Advances in Applied Microbiology, Volume 47

Seeing Red: The Story of Prodigiosin J. w. BENNETT

Department of Cell and Molecular Biology Tulane University New Orleans, Louisiana 70118

RONALD BENTLEY

Department of Biological Sciences University of Pittsburgh Pittsburgh, Pennsylvania 15260

I. Bread, Blood, and Bacteria II. Early Instances of "Blood" on Bread III. Red Bacteria and the History of Bacteriology A. Pre-Pasteurian Research B. Pigments and Paintings C. The Genus Serratia IV. Prodigiosin and Related Compounds A. Structures B. Biosynthesis V. From Saprophyte to Pathogen VI. Biological Activity of Prodigiosin and Related Compounds A. Possible Ecological Functions B. Pharmacological Activity VII. Final Comments References

I. Bread, Blood, and Bacteria Bread, b o t h l e a v e n e d a n d u n l e a v e n e d , p l a y s a crucial nutritional, religious, a n d e m o t i o n a l role in h u m a n lives. In the Old Testament, b r e a d is said to " s t r e n g t h e n e t h m a n ' s heart" (Psalms 104:151), a n d in the Lord's Prayer the request is "Give us this d a y o u r daily b r e a d " ( M a t t h e w 6:11). In J u d a i s m , u n l e a v e n e d b r e a d is the c e n t e r p i e c e of the P a s s o v e r meal. In Christianity, the Eucharist or s a c r a m e n t of the Lord's S u p p e r is c e l e b r a t e d t h r o u g h the c o n s e c r a t i o n a n d c o n s u m p t i o n of b r e a d a n d

1Biblical quotations are from the Authorized King James Version.

ADVANCES IN APPLIED MICROBIOLOGY, VOLUME 47 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved. 0065-2164/00 $25.00

2

J.w. BENNETT and RONALD BENTLEY

wine: "And as they did eat, Jesus took bread, and blessed, and brake it, and gave to them, and said, Take, eat: this is my body" (Mark 14:22). Bread is also, especially when not dried out, an excellent culture medium for the growth of many microorganisms, so much so that many present-day commercial breads contain calcium propionate "added to retard spoilage." In the pre-antibiotic era, microbial contamination of bread was used to good effect: the healing of wounds was facilitated by application of preparations made from moldy bread. A specific and early example is documented in an English herbal of 1760. Such preparations may well have contained penicillin, patulin, or other antibiotic materials formed by the fungi (Wainwright, 1990). However, in most cases when microbes use bread as a substrate for their growth, the result is spoilage. Contaminated breads can be detected by repellent flavors and distinctive coloration. Most spoilage of bread is caused by fungi: Aspergil]us niger forms black colonies, many members of the genus Penicillium are blue or green, while certain yeasts and bread molds such as Neurospora crassa form pink to red pigments. Bacteria are less commonly associated with deterioration of bread; however, under warm and humid conditions some strains of Serratia marcescens form distinctive red colonies on this substrate. The red color derives from the presence of the pigment prodigiosin and/or related materials (see later). As the bacterial colonies reach maturity, they dissolve into a fluid and viscous state with a mucilaginous appearance and an uncanny resemblance to blood. Indeed, from early times, there are many records of the appearance of "blood" on bread, beans, and other starchy foods such as polenta and potatoes. Like bread, blood is a substance with profound cultural implications beyond its physiological role. Human and animal sacrifice were practiced in many societies with the intent to propitiate the wrath of an all-powerful deity. The victim's blood was often associated with a mystical power. For the Aztecs, the sun god (Huitzilopochti) drove back the moon and stars each day. To carry out this tremendous task, he had to be nourished with human blood. In some cultures, prisoners of war were sacrificed and their blood consumed by the executioners, while in other cultures the drinking of blood was taboo (e.g., the ritual slaughter of animals by exsanguination as practiced by the Jews). The Old Testament is filled with blood imagery and stories of ritual sacrifice. During the momentous Passover devastation of all the firstborn (both men and beasts) in the land of Egypt, the Israelites were protected by the blood of an unblemished, 1-year old, male lamb spread on the side and upper door posts of their houses (Exodus 12). The paschal (Passover) lamb was a term later applied symbolically to Christ.

SEEING RED: THE STORY OF PRODIGIOSIN

3

To this day, many Christians proclaim that for believers redemption from sins is only possible by way of "the blood of the Lamb". At the Last Supper, Christ used wine to symbolize his blood: "This cup is the new testament in m y blood, which is shed for you" (Luke 22:20). The Roman Catholic faith has embraced the belief in transubstantiation, whereby the bread and wine of the Eucharist actually turn into the body and blood of Christ. This formal doctrine was specifically defined at the Fourth Lateran Council (1215) and reaffirmed at the Council of Trent (1551) (Cross and Livingstone, 1974). As noted above, mature colonies of pigmented Serratia are eerily bloodlike in appearance. More than a few microbiologists have hypothesized that the growth of these bacteria could be interpreted, in certain religious or symbolic contexts, as the miraculous appearance of blood. This paper discusses the possible role of Serratia m a r c e s c e n s in forming bloodlike material on starchy foods, and reviews many of the unusual properties of this fascinating bacterium and the red pigment(s) that it and other microorganisms form. In reviewing the historical record, we have made extensive use of previous publications (Harrison, 1924; Reid, 1936; Gaughran, 1969; Yu; 1979; Cullen, 1994). 2 II. Early Instances of "Blood" on Bread

It is impossible to know who first observed foodstuffs apparently carrying drops of blood. Red-spotted bread was probably observed in many parts of the world; however, only in European countries is there an extensive written record and only there did it come to play a role in religious controversy. The first known recorded report dates from Alexander the Great's siege of Tyre in 322 BCE. The disgruntled Macedonian troops were tired of the siege, w h e n a soldier noticed a trickle of blood inside a piece of broken bread. A soothsayer named Aristander interpreted the event as a good omen, opining that, had the droplets of blood been on the outside, the Macedonians would have been endangered. Since the flow was from the inside, it was an omen that Tyre would fall. Almost certainly, Aristander was well compensated for his ability to both calm the troops and prophesy the future. Christian Gottfried Ehrenberg (1795-1876) collected almost 100 European reports of the occurrence of blood, starting in the eleventh century; an English summary was provided by Gaughran (1969). Significantly, 2Unless otherwise apparent, the simple word "blood" will carry the meaning of "blood" or "bloodlike materials" to avoid much repetition. The genus abbreviation S. will refer to Serratia, and Streptomyces will not be abbreviated.

4

J.w. BENNETT and RONALD BENTLEY

blood was frequently observed on, or flowing from, the bread or wafers used as the Host in the Eucharistic liturgy. The prototype of many examples was recorded from Alsen, Denmark, in 1169. A village priest saw blood on a Host. Upon reporting this event to his superiors, the Chief Priest predicted the imminent shedding of Christian blood. A few days later, a pagan army overthrew churches, drove people into slavery, and killed those who resisted. Within several decades after the Alsen event, a strange myth grew up around reports of blood on Communion hosts. In 1247, near Berlin, a w o m a n removed a consecrated wafer from her mouth and sold it to Jews, who "stabbed it," resulting in the appearance of blood. The wafer was returned to the church, bringing it much fame, and the Jews were apparently unharmed. However, in other similar stories, Jews were persecuted and killed. Thus, in 1296 near Frankfurt, a purportedly stolen wafer was sold to Jews, stabbed, and yielded blood. A mob subsequently marched with banners, attacking Jews in Nuremberg, Rothenburg, Wfirzburg, and elsewhere, with a reported death toll of 10,000. Similar stereotypic reports of bloody Communion wafers led to repeated tormentations and executions of Jews for at least 200 years. Ominously, the geography of the persecutions overlapped that of the atrocities of the Holocaust several centuries later, with most of the incidents reported from cities in Germany and Poland. The number of those who perished will never be known. Scientists who believe that the explanation for bleeding hosts is the growth of Serratia are fond of quoting Scheurlen (1896), an early German observer: "dieser Saprophyt mehr menschen umgebracht hat als mancher pathogene Bacillus," or, in Isenberg's translation (1995), "This saprophyte has killed more humans than some pathogenic bacilli." In retrospect, whether or not the stories of red spots on Communion wafers were real or fabricated, the interpretation of the events and the reprisals against the Jews seem to make little sense. The Middle Ages was, of course, a time of many superstitions and profound religious belief. One wonders, however, w h y the Jews were not hailed as heroes for showing that the Host w o u l d bleed, thus providing a vivid demonstration of the truth of transubstantiation. It was nonsense to think that Jews wished to drink any sort of blood, since such an action was specifically forbidden to them. Deeper irrationalities were involved, with the incidents serving as a pretext to express a violent antisemitic prejudice. Eventually, the legend of the "blood libel" developed, which held that Jews needed the blood of children to make bread for Passover or for sorcery-related medicinal purposes; more recently, this libel was

SEEING RED: THE STORY OF PRODIGIOSIN

5

part of the many grotesqueries incorporated into Nazi propaganda (Trachtenberg, 1943). On the other hand, in a Christian context, the presence of blood on sacramental bread was interpreted in support of the doctrine of transubstantiation. In fact, one such incident became perhaps the most celebrated miracle of the thirteenth century (Cullen, 1994). According to Church history, a German priest with doubts about the doctrine of transubstantiation once celebrated mass at the Church of Santa Cristina in Bolsena, Italy in 1263. When blood dripped from the Host onto the altar linen and his vestments, his doubts were resolved and he sought absolution for his lack of faith. This event became celebrated as "The Miracle of Bolsena" and was later depicted in a Vatican fresco by Raphael. To commemorate the miracle, Pope Urban IV issued a bull that instituted the Feast of Corpus Christi and later decreed the construction of a n e w cathedral in Orvieto in which the host and vestment linens are preserved to this day. It has been suggested that the relics provide a tantalizing experimental opportunity. If enough DNA could be isolated from them, the polymerase chain reaction could be used to test the hypothesis that Serratia marcescens was involved in this medieval miracle. III. Red Bacteria and the History of Bacteriology A. PRE-PASTEURIAN RESEARCH

A giant leap forward in the understanding of microbiology in general and the formation of red-pigmented materials on foodstuffs in particular began with yet another event in Italy, this time in Legnaro (province of Padua) in 1819. The affected foodstuff was a bloody polenta (corn mush, corn porridge) found in the squalid home of a superstitious farmer named Antonio Pittarello. Eventually, more than 100 families in the region reported bloodlike materials on polenta or rice soup. A cooked chicken was described as "dripping with blood." Maleficent spirits were blamed for the event, and families who found bloodlike spots on their food were accused of evil activities. The event caused so much publicity that an official investigation was established under the direction of Dr. Vincenzo Sette, the medical officer at Pione di Sacco. Sette concluded that a fungus was responsible, while a botanist-priest, Pietro Melo, "claimed that the phenomenon was due to a spontaneous fermentation of the polenta which caused the corn meal to be transformed into a colored mucilage" (Breed and Breed, 1924). Sette eventually published his report in 1824, calling the organism Zaogalactina

6

J.w. BENNETT and RONALD BENTLEY

imetropha (from the Greek, "living slime situated on food"). On one occasion, he reddened polenta in a priest's house, thereby, according to the story, disposing of the theory that the phenomenon could only occur in the house of a sinner (it is a tribute to local piety that no one suggested the possibility that the priest was less than perfect!). Meanwhile, a pharmacist, Bartolomeo Bizio (then a student and later professor at the University of Padua), independently examined the red potenta, giving preliminary and detailed accounts of his work in 1819 and 1823 (Merlino, 1924). Bizio also classified the organism as a fungus and coined the further name Serratia marcescens. He used Serratia to honor a physicist, Serafino Serrati, who had run a steamboat on the Arno in 1787. Bizio believed that Serrati had a prior claim over "a foreigner" (presumably James Rumsey) as inventor of the steamboat and wished to honor his countryman. The second part of the binomial, marcescens, came from the appearance of the mature colonies that dissolved into "a fluid and viscous matter which has a mucilaginous appearance." Marcescens is the present participle of the Latin verb meaning "to decay or wither." Bizio performed experiments in which he used paper soaked with the red substance, or bits of red polenta, to transmit "seeds" of his fungus. As did Sette, Bizio made an honest mistake in identifying the causative bacterium as a fungus, but we are indebted to them for laying sure foundations for further investigations. Both Sette and Bizio were the first to provide evidence suggesting that the bloody material on food was due to a living organism, similar to the alga that caused pink snow on mountains, and transmissible from substrate to substrate by inoculation. Many years later, the spoilage of corn by fungal growth was investigated in connection with pellagra. With whole corn, "sometimes the embryo is colored reddish by Micrococcus prodigiosus"--one of the many binomials applied to S. marcescens (Black and Alsberg, 1910). These authors, however, did not refer to the spoilage of polenta. They are known for their discovery of penicillic acid in Penicillium puberulure and the rediscovery of mycophenolic acid in Penicillium brevicompactum (Alsberg and Black, 1913). Another chapter of the Serratia story picks up in 1848, when bloodlike spots were found on a boiled potato in a Berlin home. Ehrenberg investigated the phenomenon and became fascinated. A distinguished physician and protozoologist (he described more than 300 n e w species), Ehrenberg regarded the organism as a "tiny, oval animalcule," and renamed it Monas prodigiosa in 1849. Although aware of Sette's prior nomenclature, his historical efforts apparently did not lead him to Bizio. He came to believe that most historical accounts of bloody food could be attributed to the growth of M. prodigiosa. His colleague

SEEING RED: THE STORY OF PRODIGIOSIN

7

Scheurlen supported this belief, speaking of the "deceptive red color" of the microorganism and the fact that the hosts were a "particularly agreeable . . . m e d i u m " (Scheurlen, 1896: Gaughran, 1969). Cullen (1994) has analyzed retrospectively weather conditions that correlate humidity and warm air temperatures with historical reports of bloody foods. Another historical aspect may be noted. The laboratory culture of microorganisms on solid or semisolid media (e.g., cooked potato slices, gelatine, agar) developed towards the end of the nineteenth century; Koch's pioneering work with gelatine was reported in 1881. Although Sette and Bizio did not use pure cultures, their work with polenta is probably the first documented use of a solid m e d i u m for culturing microorganisms (Bulloch, 1938). A splendid color plate of S. marcescens growing on a potato is found in an early text (Crookshank, 1890). The organism is named as Bacterium prodigiosum with three other binomials but not including S. marcescens; it is also described picturesquely as "Blood Rain." B. PIGMENTS AND PAINTINGS At the time of Sette's work, a chemist, Pietro de Col, extracted the red pigment and used it to dye silk. He also created yet one more name, Mucot sanguineus, for the organism (Harrison, 1924). Similarly, both Sette and Bizio made ethanol extracts of the pigment and used them to dye silk and wool, sometimes with the aid of mordants. Alert to commercial possibilities, they were thwarted by the unfortunate sensitivity of the dye to light. It has to be remembered that in 1819 the major available red pigments were naturally occurring secondary metabolites derived with difficulty from insects (cochineal, kermes, lac) or from plants (madder); not until 1856 did Perkin produce the first synthetic dye, mauve. More than a century after Bizio and Sette's work, Alexander Fleming found a curious application for the red-pigmented bacterium. He made microbial "paintings" by outlining a drawing on blotting paper, placing it on a nutrient agar plate, and then inoculating with bacterial culture broths. On incubation a colored "germ painting" developed. Six of these "paintings" were reproduced on the endpapers of his biography by Andr6 Maurois. Clearly unfamiliar with the vagaries of bacterial nomenclature, Maurois used the superseded name in describing the possible colors--"the staphylococcus is yellow, the bacillus prodigiosus (sic) red, the bacillus violaceus (sic) blue" (Maurois, 1959). Even more recently, Serratia has also been used as "red ink"; at the 1956 Presidential Banquet of the American Society for Microbiologists in Houston, substitute "place cards" were fashioned from Petri plates containing

8

J.w. BENNETTand RONALDBENTLEY

appropriate media on which the names of the officers had been "written" with cultures of red-pigmented S. marcescens (Anonymous, 1999). C. THE GENUS SERRATIA The organism with this long and fascinating history is a member of the Enterobacteriaceae (aero-anaerobic, Gram-negative bacteria) and is motile (Blazevic, 1980; Grimont and Grimont, 1984, 1991). Some species and biotypes of Serratia produce reddish pigment(s) and, depending on colony age, the color ranges from dark red to pale pink. Pigment production is dependent on specific growth conditions, including m e d i u m composition, presence of certain ions and detergents, and temperature. It requires air, and the pigmentation is better developed w h e n Serratia cultures are incubated below 35°C or w h e n a low-phosphate agar without glucose (e.g., peptone-glycerol) is used. There is a strong tendency for clinical isolates to be nonpigmented and difficult to distinguish from other coliform organisms (Hejazi and Falkiner, 1997). Nonpigmented S. marcescens biotypes seem restricted to hospitalized patients, whereas pigmented biotypes are ubiquitous. In a 1978 review of the genus Serratia, three other species (in addition to S. marcescens) were recognized (Serratia liquefaciens, S. plymuthica, S. marinorubra), and a fourth, tentatively discussed as "strain 38" (Grimont and Grimont, 1978), was later named S. odorifera (Grimont et al., 1978). In the 1984 edition of Bergey's Manual of Determinative Bacteriology, S. marinorubra became S. rubidaea, and a sixth species, S. ficaria, was recognized (Holt and Krieg, 1984). In the second edition of The Prokaryotes, 10 species were mentioned and are presently known to belong in the genus Serratia (Grimont and Grimont, 1991). In addition to those already listed, the four other species are as follows: S. entomophila, S. fonticola, S. grimesii, and S. proteomaculans. Of these 10 species, only three--S, marcescens, S. plymuthica, S. rubidaea--produce prodigiosin (Grimont and Grimont, 1991). Bizio's patriotism in naming the first member of this genus as S. marcescens is admirable, and there is a pleasant euphony in the name; however, the steamboating physicist had nothing to do with the contaminated polenta. Bizio might have made a more relevant choice, for instance, the use of the peasant's name in whose house the "fungus" was found: Pittarella marcescens also has a fine ring to it! In fact, naming the red-pigmented bacterium became something of a cottage industry among early bacteriologists, with more than 20 names applied to this organism during the 100 years after Bizio's description. There was a tendency to retain the prodigious characteristic invented by

SEEING RED: THE STORYOF PRODIGIOSIN

9

Ehrenberg, probably in view of the supposed association with mira c l e s - f o r example, Bacillus prodigiosus and Bacterium prodigiosum. In 1920, the "final" report of The Committee on Classsification of the Society of American Bacteriologists recognized a possible priority for Erythrobacillus pyosepticus, which had been preserved as ATCC 275, and suggested the name Erythrobacillus prodigiosus (Grimont and Grimont, 1991). However, at that time, bacteriological nomenclature was governed by the International Botanical Code, which contained a priority principle requiring the oldest validly published name to be used. Erythrobacillus prodigiosus contradicted rules of priority and never gained acceptance outside the United States (Breed and Breed, 1924). It took the American bacteriologist Buchanan to apply the principle of priority and revive Serratia as the valid name. The first edition of Bergey's Manual of Determinative Bacteriology legitimized Bizio's priority more than a century after he had wished to honor Serrati (Bergey et al., 1923). It has been retained in subsequent editions of the Manual. Ironically, there is no proof that what is n o w called Serratia corresponds to Bizio's organism. The genus Serratia has the distinction in bacteriology of being outranked in age only by the genera Vibrio (1773) and Polyangium (1809). Even so, the acceptance of Serratia as a valid name has attracted considerable dissent. Specimens viewed by early microbiologists tended to be mixed cultures. The small size and morphological monotony of most bacteria provided few clues to the diversity of species. There was an unfortunate tendency to call all red-pigmented microorganisms Serratia simply because of their color. Red bacteria appearing on salted fish are a case in point. Such halophilic species do not ferment carbohydrates and are probably species of Halobacterium (Ayres et al., 1980). Several species of the yeast genus Rhodotorula form shiny, pink to red colonies on bread and other starchy foods and may also have been responsible for some of the incidents of bloody bread. In his delightful essay "Heretical Taxonomy for Bacteriologists," Cowan (1970) devoted a section to "The heresy of Serratia marcescens" referring to the uncertainty of applying names from the "pre-bacteriological era." Cowan felt that it was better to change the rules of nomenclature than to use pre-Pasteurian descriptions that were "in the modern sense, nothing short of the farcical" (Cowan, 1956). Later, in his posthumously published A Dictionary of Microbial Taxonomy, Cowan (1978) stated his opinion authoritatively: In my view it is a waste of time to try to find useful bacteriological information from observations made before bacteria were clearly distin-

10

J.w. BENNETT and RONALDBENTLEY

guished from algae, filamentous fungi and yeasts, and I believe we shall not lose anything by ignoring all work before the pioneer work of Pasteur. Nevertheless, many scientists, especially microbiologists, will have confidence that S. marcescens was responsible for most of the incidents involving blood on foodstuffs. There are other microorganisms forming pink or red colonies such as the yeasts Rhodotorula, Sporobolomyces salmonicolor, and Candida pulcherrima; the latter at least is ruled out since its pigment, pulcherrimin, is an iron complex and is insoluble in the usual organic solvents. Moreover, the characteristic dripping or flowing of S. marcescens cultures is not associated with the yeasts or fungi having red pigmentation. On the other hand, since other bacteria produce prodigiosin or prodigiosin-tike materials, some of the observed cases of blood-spotted food may have been due to organisms other than S. marcescens. The phenomenon was readily reproduced on "host bread" using S. marcescens cultures by Ehrenberg in the nineteenth century and more recently on polenta, unconsecrated Communion wafers (both Catholic and Protestant), and not-for-Passover matzos (Karp, 1988; Cullen, 1994); Protestant wafers gave the best results (Bennett, 1994). One of us has pointed out the lack of an appropriate control in these experiments; simple crackers or preservative-flee bread without religious significance should have been included! (Bentley, 1997). It may be noted that there have been reports of red spots on the cream layer of milk (Grimont and Grimont, 1978). Moreover, the range of materials subject to the development of red-spotted areas has been extended by the discovery of a "red spot disease" on culture beds of the kelp Laminaria japonica cultivated in the ocean around Hokkaido and used in the production of "makonbu" (Sawabe et al., 1998). The dried kelp, more colloquially known as "konbu," is usually used to flavor broth and soups, being then discarded. An aerobic, polarly flagellated, marine bacterium was identified as the causative agent of the red spot disease and the name Pseudoalteromonas bacteriolytica sp. nov. was proposed for it. It produces a prodigiosin-like pigment.

IV. Prodigiosin and Related Compounds A. STRUCTURES The major pigment of S. marcescens, originally named prodigiosine, was isolated in 1902, but a choice between alternate structural possibilities was not possible until its chemical synthesis was achieved in 1960, nearly a century and a half after the entrepreneurial hopes of Bizio and Sette had been dashed. Now termed prodigiosin, it is a very

SEEING RED: THE STORY OF PRODIGIOSIN

11

typical secondary metabolite. Prodigiosin, C20H25N3O, has an unusual structure with three pyrrole rings and is a pyrryldipyrrylmethene (Fig. 1A); two of the rings are directly linked to each other, and the third is attached by way of a methene bridge (Gerber, 1975; Williams and Qadri, 1980). It forms lustrous, square pyramidal crystals that are dark red with a green reflex; the hydrochloride C20H26C1N3 O forms crystals with a magenta color. The highly conjugated system of seven double bonds presumably accounts for the intense pigmentation. Secondary metabolites related to prodigiosin have been isolated from several bacterial genera. These related materials are frequently difficult to purify. Moreover, there has been considerable confusion with respect to naming them; to some extent, "prodigiosin" is used in the literature in a generic sense to include a family of similar materials. In devising trivial names for a group of related compounds it is useful to define a basic nucleus. Two such possibilities have been used for the prodigiosin-like materials (Gerber, 1975). The completely stripped down nucleus, devoid of all substituents, is termed "prodigiosene," while the portion common to most of the natural products, and containing a 6-methoxy substituent, is termed "prodiginine" (for structure and numbering, see Fig. 1A). The apparently bizarre numbering in which three carbons of the bipyrrole and one carbon of the monopyrrole are not numbered was devised "because substitutions on them would destroy the basic linear tripyrrole structure of prodigiosene" (Williams, 1973). Hence, prodigiosin could also be referred to as either 2-methyl-3-pentylprodiginine or 2-methyl-3-pentyl-6-methoxyprodigiosene (in some early papers, amyl is used for the preferred pentyl). Unfortunately, there has been no general agreement concerning the use of either prodigiosene or prodiginine, and, as will be seen, confusion has inevitably arisen with attempts to base nomenclature on prodigiosin itself. While there are advantages to the use of prodigiosene in connection with chemical syntheses, we believe that the use of prodiginine, as suggested by Gerber (1975) is the best solution. The few natural materials containing OH instead of O C H 3 at position 6 can conveniently be termed norprodiginines (Fig. 1A). Four structural types based on the prodiginine nucleus can be recognized: 1. Only straight chain alkyl substituents present: 1A. Alkyl substituents at both positions 2 and 3 (Fig. 1A). The prototype is prodigiosin itself with a methyl group at position 2 and a pentyl group at position 3. Higher homologues with methyl at position 2 and either hexyl or heptyl at position 3

12

J.W. BENNETT and RONALD BENTLEY

OCH 3

H

H A

/~4

H

IN R

B

H

R2 FIG. 1A,B FIG. 1. Prodigiosin and related compounds.

Structure 1A Prodigiosene Prodiginine Norprodigiosin Prodigiosin Undecylnorprodiginine Undecylprodigininea

R1 H CH30 HO CH30 HO CH30

R2 H H CH3 CH 3 CH3(CH2)lO CH3(CH2)lo

R3 H H CH3(CH2) 4 CH3(CH2) 4 H H

Structure 1B Metacycloprodigiosin

(ethyl-meta-cyclononylprodiginine) Butyl-meta-cycloheptylprodiginine has a similar meta structure

b u t w i t h only seven - - C H 2 - in the ring a n d a b u t y l substituent. Structure 1C Cycloprodigiosin hydrochloride Structure 1D C y c l o n o n y l p r o d i g i n i n e ; R -- H, n = 8 C y c l o m e t h y l d e c y l p r o d i g i n i n e ; R = CH3--, n :- 9 Tautomeric arrangements of the double b o n d systems are possible. One e x a m p l e is s h o w n in structure lB. aThis material is also referred to as u n d e c y l p r o d i g i o s i n a n d prodigiosin-25C (see text).

SEEING RED: THE STORY OF PRODIGIOSIN

13

OCH 3

OCH 3

C CH3

O~'.- (CH2)n -~ FIG. 1C,D

lB.

(along with prodigiosin) have been isolated from Pseudomonas magnesiorubra, the marine psychrophilic bacterium Vibrio psychroerythreus, and a sewage bacterium (Gerber, 1975). A river bacterium, Rugamonas rubra gen. nov., sp. nov., produces prodigiosin and (probably) the heptyl homologue (Austin and Moss, 1986). Norprodigiosin (2-methyl-3-pentylnorprodiginine) is formed by the S. marcescens mutant OF. Frequently, prodigiosin occurs as an adduct with macromolecules, typically protein. Alkyl substituents at position 2 only. Prodigiosin-like materials with an undecyl chain at position 2 were first fully characterized from certain actinomycetes (e.g., Streptomyces longisporus tuber) in 1966 (Wasserman et al., 1966; Harashima et al., 1967) and with a nonyl sidechain from Actinomadura madurae (formerly Nocardia madurae) (Gerber, 1975). The first is well known and has often been termed "undecylprodigiosin"--which would imply prodigiosin itself with an extra undecyl group. Another name, "prodigiosin-25 C" (or prodigiosin 25-C), is based on the total number of carbon atoms present with a designation of the chronological order of discovery (Harashima eta]., 1967). Thus, this name is meant to indicate that it is the third C25 material discove r e d - b u t is again not helpful in structural terms. Both of

14

J.w. BENNETT and RONALDBENTLEY these terms are used in the current literature, as is the abbreviation Red, by molecular geneticists. Confusion can be avoided by naming this material as undecylprodiginine, as suggested by Gerber (1975), and this name will be used here. As with prodigiosin, a material with - - O H at position 6 is known, and is best named as undecylnorprodiginine.

2.

3.

Ring formation between positions 2 and 4 (i.e., a meta arrangement). A structure with a cyclononyl ring linked to positions 2 and 4 and carrying an additional ethyl substituent (Fig. 1B) was isolated from Streptomyces longisporus tuber in 1969. Generally known as metacycloprodigiosin, a more appropriate name w o u l d be ethylmeta-cyclononylprodiginine. In biosynthetic terms, it appears to result from a cyclization of undecylprodiginine. Metacycloprodigiosin is probably identical with streptorubin A from Streptomyces rubrireticuli var. pimprina (Gerber, 1975). Some organisms, for example, Streptomyces hiroshimensis, produce both metacycloprodigiosin and undecylprodiginine; this is also true for an actinomycete isolated from leek roots and belonging to the Streptoverticillium baldaccii cluster (Brambilla et al., 1995). A material identified as butyl-meta-cycloheptylprodiginine has also been isolated from an actinomycete, strain B 4358 (Laatsch et al., 1991); it resembles ethyl-meta-cyclononylprodiginine, but the ring between positions 2 and 4 contains only seven - - C H 2 - - groups and carries a butyl substituent instead of an ethyl. Butyl-meta-cycloheptylprodiginine is probably identical to the material previously identified as butylcycloheptylprodiginine (butyl-ortho-cycloheptylprodiginine, with ring formation at positions 2 and 3) (Gerber, 1975), and is also identical to the antibiotic streptorubin B isolated from Streptomyces roseoverticulatus var. albosporus (Gerber, 1975; F~rstner et al., 1998). It is likely to be another cyclized form of undecylprodiginine. Ring formation between positions 3 and 4 (i.e., an ortho arrangement) and with CH3 at position 2. The only example of this structural type is a cyclized form of prodigiosin itself, usually known as cycloprodigiosin; a more informative alternative to cycloprodigiosin w o u l d be to use the name 2-methyl (methyl-ortho-cyclobutyl)prodiginine. Originally isolated from a marine bacterium, Alteromonas ruber (Gerber and Gauthier, 1979), it was assigned an incorrect structure. A revision in 1983 (Gerber, 1983; Laatsch and Thomson, 1983) indicated that the ring contains three - - C H 2 - - and one ---CH3-CH-- group. Cycloprodigiosin was also found, together

SEEING RED: THE STORY OF PRODIGIOSIN

15

with prodigiosin itself, in the anaerobic marine bacterium Vibrio gazogenes (formerly Beneckea gazogenes) (Harwood, 1978; Gerber, 1983; Laatsch and Thomson, 1983) and as its hydrochloride (Fig. 1C) in Pseudoalteromonas denitrificans, a novel marine bacterium, isolated "from the sea near Japan" (Kawauchi et al., 1997). 4. Ring formation between position 2 of the monopyrryl unit and position 10 of the dipyrryl unit. Compounds described generically as "macrocyclic prodiginines" have been isolated from Actinomadura pelletieri (formerly Nocardia pelletieri) and Actinomadura madurae (formerly N. madurae) (Gerber, 1975). These structures contain a bridge with several --CH2-- groups between the first and third pyrrole rings (Fig. 1D). Again, the nomenclature is a problem since the structure of Figure 1D with R = H and n = 8 is named as "cyclononylprodiginine"; there could easily be confusion with the previously described ethyl-meta-cyclononylprodiginine. These "macrocyclic prodiginines" are apparently unique to the two organisms named. Finally, there has been an unfortunate nomenclature confusion between the red pigment prodigiosin and a material referred to in Russian literature as "prodigiosan." The latter is a polysaccharide or lipopolysaccharide also isolated from S. marcescens. In searching Biological Abstracts Online for "prodigiosin," some abstracts containing "prodigiosan" were obtained. Moreover, in some cases, the Russian word for prodigiosan was translated as prodigiosin--to take only one example, "activation of mononuclear phagocytes by a lipopolysaccharide (prodigiosin)" (Panin et aI., 1996). B. BIOSYNTHESIS

How is the strange pyrryldipyrrylmethene structure present in prodigiosin and related compounds constructed by bacteria? Isotope tracer studies with both stable and radioactive isotopes were undertaken almost half a century ago at a time when only prodigiosin itself was known and w h e n it was believed (incorrectly) to be a tripyrrylmethene; one of us still remembers carrying out 15N assays with a mass spectrometer for this kind of work (Hubbard and Rimington, 1950). This and other studies indicated important roles for acetic acid and glycine in prodigiosin biosynthesis, and later work additionally implicated proline, serine, and alanine. Owing to difficulties in carrying out chemical degradations of labeled samples to determine location of isotopes, progress was delayed until NMR methods became available; for a summary, see Gerber et al. (1978).

16

J.w. BENNETT and RONALD BENTLEY

It was eventually learned that the A and B rings (i.e., the dipyrryl unit; see Fig. 1A) of prodigiosin, undecylprodiginine, and metacycloprodigiosin were constructed similarly, but the C ring of prodigiosin was formed differently from that of undecylprodiginine and metacycloprodigiosin. In all cases, the methyl of the --OCH3 group at position 6 derived from S-adenosylmethionine. These results were complemented by extensive studies with mutants not producing prodigiosin (Mody et al., 1990). Thus, one such mutant, mutant 933, produced methoxybipyrrolecarboxaldehyde, MBC (Fig. 2A), and mutants Wl and WF produced methylpentylpyrrole (Fig. 2B), abbreviated MAP (from the early designation as methylamylpyrrole). If mutant 933 was supplied with MAP, prodigiosin was formed; similarly, Wl and WF produced prodigiosin if supplied with MBC. This process is referred to as "syntropic pigmentation." Initially, three mutant classes were described from nonpigmented clinical strains of S. marcescens (Ding and Williams, 1983); subsequently, five more have been recognized (Mody et al., 1990). Moreover, "nonnatural" monopyrroles could be added to 933, thus leading to novel prodiginine pigments (e.g., the use of 2,4-dimethylpyrrole gave a pigment with methyls at both positions 2 and 4). When pigmented strains are grown at 37-40°C, pigment is no longer formed. These elevated temperatures apparently interfere only with production of MBC since syntropic pigmentation occurred in most cases when it was added (Katz and Sobieski, 1980). The bipyrryl unit is invariable for the entire range of prodiginine pigments. The A ring is formed from the four noncarboxyl carbons and the nitrogen atom of proline. The B ring contains the proline carboxyl carbon, one acetate unit, and two carbons and one nitrogen from serine (the serine carboxyl is lost). For prodigiosin itself the monopyrryl unit (ring C) is constructed from two carbons and the nitrogen atom of alanine (the carboxyl of this amino acid also being lost) and a tetraketide unit from four acetate units (Fig. 2B). The monopyrryl unit for formation of undecylprodiginine derives from two carbons and the nitrogen of glycine (again a decarboxylation takes place) and a heptaketide unit from seven acetates (Fig. 2C}. A condensation of glycine with other polyketides can account for formation of the other actinomycete prodiginines. An important observation, made in 1985, was that undecylprodiginine, along with other prodigiosin-like materials, was produced by Streptomyces coelicolor (Tsao et al., 1985). Prodigiosin itself had been recognized earlier in an actinomycete (Perry, 1961). However, M. Bibb (personal communication) has suggested that this identification (based largely on a u.v. spectrum) was incorrect and that Perry's material was

SEEING RED: THE STORY OF PRODIGIOSIN

17

OCH3

A

CHO

+ PRO

H

~

//~--CH20H H2N COOH SER COOH •

H2 N

NH2

HOOC'~ / CH3 ALA

R'

HN

GLY

•

CH3

R"

B Ri:

_

ri

LC. -

C •

G

•

RII

--

LOH2-OH2JuOH3

r I

|

"1

•

FIc. 2. Biosynthesis of the two components required to form prodigiosin and undecylprodiginine. Standard three-letter abbreviations are used for amino acids. In all cases, filled-in squares denote carbon derived from the methyl of acetate and filled in circles denote carbon derived from the carboxyt of acetate. In actuality, the polyketide units are most likely formed from acetate plus polymalonate condensations. The arrows at the top of the monopyrrole units indicate the molecular position involved in reaction with the aldehyde group of MBC (structure A). A. The methoxybipyrrolecarboxaldehyde, MBC, is used for prodigiosin itself as well as undecylprodiginine and related compounds. It is constructed from all five carbons and the nitrogen of proline (pro), a single acetate unit, and two carbons and the nitrogen of serine (ser; the ser COOH is lost as CO2). B. The methylpentylpyrrole, MAP, required for prodigiosin formation is constructed from two carbons and the nitrogen of alanine (ala; the ala COOH is lost as CO2) and a tetraketide unit (i.e., eight carbons) formed from acetate. C. The undecylpyrrole required for undecylprodiginine is constructed from one carbon and the nitrogen of glycine (gly; the gly COOH is lost as CO2) and a heptaketide (i.e., 14 carbons) formed from acetate.

