ADVANCES IN FOOD RESEARCH VOLUME 24
Contributors to This Volume
Sally Hudson Arnold W . Duane Brown Reiner Hamm Klau...
31 downloads
1136 Views
19MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
ADVANCES IN FOOD RESEARCH VOLUME 24
Contributors to This Volume
Sally Hudson Arnold W . Duane Brown Reiner Hamm Klaus Hofmann Walter M. Urbain Jonathan W. White, Jr. Robert L. Wickremasinghe
ADVANCES IN FOOD RESEARCH VOLUME 24
Edited by C. 0. CHICHESTER The Nutrition Foundation, Inc. New York. New York and University of Rhode Island Kingston. Rho& Island
E. M. MRAK
G . F. STEWART
University of California Davis, California
University of Calif(omia Davis, California
Editorial Board S. LEPKOVSKY EDWARD SELTZER W . M. URBAIN J . R. VICKERY
E. C. BATE-SMITH J . HAWTHORN M. A. JOSLYN J . F. KEFFORD
1978
ACADEMIC PRESS
New York
San Francisco
A Subsidiary of Harcourt Brace Jovanovich. Publishers
London
C O P Y R I C l i T @ 1978, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. N O PART O F THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY F O R M OR BY ANY MEANS, ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, OR ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FHOM T H E PUBLISHER.
ACADEMIC PRESS, INC.
111 Fifth Avenue, New York, N e w York 10003
Uiiifcd Kiii,stlorir Edifioti prrblislred b y ACADEMIC PRESS, INC. ( L O N D O N ) LTD. 24/28 Oval Road, London NW1 7DX
LIBRARY O F CONGRESS CATALOG CARD
ISBN
NUMBER: 48-7808
0-12-016424-8
PRINTkD IN 1 H E UNITED STATES 01: AMERICA
CONTENTS Contributors to Volume 24 . . . . . . . . . . . . . . . . . . . . . . . . .
vii
Sulfhydryl and Disulfide Groups in Meats Klaus Hofmann and Reiner H a m
I. 11 .
111. I \. . V. VI . VIl . VrII . IX .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods for the Determination of SH and SS Groups . . ................... SH Groups in Muscle Proteins and Their Role in the Fu n of Muscle . . . . . . . . SH and SS Content of Meats and Meat Fractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intluence of Processing on the SH and SS Groups of Meat . . . . . . . . . . . . . . . . . . . . Influence of the SH Groups on the Shelf Life of Meat and Meat Products . . . . . . . . Toxicological Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . ........................................
2 3 30 43 58 84 84 86 88 88
Histamine (7) Toxicity from Fish Products Sally Hudson Arnold and W . Duane Brown
I. 11. I11. IV . V.
Nature of the Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanisms of Formation of Histamine in Fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detection and Determination of Levels of Histamine in Fish . . . . . . . . . . . . . . . . . . . Relationship o f Spoilage to Histarnine Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unresolved Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . ................... References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
114 121 130 135 139 147
Food Irradiation Walter M . Urbain
I. I1 I11 IV . V. VI . VII
Introduction-Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......... ....... Radiation and Radiation Sources . . . . . . . . . . . . . . . . General Effects of Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications of Food Irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Economics of Food Irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Wholesomeness of Irradiated Foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ The Future of Food Irradiation . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
155
163 168 174 205 209 213 216
V
vi
CONTENTS
lea Robert L. Wickremasinghe I. U. UI . IV . V. V. VI. VI.
vu .
VIII . IX . X.
......_......... Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Composition of Tea .. . . . . . . . . . .... Changes during the Processing of Tea . . . . .. .. .......... . . Organoleptic Properties .. . ....... . .. .. . .. ....... . .. .. ........ . .. . Storage of Tea . . . . . . . . .. ..... . . . . . . . . . . . . ................. Potential By-Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clinical Effects .. .. .. .. .. .. .. . . . . . . . Host Plant-Pest Relationships .................... Instant Tea . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . Additional Research Needs . . . . . . . . . . . . . . ........................ ........................ References . . . . . . . ............................. .. ... .... .. ... ......... References.
229 232 25 I 263 266 268 269 27 1 272 213 273
Honey Jonathan W. W. White, Jr. Introduction ........................... Production and Processing ....................... Market Forms of Honey. . . . . Analysis and Composition Compositio . . . . . . . . . . . . . . . Analysis Physical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . Storage of Honey . . . . .. .. . . . . . . . vn. Nutritive Value . ....................... ......................... VUI . Uses . . . . . . . . . . . . VUI IX , Standards, Specifications, an IX X. Research Needs . .................... X References.. . . . . . . . . . . . . .. . . . . . . .................... References
288 289 295 291 333 344 352 354 358 363 364
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
375
I. U. U. Ill. IV. V. V VI. VI.
I
CONTRIBUTORS TO VOLUME 24 Numbers in parentheses indicate the pages on which the authors' contributions begin.
SALLY HUDSON ARNOLD, Institute of Marine Resources, Department of Food Science and Technology, University of California, Davis, California 95616 (113) W. DUANE BROWN, Institute of Marine Resources, Department of Food Science and Technology, University of California, Davis, California 95616 (113) REINER HAMM, Bundesanstalt fur Fleischforschung, Kulmbach, Germany ( I ) KLAUS HOFMA", Bundesanstalt fur Fleischforschung, Kulmbach, Germany (1)
WALTER M. URBAIN ,* Michigan State University, East Lansing, Michigan 48824 (155) JONATHAN W. WHITE, JR., Eastern Regional Research Center, Philadelphia, Pennsylvania I91 18 (287) ROBERT L. WICKREMASINGHE, Tea Research Institute of Sri Lanka, Coombs, Talawakelle, Sri Lunka (229)
St.
*Present address: 10645 Welk Drive, Sun City, Arizona 85351.
vii
This Page Intentionally Left Blank
ADVANCFS IN FOOD RL-%ARCH
.
VOI .
24
SULFHYDRYL AND DlSULFlDE GROUPS IN MEATS* KLAUS HOFMANN AND REINER HAMM
I . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I1 . Methods for the Determination of SH and SS Group:, . . . . . . . . . . . . . . . . . . A . General Problems in the Determination of SH Groups in Soluble and Insoluble Proteins . . . . . . . . . . . . . . . . . . . . . .................... B . Methods for the Determination of SH Group Meats . . . . . . . . . . . . . . C . Determination of SS Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. SH Groups in Muscle Proteins and Their Role in the Function of A . Myofibrillar Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Proteins of the Sarcoplasniic Reticulum (SR) . . . . . . . . . . . . . . . . . . . . . . C . Proteins of the Sarcolemma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Proteins of Mitochondria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E . Proteins of the Sarcoplasmic Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV . SH and SS Content of Meats and Meat Fractions . . . . . . . . . . . . . . . . . . . . . . A . SH and SS Content of Muscles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Cysteine plus Cystine Content of Muscles .............. C . SH Content of Organs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Fdctim Influencing the SH Content of Raw Meat . . . . . . . . . . . . . . . . . . . V . Influence of Processing on the SH and SS Groups of Meat . . . A . Influence of Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Freezing and Frozen Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . Freeze-Drying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E . Ripening of Dry Sausages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . F . Smoking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G . Irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V1 . lnfluence of the SH Groups on the Shelf Life of Meat and Meat Products . . VII . Toxicological Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX . Research Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2 3 3 6 28 30 31 41 41 42 42 43 4.5 50 52 5.5 58 58 13 77 77 80 80 81 84 84 86 88 88
.
*Dedicated to Professor Dr . Alfons Schiiberl Hannover (Germany). a pioneer in the chemistry of organic sulfur compounds .
Copyright 0 1978 by Academic Press Inc . All righn of reproduction in any form rebewed. ISBN 0-12-016424-8
.
