hettinger, Z. Kurylo-Borowska,and L. C. Craig, Ann. N.Y. Acad. Sci. 170, 1002 (1970).
560
ANTIBIOTIC BIOSYNTHESIS . . . . . . . . . . .
r- . . . .
p GUANYL) SPERMIDINE I I
/ r "/
r / N H ~ N H
OH
~
I
J
I r HO/~/NH2 ,____L--__~__J I I
J
I
I NH2 0
i
1
I I0
r~ I~
~ N H R TI L--T ....... 0
~ . . . . . .
GLY
r~,/~NH-
/ /
[42l
I
I
I r i
J
~
,8- TYR / ISER ~ . . . . . . . . . . . . . . . . . . .
EDEINE
or DAPA
AA
/ I L .........
A,
R=H
B,
R = C ( = N H ) NH2
FIG. 1. Structure of edeine A1 and B~. GLY, glycine; DAPA, ~,fl-diaminopropionic acid; DAHAA, 2,6-diamino-7-hydroxyazelaic acid ; ISER, isoserine ; #-TYR, #-tyrosine. than to the a-amino group of a, fl-diaminopropionic acid. 2 The biosynthesis of edeines resembles that of tyrocidines2 It is accomplished by two complexes of soluble enzymes of Bacillus brevis Vm4. Properties of the enzymes are reminiscent of those participating in fatty acid synthesis2 The formation of aminoacyl-AMP of constituent amino acids of edeines is catalyzed by two polyenzymes: I, M W 210,000; II, M W 180,000. Prior to the polymerization the activated amino acids are bound to the sulfhydryl group of 4'-phosphopantetheine, which is covalently bound to protein of both polyenzymes. Activation of fl-tyrosine is catalyzed by polyenzyme I, whereas activation of isoserine, a,fl-diaminopropionic acid, 2,6-diamino-7-hydroxyazelaic acid, and glycine involves polyenzyme II. The presence of both polyenzymes and spermidine are required for the formation of edeines. The mechanism of addition of spermidine to the peptide chain is not known; however, its presence is sufficient for the formation of edeine A as well as edeine B2
Assay
Methods
Activation reaction R. CH. COOH 4- ATP [
NH2
Mg2 + polyenzyme
, R. CH. CO. AMP 4- PPI
I or II
I
(1)
NH2
2 T. P. Hettinger and L. C. Craig, Biochemistry 9, 1224 (1970). 3S. G. Lee, R. Roskowski, Jr., K. Bauer, and F. Lipmann, Biochemistry 12, 398 (1973). 4F. Lynen, Biochem. J. 102, 381 (1967). 5Z. Kurylo-Borowska and J. Sedkowska, Biochim. Biophys. Acta 351, 42 (1974).
[42l
EDEINE SYNTHETASE
561
Binding reaction polyenzymeI R. CH. CO. AMP + , R- CH. CO--S--polyenzyme + AMP t or polyenzymeII [ NH~ NH~
(2)
Polymerization reaction R. CH. CO--S--polyenzyme I -I- 2?(R. CH. CO)--S--polyenzyme II r I NH2 NH~ -I- spermidine--~ edeine A +edeine B + HS-enzymes (3) Reaction (1), conducted with fl-tyrosine and polyenzyme I, or with isoserine, a,fl-diaminopropionic acid, 2,6-diamino-7-hydroxyazelaic acid, glycine, and polyenzyme II is determined by the ATP-3zPPi exchange. If experiments are conducted with polyenzyme II purified by DEAEcellulose chromatography only, it is advisable to measure the ATP-32PPi exchange dependent on isoserine, DAP, and DAHAA rather than glycine, since the contamination of this fraction with t R N A g'y and a corresponding ligase can occur. Reaction (2): binding of the amino acyl to the enzyme complex can be measured only if radioactive edeine constituent amino acids are available. Reaction (3): formation of edeine A and B is measured by incorporation of radioactive glycine and spermidine into these two compounds or microbiological assay. Reagents
Amino acids a,fl-Diaminopropionic acid, 10 m M (Calbiochem) Isoserine, 10 m M (Nutritional Biochemicals Corp.) 2,6-Diamino-7-hydroxyazelaic acid, 10 m M (obtained by hydrolysis of edeines ~) Glycine, 10 m M fl-tyrosine, 10 m M (obtained by hydrolysis of edeines ~ or synthetically 7) Reagents for the measurement of the activation reaction Tris.HC1 buffer, 1 M, pH 8.0 Mg acetate, 1 M EDTA, 0.1 M DTT, 0.1 M NaF, 0.1 M G. Roncari, Z. Kurylo-Borowska, and L. C. Craig, Biochemistry 5, 2153 (1966). 'V. M. Rodinov, A. A. Dudinskaya, V. G. Avramenko, and N. N. Suvorov, J. Ge,. Chem. USSR 28, 2279 (1958).
562
ANTIBIOTIC BIOSYNTHESIS
[42]
Na4P207, 0.1 M Na4~2P207 Bovine serum albumin (10 mg/ml) Polyenzyme I or polyenzyme II, 0.2-1.0 mg protein/ml in a solution of 0.1 M Tris, pH 7.6, 2 mM D T T Amino acid solutions: 10 mM Reagents for the biosynthesis of edeine A and B Tris buffer, 1 M, pH 7.9 Mg acetate, 1 M KC1, 1 M DTT, 0.1 M ATP, 0.1 M Phosphoenolpyruvate, 0.2 M Phosphoenolpyruvate kinase, 10 mg/ml Spermidine, 10 mM Amino acids, 10 mM [U-14C] Glycine [U-I'C] Spermidine Solution of polyenzyme I and II or of fraction A, fraction B, and fraction C. Procedure Activation o] Amino Acids. Standard reaction mixture consists of 20 ~l of 1 M Tris.HC1 buffer, pH 8.0, 1 ~l of 1 M Mg acetate, 2.5 ~l of 0.1 M DTT, 7 /~1 of 10 mM EDTA, 5 ~l of 0.1 M ATP, and 2 ~l of 1% bovine serum albumin. To this mixture add 50-150 ~l of enzymes solution (fraction I or II), 2.5 ~l of 0.1 M NaF, 5 ~l of 0.1 M Na~P207, 100,000 cpm of Na43~P20~, 25 ~1 of 10 mM amino acid and water to a total volume of 0.25 ml. The reaction is carried at 35 ° for 30 min, then terminated by addition of 100 ~l of charcoal suspension in a solution of Na4P20T and perchloric acid. s Charcoal is collected on glass fiber disks (Whatman GFA, 2.4 cm) and washed with 30 ml of water. Filters are dried, and the radioactivity of adsorbed [3-~p]ATP is measured in a liquid scintillation counter. Biosynthesis o] Edeine A and Edeine B. The reaction mixture consists of: 80 ~l of 1 M Tris buffer, pH 7.9, 20 ~l of 0.1 M DTT, 10 ~l of 1 M KC1, 10 ~l of 1 M Mg acetate, 2.5 ~l of 0.1 M ATP, 5 ~l of 0.2 M phosphoenolpyruvate, 5 ~l of (10 mg/ml) phosphoenolpyruvate kinase, 25 ~l of 0.1 M ATP, 10 ~l of each 0.1 M edeine constituent amino acid, 8 R. Calendar and p. Berg, "Procedures in Nucleic Acid Research" (G. L. Cantoni and D. R. Davies, ed.), p. 384. Harper, New York, 1966.