18

J.w. BENNETTand RONALDBENTLEY

actually undecylprodiginine. Streptomyces coelicolor also produces the aromatic polyketide actinorhodin and has been much used in molecular genetics studies (Cane, 1997). Undecylprodiginine played an important role in the first cloning of a gene, playing a defined role in the biosynthesis of an antibiotic; an O-methyltransferase gene was isolated by complementation and the color of undecylprodiginine was used as the selectable phenotype (Feitelson and Hopwood, 1983). The enzymatic product converted undecylnorprodiginine to undecylprodiginine; furthermore, two forms of the necessary enzyme, undecylnorprodiginine--S-adenosylmethionine O-methyltransferase--were detected, one with a very high molecular mass peak (Feitelson et al., 1985). The genes involved in actinorhodin and prodigiosin biosynthesis in Streptomyces coelicolor A3(2) have received very extensive investigation (see reviews by Hopwood et al., 1995; Bibb, 1996) and continue to be of interest (Chakraburtty and Bibb, 1997; White and Bibb, 1997; Guthrie et al., 1998). The regulation of prodigiosin biosynthesis is complex, being influenced by increased glucose levels and decreased by increased phosphate levels (Gyun-Kang et al., 1998). The S. marcescens genes encoding prodigiosin biosynthesis from the necessary mono- and bipyrryl units have been cloned and expressed in E. coli (Dauenhauer et al., 1984). No E. coli recombinants encoded the entire prodigiosin biosynthetic pathway. However, strain SAD400 produced prodigiosin when supplied with MBC by S. marcescens 933, while strain SAD757 required both MBC (from mutant 933) and MAP (from mutant WF). Clearly, SAD400 could form MAP, and both of the E. coli recombinants could condense the two portions of the molecule together. One strain of S. marcescens, ATCC 39006, has the unusual property of producing both prodigiosin and the [3-1actam antibiotic carbapenem (Thomson et al., 1997). Mutants defective in the production of these secondary metabolites had a mutation in a gene termed rap (for regulation of antibiotic and l~igment). It appears that this and related genes (e.g., in Erwinia and Yersinia) form a subfamily of proteins regulating diverse aspects of bacterial physiology. V. From Saprophyte to Pathogen For at least a century, S. marcescens was regarded as a harmless saprophyte. In fact, this pigmented bacterium was used extensively as a marker organism; generations of bacteriology students remember demonstrations of how a simple handshake can transmit microorganisms

SEEING RED: THE STORYOF PRODIGIOSIN

19

from one individual to another. More recently, S. m a r c e s c e n s was used as a test organism with pigskin as a substrate to evaluate topical antimicrobial action. Simulated handwashing protocols were evaluated in parallel with the i n - v i t r o model (McDonnell et al., 1999). In one dramatic public experiment to demonstrate the spread of microorganisms, the intrepid Dr. M. H. Gordon in 1906 gargled with an S. m a r c e s c e n s culture before reciting Shakespeare to the House of Commons. No MPs were present, but Petri dishes with an appropriate culture medium were placed at various distances (Yu, 1979). Recovery of red bacterial colonies demonstrated the role of speaking and coughing in spreading bacteria. Present-day microbiologists w o u l d not consider a repetition of the experiment since S. m a r c e s c e n s is now recognized as an opportunistic pathogen (see later). Another interesting prodigiosin story is the fascinating "red diaper syndrome" in Wisconsin of 1958 (Waisman and Stone, 1958). The child of a genetics professor, born uneventfully in the University of Wisconsin hospital, was brought home and a diaper delivery service was hired. The first pickup showed apparent blood stains on the diapers. Upon further examination, the child's urine and stool were normal; the bloody color developed only after the diapers were stored in a bin. The father, familiar with an abnormal tryptophan metabolism causing a "blue diaper syndrome," suspected a genetic abnormality. Eventually, of course, the guilty culprit turned out to be a strain of pigmented S. m a r c e s c e n s being used in the medical school to study aerosol techniques and genetics. The infant's intestines were heavily infected with the organism; use of sulfasuxidine and a controlled diet eventually restored a "normal" intestinal flora, but almost a year was required. Prodigiosin-producing S. m a r c e s c e n s has also been used as a marker organism in germ warfare research (Yu, 1979). In n o w notorious experiments conducted between 1950 and 1966, S. m a r c e s c e n s cultures were released by the U.S. Army on an unsuspecting population of involuntary and unwitting subjects in the New York City subways, and in locations in Calhoun County (Alabama), Key West (Florida), and San Francisco. In the Pacific Coast experiments, ships released cultures into the ocean, whereupon an aerosol was formed by wave action. Red-pigmented bacteria were recovered in air samples some 80 meters inland. When these secret experiments were finally acknowledged, the Army maintained that there had been no infections attributable to them. However, a documented outbreak of S. m a r c e s c e n s infections did occur in a San Francisco hospital in 1950-51 at the time of the aerosol experiments. One patient died from the first known case of serratial endocarditis, and his family sued the Department of Defense. While it

20

J.w. BENNETT and RONALD BENTLEY

was tempting to link the two events, later work by the Centers for Disease Control in Atlanta indicated that in 100 cases of S. m a r c e s c e n s infections in the United States, none had been caused by an organism with the same serotype and biotype as that used by the Army. The legal judgment was that the San Francisco case was probably the first in a series of new nosocomial infections rather than a consequence of the Army's program. Prior to about 1970, cases of serratial bacteremia were very rare, but in a single hospital (Stanford University Hospital) from 1968 to 1977, some 76 cases were reported (Yu et aL, 1979). S. m a r c e s c e n s is now implicated in many serious conditions; the list includes empyema, lung abscess, meningitis, osteomyelitis, peritonitis, pneumonia, sinusitis, urinary tract infection, and wound infection (Hejazi and Falkiner, 1997; von Graevenitz, 1980; Daschiner, 1980). A dramatic instance of a nosocomial outbreak occurred in Nashville, Tennessee involving an antibiotic-resistant strain infecting patients at four separate hospitals. The drug-resistant strain was isolated from the urine of a catheterized patient in April of 1973, and by late 1974 the same strain (characterized by serotype, phage type, and antibiotic-resistance pattern) had been isolated in the other hospitals. All four institutions were teaching hospitals and had regular rotations of physicians and nursing staff. A total of 210 patients were infected, 21 of them becoming bacteremic, with 8 fatalities (Williams and Qadri, 1980). Later, the organism was isolated from pooled hand rinsings of personnel. Dr. Gordon would not have been surprised. In another case, reported in 1996, an outbreak of S. m a r c e s c e n s occurred in a neurosurgery intensive care unit (Bosi et al., 1996). The responsible strain was located in a diluted hexetidine solution used as a mouthwash; the bottle of diluted antiseptic was the single source of this nosocomial outbreak. Other disinfectants (hexachlorophene, benzalkonium chloride) can also become contaminated (Yu, 1979). Medical equipment has also been implicated in the spread of S. m a r c e s c e n s in hospitals; a bizarre case involved shaving brush bristles used for personal grooming in an intensive care unit. Yet another problem arises with the use of solutions of chlorhexidine for disinfecting contact lenses. In one study, 11 of 12 strains of S. m a r c e s c e n s became adapted to the agent (Gandhi et al., 1993). Contaminated lens solutions have been associated with ocular infections (Mayo et al., 1987). The antibiotic resistance of many strains of S. m a r c e s c e n s is a serious problem (Yu, 1979; Yu et al., 1979; Farrar, 1980), with rapid horizontal transfer of drug resistance by plasmids. Although transfer to E. coli K12

SEEING RED: THE STORY OF PRODIGIOSIN

21

was inefficient, it was more effective to Klebsiella p n e u m o n i a e (Hedges, 1980). Clearly, new methods for controlling S. m a r c e s c e n s infections would be welcome. It has been suggested that a possible chemotherapeutic target might be the bacterial regulatory proteins formed by the rap gene, which were described earlier (Thomson et al., 1997). Cases of serratial endocarditis have been reported among heroin addicts in San Francisco; they were associated with a high frequency of embolic complications and a refractoriness to medical therapy. Perhaps the increased frequency with which serratial infections have been observed relates to the heavy use of antibiotics in the 1950s and the consequent development of drug-resistant strains. The absence of cases before the antibiotic era is striking. Some veterinary problems have been associated with Serratia species, and S. m a r c e s c e n s has been frequently recovered from healthy, diseased, and dead insects; both S. m a r c e s c e n s and S. liquefaciens are classified as potential insect pathogens (Grimont and Grimont, 1978). A possible role for S. m a r c e s c e n s in septic abortions in cows and buffaloes has been described (Das et al., 1988); most of the isolated strains produced prodigiosin.

VI. Biological Activity of Prodigiosin and Related Compounds A. POSSIBLE ECOLOGICAL FUNCTIONS

As typical secondary metabolites, prodigiosin and related materials have no clearly defined physiological functions in the producing organisms. However, it is possible that pigmented S. m a r c e s c e n s may have an advantage in ecological dispersion (Burger and Bennett, 1985). In studies of the drops produced by bursting air bubbles rising through bacterial suspensions, pigmented strains were enriched in the drops (relative to that of the bulk suspension) (Burger and Bennett, 1985; Syzdek, 1985). The pigmented cells appeared to have increased hydrophobicity, possibly due to the presence of prodigiosin. It was acknowledged, however, that the cell enrichment was a complex chain of events and was influenced by cultural conditions (Syzdek. 1985). Other workers indicated that clinical S. m a r c e s c e n s strains had hydrophobic properties in the absence of prodigiosin and that hydrophobicity was only shown by growth at 30°C but not at 37°C (Rosenberg et al., 1986). Pigment is not synthesized at the higher temperature. It seems clear that the cell-surface hydrophobicity of S. m a r c e s c e n s is not totally due to

22

J.W. BENNETT and RONALD BENTLEY

surface pigment. Nonpigmented cells contained an additional protein (Mr = 40,000) that may be responsible for the higher surface hydrophobicity of some nonpigmented mutants (Mallick, 1996). Color variation in Serratia has been correlated with amount of flagellar antigens, and there is an apparent coregulation of pigment and flagellin synthesis. Variation of these surface antigens may allow pathogenic stains to evade host immune systems (Paruchuri and Harshey, 1987). In a very detailed study by Van der Mei et al. (1992), S. marcescens strains were characterized by contact angle and zeta potential measurements, X-ray photoelectron spectroscopy, and infrared spectroscopy. Again, it appeared that the presence of prodigiosin did not influence the cell-surface hydrophobicity. It was suggested that the pigment was confined in deeper layers than those probed by contact angles (about 0.3-0.5 nm). Other results indicated that both pigmented and nonpigmented strains produced extracellular vesicles and had wetting activity when grown at 30°C, but not at 37°C. The wetting activity was probably important for spreading cells on the surfaces of porous or fibrous materials, especially those with hydrophobic properties (Matsuyama et al., 1986). Finally, the presence of the O-antigen may be important in adhesion of S. marcescens to plastic and glass and to human uroepithelial cells (Palomar et al., 1995).

B. PHARMACOLOGICAL ACTIVITY

In a review of pre-penicillin antibiosis, Abraham and Florey (1949) cited a report that "complete inhibition" of cholera Vibrio was achieved by an old culture of Micrococcus prodigiosus, an early name for S. marcescens, and gave instances of the antagonistic properties of "Chromobacterium prodigiosum" (i.e., S. marcescens) against other bacteria, trypanosomes (the "nagana trypanosome [Prowazek]"), protozoa, and fungi. Prodigiosin itself has been identified as having extremely broad antibiotic properties, being active against Gram-positive bacteria, protozoa, and pathogenic fungi. It has been used experimentally as a fungistatic/fungicidal agent against Coccidioides immitis, but, unfortunately, water-soluble solutions of the glutamic acid form caused venous sclerosis on injection (Williams and Hearn, 1967). Strains of Serratia plymuthica producing prodigiosin and other materials with antifungal properties were beneficial rhizobacteria for oilseed rape (Kalbe et aL, 1996).

SEEING RED: THE STORY OF PRODIGIOSIN

23

To some extent, prodigiosin prolongs the lives of mice infected with the malaria parasite Plasmodium berghei on subcutaneous administration in peanut oil (Castro, 1967). The macrocyclic structures cyclononylprodiginine (Fig. 1D, R -- H, n = 8) and cyclomethyldecylprodiginine (Fig. 1D, R = CH 3, n = 9) had activities comparable to that of prodigiosin itself, but undecylprodiginine was inactive (Gerber, 1975). At a time w h e n many malarial parasites are resistant to conventional treatment, there is an increased incentive to find new drugs. The antimalarial activity of prodigiosin makes it an attractive target for modern genetic and chemical manipulations. Prodigiosin showed significant cytotoxic activity against some cell cultures; it was particularly potent against P388 mouse leukemia (9 PS) with an IC50 (inhibitory concentration for 50% cell growth relative to untreated controls) of 3.7 x 10 -4 btg m1-1 (Boger and Patel, 1988). With L-1210 mouse lymphocytic leukemia, B16 mouse melanoma, and 9KB human epidermoid nasopharynx carcinoma, the ICso values were in the range 2.0-4.0 × 10 -2 btg m1-1. Within the last few years, a number of other interesting physiological activities have been associated with prodigiosin. In T-cell lymphoma YAC-1, undecylprodiginine strongly suppressed incorporation of [3H]acetate into the lipid fraction (Kataoka et al., 1995a). Using preparations from rat liver, undecylprodiginine had little or no inhibitory effect on fatty acid synthase, acetyl-CoA synthetase, and acetyl-CoA carboxylase. It appeared that undecylprodiginine perturbed permeation of acetate through the plasma membrane of YAC-1. Beginning in 1995, it was reported that undecylprodiginine uncoupled the vacuolar type H+-ATPase and inhibited vacuolar acidification in baby hamster kidney cells (Kataoka et al., 1995b; Ohkuma et al., 1998). The proton pump activity, but not the ATP hydrolytic activity, was inhibited in rat liver lysosomes, and glycoprotein processing was also suppressed. In addition, prodigiosin, metacycloprodigiosin, and undecylprodiginine uncouple the acidification mediated by F-type H +ATPases of both submitochondrial (rat liver) and E. coli inverted membrane vesicles (Konno eta]., 1998). Prodigiosins have an ionophoric nature and act as H+/C1- symporters (or OH-/C1- antiporters) in liposomes. Their uncoupling effect is likely due to their H+/C1- symport activity across biological membranes (Ohkuma et a]., 1998; Konno et al., 1998; Sato et al., 1998). The proton pumping activity of V-ATPase in osteoclastic cells is essential for bone resorption; undecylprodiginine and metacycloprodigiosin inhibition of acidification of vacuolar organ-

24

J.w. BENNETT and RONALD BENTLEY

elles suppresses parathyroid hormone-stimulated bone resorption (Woo et al., 1997). Undecylprodiginine (and other V-ATPase inhibitors) blocked the perforin-dependent cytotoxicity mediated by the CD8 ÷ cytotoxic T-cell clone; an acidic pH is needed to maintain the quantity and also the quality of perforin in the lytic granules (Togashi et al., 1997). The inhibitory effect of undecylprodiginine on acidification of intracellular organelles may account for some of its immunosuppressive activity (see later). In summary, these prodigiosins constitute a new group of useful probes for an analysis of vacuolar functions since their effects are selective to vacuolar pH in vivo with little or no effect on the cellular ATP level. The prodigiosin structure is small enough so that chemical modifications might lead to more active and less toxic compounds (Sato et al., 1998); a simple and elegant new synthesis of prodigiosins (D'Alessio and Rossi, 1996) and of a thienyl analogue of undecylprodiginine (D'Auria et al., 1999) have potential value in this connection. Other syntheses of prodigiosins have been summarized by F/_irstner et al. (1998). An even more remarkable property of prodigiosins is their immunosuppressive activity. Undecylprodiginine and metacycloprodigiosin, obtained from S t r e p t o m y c e s h i r o s h i m e n s i s , were found to be potent inhibitors of T-lymphocyte proliferation induced by concanavalin A and phytohemagglutinin, but were less suppressive against B-lymphocyte proliferation induced by lipopolysaccharide. There was little toxicity in mice (Nakamura et al., 1986, 1989). Cycloprodigiosin hydrochloride also had immunosuppressive properties (Kawauchi et al., 1997), and a recent paper described T-cell-specific immunosuppression by prodigiosin itself (Han et al., 1998). Metacycloprodigiosin was more effective in reducing splenic cytotoxic T-cell activity than in prolonging murine skin or cardiac allografts (Magae et al., 1996). It has also been stated that undecylprodiginine has a low therapeutic index in rats and is probably not selective for T-cell activation (Metcalfe et al., 1993). Since the action of prodigiosins is different from that of cyclosporin and rapamycin (Songia et al., 1997; Tsuji et al., 1990, 1992), it has been suggested that prodigiosin and related compounds may be useful as supplementary immunosuppressants in combined therapy (Kawauchi et al., 1997; Songia et al., 1997; Tsuji et al., 1992; Lee et al., 1998). If prodigiosin or a prodigiosin analogue emerged as a useful antimalarial drug or as an immunosuppressant in human therapy, it would provide a happy ending (or a bright new beginning) to the singular story of this already wondrous secondary metabolite.

SEEING RED: THE STORY OF PRODIGIOSIN

25

VII. Final Comments S. marcescens has played an important role in the history of bacterial taxonomy, in research on the transmission of bacterial aerosols, in the study of emerging nosocomial infections, and in the understanding of secondary metabolite biosynthesis. The prodigiosin pigments have intrigued organic chemists and pharmacologists, and may yet play roles in the treatment of infectious diseases such as malaria, and perhaps as immunosuppressant agents. However, a major reason for much of the continuing curiosity in the Serratia/prodigiosin story is the theory that these viscous, crimson bacterial colonies provide a naturalistic explanation for certain long-ago miracles involving the Eucharist, and that their appearance gives some credible explanation for the persistence of the antisemitic outrages associated with the "blood libel." The word miracle comes from the Latin miraculum, "a wonder, a marvel," and is related to the verb mirari, "to wonder, to be astonished, or amazed." There have been many extensive definitions, from the facetious to the profound. In ordinary language, the word miracle often conveys the idea of something wonderful, as in "the miracle of birth," or "penicillin is a miracle drug." To the religious community, a miracle is something more profound: an event that seems to contradict natural laws and that can be attributed to God or some other higher cause (Perschel and Perschel, 1988). Most scientists live by a creed that involves a belief that the world behaves according to the predictable laws of nature. Miracles are violations of natural laws; therefore, they do not occur. As George Santayana has written, "Miracles are propitious accidents, the causes of which are too complicated to be readily understood" (Santayana, 1930). Was the Miracle at Bolsena a divine event or merely a growth of S. marcescens? The answer to this question may be irrelevant; the crucial event (perhaps itself the miracle) was that the priest regained his faith (Vaclav, 1994). In like manner, standard accounts of the desecration of the Host, as told by microbiologists, grossly overemphasize the possible role of bacteriology, as if the invocation of a natural explanation can make medieval antisemitism more comprehensible. It is of course conceivable or even likely that prodigiosin or other microbial pigment may have played a role in some of the reports of blood on consecrated bread. If not, the dire consequences of the miracles must be kept in mind by those choosing to regard the incidents in religious terms (Isenberg, 1995).

26

J.W. BENNETT and RONALD BENTLEY

In general, too many scientists look back at these records with a smug, self-designated attitude of epistemological privilege. Ethicists, historians, philosophers, priests, rabbis, and other scholars and writers are better able to comprehend the mystery of human prejudice than are individuals using scientific facts and hypotheses alone. It is well to remember, in a paraphrase of Anatole France, that a person taking pride in being without prejudice has asserted a claim that is itself a very great prejudice--in his words, "He flattered himself on being a man without any prejudices; and this pretension itself is a very great prejudice" (France, 1938).

Acknowledgments We are grateful to Drs. F. and P. A. D. Grimont and Dr. Mervyn Bibb for reading the manuscript and for their very helpful suggestions. We thank Johanna Cullen and Ora Karp for sharing unpublished manuscripts with us, Drynda Johnston and Ann Rogers, Langley Library, University of Pittsburgh, for help in locating many references, and Mr. and Mrs. H. Katsuhisha for information on, and a sample of, "makonbu." Special thanks go to Scott Burger for stimulating our interest in prodigiosin, and to the Reverend Roger Boraas for guidance on the religious literature.

REFERENCES

Abraham, P., and Florey, H. W. (1949). Antibiotics from chromogenic bacteria. In "Antibiotics" (H. W. Florey, E. Chain, N. G. Heatley, M. A. Jennings, A. G. Sanders, E. P. Abraham, and M. E. Florey, eds.), Vol. 1, pp. 537-565. Oxford University Press, London. Alsberg, C. L., and Black, O. F. (1913). "Contribution to the Stndy of Maize Deterioration." U.S.D.A. Bureau of Plant Industry, Bulletin No. 270, Government Printing Office, Washington, DC. Anonymous (1999). A retrospective look at how far we have come. In "99th General Meeting Program Update," p. 1, American Society for Microbiology, Washington, DC. Austin, D. A., and Moss, M. D. (1986). Numerical taxonomy of red-pigmented bacteria isolated from a lowland river, with the description of a new taxon, Rugamonas rubra, new genus, and new species. J. Gen. Microbiol. 132, 1899-1910. Ayres, J. C., Mundt, J. O., and Sandine, W. E. (1980). "Microbiology of Foods." Freeman,

San Francisco. Bennett, J. W. (1994). More on the Miracle of Bolsena. Am. Soc. Microbiol. News 60, 403. Bentley, R. (1997). Secondary metabolites play primary roles in human affairs. Perspec. Biol. Med. 40, 197-221.

SEEING RED: THE STORY OF PRODIGIOSIN

27

Bergey, D. H., Harrison, F. C., Breed, R. S., and Huntoon, F. M. (1923). "Bergey's Manual of Determinative Bacteriology," 1st ed. Williams & Wilkins, Baltimore. Bibb, M. (1996). The regulation of antibiotic production in Streptomyces coelicolorA3(2). Microbiology 142, 1335-1344. Black, O. F., and Alsberg, C. L. (1910). "The Determination of the Deterioration of Maize, with Incidental Reference to Pellagra." U.S.D.A. Bureau of Plant Industry, Bulletin No. 199, Government Printing Office, Washington, DC. Blazevic, D. J. (1980). Taxonomy, isolation and identification of Serratia. In "The Genus Serratia" (A. yon Graevenitz and S. J. Rubin, eds.) pp. 4-13. CRC, Boca Raton, FL. Boger, D. L., and Patel, M. (1988). Total synthesis of prodigiosin, prodigiosene, and desmethoxyprodigiosin: Diels-Alder reactions of heterocyclic azadienes and development of an effective palladium(II)-promoted 2,2'-bipyrrole coupling procedure. J. Org. Chem. 53, 1405-1415. Bosi, C., Davin-Regli, A., Charrel, R., Rocca, B., Monnet, D., and Bollet, C. (1996). Serratia marcescens nosocomial outbreak due to contamination of hexetidine solution. J. Hosp. Infect. 33,217-224. Brambilla, U., Nasini, G., Petrolini, B., Quaroni, S., Saracchi, M., and Fedeli, L. (1995). Prodigiosin-like and other metabolites produced by a Streptoverticillium strain. Actinomycetes 6, 63-70. Breed, R. S., and Breed, M. E. (1924). The type species of the genus Serratia, commonly known as Bacillus prodigiosas. J. Bacteriol. 9, 545-557. Bulloch, W. (1938). "The History of Bacteriology." Oxford University Press, London. Burger, S. R., and Bennett, J. W. (1985). Droplet enrichment factors of pigmented and non-pigmented Serratia marcescens: Possible selective function for prodigiosin. Appl. Environ. Microbiol. 50, 487-490. Cane, D. E., ed. (1997). Polyketide and non-ribosomal polypeptide biosynthesis: A "Thematic Issue" of Chemical Reviews. Chem. Revs. 97, 2463-2706. Castro, A. J. (1967). Anti-malarial activity of prodigiosin. Nature 213,903-904. Chakraburtty, R., and Bibb, M. (1997). The ppGpp synthetase gene (relA) of Streptomyces coelicolor A3(2) plays a conditional role in antibiotic production and morphological differentiation. J. Bacteriol. 179, 5854-5861. Cowan, S. T. (1956). "Ordnung in das chaos" Migula. Can. J. Microbiol. 2, 212-219. Cowan, S. T. (1970). Heretical taxonomy for bacteriologists. J. Gen. Microhio]. 61, 145154. Cowan, S. T. (1978). "A Dictionary of Microbial Taxonomy," edited by L. R. Hill. Cambridge University Press, Cambridge. Crookshank, E. M. (1890). "Manual of Bacteriology," 3rd ed. J. H. Vail, New York. Cross, F. O., and Livingstone, E. A. (1974). The Eucharist. In "The Oxford Dictionary of the Christian Church," pp. 475-477. Oxford University Press, London. Cullen, J. C. (1994). The Miracle of Bolsena. Growth of Serratia on sacramental bread and polenta may explain incidents in medieval Italy. Am. Soc. Microbiol. News 60, 187-191. D'Alessio, R., and Rossi, A. (1996). Short synthesis of undecylprodigiosine [sic]: A new route to 2,2'-bipyrrolyl-pyrromethene systems. SYNLETT, pp. 513-514. Das, A. M., Paranjape, V. L., and Pitt, T. L. (1988). Serratia marcescens infections associated with early abortion in cows and buffaloes. Epidemiol. Infect. 101, 143-149. Daschner, F. D. (1980). The epidemiology of Serratia marcescens. In "The Genus Serratia" (A. von Graevenitz and S. J. Rubin, eds.), pp. 187-196. CRC, Boca Raton, FL.

28

J.W. BENNETT and RONALD BENTLEY

Dauenhauer, S. A., Hull, R. A., and Williams, R. P. (1984). Cloning and expression in Escherichia coli of Serratia marcescens genes encoding prodigiosin biosynthesis. J. Bacteriol. 158, 1128-1132. D'Auria, M., DeLuca, E., Mauriello, G., and Racioppi, R. (1999). A short synthesis of a thienyl analogue of undecylprodigiosin. Syn. Commun. 29, 35-42 Ding, M.-J., and Williams, R. P. (1983). Biosynthesis of prodigiosin by white strains of Serratia marcescens isolated from patients. J. Clin. Microbiol. 17, 476-480. Farrar, W. E. (1980}. Antimicrobial susceptibility of clinical isolates, synergistic effects, and [3-1actamases of Serrotia. In "The Genus Serratia" (A. von Graevenitz and S. J. Rubin, eds.), pp. 121-138. CRC, Boca Raton, FL. Feitelson, J. S., and Hopwood, D. A. (1983). Cloning of a Streptomyces gene for O-methyltransferase involved in antibiotic biosynthesis. Mol. Gen. Genet. !90, 394-398. Feitelson, J. S., Malpartida, F., and Hopwood, D. A. (1985). Genetic and biochemical characterization of the red gene cluster of Streptomyces coelicolor A3(2). J. Gen. Microbiol. 131, 2431-2441. France, A. (1938). "The Crime of Sylvestre Bonnard," p. 128. Trans. L. Hearn. The Bodley Head, London. (Original publication date, 1908.) Fiirstner, H., Szillat, H., Gabor, B., and Mynott, R. (1998). Platinum- and acid-catalyzed enyne metathesis reactions: Mechanistic studies and applications to the syntheses of streptorubin B and metacycloprodigiosin. J. Am. Chem. Soc. 120, 8305-8314. Gandhi, P. A., Sawant, A. D., Wilson, L. A., and Ahearn, D. G. (1993). Adaptation and growth of Serratia marcescens in contact lens disinfectant solutions containing chlorhexidine gluconate. Appl. Environ. Microbiol. 59, 183-188. Gaughran, E. R. L. (1969). From superstition to science: The history of a bacterium. Ann. N. Y Acad. Sci. 31, 3-24. Gerber, N. N. (1975). Prodigiosin-like pigments. Crit. Rev. Microbiol. 3,469-485. Gerber, N. N. (1983). Cycloprodigiosin from Beneckea gazogenes. Tetrahedron Lett. 24, 2797-2798. Gerber, N. N., and Gauthier, M. J. (1979). New prodigiosin-like pigments from Alteromonas rubra. Appl. Environ. Microbiol. 37, 1176-1179. Gerber, N. N., McInnes, G. A., Smith, D. G., Waiters, J. A., Wright, J. L. C., and Vining, L. C. (1978). Biosynthesis of prodiginines: 13C resonance assignments and enrichment patterns in nonyl-, cyclononyl-, methylcyclodecyl- and butylcycloheptylprodigiosin produced by actinomycete cultures supplemented with laC-labeled acetate and 15Nlabeled nitrate. Can. J. Chem. 56, 1155-1163. Grimont, F., and Grimont, P. A. D. (1991). The genus Serratia. In "The Prokaryotes" (A. Balows, H. G. Triiper, M. Dworkin, W. Harder, and K.-H. Schleifer, eds.), Vol. 3, pp. 2822-2848. Springer-Verlag, New York. Grimont, P. A. D., and Grimont, F. (1978). The genus Serratia. Annu. Rev. Microbiol. 32, 221-248. Grimont, P. A. D., and Grimont, E (1984). In "Bergey's Manual of Systematic Bacteriology" (N. R. Krieg and J. G. Holt, eds.), pp. 477-484. Williams & Wilkins, Baltimore. Grimont, P. A. D., Grimont, F., Richard, C., Davis, B. R., Steigerwalt, A. G., and Brenner, D. J. (1978). Deoxyribonucleic acid relatedness between Serratia plymuthica and other Serratia species, with a description of Serratia odorifera sp. nov. (type strain: 1CPB 3995). Int. J. Syst. Bacteriol. 28, 453-463. Guthrie, E. P., Flaxman, C. S., White, J., Hodgson, D. A., Bibb, M. J., and Chater, K. F. (1998). A response-regulator-like activator of antibiotic synthesis from Streptomyces coelicolor A3(2) with an amino-terminal domain that lacks a phosphorylation pocket. Microbiology 144, 727-738. [for erratum see Microbiology 144, 2007.]

SEEING RED: THE STORY OF PRODIGIOSIN

29

Gyun-Kang, S., Jin, W., Bibb, M., and Lee, K. J. (1998). Actinorhodin and undecylprodigiosin production in wild-type and relA mutant strains of Streptomyces coelicolor A3(2) grown in continuous culture. FEMS Microbiol. Lett. 168, 221-226. Han, S. B., Kim, H. M., Kim, Y. H., Lee, C. W., Jang, E.-S., Son, K. H., Kim, S. U., and Kim, Y. K. (1998). T-cell specific immunosuppression by prodigiosin isolated from Serratia marcescens. Int. J. Immunopharm. 20, 1-13. Harashima, K., Tsuchida, N., Tanaka, T., and Nagatsu, J. (1967). Prodigiosin-25C: Isolation and the chemical structure. Agric. Biol. Chem. 31,481-489. Harrison, E C. (1924). The miraculous micro-organism. Trans. Roy. Soc. Can., Sec. V, Set. III 18, 1-17. Harwood, C. (1978). Beneckea gazogenes sp. nov., a red, facultatively anaerobic, marine bacterium. Curr. Opin. Microbiol. 1,233-238. Hedges, R. W. (1980). R factors of Serratia. In "The Genus Serratia" (A. von Graevenitz and S. J. Rubin, eds.), pp. 139-153. CRC, Boca Raton, FL. Hejazi, A., and Falkiner, E R. (1997). Serratia marcescens. J. Med. Microbiol. 46, 903-912. Holt, J. G., and Krieg, N. R., eds. (1984). "Bergey's Manual of Systematic Bacteriology." Williams & Wilkins, Baltimore. Hopwood, D. A., Chater, K. F., and Bibb, M. (1995). Genetics of antibiotic production in Streptomyces coelicolor A3(2), a model streptomycete. Biotechnology 28, 65-102. Hubbard, R., and Rimington, C. (1950). The biosynthesis of prodigiosin, the tripyrrylmethene pigment from Bacillus prodigiosus (Serratia marcescens). Biochem. J. 46, 220-225. Isenberg, H. D. (1995). The other side of Serratia "miracles." Am. Soc. Microbiol. News 61, 155. Kalbe, C., Marten, P., and Berg, G. (1996). Strains of the genus Serratia as beneficial rhizobacteria of oilseed rape with antifungal properties. Microbiol. Res. 151,433-439. Karp, O. B. (1988). "The Color of Blood." Unpublished honors thesis, Tulane University, New Orleans. Kataoka, T., Magae, J., Yamasaki M., and Nagai, K. (1995a). Prodigiosin 25-C perturbs permeation of acetate in a cultured cell line. Biosci. Biotechnol. Biochem. 59, 18911895. Kataoka, T., Muroi, M., Ohkuma, S., Waritani, T., Magae, J., Takatsuki, A., Kondo, S., Yamasaki M., and Nagai, K. (1995b). Prodigiosin 25-C uncouples vacuolar type H÷-ATPase, inhibits vacuolar acidification and affects glycoprotein processing. FEBS Lett. 359, 53-59. Katz, D. S., and Sobieski, R. J. (1980). Production of pigment precursors in Serratia marcescens at elevated temperatures. Trans. Kansas Acad. Sci. 83, 91-94. Kawauchi, K., Shibutani, K., Yagisawa, H., Kamata, H., Nakatsuji, S., Anzai, H., Yokoyama, Y., Ikegami, Y., Moriyami, Y., and Hirata, H. (1997). A possible immunosuppressant, cycloprodigiosin hydrochloride, obtained from Pseudoalteromonas denitrificans. Biochem. Biophys. Res. Commun. 237, 543-547. Konno, H., Matsuya, H., Okamoto, M., Sato, T., Tanaka, Y., Yokoyama, K., Kataoka, T., Nagai, K., Wasserman, H. H., and Ohkuma, S. (1998). Prodigiosins uncouple mitochondrial and bacterial F-ATPases: evidence for their H+/CI- symport activity. J. Biochem. 124, 547-556. Laatsch, H., and Thomson, R. H. (1983). A revised structure for cycloprodigiosin. Tetra° hedron Lett. 24, 2701-2704. Laatsch, H., Kellner, M., and Weyland, H. (1991). Butyl-meta-cycloheptylprodiginine: A revision of the structure of the former ortho-isomer. J. Antibiot. 44, 187-191.

30

J.W. BENNETT and RONALD BENTLEY

Lee, M.-H., Yamashita, M., Tsuji, R. F., Yamasaki, M., Kataoka, T., Magae, J., and Nagai, K. (1998). Suppression of T-cell stimulating function of allogenic antigen presenting cells by prodigiosin 25-C. J. Antibiot. 51, 92-94. Magae, J., Miller, M. W., Nagai, K., and Shearer, G. M. (1996). Effect of metacycloprodigiosin, an inhibitor of killer T cells on murine skin and heart transplants. J. Antibiot. 49, 86-90. Mallick, S. A. (1996). Cell-surface hydrophobicity and its relation to outer membrane proteins of Serratia marcescens. Indian J. Exp. Biol. 34, 107-110. Matsuyama, T., Murakami, T., Fujita, M., Fujita, S., and Yano, I. (1986). Extracellular vesicle formation and biosurfactant production by Serratia marcescens. J. Gen. Microbiol. 132, 865-875. Maurois, A. (1959). "The Life of Sir Alexander Fleming, Discoverer of Penicillin," 1st ed. Dutton, New York. Mayo, M. S., Schlitzer, R. L., Ward, M. A., Wilson, L. A., and Ahearn, D. G. (1987). Association of P s e u d o m o n a s and Serratia corneal ulcers with use of contaminated solutions. J. Clin. Microbiol. 25, 1398-1400. McDonnell, G., Haines, K., Klein, D., Rippon, M., Walmsley, R., and Pretzer, D. (1999). Clinical correlation of a skin antisepsis model. J. Microbiol. M e t h o d s 35, 31-35. Merlino, C. P. (1924). Bartolomeo Bizio's letter to the most eminent priest, Angelo Bellani, concerning the phenomenon of the red-colored polenta [translated from the Italian]. J. Bacteriol. 9, 527-543. Metcalfe, S., Ashley, N., Chen, Z., and Calne, R. Y. (1993). Prodigiosin 25 C: Effect in in vitro models for T cell activation and T cell cycling and in vivo for rat heart allografls. Int. Arch. Allergy I m m u n o l . 101, 132-135. Mody, R. S., Heidarynejad, V., Patel, A. M., and Dave, P. J. (1990). Isolation and characterization of Serratia m a r c e s c e n s mutants defective in prodigiosin biosynthesis. Curr. Opin. Microbiol. 20, 95-104. Nakamura, A., Nagai, K., Ando, K., and Tamura, G. (1986). Selective suppression by prodigiosin of the mitogenic response of murine splenocytes. J. Antibiot. 39, 11551159. Nakamura, A., Magae, J., Tsuji, R. F., Yamasaki, M., and Nagai, K. (1989). Suppression of cytotoxic T cell induction in vivo by prodigiosin 25-C. Transplantation 47, 10131016. Ohkuma, S., Sato, T., Okamoto, M., Matsuya, H., Arai, K., Kataoka, T., Nagai, K., and Wasserman, H. H. (1998). Prodigiosins uncouple lysosomal vacuolar-type ATPase through promotion of H+/C1- symport. Biochem. J. 334, 731-741. Palomar, J., Leranoz, A. M., and Vinas, M. (1995). Serratia marcescens adherence: The effect of O-antigen presence. Microbios 81, 107-113. Panin, L. E., Khar'kovskii, A. V., and Usynin, I. F. (1996). The role of the mononuclear phagocyte system in regulating protein biosynthesis in rat organs and tissues in entogeny. Ontogenez 27, 280-285. Paruchuri, D. K., and Harshey, R. M. (1987). Flagellar variation in Serratia marcescens is associated with color variation. J. Bacteriol. 169, 61-65. Perry, J. J. (1961). Prodigiosin in an actinomycete. Nature 191, 77-78. Perschel, E., and Perschel, E. R. (1988). Medical miracles from a physician-scientist's viewpoint. Perspec. Biol. Med. 31, 391-404. Reid, R. D. (1936). Studies on bacterial pigmentation, I: Historical considerations. J. Bacterio]. 31, 205-210.

SEEING RED: THE STORY OF PRODIGIOSIN

31

Rosenberg, M., Blumberger, Y., Judes, H., Bar-Ness, R., Rubinstein, E., and Mazor, Y. (1986). Cell-surface hydrophobicity of pigmented and nonpigmented clinical Serratia m a r c e s c e n s strains. Infect. I m m u n . 51,932-935. Santayana, G. (1930). Introduction to Benedictus Spinoza, Ethics a n d "De lntellectus E m e n d a t i o n e . " Trans. A. Boyle. Dent and Sons, London. (Original publication date, 1910.) Sato, T., Konno, H., Tanaka, Y., Kataoka, T., Nagai, K., Wasserman, H. H., and Ohkuma, S. (1998). Prodigiosin as a new group of H÷/C1- symporters that uncouple proton translocators. J. Biol. Chem. 273, 21455-21462. Sawabe, T., Makino, H., Tatsumi, M., Nakano, K., Tajima, K., Iqbal, M. M., Yumoto, I., Ezura, Y., and Christen, R. (1998). P s e u d o a l t e r o m o n a s bacteriolytica sp. nov., a marine bacterium that is the causative agent of red spot disease of Laminaria japonica. Int. f. Syst. Bacteriol. 48, 769-774. Scheurlen, E. (1896). Geschichtliche und experimentelle Studien fiber der Prodigiosus. Arch. Hyg. 26, 1-31. Songia, S., Mortellaro, A., Taverna, S., Fornasiero, C., Scheiber, E. A., Erba, E., Colotta, E, Mantovani, A., Isetta, A.-M., and Golay, J. (1997). Characterization of the new immunosuppressive drug undecylprodigiosin in human lymphocytes: Retinoblastoma protein, cyclin-dependent kinase-2, and cyclin-dependent kinase-4 as molecular targets. J. I m m u n o l . 158, 3987-3995. Syzdek, L. D. (1985). Influence of Serratia marcescens pigmentation on cell concentrations in aerosols produced by bursting bubbles. App]. Environ. Microbiol. 49, 173178. Thomson, N. R., Cox, A., Bycroft, B. W., Stewart, G. S. A. B., Williams, P., and Salmond, G. P. C. (1997). The Rap and Hor proteins of Erwinia, Serratia and Yersinia: A novel subgroup in a growing superfamily of proteins regulating diverse physiological processes in bacterial pathogens. Mol. Microbiol. 26, 531-544. Togashi, K.-i., Kataoka, T., and Nagai, K. (1997). Characterization of a series of vacuolar type H÷-ATPase inhibitors on CTL-mediated cytotoxicity. I m m u n o l . Lett. 55,139-144. Trachtenberg, J. (1943). "The Devil and the Jews: The Medieval Conception of the Jew and Its Relation to Modern Antisemitism." Yale University Press, New Haven. Tsao, S.-W., Rudd, B. A. M., He, X.-G., Chang, C.-J., and Floss, H. G. (1985). Identification of a red pigment from Streptomyces coelicolor A3(2) as a mixture of prodigiosin derivatives. J. Antibiot. 38, 128-130. Tsuji, R. E, Yamamoto, M., Nakamura, A., Kataoka, T., Magae, J., Nagai, K., and Yamasaki, M. (1990). Selective immunosuppression of prodigiosin 25-C and FK506 in the murine immune system. J. Antibiot. 43, 1293-1301. Tsuji, R. F., Magae, J., Yamashita, M., Nagai, K., and Yamasaki, M. (1992). Immunomodulating properties of prodigiosin 25-C an antibiotic which preferentially suppresses induction of cytotoxic T cells. J. Antibiot. 45, 1295-1302. Vaclav, J. (1994). Any less a miracle? A m . Soc. Microbiol. N e w s 60, 579. Van der Mei, H. C., Cowan, M. M., Genet, M. J., Rouxhet, P. G., and Busscher, H. J. (1992). Structural and physicochemical surface properties of Serratia marcescens strains. Can. J. MicrobioI. 38, 1033-1041. yon Graevenitz, A. (1980). Infection and colonization with Serratia. In "The Genus Serratia" (A. von Graevenitz and S. J. Rubin, eds.), pp. 167-186. CRC, Boca Raton, FL. Wainwright, M. (1990). "Miracle Cure: The Story of Penicillin and the Golden Age of Antibiotics." Basil Blackwell, Oxford.