2
KLAUS H O F M A N N A N D REINER H A M M
I. INTRODUCTION Reviews on the occurrence, properties, and functional importance of SH groups in biological systems were recently presented in three comprehensive monographs (Jocelyn, 1972; Friedman, 1973; J. M. Tortschinski, 1974). However, the research on SH and SS groups in muscle tissue used for food has not yet been reviewed, although a considerable amount of work has been done in this field. SH groups are usually considered to be the most reactive functional groups in proteins (Wallenfels and Streffer, 1964); but under certain conditions the reactivity of these groups can be more or less inhibited. It is understandable, therefore, that SH groups in proteins have attracted the attention of many research workers and that the role of SH groups in proteins has been the subject of a large number of investigations. SH groups can easily be oxidized to SS groups, the SH/SS redox equilibrium 2 R-SH
+t
0 2
R-SS-R
+ H20
or 2 R-S- - 2e
* R-S-S-R
being of great biological importance. Consequently, in any discussion of the role of SH groups, a consideration of SS groups must also be included. It is the purpose of this review to discuss the methods for the determination of SH and SS groups in proteins and to assess the importance of these groups in meat quality and meat processing. Most of the work taken into consideration here is related to red meats (muscles from cattle, pigs, and sheep) and poultry; fish is only occasionally mentioned. Amino acids, cysteine and cystine, are the carriers of the SH and SS groups in proteins. Knowledge on the reactions of these amino acids is therefore indispensable for an understanding of the reactivity of SH and SS groups in biological systems. Most of the SH (and SS) groups in meat are located in the muscle proteins (see Table V). Because only a small proportion of these groups exists as low molecular SHES compounds, mainly glutathion, the research discussed in this review is primarily related to protein SH. Cysteine and cystine content is of great importance for the nutritive value of meat, as it is for most foods. Although cysteine and cystine do not belong to the essential amino acids, a deficiency of “total cystine” (sum of cysteine and cystine) in nutrition increases the requirement of one of these essential amino acids, methionine, which can be metabolized to cysteine. The methionine content of food proteins also limits their nutritive value. For this reason, a sufficient
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
3
supply of “total cysteine” helps to spare methionine in the intermediate metabolism. Any destruction of cysteine or cystine during the treatment of foods as indicated by the disappearance of SH plus SS groups represents a detrimental effect on nutritive value. It is evident, therefore, that studies of SH (and SS) changes during the storage and processing of meat are of particular interest for the nutritionist. It should be mentioned that reliable information on changes in the SH and SS content of proteins can be obtained by direct determination of these groups in the intact system. This approach is preferable to amino acid determination after hydrolysis because the problem of an exact determination of these amino acids after hydrolysis by acids or enzymes has not yet been satisfactorily solved. The importance of SH and SS groups in sensoric quality as well as for the processing of foods has been investigated for many years. This research has related mainly to milk and other dairy products (particularly cheese), cereals, doughs, and beer (Hofmann, 197 la). Corresponding work on meats started about 15 years ago. The assay of SH and SS groups is extremely difficult for a number of reasons. We will, therefore, discuss the various methods of SH and SS analysis available and examine the interpretation of the results of each in detail. Our lack of knowledge on the role of SH and SS groups in meat is mainly due to the difficulties in the determination of protein SH. It is our hope that the gaps and contradictions in our understanding of SH and SS groups which are presented in this review will initiate further research in this field.
II.
METHODS FOR THE DETERMINATION OF SH AND SS GROUPS
This chapter will present a review of methods which are used or which may be useful for the quantitative assay of SH and SS groups in animal tissues and muscle proteins. Histometrical methods for the demonstration of the location of SH groups using dyes are briefly discussed. A. GENERAL PROBLEMS IN THE DETERMINATION OF SH GROUPS IN SOLUBLE AND INSOLUBLE PROTEINS The complicated problems of SH assay in proteins have been thoroughly presented and critically discussed in several review articles (e.g., Cecil and McPhee, 1959; Benesch and Benesch, 1962; Leach, 1966; Hofmann and Hamm, 1974a; Ashworth, 1976). With nondissolved proteins, e.g., muscle homogenates, additional difficulties arise which should be examined in detail (Hamm and Hofmann, 1966a). It has been observed that, in certain proteins in the native state, the total
4
KLAUS HOFMANN AND REINER HAMM
number of sulfhydryl groups is not available to chemical reagents. Not all proteins, however, demonstrate this behavior. Moreover, the number of SH groups which react depends on several internal and external factors: the nature of the protein, the presence or absence of denaturing agents, temperature, pH value, the kind and concentration of SH reagent used, and the time of reaction. There are, therefore, several degrees of availability of protein SH groups. The difficulties in SH determination in proteins are due partly to the methods themselves and partly to differences in the reactivity of those groups. Many times the SH values determined are lower than the actual SH content; in some cases they are higher. For the confirmation of a certain result, several reagents should be applied for SH determination as has been recommended by Benesch and Benesch (1962). If the results of these various analyses differ from each other, additional investigation is necessary. In Table I the different factors leading to incorrect results are summarized. An incomplete reaction of protein SH groups can be due to three different reasons: (a) steric hindrance of SH groups by the specific structure of the protein, (b) interaction of SH with other functional groups (combined function), and (c) repulsion between hydrophilic reagents and hydrophobic groups (particularly TABLE I FACTORS LEADING TO WRONG RESULTS IN THE DETERMINATION OF SH GROUPS IN PROTEINS"
SH value too low Nonreactive (inaccessible, inavaiI able, masked, hidden) SH groupsb Incomplete reaction (slowly reacting groups; reaction time is too short. e.g., at direct titration)
SH groups are partially oxidized by oxygen (air) before or during the determination (SH -+ SS)
SH value too high Unspecific SH reagent (reaction or complex formation with other functional groups of protein) Interference by substances which consume SH reagent (e.g., ascorbic acid if an oxidizing reagent is used; or J-, Br- and S2- at titration with AgNO,) Cleavage of SS to SH groups by the SH reagent or by hydrolysis
After Hofmann and Hamm (1974a). The reasons for an unsufficient reactivity of SH groups need only a brief summary in the text because they were discussed in detail by Hofmann and Hamm (1974a). a
SULFHYDRYL AND DISULFIDE GROUPS IN MbATS
5
alkyl residues) located in the vicinity of SH groups (hydrophobic environment). Denaturing agents (urea, guanidine, dodecyl sulfate) or heating usually cause an elimination of these inhibiting influences. This might be due to an unfolding of the peptide chains of protein which makes the SH groups accessible to SH reagents. Because of the powerful reducing field required for such a reaction (Chibnall, 1943), the old supposition that SH groups are produced by the actual cleavage of the S-S linkage is no longer satisfactory. Chibnall(l943) suggested a hydrolytic cleavage of thiol ester linkages (R-CO-S-R’) during denaturation, but this type of linkage has not yet been demonstrated in proteins. In addition, the assumption that the formation of free SH groups during denaturation is due to an opening of thiazoline ring systems present in proteins (Linderstr@m-Langand Jacobsen, 1941) has not been confirmed by studies with model systems (Martin el al., 1959; Kolthoff and Shore 1964; Hofmann, I966a). According to another hypothesis, proteins contain isothiuronium residues which can react with amino groups to cause SH formation (Brush ef al., 1963). It is certain that the binding of some SH reagents (such as Ag+ ions or PCMB) to easily available SH groups can result in a denaturation of the protein which makes other hidden SH groups easily available (Bocchini et al., 1967; Jeckel and Pfleiderer, 1969). The denaturation of a protein usually makes the SH groups more easily oxidizable by oxygen. This can be prevented by the addition of EDTA (Sakai and Dan, 1959; Calcutt and Doxey, 1962). which sequesters catalyzing traces of heavy metals; 0.02 M EDTA is sufficient for the protection of SH groups (Sedlak and Lindsay, 1968) although 0.2 M EDTA was normally used in the preparation of homogenates (Tarnowski et al., 1965). According to their reactivity, protein SH groups are usually divided into three categories: fast reacting, slowly reacting, and nonreacting. However, this schematic classification is quite arbitrary. It does not take into consideration that the reactivity of an SH group is not an absolute property of this group. This reactivity depends essentially on the type of reagent used as well as other factors, e.g., pH or buffer systems. For this reason, very different SH values are often obtained after the reaction of the native protein with different reagents (Cecil and McPhee, 1959). The same is true for SH determination in meat (see Section IV). A great number of SH reagents are available (more than one hundred), which indicates the lack of a universally applicable method which is satisfying in every respect. The choice of method depends on the type of investigation in question because the determination of SH groups can have several different purposes: (1) Determination of total cystine content (cysteine plus half cystine) after complete reduction of SS groups in the course of the analysis of amino acids. Such an SH assay with the nonhydrolyzed protein prevents losses of cysteine and cystine from occurring during hydrolysis (Friedman, 1974) as already mentioned. (2) Investigation of the role of SH groups in the biological function of proteins.