[42]
EDEINE SYNTHETASE
563
SUMMARY OF THE PURIFICATION OF EDEINE BIOSYNTHETIC ENZYMES a
Stage of purification
Protein (mg)
Specific activity Purifica(ATP nmoles/mg tion protein) (fold)
A. ¢~-TyrosineActivating Enzymes 1. 30,000 g supernatant fraction 2. Ammonium sulfate precipitate (30-55 % saturation) 3. DEAE-cellulose column Fraction I Fraction II 4. Sephadex G-200 of fraction I: fraction A, MW 210,000 Sephadex G-200 of fraction II: fraction A, MW 210,000
920 300
4.0 10.0
2.5
25 18 2.5
75.0 26.0 280.0
18.7 6.5 70.0
2.0
85.0
21.2
B. Diaminopropionic Acid Activating Enzymes 1. 30,000 g supernatant fraction 2. Ammonium sulfate precipitate (30-55% saturation) 3. DEAE-cellulose column Fraction I Fraction II 4. Sephadex G-200 of fraction I Fraction B, MW 180,000 Fraction C, MW 100,000 Sephadex G-200 of fraction II Fraction B, MW 180,000 Fraction C, MW 100,000
920 300
6.2 18.0
2.8
25 18
80.0 160.0
13.2 25.8
2.0 2.0
250.0 360.0
40.0 60.0
3.0 3.0
360.0 150.0
60.0 24.2
Specific activity of the enzymes was calculated from the ATP-~2PPI exchange dependent on a,f~-diaminopropionic acid or ¢~-tyrosine. One nanomole of ATP corresponded to 200 cpm. including [U-14C]glycine (25,000 cpm/tLmole), and 10 tL1 of 0.1 M [U-14C]spermidine (25,000 cpm/~mole). Equal amounts of polyenzymes I and I I are added (50-100 ~g of each), and the volume is adjusted with water to 1 ml. If polyenzymes I and I I are purified further by Sephadcx G-200 filtration (see the table) equal amounts of fractions A, B, and C are used. After incubation at 35 ° for 30-min, samples are acidified to p H 5.5 with 1 M acetic acid. The precipitate is removed by centrifugation. The resulting supernatant solution is neutralized with 1 N N a O H and adsorbed to Dowex 50-X4 H +, on a column of 2 cm X 1 cm. The resin is washed successively with 20 ml of 0.5 N ammonium formate buffer pH 7, 20 ml of water, and 5 ml of 1 N NH4OH. The fraction
564
ANTIBIOTIC BIOSYNTHESIS
[42]
eluted by the NH40H is concentrated under vacuum to 0.4 ml. Aliquots of 100 ~l are chromatographed on Whatman 3 MM paper With isopropanol-NH~OH-H20 (4: l : l , v/v) in the presence and in the absence of standards of edeine A and edeine B. Chromatograms are dried, and the parts which were eochromatographed with standards of edeines are stained with ninhydrin. These spots are excised and counted. The parts of chromatograms developed without standards are cut into strips and placed on agar plates, inoculated with B. subtilis 168. After 3 hr at 4 ° strips are removed, and plates are incubated for 10 hr at 37 °. The zones of inhibition corresponding to the position of edeine A and B are measured and compared with those of standard edeine solutions chromatographed under the same conditions.
Purification
Preparation of Crude enzymes. The enzymes are obtained from Bacillus brevis Vm4. The organism is grown in Bactopeptone-yeast extract medium at 30 ° with shaking2 The cells are collected by centrifugation from 10 liters of a 10-12-hr-old culture. After washing 3 times with 600 ml (each time) of cold 0.1 M Tris.HC1 buffer pH 7.6 in 2 mM dithiothreitol (buffer A) the pellet (about 120 g wet weight) is suspended in 60 ml of buffer pH 7.2 consisting of morpholinopropanesulfonic acid and 2.5 mM EDTA. The cells are lysed by incubating the suspension with 36 mg of lysozyme 'and 1.2 mg of DNase at 30 ° for 20 rain and then at 0 ° for 20 min. The lysate of cells is centrifuged 90 min at 30,000 rpm at 4 °. All further operations are conducted at 4 °, unless stated otherwise. To the supernatant fraction solid ammonium sulfate is added to 30% saturation. After 1 hr the precipitate is removed by centrifugation and the concentration of ammonium sulfate of the supernatant fraction is adjusted to 55%. This is kept 1 hr at 0 ° and afterward the precipitate is collected by centrifugation. The precipitate is dissolved in a small volume of 1 M Tris.HCl buffer, pH 7.6, and dialyzed 5 hr against 2 liters of buffer A, changing the buffer twice. DEAE-Cellulose Chromatography. The dialyzed crude enzymes (about 300 rag) are applied to the column of DEAE-cellulose (2.2 X 25 cm) equilibrated with buffer A. The column is washed successively with 500 ml of buffer A and 250 ml (each): 0.1 M, 0.2 M, and 0.3 M KC1 in buffer A. Fractions of 6 ml (flow rate 40 ml/hr) are collected. Proteins and the ATP-3-~PPi exchange dependent on edeine constituent amino acids are measured in aliquots of fractions. Polyenzyme I elutes with 9Z. Kurylo-Borowskaand E. L. Tatum, Biochim. Biophys. Acta 113, 206 (1966).
[42]
EDEINE SYNTHETASE
565
0.3M K(
O.IM KCI
50
I I
1.5 30 ~
~6
L a E
1.0
"5 o .o
1
Z
3
~
Growth,
5
5
?
S
hours
FIG. 1. Relationship between growth curve and gramicidin S synthesizing activity of crude extracts. O O, Absorbance at 650 nm; • O, synthesis of gramicidin S. Reproduced from S. G. Laland and T.-L. Zimmer, Essays Biochem. 9, 31 (1973). containing 60 ml of the tryptone yeast medium described in the previous section. The flasks are incubated with aeration at 37 ° overnight and the cultures transferred to 10 liters of medium in the Microferm L a b o r a t o r y Fermenter (New Brunswick Scientific Co.). The medium contains per 10 liters: 136 g of KH2P04, 20 g of (NH4)~S04, 100 mg of CaC12.2H20, 2 g of MgSO~.7 H20, 5 mg of FeSO4.TH~O, 50 g of monosodium glutamate, and 1.5 g of each of the amino acids in gramicidin S. The p H is adjusted to 6.5 with K O H . The culture is aerated at a rate of 5 liters per minute at a pressure of 10 psi with stirring at 300 rpm. The culture is harvested after about 5 hr at an absorbance of 0.27 at 650 nm in a Spectronic 20 spectrophotometer after dilution of the culture 9 times with water. The culture is cooled in the fermentor to about 15 °, and the cells are then harvested by centrifugation at 0% All further procedures are carried out at 2-4 °. The cells are washed once with a solution of 5 m M MgC12. The washed cells are either processed at once or m a y be stored at - - 2 0 ° without any significant loss of enzyme activity. Figure 1 shows the relationship between growth curve and gramicidin S synthesizing activity of the crude extract (see next section).
[43]
GRAMICIDIN S SYNTHETASE
573
Step 2. Preparation o] Crude Extract. The cells are suspended in 300 ml of a solution of 5 mM MgCI~ containing per milliliter 10 t~moles of reduced glutathione and 0.25 t~mole of EDTA, with pH adjusted to 7.5 and disintegrated by ultrasonic treatment with a Branson Sonifier Cell Disruptor S 75 at 75 W and 20 KHz for 7 rain under cooling. The resulting suspension is centrifuged for 15 min at 3000 g and then 15 rain at 20,000 q. The supernatant after the last centrifugation is called the crude extract and contains approximately 2.5 g of protein. It may be stored at --20 ° without loss of activity. Step 8. Streptomycin Sul]ate Precipitation. To the crude extract is added dropwise under stirring a neutral solution of streptomycin sulfate, 50 mg/ml, to a final concentration of 5 mg/ml. The mixture is allowed to stand for 30 rain before it is centrifuged for 15 rain at 20,000 g. The precipitate is discarded. Step ~. Ammonium Sul]ate Precipitation. Solid (NH~).,SQ is added (final concentration 0.29 g/ml). After standing for 30 rain, the mixture is centrifuged for 15 rain at 20,000 g, the supernatant discarded and the precipitate dissolved in 50 ml of buffer A [50 mM KH~PO~, 0.25 mM EDTA, 1 mM DTT, 20% glycerol (v/v), pH 7.5]. The preparation contains approximately 10 mg of protein per milliliter and is stable at --20 °. Step 5. Chromatography on DEAE Sephadex A-50. The ammonium sulfate fraction from step 4 is applied on a DEAE Sephadex A-50 column (2.5 X 40 cm) previously equilibrated with buffer A containing 105, glycerol (instead of 205,) and 0.2 M KC1; 60 ml of the equilibrating buffer is then run through the column. The concentration of KC1 in the eluent is then increased linearly from 0.2 to 0.5 M. The light and the heavy enzyme are partially separated by this procedure (Fig. 2) in contrast to what was achieved with a former procedure where a linear gradient of phosphate buffer was used. ~° The fractions rich in the light and the heavy enzyme, respectively, were combined and precipitated by ad(lition of ammonium sulfate to 60% saturation. The precipitates were then dissolved in 5-10 ml of buffer A. Step 6. Chromatography on Sephadex G-200. Of the protein from the fraction rich in the heavy enzyme from step 5, 7.7 mg are apt)lied to a column of Sephadex G-200 (2.5 X 30 cm) equilibrated with buffer A. A separation of the light and heavy enzyme is achieved (Fig. 3) upon elution with buffer A. Fractions 13-18 are combined and contain essentially pure heavy enzyme. Similarly, light enzyme free of the heavy enzyme may be obtained by Sephadex G-200 chromatography of the light enzyme-rich fraction from step 5. The fractions containing the light and ~oj. E. Bredesen, T.-L. Berg, K. J. Figenschou, L. O. Fr0holm, and S. G. Laland, Ear. J. Biochem. 5, 433 (1968).