32

J.W. BENNETT and RONALD BENTLEY

Waisman, H. A., and Stone, W. H. (1958). The presence of Serratia marcescens as the predominant organism in the intestinal tract of the newborn: The occurrence of the "red diaper syndrome." Pediatrics 21, 8-12. Wasserman, H. H., Rodgers, G. C., and Keith, D. D. (1966). The structure and synthesis of undecylprodigiosin. A prodigiosin analogue from Streptomyces. Chem. Commun., pp. 825-826. White, J., and Bibb, M. (1997). bldA dependence of undecylprodigiosin in Streptomyces coelicolor A3(2) involves a pathway-specific regulatory cascade. J. Bacteriol. 179, 627-633. Williams, R. P. (1973). Biosynthesis of prodigiosin, a secondary metabolite of Serratia marcescens. Appl. Microbiol. 25, 396-402. Wiiliams, R. P., and Hearn, W. P. (1967). Prodigiosin. In "Antibiotics," Vol. 2: "Biosynthesis" (D. Gottlieb and P. D. Shaw, eds.), pp. 410-432, 449-451. Springer-Vertag, New York. Williams, R. P., and Qadri, S. M. H. (1980). The pigment of Serratia. In "The Genus Serratia" (A. von Graevenitz and S. J. Rubin, eds.), pp. 31-75. CRC, Boca Raton, FL. Woo, J.-T., Ohba, Y., Tagami, K., Sumitani, K., Kataoka, T., and Nagai, K. (1997). Prodigiosin 25-C and metacyctoprodigiosin suppress the bone resorption by osteoclasts. Biosci. Biotechnol. Biochem. 61,400-402. Yu, V. L. (1979). Serratia marcescens: Historical perspective and clinical review. N e w Engl. J. Med. 300, 887-983. Yu, V. L., Oakes, C. A., Axnick, K. J., and Merigan, T. C. (1979). Patient factors contributing to the emergence of gentamicin-resistant Serratia marcescens [sic]. Am. J. Med. 66, 468-472.

Microbial/Enzymatic Synthesis of Chiral Drug Intermediates RAMESH N. PATEL Bristol-Myers Squibb Pharmaceutical Research Institute New Brunswick, New Jersey 08903

I. Introduction II. AntihypertensiveDrug: Vasopeptidase Inhibitor A. EnzymaticSynthesis of BMS-199541-01 B. EnzymaticSynthesis of L-6-Hydroxynorleucine C. EnzymaticSynthesisof AllysineEthyleneAcetal III. [~3-ReceptorAgonist A. MicrobialReduction of 4-Benzyloxy-3-Methanesulfonylamino2'-Bromoacetophenone B. EnzymaticResolution of Racemic c~-MethylPhenylalanineAmides C. AsymmetricHydrolysisof Racemic Methyl-(4-Methoxyphenyl)Propanedioic Acid, Ethyl Diester IV. AnticholesterolDrugs V. Deoxyspergualin VI. AntiviralAgents VII. StereoselectiveHydrolysisof Racemic Epoxide VIII. BiocatalyticDynamicResolution: Stereoinversion of Racemic Diol IX. Resolutionof Racemic SecondaryAlcohols X. Summary References

I. Introduction Currently much attention is being focused on the interaction of small molecules with biological macromolecules. The search for selective enzyme inhibitors and receptor agonists/antagonists is key for targetoriented research in the pharmaceutical and agrochemical industries. Increasing understanding of the mechanism of drug interactions on a molecular level has led to a strong awareness of the importance of chirality as the key to the efficacy of many drug products and agrochemicals. The production of optically active chiral intermediates is a subject of increasing importance in pharmaceutical industries. Increasing regulatory pressure to market homochiral drugs by the Food and Drug Administration has driven chemoenzymatic synthesis of chiral compounds. Organic synthesis has been one of the most successful 33 ADVANCES IN APPLIED MICROBIOLOGY, VOLUME 47 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved, 0065-2164/00 $25.00

34

RAMESH N. PATEL

scientific disciplines and has enormous practical utility. There have been many advances in organic synthesis, which have led to the synthesis of natural products, drugs, agricultural chemicals, polymers, and many classes of functional molecules. This raises the question of why biocatalysis? What does biocatalysis have to offer to synthetic organic chemists? Biocatalysis offers an added dimension and an enormous opportunity to prepare industrially useful chiral compounds. One major advantage of biocatalysis over chemical catalysis is that enzyme-catalyzed reactions are stereoselective and regioselective and can be carried out at ambient temperature and atmospheric pressure, which minimizes problems of isomerization, racemization, epimerization, and rearrangements that may occur during chemical processes. Biocatalytic processes catalyzed by microbial cells and the enzymes derived therefrom can be immobilized and reused for many cycles. In addition, enzymes can be overexpressed so as to make biocatalytic processes economically efficient. The ability to design biocatalysts that would act specifically in any desired reaction will change the face of synthesis. Tailor-made enzymes with modified activity and the preparation of thermostable and pH-stable enzymes produced by random and site-directed mutagenesis will lead to the production of novel stereoselective biocatalysts. The use of enzymes in organic solvents has led to hundreds of publications on enzyme-catalyzed asymmetric synthesis and resolution processes. Molecular recognition and selective catalysis are key chemical processes in life that are embodied in enzymes. A number of review articles have been published on the use of biocatalysis in organic synthesis (Sih and Chen, 1984; Jones, 1986; Crout et al., 1994; Davies et al., 1990; Csuz and Glanzer, 1991; Crosby, 1991; Kamphuis et al., 1990a; Sih et al., 1992; Santaneillo et al., 1992; Margolin, 1993; Cole, 1994; Patel, 1997, 1998, 1999; Mori, 1995; Wong and Whitesides, 1994). This chapter provides some specific examples of the use of microbial enzymes for the synthesis of chiral drug intermediates.

II. Antihypertensive Drug: Vasopeptidase Inhibitor A. ENZYMATIC SYNTHESIS OF

BMS-199541-01

[4S-(4a,7a,10ab)] 1-octahydro-5-oxo-4-[[(phenylmethoxy)carbonyl]amino]7H-pyrido-[2,1-b] [1,3]thiazepine-7-carboxylic acid methyl ester (BMS199541-01) is a key chiral intermediate for the synthesis of Omapatrilat (BMS-186716), a new vasopeptidase inhibitor presently under development (Robl et al., 1997). Our goal was to prepare the compound by a

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

35

simpler, more convenient route using an intermediate derived from L-lysine as a readily available starting material. An enzymatic process was developed for the oxidation of the e-amino group of lysine in the dipeptide dimer N2-[N[[(phenylmethoxy)carbonyl] L-homocysteinyl] Llysine)l,l-disulfide (BMS-201391-01) to produce BMS-199541-01 (Fig. 1) using L-lysine e-aminotransferase from Sphingomonas paucimobilis SC16113 (Patel et at., 1999a). This enzyme was overexpressed in Escherichia coli and a biotransformation process was developed using the recombinant enzyme. The aminotransferase reaction required a-ketoo glutarate as the amine acceptor. The glutamate formed during this reaction was recycled back to a-ketoglutarate by glutamate oxidase from Streptomyces noursei SC6007. A selective culture technique was used to isolate microorganisms able to utilize N-a-CBZ-L-lysine as the sole source of nitrogen. Using this technique, eight different types of colonies were isolated. Cultures were grown in shake flasks, and cell extracts prepared from cell suspensions were evaluated for oxidation of the e-amino group of L-lysine in the substrate dipeptide dimer BMS-201391o01. Product (BMS199541-01) formation (0.05-0.35 mg/ml) was observed with four cultures. One of the cultures, Z-2, later identified as Sphingomonas pau-

PH

(~ CO2H

0 CO2H ,;

Dlthlothmitol or

LS

Tributylphosphine

/ I .NH2

../.5

PHN"'~U'~N"~ l,. H~ ~ H2N)

H2N

DiDeotide Monomer

O CO2H

L-lysine eaminotransferase

Dipe~)tide Dimer BMS-201391-01

SphlngomonaSor rec E.peucimobiliScoli

If (~-ketoglut,,rata~'~ k~ ) Glutamate Oxidase ~',~ Glutamate _j/ Stmptomycesnoursel

P=CBZ

oooo. BMS-199541-01

O

L

o co,.-1

x

F

o co,.-7

j

Other protecting group P= phenoxyacetyl otphenylacetyl

FIG. 1. Enzymatic conversion of dipeptide dimer BMS-201391-01 to BMS-199541-01 by L-lysine e-aminotransferase.

36

RAMESH N. PATEL

cimobi]is SC16113, exhibited higher activity (0.35 mg/ml of product formed) and was used for further studies. The low mass balance and reaction yield were due to hydrolysis of the substrate dipeptide by cell extracts. S. paucimobilis SC16113 was grown in a 700-liter fermentor containing 500 liters of medium. During fermentation, cells were harvested from 200 ml of broth by centrifugation. Cells were suspended in buffer, and cell extracts were prepared. Cell extracts were evaluated for conversion of BMS-201391-01 to BMS-199541-01. Cultures grown for 48 to 60 hours had higher specific activity compared to cells harvested at 24 or 72 hours. A specific activity (milligrams of BMS-199541-01 formed per hour per gram of protein in cell extract) of 220 was obtained for cultures grown for 60 hours. A preparative batch for biotransformation of BMS-201391-01 to BMS199541-01 using 2 liters of cell extract of S. paucimobHis SC16113 was prepared. Substrate was used at a concentration of 1.5 g/liter. A reaction yield of only 10% (0.3 g of BMS-199541-01) was obtained after 1.75 hours. The product was isolated and identified by 1H-NMR, 13C-NMR, and mass analysis. Due to the low activity of L-lysine e-aminotransferase in S. paucimobilis SC16113, we decided to purify the enzyme, determine its sequence, and overexpress the protein in a suitable host. The enzyme was purified 254-fold to a specific activity (mg product formed per hour per gram of protein) of 36,600. After Sephacryl S-200 column chromatography, the purified enzyme showed a single protein band on SDS/PAGE using a silver stain. The molecular weight of the enzyme as determined by gel-filtration techniques was 81,000 daltons, and the subunit size as determined by SDS/PAGE was 40,000 daltons, indicating that the L-lysine ~-aminotransferase is a dimeric protein. The N-terminal and internal peptide sequence (generated by Lys-peptidase treatment) of purified L-lysine aminotransferase were determined. The purified L-lysine ~-aminotransferase was evaluated for cofactor requirements. The enzyme required ¢t-ketoglutarate as an amine acceptor. NAD or NADP were not required as cofactors, indicating that the enzyme was an aminotransferase and not a dehydrogenase. A reaction yield of 70 mol% was obtained for BMS-199541-01 with the complete system. Glutamate oxidase was required to recycle glutamate back to ~-ketoglutarate. In the absence of glutamate oxidase, a 35 tool% reaction yield of BMS-199541-01 was obtained (Table I). The L-lysine ~-aminotransferase was overexpressed in E. coli strain GI724(pAL781-LAT). The enzyme was produced in a 25-liter fermentor.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

37

TABLE I COFACTORREQUIREMENTSOF L-LYSINEE-AMINOTRANSFERASE:CONVERSION Or BMS-201391-01 TO BMS-199541-01

Reaction system

BMS-201391-01 (mg/ml)

BMS-199541-01 (mg/ml)

Yield (%)

0.57 3 0.9 3 0.58

1.94 0 0.98 0 1.92

70 0 35 0 69

Complete system Minus ct-ketoglutarate Minus glutamate oxidase Minus aminotransferase Minus NAD or NADP

The complete reaction mixture in 10 ml contained 6 ml of purified L-lysine aminotransferase (Sephacryl S-200 fraction), 1 ml of i M potassium phosphate buffer, pH 7.8, containing 5 mM dithiothreitol, 1 m M EDTA, 20 mg c~-ketoglutarate and 30 mg of BMS-201391.3 ml of glutamate oxidase (7 U/ml) was added during the 5-hr reaction time. The concentrations of BMS-199541-01 and BMS-201391-01 were determined by HPLC.

The enzyme activity ranged from 1700 to 2425 units/liter of broth. The kinetics of enzyme production are shown in Figure 2. Screening of microbial cultures led to identification of Streptomyces noursei SC6007 as a source of extracellular glutamate oxidase. S. noursei SC6007 was grown in 380-liter fermentors. During fermenta-

"20

3000 "

•

(units/Lofbroth)

o

2000-

~6

== ==

_J "10

1000

v

"

Z

}-

==

O J! 0

•

,

•

10 Fermentation

,

20 Time

0

30

(hours)

F~c. 2. Fermentation of recombinant Escherichia coli: Production of L-lysine E-aminotransferase.

38

RAMESH N. PATEL

tion, cells were periodically harvested by centrifugation from 200 ml of culture broth. The supernatant solution was used for determination of extracellular glutamate oxidase activity. Glutamate oxidase activity correlated with growth of the culture in a fermentor and reached 0.75 units/ml at harvest (Fig. 3). At the end of fermentation, the fermentation broth was cooled to 8°C and cells were removed by centrifugation. Starting from the extracellular filtrate recovered after removal of cells from the fermentation broth, the glutamate oxidase was purified 260fold with a specific activity (units per milligram of protein) of 54. The purified enzyme showed a single protein band on SDS/PAGE using a silver stain. The molecular weight of the enzyme as determined by gel-filtration techniques was 125,000 daltons and the subunit size as determined by SDS/PAGE 60,000 daltons, indicating that the glutamate oxidase is a dimeric protein. The amino-terminal and internal peptide sequence of the purified enzyme were determined to allow for the synthesis of oligonucleotide probes for cloning and overexpression of the enzyme. Attempts to express the S. noursei SC6007 glutamate oxio dase using standard E. coli vectors and strains were unsuccessful. As an alternative, the SC6007 glutamate oxidase was expressed in Strep-

0.8

~'

.IL

P m t l s l v o l u m e of solids

"30 i

0.6

'20 c

o ~

0,4

o o >

10

,

~

0.2

.~ Q.

utamate oxldase 0 . 0 -0

10

20

30

40

Fermentation Time (Hours)

FIG. 3. F e r m e n t a t i o n of Streptomyces noursei SC6007: Production of glutamate oxidase.

39

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

tomyces lividans. The S. noursei SC6007 glutamate oxidase, including its native promoter sequence, was inserted into an S. lividans expression vector. Untransformed S. lividans does not have a native glutamate oxidase activity, while S. lividans transformed with the GOX expression plasmid demonstrated glutamate oxidase activity. SDS/PAGE analysis of the transformed S. lividans revealed a protein band not seen in an untransformed strain. This band was of the same molecular weight as the GOX protein purified from S. noursei SC6007, indicating that the glutamate oxidase activity present in the transformed strain arose from expression of the heterologous gene. About 0.4 units/ml of activity was detected from the S. lividans culture, indicating that the enzyme was expressed at a low level. Further research was required to overexpress this protein. Biotransformation of BMS-201391-01 to BMS-199541-01 was carried out using L-lysine s-aminotransferase from Escherichia coli GI724[pa1781-LAT] in the presence of ct-ketoglutarate and dithiothreitol (required to reduce the dipeptide dimer to a monomer). Glutamate produced during the reaction was recycled to o~-ketoglutarate by partially purified glutamate oxidase (7 units/ml) from S. noursei SC6007. Four different batches were carried out. Reaction yields of 65-70 mol% were obtained as shown in Table II. The kinetics of reaction are shown in Figure 4. Two n e w dipeptides, No[N[(phenylmethoxy)carbonyl]-L-methionyl]L-lysine (BMS-203528-01) and N,2-[S-acetyl-N-[(phenylmethoxy)carbonyl]-L-homocysteinyl]-L-lysine (BMS-204556), were evaluated as

TABLE II BIOTRANSFORMATIONOF BMS-201391-01 TO BMS-199541-01 BY L-LYSINE £-AMINOTRANSFERASEFROM ESCHERICHIACOLIGI724[pa1781-LAT] Experiment batch number

BMS-201391-01 input (g)

BMS-201391-01 remaining (g)

BMS-199541-01 01 {g)

BMS-199541-01 (mol% yield)

40455 40456 40457 40458

3 5 12 22

0.83 1.35 4.3 4.7

1.9 2.92 7.5 14.4

66.5 65 70 67

R e a c t i o n s were carried o u t as d e s c r i b e d in t h e text u s i n g cell extracts of Escherichia coli GI724[pa1781-LAT] in t h e p r e s e n c e of dithiothreitol a n d partially purified g l u t a m a t e o x i d a s e from Streptomyces noursei SC6007.

40

RAMESH N. PATEL

A .J ~,

~-

BMS-199541-01

r,

o

i

/ o 0

& i 100

BMS-201391-01

200

300

Reaction time (min)

FIG. 4. Kinetics of oxidation of dipeptide dimer BMS-201391-01 to BMS-199541-01 by L-lysine E-aminotransferase. Regeneration of c¢-ketoglutarate was carried out by glutamate oxidase.

substrates for L-lysine aminotransferase by cell-free extracts of E. coli GI724[pa1781-LAT] in the presence of a-ketoglutarate. The formation of new compounds from the enzymic reaction was investigated by liquid chromatography-mass spectrometry (LC-MS). The data indicate the of a n e w c o m p o u n d with a molecular weight of 392, which was assigned tentative structure 1. The e-NH 2 group of BMS-203528 was oxidized, and in the presence of trichloroacetic acid (TCA) the aldehyde was cyclized to the enamide with a loss of water (Fig. 5). When BMS-204556 was treated with cell-free extracts of E. coli GI724[pa1781-LAT] and o~-ketoglutarate, several new components were observed by LC-MS. The component with a molecular weight of 420.5 was assigned structure 2, formed by oxidation of the ~-NH2 group of BMS-204556 and subsequent dehydration to produce the cyclic enamide; the component with a molecular weight of 397 was proposed as desacetyl BMS-204556 3. The desacetyl BMS-204556 was then oxidized by the enzyme to BMS-199541-01 (MW = 378), as shown in Figure 6.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

0

41

COOH L-lysine e - a n g n m r a n s f e r a s e f r o m S. pm~cimobilis o r rec E. coli

\S --NH2 CH3 I

BMS-203528-01 M o l wt. 411.62

~NCbz 0 |I"COOHH+ t ....

L | L_

CH3 M o l wt. 4 1 0

NCbz""~ CH3 Molwe.392 !

not observed in M S

FIG. 5. Enzymatic oxidation of BMS-203528-01 by L-lysine e-aminotransferase. Regeneration of a-ketoglutarate was carried out by glutamate oxidase.

To reduce the cost of producing two enzymes, the transamination reactions were carried out in the absence of glutamate oxidase and with higher levels of ~-ketoglutarate. The reaction yield in the absence of glutamate oxidase averaged about 33-35 mol%. With 40 mg/ml of t~-ketoglutarate (10-fold increase in concentration) and at 40°C, the reaction yield increased to 70 mol%, equivalent to that in the presence of glutamate oxidase. Phenylacetyl- or phenoxyacetyl-protected analogues of BMS-201391-01 (Fig. 1) also served as substrates for L-lysine e-aminotransferase, giving a reaction yield of 70 mol% for the corresponding BMS-199541-01 analogues. N-a-t-bntoxycarbonyl-L-lysine and N-a-carbobenzoxy-L-lysine were also oxidized by L-lysine aminotransferase from E. coli GI724[pa1781LAT]. The chiral compounds (S)-3,4-dihydro-l,2(2H)-pyridinedicarboxylic acid, 1-(phenylmethyl)ester (BMS-202665), (S)-3,4-dihydro-l,2(2H)-pyridinedicarboxylic acid, and 1,1-dimethylethyl ester (BMS264406) were obtained as products of oxidation of N-a-CBZ-L-lysine and N-ct-BOC-L-lysine, respectively (Fig. 7). A reaction yield of 80-85 mol% was obtained for each product. In the enzymatic reaction to convert BMS-201391-01 to BMS-19954101, we used dithiothreitol (DTT) to cleave the disulfide bond of the dipeptide dimer BMS-201391 to produce the dipeptide monomer, which was the substrate for the L-lysine aminotransferase. It was observed that tributylphosphine (an inexpensive compound) was as effective as DTT for the dipeptide dimer to monomer conversion. In the presence of 10-mM tributylphosphine, 3.5 mg/ml of BMS-201391-01, 40 mg/ml ~-ketoglutarate, and 0.1 units of transaminase, a 69 mol% yield of BMS199541-01 was obtained.

42

RAMESH N. PATEL

©

z

o

o~~

.~

z

u

[

~ g

tt~ e~

¢x]

'-..

0

Z

z

0

k Z

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

NH 2 NHP Nc~-protected-L-lysine P = BOC or CBZ

L-lysine e-AminotransferDase

H20

43

COOH BMS-202665, P=CBZ BMS-264406, P= BOC

FIG. 7. Enzymatic oxidation of N-a-protected L-lysine by L-lysine e-aminotransferase. Regeneration of c¢-ketoglutarate was carried out by glutamate oxidase.

To terminate the L-lysine aminotransferase reaction during conversion of BMS-201391-01 to BMS-199541-01, 10% vol/vol trichloroacetic acid (TCA) was used. It was also observed that a much cheaper compound, methanesulfonic acid, is equally effective as TCA, giving a 70 tool% yield of BMS-199541-01. Aminotransferases have been used extensively in the synthesis of L-amino acids from the corresponding ~-ketoacids (Stirling, 1992; Kamphuis et al., 1990b). L-lysine a-ketoglutarate aminotransferase from Flavobacterium fuscum was reported by Soda et al. (1968), and they demonstrated that the product of L-lysine oxidation is 1-piperidine-6carboxylic acid. In this aminotransferase reaction, the E-amino group of L-lysine is transferred to a-ketoglutarate to yield glutamate and a-aminoadipate-8-semialdehyde, which is immediately converted into the intramolecular dehydrated compound 1-piperidine-6-carboxylic acid. The oxidation of N-o~-carbobenzoxy and Noa-t-butoxycarbonyl L-lysine by Rhodotorula graminis to produce novel chiral compounds (S)-3,4dihydro-l,2(2H)-pyridinecarboxylic acid, 1-(phenylmethyl)ester, (S)3,4-dihydro-l,2(2H)-pyridinecarboxylic acid, and 1-dimethylethyl ester has been demonstrated by Patel et al. (1999b). Soda and Misono (1968) reported that L-lysine cx-ketoglutarate aminotransferase (MW = 116,000) contained two molecules of pyridoxal phosphate as a bound prosthetic group. Hammer and Bode (1992) purified L-lysine a-ketoglutarate aminotransferase from Candida utilis and reported that it is a dimeric 83,000-dalton protein. L-lysine aminotransferase from S. paucimobi]is SC16113 was a dimeric 80,000-dalton protein. B. ENZYMATICSYNTHESISOF L-6-HYDROXYNORLEUCINE L-6-hydroxynorleucine (4, Fig. 8) is a chiral intermediate useful for the synthesis of a vasopeptidase inhibitor now in clinical trials, and for the

44

RAMESH N. PATEL

glucose

",,,.._../

gluconicacid

glucosedehydrogenase NADH

~O~COzNa OH

O HoIV~

NAD

glutamate dehydrogenase ONa

NH3~

2-keto-6-hydroxyhexanoic acid, sodiumsalt5

H

~

NH2 OH

L-6-hydroxynorleucine 4

FIG. 8. Reductive amination of sodium 2-keto-6-hydroxyhexanoic acid 5 to L-6-hydroxynorleucine 4_by glutamate dehydrogenase.

synthesis of C-7 substituted azepinones as potential intermediates for other antihypertensive metalloprotease inhibitors (Robl and Cimarusti, 1994; Robl et al., 1997). It has also been used for the synthesis of siderophores, indospicines, and peptide hormone analogues (Maurer and Miller, 1981, 1982, 1983; Bodanszky et al., 1978; Dreyfuss, 1974). Previous synthetically useful methods for obtaining this intermediate have involved synthesis of the racemic compound followed by enzymatic resolution. D-amino acid oxidase has been used to convert Damino acid to the ketoacid, leaving the L-enantiomer that was isolated by ion exchange chromatography (Kern and Reitz, 1978). In a second approach, racemic N-acetylhydroxy norleucine has been treated with L-amino acid acylase to give the L-enantiomer (Robl et al., 1997). Both of these resolution methods give a maximum 50% yield and require separation of the desired product from the unreacted enantiomer at the end of the reaction. Reductive amination of ketoacids using amino acid dehydrogenases has been shown to be a useful method for synthesis of natural and unnatural amino acids (Bommarius, 1995; Galkin et al., 1997). We have developed the synthesis and conversion of 2-keto-6-hydroxyhexanoic acid 5 to L-6-hydroxy norleucine 4_ (Fig. 8) by reductive amination using beef liver glutamate dehydrogenase and glucose dehydrogenase from a Bacillus sp. for regeneration of NADH (Hanson et al., 1999). To avoid the lengthy chemical synthesis of the ketoacid, a second route was developed to prepare the ketoacid by treatment of racemic 6-hydroxy norleucine, readily available from hydrolysis of 5-(4-hydroxybutyl) hydantoin 6, with D-amino acid oxidase from porcine kidney or Trigonopsis variabilis and catalase followed by reductive amination to convert the mixture to L-6-hydroxynorleucine.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

45

2-keto-6-hydroxyhexanoic acid 5 is converted completely to L-6hydroxy norleucine 4 by phenylalanine dehydrogenase from a Sporosarcina sp. and beef liver glutamate dehydrogenase, with formate dehydrogenase for regeneration of NADH (Hanson et al., 1999). Beef liver glutamate dehydrogenase was used for preparative reactions at a substrate concentration of 100 g/liter. As depicted in Figure 8, 2-keto6-hydroxyhexanoic acid 5, sodium salt, in equilibrium with 2-hydroxytetrahydropyran-2-carboxylic acid, sodium salt, was converted to L-6o hydroxynorleucine. The reaction requires ammonia and reduced NADH. NAD produced during the reaction was recycled to NADH by oxidation of glucose to gluconic acid using glucose dehydrogenase from Bacillus megaterium. The reaction was completed within about 3 hr with reaction yields of 89-92% and an enantiomeric excess of >98% for L-6-hydroxynorleucine. Chemical synthesis and isolation of 2-keto-6-hydroxyhexanoic acid required several steps. In a second more convenient process (Fig. 9),

o

~/

H2N~/~H C02H

H Ca{OH)p NaOH

="

H

Racemic 6-hydroxynorleucine

5-(4-hydroxybutyl) hydantoin _6 glucose

gluconic acid

glucose dehydrogenase NADH NH2

D-amino acid oxidase

0

5_

NAD

glutamate dehydrogenase O

NH3

H

~

NH~ OH

0 +

Racemic 6-hydroxynorleucine

NH2

H

~

OH © catalase + H202

H20 +

02

+ NH3

FIG. 9. Enzymatic conversion of racemic 6-hydroxynorleucine to L-6-hydroxynorleucine 4 by D-amino acid oxidase and glutamate dehydrogenase.

RAMESH N. PATEL

46

the ketoacid was prepared by treatment of racemic 6-hydroxynorleucine (produced by hydrolysis of 5-(4-hydroxy butyl)hydantoin 6) with D-amino acid oxidase and catalase. After the enantiomeric excess of the remaining L-6-hydroxynorleucine had risen to >99%, the reductive amination procedure was used to convert the mixture containing 2keto-6-hydroxy hexanoic acid and L-6-hydroxynorleucine entirely to L-6-hydroxynorleucine with yields of 91-97% and enantiomeric excess of >98%. Sigma porcine kidney D-amino acid oxidase and beef liver catalase or T. variabilis whole cells (source of oxidase and catalase) were used successfully for this transformation. C. ENZYMATIC SYNTHESIS OF ALLYSINE ETHYLENE ACETAL

(S)-2-amino-5-(1,3-dioxolan-2-yl)-pentanoic acid (allysine ethylene acetal _7) is one of three building blocks used for an alternative synthesis of omapatrilat, a vasopeptidase inhibitor (Robl et al., 1997). It was previously prepared in an eight-step synthesis from 3,4-dihydro-2Hpyran for conversion into 1-piperidine-6-carboxylic acid, an intermediate for biosynthesis of [3-1actam antibiotics (Rumbero et aL 1995). The reductive amination of ketoacid acetal 8 to acetal amino acid 7 was demonstrated using phenylalanine dehydrogenase from Thermoactinomyces intermedius (Fig. 10). The reaction requires ammonia and NADH. NAD produced during the reaction was recycled to NADH by oxidation of formate to CO2 using formate dehydrogenase (Hanson et

/--1

/--1O

o m o .nase Ammonium f o r m a ~

C02

NAD~--IcNAD OH

Phenylalanine dehydrogenase

H2N

COOH

8

FIG. 10. Reductive amination of ketoacid acetal 8 to amino acid acetal _7by phenylalanine dehydrogenase. Regeneration of NADH was carried out using formate dehydrogenase.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

47

al., 2000). An initial process was developed using heat-dried cells of T. intermedius ATCC 33205 as a source of phenylalanine dehydrogenase and heat-dried cells of methanol-grown Candida boidinii as a source of formate dehydrogenase. An improved process using phenylalanine dehydrogenase from T. intermedius expressed in E. coli BL21(DE3) (pPDH155K) (SC16144) in combination with C. boidinii as a source of formate dehydrogenase and a third-generation process using methanol-grown Pichia pastoris containing endogenous formate dehydrogenase and expressing T. intermedius phenylalanine dehydrogenase were also developed (Hanson et al., 2000). Glutamate, alanine, leucine, and phenylalanine dehydrogenases (listed in order of increasing effectiveness) converted _8 to the desired amino acid (7) (Table III). The product was identical by HPLC and MS analysis to a chemically synthesized standard. Some alternative sources of phenylalanine dehydrogenase were tested. Sporosarcina ureae strains SC16048 and SC16049 had respective specific activities of 0.996 and 0.862 g/rag for reductive amination of phenylpyruvate, but amination of 8 was m u c h slower than with the enzyme from Thermoactinomyces. Using an extract of T. intermedius ATCC 33205 as a source of phenylalanine dehydrogenase and Boehringer Mannheim formate dehydrogenase for NADH regeneration increased the estimated yield to 80%, and the process was developed using this enzyme combination. Heat-dried cells of T. intermedius and C. boidinii SC13822 grown on methanol were used for the reaction. Phenylalanine dehydrogenase activities in cells recovered from fermentation and fermentor productivities are shown in Table IV. T. intermedius gave a useful activity on a small scale (15 liters), but lysed soon after the end of the growth period, making recovery of activity difficult

TABLE III REDUCTIVE AMINATIONOF KETO ACID 8 BY AMINO ACID DEHYDROGENASE

Dehydrogenase

Source

Glutamate

Beef liver

Alanine

Bacillus s u bti]is

Leucine

Bacillus s p h a e r i c u s

Phenylalanine

Sporosarcina spp.

Amount (units)

A m i n o acid _7 produced (mM)

76

1.03

35.7

11.77

22

14.01

12.6

51.7

48

RAMESH N. PATEL TABLE IV ACTIVITIESAND PRODUCTIVITIESOF PHENYLALANINEDEHYDROGENASEAND FORMATE DEHYDROGENASEFOR VARIOUSSTRAINSGROWNIN A FERMENTOR

Enzyme Phenylalanine dehydrogenase

Formate dehydrogenase

Strain

Thermoactinomyces intermedius Escherichia coil Pichia pastoris Candida boidinii Pichia pastoris

Specific activity (U/g wet ceils)

Volumetric activity (U/liter of broth)

510

185

900

10,000 ND

24,000 14,500

94,000 25,000

9 26

120 1950

350 3200

Producivity (U/liter/week)

or impossible on a large scale (4000 liters). The problem was solved by cloning and expressing the T. intermedius phenylalanine dehydrogenase in Escherichia coli, inducible by isopropylthiogalactoside. Fermentation of T. intermedius yielded 184 units of phenylalanine dehydrogenase activity per liter of whole broth in 6 hours. At harvest the fermentor needed to be cooled rapidly, because the activity was unstable. In contrast, E. coli produced more than 19,000 units per liter of whole broth in about 14 hours and was stable at harvest. C. boidinii grown on methanol was a useful source of formate dehydrogenase, as has been shown previously (Schfitte et al., 1976). In order to recover the cells on a large scale, it was helpful to add 0.5% methanol to stabilize the cells. P. pastoris grown on methanol was also a useful source of formate dehydrogenase (Hou et al., 1982). Expression of T. intermedius phenylalanine dehydrogenase in P. pastoris, inducible by methanol, allowed both enzymes to be obtained from a single fermentation. The expression of the two activities during a P. pastoris fermentation is shown in Figure 11. Formate dehydrogenase activity per gram wet cells was 2.7-fold greater than for C. boidinii, and fermentor productivity was increased by 8.7-fold compared to C. boidinii. Fermentor productivity for phenylalanine dehydrogenase in P. pastoris was about 28% of E. coli productivity. Formate dehydrogenase has been reported to have a pH optimum of 7.5 to 8.5 (Schfitte et al., 1976). The pH optimum for reductive amination of 8 by an extract of T. intermedius was found to be about 8.7.

49

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

3000

20000

--

PDH

¸2000

FDH 10000 ¸

1000

f.

t

--

0

i

• 20

40

80

60

Fermentation Time (Hours)

FrG. 11. Fermentation of Pichia pastoris for production of recombinant phenylalanine dehydrogenase and endogenous formate dehydrogenase.

Reductive amination reactions were carried out at pH 8.0. A summary of lab scale l-liter batches is shown in Table V. The time course for a representative batch showing conversion of ketoacid 8 to amino acid 7 is presented in Figure 12 using E. coli/C, boidinii heat-dried cells. The procedure using heat-dried cells of E. coli containing cloned phenylalanine dehydrogenase and heat-dried C. boidinii was scaled up

TABLE V LABORATORYSCALE(1 LITER)BATCHESFOR REDUCTIVEAMINAT1ONREACTIONS Phenylalanine dehydrogenase source

Formate dehydrogenase source

Reaction yield of _7 (%)

EE of product _7 (%)

T. intermedius

Candida boidinii

85

>99

Escherichia coil

Candida boidinii

90

>99

Pichia pastoris

Pichia pastoris

94

>99

50

RAMESH N. PATEL 60

50 I

40-

30-

20

10

0

10

20

30

Reaction T i m e (Hours)

FIG. 12. Kinetics of p r o d u c t i o n of amino acid acetal 7 from ketoacid acetal _8 by p h e n y l a l a n i n e dehydrogenase. Regeneration of NADH was carried out using formate dehydrogenase.

(Table VI), A total of 197 kg of compound 7 was produced in three 1600-liter batches using a 5% concentration of substrate _8 with an average yield of 91.1 mol% and enantiomeric excess greater than 98%.

TABLE VI PREPARATIVESCALEBATCHESFOR REDUCTIVEAMINATIONOF KETO ACID ~l

Phenylalanine dehydrogenase source

Formate dehydrogenase source

Keto acid 8 input (ks)

Amino acid 7 output (ks)

Reaction yield of _7 (tool%)

Escherichia coli Escherichia coli Escherichia coli Pichia pastoris

Candida boidinii Candida boidinii Candida boidinfi Pichia pastoris

80.17 79.96 89.6 18.05

62.40 66.75 67.61 15.51

92 96 86 97.5

EE of amino acid _7 (%) >99 >99 >99 >99

SYNTHESIS OF CHIRALDRUGINTERMEDIATES

51

A third-generation procedure using dried recombinant P. pastor& containing T. intermedius phenylalanine dehydrogenase inducible with methanol and endogenous formate dehydrogenase induced when P. pastoris was grown in medium containing methanol allowed both enzymes to be produced during a single fermentation, and they were conveniently produced in about the right ratio for the reaction. The Pichia reaction procedure had the following modifications of the E. coli/C, boidinii procedure: concentration of substrate was increased to 100 g/liter, a quarter of the amount of NAD was used, and DTT was omitted. The procedure with P. pastoris was also scaled up to produce 15.5 kg of 7 with a 97 mol% yield and enantiomeric excess greater than 98% (Table VI) in a 180-liter batch using 10% ketoacid 8 concentration. For reusability, formate dehydrogenase could be immobilized on Eupergit C and phenylalanine dehydrogenase on Eupergit C250L. The immobilized enzymes were tested for reusability in a jacketed reactor maintained at 40°C and were used five times for conversion of 8 to 7 without much loss of activity or productivity. At the end of each reaction, the solution was drained from the reactor through a 80/400 mesh stainless steel sieve, which retained the immobilized enzymes; then the reactor was recharged with flesh substrate solution. After five reuses, the reaction rate was decreased; however, the original reaction rate was restored in the seventh test study by addition of formate dehydrogenase. T, intermedius IFO14230 (ATCC 33205) was first identified as a source of phenylalanine dehydrogenase by Ohshima et al. (1991). The enzyme was purified and characterized, then cloned and expressed in E. coli by Takada et al. (1991). The enzyme was reported to be rather specific for deamination of phenylalanine (Ohshima et al., 1988), and to carry out amination of some ketoacids at a much lower rate than amination of phenylpyruvate. In our screening, the enzyme was the most effective amino acid dehydrogenase identified for bioconversion of 8 to 7. Formate dehydrogenase from C. boidinii was introduced by Shaked and Whitesides (1980), and by Kula and Wandrey (1987) for regeneration of NADH. The advantages of this enzyme reaction are that the product CO2 is easy to remove and that the negative reduction potential (E'° -- -0.42 V) for the formate dehydrogenase reaction drives reductive amination to completion. Previously, we prepared L-[3-hydroxyvaline from ~-keto-13-hydroxyisovalerate by the enzymatic reductive amination reaction using leucine dehydrogenase from Bacillus sphaericus ATCC 4525. L-[3-hydroxyvaline is a key chiral intermediate for the synthesis of tigemonam, an antiinfective drug (Hanson et al., 1990).