6
KLAUS HOFMANN A N D REINER HAMM
(3) Study of the process of protein denaturation connected with a change (usually an increase) in the reactivity of SH groups. (4) Study of the reaction of protein SH groups with lead, mercury, other heavy metals, toxins, carcinogenic compounds, and other substances which are of relevance for environmental pollution, health hazards, and residue analysis. In this respect the protective effects of and the detoxification by SH groups in foods are of particular interest. Considering these different purposes of SH research, it is not always necessary to determine the total amount of SH groups in proteins; in some cases the determination of only the easily available SH groups of the native protein by a convenient reagent is intended (e.g., in cases 2 and 3 above). But other factors are also important for the correct selection of an SH assay method. Not all SH reagents are suitable for the SH determination in proteins, and only a small part of the reagents suggested for the SH determination in proteins can be used for the unsoluble proteins of muscle tissue (Hamm and Hofmann, 1966a). Reagents which can be used for studies on meat must fulfill the following requirements: (1) Because organic solvents cause the denaturation of muscle proteins, the SH reagent must be soluble in water. (2) The pH of the reagent solution should be approximately 7 in order to prevent denaturation by an acid or a base. (3) It must be possible to measure the excess of a reagent or a soluble reaction product after the reaction. (4) The SH reagent or the reaction product has to be stable because the reaction with undissolved proteins requires a relatively long amount of time. (5) The method has to be sensitive because of the low SH content of animal tissues. The determination of SS groups is usually based on the determination of SH groups before and after reduction of SS to SH. Here, both the complete reduction of all disulfide groups as well as the prevention of an oxidation of SH groups formed are necessary. Generally, a complete unfolding of the proteins by denaturing agents (e.g., urea, guanidinium hydrochloride) is important in order to provide a complete reduction of the S S groups.
B. METHODS FOR THE DETERMINATION OF SH GROUPS IN MEATS The number of reagents that are used or are suitable for the determination of SH groups in proteins, peptides, and amino acids is so great that it is not possible to discuss all of the possibilities in this review. Fortunately, this is also not necessary because of several excellent publications. A review on the function and analysis of SH groups in proteins was given by Heide (1955), who described in detail a modem assay based on amperometric titration with silver nitrate. Vol-
SULFHYDRYL AND DlSULFlDE GROUPS IN MEATS
7
uminous reports were presented by Cecil and McPhee (1959), Cecil (1963), and Leach (1966). Benesch and Benesch (1962) discussed the problems in the determination of SH groups, particularly in regard to the mercaptide formation with AgNO, and mercury compounds. The comprehensive publication of Lumper and Zahn ( I 965) includes important analytical aspects of the biochemistry of the disulfide exchange. Kakai: and Vejddek (1 974) summarized methods for the photometric determination of SH and SS groups, including a discussion of interfering factors. The application of organic mercury compounds for the chemical analysis of sulfur compounds (including SH and SS groups in proteins) was described by Wronski (1965). A further review article on the determination of SH and SS groups in proteins was published by Mesrob and Holesovsky (1967). The most recent information can be found in the monographs of Jocelyn (1972), Friedman (1973), and Ashworth (1976). This review first discusses the methods for the determination of SH and S S groups in meats and muscle proteins. Furthermore, those methods are considered which seem to be generally applicable for the determination of SH groups in insoluble material but which have been used thus far only for other biological systems such as flour or keratine. The determination of SH groups after complete reduction of SS groups allows the determination of the content of “total cystine” (cysteine plus cystine) without hydrolysis of the protein. The problems in cystine determination associated with the hydrolysis of proteins are described by Friedman (1974): “Direct assay of cystine by ion-exchange chromatography usually gives low values.. . . Consequently, many attempts have been made to change cystine residues to acid stable derivatives. . . . However, most of these attempts do not prevent complete destruction of the modified residues during acid hydrolysis, and the new derivatives are sometimes incompletely resolved on the chromatogram of an amino acid analyzer.” A further disadvantage in the application of protein hydrolysis is the fact that in most cases it is impossible to differentiate between SH groups and SS groups originally present. In this review the discussion of SH assay is therefore focused on methods applicable to the intact, nonhydrolized proteins. Such procedures present the additional advantage of allowing the study of the process of denaturation by following the changes in the reactivity of SH groups. The availability of the SH groups of a protein depends on its state in the protein structure and on the type of reagent used. Usually a denaturing agent [concentrated solutions of urea, guanidinium hydrochloride or, in our opinion the most suitable, diluted sodium dodecyl sulfate (also named lauryl sulfate or Duponol) solutions] has to be used in order to determine the total quantity of protein SH groups. With Ag+ ions, however, all SH groups can also be determined in the native meat protein ( H a m and Hofmann, 1965).
8
KLAUS H O F M A N N A N D REINER H A M M
The SH reagents can be divided into (1) oxidizing agents (2 RSH + RSSR), (2) mercaptide-forming agents (RSH .+ RSMe’, (RS)2Me”), (3) alkylating agents (RSH + RSR), and (4)other reagents. The reactions are usually measured either by spectrophotometry (including fluorometry) or by titration. The end-point of titration is mostly determined by means of electrometric methods (potentiometry, polarography , or amperometry). 1. Oxidizing Agents, Including Disiilfides
In most earlier research oxidizing agents known from oxidimetric methods were most often used, including iodine, ferricyanide, o-iodosobenzoate, porphyridin, 2,6-dichlorophenol-indophenol(Barron, 1951) and phosphoric tungstic acid (Folin and Morenzi, 1929).The desired reaction 2 RSH
- 2e + R-SS-R + 2H+
does not always occur in a stoichometric way because the sulfur of the SH group can be oxidized to a valence higher than that of disulfide. Furthermore, the influence of other reducing substances present in biological material (e.g., ascorbic acid or thiamine) can interfere in the reaction. More recently, the use of N-bromosuccinimide for the determination of cysteine and cystine has been described (Thibert et al., 1969;Bachhawat el af., 1973).However, this reagent is not highly specific. A different type of oxidation is the reaction of SH groups with disulfides encompassing an SH-SS exchange: R’SH
+ R-SS-R + R‘SS-R + R-SH
Numerous methods for the determination of SH groups in proteins are based on this type of reaction. Mirsky and Anson (1935)have treated protein SH groups with cystine and determined the obtained cysteine by reaction with phosphoric tungstic acid. Protein-bound SH groups in tissue slices can be visualized with 2,2’-dihydroxy-6,6’-dinaphtyl disulfide; a procedure introduced by Barnett and Seligman (1952). After the SH-SS exchange, the naphtol residue bound to the protein is coupled with tetrazotized diorthoanisidine to a red dye. It is also possible to extract the excess reagent from the tissue and to determine it after coupling to the red dye (Flesch et al., 1954). Ellman (1959) introduced the disulfide 5,5’-dithiobis (2-nitrobenzoic acid) (DTNB, Ellman’s reagent) which is now widely used for the determination of SH and SS groups in tissues (Ellman. 1959;Gabay et al., 1968;Khan el al., 1968;Sedlak and Lindsay, 1968;Dube, 1969; Caldwell and Lineweaver, 1969; Dzinleski et al., 1969; Yuan, 1970; Randall and Bratzler, 1970;Buttkus, 1971;Bowers, 1972;Boyne and Ellman, 1972;Dub6 et al., 1972;Habeeb, 1972;Hay et al., 1973;Miller and Spencer,
9
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
197.5; Usunov and Zolova, 1976). In basic solution, the disulfide DTNB (I) reacts quickly with dissociated SH groups in proteins as follows:
COO-
COO-
+
protein-SS &NO2-
[-S&NOz -
-
S
COO-
e p Q
-
(11)
H s -&Noz (111)
The resulting anion of the mesomeric 3-carboxy-4-nitro-thio-phenolate(II)* imparts the solution a lemon-yellow color which can be photometrically measured at 412 nm. The advantages of this reagent are its high specificity for SH groups and the high sensitivity of the resulting color reaction (eM = 13,600). As Klotz and Carver (1961) have pointed out, the stoichemistry of this method is not clear because Ellman's reagent (E-SS-E) can react also in another way: 2 Protein-SH
+ E-SS-E
---f
Protein-SS-Protein
+ 2E-SH
But this possibility involves no disadvantage for SH determination, because, in this case, 1 mole of the yellow thiophenolate is also formed from 1 mole proteinSH. The same is true if the mixed disulfide, primarily formed, reacts with a further protein SH group (Habeeb, 1972): Protein-SS-E
+ Protein-SH
-
Protein-SS-Protein
+ E-SH
The reaction of DTNB with protein SH groups is faster and more complete when the protein is allowed to be denatured by 8 M urea solution (Srere, 196.5). *Only by utilizing this nomenclature is it clear that the SH group is present in the ionized state (as thiophenolate). This is important because the nondissociated reaction product (111) is colorless (it fades by acidification). Consequently the Ellman's reagent should be termed bis(3-carboxy-4nitrophenyl) disulfide.