574
[43]
ANTIBIOTIC B I O S Y N T H E S I S 2.0
8
1.5
0
0.5
1.0 o
'>
< a: O.g 2
L~ 10
40
Fraction
60
80
100
120
{12
number
Fro. 2. Chromatography of ammonium sulfate fraction on DEAE-Sephadex A-50. Of the protein from step 4, 450 mg was applied to the column (2.5 X 40 cm). Fractions of 4.5 ml were collected upon elution with a linear 0.2-0.5 M KC1 gradient in buffer A containing 10% glycerol at a rate of 25 ml/hr. - - , Absorbance at 280 n m ; • • , D-phenylalanine stimulated ATP-~PP~ exchange; 0 - - 0 , L-ornithine stimulated ATP-'PP~ exchange. 1600
120o
E
e
>
?, "a o
0.3 400
~"
02
0.1 '~
lO
20 Fraction
30
1,0
number
FIG. 3. Chromatography on Sephadex G-200 of the fraction rich in heavy enzyme from step 5. Protein, 7.7 mg, was applied to the column (2.5 X30 cm) and eluted with buffer A. Fractions of 3 ml were collected at a rate of 10 ml/hr. - - , Absorbance at 280 rim; • 0 , D-phenylalanine stimulated ATP-~PPI exchange; O O, t-ornithine stimulated ATP-~PPI exchange.
[43l
GRAMICIDIN S SYNTHETASE
575
7 E
o. u
._> O o ¢~
] J,\ 0
10
~
30
Fraction number
FIG. 4. Affinity chromatography of gramicidin S synthetase. Protein, 0.5 mg in 3 ml from step 5, to which was added 7.5 #moles of ATP, was applied to a column (1 X 8 cm) of 3,'3-diaminodipropylamine-substituted Sepharose with L-proline as ligand. The column was eluted with buffer A at a rate of 7 ml/hr, and 2.5-ml fractions were collected. At the points shown by arrows 0.1 M and 0.5 M KCI, respectively, were added to the elution buffer. • O, D-phenylalanine stimulated ATP-32PPI exchange; 0 O, L-ornithine stimulated ATP22PPt exchange.
the heavy enzyme, respectively, are concentrated against buffer A using Sartorius collodion bags to concentrations of approximately 1 mg/ml. The light enzyme has been purified by Yamada and Kurahashi ~ to a nearly homogeneous state by a different purification procedure.
Step 7. Alternative Method ]or the Separation o] Light and Heavy Enzyme: Affinity Chromatography. 3,3'-Diaminodipropylamine substituted Sepharose to which one of the amino acids in gramicidin S is coupled through its carboxyl group, has been used to study the affinity between the light and the heavy enzyme of gramicidin S synthetase. 11 The column with proline (1 ffmole of proline per milliliter of Sepharose) attached provides a rapid and convenient method for the separation of the two enzymes (Fig. 4). It should be noted that when a DEAE-Sephadex preparation rich in light enzyme is used, a substantial amount of light enzyme is not absorbed and is consequently eluted before the addition of KCl. " L. Pass, T.-L. Zimmer, and S. G. Laland, Eur. J. Biochem. 40, 43 (1973).
576
ANTIBIOTIC BIOSYNTHESIS
[43]
TABLE I GRAMICIDIN S SYNTHESIZING ACTIVITY AND ATP-a2PPI EXCHANGE ACTIVITY DURING PURIFICATION OF GRAMICIDIN S SYNTHETASE
Fraction
Protein (mg)
Gramicidin S synthesizing activity (%)
Crude extract Ammonium sulfate fraction DEAE Sephadex A-50 Sephadex G 200
3000 500 25 20
100" 90 30 3
ATP-3~PPi exchange activity (%) --100 80
Under the conditions used for measuring gramicidin S synthesis, the crude extract from 10 liters of Bacillus breviswill incorporate about 10 to 15 × 106 cpln into gramicidin S.
Comments on the Purification of Gramicidin Synthetase. The biosynthesis of gramicidin S is the result of the concomitant functioning of a large number of catalytic activities. For instance, in the case of the heavy enzyme it has been estimated that about 18-20 different catalytic activities are involved. 1~ During the purification of gramicidin S synthetase which gives an almost homogeneous preparation of heavy enzyme, some 97% of its ability to synthesize the antibiotic is lost (Table I). The loss is particularly great during fractionation on the D E A E Sephadex G-50 and Sephadex G-200. Of the many catalytic activities involved in the biosynthesis, the catalytic activities responsible for amino acid activation are remarkably stable. As the crude extract and the ammonium sulfate fraction are unsuitable for determining the absolute levels of specific amino acid stimulated activation due to the high level of background ATP-32PPi exchange, the degree of purification of the synthetase cannot be determined by this method.
Properties of Gramicidin S Synthetase Purity o] the Enzymes. The heavy enzyme is found to be essentially pure as determined by polyacrylamide gel electrophoresis. The same technique reveals that the light enzyme preparation (by our procedure) is contaminated to a degree of about 10% by other proteins. 1~S. G. Laland, ~. Fr0yshov, C. Gilhuus-Moe, and T.-L. Zimmer, Nature (London) New Biol. 239, 43 (1972).
[43]
GRAMICIDIN S SYNTHETASE
577
TABLE II PHYSICAL PROPERTIES OF GRAMICIDIN S SYNTHETASI~ Properties MW s2o,w (S)
Heavy enzyme 240,000 280,000 b 11.6
Light enzyme 100,00() 6.1"
H. Yamada and K. Kurahashi, J. Biochem. 66, 529 (1969). b It. Kleinkauf, W. Gevers, and F. Lipmann, Proc. Nat. Acad. Sci. U.S. 62, 226 (1969).
Stability. The purified enzyme preparation may be stored at --20 ° for several months without loss of amino activating as well as gramicidin S synthesizing activities. Physical Properties. Our recent results have shown the molecular weight and the sedimentation Constant to be 240,000 and 11.6 S, respeelively as determined by sucrose gradient centrifugation (Table II). Polyacrylamide disc electrophoresis confirms this molecular weight. Presence o] 4'-Phosphopantetheine. Treatment of the essentially pure heavy enzyme with 0.2 M KOH at 37 ° for 1 hr 1~ liberates 1 mole of 4'-phosplIopantetheine from 1 mole of enzyme1'-' as determined by a microbiological assay using Lactobhcillus helveticus 80 (ATCC 12046)." 4'-Phosphopantetheine is believed to play an essential role in the stepwise polymerization process leading to gramieidin S. ~ Specificity. (a) AMINO ACID ACTIVATION. (i). The light enzyme. The light enzyme will activate both the L- and the D-isomers of phenylalanine, the former giving the highest rate of reaction. Phenylalanine analogs such as p-fluorophenylalanine, fl-2-thienylserine, and fl-phenylserine stimulate the ATP-:t~PPi exchange reaction, '~ whereas tyrosine and tryptophan as well as phenylpyruvic acid do not. (ii). The heavy enzyme. The following amino acids have been found in our laboratory to stimulate the ATP-:'~PP~ exchange reaction: azetidine-2-carboxylic acid, isoleucine, norvaline, norleucine 'rod D-leucine. Lysine and arginine do not stimulate the reaction. The D-isomers of proline, valine, and ornithine are not activated. '~ (t)) SYNTHESIS Of GRAMICIDINS. The following amino acids may be incorporated into gramicidin S: p-fluorophenylalanine, azetidine-2-carboxylic acid, norvaline, norleucine, and isoleucine. ~P. W. Majerus, A. W. Alberts, and P, R Vagelos, Proc. Nat. Acad. Sci. 1;.S. 53, 410 (1965). ~J. A. Craig and E. E. Snell, J. B~wteriol. 61, 283 (1951). '~ D. C. Leung and R. Baxter, Biochim. Biophgs. Acta 279, 34 (1972).
578
[43]
ANTIBIOTIC BIOSYNTHESIS TABLE I I I KINETIC CONSTANTS OF GRAMICIDIN S SYNTHETASE Amino acid
K ~ for amino acid (raM)
L-Phe D-Phe L-Pro I,-Val I,-0rn L-Leu
0.0450. 042" 0.23 0.31 0.17 0.10
K m for A T P (raM)
0.71, 0.711.40 2.33 1.72 3.10
Ki for A M P a (raM)
1.2 0.08 0.36 0.54 0.38 0.42
a T. Kristensen, C. C. Gilhuus-Moe, T.-L. Zimmer, and S. G. Laland, Eur. J. Biochem. 84, 548 (1973).