52

RAMESH N. PATEL

III. 133-Receptor Agonist [3-adrenoceptors have been classified as 131 and 132 (Land et al., 1967). Increased heart rate is the primary consequence of (31-receptor stimulation, while bronchodilation and smooth muscle relaxation are mediated from [32 receptor stimulation. Rat adipocyte lipolysis was initially thought to be a (51-mediated process (Land et al., 1967). However, recent results indicate that the receptor-mediated lipolysis involves neither I~1 nor [32, but "atypical" receptors, later called 133-adrenergic receptors (Arch, 1997). 133-adrenergic receptors are found on the cell surface of both white and brown adipocytes and are responsible for lipolysis, thermogenesis, and relaxation of intestinal smooth muscle (Arch et al., 1984). Consequently, several research groups are engaged in developing selective [~3 agonists for the treatment of gastrointestinal disorders, type II diabetes, and obesity (Wilson et al., 1984; Bloom et a]., 1989; Fisher et al., 1994; Sher, 1994). Efficient biocatalytic synthesis of chiral intermediates required for total chemical synthesis of [33 receptor agonist have been reported (Patel et al., 1998). These include: (a) microbial reduction of 4-benzyloxy-3-methanesulfonylamino-2'-bromoacetophenone 9 to corresponding (R)-alcohol 10 by Sphingomonas paucimobilis SC16113, (b) enzymatic resolution of racemic a-methyl phenylalanine amide 11 and c~-methyl-4-hydroxyphenylalanine amide 13 by amidase from Mycobacterium neoaurum ATCC 25795 to prepare the corresponding (S)-amino acids 12 and 14, and (c) asymmetric hydrolysis of methyl-(4-methoxyphenyl)propanedioic acid ethyl diester 15 to the corresponding (S)-monoester 16 by pig liver esterase. A. MICROBIAL REDUCTION OF 4-BENZYLOXY-3-METHANESULFONYLAMINO-2'-BROMOACETOPHENONE

Microbial reduction of 4-benzyloxy-3-methanesulfonylamino-2'-bromoacetophenone 9- to the corresponding (R)-alcohol 10 was demonstrated using S. paucimobilis SC16113 (Fig. 13). Among cultures evaluated, Hansenula anamola SC13833, Hansenula anamola SC16142, Rhodococcus rhodochrous ATCC 14347, and S. paucimobilis SC16113 gave the desired alcohol 10 in >96% enantiomeric excess and >15% reaction yield. S. paucimobilis SC16113 in the initial screening catalyzed the efficient conversion of ketone 9_to the desired chiral alcohol 10 with 58% reaction yield and >99.5% enantiomeric excess. Since substrate 9 is insoluble in water, the effect of solvents to dissolve substrate 9 and supply it in the biotransformation reaction mixture was evaluated. Dimethylformamide at 2-5% concentrations

SYNTHESIS OF CHIRALDRUG INTERMEDIATES

53

O / . . ~ ~ O

~HSO2CH3

OH Bspingomonaspaucimobilis

SC16113

OJ ~ NH B r

~"

SO2CH3 Product10 (R)-Alcohol

Substrate9 Ketone

OH

OH H o

H

(

/CH2R

CH3 ~I'~OCH2CO2H CI

BRL37344

NHSO2CH3~ J BMS-210620 OCH3

FIc. 13. Stereoselective reduction of 4-benzyloxy-3-methanesulfonylamino-2'-bromoacetophenone 9 to corresponding [R)-alcohol 10 by Sphingomonas paucimobilis SC16113. Structure of antiviral compounds BRL-37344 and BMS-210620,

was found to be the best cosolvent to supply the substrate in the biotransformation process. The fermentation of S. paucimobilis SC16113 was carried out in a 750-liter fermentor. From each fermentation batch, about 60 kg of wet cell paste was collected. Cells harvested from the fermentor were used to conduct the biotransformation in 1-, 10-, and 210-liter preparative batches under aerobic or anaerobic conditions. The cells were suspended in 80-mM potassium phosphate buffer (pH 6.0) to 20% (wt/vol, wet cells) concentration. Compound 9 (1-2 g/liter) and glucose (25 g/liter) were added to the fermentor, and the reduction reaction was carried out at 37°C. In some batches, the microfiltered and diafiltered ceils were used directly in the bioreduction process. In all biotransformation batches, a reaction yield of >85% and an enantiomeric excess of >98% were obtained. The isolation of chiral alcohol 10 from the 200-liter preparative batch was carried out to obtain 100 g of product 10. The isolated 10 gave a homogeneity index (HI) of 83% and an enantiomeric excess of 99.5% as analyzed by chiral HPLC. The MS and NMR data of isolated compound 10 and standard compound 10 were virtually identical. In an alternate process, frozen cells of S. paucimobilis SC16113 were used with resin-adsorbed (XAD-16 resin) substrate at 5 and 10 g/liter substrate concentrations. In this process, an average reaction yield of 85% and an enantiomeric excess of >99% were obtained for chiral

54

RAMESH N. PATEL

alcohol 10. At the end of the biotransformation, the reaction mixture was filtered on a 100-mesh (150 m) stainless steel screen, and the resin retained by the screen was washed with 2 liters of water. The product was then desorbed from the resin and crystallized in an overall 75 mol% yield with 91% homogeneity and 99.8% enantiomeric excess. The reduction of compound 9 to compound 10 was also carried out using cell extracts of S. paucimobilis SC16113. Glucose dehydrogenase was used to regenerate the cofactor NADPH required for the reduction. After a 90-min reaction time, 80% conversion of ketone 9 to chiral alcohol 10 was obtained. B. ENZYMATICRESOLUTION OF RACEMIC m-METHYL PHENYLALANINE AMIDES

The enzymatic resolution of racemic m-methyl phenylalanine amide 11 and ct-methyl-4-hydroxyphenylalanine amide 13 to the corresponding (S)-amino acids 12 and 14 (Fig. 14), respectively, by an amidase from Mycobacterium neoaurum ATCC 25795 was demonstrated by Patel et

0

0

M. neoaurumATCC2795

o

~

+

Product 12

Substrate 11

O H3C, ~ t

O NH2

H3C

M. neoaurumATCC25795

¢~ -

~ "OH

~.~

OCH 3

OCH 3

Substrate 13

Product 14

..,,,.~ R NH2

4-

H3

OCH 3

FIG. 14. Enantioselective enzymatic hydrolysis of ct-methyl phenylalanine amide 11 and c~-methyl-4-hydroxyphenylalanineamide 13 to corresponding (S)-amino acids by amidase from Mycobacterium neoaurum ATCC 25795.

SYNTHESIS OF CHIRALDRUG INTERMEDIATES

55

al. (1998). The chiral amino acids are intermediates for synthesis of a [33-receptor agonist (Bloom et al., 1989; Baroni eta]., 1994). The cells (10% wt/vol, wet cells) of M. n e o a u r u m ATCC 25795 were

evaluated for biotransformation of compound 11 to compound 12. The reaction was completed in 75 min with a reaction yield of 48 mol% (theoretical maximum = 50%) and an enantiomeric excess of 95% for the desired product 12. Freeze-dried cells of M. n e o a u r u m ATCC 25795 were suspended in 100-mMpotassium phosphate buffer (pH 7.0) at 1% concentration, and cell suspensions were used for biotransformation of compound 11. The reaction was completed in 60 min with a reaction yield of 49.5 mol% (theoretical maximum = 50%) and an enantiomeric excess of 99% for the desired product 12 (Fig. 14). Biotransformation of compound 11 was also carried out using a purified amidase. A reaction yield of 49 mol% and an enantiomeric excess of 99.8% were obtained for desired product 12 after a 60-min reaction time. Freeze-dried cells ofM. n e o a u r u m ATCC 25795 and partially purified amidase were used for biotransformation of compound 13. A reaction yield of 49 mol% and an enantiomeric excess of 78% were obtained for the desired product 14 using freeze-dried cells. The reaction was completed within 50 hours. Using partially purified amidase, a reaction yield of 49 tool% and a higher enantiomeric excess of 94% were obtained for desired product 14 after a 70-hr reaction time.

C. ASYMMETRICHYDROLYSISOF RACEMICMETHYL-(4-METHOXYPHENYL)-PROPANEDIOICACID, ETHYL DIESTER

The enzymatic asymmetric hydrolysis of methyl-(4-methoxyphenyl)propanedioic acid ethyl diester 15 to the corresponding (S)-monoester 16 by pig liver esterase has been demonstrated (Fig. 15). Chiral (S)monoester is a key intermediate for the synthesis of [33-receptor agonists. Various organic solvents were tested for the PLE-catalyzed asymmetric hydrolysis of diester 15 in a biphasic system. The results (Table VII) indicate that the reaction yields and enantiomeric excess of monoester 16 were dependent on the solvent used in asymmetric hydrolysis. Tetrahydrofuran, methyl ethyl ketone (MEK), methylisobutyl ketone (MIBK), hexane, and dichloromethane inhibited PLE. Lower reaction yields (28-56 mol%) and lower enantiomeric excess (59-72%) were obtained using t-butylmethyl ether, dimethylformamide (DMF), and

56

R A M E S H N. PATEL

/~

CO2C2H5 Pig LiverEsterase

H3CO-

~3CO-

...(.j

Diester 15

S-(-)-Monoester 16

FIG. 15. A s y m m e t r i c h y d r o l y s i s of r a c e m i c m e t h y l - ( 4 - m e t h o x y p h e n y l ) - p r o p a n e d i o i c acid e t h y l diester 15 to t h e c o r r e s p o n d i n g (S)-monoester b y pig liver esterase.

TABLE VII EFFECT OF SOLVENTON ASYMMETRIC HYDROLYSISOF METHYL-(4-METHOXYPHENYL)

PROPANEDIOIC ACID EHYL DIESTER 15 Enantiomeric Reaction time (hours)

Diester 15 (mg/ml)

Monoester 16 (mg/ml)

Yield (mol%)

of m o n o ester 16(%)

Methanol

22

0

0.65

37

92

Ethanol

22

0

1.7

96.7

96

Acetonitrile

22

0

0.5

28.2

59.3

Dimethylformamide

22

0

0.85

48.3

68.5

Dimethylsulfoxide

22

0.61

1

56.9

72

Acetone

22

0

1.44

81.9

65.1 82.1

Solvent

excess

Methylethylketone

48

0

1.36

77.3

Methylisobutylketone

64

2.01

0

0

-

t-butylmethylether

22

O. 76

0.8

46

64.4

Tetrahydrofuran

48

2

0

0

-

Toluene

22

0.18

0.59

33.6

91

Hexane

64

2.05

0

0

-

SYNTHESIS OF CHIRALDRUGINTERMEDIATES

57

dimethylsulfoxide (DMSO) as cosolvents. Higher enantiomeric excesses (>91%) were obtained using methanol, ethanol, and toluene as cosolvents. Ethanol gave the highest reaction yield (96.7%) and enantiomeric excess (96%) for monoester 16. The effect of temperature and pH were evaluated for the PLE-catalyzed hydrolysis of diester 15 in a biphasic system using ethanol as a cosolvent. It was observed that the enantiomeric excess of desired monoester 16 was increased by lowering the temperature from 25 to 10°C. The optimum pH for asymmetric hydrolysis of diester 15 in a biphasic system using ethanol as a cosolvent was 7.2 at 10°C. A semipreparative-scale asymmetric hydrolysis of diester 15 was carried out in a biphasic system using 10% ethanol as a cosolvent. Substrate (3 g) was used in a 300-ml reaction mixture. The reaction was carried out at 10°C, with 125-rpm agitation, and at a pH of 7.2 for 11 hours. A reaction yield of 96 mol% and an enantiomeric excess of 96.9% were obtained. From the reaction mixture, 2.6 g of monoester 16 were isolated in an 86.3 mol% overall yield. The enantiomeric excess of isolated S-(-)-monoester 16 was 96.9%. 1H NMR and MS of isolated product were consistent with monoester 13, and the specific rotation of monoester [~]D was --14.4 (c = 1.1 in methanol).

IV. Anticholesterol Drugs Pravastatin 17 and mevastatin 18 are anticholesterol drugs that act by competitively inhibiting HMG-CoA reductase (Endo et al., 1976a). Pravastatin sodium is produced in two fermentation steps. The first step is production of compound ML-236B by Penicillium citrinum (Endo et al., 1976b; Hosobuchi eta]., 1993a, 1993b). Purified compound is converted to its sodium salt 19 with sodium hydroxide and in the second step is hydroxylated to pravastatin sodium 17 (Fig. 16) by Streptomyces carbophilus (Serizawa et al., 1983). A cytochrome P450-containing enzyme system has been demonstrated from the S. carbophilus that catalyzed the hydroxylation reaction (Matsuoka et al., 1989). Squalene synthase is the first pathway-specific enzyme in the biosynthesis of cholesterol and catalyzes head-to-head condensation of two molecules of farnesyl pyrophosphate (FPP) to form squalene 20. It has been implicated in the transformation of FPP into presqualene pyro-

58

RAMESHN. PATEL A

.,,,OH

NaOOC" "~.... HOy

NaOOC/~.,.,,~OH

1 if,,. H 11.. ~ C H 3 CH3:" \V " 7) are better, and there was less hydrolysis (e.g., 38 and 30% in 24 hours at pH 8 and 9, respectively). Therefore, pH 8.0 was selected for conducting enzymatic hydrolysis. Even at pH 8, 19% of racemic epoxide 28 was hydrolyzed in 4 hours. Therefore, it was necessary to find a microorganism that hydrolyzes the racemic epoxide with high stereospecificity at a faster rate to prevent (or at least minimize) loss of unreacted desired S-epoxide 28 by chemical hydrolysis. Several fungi, yeast, and bacterial cultures were screened for stereospecific hydrolysis of the racemic epoxide. Two A. niger strains (SC16310, SC16311) and Rhodotorula glutinis SC16293 selectively hydrolyzed the R-epoxide, leaving behind S-epoxide 28. The enantiomer ratio (E) values (Chen et aL, 1982) for these microorganisms were -25. Unreacted S-epoxide 28 was obtained in >95% enantiomeric excess and at a 45% yield (theoretical maximum = 50%). Rhodococcus equi SC15835 did not hydrolyze the epoxide. Nocardia salmonicolor SC6310 hydrolyzed the racemic epoxide at a slow rate, and the enantiomeric excess of the S-epoxide was only 30%.

~spRhOdotorula glutinis?--~ ergillus~ger ~~V + 29

2~

OH-"OH ao

FIG. 20. Enantioselective hydrolysis of racemic epoxide 29 to corresponding (R)-diol 30 and unreacted (S)-epoxide 28.

65

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

From the initial screening studies, R. glutinis SC16293 and two A. niger strains, SC16310 and SC16311, were selected for further research. Hydrolysis of racemic epoxide by R. glutinis SC16293 was carried out. The desired S-epoxide 28 was obtained in 40% yield and >95% enantiomeric excess when the substrate was used at 2 g/liter and cells were used at 100 g/liter concentrations. Several solvents at 10% (vol/vol) were evaluated in an aqueous reaction mixture to improve the enantiomeric excess and yield (Table IX). Two solvents--cyclohexane and 1,1,2-trichloro-trifluoroethane (where the epoxide was not very soluble)--were used in higher amounts. Solvents had significant effects on both the extent of hydrolysis and the enantiomeric excess of unreacted S-epoxide 28. Most solvents, except for methyl tert-butyl ether (MTBE), gave lower enantiomeric excess than that of reactions catalyzed in buffer without any solvent supplement. The extent of hydrolysis in the presence of solvents was always lower than that in buffer. MTBE gave excellent results. A reaction yield of 45% (theoretical maximum = 50%) and an enantiomeric excess of 98.9% were obtained for unreacted S-epoxide 28. The hydrolysis reaction in the presence of MTBE gave an E value of 68. Two A. niger strains--SC16310 and SC16311--were evaluated in terms of their potential for stereospecific hydrolysis of the racemic epoxide. Both strains gave an enantiomeric excess of 97% and a yield of 45% of the remaining S-epoxide 28 when substrate was used at a 2

TABLE IX ENANTIOSELECTWEHYDROLYSISOF RACEMICEPOXIDE 2~ BY RHODOTORULA GLUTINIS SC16293 IN A BIPHASIC SYSTEM

Solvent

Reaction time (hours)

Remaining epoxide (%)

EE of (S)-epoxide 28 (%)

E value

Buffer

7

37

96.6

Cyclohexane

5

53

45.5

14 5

Toluene

5

66

45.9

29

1,1,2-trichlorotrifluoroethane

5

76

31.5

511

M e t h y l tert-butyl e t h e r (MTBE)

5

45

98.9

68

Methyl isobutyl ketone

5

68

22.7

4

n-Butanol

5

81

3.1

1

Dimethylsulfoxide

5

46

83.5

14

Dimethyl formamide

5

43

80

10

66

RAMESHN. PATEL

g/liter concentration. At a higher substrate concentration (5 g/liter) using a 100 g/liter cell concentration, a reaction yield of 51% and enantiomeric excess of 84% were obtained with SC16311.

VIII. Biocatalytic Dynamic Resolution: Stereoinversion of Racemic Diol One of the most often-used techniques for development of chiral compounds involves biocatalytic resolution. Though these kinetic resolution processes often provide compounds with high enantiomeric excess, the maximum theoretical yield of product or substrate is only 50%. In many cases, since the reaction mixture contains a 50:50 mixture of reactant and product with only slight difference in properties (e.g., hydrophobic alcohol and its acetate), separation becomes very difficult and impractical. These problems of kinetic resolution can be solved by employing a "dynamic resolution" process. The dynamic resolution process for alcohol is essentially a stereoinversion process. Only one enantiomer of the alcohol is enantiospecifically oxidized to the ketone, while the other enantiomer of the alcohol remains unchanged. The ketone is not isolated but is reduced to the opposite enantiomer of the alcohol during the process. The net result is conversion of the racemic alcohol to one enantiomer of the alcohol in high (theoretical maximum = 100%) yield. Dynamic resolution thus overcomes the limitation on maximum theoretical yield (50%) encountered during kinetic resolution of alcohol with enzymes. Only a handful of reports have appeared in the more recent literature on dynamic resolution of alcohols (Buisson et al., 1992; Nakamura et al., 1995; Fantin et al., 1995; Takahashi et al., 1995; Shimizu et al., 1987a, 1987b; Hasegawa et al., 1990; Stecher and Faber, 1997). Geotrichum c a n d i d u m , Candida parapsilosis, and a few other species are reported to be effective in such processes. Dynamic resolution involving a biocatalyst and metal-catalyzed in-situ racemizations has also been reported with limited success (Allen and Williams, 1996; Dinh et al., 1996). Chiral S-diol 31 (S-l-{2',3'-dihydrobenzo[b]furan-4'-yl)-ethane-l,2diol) is a key intermediate for a new prospective circadian modulator drug candidate (Catt et al., 1998, 1999). Dynamic resolution of racemic diol RS-l-{2',3"-dihydrobenzo[b]furan-4'-yll-ethane-l,2-diol32 to S-diol S-l-{2',3"-dihydrobenzo[b]furan-4'-yll-ethane-l,2-diol 31 (Fig. 21) was demonstrated by Goswami et al. (1999b).

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

~

OH RS-Diol OH 32

O

67

H OH S-Diol 31

T Reduction _ Oxidation OH

OH 0

R-Diol 34

Ketone 33

FIG. 21. Biocatalytic dynamic resolution. Stereoinversion ofracemic diol 32 to (S)-diol 31 by Candida boidinii and Pichia methanolica.

Seven cultures were selected from the screening of 20 microorganisms as leading candidates for dynamic resolution. These were Candida boidinii SC13821, SC13822, SC16115, Pichia methanolica SC13825, SC13860, and Hansenula polymorpha SC13895, SC13896. The relative proportions of S-diol 31 increased with time in biotransformations with the above cultures. At the end of 1 week, the enantiomeric excess of the remaining S-diol 31 was found to be in the range 87-100% with these microorganisms. Only two microorganisms, Candida parapsilosis SC16346 and Arthrobacter simplex SC6379, gave a higher yield of R-diol. A new c o m p o u n d was formed during these biotransformations, as seen by the appearance of a new peak in the HPLC of reaction mixture. This c o m p o u n d was slightly less polar than the diol. The identity of this compound was established as hydroxy ketone 33 from an LC-MS peak at mass 178. The starting RS-diol showed a mass peak at 180 by LC-MS. The area of the HPLC peak for hydroxy ketone 33 at first increased with time, reached a maximum, and then decreased. This w o u l d be expected from the proposed pathway of dynamic resolution (Fig. 21). Hydroxy ketone 33 was first formed by oxidation of R-diol 34, and then subsequently reduced back to diol, but only to S-diol 31.

68

RAMESH N. PATEL

The quantity and enantiomeric excess of the diol at various times were followed very carefully for transformation of RS-diol by the seven microorganisms described above. The reactions were also conducted with and without glucose to investigate the effect of glucose on the course of biotransformation (Table X). C. boidinii SC13822, C. boidinii SC16115, and P. methanolica SC13860 transformed RS-diol 32 in 3-4 days, and S-diol 31 was obtained with a yield in the range 62-71% and enantiomeric excess in the range 90-100%.

IX. Resolution of Racemic Secondary Alcohols The current interest in enzymatic production of chiral compounds lies in preparation of intermediates for pharmaceutical synthesis. S-(+)-2-pentanol is a key chiral intermediate required for synthesis of anti-Alzheimer's drugs that inhibit [3-amyloid peptide release and/or synthesis (Audia et al., 1996; Hamilton et al., 1996). The enzymatic

TABLE X DYNAMICRESOLUTION: STEREOINVERSIONOF RACEMICDIOL 32 TO CHIRAL (S)-DIOL 31

Medium

Reaction time (days)

Remaining diol

(S)-diol

(%)

31 (%)

EE of

Microorganism

Strain

Candida boidiniii

13821

Buffer Buffer + glucose

4 4

70 74

87 54

Candida boidinii

13822

Buffer Buffer+ glucose

4 4

66 62

90 100

Candida boidinii

16115

Buffer Buffer + glucose

3 4 3

74 64 71

95 100 94

Pichia methanolica

13825

Buffer Buffer + glucose

4 4

83 72

63 87

Pichia methanolica

13860

Buffer

3 4 2 3

65 46 57 67

100 100 89 100

Buffer + glucose

Hansenula polymorpha

13895

Buffer Buffer + glucose

4 3

84 100

44 32

Hansenula polymorpha

13896

Buffer Buffer + glucose

4 3

73 74

60 52

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

69

resolution of racemic 2-pentanol and 2-heptanol by lipase B from Candida antarctica has been demonstrated by Patel et al. (1999b). Commercially available lipases were screened for stereoselective acetylation of racemic 2-pentanol in an organic solvent (hexane) in the presence of vinyl acetate as an acyl donor. C. antarctica lipase B efficiently catalyzed enantioselective acetylation of racemic 2-pentanol. Reaction yields of 49% (theoretical maximum = 50%) and an enantiomeric excess of 99% were obtained for S-(+)-2-pentanol. Preparativescale acetylation of racemic 2-pentanol was carried out in an organic solvent (heptane) in the presence of vinyl acetate as an acyl donor using lipase B (Table XI). At the end of the reaction, 44.5 g of S-(+)-2-pentanol were estimated by HPLC analysis, with an enantiomeric excess of 98%. Among acylating agents tested, succinic anhydride was found to be the best choice due to easy recovery of (S)-2-pentanol at reaction end. Reactions were carried out using racemic 2-pentanol as solvent as well as substrate. Using 0.68 mole equivalent of succinic anhydride (Fig. 22) and 13 g of lipase B per kilogram of racemic 2-pentanol, a reaction yield of 43 mol% (theoretical maximum = 50%) and an enantiomeric excess of >98% were obtained for (S)°2-pentanol. Product was isolated in overall 36% yield (theoretical maximum = 50%). The results from three preparative batches are shown in Table XII. As described earlier, resolution of 2-heptanol was also carried out using lipase B. Reactions were carried out using racemic 2-heptanol as solvent as well as substrate. Using 0.68 mole equivalent of succinic anhydride and 13 g of lipase B

TABLE XI PREPARATIVESCALE ENZYMATICACETYLATIONOF RACEMIC 2-PENTANOL USING LIPASE B FROM CANDIDAANTARCTICA Reaction time (hours)

(S)-2-pentanol (g/liter)

(R)-2-pentanol (g/liter)

0 2 4 6

50 50 46 44.5

50 20 10 0.02

EE of (S)-2-pentanol (%) 0 32 60 98

Reaction m i x t u r e in 1 liter of h e p t a n e c o n t a i n i n g 100 g of r a c e m i c 2-pentanol, 1.02 m o l e e q u i v a l e n t of v i n y l acetate, a n d 1 g of lipase B from C. antarctica. T h e reaction w a s carried o u t at 35°C a n d 150 r p m .

70

RAMESH N. PATEL

OH

Racemic 2-pentanol

OH

Lipase B from C. antarctica

S-(+)-2-pentanol

o

°

R-(-)-2-pentylhemisuccinate

FIc. 22. Enzymatic resolution of racemic 2-pentanol to S-(+)-2-pentanol by Candida antarctica lipase.

TABLE XII ENZYMATICACYLATIONOF SECONDARYALCOHOLSUSINGSUCCINICANHYDRIDE ANDLIPASEB FROMCANDIDAANTARCTICA

Batch number 2-pentanoh 132 133 136

Batch number 2-heptanoh 140 141

2-pentanol input (kg)

(S)-2-pentanol (% yield)

EE of (S)-2pentanol (%)

0.5 0.5 0.9

42.0 43.5 43.7

>99 >98 99

2-heptanol input (kg)

0.1 0.5

(S)-2-heptanol (% yield)

43.0 44.5

EE of (S)-2heptanol (%)

>99 >99

Reaction mixture contained racemic 2-pentanol or 2-heptanol as solvent as well as substrate; 0.68 mole equivalent of succinic anhydride and 13 g of lipase B per kg of substrate input. The reaction was carried out at 38°C and 150 rpm.

per kilogram of racemic 2-heptanol, a reaction yield of 44 mol% (theoretical maximum = 50%) and an enantiomeric excess of >99% were o b t a i n e d f o r S - ( + ) - 2 - h e p t a n o l ( T a b l e XII). P r o d u c t w a s i s o l a t e d i n a n overall 40% yield (theoretical maximum = 50%).

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

71

X. Summary Biocatalytic processes were used to prepare chira] intermediates for pharmaceuticals. These include the following processes. Enzymatic synthesis of [4S-(4a,7a,10ab)]l-octahydro-5-oxo-4-[[(phenylmethoxy) carbonyl]amino]-7H-pyrido-[2,1-b] [1,3]thiazepine-7-carboxylic acid methyl ester (BMS-199541-01), a key chiral intermediate for synthesis of a new vasopeptidase inhibitor. Enzymatic oxidation of the e-amino group of lysine in dipeptide dimer N2-[N[[(phenylmethoxy)carbonyl] L-homocysteinyl] L-lysine)l,l-disulfide (BMS-201391-01) to produce BMS-199541-01 using a novel L-lysine ~-aminotransferase from S. paucimobilis SC16113 was demonstrated. This enzyme was overexpressed in E. coli, and a process was developed using recombinant enzyme. The aminotransferase reaction required ~-ketoglutarate as the amine acceptor. Glutamate formed during this reaction was recycled back to ~-ketoglutarate by glutamate oxidase from S. noursei SC6007. Synthesis and enzymatic conversion of 2-keto-6-hydroxyhexanoic acid 5 to L-6-hydroxy norleucine 4 was demonstrated by reductive amination using beef liver glutamate dehydrogenase. To avoid the lengthy chemical synthesis of ketoacid 5, a second route was developed to prepare the ketoacid by treatment of racemic 6-hydroxy norleucine (readily available from hydrolysis of 5-(4-hydroxybutyl) hydantoin, 6) with D-amino acid oxidase from porcine kidney or T. variabilis followed by reductive amination to convert the mixture to L-6-hydroxynorleucine in 98% yield and 99% enantiomeric excess. Enzymatic synthesis of (S)-2-amino-5-(1,3-dioxolan-2-yl)-pentanoic acid (allysine ethylene acetal, 7), one of three building blocks used for synthesis of a vasopeptidase inhibitor, was demonstrated using phenylalanine dehydrogenase from T. intermedius. The reaction requires ammonia and NADH. NAD produced during the reaction was recycled to NADH by oxidation of formate to CO2 using formate dehydrogenase. Efficient synthesis of chiral intermediates required for total chemical synthesis of a ~3 receptor agonist was demonstrated. These include: (a) microbial reduction of 4-benzyloxy-3-methanesulfonylamino-2"-bromoacetophenone 9 to corresponding (R)-alcohol 10 by S. paucimobilis SC16113, (b) enzymatic resolution of racemic m-methyl phenylalanine amide 11 and ~-methyl-4-hydroxyphenylalanine amide 13 by amidase from M. neoaurum ATCC 25795 to prepare corresponding (S)-amino acids 12 and 14, and (c) asymmetric hydrolysis of methyl-(4methoxyphenyl)-propanedioic acid ethyl diester 15 to corresponding (S)-monoester 16 by pig liver esterase.

72

RAMESH N. PATEL

(S)[1-(acetoxyl)-4-(3-phenyl)butyl]phosphonic acid diethyl ester 21, a key chiral intermediate required for total chemical synthesis of BMS188494 (an anticholesterol drug) was prepared by stereoselective acetylation of racemic [1-(hydroxy)-4-(3-phenyl)butyl]phosphonic acid diethyl ester 22 using G. candidum lipase. Lipase-catalyzed stereoselective acetylation of racemic 7-[N,N'-bis(benzyloxy-carbonyl)N-(guanidinoheptanoyl)]-~-hydroxy-glycine 24 to corresponding S-(-)-acetate 25 was demonstrated. S-(-)-acetate 25 is a key intermediate for total chemical synthesis of (-)-15-deoxyspergualin 23, an immunosuppressive agent and antitumor antibiotic. Stereoselective microbial reduction of (1S)[3-chloro-2-oxo-l-(phenylmethyl)propyl] carbamic acid, 1,1-dimethyl-ethyl ester 26 to corresponding chiral alcohol 27a (a key chiral intermediate for HIV protease inhibitors) was also demonstrated. Stereospecific enzymatic hydrolysis of racemic epoxide RS-I-{2',3'dihydro benzo[b]furan-4'-yl}-l,2-oxirane 29 the corresponding R-diol 30 and unreacted chiral S-epoxide 28 was demonstrated using R. glutinis and A. niger. Dynamic resolution of racemic diol RS-l-{2',3'-dihydrobenzo[b]furan-4'-yll-ethane-l,2-diol 32 to corresponding S-diol S-l-{2',3'-dihydrobenzo[b]furan-4'-yll-ethane-l,2-diol 31 was demonstrated using C. boidinii and P. methanolica. Chiral (S)-epoxide 28 and (S)-diol 31 are key intermediates for a new prospective circadian modulator drug. Enzymatic resolution of racemic 2-pentanol and 2-heptanol by lipase B from Candida antarctica was demonstrated. S-(+)-2-pentanol is a key chiral intermediate required for synthesis of anti-Alzheimer's drugs.

REFERENCES

Allen, J. V., and Williams, J. M. J. (1996). Dynamic kinetic resolution with enzyme and palladium combinations. Tetrahedron Lett. 37, 1859-1862. Arch, J. R. S. (1997). [~3-adrenoceptors and other putative atypical [3-adrenoceptors. Pharmacol. Rev. Commun. 9, 141-148. Arch, J. R. S., Ainsworth, A. T., Cawthrone, M. A., Piercy, V., Sennitt, M. V., Thody, V. E, Wilson, C., and Wilson, S. (1984). Atypical [3-adrenoceptor on brown adipocytes as target for anti-obesity drugs. Nature 309, 163-165. Audia, J. E., Britton, T. C., Droste, J. J., Folmer, B. K., Huffman, G. W., Varghese, J, Latimar, L. H., Mabry, T. E., and Nissen, J. S. (1996). "Preparation of N-(Phenylacetyl)di- and Tripeptide Derivatives for Inhibiting [~-Amyloid Peptide Release." U.S. Patent 96-755442. Baroni, M., Cecchi, R., and Croci, T. (1994). "Synthesis of ~-3 Receptor Agonists." European Patent Application 626367 A1941130.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

73

Barrish, J. C., Gordon, E., Alam, M., Lin, R-F., Bisacchi, G. S., Cheng, P. T. W., Fritz, A. W., Greytok, J. A., Hermsmeier, M. A., Humphreys, W. G., Lis, K. A., Marella, M. A., Merchant, Z., Mitt, T., Morrison, R. A., Obermeier, M. T., Pluscec, J., Skoog, M., Slusarchyk, W. A., Spergel, S. Stevenson, J. M., Sun, C. Q., Sundeen, J. E., Taunk, P., Tino, J. A., Warrack, B. M., Colono, R., and Zahler, R. (1994). Aminodiol HIV protease inhibitors, 1: Design, synthesis, and preliminary SAR. J. Med. Chem. 37, 1758-1771. Biller, S. A., and Majnin, D. (1995). "Preparation of a-Phosphonoalkanoates as Squalene Synthase Inhibitors." U.S. Patent 95-5428028. Biller, S. A., Sofia, M. J., DeLange, B., Foster, C., Gordon, E. M., Harrity, T., Rich, L. C., and Ciosek, C. P. (1991a). The first potent inhibitor of squalene synthase: A profound contribution of an ether oxygen to inhibitor-enzyme interaction. J. Am. Chem. Soc. 113, 8522-8527. Biller, S. A., Foster, C., Gordon, E. M., Harrity, T., Rich, L. C., Maretta, J., and Ciosek, C. P. (1991b). Isoprenyl phosphinylformates: New inhibitors of squalene synthase. J. Med. Chem. Soc. 34, 1912-1918. Bloom, J. D., Datta, M. D., Johnson, B. D., Wissner, A., Bruns, M. G., Largis, E. E., Dolan, J. A., and Claus, T. H. (1989). Disodium (R,R)5-[2-(3-chlorophenyl)-2-hydroxyethyl]amino]propyl]-l,3-benzodioxole-2,2-dicarboxylate: A potent [5-adrenergic agonist virtually specific for ~3 receptors. J. Med. Chem. 35, 3081-3084. Bodanszky, M., Martinez, J., Priestly, G. P., Gardner, J. D., and Mutt, V. (1978). Synthesis and properties of a biologically active analog of the C-terminal heptapeptide with e-hydroxynorleucine sulfate replacing tyrosine sulfate. J. Med. Chem. 21, 1030-1035. Bommarius, A. S. (1995). In "Enzyme Catalysis in Organic Synthesis" (K. Drauz and H. Waldmann, ed.), Vol. 2, pp. 633-641. VCH, Weinheim. Buisson, D., Azerad, R., Sanner, C., and Larcheveque, M. (1992). A study of the stereocontrolled reduction of aliphatic [3-ketoesters by Geotrichum candidum. Biocatalysis 5,249-265. Catt, J. D., Johnson, G., Keavy, D. J., Mattson, R. J., Parker, M. F. Takaki, K. S., and Yevich, J. P. (1998). PCT International Application WO98/25606. Catt, J. D., Johnson, G., Keavy, D. J., Mattson, R. J., Parker, M. F., Takaki, K. S., and Yevich, J. P. (1999). "Preparation of Benzofuran and Dihydrobenzofuran Melatonergic Agents." U.S. Patent 99-5856529. Chen, C.-S., Fujimoto, Y., Girdaukas, G., and Sih, C. J. (1982). Quantitative analyses of biochemical kinetic resolutions of enantiomers. J. Am. Chem. Soc. 104, 7294-7299. Cole, D. C. (1994). Recent stereoselective synthetic approaches to [3-amino acids. Tetrahedron 32, 9517-9582. Crosby, J. (1991). Synthesis of optically active compounds: A large-scale perspectives. Tetrahedron 47, 4789-4846. Crout, D. H. G., Davies, S., Heath, R. J., Miles, C. O., Rathbone, D. R., and Swoboda, B. E. P. (1994). Application of hydrolytic and decarboxylating enzymes in biotransformations. Biocatalysis 9, 1-30. Csuz, R., and Glanzer, B. I. (1991). Baker's yeast-mediated transformations in organic chemistry. Chem. Rev. 96, 556-566. Davies, H. G., Green, R. H., Kelly, D. R., and Roberts, S. M. (1990). Recent advances in the generation of chiral intermediates using enzymes. Biotechnology 10, 129-152. Dinh, P. M., Howarth, J. A., Hudnott, A. R., and Williams, J. M. J. (1996). Catalytic racemization of alcohols: Applications to enzymic resolution reactions. Tetrahedron Lett. 37, 7623-7626.