10
KLAUS HOFMANN AND REINER HAMM
Therefore, the determination of total protein sulfhydryl requires that the protein be denatured, preferably with sodium dodecylsulfate (Diez et al., 1964). The pH of the reaction mixture has a marked effect on the rate of color development. At pH 8, color development is complete in 5 minutes for all proteins tested (Beveridge ef d., 1974). Although the color reaction also occurs in solutions of weak acidity (above 4.7, Sedlak and Lindsay, 1968), pH values not lower than 8 are required in order to obtain complete reaction; Ellman (1959) suggested pH 8.2. Higher pH values accelerate the SH disulfide exchange (Lumper and Zahn, 1965), but they also induce a hydrolytic cleavage of the reagent causing a strong increase of the blank. According to our own experience, the original light-yellow color continuously deepens even in solutions of DTNB in phosphate or tris buffer of pH 7.5. We therefore recommend solutions of DTNB in ethanol, without addition of buffer, which are stable for a longer time (Hofmann and Bliichel, unpublished observations). Sedlak and Lindsay (1968) used methanol as a solvent for DTNB probably for the same reason. Calvin (1954) has shown that in basic solution two symmetrical disulfides can react with each other and form a mixed disulfide. In the case of the protein disulfide (PSSP) and Ellman’s reagent (ESSE), the reaction would be as follows: PSSP
+ ESSE + 2 PSSE
This reaction would not influence the results of the estimation of SH groups with ESSE because no thiolate anion (ES-) is formed. Furthermore, after Robyt ef al. (1971), mixed disulfides may also be formed by a series of reactions between ESSE and proteins which contain SH and SS groups. After the initial reaction PS-
+ ESSE + PSSE + ES
the released thiol ES- reacts with protein SS*, forming a second molecule PSSE and PS- which reacts again with ESSE. Altogether 3 moles PSSE and 1 mole ES- result. The derivatized protein PSSE can be separated, and the addition of a thiol (dithiothreitol) or adjustment of the pH to 10.5 releases a corresponding amount of ES- which can be measured at 412 nm. With this method Robyt et al. (1971) estimated the number of SH and SS groups in several proteins. Although ES- reacts immediately with protein SS groups, the number of ES- ions released is equal to the number of protein SH groups. Thus SS groups do not limit the specificity of the estimation of SH groups by Ellman’s reagent as it was postulated by Jocelyn (1972). Diez et al. (1964) concluded from their experiments that Ellman’s reagent in comparison with SH reagents such as N-ethylmaleimide, p-chloromercuri*In contrast, Weitzman (1975) reported, that thionitrobenzoate (ES-) would not react with disulfide groups in proteins. Therefore, the postulated mechanism of reaction seems to be in question.
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
11
benzoate, and iodine is the most useful reagent for routine determination of the total content of protein SH groups. A turbidity in the reaction solution can be eliminated either by filtration (Sedlak and Lindsay, 1968; Boyne and Ellman, 1972) or by centrifugation in the ultracentrifuge (Hofmann and Bliichel, unpublished observations). The precipitatin of proteins by TCA is not recommended (Dube, 1969). Dube found that the yellow color of the solution, obtained after reaction of GSH with DTNB, disappears after the addition of TCA, but comes back completely after adjusting the pH to 8; however, this is not the case in reaction mixtures of meat proteins and Ellman’s reagent because the formed carboxy-nitro-thiophenolateis partly adsorbed by the precipitate. This can be recognized by the strong yellow color of the washed precipitate after adjusting to pH 8 (Hofmann and Bluchel, unpublished observations). Since the reagent itself, as well as the tissue suspension after separation of the insoluble constituents show some absorbance, both blanks have to be measured separately and to be subtracted from the absorbance of the sample solution. Contrary to Dube (1969), we found that, in the determination of SH groups in meat with DTNB, myoglobin does not seriously disturb the measurement because the absorbance caused by myoglobin is extremely small, particularly with pork. Even with beef, the blank absorbance amounts to not more than 10% of the total absorbance (Hofmann and Bliichel, unpublished observations). It is often necessary to differentiate between protein-bound and nonbound SH groups in tissues. The latter groups are soluble in TCA and can be determined after the precipitation of proteins; the former can be calculated as the difference between the total SH content, measured after denaturation of the proteins by SDS, and the nonprotein SH content. Furthermore, the content of “available protein SH” can be determined by DTNB using the nondenatured sample (Habeeb, 1972). Under special conditions, DTNB reacts quickly with nonprotein SH and slowly with protein SH so that both types of SH groups can be determined in the same mixture. After Gabay et al. (1968), tissue homogenates are allowed to react with DTNB for 2 minutes at pH 6.8 giving nonprotein SH, and for 20 minutes at pH 7.6 giving total SH. Protein SH is then estimated by subtracting nonprotein SH from total SH. Boyne and Ellman (1972) described a kinetic analysis which allows the differentiation between (a) soluble, rapidly reactive SH (GSH-like), (b) soluble, slowly reactive SH (BSA-like*), (c) soluble, unreactive SH, (d) insoluble, reactive SH, and (e) insoluble, unreactive SH. Butterworth et al. (1967) suggest another method for the determination of protein SH in the presence of nonprotein SH: from the mixed protein disulfide, which is separated from the system after reaction with DTNB, the thiophenolate anion is released by the addition of dithiothreitol and then measured. *BSA
=
bovine serum albumin.
12
KLAUS HOFMANN AND REINER HAMM
Under the conditions used in the SH determination, DTNB also reacts with sulfite, thiosulfate (Man and Bryant, 1974), hydrogen sulfite, cyanide, and sulfide (Benedict and Stedman, 1970); therefore, such compounds interfere in the SH determination. Principally, all substances carrying a sulfur containing anion at pH 8 react with DTNB. Thiamine is another interfering factor which has not been taken into consideration up to now. In basic solution a cleavage of the thiamine ring occurs resulting in the formation of a S - group (Zima and Williams, 1940; Vogel and Knobloch, 1953):
By the reaction of thiamine with DTNB at pH values as low as 8.2 (i.e,, under the conditions of the SH determination), a yellow color is obtained. Actually, this reaction can be used for the determination of thiamine (Hofmann, 1974~). However, the thiamine content of meat is not sufficient to cause a noticeable error in the SH determination by DTNB. The maximum thiamine content of meat is about 0.9 mg per 100 gm of meat (Schweigert and Payne, 1956), whereas the total SH content varies between 60 and 80 mg SH per 100 gm of meat (see Section IV, A). For certain investigations, DTNB cannot be used. An example is the study of the reaction of patulin with the SH groups of meat (Hofmann et al., 1971): Patulin
+ RSH
-+
Patulin-SR
Following an addition of DTNB to measure the SH concentration after a certain time of reaction, first an equivalent amount of E-SH is obtained (see above). The E-SH, however, can react with patulin, and, consequently, the SH content measured in this manner will be too low. The determination of SH groups with other disulfides was suggested by Bitny-Szlachto et al. (1963), Drabikowski and Bitny-Szlachto (1963, I964), Kakol et al. (1964), Drabikowski and Nowak (1965), Grassetti and Murray (1969), Kakol (197I), and Swatditat and Tsen (1972). The use of thiamine disulfide (Kiermeier and Hamed, 1962; Kohno, 1965, 1966) allows a particularly sensitive SH assay: The thiamine formed by the SH-SS exchange is measured with the fluorometric thiochrome method. Here the relatively great error (+ 10%) of the thiochrorne method is a disadvantage. Bis(p-nitropheny1)-disulfide(Maier, 1969) is suitable for the determination of the sum of volatile mercaptans and hydrogen sulfide which arise, e.g., during heating of meat (see Section V, A, 4).