Inhibitors o] Gramicidin S Synthesis. The heavy enzyme contains several SH-groups essential for activity. Reagents known to react with sulfhydryl groups are strong inhibitors of gramicidin S synthesis. Amino acid derivatives which show affinity for the activation sites, such as the amino acid amides, are also inhibitory. The D-isomer of leucine which is activated and becomes thio ester bound to the heavy enzyme ~6 cannot be incorporated into gramicidin S and therefore inhibits total synthesis. Ammonium ions inhibit gramicidin S synthesis presumably by reacting with the activated amino acids to form the corresponding amides. 17 Kinetic Constants. The Km values for the ATP-S-~PPi exchange reaction with respect to ATP and the individual amino acids, as well as the Ki for AMP, have been determined (Table III).ls Characteristics o] the Light Enzyme. The light enzyme catalyzes an ATP-dependent racemization of phenylalanine. The enzyme does not require pyridoxal phosphate or FAD for activity and is not influenced by known inhibitors of pyridoxal-containing enzymes such as hydroxylamine, KCN, or isonicotinic hydrazide. 9 The results seem to indicate that racemization takes place after the amino acid has become thioester-bound to the enzyme 19 ATP L-Phe ~ L - P h e - A M P PPi
,~b-Phe-S
enzyme ~ D-Phe-S enzyme
AMP
~H. Saxholm, T.-L. Zimmer, and S. G. Laland, Eur. J. Biochem. 30, 138 (1972). ~7S. Tomino, M. Yamada, H. Itoh, and K. Kurahashi, Biochemistry 8, 2552 (1967). 1ST. Kristensen, C. C. Gilhuus-Moe, T.-L. Zimmer, and S. G. Laland, Eur. J. Biochem. 34, 548 (1973). 19H. Takahashi, E. Sato, and K. Kurahashi, J. Biochem. 69, 973 (1971).
[44l
POLYMYXIN SYNTHETASE
579
The biosynthesis of tyrocidine, a related cyclic decapeptide produced by a different strain of B. brevis, is also initiated by phenylalanine in a similar reaction. It has been found in this laboratory and by others 2° that the light enzyme of gramicidin S synthetase may substitute for the light enzyme of tyrocidine synthetase and vice versa in antibiotic formation. ~°K. Fujikawa, Y. Sakamoto, and K. Kurahashi, Y. Biochem. 69, 869 (1971).
[44] P o l y m y x i n S y n t h e t a s e : L - 2 , 4 - D i a m i n o b u t y r a t e Activating Enzyme By HENRY PAULUS
L-2,4-Diaminobutyrate -t- ATP ~- i-2,4-diaminobutyryl-AMP + PPI The sequence of reactions leading to the biosynthesis of polymyxin has not yet been elucidated. However, the following lines of evidence implicate an enzyme that activates L-2,4-diaminobutyrate in the biosynthetic process: (1) The level of the L-2,4-diaminobutyrate activating enzyme parallels the rate of polymyxin B production during the growth cycle of Bacillus polymyxa Pfizer 24591; (2) the L-2,4-diaminobutyrate activating enzyme is also found in B. polymyxa ATCC 10401, which produces polymyxin D, and in B. circulans ATCC 14040, which produces circulin, while it is absent from mutants of B. polymyxa Pfizer 2459 which have lost the ability to produce polymyxin B1; and (3) the enzyme is highly specific for L-2,4-diaminobutyrate which constitutes 5 or 6 of the 10 amino acid residues of polymyxin and circulin (Fig. 1) but is absent from other cellular constituents of B. polymyxa. 1,2 Assay Method 1
The assay is based on that of the aminoacyl RNA synthetases 3 and relies on the L-2,4-diaminobutyrate-dependent exchange of a2PPi with ATP, measured by the formation of charcoal-adsorbable radioactivity. In order to avoid interference by amino acid acyl RNA synthetases, the L-2,4diaminobutyrate should be free of contaminating amino acids, and crude bacterial extracts should be freshly dialyzed. 1K. Jayaraman, J. Monreal, and H. Paulus, Biochim. Biophys. Acta 185, 447 (1969). 2M. Brenner, E. Gray, and H. Paulus, Biochim. Biophys. Acta 90, 401 (1964). F. H. Bergmann, this series, Vol. 5, p. 708.
[44l
POLYMYXIN SYNTHETASE
579
The biosynthesis of tyrocidine, a related cyclic decapeptide produced by a different strain of B. brevis, is also initiated by phenylalanine in a similar reaction. It has been found in this laboratory and by others 2° that the light enzyme of gramicidin S synthetase may substitute for the light enzyme of tyrocidine synthetase and vice versa in antibiotic formation. ~°K. Fujikawa, Y. Sakamoto, and K. Kurahashi, Y. Biochem. 69, 869 (1971).
[44] P o l y m y x i n S y n t h e t a s e : L - 2 , 4 - D i a m i n o b u t y r a t e Activating Enzyme By HENRY PAULUS
L-2,4-Diaminobutyrate -t- ATP ~- i-2,4-diaminobutyryl-AMP + PPI The sequence of reactions leading to the biosynthesis of polymyxin has not yet been elucidated. However, the following lines of evidence implicate an enzyme that activates L-2,4-diaminobutyrate in the biosynthetic process: (1) The level of the L-2,4-diaminobutyrate activating enzyme parallels the rate of polymyxin B production during the growth cycle of Bacillus polymyxa Pfizer 24591; (2) the L-2,4-diaminobutyrate activating enzyme is also found in B. polymyxa ATCC 10401, which produces polymyxin D, and in B. circulans ATCC 14040, which produces circulin, while it is absent from mutants of B. polymyxa Pfizer 2459 which have lost the ability to produce polymyxin B1; and (3) the enzyme is highly specific for L-2,4-diaminobutyrate which constitutes 5 or 6 of the 10 amino acid residues of polymyxin and circulin (Fig. 1) but is absent from other cellular constituents of B. polymyxa. 1,2 Assay Method 1
The assay is based on that of the aminoacyl RNA synthetases 3 and relies on the L-2,4-diaminobutyrate-dependent exchange of a2PPi with ATP, measured by the formation of charcoal-adsorbable radioactivity. In order to avoid interference by amino acid acyl RNA synthetases, the L-2,4diaminobutyrate should be free of contaminating amino acids, and crude bacterial extracts should be freshly dialyzed. 1K. Jayaraman, J. Monreal, and H. Paulus, Biochim. Biophys. Acta 185, 447 (1969). 2M. Brenner, E. Gray, and H. Paulus, Biochim. Biophys. Acta 90, 401 (1964). F. H. Bergmann, this series, Vol. 5, p. 708.
580
H
MOA
[44]
ANTIBIOTIC BIOSYNTHESIS
NH
~ L-DBA
~ L-Thr~p
1
NH z
]----~L-DBA~L-DBA
2
3
4
NH 2
~'{ 5
.
[~[
6
NH~
]~L-DBA~-DBA~L-Thr 7
8
9
10
FI¢. 1. Structure of some polymyxins. MOA, (+)-6-methyloctanoate; DBA, 2,4-diaminobutyrate. Antibiotic Polymyxin Polymyxin Polymyxin Polymyxin Circulin A
A, B1 D1 E~
Residue 3
Residue 6
Residue 7
D-DBA L-DBA L-ser L-DBA L-DBA
D-Leu D-Phe D-Leu D-Leu D-Leu
L-Thr L-Leu L-Leu L-Leu L-Ile
Reagents Potassium phosphate, 0.1 M, pH 7.5. 32PPi, 20 raM, obtained commercially or prepared by the pyrolysis of [a2P]K~HP04 in a platinum crucible2 The presence of small amounts (5-10%) of [32p]orthophosphate will not interfere with the assay, and purification by anion exchange chromatography is therefore not necessary. ATP, 20 mM MgCI~, 50 mM L-2,4-Diaminobutyrate, 10 raM, Commercial preparations are frequently contaminated with other amino acids ~ and must therefore be further purified by chromatography on Dowex 50 or by crystallization as the dipicrate and then as the monohydrochloride2 AIternatively, L-2,4-diaminobutyrate can be conveniently prepared by the Schmidt degradation of L-glutamic acid and crystallized as above2 ,~ Enzyme. Ultrasonic or lysozyme extract as described below. HCIO~, 1 N Norit A charcoal suspension in water (10 mg/ml)
Procedure. The incubation mixtures contain 0.1 ml of phosphate buffer, 0.1 ml of 3~PPi, 0.1 ml of ATP, 0.1 ml of MgCI~, 0.1 ml of L-2,4-diaminobutyrate, enzyme and water to give a final volume of li0 ml. The reaction is carried out for 30 min at 37 ° for the assay of the 4 M. DiGirolamo, O. Cifferi, and A, B. DiGirolamo, J. Biol. Chem. 239, 502 (1964). D. W. Adamson, J. Chem. Soc. London, p. 1564 (1939). 6 H. Paulus and E. Gray, J. Biol. Chem. 239, 865 (1964).
[44l
POLYMYXIN SYNTHETASE
581
solubilized enzyme or at 25 ° with the membrane-bound enzyme and is terminated by the addition of 1 ml of cold HCI04 and 1 ml of Norit A charcoal suspension. After 2 hr at 0 ° with occasional shaking, the mixtures are filtered on Millipore filters (0.45 t~m) and the charcoal is washed with five 10-ml portions of water. The filters are transferred to glass vials and dried at 100°; their radioactivity is determined in a liquid scintillation spectrometer with 5 ml of toluene containing 0.4% Omnifluor (New England Nuclear). The results are corrected by subtracting the radioactivity observed in controls from which L-2,4-diaminobutyrate had been omitted. The assay is linear with respect to enzyme concentration and time for at least 45 min if the exchange does not exceed 0.2 umole. A unit of activity is defined as the amount of enzyme that catalyzes a 3~PP~-ATP exchange of 1 umole per minute under these conditions.