74

RAMESH N. PATEL

Dreyfuss, P. (1974). Synthesis and some pharmacological properties of 8-E-hydroxy norleucine-vasopressin.J. Med. Chem. 17,252-255. Endo, A., Kuroda, M., and Tsujita, Y. (1976a). ML-236A, ML-236B, and ML-236C, new inhibitors of cholesterogenesis produced by Penicillum citrinium. J. Antibiot. 29, 13461348. Endo, A., Kuroda, M., and Tanzawa, K. (1976b). Competitive inhibition of 3-hydroxy-3methylglutaryl coenzyme A reductase by ML-236A and ML-236B fungal metabolites, having hypocholesterolemic activity. FEBS Lett. 72,323-326. Faber, K., Mischitz, M., and Kroutil, W. (1996). Microbial epoxide hydrolases. Acta Chim. Scand. 50, 249-258. Fantin, G., Fogagnolo, M., Giovannini, P. P., Medici, A., and Pedrini, P. (1995). Combined microbial oxidation and reduction: A new approach to the high-yield synthesis of homochiral unsaturated secondary alcohols from racemates. Tetrahedron: Asymmetry 6, 3047-3053. Fisher, L. G., Sher, P. M., Skwish, S., Michael, I. M., Seiler, S. M., and Dickinson, K. E. J. (1994). BMS-187257, a potent, selective, and novel heterocyclic ~3 adrenergic receptor agonist. Bioorg. Med. Chem. Lett. 6, 2253-2258. Galkin, A., Knlakova, L., Yoshimura, T., Soda, K., and Esaki, N. (1997). Synthesis of optically active amino acids from alpha-ketoacids with Escherichia coli cells expressing heterologous genes. AppL Environ. Microbiol. 63, 4651-4656. Goswami, A., Totleben, M. J., Singh A., and Patel, R. N. (1999a). Stereospecific enzymatic hydrolysis of racemic epoxide: A process for making chiral epoxide. Tetrahedron: Asymmetry 10, 3167-3175. Goswami, A., Mirfakhrae, K. D., and Patel, R. N. (1999b). Deracemization of racemic 1,2-diol by biocatalytic stereoinversion. Tetrahedron: Asymmetry 10, 4239-4244. Grogan, G., Rippe, C., and Willetts, A. (1997). Biohydrolysis of substituted styrene oxides by Beauveria densa CMC 3240. J. MoL Cat. B: Enzymatic 3,253-257. Hasegawa, J., Ogura, M., Tsuda, S., Maemoto, S., Kutsuki, H., and Ohashi, T. (1990). High-yield production of optically active 1,2-diols from the corresponding racemates by microbial stereoinversion. Agric. Biol. Chem. 54, 1819-1827. Hamilton, G. S., Li, J.-H., and Steiner, J. P. (1996). "Preparation of Bridged Heterocyclic Derivatives for Treatment of Neurological and Other Disorders." U.S. Patent 965721256. Hammer, T., and Bode, R. (1992). Enzymatic production of a-aminoadipate-5-semialdehyde and related compounds by lysine e-dehydrogenase from Candida albicans. J. Basic MicrobioL 1, 21-27. Hanson, R., Singh, J., Kissick, T. R, Patel, R. N., Szarka, L. J., and Mueller, R. H. (1990). Synthesis of L-~-hydroxyvaline from c~-keto-~-hydroxyisovalerate using leucine dehydrogenase from Bacillus species. Bioorg. Chem. 18, 116-130. Hanson, R., Schwinden, M. D., Banerjee, A., Brzozowski, D., Chen, B.-C, Patel, B. P., McNamee, C., Kodersha, G., Kronenthal, D., Patel, R. N., and Szarka, L. J. (1999). Enzymatic synthesis of L-6-hydroxynorleucine.Bioorg. Med. Chem. 7, 2247-2252. Hanson, R., Howell, J., LaPorte, T., Donovan M., Cazzulino, D., Zannella, V., Montana, M., Nanduri, V., Schwarz, S., Eiring, R., Durand, S., Wasylyk, J., Parler, L., Liu, M., Okuniewicz, F., Chen, B.-C., Harris, J., Natalie, K., Ramig, K., Swaminathan, S., Rosso, V., Pack, S., Lotz, B., Bernot, P., Rusowicz, A., Lust, D., Tse, K., Venit, J., Szarka, L., and Patel, R. N. (2000). Enzymatic synthesis of allysine ethylene acetal. Enzyme Microb. Technol. 26, 348-358. Hechtberger, P., Wirnsberger, G., Mischitz, M., Klempeir, N., and Faber, K. (1993). Asymmetric hydrolysis of epoxides using an immobilized enzyme preparation from Rhodococcus spp. Tetrahedron, Asymmetry 4, 1161-1164.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

75

Hosobuchi, M., Kurosawa, K., and Yoshikawa, H. (1993a). Application of computers on monitoring and control of fermentation process: Microbial conversion of ML-236B sodium to pravastatin. Biotechno]. Bioeng. 42, 815-820. Hosobuchi, M., Shioiri, T., Ohyama, J., Arai, M., Iwado, S., and Yoshikawa, H.. (1993b). Production of ML-236B, and inhibitor of 3-hydroxy-3-methylglutaryl CoA reductase by Penicillium citrinum. Biosci. Biotechnol. Biochem. 57, 1414-1419. Hou, C. T., Patel, R. N., Laskin, A. I., and Barnabe, N. (1982). NAD-linked formate dehydrogenase from methanol-grown Pichia pastoris NRRL-Y-7556. Arch. Biochem. Biophys. 216, 296-305. Iwasawa, H., Kondo, S., Ikeda, D., Takeuchi, T., and Umezawa, H. (1982). Synthesis of (-)-15 deoxyspergualin and (-)-spergualin 15-phosphate. J. Antibiot. 35, 1665-1669. Jones, J. B. (1986). Enzymes in organic synthesis. Tetrahedron 42, 3351-3403. Kamphuis, J., Boesten, W. H. J., Broxterman, Q. B., Hermes, H. F. M., van Balken, J. A. M. Meijer, E. M., and Schoemaker. H. E. (1990a). New developments in chemoenzymatic production of amino acids. Adv. Biochem. Eng. Biotechnol. 42, 133-186. Kamphuis, J., Hermes, H. F. M., van Balken, J. A. M., Schoemaker, H. E., Boesten, W. H. J., and Meijer, M. E. (1990b). In "Production of Amino Acids and Derivatives" (D. Rozzell and F. Wagner, ed.), pp. 117-200. Hanser, Munich. Kern, B. A., and Reitz, R. H. (1978). A general method for the preparation of the D- or L-stereoisomers of 5-substituted hydantoins. Agric. Biol. Chem. 42, 1275-1278. Kula, M. R., and Wandrey, C. (1987). Continuous enzymatic transformation in an enzyme-membrane reactor with simultaneous NADH regeneration. Methods Enzymol. 136, 9-21. Kondo, S., Iwasawa, H., Ikeda, D., Umeda, Y., Ikeda, H., Inuma, H., and Umezawa, H. (1981). The total synthesis of spergualin, an antitumor antibiotic. Antibiotics 34, 1625-1627. Kroutil, W., Mischitz, M., Plachota, P., and Faber, K. (1996). Deracemization of cis-2,3epoxyheptane via enantioconvergent biocatalytic hydrolysis using Nocardia EH1epoxide hydrolase. Tetrahedron Lett. 37, 8379-8382. Kroutil, W., Mischitz, M., and Faber, K. (1997a). Deracemization of 2,3-disubstituted oxiranes via biocatalytic hydrolysis using bacterial epoxide hydrolases: kinetics of an enantioconvergent process. J. Chem. Soc., Perkin Trans. 1, 3629-3636. Kroutil, W, Osprian, I., Mischitz, M., and Faber, K. (1997b). Chemoenzymatic synthesis of (S)-(-)-frontalin using bacterial epoxide hydrolases. Synthesis 2, 156-158. Lands, M. A., Arnold, A., McAuliff, J. P., Luduena, F. P., and Brown, T. G. (1967). Differentiation of receptor systems activated by sympathomimetic amines. Nature 214, 597-598. Lawrence, M., Biller, S. A., and Fryszman, O. M. (1991). "Isoprenoid Phosphinylformic Acid Squalene Synthase Inhibitors and Methods for Preparing the Same." U.S. Patent 91-5447922. Maeda, K., Umeda, Y., and Saino, T. (1993). Synthesis and background chemistry of 15-deoxyspergualin. Ann. N. Y. Acad. ScL 685, 123-125. Margolin, A. L. (1993). Enzymes in the synthesis of chiral drugs. Enzyme Microb. Technol. 15, 266-280. Matsuoka, T., Miyakoshi, S., Tanzawa, K., Nakahara, K., Hosobuchi, M., and Serizawa, N. (1989). Purification and characterization of cytochrome P450sca from Streptomyces carbophilus. ML-236B (compactin) induces a cytochrome P450sca in Streptomyces carbophilus that hydroxylates ML-236 B to pravastatin sodium (CS-514), a tissue-selective inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase. Eur. J. Biochem. 184, 707-713.

76

RAMESH N. PATEL

Maurer, P. J., and Miller, M. J. (1981). Mycobactins: synthesis of (-)-cobaltin T from ~-hydroxynorleucine. J. Org. Chem. 46, 2835-2836. Maurer, P. J., and Miller, M. J. (1982). Microbial iron chelators: Total synthesis of aerobactin and its constituent amino acid, N6-acetyl-N6-hydroxylysine. J. Am. Chem. Soc. 104, 3096-3101. Maurer, P. J., and Miller, M. J. (1983). Total synthesis of a mycobactin: Mycobactin $2. J. Am. Chem. Soc. 105, 240-245. Mori, K. (1995). Biochemical methods in enantioselective synthesis of bioactive natural products. SYNLETT 11, 1097-1109. Nakamura, K., Inoue, Y., Matsuda, T., and Ohno, A. (1995). Microbial deracemization of 1-arylethanol. Tetrahedron Lett. 36, 6263-6266. Nellaiah, H., Morisseau, C., Archelas, A., Furstoss, R., and Baratti, J. (1996). Enantioselective hydrolysis of p-nitrostyrene oxide by an epoxide hydrolase preparation from Aspergillus niger. Biotechnol. Bioeng. 49, 70-77. Ohshima, T., Sugimoto, H., and Soda, K. (1988). Selective enzymic determination of L-phenylalanine and phenylpyruvate. Anal. Lett. 21, 2205-2215. Ohshima, T., Takada, H., Yoshimura, T., Esaki, N., and Soda, K. (1991). Distribution, purification, and characterization of thermostable phenylalanine dehydrogenase from thermophilic actinomycetes. J. Bacteriol. 173, 3943-3948. Patel, R. N. (1997). Stereoselective biotransformations in synthesis of some chiral pharmaceutical intermediates. Adv. Appl. Microbiol. 43, 91-140. Patel, R. N. (1998). Tour de paclitaxel: Biocatalysis for semisynthesis. Ann. Rev. Microbiol. 98, 361-395. Patel, R. N. (1999). Biocatalysis for synthesis of chiral drug intermediates. In "Stereoselective Biocatalysis" (R. Patel, ed.), pp 87-130. Marcel, New York. Patel, R. N., Banerjee, A., and Szarka, L. J. (1997a). Stereoselective acetylation of [1-(hydroxy)-4-(3-phenoxyphenyl)butyl]phosphonic acid diethyl ester. Tetrahedron: Asymmetry 8, 1055-1059. Patel, R. N., Banerjee, A., and Szarka, L. J. (1997b). Stereoselective acetylation of racemic 7-[N,N'-bis-benzyloxycarbonyl)-N-(guanidinoheptanoyl)]-a-hydroxyglycine.Tetrahedron Asymmetry 8, 1767-1771. Patel, R. N., Banerjee, A., McNamee C., Brzozowski, D., and Szarka, L. J. (1997c). Preparation of chiral synthon for HIV protease inhibitor: Stereoselective microbial reduction of N-protected a-aminochloroketone. Tetrahedron: Asymmetry 8, 2547-2552. Patel, R. N., Banerjee, A., Chu, L., Brzozowski, D., Nanduri, V., and Szarka, L. J. (1998). Microbial synthesis of chiral intermediates for [~3-receptor agonists. J. Am. Oil Chem. Soc. 75, 1473-1482. Patel, R. N., Banerjee, A., Nanduri, V., Goldberg, S., Johnston, R., Hanson, R., McNamee, C., Brzozowski, D., Tully, T., Ko, R., LaPorte, T., Cazzulino, D., Swaminathan, S., Parker, L., and Venit, J. (1999a). Biocatalytic preparation of a chiral synthon for a vasopeptidase inhibitor: Enzymatic conversion of N2-{N-[(phenylmethoxy)carbonyl]L-homocysteinyl]-L-lysine (11')-disulfide to [4S-(4a,7a,10ab)]l-octahydro-5•x••4•[(pheny•meth•xy)carb•ny•]amin•].7H.pyrid••[2•1•b][1•3]thiazepin•7•carb•xy• ic acid methyl ester by a novel L-lysine e-aminotransferase. Enzyme Microb. Technol. In press. Patel, R. N., Banerjee, A., Hanson, R. L, Brzozowski, D. B., Parker, L. W., and Szarka, L. J. (1999b). Oxidation of N-a-protected-L-lysine by Rhodotorula graminis to produce novel chiral compounds. Tetrahedron: Asymmetry 10, 31-36. Patel, R. N., Banerjee, A., Nanduri, V., and Comezoglu, F. T. (1999c). Enzymatic resolution of secondary alcohols. Biocatal. Biotransform. Manuscript submitted.

SYNTHESIS OF CHIRAL DRUG INTERMEDIATES

77

Pedragosa-Moreau, S., Archelas, A. R., and Furstoss, R. (1993). Microbiological transformations, 28: Enantioselective epoxide hydrolyses as a preparative access to both enantiomers of styrene oxide. J. Org. Chem. 58, 5533-5536. Pedragosa-Moreau, S., Archelas, A., and Furstoss, R. (1994). Microbiological transformations, 29: Enantioselective hydrolysis of epoxides use microorganisms: A mechanistic study. Biorg. Med. Chem. 2,609-616. Pedragosa-Moreau, S., Morisseau, C., Zylber, J., Archelas, A , Baratti, J., and Furstoss, R. (1996). Microbiological transformations, 33: Fungal epoxide hydrolases applied to the synthesis of enantiopure para-substituted styrene oxides. A mechanistic approach. J. Org. Chem. 61, 7402-7407. Pedragosa-Moreau, S., Morisseau, C., Baratti, J., Zylber, J., Archelas, A., and Furstoss, R. (1997). Microbiological transformations, 37: An enantioconvergent synthesis of the [~-blocker Nifenalol® using a combined chemoenzymatic approach. Tetrahedron 53, 9707-9714. Robl, J. A., and Cimarusti, M. P. (1994). A synthetic route for the generation of C-7 substituted azepinones. Tetrahedron Lett. 35, 1393-1396. Robl, J. A., Sun, C., Stevenson, J., Ryono, D. E., Simpkins, L. M., Cimarusti, M. A., Dejneka, T., Slusarchyk, W. A., Chao, S., Stratton, L., Misra, R. N., Bednarz, M. S., Asaad, M. M., Cheung, H. S., Aboa-Offei, B. E., Smith, P. L., Mathers, P. D., Fox, M., Schaeffer, T. R., Seymour, A. A., and Trippodo, N. C. (1997). Dual metalloprotease inhibitors: Mercaptoacetyl-based fused heterocyclic dipeptide mimetics as inhibitors of angiotensin-converting enzyme and neutral endopeptidase. J. Med. Chem. 40, 1570-1577. Rumbero, A., Martin, J. C., Lumbreras, M. A., Liras, P., and Esmahan, C. (1995). Chemical synthesis of allysine ethylene acetal and conversion in situ into 1-piperidine-6-carboxylic acid: Key intermediate of the a-aminoadipic acid for [3-1actam antibiotic biosynthesis. Bioorg. Med. Chem. 3, 1237-1240. Santaneillo, E., Ferraboschi, P., Grisenti, P., and Manzocchi, A. (1992). The biocatalytic approach to the preparation of enantiomerically pure chiral building blocks. Chem. Rev. 92, 1071-1140. Schfitte, H. Flossdorf, J. Sahm, H., and Kula, M.-R. (1976). Purification and properties of formaldehyde dehydrogenase and formate dehydrogenase from Candida boidinii. Eur. J. Biochem., 62, 151-160. Serizawa, N., Serizawa, S., Nakagawa, K., Furuya, K., Okazaki, T., and Tarahara, A. (1983). Microbial hydroxylation of ML-236B (compactin): Studies on microorganisms capable of 3~-hydroxylation of M1-236B. J. Antibiot. 36, 887-891. Shaked, Z., and Whitesides, G. M. (1980). Enzyme-catalyzed organic synthesis: NADH regeneration by using formate dehydrogenase. J. Am. Chem. Soc. 102, 7104-7105. Sher, P. M. (1994). "Preparation of 5-Substituted Benzo-l,3-Dioxo]es as [33 Adrenergic Receptor Agonists." U.S. Patent Application 5321036, 6o14-94. Shimizu, S., Hattori, S., Hata, H., and Yamada, H. (1987a). Stereoselective enzymic oxidation and reduction system for the production of D-(-)-pantoyl lactone from a racemic mixture of pantoyl lactone. Enzyme Microb. TechnoI. 9, 411-416. Shimizu, S., Hattori, S., Hata, H., and Yamada, H. (1987b). One-steo microbial conversion of a racemic mixture of pantoyl lactone to optically active D-(-)-pantoyl lactone. AppI. Environ. Microbiol. 53,519-522. Sih, C. J., and Chen, C. S. (1984). Microbial asymmetric synthesis: Enantioselective reduction of ketones. Angew. Chem. 96, 556-566. Sih, C. J., Gu, Q.-M., Holdgrun, X., and Harris, K. (1992). Optically active compounds via biocatalytic methods. Chirality 4, 91-97.

78

RAMESH N. PATEL

Soda, K., and Misono, H. (1968). L-lysine ~-ketoglutarate aminotransferase, II: Purification, crystallization, and properties. Biochemistry 7, 4110-4119. Soda, K., Misono, H., and Yamamoto, T. (1968). L-lysine c~-ketoglutarate aminotransferase, h Identification of product, 1-piperidine-6-carboxylic acid. Biochemistry 7, 41024109. Stecher, H., and Faber, K. (1997). Biocatalytic deracemization techniques: Dynamic resolutions and stereoinversion. Synthesis 1, 1-16. Steiger, A., Pyun, H. J., and Coates, R. M. (1992). Synthesis and characterization of aza analog inhibitors of squalene and geranylgeranyl diphosphate synthesis. J. Org. Chem. 57, 3444-3450. Stirling D. I. (1992). The use of aminotransferases for the production of chiral amino acids and amines. In "Chirality in Industry" (A. N. Collins, G. N. Sheldrake, and J. Crosby, eds.), pp 209-222. Wiley, New York. Takada, H., Yoshimura, T., Ohshima, T., Esaki, N., and Soda, K. (1991). Thermostable phenylalanine dehydrogenase of Thermoactinomyces intermedius: Cloning, expression, and sequencing of its gene. J. Biochem., 109, 371-376. Takahashi, E., Nakamichi, K., and Furui, M. (1995). R-(-)-mandelic acid production from racemic mandelic acids using Pseudomonas polycolor IFO 3918 and Micrococcus freudenreichi Ferm-P 13221. J. Ferment. Bioeng. 80, 247-250. Takeuchi, T., Iinuma, H., Kunimoto, S., Masuda, T., Ishizuka, M., Hamada, M., Naganawa, H., Kondo, S., and Umezawa, H. (1981). A new antitumor antibiotic, spergualin: Isolation and antitnmor activity. J. Antibiot. 34, 1619-1621. Umezawa, H., Kondo, S., Iinuma, H., Kunimoto, Y., Iwasawa, H., Ikeda, D., and Takeuchi, T. (1981). Structure of an antitumor antibiotic, spergualin. J. Antibiot. 34, 1622-1624. Umeda Y., Moriguchi, M., Katsushige, I., Kuroda, H., Nakamnra, T., Fujii, A., Takeuchi, T., and Humezawa, H. (1987). Synthesis and antitumor activity of spergualin analogues, III: Novel method for synthesis of optically active 15-deoxyspergualin and 15-deoxy-11-O-methylspergualin. J. Antibiot. 40, 1316-1324. Valentijn, A. R. P. M., de Harm, R., de Kant, E., van der Marel, G. A., Cohen, L. H., van Boom, J. H. (1995). Synthesis of a potential enzyme-specific inhibitor of squalene synthase. Red. Trans. Chim. Pays-Bas 114, 332-336. Wandel, U., Mischitz, M., Kroutil, W., and Faber, K. (1995). High selective asymmetric hydrolysis of 2,2-disubstituted epoxides using lyophilized cells of Rhodococcus sp. NCIMB 11216. J. Chem. Soc., Perkin Trans. 1, 735-736. Weijers, C. A. G. M. (1997). Enantioselective hydrolysis of aryl, alicyclic and apliphatic epoxides by Rhodotorula glutinis. Tetrahedron: Asymmetry 8, 639-647. Wilson, C., Wilson, S., Piercy, V., Sennitt, M. V., and Arch, J. R. S. (1984). The rat lipolytic beta-adrenoceptor: studies using novel [~-adrenoceptor agonist. Eur. J. Pharmacol. 100, 309-319. Wong, C.-H., and Whitesides, G. M. (1994). "Enzymes in Synthetic Organic Chemistry." Tetrahedron Organic Chemistry Series, VoL 12. Elsevier, New York. Zhang, J., Reddy, J., Roberge, C., Senanayake, C., Greasham, R., and Chartrain, M. (1995). Chiral bioresolution of racemic indene oxide by fungal epoxide hydrolases. J. Ferment. Bioeng. 80, 244-246.

Recent Developments in the Molecular Genetics of the Erythromycin-Producing Organism

Saccharopolyspora erythraea THOMAS

J.

VANDEN BOOM

Abbott Laboratories Fermentation Microbiology Research and Development North Chicago, Illinois 60064

I. II. III. IV.

V.

VI.

VII. VIII.

IX.

X.

Introduction Background Experimental Properties of S. erythraea Strains Characterization of the S. erythraea Genome A. Physical-Genetic Mapping of the Chromosome B. Genomic Polymorphisms in Industrially Improved S. erythraea Strains Introduction of DNA into S. erythraea A. Sonication-Dependent Electroporation B. Electroporation of Germinating Spores Transcriptional Organization and Regulation of the Erythromycin Biosynthetic Gene Cluster A. Previous Transcriptional Studies of the eryCI, ermE, and eryG Genes B. Construction and Analysis of Transcriptional Mutants in S. erythraea C. Erythromycin Biosynthetic Gene Cluster Promoters D. Transcriptional Overview of the ery Gene Cluster New Molecular Genetic Tools for Studying Gene Expression in S. erythraea Genetic-EngineeringApproaches to Industrial Strain Improvement A. Construction of High-Productivity Source Strains for Naturally Occurring Erythromycin Intermediates B. Two-Step Genetic-EngineeringApproaches for Optimization of Novel Macrolide Production in S. erythraea C. Introduction of the Vitreoscilla hemoglobin gene into S. erythraea Combinatorial Biosynthesis A. Manipulation of ery Biosynthetic Genes in Heterologous Streptomyces Hosts B. Manipulation of ery Biosynthetic Genes in S. erythraea Future Prospects References

I. Introduction N e a r l y 50 y e a r s s i n c e t h e m a c r o l i d e a n t i b i o t i c e r y t h r o m y c i n w a s first d e s c r i b e d ( M c G u i r e et al., 1952), t h e p r o d u c i n g m i c r o o r g a n i s m Saccharopolyspora erythraea r e m a i n s t h e s u b j e c t of k e e n i n d u s t r i a l i n t e r est. A n u m b e r of factors h a v e c o n t r i b u t e d to t h e o n g o i n g i n d u s t r i a l 79 ADVANCESINAPPLIEDMICROBIOLOGY.VOLUME47 Copyright©2000byAcademicPress Allrightsofreproductionin anyformreserved. 0065-2164/00$25.00

80

THOMAS J. VANDENBOOM

research interest in both erythromycin and S. erythraea. Prominent among these is the fact that various dosage forms of erythromycin continue to enjoy widespread use globally for a variety of indications due to the excellent record of therapeutic efficacy and safety achieved by erythromycin-derived products. Moreover, the introduction of second-generation semisynthetic erythromycin derivatives in the early 1990s created additional demands for bulk erythromycin A as the starting raw material for these products. The two major commercial second-generation erythromycin species, clarithromycin and azithromycin, are shown in Figure 1. The emergence of clinical isolates resistant to the second-generation macrolide antibiotics (Weisblum, 1998) has fueled continuing research in a number of industrial laboratories to develop third-generation erythromycin derivatives (Ma et al., 1999; Agouridas et al., 1998; Phan et al., 1997). Perhaps the most promising class of third-generation candidates currently in clinical development is the 3-oxo-erythromycin derivatives, or "ketolides." Two leading clinical candidates in this class, ABT-773 and HMR-3647, are also shown in Figure 1. Finally, there appears to be growing interest in the genes involved in erythromycin biosynthesis in the emerging field of combinatorial biosynthesis. Both the type I polyketide synthase (PKS) from S. erythraea (for reviews, see Hutchinson, 1998, 1999; Cane et al., 1998) and the related desosamine deoxysugar biosynthesis genes from Streptomyces venezuelae (Zhao et aL, 1998) have been successfully manipulated to produce hybrid microbial metabolites. In this review, I discuss recent advances in the molecular genetics of S. erythraea, with particular emphasis on current topics of industrial interest. Our present knowledge of the S. erythraea genome, as well as recent advances in molecular genetic methods applicable to wild-type and industrially improved strains of this organism, are considered here. In addition, this review summarizes recent studies on the transcriptional organization and regulation of the erythromycin biosynthetic gene cluster. These studies have improved our understanding of erythromycin gene expression in this organism and provide a foundation for future genetic manipulations of this industrially significant metabolic pathway. Finally, I briefly consider genetic-engineering approaches to erythromycin strain improvement and the role of S. erythraea and erythromycin biosynthetic genes in the emerging field of combinatorial biosynthesis. S. erythraea has received considerable attention as a model system for the study of polyketide biosynthesis. This topic is beyond the scope of this review. A brief overview of the biosynthesis of the erythromycin polyketide backbone is included herein simply as background for this review. The interested reader is referred to several

MOLECULAR GENETICS OF

Saccharopolyspora erythraea \

81

/ N

.o,,,? 'o'

CS

~ 2 oH

Erythromycin A

\

.-" N

I OH

"T~

/

N

~J

0

i

HO,,, N o.

O

HO Z

O~'

O'"

O

H

H O

Azithromycin

~[~

-

~, '~

Clarithromycin

~N

...--

,,,

\N/

~..o.....L~... O O

HMR-3647

ABT-773

FIG. 1. Erythromycin A and erythromycin-derived second- and third-generation anti-

biotics.

recent reviews that have appeared elsewhere for additional coverage of this topic (Staunton and Wilkinson, 1997; Katz, 1997; Khosla et al., 1999).

II. Background The Gram-positive actinomycete Saccharopolyspora erythraea produces the clinically significant macrolide antibiotic erythromycin A. The erythromycin biosynthetic gene cluster has been localized near one

82

THOMAS J. VANDENBOOM

end of the linear S. erythraea chromosome (Reeves eta]., 1998). The gene cluster has been cloned and sequenced and contains at least 20 genes involved in the formation and modification of the 14-membered macrolide 6-deoxyerythronolide B (6-DEB) and in the synthesis, attachment, and modification of the two deoxysugars desosamine and mycarose (Salahbey et al., 1998; for reviews, see Staunton and Wilkinson, 1997; Katz, 1997). Functions for the majority of genes located in this cluster have been proposed based on an analysis of blocked mutants constructed through targeted gene inactivation. A schematic view of the erythromycin biosynthetic pathway through the first bioactive erythromycin intermediate erythromycin D is shown in Figure 2. During erythromycin biosynthesis, the aglycone backbone 6-DEB is produced by a type I modular PKS from one propionyl-CoA and six (2S)-methylmalonyl-CoA molecules in a process closely resembling fatty acid biosynthesis (Staunton and Wilkinson, 1997). The 6DEB synthase (DEBS) is encoded by three large genes, designated eryAI, eryAII, and eryAIII, located roughly in the center of the biosynthetic gene cluster. The three multifunctional enzymes encoded by these genes each contains two modules, or sets of enzymatic activities, responsible for a single round of polyketide chain extension. The catalytic activities present in these modules dictate the stereochemistry and extent of reduction during each round of chain extension. In addition, the specificity of the initial loading module dictates the preferred starter units used by the DEBS enzyme (Weissman et al., 1998b). Following synthesis of the 6-DEB polyketide backbone, a specific hydroxylation occurs at the C-6 position to produce erythronolide B (EB). The C-6 hydroxylase responsible for this reaction is encoded by the eryF gene (Weber et al., 1991). EB is then modified by sequential attachment of mycarose and desosamine at the C-3 and C-5 hydroxyl groups, respectively, to produce the first bioactive intermediate, erythromycin D. Mutations affecting the synthesis and attachment of mycarose define eryB genes and result in phenotypic accumulation of the aglycone EB, whereas mutations affecting the synthesis and attachment of desosamine define e ~ C genes and result in phenotypic accumulation of 3-a-mycarosyl erythronolide B. The terminal steps of the erythromycin biosynthetic pathway form a metabolic grid in which erythromycin D is converted to erythromycin A by two alternative pathways (Fig. 3). Two modification enzymes, a specific mycarosyl O-methyltransferase encoded by the eryG gene (Paulus et al., 1990; Haydock et al., 1991) and a C-12 hydroxylase encoded by the eryK gene (Stassi et al., 1993), compete for the erythromycin D

MOLECULAR GENETICS OF

Saccharopolyspora erythraea

83

Z

o

:~

t~

e 0

w •

~

~0

"o

£ o ~

o

2-~ ©

o

o:

113

o

"

i

©

g o

9 <E o~ o-c: o

r7

84

THOMAS J. VANDENBOOM

j ~ , ERYC

ERY D

ERYA eryG~~

,~ryK'~ ERYB

O

OH3 :

. cY o.

HO .\r~

H3C\ /CH3 N HO-..,~

r ]

I ~ .G.% .,.,,,''~O"'~C H3

H3Rc'~ 1 ~ O 3 r,,-o~/CH3 OH2 i H2 u"~["~CH3 ~ . "'OH CH3 O H3C OR2

Erythromycin

R1

ERY A ERY B ERY C ERY D

OH H OH H

R2 CH3 CH3 H H

F]c. 3. Terminalbiosynthetic steps in the erythromycinA biosynthetic pathway.

pathway intermediate. Initial O-methylation of the mycarose moiety on erythromycin D leads to the bioactive intermediate erythromycin B. Subsequent C-12 hydroxylation of erythromycin B leads to erythromycin A. Alternatively, these reactions can be reversed. Initial C-12 hydroxylation of erythromycin D leads to the bioactive intermediate erythromycin C, which in turn is converted to erythromycin A by the specific mycarosyl O-methyltransferase. The latter route has been suggested as the preferred pathway based on kinetic studies of the C-12 hydroxylase enzyme (Lambalot eta]., 1995). II1. Experimental Properties of S.

erythraea Strains

Wild-type strains of S. erythraea (e.g., NRRL2338) are readily manipulated using minor variations of the molecular genetic methods devel-

MOLECULAR GENETICS OF

Saccharopolyspora erythraea

85

oped for the better-characterized Streptomyces species S. coelicolor and S. lividans (Hopwood et al., 1985). Strain NRRL2338, and closely related S. erythraea strains (e.g., ER720), have been widely used in molecular genetic studies to elucidate gene-function relationships in the erythromycin biosynthetic pathway. In addition, molecular genetic manipulations of the erythromycin polyketide synthase in these genetic backgrounds has led to isolation of a number of novel macrolide compounds (see below) and have provided some insights into the function of individual enzymatic domains in this complex enzyme system (Katz, 1997). Several variants of wild-type strain NRRL2338 are in use within the S. erythraea research community, including the NRRL2338 "red variant strain" used by Leadlay and coworkers (Hessler et al., 1997). These variants may differ slightly in their experimental handling properties (e.g., growth rate) and erythromycin productivities. In contrast to the wild-type variants noted above, a wide range of mutant strains of S. erythraea developed for the large-scale fermentative production of erythromycin have been less amenable to established streptomycete and wild-type S. erythraea molecular genetic methods (Fitzgerald eta]., 1998; Katz, 1997; Brunker et a]., 1998). Significant differences in protoplast formation and regeneration, recombination, genome structure, plasmid maintenance, drug resistance, and genetic stability have been observed between wild-type and certain industrially improved strains of this organism (unpublished observations). Industrially improved strains (e.g., CA340) represent the product of numerous cycles of mutagenesis and screening for improved erythromycin titers (or other desirable fermentation properties). Practical experimental differences between wild-type and industrially improved strains likely result from secondary mutations present in heavily mutagenized improved genetic backgrounds or reflect the pleiotropic nature of certain titer-enhancing mutations. These strain differences, although poorly understood, have provided an incentive for further development of molecular genetic tools applicable to the full range of wild-type and improved S. erythraea strains encountered in industrial applications. IV. Characterization of the S. erythraea Genome

A.

PHYSICAL-GENETIC MAPPING OF THE CHROMOSOME

A physical map of AseI and DraI restriction sites in the chromosome of S. er~hraea strain NRRL2338 was completed using high-resolution PFGE (Reeves et al., 1998). Summation of individual AseI, DraI, and AseI-DraI fragments revealed a chromosome size of roughly 8 Mb. This

86

THOMAS J. VANDEN BOOM

genome size is comparable to several previously characterized Streptomyces chromosomes (Kieser et al., 1992; LeBlond et al., 1993; Lezhava eta]., 1995; Pandza et al., 1997). The S. erythraea chromosome also shares several other features in common with previously characterized Streptomyces chromosomes. These features include a linear topology, the localization of genes involved in secondary metabolism near one end of the chromosome, and evidence for large genetically unstable regions of DNA (LeBlond et al., 1996; Aigle et al., 1996). In contrast to previously described Streptomyces chromosomes, no readily detectable terminal-inverted-repeat (TIR) sequences were observed in the S. erythraea chromosome when chromosomal end restriction fragments were hybridized to total AseI- and DraI-digested genomic DNA. It remains to be determined if short TIR sequences are present in this organism. Such TIR sequences would likely be undetectable using the large end restriction fragment probes employed in these experiments (Reeves et al., 1998). A total of 15 genetic loci have been mapped to specific AseI and DraI restriction fragments. The erythromycin biosynthetic gene cluster has been localized to an approximately 700-kb AseI-DraI restriction fragment located within 1.25 Mb from one end of the linear S. erythraea chromosome. Interestingly, several additional unlinked genes possibly involved in erythromycin biosynthesis or resistance, including ertX, a putative ABC-type transporter (O'Neill et al., 1995), gdh, a thymidine diphosphoglucose-4,6-dehydratase, and kde, a putative thymidine diphospho-4-keto-6-deoxyglucose 3,5-epimerase (Linton et al., 1995), have been localized to the same 7O0-kb region of the chromosome. In addition, this region also contains the attB site for the integrative S. erythraea plasmid pSE101 (Brown et al., 1988). B. GENOMIC POLYMORPHISMS IN INDUSTRIALLY IMPROVED S. ERYTHRAEA STRAINS

The AseI, DraI, and AseI-DraI chromosomal restriction digest profiles appear similar between wild-type and industrially improved strains of S. erythraea, except for two notable genomic polymorphisms. AseI chromosomal digests of strain CA340 reveal the loss of the 48-kb AseIN fragment and the appearance of an approximately 75-kb novel restriction fragment (Reeves et al., 1998). Strain CA340 produces roughly 10-fold more erythromycin than wild-type strain NRRL2338. It is at present not known if this AseI polymorphism is related to this productivity increase. In addition, more recent industrially improved erythro-

MOLECULARGENETICSOF Saccharopolyspora erythraea

87

mycin production strains, derived from strain CA340, harbor an additional roughly 150-kb chromosomal deletion (unpublished observation). Again, it is not known if this chromosomal polymorphism is directly related to the concomitant increase in erythromycin productivity observed in this genetic lineage. However, several genetic explanations seem plausible to explain the correlation between these genomic rearrangements and the phenotypic improvement in erythromycin biosynthesis: (1) competing secondary metabolic pathways might be deleted, (2) a negative trans-acting regulator might be deleted, or (3) a positive trans-acting regulator might be activated (or duplicated) as a result of the chromosomal rearrangement. Additional studies of these S. erythraea chromosomal polymorphisms should clarify the competing hypotheses outlined above, improve our understanding of the fluidity of the S. erythraea genome, and potentially facilitate future rational strain development efforts with this organism. The identification of the dispensable gene set present in the large 150-kb deleted region of the S. erythraea chromosome might also contribute to molecular genetic strain development approaches in other polyketide producing actinomycetes. V. Introduction of DNA into S. erythraea

Protoplast transformation techniques have proven effective for introduction of plasmid DNA into wild-type strains of S. erythraea at efficiencies of 105 to 106 transformants per ~g of replicating plasmid DNA (Yamamoto et al., 1986). Difficulties in extending this technique to industrially improved fermentation strains of S. erythraea led to the development of two complementary electroporation methods for introduction of DNA into this organism (Fitzgerald et al., 1998; English et aL, 1998). Although electroporation has found widespread application for introduction of DNA (and other macromolecules) into a broad range of cell types, there remain few reports describing application of this technology to industrially important filamentous organisms (Pigac and Schrempf, 1995; Tyurin and Livshits, 1996). A. SONICATION-DEPENDENTELECTROPORATION The development of electroporation techniques for S. erythraea was facilitated by the availability of a virulent bacteriophage for this host, designated CABT1, from the Abbott Laboratories Culture Collection (Fitzgerald et al., 1998). Concentrated preparations of the dsDNA

88

THOMAS J. VANDEN BOOM

genome of this bacteriophage are readily obtained using standard phage propagation and purification techniques (Sambrook et al., 1989). This reagent permitted development of a sensitive electrotransfection assay and avoided potential problems with S. erythraea host restriction of shuttle vectors propagated in heterologous genetic backgrounds. During attempts to prepare well-dispersed homogeneous suspensions of vegetative S. erythraea cultures in our laboratory, we made the fortuitous observation that sonication treatment rendered this organism electrocompetent. Subsequent experimentation demonstrated that the observed electrocompetence was strictly sonication dependent for the vegetative cultures being examined. Culture preparation, sonication, and electroporation conditions were optimized using the electrotransfection assay to achieve electroporation efficiencies of 1.2 x 103 plaque forming units per microgram of CABT1 DNA. A plasmid-based electrotransformation assay was also developed to optimize sonicationdependent electroporation conditions for plasmid DNA uptake. This system utilized an Escherichia coli-Streptomyces shuttle vector, designated pCD1, derived from the pJV1 replicon of Streptomyces phaeochromogenes (Bailey et al., 1986; Fitzgerald et al., 1998). This vector is poorly maintained by S. erythraea but permits recovery and scoring of primary thiostrepton-resistant transformants prior to eventual loss of the plasmid. The electrotransformation efficiency obtained with this plasmid was 1.0 × 104 thiostrepton-resistant transformants per microgram of pCD1 DNA. Interestingly, the positive effect of sonication on electrocompetence was eliminated when the sonicated hyphal fragments were returned to culture tubes and incubated with shaking for 60 min prior to electroporation, suggesting that the physical alteration responsible for electrocompetence was eliminated or repaired during this period. It is tempting to speculate that the mechanical disruption of long vegetative hyphal fragments by ultrasound treatment in this procedure perhaps exposes interior hyphal membranes more susceptible to electroporation mediated DNA transfer. Alternatively, the sonication treatment might alter the normal outer cell wall to facilitate DNA entry into the cell. B. ELECTROPORATION OF GERMINATING SPORES

The amount of nonviable cellular material in the sonicated S. erythraea culture preparations prompted us to look for alternative (sonication-independent) conditions to achieve electrocompetence. Toward this end,

MOLECULAR GENETICS OF

Saccharopolyspora ezythraea

89

cellular preparations from different stages of the developmental life cycle of S. erythraea were examined (English et al., 1998). The life cycle of S. erythraea is typical of other filamentous actinomycetes and has been reviewed elsewhere (Chater, 1998). This approach led to the discovery that germinating spores (germlings) of this organism move through a remarkably brief window of electrocompetence during outgrowth to vegetative hyphae. Under the culture conditions reported, the optimum outgrowth period for the harvest of electrocompetent germlings was between roughly 16 and 18 hours. The cellular physiological and/or morphological changes occurring during this period that result in electrocompetence are at present not understood. The utility of this method was demonstrated by constructing a targeted gene disruption of the pccA locus in the industrially improved S. erythraea strain CA340 (English et a]., 1998). The pccA gene encodes the biotinylated c~-subunit of the propionyl-CoA carboxylase from this organism. Electroporation efficiencies of 4-8 x 103 transformants per microgram of DNA were obtained with the suicide vector pJAY4 used in these experiments. Taken together with the protoplast transformation technique of Yamamoto et al. (1986), the two electroporation procedures reviewed here permit introduction of DNA into the full range of S. erj/thraea strains we have examined.

Vl. Transcriptional Organization and Regulation of the Erythromycin Biosynthetic Gene Cluster A. PREVIOUS TRANSCRIPTIONAL STUDIES OF THE ERYCI, ERME, AND ERYG GENES

Although DNA sequence analysis of the erythromycin biosynthetic gene cluster has provided some insights into the potential transcriptional organization of this gene cluster (Dhillon et al., 1989; Cortes et al., 1990; Donadio et al., 1991; Haydock et al., 1991; Donadio and Katz, 1992; Stassi et a]., 1993; Gaisser et al., 1997, 1998; Summers et al., 1997), detailed transcriptional studies have been lacking. Previous detailed transcriptional studies of ery genes have been limited to analysis of the eryCI-ermE region of the ery gene cluster reported by Bibb et al. (1994). A preliminary transcriptional study of the eryG gene has also been reported (Weber eta]., 1989). The regulatory region involved in divergent transcription of the eryCI and ermE genes contains two ermE and two eryCI promoters (Bibb et al., 1994). The ermE gene encodes an N-methyltransferase, which confers resistance to erythromycin in the producing host through methyla-

90

THOMAS J. VANDENBOOM

tion of the S. erythraea 23S rRNA (Thompson et al., 1982). $1 nuclease and exonuclease VII transcript mapping experiments identified two transcriptional start sites for ermE, designated ermEpl and ermEp2. These experiments were performed using RNA isolated from Streptomyces lividans strain TK24 containing the eryCI-ermE promoter region cloned on high-copy-number plasmids. The divergently transcribed erythromycin biosynthetic gene eryCI is also transcribed from two promoters, designated eryCIpl and eryCIp2. The eryCI gene is thought to encode an aminotransferase involved in the synthesis of desosamine. Interestingly, transcription from ermEpl is initiated at the same position as eryCIp2, but on the opposing DNA strand. The existence of tandemly arranged, divergently transcribed overlapping promoters in this region led to the suggestion by these workers that a high degree of coordinate regulation occurs with these genes. The transcriptional start site for ermEpl is located immediately adjacent to the ermE translational start codon. This results in transcription of a leaderless message for the ermE gene. The ermEpl, ermEp2, and eryCIpl promoters contain recognizable -10 and -35 regions. In contrast, no recognizable -10 or -35 sequences were evident in the eryCIp2 promoter region, suggesting the involvement of alternative transcription factors in initiation of transcription from this promoter. Transcription of the eryG gene, which encodes the mycarosyl Omethyltransferase enzyme, was previously examined by Weber et al. (1989). Using Northern hybridization experiments, these workers identified an approximately 1.3-kb eryG transcript. A possible promoter region immediately upstream was reported based on an analysis of a cloned S. erythraea DNA fragment in a luxAB reporter system in S. lividans. More recent evidence from our laboratory has identified, in addition to the approximately 1.3-kb transcript described earlier, an additional 2.6-kb eryG transcript identifiable in Northern hybridization experiments (Reeves et al., 1999). This larger transcript contains the eryBII gene located immediately upstream of eryG. From our analysis of transcriptional terminator mutations in this region (see below), the stable eryG and eryBII-eryG messages detected in Northern hybridization experiments appear to be derived primarily from one of two very large overlapping polycistronic transcripts in this region. There is no evidence for a functional promoter in S. erythraea immediately upstream of either the erzG or eryBII gene. The previous results from Weber and coworkers (1989) might be attributable to the use of the high-copy-number luxAB reporter group plasmid and heterologous genetic background used in the earlier experiments.