SULFHYDRYL AND DISULFlDE GROUPS I N MEATS
13
2 . Mercaptide-Forming Reagents and Amperometric Titration Unsatisfactory results with oxidation methods lead to the study of reactions with heavy metals, which react with SH compounds forming undissociated mercaptides in which the H of the SH group is replaced by a heavy metal. Reagents frequently used for the determination of SH groups in tissues are silver salts, mercury salts, and organic mercury compounds of the composition R-HgX(R = alkyl or aryl residue; X = halogen or OH). The procedures preferred are titrations with amperometric or potentiometric end-point determination using platinum or dropping mercury electrodes. This type of indication is based on the reducibility of heavy metal ions at the surface of the electrode. AS to the theory and application of the use of electrometric methods in biochemistry, we refer to the special literature (Kolthoff and Lingane, 1952; Konopik, 1953; Ewing, 1960; Brezina and Zuman, 1956; Purdy, 1965). a. The Amperometric Titration with Silver Nitrate. The amperometric titration of SH compounds with AgNO, was introduced by Kolthoff and Harris (1946) and first applied to the investigation of proteins by Benesch and Benesch (1948). Benesch et al. (1955) improved this assay essentially by using tris(hydroxymethy1)aminomethane as a buffering agent instead of ammonia and by the use of Hg/HgO/Ba(OH), reference electrode. Due to the potential of this electrode (-0.1 V against saturated calomel electrode), air oxygen cannot be reduced at the indicator electrode and, therefore, a complete elimination of oxygen (which could not be realized) is unnecessary (Kolthoff et al., 1965b). (Another reason for eliminating oxygen is the prevention of SH oxidation; such an oxidation. however, can be extensively prevented by addition of EDTA.) From that time, the amperometric titration with AgNO, was widely used for the determination of SH-(and SS-) groups in proteins and tissues. Important work on the methodology-other than animal tissues-was carried out by Rosenberg el a f . (1950), Heide (1953, Kolthoff et al. ( I 957, 1965b), Stauff and Duden (1958), Carter (1959), Staib andTurba (1956), Gruen and Harrap (1971), Harrap and Gruen (1971) (Ag/S specific ion electrode), and Mildner et ul. (1972). The great stability of the titration agent AgNO,, the simplicity of the apparatus (which can be easily built in each laboratory), the exact end-point determination (see Fig. I ) , and the small number of interfering factors involved are the reasons that this method was applied to practically all SH-containing materials including tissues. Animal tissues which have been investigated include liver, kidney, heart (Bhattacharya, 1958, 1959; Lastovskaya, 1969; Pavlyuk and Genyk, 1970), tumor tissues (Neogy et al.. 1961a,b), nerves (Krasnov, 1962), myofibrils from skeletal muscle (Hofmann, 1964, 1971a; Hamm and Hofmann, 1965, 1 9 6 6 ~Tinbergen, ; 1970), whole muscle tissue (Krylova and Kusnezowa, 1964; Hamm and Hofmann, 1965; Lastovskaya, 1969; Hofmann et al., 1969, 1974; Bolshakov and Mitrofanov, 1970; Bognar, 1971a; Hofmann, 1971b; Bow-
14
KLAUS HOFMANN AND REINER HAMM
FIG. 1. Amperometric titration curves: (A) Direct titration of SH groups with silver nitrate. Values of current (marked by circles) recorded 30 seconds after each addition of the reagent (0.1 ml AgNO,). (B) Titration of SH protein as in A, but with a continuous registration of the current (notice the decrease in current after the early additions of reagent which indicates the retarded reaction of the SH groups of protein). (C) AgNO, is titrated with an SH compound or KJ (reverse SH titration).
ers, 1972; Kortz, 1973; RaheliC et al., 1974), canned meat products (Bem et al., 1970; SusiC et al., 1974), and freeze dried meat (Potthast, 1972). Furthermore, the SH and S S groups in wheat flour were determined by amperometric (Rohrlich and Essner, 1966) and potentiometric titration (Kiihbauch and Wunsch, 1971). When compared with the amperometric titration, the potentiometric titration has two notable disadvantages: (a) the end-point determination is less exact because the point of inflection of the S-shaped titration curve is not very sharp, and (b) the air oxygen has to be removed as completely as possible because the use of a stronger negative electrode (calomel electrode) is necessary, the potential of which allows the reduction of oxygen. Several authors used the amperometric titration with AgN03 for the determination of SH groups in the isolated muscle proteins such as actin, myosin, tropomyosin, etc. (Azzoneef al., 1956; Staib and Turba, 1956; Tortschinski, 1959; Poglasov and Baev, 1960; Kofman, 1963; Berg et al., 1965; Hamm and Hofmann, 1965; Lytvynenko er al., 1963; Lusty and Fasold, 1969; Yuan, 1970; Brennock and Read, 1972; Hofmann, 1972a). The principle of the amperometric titration is as follows: a platinum wire indicator electrode is placed in the solution to be titrated; an electric connection is made with the reference electrode. The diffusion current flowing through the cell (caused by free Ag+ ions) is read on a microamperometer. During the titration of SH groups with silver nitrate, the current is near zero until the end-point because the Ag+ ions added are still being consumed by the formation of silver mercaptide: R-SH
+ Ag+ * R-SAg + Ht
After the end-point is reached, the diffusion current of silver rises because of an excess of silver ions in the solution. This diffusion current is proportional to the concentration of silver ions. By plotting the current reading during the titration
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
15
against the volume of silver nitrate solution, two straight lines are obtained which intersect at the end-point (Fig. 1A). By replacing the microamperometer with the recorder of a polarograph, a stepwise course of the current is obtained corresponding to the added portions of silver nitrate solution (Fig. 1B) (Beneschet af., 1955; Hofmann, 1972b). Instead of the rotating platinum electrodes which were originally preferred (Kolthoff and Harris, 1946), the use of a more simple stationary platinum electrode is recommended (Staib and Turba, 1956; Carter, 1959; Hofmann , 1964; Mallik, 1965; Hamm and Hofmann, 1966c) whereby the medium is agitated by means of a magnetic stirrer or by the use of a rotating beaker (King and Morris, 1967). The stationary electrode is even more sensitive than the rotating platinum electrode (Mallik, 1965; Richmond and Somers, 1966). If only small volumes of fluid are available for the demonstration, the titration can be carried out using a vibrating platinum electrode (Rosenberg et af., 1950). In addition to the platinum electrode, an AgS specific electrode was also suggested (Gruen and Harrap, 1971; Harrap and Gruen, 1971). Figure 2 shows the schema of a titrations apparatus with a stationary indicator electrode (after Staib and Turba, 1956; modified by Hamm and Hofmann, 1966~).The apparatus is only slightly susceptible to interference; the platinum electrode has to be cleaned only after a great number of titrations by dipping it into warm 20% nitric acid (contrary to information in the literature, e.g., Leach, 1966, p. 19). A direct titration of tissue homogenates is difficult because tissue particles adhere to the indicator electrode and cause an irregular influence on the diffusion current. Therefore, no regular titration curves are obtained (Hamm and Hofmann, 1966c; Bolshakov and Mitrofanov, 1970; Pavlyuk and Genyk, 1970). The direct titration of dissolved proteins with slowly reacting SH groups also implies problems. These difficulties can be overcome by the “indirect titration”: In this procedure an excess of AgNO, is first added to the sample; after the time required for a complete reaction, the excess of AgNO, is titrated with a SH compound (Neogy et al., 1961 a; Lusty and Fasoid, 1969; Bolshakov and Mitrofanov, 1970) or with KJ (Hofmann, 1969, 1971b; Bem et af., 1970; SusiC et af., 1974) (see titration curve Fig. 1C). The “double-indirect titration” procedure (Hofmann, 1964; Hamm and Hofmann, 1966c), which also allows the use of SH reagents such as NEM or PCMB (Hofmann and Hamm, 1967b; Hofmann, 197 Ic), is even more variable. The applicability of phenylmercuric acetate for this method was also examined (Mildner et al., 1972). This method of “double-indirect titration” should be briefly described because of its general applicability. The meat or protein sample, which should contain between 0.2 and 0.8 pmole SH (e.g., 25 mg tissue) is first incubated with I .O pmole of the SH reagent (AgNO,, PCMB, NEM, etc.). After the reaction, the mixture is filtered and 1.0 pmole of SHglutathion is added to the filtrate. The glutathion remaining after reaction with the excessive SH reagent is finally titrated amperometrically with IOp3M AgNO,
16
KLAUS HOFMANN AND REINER HAMM
iE 'I
17-
-4 IndKO tor
Reference electrode
2 3
FIG. 2 . Schema of the apparatus for the amperometric titration of SH groups (Hamm and Hofmann. 1966~).( I ) Indicator electrode (glass tube with platinum wire. 0.5 mm in diameter and I cm in free length). (2) Reference electrode (platinum wire, 0.5 mm in diameter and 2 cm in free length. dipped in mercury which is covered by a thin layer of HgO and Ba(OH), and a solution saturated by both substances. (3) Conductive connection (saturated KCI solution). (4) Porous diaphragm (clay). (5) Measuring instrument. (6) Magnetic stirrer with a constant speed of rotation. (7) Coated magnetic rod. The platinum wires of the two electrodes can be connected with the measuring instrument by either a drop of mercury or by welding them together directly. The content of the electrode vessels is 100 ml each; the volume of the sample solution 36 ml. Through the open tube in the top of the titration vessel, an inert gas can be led into the sample solution by means of a small rubber tube.
solution in tris buffer pH 7.4 (Fig. 3). As Fig. 3 shows, the consumption of AgNO, in the double-indirect titration procedure corresponds exactly to the amount of SH present. Therefore, the result is the same as would be obtained with a direct titration, if that method could be applied at all. Another essential advantage of this procedure is the fact that the titration of GSH with Ag+ ions results in an ideal titration curve, as shown in Fig. IA and B. With the direct titration of proteins, however, curves are sometimes obtained, the first part of which is not horizontal but at a slight incline. The problem of the evaluation of such titration curves was recently discussed by Hofmann and Hamm (1974b). We have applied this method to investigations about the role of SH groups in meat during heating, curing, freezing, storage, etc. (Hofmann, 1971d). The method has also been successfully used by other research workers for the deter-
17
SULFNYDRYL A N D DISULFIDE GROUPS IN MEATS
I
I
1 - 1
SH reagent ( 1 ml lO-3M
21
prnole)
protein SH (eg. 0 6 p r n ) remaining
m -
SH reagent (04p m f
SH glutathion (1 mllO-3 M
91 pm)
remaining SH glutathion (0.6 +m)
AgN03 consumed during titration (0.6+m)
L ’
I
I
1 pMd
1
*
FIG. 3 . Principle of the determination of SH groups by the double-indirect titration method (Hofmann. 1964; Hamm and Hofmann, 1 9 6 6 ~ )At . top of illustration, SH reagent. is valid for SH reagents which bind I SH group per molecule. c . g . . AgNO,. NEM. o r PCMB.