Enzyme Preparation Reagents Medium for starter cultures. Per liter: glucose, 10 g; Bacto yeast extract, 5 g; ammonium sulfate (Schwarz-Mann enzyme grade), 20 g; K,,HPQ, 2.6 g; MgSO~-7H20, 0.5 g; NaC1, 50 rag; and FeSO.7H20, 10 mg Growth medium. Per liter: ammonium sulfate (Schwarz-Mann enzyme grade), 42 g; MgSO~.7H,_,O, 0.2 g; NaC1, 0.1 g; CaCI~, 0.1 g ; FeSO4- 7H~O, 10 mg; ZnSQ 10 mg; MnSO~. H..,O, 7.5 mg; biotin, 0.5 ug; 0.5 M potassium phosphate, pH 7.5, 40 ml; and glucose (500 g/liter), 10 ml (added separately as a sterile solution to the autoclaved medium) Buffer A: 10 mM Tris.HC1, pH 7.5; 10 mM KC1; 2 mM EDTA and 2 mM 2-mercaptoethanol Buffer B: 10raM Tris.HC1, pH 7.5; 10 mM KC1; 2 mM MgCI~ and 2 mM 2-mercaptoethanol Pancreatic deoxyribonuclease I (crystalline) Egg white lysozyme (twice crystallized) Potassimn phosphate, 5 raM, pH 7.5 Growth of Bacteria. 1,6 The bacterial strain used is a well sporulating derivative of B. polymyxa Pfizer 2459, selected by passing the culture through several sporulation cycles. The strain is best maintained at --15 ° as a suspension of heat-activated spores. Starter medimn is inoculated with spores and, after overnight growth, 5 ml are transferred to 1 liter of growth medium in a 2-liter Erlenmeyer flask on a rotary shaker at
582
[441
ANTIBIOTIC BIOSYNTHESIS SUBCELLULAR D~STmBUTION OF AMINO ACID ACTIVATING ENZYMI~;S
Activation of Method of c e l l disruption Assayed at 25° Procedure A (lysozyme lysis) Assayed at 37° Procedure B (sonication)
Subcellular fraction
L-Diaminobutyrate
Soluble Particles Washed particles
1.6 1.8 1.1
Soluble Particles
8.7 1.5
L-Leucine L-Methionine
Units per g of cells 6.2 1.3 0.06 27 2.3
0.5 0.7 0.05 18 0.7
37 ° . Near the end of the exponential phase of growth, the cells are harvested by centrifugation at 4 ° and washed once with the buffer to be used for cell disruption. Preparation o] Extracts. 1 The subcellular distribution of the L-2,4diaminobutyrate activating enzyme depends on the method of cell disruption. Mild lysis of protoplasts prepared with lysozyme (Procedure A) yields primarily membrane-bound enzyme, which can be easily separated from most other amino acid activating enzymes by washing of the membrane fragments. In contrast, ultrasonic disruption (Procedure B) leads to almost complete solubilization of tile L-2,4-diaminobutyrate activating enzyme, which can then be purified further by conventional enzyme fractionation procedures. Procedure A. Freshly harvested cells of B. polymyxa (1 g) are suspended in 50 ml of buffer A and treated with lysozyme (25 mg) at 25 ° for 20 min. The suspension is cooled to 4 ° and centrifuged at this temperature for 20 min at 30,000 g to yield a soluble fraction and a particulate fraction. The latter is resuspended in 25 ml of buffer A and centrifuged again at 30,000 g for 20 min. The particles are then suspended in 25 ml of buffer B and treated with 1 mg of pancreatic deoxyribonuclease for 10 min at 25 °. The sample is again cooled to 4 ° centrifuged and washed (with buffer B) as before, and the final membrane preparation is then resuspended in a small volume of buffer B. As shown in the table, this fraction contains about one-third of the total L-2,4-diaminobutyrate activating activity and only very low levels of other amino acid activating enzymes. Procedure B. All steps are carried out at 4 °. Washed cells of B. polymyxa, either freshly harvested or after storage at --15 °, are suspended in 3 volumes of 5 m M potassium phosphate buffer, p H 7.5, and
[44l
POLYMYXIN SYNTHETASE
583
sonicated in 10-ml portions for 10 min each with an MSE model 60 ultrasonic disintegrator. After centrifugation at 30,000 g for 20 min, most of the L-2,4-diaminobutyrate activating activity, together with other amino acid-activating enzymes, is found in the supernatant fraction (see the table). A 3-fold purification can be achieved without loss of activity by centrifugation at 100,000 g for 90 rain, followed by precipitation of the enzyme by addition of solid ammonium sulfate (25 g/100 ml). When the ammonium sulfate precipitate is subjected to gel filtration on Sephadex G-200 in 5 mM potassium phosphate, pH 7.5, the a-2,4-diaminobutyrate-activating enzyme emerges near the void volume of the column while most other amino acid-activating enzymes elute somewhat later. The preparation so obtained is essentially free of the activating enzymes for basic amino acids and is therefore suitable for the assay of L-2,4-diaminobutyrate in mixtures of basic amino acids. 2
Properties Stability. 1 The particulate enzyme preparation obtained by Procedure A is quite unstable at elevated temperatures. In buffer B, its half-life is 5 min at 37 ° and several hours at 25 °. The solubilized enzyme (Procedure B) is relatively stable, even at 37% and has been stored in 40% ethylene glycol at --15 ° for a year without loss of activity. This contrasts with the extreme lability of the enzyme from other polymyxin-producing organisms (B. polymyxa ATCC 10401 and B. circulans ATCC 14040), where sonication leads to complete destruction of enzyme activity and enzyme preparations obtained by milder methods of cell disruption (alumina grinding or Hughes' press treatment) are completely inactivated upon overnight storage at 4 ° . Molecular Properties. 1 On isopycnic centrifugation in sucrose gradients, the particulate L-2,4-diaminobutyrate-activating enzyme obtained after protoplast lysis (Procedure A) bands at a density of 1.20-1.28 g/ml, indicative of an association with the cytoplasmic membrane. The soluble enzyme (Procedure B) appears polydisperse upon zone sedimentation in sucrose gradients, with an average molecular weight of about 300,000. After treatment with streptomycin, the soluble enzyme yields a sharper peak on sucrose gradients at a position corresponding to a molecular weight of about 100,000, but, on account of its instability, this form of the enzyme has not been studied further. Catalytic Properties. 1,~ The apparent K,,, for L-2,4-diaminobutyrate is 0.6 mM for both the solubilized and the membrane-bound enzyme. No
H. Paulus, unpublished experiments.
584
ANTIBIOTIC BIOSrNTHESlS
[44]
reaction is observed with related compounds such as D-2,4-diaminobutyrate, b-ornithine, and L-lysine. L-2,4-Diaminobutanol is a competitive inhibitor with a K~ of 1.5 raM. No pyrophosphate exchange is observed with GTP, UTP, or CTP in the place of ATP. The enzyme has an absolute requirement for Mg 2÷ ion, with maximum activity at 5-10 mM Mg 2+. Its pH optimum is at 7.0. Other Considerations Since other steps in the biosynthesis of polymyxin have not yet been elucidated, it is not possible to define the place of the L-2,4-diaminobutyrate activating enzyme in the biosynthetic scheme at this time. Nevertheless, it is possible to make a comparison with the biosynthesis of gramicidins and tyrocidine,9 which also involves the activation of the constituent amino acids by specific enzymes. These enzymes catalyze the formation of the appropriate aminoacyl adenylates, from which the amino acid is then transferred to a sufhydryl group on the enzyme. The activating enzymes thus have two catalytic functions, the formation of aminoacyl adenylates and of aminoacyl thiol esters, which are reflected in the amino acid-dependent exchange of ATP with PPi and of ATP with A M P . ~° While the L-2,4-diaminobutyrate activating enzyme catalyzes an ATP-PPi exchange, attempts to demonstrate a diaminobutyrate-dependent exchange of ATP with AMP have been unsuccessfulJ The failure to observe such an exchange reaction might imply that thiolester formation does not occur on the activating enzyme itself but, rather, that the enzyme-bound diaminobutyryl-AMP is the substrate for a separate diaminobutyryltransferase. In this case, the role of the L-2,4-diaminobutyrate activating enzyme would be analogous to that of the D-alanine activating enzyme in the synthesis of teichoic acid by L a c t o b a c i l l u s c a s e i Y
8This volume [43]. This volume [45]. 1oF. Lipmann, Science 173, 875 (1971). 1~R. Linzer and F. C. l'Teuhaus,J. Biol. Chem. 248, 3196 (19737.