MOLECULARGENETICSOF Saccharopolyspora erythraea

91

B. CONSTRUCTIONAND ANALYSISOF TRANSCRIPTIONAL MUTANTSIN S. ERYTHRAEA The development of a novel transcriptional terminator cartridge for the targeted construction of polar transcriptional mutants in S. erythraea has facilitated transcriptional studies of the erythromycin gene cluster (Reeves et al., 1999). These studies have improved our understanding of the transcriptional organization of this industrially significant biosynthetic gene locus. The transcriptional terminator cartridge, designated trm, contains a 227-bp transcriptional terminator sequence obtained from the cloned S. erythraea ribosomal RNA rrnD operon. This terminator sequence is flanked by convenient multiple cloning site sequences to permit in-vitro insertion of the cartridge into targeted genes of interest. A native S. erythraea terminator sequence was selected to ensure functionality in this host. Two regions of predicted secondary structure are present in the trm sequence. The first is a stem-loop structure consisting of a 17-bp stem with a short 4-base loop. This stem-loop is immediately followed by a thymidine-rich region, characteristic of rho-independent terminators (Deng et al., 1987). The calculated AG of the stem-loop is -30 kcal/mol. The second predicted region of secondary structure in this sequence is a stem-loop consisting of an 18-bp stem with a 4-base loop, located 30 bp downstream from the first stem-loop. This structure has a predicted AG o f - 2 4 kcal/mol. The t~m cartridge was cloned into targeted ery genes and introduced into the corresponding S. erythraea chromosomal loci via homologous recombination using either a pWHM3-derived (Vara et al., 1989) or pJVI-derived (Bailey et al., 1986) integration vector. Several additional features make S. erythraea an attractive experimental system for molecular genetic studies of the erythromycin biosynthetic pathway. These include: (1) the availability of cloned ery genes and DNA sequences, (2) the availability of DNA transformation and gene replacement methods for this organism, (3) the availability of purified erythromycin pathway intermediates, and (4) the ability of this organism to utilize exogenously supplied pathway intermediates in erythromycin biosynthesis. In our genetic studies of the ery gene cluster, biotransformation experiments with erythromycin pathway intermediates provided a simple, but powerful method to analyze the effects of specific trn~ transcriptional terminator insertion mutations (designated ery::t~n~) on expression of downstream ery genes. A variety of molecular genetic and biochemical methods, including Northern hybridizations, Western blotting, and $1 nuclease protection assays, were

92

THOMASJ. VANDENBOOM

also employed in the characterization of the ery::trnn mutants. Two examples of the recently described S. erythraea tm~ insertion mutants (Reeves et a]., 1999) are reviewed here to illustrate the use of this methodology in the study of ery gene expression in wild-type and industrially improved genetic backgrounds. Based on these studies, the tm~ genetic element should also be of general use for the study of other genetic loci in this host. The tmn transcriptional terminator element was introduced in vitro into a cloned segment of the eryAI gene. This disrupted gene sequence was then introduced into the chromosome of wild-type strain NRRL2338 via homologous recombination to generate mutant NRRL2338 eryAI::tm~. The eryAI::tmn insertion was also introduced into the industrially improved genetic background of strain CA340 to generate mutant CA340 eryAI::tm~. As predicted, these eryA blocked mutants do not produce erythromycin A or any erythromycin pathway intermediates, consistent with previously characterized blocked mutants in the polyketide synthase genes. The central region of the erythromycin biosynthetic gene cluster contains, in addition to the three large eryA genes, four additional downstream genes. These seven genes, which include (in order) eryAL eryAII, eryAIII, eryCII, eryCIII, eryBII, and eryG, span approximately 35 kb of the erythromycin gene cluster. Several lines of evidence obtained from our analyses of the eryAI::tm~ mutants suggest that this set of genes is transcribed as a very large 35-kb polycistronic message. In biotransformation experiments, the eryAI::tmn mutation in both NRRL2338 and CA340 shows a polar effect on downstream ery genes, including the eryG Oomethyltransferase gene. Cultures of the eryAI::t~ mutant, when supplemented with erythromycin C, show greatly reduced levels of bioconversion of the erythromycin C to erythromycin A. Similar results were obtained in feeding experiments with other pathway intermediates, indicating an additional polar effect on both downstream eryB and eryC genes, as predicted. $1 nuclease protection assays using an eryG probe were also performed using total RNA extracted from the NRRL2338 eryAI::trnn and CA340 eryAI::i~nn mutant strains. A significant reduction of the eryG signal was observed in the insertion mutant strains relative to the parental controls, hterestingly, the polar effect of eryAI::tr~ insertion on eryG transcript levels appeared to be more pronounced in the CA340 industrially improved genetic background. This suggests that the promoter region immediately upstream of eryAI plays a more significant role in expression of the downstream eryG gene in the industrially improved strain CA340 than in the wild-type strain NRRL2338. Western blot experiments were also performed on wild-type strain NRRL2338 and the

MOLECULARGENETICSOF Saccharopolysporaerythraea

93

NRRL2338 eryAI::tm~ mutant to determine if the eryG O-methyltransferase protein was present in cell free extracts of these strains. A crossreacting band was readily detected in extracts of the parental control strain, but not in similarly prepared extracts of the eryAI::trnn mutant. In order to examine whether the eryAI::tm, insertion was having an effect other than on termination of transcription, an independent mutant strain was constructed using oligonucleotide-directed mutagenesis. In this mutant, designated ARR50, the predicted eryAI promoter ATrich -10 hexamer sequence (TATTGT) was replaced with the Sinai restriction enzyme recognition sequence CCCGGG. The polar phenotype of this strain was identical to that of the eryAI::trnn insertion mutant. Taken together, these experiments provide strong evidence that the eryAI, eryAII, eryAIII, eryBII, eryCII, eryCIII, and eryG genes are primarily cotranscribed from a promoter upstream of eryAL In addition, these experiments demonstrate the general utility of this new genetic element in the study of complex transcriptional units in this organism. C. ERYTtIROMYCINBIOSYNTHETICGENECLUSTERPROMOTERS The transcription start sites for seven additional ery gene cluster promoters, located in four ery gene cluster regulatory regions, have been reported (Reeves et al., 1999). These promoter regions include the 224-bp eryAI-eryBIV intergenic region, the 188-bp eryBI-eryBIII intergenic region, the 83-bp eryCVI-eryBVIintergenic region, and the region immediately upstream of the eryK gene. The 224-bp eryAI-eryBIV intergenic region contains an eryAI promoter divergently transcribed from two eryBIV promoters, designated eryBIVP1 and eryBIVP2. The eryAI transcription start site was located 27 bp upstream of the translation start codon for this gene. Based on an analysis of the eryAI::trnnmutant described above, the eryAI-containing transcript extends from eryAI to eryG. The transcription start sites for the eryBIVP1 and eryBIVP2 promoters are located 84-88 and 132 bp upstream of the predicted translational start codon for eryBIV, respectively. The eryBIVP1 promoter appears to be the major rightward promoter (as shown in Fig. 4) for this region based on the relative abundance of the respective Sl-protected fragments. The eryBIV-containing transcript is thought to extend from eryBIVto eryBVII, based on an analysis of ery::tmn insertion mutations in this region. This segment of the ery gene cluster also contains a smaller overlapping polycistronic transcript that extends from eryBVI through eryCVI. The promoter for this transcript is located in the 83-bp eryBVI-eryCVIintergenic region. The -35 region of the minor eryBIV promoter, eryBIVP2, is predicted

94

T H O M A S J. V A N D E N B O O M

~0

~3

! .I:i. oa°t "@~O

MOLECULAR GENETICS OF Saccharopolyspora erytt~raea

95

to overlap with the -35 region of the divergent eryAI promoter. On the basis of our transcriptional analysis of the cry gene cluster, the three promoters identified in this 224-bp regulatory region account for the transcription of roughly 80% of the gene cluster, or 14 of the 20 identified cry genes. The sequence of this regulatory region was determined for an industrially improved strain of S. erythraea and compared to the wild-type strain NRRL2338 sequence. Surprisingly, no mutational changes were identified in this regulatory region in the industrially improved genetic background (unpublished observation) despite the significant increase in cry transcript levels and erythromycin titers observed in this genetic strain lineage (Reeves eta]., 1999). The transcriptional activity of DNA fragments containing the 224-bp eryAI-eryBIV regulatory region was examined using a kanamycin/neomycin phosphotransferase (APH) reporter group in both S. erythraea strain NRRL2338 and S. lividans strain TK24 (Atkins and Baumberg, 1998). In S. lividans, appreciable APH activity was detected in APH fusion strains regardless of the orientation of the S. erythraea eryAI-eryBIV DNA fragment. This finding is consistent with the presence of divergently transcribed promoters in this region. The function of these promoters in S. lividans further supports the notion that a pathway-specific activator is not involved in regulation of the cry gene cluster and is consistent with the previous observation that the ermE and eryCI promoters also function in this heterologous host (Bibb et al., 1994). The 188-bp eryBI-eryBIII intergenic region contains two divergently transcribed promoters. Two potential transcriptional start sites, located 1-2 bp upstream of the predicted start codon, were identified upstream of the eryBIII gene. An eryBIII::tr,,n insertion mutant displayed a polar effect on the downstream eryF gene, indicating that the eryF gene is cotranscribed with eryBIII. The start site for the divergently transcribed monocistronic eryBI message is located 17-18 bp upstream of the predicted translational start for the eryBI gene. The transcription start site for the eryK gene, which encodes the P450-dependent C-12 hydroxylase, was localized to a region 45-50 bp upstream of the predicted TTG start codon for this gene and 9-14 bp from the predicted termination codon of the adjacent orf21 coding region (Pereda et al., 1997). This gene, which is located on one end of the cry gene cluster, is transcribed as a monocistronic message. D. TRANSCRIPTIONAL OVERVIEW OF THE ERY GENE CLUSTER

A current transcriptional map of the erythromycin biosynthetic gene cluster is shown in Figure 4. The monocistronic messages for eryCI and

96

THOMAS J. VANDENBOOM

ermE, described earlier, along with the more recently characterized eryBI and eryK transcripts (Reeves et al., 1999), represent the exception in this gene cluster, as the majority of ery genes appear to be transcribed as large polycistronic messages. This includes the previously reported eryG gene described above. Four polycistronic messages, including a large transcript of approximately 35 kb, account for the expression of roughly 85% of the gene cluster (Reeves et al., 1999). The two largest transcripts in the gene cluster extend divergently from the eryA-eryBIV intergenic region. The large 35-kb transcript includes the eryAL eryAII, eryAIII, eryCII, eryCIII, eryBII, and eryG genes. The divergently expressed transcript includes the eryBIV, eryBV, eryCVI, eryBVI, eryCIV, eryCV, and eryBVII genes. A second promoter upstream of the eryBVI gene produces an overlapping transcript that includes the eryBVI, eryCIV, eryCV, and eryBVII genes. Finally, a bicistronic message appears to be involved in expression of the eryBIII and eryF genes, and is divergently expressed from the monocistronic eryBI message.

VII. New Molecular Genetic Tools for Studying Gene Expression in S. erythraea

The gene encoding the native S. erythraea c¢-galactosidase enzyme, designated melA, has been cloned and sequenced (manuscript in preparation). The MelA enzyme was purified to near homogeneity and the N-terminal amino-acid sequence determined. Oligonucleotide primers based on the N-terminal amino-acid sequence and a conserved downstream region within the a-galactosidase gene were used to generate a 640-bp PCR product probe. Using this probe, the complete ct-galactosidase gene was identified in a lambda phage library of S. erythraea chromosomal DNA. The S. erythraea melA gene has been used to construct a reporter system for the study of gene expression in this organism. This system includes two components: (1) a parental S. erythraea strain containing a chromosomal deletion of the melA gene, and (2) a promoterless melA gene engineered into a pJVl-based integration vector. The melA integration plasmid is designated pDPE185. The promoterless melA gene cartridge present in pDPE185 is preceded by stop codons engineered in all three frames and includes a convenient upstream multiple cloning site. A ribosomal RNA terminator is present both upstream and downstream of the melA gene. Expression of the melA gene is conveniently monitored in liquid and solid agar cultures using either p-nitrophenol-

MOLECULAR GENETICS OF

Saccharopolyspora erythraea

97

0~-D-galactopyranoside (pNP{xG) or 5-bromo-4-chloro-3-indoyl-~-D-galactoside (X-~-gal), respectively. The function of this reporter group system was examined using the ermE* promoter region. The ermE* promoter fragment was inserted into the multiple cloning site of pDPE185. The resulting plasmid was introduced into the S. erythraea melA deletion host strain. Strains containing the ermE* promoter upstream of the melA gene had 10- to 15-fold more o~-galactosidase activity, as measured using the pNPRGbased assay, than the parental control strain. The melA reporter group system should be useful for constructing new S. erythraea vectors (e.g., insertional inactivation applications), for identifying new promoters, and in the study of gene expression in this organism (Satter, 1998).

VIII. Genetic-Engineering Approaches to Industrial Strain Improvement A. CONSTRUCTION OF HIGH-PRODUCTIVITY SOURCE STRAINS FOR NATURALLY OCCURRING ERYTHROMYCIN INTERMEDIATES

Genetic-engineering approaches have been used to introduce null mutations in the eryG and eryK genes, both individually and in combination, into various wild-type and industrially improved strains of S. erythraea. The analysis of mutants harboring disruptions of these genetic loci have permitted assignment of biochemical functions to these gene products (Paulus et al., 1990; Stassi eta]., 1993). In addition, genetically engineered eryG, eryK, and eryG/eryK null mutants represent important fermentation source strains for erythromycins C, B, and D, respectively (English et al., 1998). These erythromycin intermediates represent potential starting materials for development of new semisynthetic fermentation-based products (Faghih eta]., 1998). Naturally occurring erythromycin intermediates also find routine use among erythromycin manufacturers as analytical reference standards. Introduction of these mutant alleles into industrially improved erythromycin A-producing strains had little or no impact on overall macrolide titers {unpublished observation). This approach illustrates the utility of exploiting macrolide productivity levels of existing erythromycin A-producing industrial strains for production of other naturally occurring erythromycin intermediates. Moreover, this approach should also be applicable to the wide range of novel polyketide derivatives being generated through combinatorial biosynthesis methods with this organism (see below).

98

THOMAS J. VANDEN BOOM

B. TWO-STEP GENETIC ENGINEERING APPROACHES FOR OPTIMIZATION OF NOVEL MACROLIDE PRODUCTION IN S. ERYTHRAEA

Two reports by Stassi and coworkers (1998a, 1998b) illustrate the potential of rational strain development approaches for optimization of novel polyketide yields in genetically engineered strains of S. erythraea. The first report involves construction of a double mutant with disruptions in both the eryF gene, encoding the erythromycin C-6 hydroxylase, and the eryK gene, encoding the erythromycin C-12 hydroxylase. The predicted product of this double mutant is 6,12-dideoxyerythromycin A. Fermentations of the double eryF/eryK S. erythraea mutant produced a mixture of the desired product--6,12-dideoxyerythromycin A--and the immediate biosynthetic precursor--6,12-dideoxyerythromycin D--with the precursor representing the dominant product in the fermentation. In order to facilitate conversion of the 6,12-dideoxyerythromycin D precursor to the desired end-product, an additional copy of the eryG gene, encoding the 3"-O-methyltransferase enzyme responsible for this final biosynthetic step, was engineered into the eryF/eryK double mutant. The additional copy of the eryG gene under the control of the ermE* promoter was integrated into a chromosomal locus unlinked to the erythromycin gene cluster. The eryG diploid strain had significantly higher specific activities of EryG throughout the course of a 5-day fermentation. This increase in EryG activity resulted in quantitative conversion of the 6,12-dideoxyerythromycin D precursor to the desired 6,12-dideoxyerythromycin A product in fermentations of the eryF/eryK eryC--diploid strain. The second report involves addition of a heterologous primary metabolic enzyme to S. erythraea to improve precursor availability for a genetically engineered hybrid PKS. The ethylmalonate-specific acyltransferase domain from module five of the niddamycin PKS of Streptomyces caelestis (Kakavas et al., 1997) was substituted for the methylmalonate-specific acyltransferase domain present in module four of the erythromycin PKS (Stassi et al., 1998b). The predicted product of this mutant is 6-desmethyl-6-ethylerythromycin A. Interestingly, S. erythraea strains harboring this hybrid PKS still produced erythromycin A instead of the predicted hybrid polyketide product. The authors attributed this surprising result to the relaxed substrate specificity of the niddamycin acyltransferase domain used in this mutant construction and the limited intracellular availability of the ethylmalonyl-CoA precursor (relative to methylmalonyl-CoA) in this organism. In order to increase the available ethylmalonyl-CoA pools in this hybrid PKS mutant, the ccr gene from Streptomyces collinus expressed from the ermE*

MOLECULAR GENETICS OF Saccharopolyspora erflhraea

99

promoter was integrated at the same unlinked genetic locus used in the eryF/eryK mutant described above (Stassi et al., 1998a). The ccr gene encodes the crotonyl-CoA reductase enzyme responsible for the last step in the reductive biosynthesis of butyryl-CoA from two molecules of acetyl-CoA (Wallace et al., 1996) in S. collinus. Butyryl-CoA can then be carboxylated to form the desired ethyl malonyl-CoA precursor (Wallace et al., 1997). S. erythraea strains containing the S. collinus ccr gene had roughly 20-fold higher Ccr activity than the parental control strain. The addition of this activity to the PKS mutant led to production of the desired 6-desmethyl-6-ethylerythromycin A compound as the predominant product in fermentations of this strain. Taken together, these examples clearly demonstrate the utility of two-step genetic-engineering approaches to optimize both primary and secondary metabolic activities required for production of novel polyketides in S. erythraea. C. INTRODUCTION OF THE VITREOSCILLA HEMOGLOBIN GENE INTO S. ERYTHRAEA

Introduction of the Vitreoscilla hemoglobin gene (vhb) into an industrial strain of S. erythraea has been reported (Minas et al., 1998; Brunker et al., 1998). The vhb gene has been introduced into a variety of microorganisms, including Acremonium chrysogenum (DeModena et al., 1993), Escherichia coli (Khosla and Bailey, 1988), Corynebacterium glutamicum (Sander et al., 1994), Bacillus subtilis (Kallio and Bailey, 1996), and S. coelicolor (Magnolo et al., 1991) in attempts to improve oxygen metabolism and product yields in these organisms. However, the role of this gene in enhancing specific product yields in these systems remains speculative. The dramatic 70% improvement in erythromycin titers reported by these authors is difficult to assess due to the lack of information provided on the industrially improved S. erythraea strain used in these experiments. Nonetheless, this work represents the first reported attempt to improve erythromycin titers in S. erythraea through introduction of a heterologous gene into this organism. Functional expression of the vhb gene in the genetically engineered S. erythraea strain was confirmed by CO-difference spectrum assays. In the absence of selective pressure, the vhb integrant appeared to be genetically stable through at least a 9-day fermentation cycle. The molecular role of the Vhb protein in improving erythromycin titers in this strain is at present poorly understood. However, the improvement in erythromycin production observed in the genetically modified industrial strain did not appear to be due to simply an increase in biomass yield or to the pattern of mycelial fragmentation. The commercial

100

THOMAS J. VANDEN BOOM

utility of the S. erythraea::vhb strain described by these workers remains to be determined. Due to differences in industrial S. erythraea strain lineages, it will be of interest to determine whether the observed improvement in erythromycin yields is of general utility or limited to the model strain used by these workers.

IX. Combinatorial Biosynthesis In the past decade, intensive work in a number of laboratories (for reviews, see Staunton and Wilkinson, 1997; Katz, 1997) has led to significant advances in our understanding of the genetics and biochemistry of the erythromycin type I polyketide synthase in S. erythraea. Early molecular genetic studies of this system revealed the modular structure of this complex enzyme system and provided significant insights into the functional roles of individual enzymatic domains within the eryA-encoded PKS (Cortes eta]., 1990; Donadio et al., 1991). These studies also led to a series of successful genetic-engineering efforts involving targeted gene replacements of specific PKS modules to produce a number of novel erythromycin derivatives (reviewed by Katz, 1997). Although providing an important validation of this genetic-engineering approach to produce new molecular structures, this methodology suffered from several significant limitations. The construction of individual mutants was labor intensive and involved single mutational changes. In addition, mutants were produced in wild-type S. erythraea genetic backgrounds, leading to relatively low titers of the novel erythromycin derivatives produced in these strains. The development of combinatorial genetic-engineering approaches that permit introduction of multiple genetic alterations into the ery PKS represents a significant technical advance and illustrates the potential of this gene cluster for production of a wide range of polyketide compounds (McDaniel et al., 1999). Parallel advances in precursor-directed biosynthesis of novel polyketide derivatives using genetically engineered variants of erythromycin PKS further demonstrate the promise of S. erythraea and the role of erythromycin biosynthetic genes in the emerging field of combinatorial biosynthesis (Marsden et a]., 1998; Pacey et al., 1998; Weissman et al., 1998a). These technical developments have generated considerable interest of late, as reflected in the number of reviews that have appeared on this topic (Hutchinson, 1998, 1999; Cane et al., 1998; Hallis and Liu, 1999). In the context of this review, I consider briefly here the complementary combinatorial biosynthesis experimental platforms that have emerged for manipulation of erythromycin biosynthetic genes in both the native host S. erythraea

MOLECULARGENETICSOF Saccharopolyspora erythraea

101

and heterologous bacterial hosts. Continued development of appropriate experimental platforms to fully exploit the potential of S. erythraea and the ery gene cluster in combinatorial biosynthesis applications will play a significant role in successfully translating this exciting new technology into therapeutically useful and commercially viable products. Ultimately, the success of these approaches in drug discovery programs will be measured not by the number of new chemical entities produced in submilligram quantities, but rather by the number of viable therapeutic leads that move into preclinical and clinical testing. A. MANIPULATIONOF ERY BIOSYNTHETICGENESIN HETEROLOGOUSSTREPTOMYCESHOSTS The milestone report by McDaniel et al. (1999) illustrates the successful use of a heterologous system to manipulate the erythromycin PKS system. These authors report production of >100 6-DEB derivatives through construction of erythromycin PKS mutants harboring multiple combinations of individual eryA mutations in either S. coelicolor or S. lividans. The mutations include AT substitution, KR deletion, KR gainof-function, and KR stereochemical alterations. The use of S. coelicolor and S. lividans offers numerous experimental advantages because of the well-developed molecular genetic systems in these organisms (Hopwood et al., 1985; Gusek and Kinsela, 1992; Hopwood, 1997). This includes a wide range of plasmid vectors and gene expression tools, well-established protoplast transformation methods, and available genomic information (Kieser et al., 1992). These genetic tools facilitate construction of the numerous polyketide synthase mutants required to produce a library of 6-DEB polyketide derivatives. The host strains used by McDaniel and coworkers, S. coelicolor strain CH999 and S. lividans strain K4-114, contain a chromosomal deletion of the entire actinorhodin polyketide gene cluster (Ziermann and Betlach, 1999). Deletion of the actinorhodin biosynthetic genes in these hosts simplifies quantitative analysis of the novel polyketides produced and reduces competing metabolic demands for production of extrachromosomally encoded PKS proteins. The erythromycin PKS genes are expressed in these strains from the pCK7 expression plasmid, which has been reviewed elsewhere (Katz, 1997). Several limitations of this heterologous experimental platform should also be considered. First, the S. coelicolor system described by McDaniel et al. (1999) involved only the polyketide synthase genes from S. erythraea to produce 6-DEB derivatives. The erythromycin biosynthetic genes required for biosynthesis and attachment of the two

102

THOMAS J. VANDEN BOOM

deoxysugars mycarose and desosamine, and for the hydroxylation and O-methylation modification reactions were not provided in the heterologous host. This required separate introduction of the PKS mutations into S. erythraea to examine the impact of the polyketide modifications on the downstream biosynthetic reactions. The importance of the glycosylations and further macrolide ring and sugar modifications on the bioactivity of erythromycin derivatives underscores the importance of this step in evaluating this combinatorial library. Presumably, the genetic versatility of S. coelicolor and S. ]ividans would permit further genetic engineering of these hosts to include these ancillary reactions if desired. However, further development of the native S. erythraea host might represent a more attractive option since this organism already contains the required secondary metabolic pathways necessary to supply biosynthetic precursors for these pathways. Second, the yield of 6-DEB in the S. coelicolor control strain harboring the wild-type erythromycin PKS genes was approximately 20 mg/liter. All mutational changes in the PKS genes resulted in lower yields from this parental baseline, presumably reflecting differences in substrate specificity and processivity in the downstream polyketide biosynthetic reactions. The authors also noted that yield losses correlated with particular single mutations appeared to be additive when engineered in multiple combinations into the hybrid PKS. The resulting yields of the 6-DEB derivatives produced in the S. coelicolor library ranged from 2)-[D-Galp~ [~(1-->6)]-[3-D-fructose obtained from the batch reaction. This regioselectivity was probably due to immediate binding of trisaccharide products to the activated carbon column, thereby preventing attainment of equilibrium. The product spectrum was thus a consequence of differences in the activation energies of formation. Immobilized enzymes have also been used to synthesize oligosaccharides by the kinetic approach. Larsson et al. (1987) used immobilized

326

s.G. PRAPULLAet al.

~-galactosidase (on Sepharose CL-4B) to hydrolyze lactose in the presence of N-acetylgalactosamine, resulting in synthesis of D-Gal-[3(1-~6)D-Gal-NAC. Mozaffar et al. (1988) immobilized Bacillus circulans ~galactosidase on porous silica gel by crosslinkage with glutaraldehyde, and synthesized a range of oligosaccharides during lactose hydrolysis. Kery et al. (1991) reported the successful use of cellulose beads as a support matrix for immobilized [3-galactosidase. They synthesized a mixture of 6- and 3-1inked disaccharides and trisaccharides from p-nitrophenyl-~-galactoside as a glycosyl donor and methyl-c~-galactoside as an acceptor. Hayashi et al. (1991) immobilized a fructosyl-transferring enzyme from Aureobasidium sp. ATCC 20524, which produces 1-kestose from sucrose, onto a Shirasu (volcanic ash) porous glass (1142 U/g support). The packed column was stable for more than 30 days during continuous operation. Yun et al. (1992) investigated continuous production of fructooligosaccharides employing a packed-bed reactor charged with cells of A. pullulans KFCC 10245 immobilized on calcium alginate. They optimized conditions for reactor operation at a feed concentration of 860 g/liter, a feed rate (as superficial space velocity) of 0.2 per hour, and a temperature of 50°C. Under these optimal conditions they obtained a productivity of 180 g/liter/hr, with initial activity maintained for more than 100 days. They successfully scaled up the reactor to a production scale of 1000 liters. Yun and Song (1996) carried out continuous production of fructooligosaccharides from sucrose using fructosyltransferase immobilized on a highly porous resin, Diaion HPA 25. The optimal operation conditions of the immobilized enzyme column were a 600 g/liter of sucrose feed concentration and a feed rate as superficial space velocity of 2.7 per hour at pH 5.5 and 55°C. An 8% loss of initial activity was reported after 30 days of continuous operation when the column was rn at 50°C; a productivity of 1174 g/liter/hr was reported during this period. D. USE OF LIPID-COATEDGLYCOSIDASES The addition of water-miscible organic solvents increases the yield of transglycosylation catalyzed by glycosidases in a reverse reaction (Mori et al., 1997). However, the transglycosylation yields in these examples are usually low because the hydrolysis reactions proceed fast relative to transglycosylation in homogeneous aqueous organic media. Mori et al. (1997) prepared a lipid-coated [~-D-galactosidase in which the enzyme surface is covered with a lipid monolayer, and two long lipophilic alkyl tails serve to solubilize the enzyme in organic solvents. In a two-phase aqueous organic system, the lipid-coated enzyme exists

MICROBIAL PRODUCTION OF OLIGOSACCHARIDES

327

in the organic (2-propyl ether) phase and acts as an efficient transglycosylation catalyst for various hydrophobic alcohols with lactose in the aqueous buffer solution. Native [3-D-galactosidase, however, gave poor yields of both transglycosylation and hydrolysis reactions in a twophase system due to denaturation of the enzyme at the interface. E. USE OF RECOMBINANTS

Large-scale synthesis of specific tailor-made oligosaccharides is increasingly becoming essential. This has increased the use of glycosyltransferases over the glycosyl hydrolases, as the former display high regiospecificity for the acceptor and the donor substrate. The major drawback in utilization of glycosyltransferases is a lack of availability and the prohibitive cost of the sugar nucleotides used as activated sugar donors. To tackle this drawback, Samain et al. (1997) used growing bacterial cells as natural minireactors for continuous regeneration of sugar nucleotides and used the intracellular pool of sugar nucleosides as the substrate for in-vivo synthesis of "recombinant" oligosaccharides by recombinant glycosyltransferases. The method reported by them described cultivation of E. coli harboring genes for oligosaccharide synthesis. They produced pento-Noacetyl chitopentose and its deacetylated derivative tetra-N-acetyl-chitopentose in high yields (up to 2.5 g/liter) by cultivating at high density cells of E. coli expressing nodC or nodBC genes (nodC encodes for chitooligosaccharide synthesis and nodBC for chitooligosaccharide N-deacetylase). V. Assays and Structural Determination of Oligosaccharides Measurement of enzyme activity and determination of oligosaccharide concentration and yield require a knowledge of the methods for estimation of the compound itself. Assays for oligosaccharides have largely been reported using chromatographic methods, primarily by high-performance liquid chromatography (HPLC). Other techniques reported by various authors include paper chromatography, thin-layer chromatography (TLC), mass spectrometry (MS), and nuclear magnetic resonance (NMR).

A. HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY Advances in high-performance liquid chromatography (HPLC) with particular reference to column and instrument technology have led to speedy analysis of oligosaccharides (Table I/I). Jeon and Mantha (1985) reported that both polar-bonded-phase and resin-based HPLC columns are commonly used for carbohydrate separation. The polar-bonded phases are efficient, and carbohydrates elute in order of increasing

328

S. G. P R A P U L L A e t a l . T A B L E III ANALYSIS OF OLIGOSACCHARIDES USING H P L C

Column Aminospheri-5 (4.6 × 250 mm) Spherogel carbohydrate (7.5 x 30 cm) Shimadzu PNH2 (4.6 × 150 mm)

Column temperature

Mobile phases

Flow rate ml/min

Detection system

Source

Ambient

Acet onitrile:distilled water 75:25 (v/v)

2.0

Refractive index (RI)

Jeon & Mantha (1985)

80°C

Distilled water

0.6

RI

Jeon & Mantha (1985) (1985)

Ambient

Acetonitrile:distilled water 75:25 (v/v)

1.0

RI

Hidaka etal. (1988)

Acetonitrile:distilled water 75:25 (v/v)

1.5

RI

Jung etal. (1987) Duan eta]. (1994)

Bondapak carbohydrate (10.4 × 300 mm) Wakopack WB-T-130E (7.8 × 300 mm)

60°C

Distilled water

0.2

RI

Hayashi et el. (1991)

Shodex Ionopak ks-801 (8 × 300 nm)

7O°C

Distilled water

1.0

RI

Nakao et el. (1993)

Asahipak NH2P-50 (5 × 250 mm)

25°C

Acetonitrile:distilled water 75:25 (v/v)

1.0

RI

Nakao eta]. (1993)

A m i n e x HP X 42C (7.8 × 300 mm)

85°C

Water

1.0

RI

Yun etal. (1995, 1997), Yun & Song (1993, 1996)

A m i n e x HP X 42C (7.8 x 300 mm)

85°C

Acetonitrile:distilled water 75:25 (v/v)

0.75

RI

Dumortier etal. (1994)

Waters p Bondasphere 5 p NH2 - IoOA (3.9 × 150 mm)

30°C

Acetonitrile:distilled water 75:25 (v/v)

1.0

RI

Takeda etal. (1994)

Sperisorb NH2 (4.6 × 250 ram)

Acetonitrile:distilled water 75:25 (v/v)

1.8

RI

Barthomeuf & Pourrat (1995)

Nucleosil capcell 5 pm

A m m o n i u m acetate 50 m M

0.6

RI

Samain et al. (1997)

Capcell Pak C18 (3.9 x 250 mm)

Distilled water

-

*

Barthomeuf etal. (1997)

*168-diode array spectrophotometer.

MICROBIALPRODUCTIONOF OLIGOSACCHARIDES

329

monosaccharide chain length. On the other hand, components elute in order of decreasing molecular size from resin-based columns. Resinbased columns are increasingly being studied for use in analysis of small oligosaccharides (with resolution of polymers limited to a degree of polymerization equal to 4). Concentrations of oligosaccharides can be determined by weighing the individual peaks and comparing them to standard curves. This method has been widely used. The accuracy, sensitivity, reproducibility, and ease of operation with which the samples can be analyzed has made this a very popular technique. B. PAPERCHROMATOGRAPHY Toba et al. (1980) detailed a method for estimation of oligosaccharides by paper chromatography using sheets (40 x 40 cm) of Toyo No. 514 paper. The papers were developed in a butanol:pyridine:water (6:4:3) medium. Sugars were detected by an aniline hydrogen phthalate (AHP) reagent. Toba and Adachi (1978) described two-dimensional analysis of oligosaccharides by paper chromatoelectrophoresis. They carried out detection and characterization of oligosaccharides with the use of diphenyl amine aniline phosphoric acid (DAAP), triphenyl tetrazolium chloride (TTC), and AHP. This method is simple and is now widely used for separation of individual oligosaccharides, but it lacks sophistication and accuracy and is more suited to qualitative analysis rather than quantification of reaction products. Paper chromatoelectrophoresis is an improved version of this method, it is not routinely used because of its high cost. C. GAS-LIQUIDCHROMATOGRAPHY Toba and Adachi (1978) carried out gas-liquid chromatography (GLC) of oligosaccharide trimethylsilyl (TMS) esters with a Hitachi model 063 gas chromatograph fitted with a hydrogen flame ionization detector and a stainless steel column (2 m x 3 mm). Nitrogen was used as a carrier gas at a flow rate of 30 ml/min, with temperatures rising from 150 to 300°C at 3 °C/rain. They obtained trimethyl derivatives of the transfer reaction products by shaking the reaction products with pyridine:hexamethyldisilazane:trifluoroacetic acid (10:9:1, v/v), and separating on a column of Chromosorb W coated with 1.5% SE-52. They also made a detailed GLC analysis of the methyl glycosides using a stainless steel

330

s.G. PRAPULLAet al.

column (2 m X 3 mm) packed with 8% diethylene glycol succinate on Diasolid L operated at 200°C, with nitrogen as a carrier gas at 20 ml/min. They achieved methylation of oligosaccharides by shaking 0.5-2.0 mg of sample with 0.2 ml methyl iodide, 0.2 ml N,N-dimethylformamide, and 0.2 g silver oxide at room temperature in the dark for 18 hours. The mixture was then filtered, the residue washed with chloroform, and the filtrate evaporated to dryness. The products were boiled with 5% methanolic hydrogen chloride for 8 hours. The resulting methylglycosides were identified by comparing their retention times with that of standards. Burvall e t al. (1979) reported the use of a 2-m glass column packed with 3% ECNSS-M on Gas Chrom Q (100-200 mesh) at a specified column temperature for sugar alditol acetates and at 160°C for partially methylated alditol acetates. For permethylated alditol derivatives of dito tetrasaccharides, they recommended the use of a glass capillary column (25 m x 0.25 mm) wall coated with SE-30 at a specified column temperature and at 160°C for partially methylated alditol acetates. Stevenson e t al. (1993) prepared samples for GLC analysis by adding 10 gl of quenched reaction sample to acetonitrile (100 gl) and evaporating to dryness at 35°C under a stream of air in a block heater. The solid residue was trifluoroacetylated by addition of dry pyridine (5 gl), dichloromethane (65 ~1), and trifluoroacetic anhydride (30 ~1). They injected 1 ~1 of sample after incubation at 35°C for 15 min onto a gas chromatograph fitted with an S.G.EBP1 (0.22 mm x 25 m) capillary column and a flame ionization detector. They maintained injector and detector temperatures at 300°C and oven temperature at 120°C for 2.5 min after injection, then increased by 10°C/min to 270°C. This is a powerful technique for both qualitative and quantitative analysis with a high degree of accuracy. However, reaction products need to be converted to trimethyl derivatives, which is a laborious process. Much care needs to be exercised during methylation of sugars; otherwise, incomplete methylation leads to erroneous results and affects column performance. D. THIN-LAYERCHROMATOGRAPHY Wierzbicki and Kosikowski (1973) recommended using 0.25 mm thick plates of silica gel for analysis of oligosaccharides. A mixture of n-butanol:acetic acid:diethylether:water (9:6:3:1) was used as the developing solvent and the color-developing reagent was prepared by dissolv-

MICROBIALPRODUCTIONOF OLIGOSACCHARIDES

331

ing 5 g of benzidine in acidified ethanol. After development, the plates were dried in an oven at 100°C for 10-15 minutes, sprayed with color reagent, and heated to 100°C for 25 minutes. Mono- and oligosaccharides gave yellow brown spots under daylight and green spots under short-wavelength UV. Balken et al. (1991) qualitatively determined oligosaccharides by TLC on silica gel plates using of chloroform:ammonium hydroxide:methanol solvent system (60:20:54). Spots were visualized by spraying with 1-naphthol or phosphomolybdate. Prakash et al. (1989) suggested a similar process with use of ethyl acetate:acetic acid:water (2:1:1, v/v/v) as the solvent system. Quantitative estimation was done by extracting individual spots developed on TLC using a mixture of 40% trichloroacetic acid in water, acetic acid, and ethanol (1:1:8) and measuring the optical density at 400 mn. Concentrations were determined from a standard curve prepared with lactose (Wierzbicki and Kosikowski, 1973). Stevenson et al. (1993) performed TLC using silica gel 60 F254 on aluminum foil plates with ethyl acetate:pyridine:water (33:13:4). Compounds were visualized by dipping the plates in a mixture of 5% (v/v) concentrated sulfuric acid and 2% (w/v) anisaldehyde in 95% ethanol and heating with hot air until spots appeared. Trincone et al. (1991) monitored formation of phenyl iS-D-glycosides by carrying out TLC using chloroform:water (8:2, v/v) as the eluant. TLC is one of the simplest methods for analysis. However, the accuracy with which the eluted spots can be extracted is poor, so that it does not find wide application. F,. 13C-NMR ANALYSIS This technique has largely been used for elucidation of the structure elucidation of the oligosaccharides. Prakash et al. (1989) obtained proton-decoupled 13C-NMR spectra of oligosaccharides at 25 MHz with a JEOL FX100 operated in pulsed Fourier transform mode at 30°C using D20 as a solvent. Chemical shifts were referenced to an external standard 2,2,3,3-tetra-deuterio-4,4-dimethyl-4-sila pentanoate (TSP-d4). Barthomeuf et al. (1997) carried out NMR analysis of oligosaccharides at 100 MHz using a Bruker AC 400 spectrophotometer and a sample containing approximately 5 mg/ml sugar in deuterium oxide:water (1:2, v/v). They expressed the results obtained in ppm from the signal of trimethyl silane (TMS) and referenced chemical shifts using acetone as an internal standard. Samples used for NMR analysis were obtained by semipreparative HPLC. NMR is one of the most powerful techniques

332

s . G . PRAPULLA et al.

available for structural elucidation. The cost of the equipment is its major constraint. Moreover, the protocol for sample preparation is much more elaborate, and it is necessary to carry out semipreparative HPLC before samples are subjected to NMR analysis. F. FAB MASS SPECTROMETRY Prakash eta]. (1989) used FAB mass sPectrometry for elucidation of the major galactooligosaccharides formed by Trichoderma harzianum. They recorded positive-ion fast atom bombardment mass spectra in a JMS DX300 mass spectrometer. The sample and stainless steel probe tip were introduced into the FAB source and bombarded by a 3-keV argon beam. They dissolved the oligosaccharide sample (-5 mg) in about 5% sodium acetate and loaded it on a glycerol-coated probe tip. Similar to NMR, this method is more applicable for structural elucidation rather than routine analysis. Elaborate sample preparation steps and the initial cost of the equipment are limitations on its wide application.