mination of SH groups in meat and myofibrils (Tinbergen, 1970; Bognar, I97 1 a; Potthast. 1972; Bolshakov and Mitrofanov, 1970; RaheliC et al., 1974). Doubleindirect titration permits a great variety in both reaction conditions and in the types of the SH reagent primarily used. The application of different reagents using the same technique is recommended for checking results obtained with one special reagent, for example AgNO,. This possibility is of particular interest because the specificity of SH reagents is not always certain. As a result, the specificity of Ag+ ions for SH groups in proteins has often been questioned (cf. Kolthoff and Stricks. 1950; Cecil and McPhee, 1959; Benesch and Benesch, 1962; Leach, 1966; Burton 1958), although this criticism was founded on investigations with low molecular compounds (mainly with cysteine). This problem was discussed in detail by Hofmann and Hamm (1974a). They concluded that results obtained with low molecular SH compounds cannot simply be transferred to proteins. On the contrary, studies using proteins with defined SH and SS content showed that under the conditions of the amperometric titration (tris buffer pH 7.4), Ag+ ions can be considered as being specific for protein SH groups (Hofmann and Hamm, 1975). It should be mentioned, however. that the direct titration of SH proteins with AgNO, in the presence of 8 M urea may lead to overly high SH values. This occurs because of a cleavage of SS groups with the excess of silver ions as Kolthoff et al. (1965b) found with bovine serum albumine. In absence of urea, however, they obtained correct SH values. In order to increase the specificity of the titration, a “blank” titration has been carried o u t using an excess of PCMB to block the protein SH groups (Bhattacharya, 1958, 1959). But the “blank” obtained in this way seems to be questionable because PCMB reacts more slowly and less completely with the SH groups of tissue than
18
KLAUS HOFMANN AND REINER HAMM
AgNO, (Hamm and Hofmann, 1967). Therefore, instead of a real blank, it is the SH groups unavailable for PCMB which are determined. From results obtained with several cysteine derivatives it can be concluded that Ag+ ions and mercaptide form a complex: R-SAg = x Ag+ -+ (R-SAg)Ag$
Therefore, the possibility of an excess of Ag+ consumption during the titration depends on the conformation of the SH compound. Complex formation apparently occurs if a cysteine derivative of Type I or I1 is present, while Type 111 does not form complexes (after Hofmann and Hamm, 1974a). +
HaN-CH-CGR
I
r* SH
Type 1
R-NH-CH-COO-
I
TH2
R-HN-CH-C&R
I
CHp
I
SH
SH
Type I1
Type 111
The use of AgNO, for the SH determination in muscle tissue does not result in inflated SH values as has been shown by comparison with other SH reagents (Tinbergen, 1970; Hofmann, 1 9 7 1 ~ cf. ; Section IV, A). During elaboration of the procedure for the SH determination in tissues by amperometric titration with AgN03 (described previously), Hamm and Hofmann (1 966c) investigated several possible influences. The results are as follows: (1) Between 10" and 40" C, the temperature of the GSH solution during titration has no influence on the titration end-point.* At 5"C, however, an elevated end-point was observed. (2) The titration of GSH in 8 M urea leads to the same result as the titration in the absence of urea. (3) An increase in the excess of AgN0, (over SH) from 100 to 200% does not influence the result of the indirect titration of myofibrils. (4) After 1 hour's reaction of AgNO, with myofibrils and subsequent titration, the same results are obtained whether these procedures are carried out in the presence or in the absence of air (under nitrogen). (5) Denaturation of the myofibrils with 8 M urea does not result in a significant increase of SH groups; therefore, all SH groups present in the native muscle proteins seem to react with AgNO,. (6) The recovery of GSH added to myofibrils and to the total tissue was 97% and 100% respectively. The accuracy of this amperometric titration is 97-98% (Krasnov, 1962). The error in the estimation of SH in animal tissues with this method has been deter*This was also found by Hoch and Vallee (1960).
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
19
mined to be from I .5% (Pavlyuk and Genyk, 1970) to 3% (Hamm and Hofmann, 1966~). The presence of Ca2+,Mg’+, Zn”, Fe3+, NO,-, Pod3-, lactate, and ATP at levels found in meat and meat products does not influence the amperometric titration of GSH with AgNO,. However, the presence of Cu2+and Mn2+ ions at higher levels results in low SH values, probably by a catalytic oxidation of the SH groups which cannot be prevented by the addition of 0.5% EDTA (Hofmann, 1970). Concentrations of sodium chloride higher than 0.1 M cause a change in the normal titration curve and, therefore, make the determination of the end-point less exact. Titrations in a 0.6 M KCI solution also result in abnormal curves; thus the amperometric titration of actomyosin in this solution is not useful (Hofmann, 1970). Finally it should be mentioned that free amino acids other than cysteine, which are always present in tissues, do not interfere in the amperometric titration (Benesch er al., 1955; Hofmann and Hamm, 1975). The same is valid for NAD, hemin chloride, ascorbic acid, and oxidized glutathion (Benesch et al., 1955). b. Mercury Compounds as SH Reagents. Numerous Hg compounds have been suggested for the determination of SH groups in proteins; we will discuss only the most important of these compounds. In addition to Hg2+ salts, organic mercury compounds have been used. Here the organic residue R‘ in reaction ( I ) contains hydrophilic groups for increasing the solubility in water and chromophoric groups for producing a measurable color. The reactions of protein SH (I) with an excess of the mercurial reagent (11) are as follows:
+
(I)
x R-SH
(2)
2~ R-SH
(1)
(X
+ Y)
R’-Hg+
+ (X + y) Hg2+ (W
-H+ + - H+ 4
x R-S-Hg-R‘ x (R-S)2Hg
(111)
+ yR’-Hg+
+ y Hg2+ (IV)
Three different principles are used for measuring these reactions: (a) Determination of the reaction product (111) in reaction (1) and (2): (i) Measurement of the increase in optical density at 250-255 nm which occurs if p-chloromercuribenzoate (PCMB) reacts with SH groups (Boyer, 1954). This method is only possible if the SH protein is dissolved and if the solution remains clear during reaction with PCMB. In order to correct for slight changes in opacity, the difference in readings at 255 and 320 nm was used (Yasui et al., 1968). By stepwise addition of the mercurial, a spectrophotometric titration can also be carried out (Katz and Mommaerts, 1962; Tonomura and Yoshimura, 1962; Arai and Watanabe, 1968). (ii) By coloring the tissue or insoluble proteins with 1-(4-chloromercuriphenylazo)-naphtol-2 (Flesch and Kun, 1950; Burley, 1954; Szydlowska et al., 1967), a subsequent, semiquantitative evaluation is possible. Further reagents suggested
20
KLAUS HOFMANN AND REINER HAMM
for labeling SH groups are 2-chloro-mercuri-4-phenylazophenol and 2-chloromercuri-4-(p-nitrophenylazo)-phenol(Chang and Liener, 1964), and fluorescein1,3,6,8-tetramercuric acetate as a sensitive spot reagent (Havir et al., 1966). (iii) Reaction of the SH proteins with I4C-PCMB and measurement of the radioactivity of the product (Erwin and Pedersen, 1968; Krabow and Golosby, 1971). (b) Determination of the reagent in excess (IV): ( i ) Measurement of the absorption of the reagent used, e.g., of PCMB at 232 nm after separation from the insoluble proteins (Hamm and Hofmann, 1967). (ii) Transformation of the reagent in excess into a colored complex, e.g., by the reaction of PCMB with dithizon (Fridovich and Handler, 1957; Sasago et al., 1963), or the reaction of o-hydroxymercuribenzoic acid with thiofluorescein (absorption maximum of 588 nm) (Wronski, 1967). ( i i i ) Titration of the reagent in excess with cysteine after reaction with the tissue using sodium nitroprusside as an indicator. The end-point of the titration is the appearance of a red color (MacDonnell et al., 1951; Zahn et al., 1962). Another possibility is the potentiometric titration of the excess of PCMB with cysteine (Calcutt and Doxey, 1959, 1961; Calcutt, 1961; Calcutt et al., 1961; Doxey, 1961) and the indirect back-titration of excessive SH reagents such as phenyl mercuriacetate (Mildner et al., 1972) and other suitable Hg reagents using the technique described by Hamm and Hofmann (1966~). (iv) Polarographic determination of the reagent in excess such as CH,HgJ or CH,HgCl (Maclaren et al., 1960; Leach, 1960a,b; Hird and Yates, 1961; Jamieson et al., 1963; Forbes and Hamlin, 1968; Mrowetz and Klostermeyer, 1972; Mrowetz et al., 1972; Marsalova and Roozen, 1973). (c) Estimation of reagent consumed (11) minus (IV): Direct titration of the protein SH groups in the presence of sodium nitroprusside until the purple color disappears. Reagents suggested for this titration are CH,HgNO, (Katchalski et al., 1957; Barany et al., 1964; Dworschak, 1970), NEM (Tsao and Bailey, 1953; Connell, 1957), PCMB (Connell, 1960a,b) and phenyl mercurihydroxyde (Meichelbeck, 1963). The titrations must be carried out at a temperature near 0°C because of the instability of the color. Furthermore, titrations with o-hydroxy mercuribenzoic acid and thiofluorescein (with which the color changes from blue to clear) (Wronski, 1963) or with salyrganic acid (mersalyl) and azopyridine (pyridine-2-azo-p-dimethylaniline)(Klotz and Carver, 1961; Ehrlich, 1967; Parker and Kilbert, 1970) as indicators are possible. Amperometric titration with HgClz is often used because of its very exact end-point determination (Kolthoff et al., 1957, I965b; Oganessjan and Dschanibekova, 1958; Matsumoto et al., 1960; Sullivan et al., 1963). Amperometric titrations can also be carried out with methylmercuric and ethylmercuric chloride (Kolthoff and Tan, 1965; Kolthoff et al., 1965b). Two advantages