[45[
TYROCIDINE SYNTHETASE SYSTEM
5S;~
[45] Tyrocidine Synthetase System B y Su~'G G. LEE and FRITZ LIPMANN
I. Biosynthesis of Tyrocidine During our work on the biosynthesis of gramicidin S by another Bacillus brevis (see Zimmer and Laland, this volume [43]), the synthesis of which is, in prineiple, quite analogous to that of tyrocidinc, it was first observed t h a t the activation of amino acids in this antibiotic polypeptide synthesis was a two-phase process. 1,2 Pyrophosphate exchange with A T P t h a t was dependent on amino acid activation had already been described by Kurahashi, :~ but we soon realized t h a t the exchange reaction indicating an aminoaeyl-adenylate formation, as in the activation for ribosomal synthesis, was followed by a secondary transfer. I t was then found t h a t using [ 3 H ] A M P and 14C-labeled amino acids for the amino a e i d - A T P reaction, the Sephadex G-50 filtrate contained two kinds of amino acids, only half of which was obviously bound to 13H]AMP as it was lost on triehloroaeetie acid precipitation together with [:~H]AMP. "-,4 The trichloroaeetie acid precipitate, however, contained the other half of the ['~C]amino acid, and tests for its binding to the protein showed it to be linked in thioester linkage. This observation seems to have been the first clear indication of the amino acids being linked by thioester bonds to the enzyme protein before polymerization. Polymerization apparently occurred without release of the terminal amino acid, since enzyme-bound polypeptides, isolated first by Laland and his collaborators ~ and then by Gevers et at.,'-' were shown to be linked by thioester bonds to the enzymes. The similarity between this type of activation and t h a t of acetate and of growing f a t t y acid chains in f a t t y acid synthesis to protein-bound pantetheine, 6,~ suggested ~H. Kleinkauf, W. Gevers, and F. Lipmann, Proc. Nat. Aead. Sci. U.S. 62, 226 (1968). z W. Gevers, H. Kleinkauf, and F. Lipmann, Proc. Nat. Acad. Sci. U.N. 63, 1335 (1969). H. Itoh, M. Yamada, S. Tomino, and K. Kurahashi, d. Biocl~ern. (Tokyo) 64, 259 (1968). W. Gevers, H. Kleinkauf, and F. Lipmann, Proc. Nat. Acad. Sci. U.S. 60, 269 (1968). 5T. Ljones, O. Vaage, T. L. Zimmer, L. O. FrOholm, and S. G. Laland, FEBS Lctt. 1, 339 (1968). " P. N. Majerus, A. W. Alberts, and P. R. Vagelos, Proc. Nat. Acad. Sci. U.S. 53, 410 (1965). F. Lynen, D. Oesterhelt, E. Sehweizer, and K. Willeeke, i~* "Cellular Compartmentalization and Control of Fatty Acid Metabolism" (F. C. Gran, ed.), p. 1. Universitetsforlaget, Oslo, 1968.
586
ANTIBIOTIC BIOSrNTHESIS
~ediate
@
enzyme
/ D~-Phe ,---! I
\,,o
•
[45]
q.Tjs
DLPhe\ ,, \ . . ~
./]
eovy enz~0 FIG. 1. Structure of tyrocidine. that pantetheine might possibly be involved in the polymerization reaction of amino acids in antibiotic synthesis. Using an inefficient procedure for pantetheine determination, our early attempts to find it were frustrated2 However, in the meantime, Laland and his group 9 had found pantetheine in the gramicidin S system; they made the important observation that, only in the larger enzyme, 1 mole of pantetheine was present. Most likely, then, pantetheine was involved in the polymerization reaction since this enzyme had been found to thioesterify four amino acids which only polymerized after induction by the smaller phenylalanineactivating enzyme. Such a function of pantetheine in polymerization was preliminarily indicated by experiments with gramicidin S and tyrocidine by comparing the location of amino acid marker with that of pantothenic acid in enzyme systems charged with single amino acids and polypeptides. Fragmentation of the charged enzymes with pepsin after the reaction showed coincidence of pantothenic acid in the fragments containing peptides but not in those containing single amino acids. 1° More definitive experiments on the functioning of pantetheine and the isolation of pantetheine-peptidyl-carrier proteins from the two polyenzymes involved in tyrocidine synthesis has now confirmed the peptidyl carrier function. 11 The isolation procedure of a peptidyl carrier fragment from the intermediate and heavy enzymes for tyrocidine synthesis will be described in the second part of this report; the tyrocidine-synthesizing system will be described in detail in the first part. In summary, the synthesis of the cyclic decapeptide (Fig. 1) is obtained by the assembly of three separable 8It. Kleinkauf, W. Gevers, R. Roskoski, Jr., and F. Lipmann, Biochem. Biophys. Res. Commun. 4, 1218 (1970). gS. G. Laland, O. Frfyshov, C. Gilhuus-Moe, and T. L. Zimmer, Nature (London) New Biol. 239, 43 (1972). 1oI-I. Kleinkauf, R. Roskoski, Jr., and F. Lipmann, Proc. Nat. Acad. Sci. U.S. 68, 2069 (1971). 11S. G. Lee and F. Lipmann, Proc. Nat. Acad. Sci. U~. 71, 607 (1974).
[45]
TYROCIDINE SYNTHETASE SYSTEM
587
enzymes with molecular weights of 100, 230, and 440 X 103. They activate, respectively, one, three, and six amino acids, indicated by the brackets in Fig. 1, which are added in the direction of the arrows to form a thioester-linked growing polypeptide chain beginning with D-phenylalanine. The decapeptide is released from the enzyme by ring closure between Leu ~ S- E and the H~N. Phe at the N-terminal end of the enzymebound chain.
Growing Procedure for Obtaining Enzyme-Rich B a c i l l u s b r e v i s Materials 1. Germination medium: 1% skim milk, 0.1% yeast extract, 0.05% K2HPQ, 0.05% KH~PO4, 0.05% MgSO4.7H20, 0.01% NaC1, 0.002% CaCI2"2H20, 0.001% MnSO4.4H~O, and 0.001% FeSQ. 7H20 in distilled water. 2. Growth medium: 1% Bacto-peptone (Difco), 1% beef extract (Difco), and 0.25% NaC1 in distilled water. The pH of the medium is adjusted to 7.0 with KOH. 3. [l~C]Ornithine (New England Nuclear Corp.)
General Procedures. A batch of about 2 mg of spores of B. brevis (ATCC 8185), either in lyophilized form or in colonies on an agar slant, is transferred to 200 ml of the germination medium in a 1000-ml flask and incubated for 8-10 hr at 37 ° in a New Brunswick Controlled Environment Incubator Shaker at 300 rpm to convert the spores into vegetative cells. For small-scale growth, 2 ml of the vegetative cell culture is transferred to several 2-liter flasks, each containing 500 ml of growth medium, and incubation is continued at 37 ° in the rotary shaker at 300 rpm. For larger scale growth, a New Brunswick Fermentor, model CMG-314, is used. About 50 ml of the vegetative cell culture inoculum is used per 10 liters of growth medium, and the cells are grown at 37 ° with vigorous stirring and aeration. For a good yield of active enzymes it is crucial to harvest the organism at the time of maximum enzyme production, and keen attention should be paid to the detection of this time. For this purpose, the growth of the organism is monitored by measuring the optical density (OD) of the bacterial suspension, usually diluted 10-fold with 1% NaC1, and the density is measured at 30-min intervals at 600 nm. Beginning at the time when the OD of the culture reaches 1.5, 2 ml of tile culture is taken at 30-min intervals and incubated in a 50-ml flask with 1 tLCi of [14C]ornithine for 10 rain at 37 ° in the rotary shaker at 300 rpm. A
588
ANTIBIOTIC BIOSYNTHESIS
[45]
l0 8 6 4 E c 3 O
$2 a o
1.0 - 0.8 / i 0.6
~ o.4 ~ 0.3 0.2 0.12 i
4
6
8
IO
12
Time o f f e r inoculefion, hr
FIG. 2. Cell growth (E] [~), tyroeidine synthesis (0' ornithine uptake (O---O, epm X 10-').