Vl. Applications of Oligosaccharides Oligosaccharides have application in both nonfood and food-related fields. A. NONFOOD APPLICATIONS

The most important nonfood applications of oligosaccharides are in the medical and biotechnological fields. These compounds play a wide range of roles within living cells, often acting as recognition molecules. Microbial production allows not only easy availability of oligosaccharides but also facilitates synthesis of "tailor-made" compounds, thus giving additional thrust to studies focused on elucidating the molecular mechanism of a variety of biological functions. Oligosaccharides, generally as glycoconjugates, have been implicated in cell-cell, cell-virus, and cell-bacteria interactions (Rademacher et al., 1988; Schnaar, 1991). The lipopolysaccharide O antigen found in Gram-negative bacteria, the blood group determinants, and several tumor-associated antigens have potential as diagnostic tools. Oligosaccharides regulate cellular differentiation and act as inducers of disease-resistance responses following invasion by fllngal pathogens (McNeil et al., 1984). Bacterial cell-surface oligosaccharides are reported to be

MICROBIAL PRODUCTION OF OLIGOSACCHARIDES

333

responsible for host recognition and binding following infections of root hair with nitrogen-fixing bacteria (Lerouge et al., 1990). Hedbys et al. (1984) suggested the potential use of 6-O-~-D-galactopyranosyl-2acetamido-2-deoxy-D-galactose, formed by the transgalactosylase activity of E. coli [5-galactosidase, as a cell surface receptor. Ooi et al. (1985) used the [~-galactosidase of A. oryzae to synthesize glycosides of digoxigenin. A [3-galactosyl-serine-glycoprotein component was synthesized separately by Cantacuzene et al. (1991) and Sauerbrei and Thiem (1992) using the ~-galactosidase of E. coli and A. niger. Hedbys et al. (1989) were able to synthesize the blood group determinant Gal-[31-3Glc-NAC[~-SEt using [~-galactosidase from E. coil A Salmonella-inhibiting therapeutic composition containing fructooligosaccharide (largely Neosugar) as the effective compound was patented by Coors Biotechnology (Speights, 1990; Speights et al., 1991). Other nonfood applications of oligosaccharides include their use in drug delivery, cosmetics, and mouthwash. A Japanese company has patented the use of an enzymatically synthesized, nonreducing, neotrehalose-like oligosaccharide for use in cosmetics and pharmaceuticals. This compound is stable, not susceptible to crystallization, and has appropriate viscosity Aga eta]., 1995). Nonfermentable oligosaccharides have been used to manufacture a high-density liquid preparation of yeast (>800 g/liter) with improved activity retention, easy dissolution, uniform suspendability, and greater tolerance to repeated freeze-thaw operations (Suoranta, 1991). Chitooligosaccharides are also emerging as an important class of oligosaccharides with a diversity of applications, including use as antimicrobials, plant growth inhibitors, and cosmetic components (Crittenden and Playne, 1996). B. FOODAPPLICATIONS Oligosaccharides have grown immensely popular as food ingredients, particularly in Japan and Europe, largely due to the health benefits associated with their consumption. Several foods containing oligosaccharides as the functional ingredient have obtained FOSHU (Foods of Specific Health Use) status under Japanese federal legislation. In fact, foods incorporating oligosaccharides comprise 34 of the 58 approved foods in the 1996 FOSHU list. Oligosaccharides have excellent functional properties and improve the general colonic environment. They are bifidogenic factors and improve the growth of indigenous bifidobacteria (Crittenden and Playne,

334

s.G. PRAPULLAet al.

1996). The growth of bifidobacteria has many beneficial effects, such as improved digestion and absorption, increased vitamin availability, and prevention of gut colonization by pathogens and putrefactive bacteria. The current interest in application of bifidobacteria to improve colonic health has made oligosaccharides quite popular. Oligosaccharides are being included in probiotic yogurt and yogurt drinks to produce synbiotic products. "Bifiel," produced by Yakult Honsha, Japan (Kan et al., 1989) contains galactooligosaccharides, while fructooligosaccharides have been incorporated into "Symbalance" (Toni Milch, Switzerland), "Fyos" (Nutricia, Belgium), and "Fysiq" (Mona, Netherlands). Barthomeuf et al. (1997) reported that bifidogenic activity is optimum with short oligofructosaccharides in which the fructosyl units (n -- 2-8), are bound by a 132-1 linkage. New physiological effects of oligosaccharides consumption continue to be elucidated, including possible protection against the development of colon cancer. Oligosaccharides help prevent constipation because they are indigestible and so have an effect similar to dietary fiber. Apart from general enhancement of the colonic environment, oligosaccharides have been shown to exhibit protective action against diarrhea caused by the heat-stable enterotoxin of E. coli (Sta). Sta c a u s e s diarrhea by stimulating intestinally bound guanylate cyclase. Using the T84 human colon carcinoma cell line, these studies showed that fucosylated oligosaccharides could inhibit Sta-stimulated guanylate cyclase by 6080%, whereas other oligosaccharides exhibited a 17-18% inhibitory effect on the enzyme (Daniel et al., 1996). Oligosaccharides are incorporated into several health drinks and athletic beverages. A general-purpose sports beverage made up of oligosaccharides (20 g/liter), sucrose (55 g/liter), citric acid (1.8 g/liter), citric aroma (1.0 g/liter), and NaC1 (1.0 g/liter) has been patented (Korduner et al., 1982). Another product especially adapted for rapid administration of water and carbohydrates during heavy muscle work has been patented by Pripps Bryggerier (Gyllang et al., 1986). It is a monosaccharide-free solution containing 3-25% (by weight) of a mixture of soluble oligosaccharides. The same company holds an international patent (Newsholme et al., 1988) for an instant beverage mix for athletes with 0-750 g/liter oligosaccharides with 0-55 g/liter monosaccharides and 2-40 g/liter amino acids. A Japanese company holds an international patent (Takaichi et al., 1991) for a liquid nutrient product containing three to six oligosaccharides.

MICROBIALPRODUCTIONOF OLIGOSACCHARIDES

335

1. Oligosaccharides in Processed Foods

Oligosaccharides are used in the confectionery, beverage, and bakery industries not as pure products, but as mixtures containing oligosaccharides with various degrees of polymerization. The choice of an appropriate mixture for a particular food application depends on the specific physicochemical and physiological properties of the oligosaccharides, which in turn depends on their structure. Oligosaccharides are typically 30-60% as sweet as sucrose, which makes them desirable in foods when a low sweetness level is used to enhance other food flavors. In conjunction with intense artificial sweeteners, they help to mask certain aftertastes. The primary advantage of using oligosaccharides is that they provide bulking properties almost identical to that of sucrose and, being indigestible, they are safe for consumption by individuals suffering from insulin-dependent diabetes. Oligosaccharides are therefore being widely used in confections, jams, marmalades, and desserts (Crittenden and Playne, 1996) as low-calorie, noncariogenic sugar substitutes. The use of 15-40% oligosaccharides in confectionery syrup was patented by the Corn Products Company (Walon, 1971). Hayashibira (1974) patented the use of fructooligosaccharides as sweetening agents in the manufacture of frozen dairy desserts, ice creams, sweetened condensed milk, sweetened dry milk, and other dairy products. The use of oligosaccharides with one to four fructose units linked to sucrose as low cariogenic sweeteners was patented by Meiji Seika Kaisha (Adachi and Hidaka, 1981) in 1981. The same company holds an American patent (Adachi and Hidaka, 1987) for the use of microbially produced fructooligosaccharides as low-calorie noncariogenic sweeteners. Tate and Lyle, a British company, patented (Beyts, 1989) the use of glucooligosaccharides in a synergistic mixture of sweeteners to be used in dietetic foods, beverages, bakery products, and confections. The use of fructooligosaccharides as normal and low-calorie sweeteners is covered by a French patent (Biton et al., 1989) held by Ronssel-Uclaf of South Africa. Meiji Seika patented a process for production of a candy that has pressurized gas entrapped within it along with incorporation of oligosaccharides. On consumption, the candy releases the gas, generating a pleasant taste in the mouth (Sumi et a]., 1991). Roquette Fr~res patented (Mentink and Serpelloni, 1992) a low-calorie chocolate in which the sweetener is an oligosaccharide.

336

s.G. PRAPULLAet al.

Raffinerie Tirlemontoise, a Belgium company, obtained an international patent (de Soete and Freund, 1993) for a low-calorie chocolate confectionery product in which sugar is wholly or partially replaced by linear or branched fructooligosaccharides with or without the addition of intense sweeteners. Unilever holds a European patent (Plug, 1995) describing the use of oligosaccharides in foods. Due to their noncariogenicity, low calorific value, and viscosity, oligosaccharides are being increasingly incorporated into a variety of chewing gums. The Wrigley Company (United States) has developed a petroleum wax-free chewing gum incorporating oligosaccharides that act as binders. The oligosaccharides described were fructooligosaccharide, isomaltose, and oligofructose (Yatka et al., 1994). The same company holds a patent (Yatka et al., 1995) for the manufacture of other chewing gums containing oligosaccharides. Embodiments of the product include use of oligosaccharides as a rolling compound applied to the product, use as a coating such as a hard shell, for pellet gum, and as center fill in chewing gum. The patent also covers co-drying of oligofructose with other sweeteners and evaporation to make syrups for use as encapsulating agents for high-intensity sweeteners or flavors used in gums. The various functional and physicochemical properties of oligosaccharides have popularized their use in the bakery industry, especially since they have also been shown to act as strong inhibitors of starch retrogradation (Nakakuki, 1993). A 1988 patent (Kono et al., 1988) describes oligosaccharides originating from agar and/or carrageenan as being effective in preventing retrogradation of gelatinized starch. Yakult Honsa (Japan) described a method for producing bread with a galactooligosaccharide incorporated at 2.6% by weight. They also hold a patent for the process of making bread containing Gal-(Gal)n (Sonoike et al., 1992). Partially esterified oligosaccharides are suitable as fat substitutes since they have good sensory properties and have no calorific value, as they are not substantially hydrolyzed in the digestive tract. They have excellent mouth feel and many other properties similar to vegetable oils and fats (White, 1990, 1991). Carboxylic acid esters of oligosaccharides have been used in the preparation of edible water-in-oil and oil-in-water emulsions with reduced energy and fat content, as in the preparation of margarine, cream, and mayonnaise. A novel branched oligosaccharide synthesized by Aspergillus s y d o w i fructosyltransferase has also been reported to be useful as a food additive (Mieth et al., 1985).

MICROBIAL PRODUCTION OF OLIGOSACCHARIDES

337

VII. Conclusions

Oligosaccharides have emerged as functional foods with varied applications in response to an increasing consumer demand for healthier foods. The functional properties of oligosaccharides make them unique in food applications, and their popularity and demand is continually increasing. In order to meet the rising demand, various research groups have successfully produced them using microbes. Microbial production of oligosaccharides is an effective alternative to chemical synthesis. A whole range of novel oligosaccharides are being developed. Research efforts directed at production of novel oligosaccharides with varied functional properties will be a future trend. Application of novel production techniques for better yields is also envisaged. There is tremendous potential for broader elucidation of the physiological effects of oligosaccharides. Unique physicochemical properties, combined with health benefits, make the oligosaccharides interesting from the point of view of research and development, along with the large scope of potential commercialization.

Acknowledgments The authors thank Mr. M. N. Ramesh, a scientist in the Food Engineering Department, who helped us compile and prepare this chapter.

REFERENCES

Adachi, T., and Hidaka, H. (1981). "A Sweetener." U.K. Patent Application 2072679A. Adachi, T., and Hidaka, H. (1987). "Sweetener." U.S. Patent Application US4681771. Aga, H., Shibuya, T., Sugimoto, T., and Miyake, T. (1995). "Non-Reducing Oligosaccharide with Neotrehalose Structure, and Its Production and Uses." European Patent Application, EP0662479A2, JP93-342187 (19931215). Ajisaka, K., and Fujimoto, H. (1989). Regioselective synthesis of trisaccharides by use of the reversed hydrolysis activity of (z- and [~-D-galactosidase. Carbohydr. Res. 185, 139-146. Ajisaka, K., Nishida, H., and Fujimoto, H. (1987). Use of an activated carbon column for the synthesis of disaccharides by use of a reverse hydrolysis activity of [~-galactosidase. Biotechnol. Lett. 9, 387-392. Balken, J. A. M., Van Dooren, Th. J. G. M., Van den Tweel, W. J. J., Kamphuis, J., and Meijer, E. M. (1991). Production of 1-kestose with intact mycelium of Aspergillus phoenicis containing sucrose-lF-fructosyltransferase. App]. Microbiol. Biotechnol. 35, 216-221. Barthomeuf, C., and Pourrat, H. (1995). Production of high-content fructooligosaccharides by an enzymatic system from Penicillium rugulosum. Biotechno]. Lett. 17, 911-916.

338

S.G. PRAPULLA et al.

Barthomeuf, C, Grizard, D., and Teulade, J. C. (1997). Assay and structural determination of fructooligosaccharides synthesized by an enzymatic system from Penicillium rugu]osum. Biotechnol. Tech. 11,845-848. Bartlett, T. J., Rastall, R. A., Rees, N. H., Adlard, M. W., and Bucke, C. (1992). The application of aqueous two-phase systems to oligosaccharide synthesis by c~-mannosidase catalyzed glycosyl transfer reactions. J. Chem. Technol. Biotechnol. 55, 73-78. Beyts, P. K. (1989). "Sweetening Compositions." UK Patent Application GB2210545A, GB8723423 (19871006). Biton, J., GellL G., Michel, J. M., Paul, F. B., and Monsan, P. F. (1989). "Use of Fructosyl Oligosaccharides as Low-Calorie Sweeteners, and Foods, Dietetic Products and Beverages Containing Them." French Patent Application FR2629985A1, FR88-04944 (19880414). Bonnin, E., and Thibault, J. F. (1996). Galactooligosaccharide production by transfer reaction of an exogalactanase. Enzyme Microb. Technol. 19, 99-106. Bucke, C., and Rastall, R. A. (1990). Synthesizing sugars by enzymes in reverse. Chem. Britain 26, 675-678. Burvall, A., Asp, N. G., and Dalquist, A. (1979). Oligosaccharide formation during hydrolysis of lactose with Saccharomyces lactis lactose (Maxilact), l: Quantitative aspects. Food Chem. 4, 243-250. Cantacnzene, D., Attal, S., and Bay, S. (1991). Synthesis of c~- and [3-glycopyranosylserine derivatives by transglycosylation. Biomed. Biochim. Acta 50, 231-236. Chen, W., and Liu, C. (1996). Production of [3-fructofuranosidase by Aspergillus japonicus. Enzyme Microb. Teehnol. 18, 153-160. Crittenden, R. G., and Playne, M. J. (1996). Production, properties and applications of food-grade oligosaccharides. Trends Food Sci. Technol. 7, 353-360. Croft-Hill, A. (1898). Reversible zymohydrolysis. J. Chem. Soc. 63,634-658. Daniel, D. G., William, A. S., Blessing, L. L., Busta, F. F., and Brady, L. J. (1996). Probiotics, cecal microflora and aberrant crypts in the rat colon. J. Nutr. 126, 1362-1371. de Soete, J., and Freund, D. (1993). "Reduced-Calorie Chocolate Confectionery Compositions." PCT International Patent Application WO93/02566A1. Dickson, R. C., Dickson, L. R., and Markin, J. S. (1979). Purification and properties of an inducible [5-galactosidase isolated from the yeast Kluyveromyces lactis. J. Bacteriol. 12, 51-61. Duan, K. J., Chen, J. S., and Sheu, D. C. (1994). Kinetic studies and mathematical model for enzymatic production of fructooligosaccharides from sucrose. Enzyme Microb. Technol. 16, 334-339. Dumortier, V., Brassart, C., and Bouquelet, S. (1994). Purification and properties of a [5-D-galactosidase from B. bifidum exhibiting a transgalactosylation reaction. Biotechnol. Appl. Biochem. 19, 341-354. Fujita, K., Hara, K., Hashimoto, H., and Kitahata, S. (1990). Purification and some properties of ~-fructofuranosidase I from Arthrobacter sp. K1. Agric. Biol. Chem. 54(4), 913-919. Fujita, K., Kuwahara, N., Tanimoto, T., Koizumi, K., lizuka, M., Minamiura, N., Furuichi, K., and Kitahata, S. (1994). Chemical structures of heterooligosaccharides produced by Arthrobacter sp. K71 [5-fructofuranosidase. Biosci. Biotechnol. Biochem. 58,239243.

MICROBIAL PRODUCTION OF OLIGOSACCHARIDES

339

Furuichi, K., and Kitahata, S. (1994). Chemical structures of heterooligosaccharides produced by Arthrobacter sp. K-1 [~-fructofuranosidase. Biosci. Biotechnol. Biochem. 58, 239-243. Garegg, P. J. (1990). Artificial antigens: Prospects for medicines. Chem. Britain 26, 669674. Gross, D. (1962). Nonreducing oligosaccharide synthesis. In "Methods in Carbohydrate Chemistry" (R. L. Whistler and M. L. Wolfrom, eds.), Vol. 10, pp. 360-364. Academic Press, New York. Gyllang, H. E., Ekblom, B. T., Wesshagen, V. V., and Ericson, K. Y. (1986). "Beverage Product." European Patent Application EPO181988A1. Hayashi, S., Ito, K., Nonoguchi, M., Takasaki, Y., and Imada, K. (1991). Immobilization of fructosyl-transferring enzyme from Aureobasidium sp. on Shirasu porous glass. J. Ferment. Bioeng. 72, 68-70. Hayashibara, K. (1974). "Sweetening Compositions." Japanese Patent Application JP4940950 (4940950). Hedbys, L., Larsson, P. O., and Mosbach, K. (1984). Synthesis of the disaccharide 6-O-[~D-galactophyranosyl-2-acetoamido-2-deoxy-D-galactose using immobilized galactosidase. Biochem. Biophys. Res. Commun. 123, 8-15. Hedbys, L., Johansson, E., Mosbach, K., Larsson, P. O., Gunnarsson, A., Svensson, S., and Lonn, H. (1989). Synthesis of Gal~I-3-Glu-NAC and Gal~I-3-Glc-NAC-~-SEt by an enzymatic method comprising the sequential use of [5-galactosidases from bovine testes and Escherichia coil Glycoconjugate J. 6, 161-168. Hidaka, H., Hirayama, M., and Sumi, N. (1988). A fructooligosaccharide-producing enzyme from Aspergillus niger ATCC 20611. Agric. Biol. Chore. 52, 1181-1187. Ichikawa, Y., Look, G. C., and Wong, C. H. (1992). Enzyme-catalyzed oligosaccharide synthesis. Anal. Biochem. 202, 215-238. Jeon, I. J., and Mantha, V. R. (1985). High-performance liquid chromatography analysis of oligosaccharides formed during ~-galactosidase action an lactose. J. Dairy Sci, 68, 581-588. Johansson, E., Hedbys, L., Larsson, P. O., Mosbach, K., Gunnarsson, A., and Svensson, S. (1989). Studies of the reversed c~-mannosidase reaction at high concentrations of mannose. Enzyme Microb. Technol. 11,347-352. Jnng, K. H., Li, J. Y., Yoo, S. J., Lee, J. H., and Yoo, M. Y. (1987). Production of fructosyltransferase from Aureobasidium pullalans. Biotechnol. Lett. 9, 703-708. Kan, T., Kobayashi, Y., Sonoike, Y., Terashima, T., and Mntai, M. (1989). "Liquid Food." European Patent Application EP0307523A1. Kery, V., Kucar, S., Matulovd, M., and Haplov~, J. (1991). Immobilization of [3-D-galactosidase from Escherichia coli on cellulose beads and its use for the synthesis of disaccharide derivatives. Carbohydr. Res. 209, 83-87. Kono, T., Tokunaga, T., Yamaguchi, G., Kitagawa, H., and Hiragai, T. (1988). "Method for Preventing Deterioration of Starch-Containing Foods." U.S. Patent Application US4748032. Korduner, H. M., Gyllang, H. E., and Ericson, K. Y. (1982). "Beverage Product." U.S. Patent Application US4312856. Larsson, P. O., Hedbys, L., Svensson, S., and Mosbach, K. (1987). Disaccharide synthesis with immobilized [~-galactosidase. In "Methods in Enzymology" (J. J. Langore, and H. von Vunakis, eds.), Vol. 136, pp. 230-233. Academic Press, New York.

340

S.G. PRAPULLA et al.

Lerouge, P., Roche, P., Faucher, C., Maillet, F., Truchet, G., Prome, J. C., and Denairie, J. (1990). Symbiotic host-specificity of Rhizobium meliloti is determined by a sulfated and acylated glucosamine oligosaccharide signal. Nature 344, 781-784. Lopez-Leiva, M. H., and Guzman, M, (1995). Formation of oligosaccharides during enzymatic hydrolysis of milk whey permeates. Process Biochem. 30, 757-762. McNeil, M., Darvil], A. G., Fry, S. C., and Albersheim, P. (1984). Structure and Function af the primary cell walls of plants. Annu. Rev. Biochem. 53, 625-663. Mentink, L., and Serpelloni, M. (1992). "Low-Calorie Chocolate." French Patent Application FR2676164A1, FR91-05511 (19910506). Mieth, B., Engst, W., Elsner, A., and Behrens, H. (1985). "Preparation of Food Emulsions." German Democratic Republic Patent Application DD228457 (DD228457). Mitsuo, N., Takeichi, H., and Satoh, T. (1984). Synthesis of [3-alkyl glucosides by enzymatic transglucosylation. Chem. Pharm. Bull. 32, 1183-1187. Mochizuki, K., and Sakurai, M. (1990). "Method for Producing Pressurized Gas-Entrapping Candy." U.S. Patent Application US4935189, JP87-325445. Mori, T., Fujita, S., and Okahata, Y. (1997). Transglycosylation in a two-phase aqueousorganic system with catalysis by a lipid coated ~-D-galactosidase. Carbohydr. Res. 298, 65-73. Mozaffar, A., Nakanishi, K., and Matsuno, R. (1988). Production of oligosaccharides by glutaraldehyde-treated and -immobilized [3-galactosidases from Bacillus circulans. Biotechno]. Lett. 10, 805-808. Nakagawa, R., and Kitagawa, N. (1996). "Method for Cold Storage of Vegetables." U.S. Patent Application WO96/14754A1. Nakakuki, T., ed. (1993). Oligosaccharides: Production, properties and applications. In "Japanese Technology Review," Vol. 3. Gordon and Breach, Singapore. Nakao, M., Harada, M., Kodama, Y., Nakayama, T., Shibano, Y., and Amachi, T. (1993). Purification and characterization of a thermostable [~-galactosidase with high transgalactosylation activity from Saccharopolyspora rectivirigula. AppI. Microbiol. Biotechnol. 40, 657-663. Newsholme, E. A., Blomstrand, E., Ekblom, B., and Ericson, K. (1988). "Use of Amino Acids for the Preparation of a Beverage." PCT International Patent Application WO 88/09132A1. Nilsson, K. G. I. (1986). A comparison of the enzyme-catalyzed formation of peptides and oligosaccharides in various hydroorganic solutions using the non-equilibrium approach. In "Biocatalysts in Organic Media" (C. Laane, J. Tramper, and M. D. Lilly, eds.). Elsevier, Amsterdam. Nilsson, K. G. I. (1987). A simple strategy for changing the regioselectivity of glycosidasecatalysed formation of disaccharides. Carbohydr. Bes. 167, 95-103. Novak, M., Stanek, J., and Hollerova, I. (1996). Formation of oligofructosides during gluconate production by A. niger. Biotechnol. Lett. 18, 211-212. Ogawa, K., Uejima, O., Nakakuki, T., Usui, T., and Kainuma, K. (1990). Enzymatic synthesis of p-nitrophenyl-a-maltoheptaoside by transglucosylation of maltoheptoseforming amylase. Agric. Biol. Chem. 54, 581-586. Ooi, Y., Hashimoto, T., Mitsuo, N., and Satoh, T. (1985). Enzymic formation of [3-alkyl glycosides by ~-galactosidase from A. oryzae and its application to the synthesis of chemically unstable cardiac glycosides. Chem. Pharm. Bull. 33, 1808-1814. Palcic, M. M., Srivastava, O. P., and Hindsgaul, O. (1987). Transfer of D-galactosyl groups to 6-O-substituted 2-acetoamido-2-deoxy-D-glucoseresidues by use of bovine D-galactosyltransferase. Carbohydr. Res. 159, 315-340.

MICROBIAL PRODUCTION OF OLIGOSACCHARIDES

341

Patel, V., Saunders, G, and Bucke, C. (1994). Production of fructooligosaccharides by Fusarium oxysporum. Biotechno]. Lett. 16, 1139-1144. Pivarnik, L. F., and Rand, A. G. (1991). Transgalactosidase activity of Kluyveromyces lactis derived ~-galactosidase. Abstr. Annu. Mtg. Inst. Food TechnoL Abstr. #579. Pivarnik, L. F., Senecal, A. G., and Rand, A. G. (1995). Hydrolytic and transgalactosylic activities of commercial [~-galactosidase (lactase) in food processing. In "Advances in Food and Nutrition Research" (J. E. Kinsella and S. L. Taylor, eds.), Vol. 38, pp. 1-90. Academic Press, New York. Plug, J. P. (1995). "Oligosaccharides and Their Use As Food Ingredients." European Patent Application EPO664299A1, EP93202933 (19931020). Prakash, S. B., Suyama, K., Itoh, T., and Adachi, S. (1989). Structure elucidation of major galactooligosaccharides formed by growing culture of Trichoderma harzianum. J. Agric. Food Chem. 37, 334-337. Rademacher, T. W, Parekh, R. B., and Dwek, R. A. (1988). Glycobiology. Annu. Rev. Biochem. 57, 785-838. Rastall, R. A., and Bucke, C. (1992). Enzymatic synthesis of oligosaccharides. In "Biotechnology and Genetic Engineering Reviews" (M. P. Tombs, ed.), Vol. 10, pp. 253-281. Intercept, London. Rastall, R. A., Rees, N. H., Wait, M. R., Adlard, M. W., and Bucke, C. (1992). c~-mannosidase-catalysed synthesis of novel manno-, lyxo- and heteromannooligosaccharides: A comparison of kinetically and thermodynamically mediated approaches. Enzyme Microb. Technol. 14, 53-57. Rosevear, P. R., Nufiez, H. A., and Barker, R. (1980). Synthesis and solution conformation of the type 2 blood group oligosaccharide C~-L-Fuc(1-2)-[3-D-Gal(1-4)-~-D-GlcNAC. Biochemistry 19,489-492. Samain, E., Drouillard, S., Heyraud, A., Driguez, H., and Geremia, R. A. (1997). Gram scale synthesis of recombinant chitooligosaccharides in Escherichia coil Carbohy&: Res. 302, 15-42. Sauerbrei, B., and Thiem, J. (1992). Galactosylation and glycosylation by use of [~-galactosidase. Tetrahedron Lett. 33,201-204. Schnaar, R. L. (1991). Glycosphingolipids in cell surface recognition. Glycobiology 1, 477-485. Shiffin, S., and Hunn, G. (1969). Effect of alcohols on the enzymatic activity and the subunit association of [3-galactosidase. Arch. Biochem. Biophys. 130, 530-535. Shinoyama, H., Kamiyama, Y., and Yasui, T. (1988). Enzymatic synthesis of alkyl [~-xylosides from xylobiose by application of the transxylosyl reaction of Aspergillus niger [3-xylosidases. Agric. Biol. Chem. 53, 2197-2202. Sonoike, Y., Kobayashi, Y., Kato, H., and Kan, T. (1992). "Method for Producing Bread Containing Oligosaccharide." U.S. Patent Application US5118521, JP88-125874 (19890519). Speights, R. M. (1990). "Method for Inhibiting the Growth of Salmonella." U.S. Patent Application US4902674, US107115 (19871013). Speights, R. M., Perna, P. J., and Downing, S. L. (1991). "Method for Inhibiting the Growth of Salmonella." U.S. Patent Application US4987124, US433506 (19891108). Stevenson, D. E., Stanley, R. A., and Fumeause, R. H. (1993). Optimization of alkyl [~-D-galactopyranoside synthesis from lactose using commercially available [5-galactosidase. Biotechnol. Bioeng. 42, 657-666.

342

s . G . PRAPULLA et a].

Stevenson, D. E., Stanley, R. A., and, Furneaux, R. H. (1996). Oligosaccharide and alkyl ~-galactopyranoside synthesis from lactose with Caldocellum saccharolyticum glycosidase. Enzyme Microb. Technol. 18, 544-549. Stokes, K. M., and Wilson, I. B. (1972). A common intermediate in the hydrolysis of galactosidase by [3-galactosidase from Escherichia coil Biochemistry 11, 1061-1064. Sumi, N., Mochizuki, K., Moriyama, Y., and Sakurai, M. (1991). "Method of Producing Pressurized Gas-Entrapping Candy." U.S. Patent Application US5023098, US432821 (19891107). Suoranta, K. (1991). "A Novel High-Density Yeast Preparation: A Method for Producing the Same, and the Use Thereof." PCT International Patent Application WO 91/12315A1. Takaichi, A., Azuma, Y., and Otsuka, I. (1991). "Liquid Nutrient Composition." PCT International Patent Application WO91/03948A1, JP89-239524 (19890914). Takaku, H. (1988). In "Handbook of Amylase and Related Enzymes of the Amylase Research Society of Japan," pp. 215-217. Pergamon, Oxford. Takeda, H., Sato, K., Kinoshita, S., and Sasaki, H. (1994). Production of 1-kestose by Scopu]ariopsis brevicaulis. J. Ferment. Bioeng. 77, 386-389. Toba, T., and Adachi, S. (1978). Hydrolysis of lactose by microbial [~-galactosidases: Formation of oligosaccharides with special reference to 2-O-[3-D-galactopyranosyl-Dglucose. J. Dairy Sci. 61,333-338. Toba, T., Tomita, Y., Itoh, T., and Adachi, S. (1980). [3-galactosidases of lactic acid bacteria: Characterization by oligosaccharides formed during hydrolysis of lactose. J. Dairy Sci. 64, 185-192. Toba, T., Watanabe, A., and Adachi, S. (1982). Allolactose and 6-O-~-galactopyranosyl-ogalactose in commercial yogurt. J. Dairy Sci. 65, 702-706. Tomomatsu, H. (1994). Health effects of oligosaccharides. Food Technol. 61, 61-65. Toone, E. J., Simon, E. S., Bednarski, M. D., and Whitesides, G. A. (1989). Enzyme-catalyzed synthesis of carbohydrates. Tetrahedron 45, 5365-5422. Trincone, A., Nicolaus, B., Lama, L., Morzillo, P., De Rosa, M., and Gambacorta, A. (1991). Enzyme-catalyzed synthesis of alkyl-[3-D-glycosides with crude homogenate of Sulfolobus solfataricus. BiotechnoL Lett. 13, 235-240. Usami, S., Ishii, T., Kirimura, K., Vehara, K., and Chen, J. (1991). Production of [~-fructofuranosidase showing fructose-transferring activity by Penicil]ium firequentans (P. glabrum). J. Ferment. Bioeng. 72, 303-305. Usui, T., and Murata, T. (1988). Enzymatic synthesis of p-nitrophenyl c~-maltopentoside in an aqueous methanol solvent system by maltoteraose forming amylase: A substrate for human amylase in serum. J. Biochem. 103, 969-972. Viratelle, O., Tenu, J. P., Gamier, J., and Yon, J. (1969). A preliminary study of the nucleophilic competition in [3-galactosidase-catalyzed reactions. Biochem. Biophys. Res. Commun. 37, 1036-1041. Walon, R. (1971). "Confectionery Syrup." Canadian Patent 868304. White, F. (1990). "Partially Esterified Polysaccharide (pep) Fat Substitutes." U.S. Patent US4959466, US147806 (19880125). White, J. F. (1991). "Partially Esterified Polysaccharide (pep) Fat Substitutes." European Patent Application EP0463245A1, EP90306871 (19900622). Wierzbicki, L. E., and Kosikowski, F. V. (1973). Formation of oligosaccharides during [3-galactosidase action on lactose. J. Dairy Sci. 56, 1400-1404. Wiezemann, S., Denzell, K., Schilling, G., and Gross, G. G. (1988). Enzymatic synthesis of 1-O-pheny] carboxyl-[3-D-glucose esters. Bioorg. Chem. 16, 29-32.

MICROBIAL PRODUCTION OF OLIGOSACCHARIDES

343

Wong, C. H., Haynie, S. L., and Whitesides, G. M. (1982). Enzyme-catalyzed synthesis of N-acetyllactosamine with in situ regeneration of uridine 5'-diphosphate glucose and uridine 5'-diphosphate galactose. J. Org. Chem. 47, 5416-5418. Yatka, R. J., Reed, M. A., and Broderick, K. B. (1994). "Wax-Free Chewing Gum Including Special Oligosaccharide Binders." U.S. Patent Application US5342631, US996914 (19921229). Yatka, R. J., Richey, L. C., and Meyers, M. A. (1995). "Chewing Gum Products Using Oligofructose." U.S. Patent Application US5431929, US244845 (19940728). Yun, J. W., and Song, S. K. (1993). The production of high content fructooligosaccharides from sucrose by the mixed-enzyme system of fructosyltransferase and glucose oxidose. Biotechnol. Lett. 15, 573-576. Yun, J. W., and Song, S. K. (1996). Continuous production of fructooligosaccharides using fructosyltransferase immobilized on ion-exchange resin. Biotechnol. Bioprocess. Eng. 1, 18-21. Yun, J. W., Jung, K. H., Jeon, Y. J., and Lee, J. H. (1992). Continuous production of fructooligosaccharides by immobilized cells of Aureobasidium pullulans. J. Microbiol. Biotechnol. 2, 98-101. Yun, J. W., Lee, M. G., and Song, S. K. (1994a). Batch production of high-content fructooligosaccharides from sucrose by the mixed-enzyme system of [~-fructofuranosidase and glucose oxidase. J. Ferment. Bioeng. 77, 159-163. Yun, J. W., Lee, M. G., and Song, S. K. (1994b). Production of panose from maltose by intact cells of Aureobasidium pullulans. Biotechno]. Lett. 16, 359-362. Yun, J. W., Kang, S. C., and Song, S. K. (1995). Continuous production of fructooligosaccharides from sucrose by immobilized fructosyltransferase. Biotechnol. Tech. 9,805808. Yun, J. W., Kim, D., Moon, H., Song, C., and Song, S. (1997). Simultaneous formation of fructosyl- and glucosyltransferase in Aureobasidium pullulans. J. Microbiol. Biotechnol. 7, 204-208.