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
21
of these two mercurials are the ease of preparation of extremely pure ethyl mercuric salts and the high level of water solubility of methylmercurinitrate. A disadvantage of the use of Hg2+ salts for SH determination is the fact that it is not certain whether the bivalent Hg cation reacts with one SH group only, forming RSHgX(X = halogen). or with two SH groups, forming (RS),Hg (Benesch and Benesch, 1962). Therefore, the application of organic mercurials, which can react with one SH group only, is generally preferred. PCMB is the most frequently employed compound of this type. It is extremely stable and, contrary to most of the other mercurials, has a low level of toxicity. Its low solubility in water and the difficulty of preparation of PCMB of high purity are disadvantages of this reagent (Benesch and Benesch, 1962). Impurities in commercial preparations can cause a catalytic oxidation of SH groups as has been demonstrated in the following example. When Bendall (personal communication) tried to block a part of the SH groups of the isolated myosin with PCMB, he found that the amount of SH groups decreased more strongly than had been expected from the amount of PCMB added. Hofmann and Bendall (unpublished observations) used the following experiment in order to find the reason for this phenomenon: They added to GSH solution in tris buffer pH 7.4 (a) unpurified PCMB, (b) purified PCMB (twice recrystallized in NaOH), and (c) unpurified PCMB and EDTA (final concentration 0.01 M ) . The molar proportion SH: PCMB was 4: 1 in all cases. A stream of air was passed through each solution for 30 minutes. Before and after aeration the SH content in an aliquot of the solutions was determined by Ellman’s reagent. The solution of the unpurified PCMB (without EDTA) showed a decrease in the SH content of 21%, the other two solutions a decrease of 2 and I % , respectively. This result indicates that the unpurified preparation contained traces of heavy metals which are known to catalyze the oxidation of SH groups. The spectrophotometric assay of SH groups in myofibrils by PCMB (estimation of the excess of the reagent by measuring the optical density at 232 nm) involves a relatively high error factor (12.6%) (Hamm and Hofmann, 1967); therefore, this method does not allow an exact SH determination in tissue or in myofibrils. Finally some methods should be mentioned which may be of interest for analytical as well as for preparative purposes. An insoluble reagent formed by the binding of PCMB to Dowex-2 resin has been used for the selective removal of thiols from solutions or from tissue homogenates (McCormack er al., 1960). It was possible to remove the bound thiol from the reagent by exchange with other SH compounds. This reagent might be valuable for concentrating or isolating these thiols. An organomercurial-polysaccharide has been synthesized and successfully applied to the separation of protein mixtures into an SH-fraction and a fraction containing no SH groups (Eldjam and Jellum, 1963). This material has been applied as a purification step in the isolation of SH enzymes as well as for concentrating dilute solutions of these enzymes. It is particularly well suited for
22
KLAUS HOFMANN AND REINER HAMM
the chromatographic fractionation of individual SH proteins. Furthermore, an organomercurial resin has been prepared which is capable of binding low molecular SH compounds and SH proteins, which can then be recovered almost quantitatively by elution with cysteine (Liener, 1967).
3.
N-Ethylmaleimideand Its Derivatives
N-ethylmaleimide (NEM) contains a reactive double bond causing an absorption maximum in the UV range of 300-302 nm. At pH values around 7, an addition of SH groups to the double bond occurs.
R-SH
+
R-S,H
HC=CH O(L,,,CO I
I -
C-CH,
I
1
,co
oc, I?
During this reaction the absorption maximum in the UV disappears because of the transition of the double bond into a single bond. The measurement of this decrease in absorbence at 300 nm permits a quantitative determination of SH groups. Principal studies on this method were camed out by Friedmann (1952), Gregory (1955), Alexander (1958), Roberts and Rouser (1958), and Leslie (1965). The formation of additive compounds by the reaction of thiols with NEM was demonstrated by Smyth et al. (1960) and Lee and Samuels (1961). The extreme stability of the C-S bond in the reaction product is an important advantage of this method. But NEM also shows some disadvantages which limit its application: (a) the sensitivity of the measurement is relatively small (eM = 620 at 300 mm); (b) in an aqueous solution, NEM is gradually disintegrated because of a hydrolytic cleavage of the CO-NH bond and the product of this hydrolysis reacts very slowly with SH groups (Gregory, 1955); (c) under certain conditions, several non-SH containing amino acids may react with NEM (Riggs, 1961), therefore, NEM is not always specific; (d) if proteins are not removed from the reaction mixture before SH determination, high blanks may appear because of the high absorbance of proteins around 300 nm. Nevertheless, NEM is widely used for blocking or determining protein SH. According to our experience, NEM is a suitable SH reagent provided that there is not too great an excess of the reagent and that the reaction time is not too long. Leslie et al., (1962) found NEM suitable for the determination of mercapto groups in proteins when they were denatured. Using NEM for the determination of SH groups in tissue and muscle proteins, Hamm and Hofmann (1966b) came to the following conclusion. The titration after Tsao and Bailey (1953) with sodium nitroprusside as an indicator is not appropriate because the change in color is indistinct due to the slow
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
23
reaction of the SH groups. But the following spectrophotometric procedure is quite suitable: 0.5 gm myofibrils of finely minced tissue (containing about 0.1 gm protein) is weighed into a 12-ml test tube and mixed with 5 ml 0.1 M phosphate buffer pH 6.0 and 5 m12 10-3M aqueous NEM solution. After shaking for 2 hours at 2 5 T , the main part of the proteins is removed by centrifuging at 15,000 X g for 5 minutes. The remaining proteins are precipitated in 5 ml of the clear supernatant by the addition of 1 ml 20% TCA. After centrifugation (15,000 X g for 5 minutes), the optical density of the supernatant is measured against a blank at 300 nm. In the calculation of the SH content, the partial hydrolysis of NEM (decrease in absorbence 0.003) and the moisture content of the sample (0.4. lop3 1) have to be taken into consideration. The SH content is calculated from the extinction E measured in a 1-cm cuvette by the following equation: pM01 SH = 10.00 - 20.12 x (E + 0.003). The presence of TCA does not interfere in the measurement of optical density. The protein precipitate does not include measurable amounts of NEM, provided that the water insoluble proteins are removed before as described above. With this method 3.5 moles SH/105 gm protein were found in myofibrils, whereas with AgN03 9.2 moles SH/l05 gm protein were determined in the same preparation. Therefore, NEM reacts only with a part of the SH groups which are defined as “easily available SH groups” of muscle proteins. A preceding denaturation of proteins, either by 8 M urea or by heating (cf. Section V,A, l), results in an increase in SH groups reacting with NEM; but the total number of SH groups in myofibrils which react with AgNO, cannot be achieved (Hamm and Hofmann , 1966b). An almost equal number of SH groups detectable with AgNO, can be determined by reaction with NEM if the NEM reacts at pH 7.4instead of pH 6.0 and if the subsequent SH determination is carried out by indirect amperometric titration (Hofmann, 1 9 7 1 ~ )It. follows from this result that at a higher pH the SH groups are more easily available for NEM. But the determination of the partially available SH groups with NEM at pH 6.0 is still of value because it allows one to follow the process of protein denaturation (unfolding) (see Section V,A, 1). Weitzman and Tyler (1971) found that NEM gives a well-defined polarographic reduction wave and that measurement of this can form the basis of a more sensitive procedure for estimating SH groups. Numerous N-substitution products of maleimide were synthetized in order to obtain derivatives which are more sensitive for optical measurement than NEM. These derivatives cannot be discussed in detail. Some of them were used for studying protein conformation because the reaction products of SH groups with NEM derivatives are resistant against protein hydrolysis by acids. Holbrook et af. (1966) studied the influence of the size of the substituents in N-substituted maleimides on their reactivity against SH groups in proteins. They found that the larger the substituent the slower the reaction. Colored or fluorescent compounds were obtained by the