O, cpm X 10-s), and
0.5-1.0-ml aliquot is withdrawn and the cells are collected on a 0.80 ~m pore Millipore filter. The filter is then washed in three 2-ml portions of the growth medium and dried quickly, and the radioactivity is counted. As shown in Fig. 2, the uptake of [14C]ornithine by the cells remains near background level, more or less in parallel with the cell density, until the appearance of tyrocidine-producing enzymes. When these enzymes appear, the capacity for ornithine uptake increases abruptly. At that time, the cells are chilled rapidly with crushed ice and harvested by centrifugation. Any delay in the chilling process causes loss of active enzymes. Generally, about 30-40% of the total ornithine is absorbed in 10 min at the peak of uptake; a maximum yield of enzymes is obtained when the cells are harvested at the time when about 10-15% of the total ornithine is taken up in 10 min. At first, the ornithine-dependent PPi-ATP exchange was measured in a cell homogenate as indicator for tyrocidine synthesis, until the abrupt upshift of the ornithine uptake was found to present a faster assay. The Enzymic Synthesis of Tyrocidine
Reagents 1. Amino acid stock solution: 20 mM D-phenylalanine, L-proline, and L-ornithine to test for the 100,000, 230,000, and 440,000 molecular weight enzymes, respectively
[451
TYROCIDINE SYNTHETASE SYSTEM
589
2. 32PPi stock solution: 10 mM PP~ (pH adjusted to 7.7) containing 20 ~Ci of ~PP~ per milliliter 3. ATP-buffer stock solution: 10 mM ATP, 50 mM MgCl~, 2.5 mM EDTA, 5 mM dithiothreitol, 100 mM triethanolamine buffer (pH 7.7), and 25% sucrose 4. Bovine serum albumin solution: 0.4% bovine serum albumin in 0.01 N NaOH 5. Charcoal suspension: mix 50 ml (wet volume) of acid-washed activated charcoal with 250 ml of 14% perchloric acid containing 0.4 M PPi ; add water to make up 1 liter. 6. PP~ washing solution: 0.1 M PP~, pH adiusted to 8.0 with HCI 7. Tyrocidine-constitutivc amino acid mixture: 1 mM each of L-phenylalanine, L-asparagine, L-glutamine, L-tyrosine, L-valine, L-ornithine, L-leucine, and [~4C]proline (20 ~Ci/ml)
Assay Procedures
AYP-PP~ Exchange. The reaction medium is prepared by mixing one part each of amino acid stock solution, 32PPi stock solution, bovine serum albumin solution, and ATP-buffer stock solution. When mixing these solutions, the 32ppi solution is added immediately before use because precipitation of Mg2P207 occurs from the medium after a long period of standing. The exchange reaction is carried out for 15 min by incubating 0.025 ml of enzyme with 0.1 ml of reaction medium at 37°; the reaction is stopped by adding 0.5 ml of charcoal suspension to the reaction vessel. The charcoal is collected on a glass filter; the filter is then washed with five 3-ml portions of 0.1 M pyrophosphate washing solution, followed by two 3-ml portions of water, and is dried and the radioactivty counted. For crude preparations, dilute solutions should be used since the exchange activity tends to decline in concentrated preparations. This is probably due to ATPase because a new addition of ATP reestablishes the exchange. Tyrocidine Synthesis. After Sephadex G-200 filtration (see below), a 0.15-ml mixture of the three complementary enzymes of tyrocidine synthesis is added to a reaction vessel containing 0.05 ml of tyrocidine-constitutive amino acids, including 1 ~Ci of [14C]proline (reagent 7) and 0.05 ml of ATP-buffer stock solution (reagent 3). The mixture is incubated for 30 rain at 37°; at the end of the reaction, 5% triehloroacetic acid (TCA) is added, and the precipitate is collected on a Millipore filter (0.45 t~m pore). The filter is washed with five 2-ml portions of 5% TCA, then dried, and the radioactivity is counted. For the blank, in place of the complete reagent 7, [14C]proline only is used in the reaction. In the latter case, the radioactivity incorporated into the TCA precipitate in
590
ANTIBIOTIC BIOSYNTHESIS
[45]
the blank represents the small amount of [~4C]proline thioesterified to the intermediate enzyme. This method is good for orientation, but for a more reliable measure, tyrocidine should be separated from the reaction mixture by extracting it with an equal volume of n-butanol:chloroform (4:1, v/v) and chromatographing it on a silica thin-layer plates using ethyl acetate:pyridine:acetic acid:water (90:30:16:9, v/v). The exact proportion of the three enzymes for a maximum rate of tyrocidine synthesis has not been established. It is observed, however, that a combination of a catalytic amount of the light enzyme and an excess of intermediate over heavy enzyme gives a much higher rate of tyrocidine synthesis than an equimolar mixture of the enzymes.
Purification of Tyrocidine-Synthesizing Enzymes
Preparation o/Extracts A batch of 80 g of frozen cells is broken up and thawed by blowing air over it at room temperature. The paste is suspended in 300 ml of 20 mM triethanolamine buffer, pH 7.7, containing 0.5 mM EDTA and 1 mM dithiothreitol (buffer B), and to this is added 80 mg of lysozyme in 20 ml of buffer B. The cells are lysed by incubation of the mixture for 5-10 min. The cells harvested at the peak of tyrocidine-synthesizing activity lyse within a few minutes, but older cells require a longer time. To the viscous mixture of the cell lysate, 50 tLg of DNase and 2 ml of 1 M MgC12 are added, and it is incubated for 1-2 min at room temperature until the disappearance of viscosity. Prolonged incubation of the lysate, especially after the addition of DNase, should be avoided since this causes dissociation of polyenzymes into inactive subunits. The lysate is chilled rapidly to 4-6 ° and centrifuged for 15 rain at 20,000 g. Solid ammonium sulfate is added to the supernatant to 33% saturation, and the precipitate is removed by centrifugation at 20,000 g for 10 min. More ammonium sulfate is added to 45% saturation, and the precipitate is again collected by centrifugation at 20,000 g for 10 min. The pellets are dissolved in buffer B (about 100-150 mg of protein per milliliter of buffer), and insoluble material is removed by centrifugation. When a cell lysate prepared from older cells is subjected to centrifugation, much of the enzyme activity precipitates with cell debris. This activity can be partly recovered by repeating washing.
Sephadex G-200 Filtration The 33-45% ammonium sulfate fraction is applied to a Sephadex G-200 column equilibrated with buffer B containing 0.1 M KC1 and is
[45]
TYROCIDINE
¢~
IOO
."'.~,
d;
,~
¢~ x:J
:
;
?
:
SYNTHETASE
A
~--Protine
:"i
•..
.."
* L I
k
D
,.'"%.,
".
"'.
Ormthine
:
I
~o
ohenylo~onine ".
"', ",
t~
! ].j
591
SYSTEM
0CO
'.
°'z x.
f
°
50
IO0 Froction
IbO
200
number
Fro. 3. Separation of the three complementaryfractions, heavy (HE), intermediate (IE), and light (LE) enzymes,on Sephadex G-200 chromatography. eluted with the same solution. For 2 g of protein, a column of 60 X 7 cm gives a satisfactory resolution of enzymes and an adequate flow rate. As shown in Fig. 3, the heavy and intermediate enzymes are located in the first and second peaks by measuring L-ornithine- and L-proline-dependent ATP-PPi exchange activities; the third and fourth peaks are assayed by D-phenylalanine-dependent ATP-PPi exchange. The third peak corresponds to the light enzyme, and the fourth peak, which may be bigger than shown in the figure, contains dissociation products of the polyenzymes. To the heavy, intermediate, and light enzyme peak fractions, solid ammonimn sulfate is added to 50% saturation, and precipitates are collected by centrifugation. The pellets are dissolved in a small volume of buffer B and passed through G-50 columns equilibrated with the same buffer. Sucrose is added to the enzyme solutions eluted from the columns to 5% ; the solutions are stored in liquid nitrogen until used for further purification.
Purification o] the Light Enzyme The light enzyme peak from G-200 chromatography is applied to a 12 X 2 cm DEAE-cellulose column equilibrated with buffer B containing 5% sucrose. The column is eluted with 100 ml of buffer B containing 5% sucrose and 0.1 M KC1, then further eluted with 500 ml of a KC1 gradient (0.1-0.4 M) in buffer B containing 5% sucrose. The light enzyme, which elutes between 0.16 and 0.18 M KC1, is assayed by measuring D-phenylalanine-dependent ATP-PPi exchange activity, and the enzyme
592
ANTIBIOTIC BIOSYNTHESIS
[45]
solution is concentrated with a Diaflo ultrafiltration apparatus to a few milliliters. The enzyme, which is about 30% pure at this stage, may be purified to homogeneity by hydroxyapatite chromatography and sucrose gradient centrifugation. The enzyme elutes at 0.03-0.04 M phosphate when applied to a hydroxyapatite column (30 X 1 cm) and eluted with a phosphate gradient (0.005-0.1 M), pH 7.3, containing 5% sucrose and 1 mM dithiothreitol. The enzyme solution is concentrated to 0.5 ml by means of the Diaflo apparatus, and 0.1 ml is layered on each of three 10-30% sucrose gradients in buffer B and 0.1 M KC1, which are then centrifuged for 7 hr at 50,000 rpm in a Beckman SW 50 rotor. Under these conditions, the peak activity of this enzyme sediments a little more than one quarter from the top.