Index

A ABT-773, 80-81

Acetobacterium woodii, fatty acid hydration by, 207 Acetyl esterase, antibiotic resistance and, 147 Acriflavine, 132 Actinomadura madurae, prodigiosins, 15 Actinomadura pel]etieri, prodigiosins, 15 Actinomycetes, oligosaccharide biosynthesis by, 319 Actinomycin, biosynthesis of, 141 Actinorhodin, 18 Acyl carrier protein (ACP), 121 Adriamycin, 125 AIB factor. See Anti-isobntyrate factor Air pollution, by organochlorine pesticides, 274 Alcohols, biocatalytic dynamic resolution of, 66-70 Algae, degradation of lindane by, 272, 287-288 Allysine ethylene acetal, enzymatic synthesis of, 46-51, 71 Allzyme phytase 115,183 Alteromonas ruber, prodigiosins, 14 Amidase, enzymatic synthesis of chiral drug intermediates, 54-55 Amycolotopsis orientalis, bioactive products from, 120 Anabaena, degradation of lindane by, 288 Animal feed, phytase and, 159-160, 175-177 Anthracyclines biosynthesis of, 125 hybrid bioactive products of, 137 Anti-Alzheimer's drug, enzymatic synthesis of chiral drug intermediates, 66, 72 Antibiotic properties, of prodigiosin, 22-23 Antibiotic resistance, 144-148

Anticholesterol drugs, enzymatic synthesis of chiral drug intermediates, 57-60, 72 Antihypertensive drugs enzymatic synthesis of allysine ethylene acetal, 46-51, 71 BMS-199541-01, 34-43 L-6-hydroxynorleucine, 43-46 Anti-isobutyrate (AIB) factor, 143-144 Antimalarial activity, of prodigiosin, 23 Antiviral agents, enzymatic synthesis of chiral drug intermediates, 61-63 Apase6, 181 aphA gene, 181 appA gene, 170 Aquaculture, phytase use in, 186-187 Arabidopsis thaliana phytase, 171-172 Arthrobacter globiformis, degradation of lindane by, 288 Arthrobacter simplex, enzymatic synthesis of chiral drug intermediates, 67 Arthrobacter spp., oligosaccharide biosynthesis by, 314 Arxula adeninivorans phytase, 173,176 Aspergillus fumigatus phytase, 176, 180, 182

Aspergillus japonicus, fructooligosaccharide biosynthesis by, 310

Aspergillus niger enzymatic synthesis of chiral drug intermediates, 63-65 oligosaccharide biosynthesis by, 310-311,314, 316-317 Aspergillus niger phytase, 160-170, 175-182, 184 Aspergillus oryzae, oligosaccharide biosynthesis by, 317-319 Aspergillus phoenicis, 1-kestose biosynthesis by, 309-310 Aspergillus sydowi, oligosaccharide biosynthesis by, 336 Aspergillus terreus phytase, 176

345

346

Aureobasidium pullulans,

INDEX

~-lactams antibiotic resistance, 148 biosynthesis of, 118-120 Bialaphos, 131 Bifidobacterium bifidum, oligosaccharide biosynthesis by, 319 Bifidobacteriam spp., in human gastrointestinal tract, 304,333-334 Bioactive products from Streptomyces, 113-114 biosynthesis of, 115-131 B derivatives of, 133-137 molecular genetics of, 131-133 [~3-adrenergic receptors, 52 regulation of production, 137-144 [~3-receptor agonists, enzymatic synthesis resistance to, 144-147 of, 52-57, 71 Biocatalysis. See Microbial/enzymatic Bacillus amyloliquefaciens phytase, 172 synthesis Bacillus brevis, bioactive products Biocatalytic dynamic resolution, of from, 115 racemic diol, 66-68, 72 Bacillus licheniformis, bioactive products Biodegradation from, 115 of lindane Bacillus megaterium, enzymatic by algae and cyanobacteria, 272,277, synthesis of chiral drug 287-288 intermediates, 45 by bacteria, 272, 277,279-285 Bacillus pumilus, fatty acid hydration by, by fungi, 285-287 207 Bio-Feed phytase, 183 Bacillus subtilis phytase, 172, 176 Biorefining, 222 Bacitracins, biosynthesis of, 115-116 Bizio, Bartolomeo, 6, 8 Bacitracin synthetase, 117 Bleomycin, 115 Bacteria. See also Bioactive products; Blood, cultural history of, 2 Microbial/enzymatic synthesis; "Blue diaper syndrome," 19 Secondary metabolites BMS-186318, biotransformation of, 61 degradation of lindane by, 272,279-285 BMS-186716, microbial/enzymatic genetically engineered microorganisms synthesis of, 34 (GEMs), 288-289 BMS-188494, biotransformation of, 58, 72 oligosaccharide biosynthesis by, BMS-199541-01, biotransformation of, 309-315,318-319 34-43, 71 prodigiosins, 2-26 BMS-201391-01, biotransformation of, xylose-fermenting, 226-227, 233 35-43, 71 Bacterium prodigiosum, prodigiosins, 7 BMS-202665, biotransformation of, 41, 43 Bacteroides polypragmatus, xylose BMS-203528-01, biotransformation of, fermentation with, 233 39-41 Barley, trihydroxy fatty acids from, 213 BMS-204556, biotransformation of, Basidiomycetes, biodegradation of 39-40, 42 organochlorine compounds by, BMS-264406, biotransformation of, 41, 43 272,286 Bread, red spoilage of, 2-5 4-Benzyloxy-3-methanesulfonylamino-2'-br Butyl-meta-cycloheptylprodiginine, 14 omoacetophenone, microbial 7-Bntyrolactone, 142 reduction of, 52-54 oligosaccharide biosynthesis by, 305,313-314,320 Autoregulators, bioactive products from Streptomyces, 143 Avermectins, biosynthesis of, 128 Azithromycin, 80-81

INDEX C

Caldariomyces fumigo, enzymatic synthesis of chiral drug intermediates, 61-63 Caldocellus saccharolyticum, oligosaccharide biosynthesis by, 318-319 Candicidin, biosynthesis of, 127 Candida antarctica, enzymatic synthesis of chiral drug intermediates, 69-72 Candida boidinii enzymatic synthesis of chiral drug intermediates, 46-51, 61-63, 66-68, 72 xylose fermentation by, 228 Candida parapsilosis, enzymatic synthesis of chiral drug intermediates, 66-67 Candida pulcherrima, red pigmentation of, 10 Candida shehatae, xylose fermentation by, 228-229, 233-239 Candida tropicalis, xylose fermentation by, 228 Candida utilis, enzymatic synthesis of chiral drug intermediates, 43 Carbazomycinal, biosynthesis of, 143 ccr gene, 98-99 Cephalosporins biosynthesis of, 136, 139, 141 degradation of, 147 structure of, 118-119 Cephalosporium acremonium, bioactive products from, 140 Ceriporiopsis subvermispora, 287 Chiral drug intermediates microbial/enzymatic synthesis of anticholesterol drugs, 57-60, 71-72 antihypertensives, 34-51 antiviral agents, 61-63 [~3-receptor agonists, 52-57, 71 biocatalytic dynamic resolution, 66-68 deoxyspergualin, 60-61, 72 racemic diol stereoinversion, 66-68, 72 racemic epoxide hydrolysis, 63-66 racemic secondary alcohol resolution, 68-72

347

Chitooligosaccharides, purification of, 322-323 Chlamydomonas reinhardtii, lindane degradation by, 287 Chloramphenicel, biosynthesis of, 129-130 Chlorella vulgaris, lindane degradation by, 287 1-Chloroalkane helihydrolase (DhaA), 279 Chlortetracycline biosynthesis of, 122,124, 141 structure of, 123 Chromatography, oligosaccharide assay by, 327-331 Chromobacterium prodigiosum, 22 Clarithromycin, 80-81 Clavibacter spp., ALA2, 212,215-216 Clavulanic acid, biosynthesis of, 118, 120 Clostridium acetobutilicum, xylose fermentation by, 227 Clostridium butyricum, lindane degradation by, 279-280 Clostridium pasteurianum, lindane degradation by, 280 Clostridium rectum, lindane degradation by, 278 Clostridium saccharolyticum, xylose fermentation with, 233 Clostridium spp., lindane degradation by, 279-280 Clostridium thermocellum, xylose fermentation by, 227 Combinatorial biosynthesis, 80 of bioactive products, 137 Saccharopolyspora erythraea, 100-103 Coriolus bulleri, 287 Coriolus hirsutus, 286 Coriolus versicolor, 287 Corynebacterium spp., oleic acid hydration by, 203 Coupling sugar, 301 Cyanobacteria, degradation of lindane by, 287-288 Cyath us bulleri, 286 Cycl gene, 230 Cyclodextrins, 302 Cyclononylprodiginine, 15 Cycloprodigiosin, 14

348

INDEX

Daunorubicin, 125 6-DEB synthase (6-DEBS), 82 de Col, Pietro, 7 DehH1, 279 6-Deoxyerythronolide B (6-DEB), 82 Deoxyspergualin, enzymatic synthesis of chiral drug intermediates, 60-61, 72 Desosamine deoxysugar biosynthesis genes, from Streptomyces venezuelae, 80 DhaA, 279 DhIA, 279 DHOD. See 10,12-Dihydroxy-8(E)-octadecenoic acid Dihydrogranatirhodin, 137 Dihydromederrhodin, 137 7,10-Dihydroxy-8(E)-octadecenoic acid (DOD), biosynthesis, 208-210, 212 lO,12-Dihydroxy-8(E/-octadecenoic acid (DHOD), biosynthesis of, 211-212 Dihydroxy unsaturated fatty acids, biotransformation of, 208-212 Dinoflagellates, phytic acid in, 174-175 Diols, racemic, biocatalytic dynamic resolution of, 66-68, 72 DNA, introduction into Saccharopolyspora erythraea, 87-89 DOD. See 7,10-Dihydroxy-8(E)-octadecenoic acid Doxorubicin, 125 Dyes, red bacteria, 7

Electroporation, DNA introduction into Saccharopolyspora erythraea, 87-89 Environment, phosphorus levels in, 160, 174-175 Enzymatic/microbial synthesis. See Microbial/enzymatic synthesis Epoxide, racemic, biocatalyzed stereoselective hydrolysis of, 63-66 ermE gene, 89-90, 97 ertX gene, 75 eryCI gene, 89-90

eryG gene, 90, 98 Erythromycin, 79 biosynthesis of, 82-84 from Saccharopolyspora erythreae, 85-100, 126, 133 chemical structure of, 81 Erythromycin biosynthetic gene cluster, 80-82, 86, 89-96 Escherichia cob HAP phytase from, 170-171 xylose fermentation by, 227,233 Esterase, enzymatic synthesis of chiral drug intermediates, 55-57 Ethanol biosynthesis of bacteria, 226-227, 233 critical parameters, 232,234-241 ethanol tolerance, 241-255 heat shock proteins, 253-254 Pichia spp., 229-232 thermal tolerance, 241-255 trehalose, 252-253 yeasts and fungi, 226,228-229, 233 uses of, 222 Ethyl-meta-cyclononylprodiginine, 14

FAB mass spectrometry, of oligosaccharides, 332 Factor A, 142 Factor B, 142-143 Factor C, 143 Factor I, 143 Farnesyl pyrophosphate (FPP), 57-58 Fatty acids, unsaturated, biotransformation of, 201-216 Flavobacterium fuscum, enzymatic synthesis of chiral drug intermediates, 43 Ftavobacterium spp. oleic acid hydration by, 205 sp. DS5,210 Foods oligosaccharides in, 303-304,333-337 red spoilage of, 2-6 Formate dehydrogenase, enzymatic synthesis of chiral drug intermediates, 47, 48-49

INDEX FPP. See Farnesyl pyrophosphate Frnctooligosaccharides, 301,309 biosynthesis of, 309-315 properties of, 303 purification of, 322 uses of, 335-336 Fructosyltransferase, 305 Fucosyltransferase, 307-308 Fungi degradation of lindane by, 285-287 galactooligosaccharide biosynthesis by, 316-317 phytase from, 160-170, 183 xylose-fermenting, 226,228-229, 233 Fusarium oxysporum fructooligosaccharide biosynthesis by, 311 xylose fermentation by, 228, 233

G Galactooligosaccharides, 300 biosynthesis of, 315-319 purification of, 322 Galactosides, 315-316, 320 Galactosyltransferase, 307 Gas-liquid chromatography (GLC), oligosaccharide assay by, 329-330 gdh gene, 75 Genetically engineered microorganisms (GEMs), 288-289 Genistein, 135 Gentiooligosaccharides, 302 Geotrichum candidum, enzymatic synthesis of chiral drug intermediates, 58-59, 66, 72 Germicidin, biosynthesis of, 143 GLC. See Gas-liquid chromatography Glucomannan, 224 Glucooligosaccharides, uses of, 335 Glucose dehydrogenase, enzymatic synthesis of chiral drug intermediates, 54 Glutamate dehydrogenase, enzymatic synthesis of chiral drug intermediates, 44-46 Glutathione-dependent reductive dehalogenase (LinD), 278 Glycopeptides, biosynthesis of, 120

349

Glycosidases lipid-coated, 326-327 oligosaccharide biosynthesis and, 306-307,320 Glycosyl sucrose, 301 Glycosyltransferases, 307-308 Gramicidins, biosynthesis of, 115-116 Gramicidin S synthetase, 117

H Haloacetate dehalogenase (DehH1), 279 Haloalkane dehalogenase (DhlA), 279 Hansen ula polymorpha enzymatic synthesis of chiral drug intermediates, 66, 67 phytase production from, 183 HAPs. See Histidine acid phosphatases HCH. See Hexachlorocyclohexane Heat-shock proteins, 253-254 Hexachlorocyclohexane (HCH) biodegradation of by algae and cyanobacteria, 272,277, 287-288 by bacteria, 272,277, 279-285 by fungi, 285-287 environmental contamination by, 272-274 producers of, 271 structure of, 270-271 toxicological effects of, 274-277 High-performance liquid chromatography (HPLC), oligosaccharide assay by, 327-329 Histidine acid phosphatases (HAPs), 161-172 HIV protease inhibitors, enzymatic synthesis of chiral drug intermediates, 61, 72 HMR-3647, 80-81 HOD. See lO-Hydroxy-8-octadecenoic acid HPLC. See High-performance liquid chromatography Hxt gene, 240 Hydrolytic dehalogenases, 278-279 10-Hydroxy fatty acids, biosynthesis of, 207-208

350 Hydroxy fatty acids dihydroxy, 208-212 monohydroxy, 202-207 strain ALA2 system, 212,215-216 trihydroxy, 212-215 uses of, 201 L-6-Hydroxynorleucine, enzymatic synthesis of, 43-46, 71 10-Hydroxy-8-octadecenoic acid (HOD), biosynthesis of, 210 10-Hydroxystearic acid, biosynthesis of, 202-206

Immunosuppressive activity, of prodigiosin, 24 Isomaltooligosaccharides, 302,320 Isomaltulose oligosaccharides, 301 Ivermectin, 128-129

&de gene, 75 Kelp, "red spot disease," 10 1-Kestose, 303,309-310 Ketolides, 80-81 Klehsiella aerogenes phytase, 172-173 Klebsiella oxytoca, xylose fermentation by, 227, 233 Klebsiella terrigena phytase, 172-173 Kluyveromyces fragilis, oligosaccharide biosynthesis by, 319 Kluyveromyces lactis, oligosaccharide biosynthesis by, 319 Konbu, "red spot disease," 10

[~-Lactams antibiotic resistance, 148 biosynthesis of, 118-120 Lactobacillus casei, xylose fermentation by, 227

INDEX

Lactobacillus pentoaceticus, xylose fermentation by, 227 Lactobacillus pentosus, xylose fermentation by, 227 Lactohacillus plantarum, xylose fermentation by, 227 Lactobacillus spp., oligosaccharide biosynthesis by, 318 Lactobacillus xylosus, xylose fermentation by, 227 Lactose, galactoside synthesis using, 320-321 Lactosucrose, 301 Lactulose, 300, 305 Lesquerella hydroxy fatty acids, 201-202 Lignocellulose, 222-224 biorefining of, 222-223 aeration, 238-240 ethanol tolerance, 241-255 heat-shock proteins, 253-254 nutrient uptake, 240-241 p/I, 236-238 pretreatment, 224 simultaneous saccharification and fermentation (SSF), 225-226 temperature, 234-236 thermal tolerance, 241-255 trehalose, 252-253 xylose-fermenting microbes, 226-233 LinB, 279,285 /rinD, 278 Lindane, 270-271 biodegradation of by algae and cyanobacteria, 272,277, 287-288 by bacteria, 272,277, 279-285 by fungi, 285-287 environmental contamination by, 272-274 toxicological effects of, 274-277 Linoleic acid, biocatalyzed hydration of, 207 Linolenic acid, biocatalyzed hydration of, 207 Lipase, enzymatic synthesis of chiral drug intermediates, 60-61, 69, 72 L-Lysine-e-aminotransferase, biotransformation of, 35-37, 39-43

INDEX M

Macrocyclic prodiginines, 15 Macrolides, 80-81, 125-126 biosynthesis of, 125-127 derivatives of, 136 hybrid macrolides, 137 structure of, 126 Maltooligosaccharides, 302 Maxilact, 319 Mederrhodin, 137 MelA enzyme, 96 melA gene, 96-97 Metacycloprodigiosin, 12, 14, 23-24 6-Methoxycarbazomycinal, biosynthesis of, 143 Methyl-(4-methoxyphenyl)-propanedioic acid ethyl diester, racemic, asymmetric hydrolysis of, 55-57 c~-Methyl phenylalanine amide, racemic, enzymatic resolution of, 54-55 6-Methyl salicylic acid (6MS), biosynthesis of, 121-122 6-Methyl salicylic acid synthase, 122 Microalgae, biodegradation of organochlorine compounds by, 272 Microbial/enzymatic synthesis chiral drug intermediates, 33-34, 71-72 anticholesterol drugs, 57-60, 72 antihypertensives, 34-51 antiviral agents, 61-64 [~3-receptor agonists, 52-57, 71 biocatalytic dynamic resolution, 66-68 deoxyspergualin, 60-61, 72 racemic diol stereoinversion, 66-68, 72 racemic epoxide hydrolysis, 63-66 racemic secondary alcohol resolution, 68-72 of ethanol bacteria, 226, 227, 233 critical parameters, 232, 234-241 ethanol tolerance, 241-255 heat shock proteins, 253-254 Pichia spp., 229-232 thermal tolerance, 241-255 trehalose, 252-253 yeasts and fungi, 226, 228-229, 233

351

of oligosaccharides, 299-300, 308-309, 337 enzymatic mechanism, 305-308 fructooligosaccharides, 309-315,322 galactooligosaccharides, 315-319,322 immobilized systems, 324-326 isomaltooligosaccharides, 320 lipid-coated glycosidases, 326-327 novel oligosaccharides, 320-321 organic solvents for, 323-324 purification of, 322-323 recombinants for, 327 two-phase systems, 324 of unsaturated fatty acids, 201-202 dihydroxy, 208-212 monohydroxy, 202-207 strain ALA2 system, 212,215-216 trihydroxy, 212-215 Microbial products, 113-114 Micrococcus prodigiosus, 6, 22 Micromonospora purpurea, secondary metabolite production by, 132 Micromonospora rosaria, secondary metabolite production by, 132 Minocycline, derivatives of, 136 Monas prodigiosa, 6 Monensin, 129 Monohydroxy fatty acids, biotransformation of, 202-207 Mortierella ramanniana, enzymatic synthesis of chiral drug intermediates, 61-63 Mucor sanguineus, 7 Mucor spp., xylose fermentation by, 228 Myceliophthora thermophila, phytase genes, 176 Mycobacterium fortuitum, oleic acid hydration by, 204 Mycobacterium neoaurum, enzymatic synthesis of chiral drug intermediates, 54-55, 71 Myo-inositol phosphate, production of, 187-188

N

Natuphos, 183 Neokestose, 303,305 Neosugar, 305

352

INDEX

Neurospora crassa, xylose fermentation by, 233 NMR analysis, of oligosaccharides, 331-332 Nocardia aurantia, oleic acid hydration by, 204 Nocardia cholesterolicum, fatty acid hydration by, 203,204,207 Nocardia madurae, prodigiosins, 15 Nocardia pelletieri, prodigiosins, 15 Nocardia salmonicolor, enzymatic synthesis of chiral drug intermediates, 64 Norprodigiosin, 12-13 Nostoe ellipsossorum, degradation of lindane by, 288 Nystatin, structure of, 127

in chewing gums, 336 classification of, 300-303 foods, use in, 303-304,333-337 functions of, 332 health benefits of, 304-305 properties of, 303-304 Omapatrilat, microbial/enzymatic synthesis of chiral intermediates, 34, 46 Organochlorine compounds biodegradation of by algae and cyanobacteria, 272,277, 287-288 by bacteria, 272,277, 279-285 by fungi, 285-287 degradation of, 272 toxicological effects of, 274-277 Oxytetracyclines (OTC), biosynthesis of, 122,140, 141

O Occupational health concerns, phytase production, 188-189 Oleandomycin, 126 Oleandrose, 128 Oleic acid biotransformation to 10-HSA, 202-206 hydroxylation of, 207 Oligosaccharides, 299-300,337 applications of, 332-336 assay of, 327 FAB mass spectrometry, 332 GLC, 329-330 HPLC, 327-328 NMR, 331-332 paper chromatography, 329 TLC, 330-331 biosynthesis of, 308-309 enzymatic mechanisms, 305-308 fructooligosaccharides, 309-315,322 galactooligosaccharides, 315-319, 322 immobilized systems, 324-326 isomaltooligosaccharides, 320 lipid-coated glycosidases, 326-327 novel oligosaccharides, 320-321 organic solvents for, 323-324 purification of, 322-323 recombinants for, 327 two-phase systems, 324

Pachysolen tannophilus, xylose fermentation by, 228,229-231, 233-234, 237, 239 Palatinose oligosaccharides, 301 Paper chromatography, oligosaccharide assay by, 329 Paramecium phytase, 174 Penicillin amidases, antibiotic resistance and, 147 Penicillins biosynthesis of, 118-120, 136 degradation of, 147 Penici]lium citrinum, enzymatic synthesis of chiral drug intermediates, 57

Penici]lium frequentans, fructooligosaccharide biosynthesis by, 311 Penicillium patulum, bioactive products from, 121

Penicillius rugulosum, fruct ooligosaccharide biosynthesis by, 311-313 Peniophora lycii phytase, 165, 183 2-Pentanol, biocatalytic dynamic resolution of, 69, 72

INDEX Peptide antibiotics, biosynthesis of, 115-120 Peroxidase, semisynthesis of, 188 Pesticides biodegradation of by algae and cyanobacteria, 272, 277,287-288 by bacteria, 277, 279-285 by fungi, 285-287 Phanerochaete chrysosporium, environmental pollutant degradation by, 286-287 Phanerochaete sordida, environmental pollutant degradation by, 286 Phenylalanine dehydrogenase, enzymatic synthesis of chiral drug intermediates, 46-51 Phenyl galactoside, 320 Phlebia brevispora, 287 Phlebia radiata, 287 pho genes, 171 Phosphorus levels, environmental, 160, 174-175 PhyA, 161-166, 176 phyA gene, 164 PhyB, 166-170, 182 phyB gene, 166 PhyC, 172 Phytase, 158, 189-192 animal feed and, 159-160, 175-177, 183 Bacillus phytase, 172 bioengineering, 175-183,185 E. coli HAP phytase, 170-171 environmental phosphorus levels and, 160, 174-175 enzyme production, 183 enzyme specificity of, 179 functions of, 159 fungal phytase, 160, 183 heat tolerance of, 175-177 histidine acid phosphatases, 161-172 Klebsiella phytase, 172-173 occupational health issues, 188-189 pH optimum of, 177-179 PhyA, 161-166, 176 PhyB, 166-170 PhyC, 172 plants, 171-174, 184-185 substrate specificity of, 179

353

synergistic effect of, 180-183 temperature optimum of, 177-179 in transgenic animals, 185-186 uses of, 159-160, 186-188 yeast phytase, 171,173,183-184 Phytic acid, 159, 175 PHYT I gene, 171 PHYT II gene, 171 Pichia pastor& enzymatic synthesis of chiral drug intermediates, 47-51 phytase production from, 183 Pichia stipitis, xylose fermentation by, 228, 229-231,233-239,249-251 Picromycin, 126 Pithia methanolica, enzymatic synthesis of chiral drug intermediates, 66-68, 72 PKS. See Polyketide synthase Plants phytase from, 171-174, 184-185 phytic acid in, 159 Pleurotus sajor-caju, 286 Pmal gene, 248 Pma2 gene, 248 Polenta, red spoilage of, 5-6 Polyenes, 125-127 Polyethers, bioactivity of, 129 Polyketides, biosynthesis of, 120-129, 132 Polyketide synthase (PKS), 80, 98, 132-133 Polymorphisms, of Saccharopolyspora erythraea, 86-87 Pravastatin, enzymatic synthesis of chiral drug intermediates, 57-60 Prodiginine, 11-12 Prodigiosan, 15 Prodigiosene, 11-12 Prodigiosin, 8, 10-26 antibiotic properties, 22-23 antimalarial activity, 23 biosynthesis of, 15-18 ecological functions, 21-22 immunosuppressive activity, 24 pharmacological activity, 22-24 Serratia spp. producing, 8 structure, 10-15 Prodigiosin 25-C, 13 Protease inhibitors, enzymatic synthesis of chiral drug intermediates, 61, 72

354

INDEX

PsAdhl gene, 230 PsAdh2 gene, 230 Pseudoalteromonas bacteriolytica, "red spot disease," 10

Pseudoalteromonas denitrificans, prodigiosins, 15

Pseudomonas enzymatic synthesis of chiral drug intermediates, 60-61 strain 42A2, 210-211

Pseudomonas magnesiorubra, prodigiosins, 13

Pseudomonas paucimobilis, lindane

Rhodococcus rhodochrous, fatty acid hydration by, 203,207

Rhodotorula glutinis, enzymatic synthesis of chiral drug intermediates, 63-66 Rhodotorula graminis, enzymatic synthesis of chiral drug intermediates, 43 Rice, trihydroxy fatty acids from, 213-214 Ricinoleic acid, 201 Rifamycin, biosynthesis of, 143 Ronozyme P, 183 Rugomonas rubra, prodigiosins, 13

degradation by, 280, 285

Pseudomonas putida, lindane degradation by, 280,285

Pseudomonas spp. lindane degradation by, 280-281 strain PR3,208, 211 PsStul gene, 240 Pulcherrimin, 10 Pyruvate decarboxylase, 230

Saccharomyces cerevisiae bioactive products from, 143 expression of Pichia genes in, 231-232 fermentation by, 226,233,240, 242-248 yeast phytase, 171 Saccharomyces lactis, oligosaccharide biosynthesis by, 319

Saccharopolyspora erythraea R

Racemic diol, biocatalytic dynamic resolution of, 66-68, 72 Racemic epoxide, biocatalyzed stereoselective hydrolysis of, 63-66 Racemic methyl-(4-methoxyphenyl)-propanedioic acid ethyl diester, asymmetric hydrolysis of, 55-57 Racemic c~-methyl phenylalanine amide, enzymatic resolution of, 54-55 Racemic secondary alcohols, biocatalytic dynamic resolution of, 66-68 rap gene, 18 Recombinant oligosaccharides, in-vivo synthesis of, 327 Red bacteria history of, 2-9 naming, 6-10 pigments and paintings, 7-8 "Red diaper syndrome," 19 "Red spot disease," 10 Relomycin, structure of, 126 Response-regulator protein, 141

combinatorial biosynthesis with, 80, 100-103, 137 experimental properties of, 84-85 molecular genetics of, 85-105,133 erythromycin biosynthetic gene cluster, 80, 81-82, 86, 89-96 industrial strain improvement, 97-100 introduction of DNA into, 87-89 mapping, 85-86 new tools for, 96-97 polymorphisms, 86-87 optimization of macrolide production, 98-99 polyketide synthase from, 80

Saccharopolyspora rectivirgula, oligosaccharide biosynthesis by, 319 Salinomycin, 129

Schizosaccharomyces pombe fermentation by, 226,233,252 yeast phytase, 171 Schwanniomyces castellii phytase, 173,176 Schwanniomyces occidenta]is phytase, 173

INDEX

35 5

Scopulariopsis brevicaulis,

Sphingomonas paucimobilis

fructooligosaccharide biosynthesis by, 311 Sebacic acid, 201 Secondary alcohols, biocatalytic dynamic resolution of, 68-70 Secondary metabolites, 114 biosynthesis of bialaphos, 131 chloramphenicol, 129-130 peptide antibiotics, 115-120 polyketides, 120-129, 132 streptomycin, 130-131 derivatives of, 135-137 hybrid bioactive products, 137 isolation of, 133-134 microbial producers of, 131-134 regulation of production, 137-144 resistance to, 144-148 screening for, 135 Sensor-transmitter protein, 141 Serrati, Serafino, 6

enzymatic synthesis of chiral drug intermediates antihypertensives, 35-36, 43 [~3-receptor agonist, 52-56, 71 lindane degradation by, 278,281,283 Squalene synthase, enzymatic synthesis of chiral drug intermediates, 58-60 SSF. See Simultaneous saccharification and fermentation

Serratia ficaria, 8 Serratia liquefaciens, 8, 21 Serratia marcescens, 25 antibiotic resistance, 20-21 cell-surface hydrophobicity, 22 prodigiosin, 8, 10, 15-16, 18, 21-22 Serratia marcescens infection, 19-21

Serratia marinorubra, 8 Serratia odorifera, 8 Serratia plymuthica, 8, 22 Serratia rubidaea, 8 Serratia spp., 8-10 color variations in, 22 prodigiosin, 8, 10-26 Simultaneous saccharification and fermentation (SSF), 225-226 Soil amendments, phytase as, 187 Sonication-dependent electroporation, DNA introduction into Saccharopolyspora erythraea, 87-88 Soybean oligosaccharides, 302-303,305 Spergualin, enzymatic synthesis of chiral drug intermediates, 60-61, 72 Sphingobacterium spp., oleic acid hydration by, 205

Streptococcus thermophilus, oligosaccharide biosynthesis by, 318

Streptomyces aureofaciens, bioactive products from, 122,139, 141,144, 146-147 Streptomyces avermectilis, bioactive products from, 143 Streptomyces carbophilus, enzymatic synthesis of chiral drug intermediates, 57 Streptomyces cinnamonensis, bioactive products from, 143 Streptomyces clavuligerus, bioactive products from, 118, 140 Streptomyces coelicolor, prodigiosins, 16, 18 Streptomyces collinus, ccr gene, 98-99 Streptomyces fradiae, bioactive products from, 127,144 Streptomyces galilaeus, combinatorial biosynthesis with, 137 Streptomyces griseus, bioactive products from, 128,142-143

Streptomyces hiroshimensis, prodigiosins, 14

Streptomyces hydroscopicus, bioactive products from, 131

Streptomyces lividans bioengineering, 145-146 enzymatic synthesis of chiral drug intermediates, 39 Streptomyces longisporus ruber, prodigiosins, 13-14 Streptomyces nodosus, enzymatic synthesis of chiral drug intermediates, 61-63 Streptomyces noursei, enzymatic synthesis of chiral drug intermediates, 35, 37-39, 71

356

INDEX

Streptomyces parvulus, bioactive products from, 141

Streptomyces purpurascens, combinatorial biosynthesis with, 137

Streptomyces rimosus bioactive products from, 122,140-141 bioengineering, 146 Streptomyces rubriretuculi var. pimprina, prodigiosins, 14 Streptomyces spp. bioactive products biosynthesis of, 115-131 derivatives of, 133-137 molecular genetics of, 131-133 regulation of production, 137-144 resistance to, 144-147 bioactive products from, 125 growth phases of, 137-138

Streptomyees venezuelae bioactive products from, 143 chloramphenicol biosynthesis from, 129 desosamine deoxysugar biosynthesis genes, 80

Streptomyces viridochromogenes, bioactive products from, 131, 143 Streptomycin, biosynthesis of, 130-131 Streptorubin A, 14

StreptoverticiHium baldaccii, prodigiosins, 14

Streptoverticillium spp., bioactive products from, 143

Sulfiblobus solfataricus, 321 Syntropic pigmentation, 16

Talaromyces lanuginosus phytase, 176 Talaromyces thermophi]is, phytase genes, 176 Tetrachlorocyclohexadiene dehalogenase (LinB), 279 Tetracyclines, 125 biosynthesis of, 122-125, 140-141 structure of, 122-123

Thermoactinomyces intermedius, enzymatic synthesis of chiral drug intermediates, 46-48, 51, 71

Thin-layer chromatography (TLC), oligosaccharide assay by, 330-331 THOA. See 9,12,13-Trihydroxy10(E}-octadecenoic acid Tigemonam, chiral intermediate of, 51 Tk! genes, 232 Tkt genes, 232 TLC. See Thin-layer chromatography TOD. See 7,10,12-Trihydroxy-8(E}octadecenoic acid Trametes hirsutus, 286-287 Transcriptional terminator cartridge, in Saccharopolysporo erythraea, 91-92 Transgenic animals, phytase, 185-186 Transgenic plants, phytase, 185 Trehalose, 252-253 Trichoderma hardanum, oligosaccharide biosynthesis by, 317 Trigonopsis variabilis, enzymatic synthesis of chiral drug intermediates, 44, 46 8,9,13-Trihydroxy docosanoic acid, biosynthesis of, 212 7,10,12-Trihydroxy-8(E)-octadecenoic acid (TOD), biosynthesis of, 211 9,12,13-Trihydroxy-10(E)-octadecenoic acid (THOA), biosynthesis of, 212-215 12,13,17-Trihydroxy-(Z)-octadecenoic acid, biosynthesis of, 212 Trihydroxy unsaturated fatty acids, biotransformation of, 212-215 Tylosin biosynthesis of, 127 structure of, 126

U Undecylprodiginine, 12, 16, 18, 23-24 Undecylprodigiosin, 12-13 Unsaturated fatty acids, 201-202 dihydroxy, 208-212 monohydroxy, 202-207 strain ALA2 system, 212,215-216 trihydroxy, 212-215 uses of, 201

INDEX

357 X

V Vancomycin biosynthesis of, 120 structure of, 121 Vasopeptidase inhibitor enzymatic synthesis of, 71 allysine ethylene acetal, 46-51, 71 BMS-199541-01, 34-43 L-6-hydroxynorleucine, 43-46 vhb gene, introduction into Saccharopolyspora erythraea, 99-100 Vibrio gazogenes, prodigiosins, 15 Vibrio psychroerythreus, prodigiosins, 13 Vitreoscilla hemoglobin gene, introduction into Saccharopolyspora erythraea, 99-100

W

Water pollution, by organochlorine pesticides, 273-274 White-rot fungi, biodegradation of organochlorine compounds by, 272, 286-287

Xylitol dehydrogenase, 230 Xylooligosaccharides, 303 Xylose, 222-223 bioconversion of bacteria, 226-227,233 critical parameters, 232,234-241 ethanol tolerance, 241-255 heat shock proteins, 253-254 Pichia spp., 229-232 thermal tolerance, 241-255 trehalose, 252-253 yeasts and fungi, 226, 228-229, 233 lignocellulose, 223-226 Y Yeast phytase, 171,173 Yeasts oligosaccharide biosynthesis by, 319 xylose-fermenting, 226, 228-229,233 Z

Zymomonas mobilis, xylose fermentation by, 227,233

CONTENTS OF PREVIOUS VOLUMES

Volume 37

Microbial Degradation of the Nitroaromatic Compounds Frank K. Higson An Evaluation of Bacterial Standards and Disinfection Practices Used for the Assessment and Treatment of Stormwater Marie L. O'Shea and Richard Field Haloperoxidases: Their Properties and Their Use in Organic Synthesis M. C. R. Franssen and H. C. van der Plas Medicinal Benefits of the Mushroom Ganoderma S. C. Jong and J. M. Birmingham Microbial Degradation of Biphenyl and Its Derivatives Frank K. Higson The Sensitivities of Biocatalysts to Hydrodynamic Shear Stress Ale~ Prokop and Rakesh K. Bajpai Biopotentialities of the Basidiomacromycetes Somasundaram Rajarathnam, Mysore Nanjarajurs Shashirekha, and Zakia Bano INDEX Volume 38

Selected Methods for the Detection and Assessment of Ecological Effects Resulting from the Release of Genetically Engineered Microorganisms to the Terrestrial Environment G. Stotzky, M. W. Broder, J. D. Doyle, and R. A. Jones

Biochemical Engineering Aspects of Solid-State Fermentation M. V Ramana Murthy, N. G. Karanth, and K. S. M. S. Raghava Rao The New Antibody Technologies Erik P. Lillehoj and Vedpal S. Malik Anoxygenic Phototrophic Bacteria: Physiology and Advances in Hydrogen Production Technology K. Sasikala, Ch. V. Ramana, R Rahuveer Rao, and K. L. Kovacs INDEX

Volume 39

Asepsis in Bioreactors M. C. Sharma and A. K. Gartu Lipids of n-Alkane-UtilizingMicroorganisms and Their Application Potential Samir S. Radwan and Naser A. Sorkhoh Microbial Pentose Utilization Prashant Mishra and Ajay Singh Medicinal and Therapeutic Value of the Shiitake Mushroom S. C. Jong and J. M. Birmingham Yeast Lipid Biotechnology Z. Jacob Pectin, Pectinase, and Protopectinase: Production, Properties, and Applications Takuo Sakai, Tatsuji Sakamoto, Johan Hallaert, and Erick J. Vandamme 359

360

CONTENTS OF PREVIOUS VOLUMES

Physicochemical and Biological Treatments for Enzymatic/Microbial Conversion of Lignocellulosic Biomass Purnendu Ghosh and Ajay Singh

Volume 41

INDEX

Improving Productivity of Heterologous Proteins in Recombinant Saccharomyces cerevisiae Fermentations Amit Vasavada

Volume 40

Microbial Cellulases: Protein Architecture, Molecular Properties, and Biosynthesis Ajay Singh and K3"yoshi Hayashi Factors Inhibiting and Stimulating Bacterial Growth in Milk: An Historical Perspective D. K. O'Toole Challenges in Commercial Biotechnology. Part I. Product, Process, and Market Discovery Ale~ Prokop Challenges in Commercial Biotechnology. Part II. Product, Process, and Market Development Ale~ Prokop Effects of Genetically Engineered Microorganisms on Microbial Populations and Processes in Natural Habitats Jack D. Doyle, Guenther Stotzky, Gwendolyn McClung, and Charles W. Hendricks

Microbial Oxidation of Unsaturated Fatty Acids Ching T. Hou

Manipulations of Catabolic Genes for the Degradation and Detoxification of Xenobiotics Rup Lal, Sukanya Lal, P. S. Dhanaraj, and D. M. Saxena Aqueous Two-Phase Extraction for Downstream Processing of Enzymes/Proteins K. S. M. S. Raghava Rao, N. K. Rastogi, M. K. Gowthaman, and N. G. Karanth Biotechnological Potentials of Anoxygenic Phototrophic Bacteria. Part I. Production of Single Cell Protein, Vitamins, Ubiquinones, Hormones, and Enzymes and Use in Waste Treatment Ch. Sasikala and Ch. V. Ramana Biotechnological Potentials of Anoxygenic Phototrophic Bacteria. Part II. Biopolyesters, Biopesticide, Biofuel, and Biofertilizer Ch. Sasikala and Ch. V. Ramana INDEX

Detection, Isolation, and Stability of Megaplasmic-Encoded Chloroaromatic Herbicide-Degrading Genes within Pseudomonas Species Douglas J. Cork and Amjad Khalil INDEX

Volume 42

The Insecticidal Proteins of Bacillus th uringiensis P. Ananda Kumar, R. P. Sharma, and V S. Malik

CONTENTS OF PREVIOUS VOLUMES Microbiological Production of Lactic Acid

361

Volume 44

John H. Litchfield

Biodegradable Polyesters

Biologically Active Fungal Metabolites Cedric Pearce

Ch. Sasikala

The Utility of Strains of Morphological Group II Bacillus Samuel Singer

Old and New Synthetic Capacities of Baker's Yeast P. D'Arrigo, G. Pedrocchi-Fantoni, and S. Servi

Phytase Rudy J. Wodzinski and A. H. J. Ullah INDEX

Investigation of the Carbon- and Sulfur-Oxidizing Capabilities of Microorganisms by Active-Site Modeling Herbert L. Holland

Volume 43

Production of Acetic Acid by Clostridium thermoaceticum Munir Cheryan, Sarad Parekh, Minish Shah, and Kusuma Witjitra

Contact Lenses, Disinfectants, and Acanthamoeba Keratitis Donald G. Ahearn and Manal M. Gabriel

Marine Microorganisms as a Source of New Natural Products V. S. Bernan, M. Greenstein, and W. M. Maiese

Stereoselective Biotransformations in Synthesis of Some Pharmaceutical Intermediates Ramesh N. Patel

Microbial Xylanolytic Enzyme System: Properties and Applications Pratima Bajpai

Oleaginous Microorganisms: An Assessment of the Potential Jacek Leman INDEX

Microbial Synthesis of D-Ribose: Metabolic Deregulation and Fermentation Process R de Wulf and E. J. Vandamme

Production and Application of Tannin Acyl Hydrolase: State of the Art B K. Lekha and B. K. Lonsane

Ethanol Production from Agricultural Biomass Substrates Rodney /. Botbast and Bada] C. Saha

Thermal Processing of Foods, A Retrospective, Part I: Uncertainties in Thermal Processing and Statistical Analysis M. N. Ramesh, S. G. Prapulla, M. A. Kumar, and M. Mahadevaiah

Thermal Processing of Foods, A Retrospective, Part II: On-Line Methods for Ensuring Commercial Sterility M. N. Ramesh, M. A. Kumar, S. G. Prapulla, and M. Mahadevaiah INDEX

362

CONTENTS OF PREVIOUS VOLUMES

Volume 45

One Gene to Whole Pathway: The Role of Norsolorinic Acid in Aflatoxin Research J. W. Bennett, P.-K. Chang, and D. Bhatnagar Formation of Flavor Compounds in Cheese P. F Fox and J. M. Wallace

Breathing Manganese and Iron: Solid-State Respiration Kenneth H. Nealson and Brenda Little Enzymatic Deinking Pratima Bajpai Microbial Production of Docosahexaenoic Acid (DHA, C22:6) Ajay Singh and Owen P. Ward

The Role of Microorganisms in Soy Sauce Production Desmond K. O'Toole

INDEX

Gene Transfer Among Bacteria in Natural Environments Xiaoming }Tin and G. Stotzky

Volume 46

Cumulative Subject Index