Purification of the Intermediate Enzyme The intermediate enzyme from Sephadex G-200 chromatography is applied to a hydroxyapatite column equilibrated with 10 mM phosphate buffer, pH 7.3, containing 5% sucrose and 1 mM dithiothreitol, and the column is eluted with phosphate gradient (0.01-0.15 M), pH 7.3, containing 5% sucrose and 1 mM dithiothreitol. For 200-800 mg of protein, a 50 X 2 cm-column and a l-liter gradient are used. The intermediate enzyme elutes at 0.09-0.11 M phosphate when a 50-cm column is used, but at 0.07-0.08 and 0.05-0.06 M, respectively, when 20-cm and 12 cm columns are used, indicating that the elution concentration depends on the length of the column. The intermediate enzyme eluted from the column is located by measuring L-proline-dependent ATP-PPi exchange activity and is concentrated to a few milliliters by means of the Diaflo apparatus; this is then passed through a Sephadex G-50 column equilibrated with buffer B and applied to a DEAE-cellulose column (12 X 2 cm) equilibrated with buffer B containing 5% sucrose. The column is eluted with 50 ml of buffer B containing 5% sucrose and 0.1 M KC1, and then further eluted with 500 ml of KC1 gradient (0.1-0.5 M) in buffer B and 5% sucrose. The enzyme that comes off the column between 0.23 and 0.25 M KC1 is concentrated with the Diaflo apparatus to 0.5 ml, and 0.1 ml of this is layered on each of three 10-30% sucrose gradients in buffer B and 0.1 M KC1; the gradients are centrifuged for 7 hr at 50,000 rpm using a Beckman SW 50 rotor. Under these conditions, the intermediate enzyme sediments a little less than half way from the top.
Purification o] the Heavy Enzyme The heavy enzyme from G-200 chromatography is purified by hydroxyapatite chromatography in a manner identical to that described for
[45]
TYROCIDINE SYNTHETASE SYSTEM
dE
0.01M
[1_
%
50
~-
7D
o
593
4O
0.15M •
" ~ - - O r nithine
I/
~, 50
Io E
c © CO (M
05 5
2o !
I
I
©
10,
o E
20
I00 Froclion
150 number
00
FIG. 4. Hydroxyapatite chromatography of heavy enzyme fraction from Sephadex G-200 chromatography.
the intermediate enzyme. The heavy enzyme elutes at 0.05-0.06 M phosphate (Fig. 4). The elution concentration does not seem to be dependent on the column length. The enzyme solution is concentrated to about 1 ml and subjected to sucrose gradient centrifugation as described for the intermediate enzyme. Under such conditions, the heavy enzyme sediments a little less than two-thirds from the top (Fig. 5). The heavy enzyme eluted from the G-200 column is often turbid. Since the presence of much
,%
£3[3_ c~
to
-?
c c
g
100
_o × E Q.
~~I.
50
0rnithJne
i
"o
~/~/ t. ph~nylalonine
Froction
20 number
50
FIG. 5. Sucrose gradient centrifugation of heavy enzyme fraction from hydroxyapatite chromatography.
594
ANTIBIOTIC BIOSYNTHESIS
[45I
TABLE I PURIFICATION OF THE ~NTERMEDIATE AND HEAVY ENZYMES OF TYROCIDINE SYNTHESIS
Stages
Protein (mg)
Intermediate 1. 20,000 g supernatant 2. Ammonium sulfate (33-45% saturation) 3. Sephadex G-200 4. Hydroxyapatite 5. DEAE-cellulose 6. Sucrose gradient
Specific activity (cpm/mg protein)
Yield (%)
Proline-dependent 26,000 7,400 780 72 11 3.8
Heavy 1. 20,000 g supernatant 2. Ammonium sulfate (33-45% saturation) 3. Sephadex G-200 4. Hydroxyapatite 5. Sucrose gradient
Amino aciddependent ATP-3~PPi exchange (cpm)
3.1 X l0 g 2.4 X 109
1.2 X 105 3.2 X 105
77
1.8 9.3 5.1 2.8
2.3 1.3 4.6 7.4
106 107 107 107
58 30 16 9
1.1 X 109 7.2 >< 108
4.2 X 104 9.7 X 104
65
5.9 X 108 3.1 X 108 1.5 X 108
1.3 X 106 1.2 X 107 2.6 X 107
54 28 14
X X X X
109 108 108 108
X X X X
Ornithine-dependent 26,000 7,400 460 26 "5.8
turbid material interferes with hydroxyapatite chromatography, it is precipitated with ammonium sulfate (50% saturation) and the pellets obtained by centrifugation are dissolved in about 5 volumes of buffer B; the enzyme is passed through 15% sucrose and layered on 40% sucrose by high-speed centrifugation. Usually, 5 ml of 40% sucrose in buffer B and 0.1 M KC1 is layered on the bottom of 38-ml tubes, 10 ml of 15% sucrose in the same buffer solution is layered next, and then the enzyme solution is layered on the top. The tubes are centrifuged for 6 hr at 60,000 rpm using a Beckman 60 titanium fixed-angle rotor. The enzyme sediments through the 15% sucrose and is layered on the 40% sucrose cushion. Table I shows the progress of purification. Acrylamide gel electrophoresis (Fig. 6) indicates the purity of the stage 6 intermediate and stage 5 heavy enzymes to be approximately 90%. In Table II, the amino acids which are activated by the purified enzymes indicate their specificity. With regard to aromatic amino acids, 12 the heavy enzyme shows 12R. Roskoski, Jr., H. Kleinkauf, W. Gevers, and F. Lipmann, Biochemistry 9, 4846 (1970).
[45]
TYROCIDINE SYNTHETASE SYSTEM
595
IE HE
FIG. 6. Gel electrophoresis of intermediate (IE) and heavy (HE) enzymes from sucrose gradient centrifugation. a preference for tyrosine and the intermediate for phenylalanine and tryptophan (data not included). II. Dissociation Products of the Polyenzymes of Tyrocidine Synthesis Of the three complementary enzymes of tyrocidine synthesis, the intermediate and heavy enzymes are polyenzymes, each composed of amino-acid activating subunits of 70,000 molecular weight and of a 4'phosphopantetheine-containing protein of 17,000 molecular weight which carries the growing nascent peptides. The intermediate enzyme (230,000 molecular weight), which activates the second, third, and fourth amino acids of tyrocidine, contains three amino acid activating subunits; the heavy enzyme (440,000 molecular weight), which activates the fifth to tenth amino acids of tyrocidine, contains six amino acid activatin~ subunits. The dissociation of these polyenzymes may be achieved either by
596
ANTIBIOTIC BIOSYNTHESIS
[45]
TABLE II AMINO ACID-DEPENDENT ATP-a~PPi EXCHANGE WITH THE PURIFIED ENZYMES OF TYROCIDINE SYNTHESIS a
[3*P]ATP formed (nmoles) Amino acids
Light
Intermediate
Heavy
~Phenylalanine D-Phenylalanine I~Proline L-Asparagine L-Glutamine L-Tyrosine ~Valine L-Ornithine L-Leucine
34.3 22.4 0 0 0 NM b 0 0 0
12.3 2.1 30.2 0 0 2.7 0 0 0
2.2 0 0 20.0 11.6 13.6 38.2 15.4 24.8
The last stages of purified light enzyme (1.6 ~g), intermediate enzyme (2.4 ug), and heavy enzyme (4.1 ~g) were incubated for 15 min at 37° with 4 mM of the designated amino acids and 2 mM ATP, 2 mM ~2PPi (0.32 #Ci), and 5 mM KC1 in a reaction volume of 0.1 ml. The exchanges were assayed as described in the text. The activation of L-tyrosine by the intermediate, and L-phenylalanine by the heavy enzyme, indicate the interehangeability of aromatic amino acids, although with ~ selective affinity [Reprinted with permission from R. Roskoski, Jr., H. Kleinkauf, W. Gevers and F. Lipmann, Biochemistry 9, 4846 (1970) Copyright by the American Chemical Society]. b Not measured. extended incubation of crude bacterial extracts or by incubating the purified polyenzymes with a Triton extract made from the insoluble fraction of the bacterial lysate.
Procedure/or Dissociating Polyenzymes 13 Dissociation o/Polyenzymes in Crude Extracts. A batch of 5-100 g of frozen B. brevis (ATCC 8185) cells is thawed and suspended in 4 volumes (w/v) of buffer B, and the cells are lysed by incubation at 37 ° for 5-10 min with lysozyme (1 m g / g cells). To the lysate are added D N a s e (1/~g per gram of cells) and MgCl~ (5 m M final concentration), and the mixture is incubated for 30-60 min. Autolysis of the polyenzymes is more pronounced after addition of the DNase. Figure 7 shows a sucrose gradient analysis of the progress of autolysis of the h e a v y enzyme with 1'S. G. Lee, R. Roskoski, Jr., K. Bauer, and F. Lipmann, Biochemistry 12, 398 (1973).
[45]
SYNTHETASE SYSTEM
TYROCIDINE
597
70
o x: o
D
m
60
a_: ?
13-
"n
'
50 ?%" - -
8-
×