24
KLAUS HOFMANN A N D REINER HAMM
introduction of chromophoric groups into the NEM molecule (Witter and Tuppy, 1960; Riordan and Vallee, 1972; Kanaoka et a / . , 1967, 1968, 1973; Sekine and Ando, 1972; Nara and Tuzimura, 1973). Such derivatives can be used for the histochemical demonstration of SH groups (Tsou et al., 1955). Furthermore, NEM and its derivatives were labeled by substituents containing I4C (Kielley and Barnett, 1961; Tkachuk and Hlynka, 1963; Lee and Samuels, 1964; Lee and Lai, 1967), 3?S and 35S(Merz et a l . , 1965). Maleimides were used for the determination of SH groups in myofibrils (Hamni and Hofmann, 1965; Hofmann, 1971 c), poultry muscles (Gawronski et a l . , 1967), fish muscle (Connell, 1957), and in muscle proteins, particularly in myosin and actin (Tsao and Bailey, 1953; Kielley and Barnett, 1961; Katz and Mommaerts, 1962; Groschel-Stewart and Turba, 1963; Martonosi, 1968; Seidel, 1969). According to Schoberl (1958), in addition to maleimides, vinylsulfones, which also bear reactive double bonds, are very suitable for the blocking and determination of protein SH groups, e.g., in keratins. 4 . Other SH Reagents Of the reaction types not described in the preceding sections, the nitroprusside reaction has been known for the longest time. The sensitivity of its reaction with SH groups in alkaline solution forming a red color is very great: (Fe(CN),NO]'-
+ R-S- + [Fe(CN),NO.S-R]'
In a strong alkaline solution, the color fades very quickly, while in an ammoniacal solution, it is quite stable. However, the color stability is insufficient for an exact quantitative determination of SH groups. Nevertheless, sodium nitroprusside has often been used for the assay of SH groups in meat (Chajuss and Spencer, 1962b; Khan et a l . , 1963; Khan and van den Berg, 1964, 1965; Motoc and Banu, 1968; Davidkovli and Davidek, 1971; Khan and Nakamura, 1971). Numerous efforts for increasing the color stability with additives or variation of reaction conditions have had only limited success. An addition of cyanide improves the color stability (Schoberl and H a m , 1948); but in this case a cleavage of SS groups occurs which might result in overly high SH values. Sodium nitroprusside has also been used as an indicator for the titration of SH groups of proteins with other SH reagents (Connell, 1960a,b); but such procedures imply problems which must not be overlooked. The discoloration of the protein-SH-nitroprusside complex in such titration methods is due to a replacement of the nitroprusside anion by the added SH reagent: [Fe(CN),NO. S-Protein]
+ reagent
-
[Fe(CN),NO]'-
+ reagent-S-Protein
SULFHYDRYL A N D DISULFIDE GROUPS IN MEATS
25
The sodium nitroprusside often reacts with only a limited number of the protein SH groups. Consequently, the change in color from red to almost clear will have already occurred if the titration reagent has reacted with the SH groups that participate in the nitroprusside color complex of the protein. So, rather than all the SH groups reacting with the titration reagent, only the groups which react with sodium nitroprusside are determined. Sodium nitroprusside is specific for SH groups in proteins as well as generally in animal tissues. The sodium nitroprusside reaction provides an excellent and very sensitive spot test for the qualitative indication of SH groups. With meats, the red myoglobin color disturbs the test; but with washed muscle tissue, myofibrils. or with meat which is pale by nature (e.g., poultry or fish muscle), good results are obtained. The SH test with sodium nitroprusside can be carried out also in ZnCI, solution (instead of ammonia) which causes a low acidification of the reaction mixture (Pohloudek-Fabini and Papke, 1964). Myofibrils take o n an intensively pink color, whereas, with solutions of SH compounds such as cysteine or glutathion, a pink precipitate is formed. The use of B mixture of I % sodium nitroprusside solution and a 30% ZnClz solution has been recommended (Hofmann, 1965a). An intensive color reaction is obtained with native muscle proteins. Serum albumin or egg albumin reacts only after denaturation by urea or heating (Hofmann, 1965a, 1966a). Contrary to the SH test in ammoniacal solution, creatine and acetone do not interfere. Sulfite and thioesters, however. do react. The presence of sufficient amounts of ammonium ions or EDTA prevents the color reaction of nitroprusside ZnC1, by the formation of Zn complexes (Hofmann, 1965a). This result suggests that zinc participates in the color formation. The spot test with iodine azide represents a very sensitive SH reaction (see Feigl, 1960): 2NaN,
+ JB
('")
>
3 N4
+ 2 NaJ
This reaction, which occurs in a sufficient rate only in presence of SH groups, can be recognized by the formation of small gas bubbles (nitrogen). SH reagents which have been known for a long time are iodoacetamide, iodoacetate, and iodoacetic acid. They are used for the alkylation of SH groups of the cysteine residues in proteins. After hydrolysis of the S-alkylated protein the cysteine residues are present as carboxymethyl cysteine which can be determined in the course of the amino acid analysis. Under certain conditions these reagents can react with methionine, histidine, and lysine residues of proteins (Gundlach ez ul., 1959). In many investigations, iodoacetamide, iodoacetate, and iodoacetic acid have been used for blocking and determination of SH groups in muscle proteins (Mirsky and Anson, 1935; Mirsky, 1936; Barany and Barany. 1959; Barany et al., 1964; Stracher, 1964).
26
K L A U S H O F M A N N A N D REINER H A M M
4-Iodobutanesulfonate was also recommended as an SH alkylating agent in the analysis of amino acids (Jermyn, 1966). 4,4’-Bisdimethylaminodiphenylcarbinol is a new, highly sensitive SH reagent, which shows in an acidic solution an intensive blue color (cM = 70800 at 612 nm). The product of the reaction between the dye cation and SH groups is colorless, probably because of the following reaction (Rohrbach et al., 1973): 1+
r
colorless
This decrease of optical density caused by SH groups can be used for the quantitative determination of protein SH. Urea and cyanate disturb this reaction because they also cause a discoloration; 4 M guanidine, however, does not interfere, Another sensitive test for cysteine, glutathion, and other SH compounds is possible by means of chloranil (I), bromanil (11). or 2, 3 dichloronaphtoquinone (111) (Hofmann, 1965b).
(11
(11)
(1x1)
For this test, the SH containing solution is alkalized by the addition of K,CO, and shaken with a solution of I, 11, or 111 in chloroform. After the separation of the aqueous and organic phases, the colored reaction products appear in one of these two phases (see Table 11). Disulfides and SH-free amino acids do not interfere in this reaction. The color of the solutions is very stable and is suitable for quantitative SH determination (Hofmann, 1965b).
27
SULFHYDRYL AND DISULFIDE GROUPS I N MhATS TABLE I1 COLOR REACTION OF HALOGENIZED QUINONES WITH SH COMPOUNDS I N ALKALINE (K,C03) SOLUTION“.”
SH compound
Chloranil bromanil
2.3-Dichloronaphtoquinone
Cysteine Glutathion Thioglycolic acid Eth ylmercaptane Thiamine (SH form) Thiamine pyrophosphate
Green (H,O) Reddish brown (H,O) Yellow (H,O) Red (CHCI:,) Reddish brown (CHCI:,) Yellowish brown (H,O)
Yellow (H,O) Pink (H,O) Orange (H,O) Yellow (CHCl3) Yellow (CHCI,) Pink (H,O)
(’ From Hofrnann (1965b). The color appears in the phase indicated in parentheses
Fontana et al. (1968) suggested azobenzene-2-sulfenyl bromide as an SH reagent specifically for cysteinyl residues; in this reaction asymmetric disulfids are formed:
/ S -Br
/
S-S-R
Friedman (1973, 1974) found that protein SH groups can be transferred quantitatively to acid-stable S-P-pyridyl ethyl cysteine residues by reaction with 4-vinylpyridine. The corresponding cysteine derivatives released by hydrolysis elute as discrete peaks from an amino acid analyzer. A related assay was also developed in which half-cystine residues are changed to S-P-(2-quinolylethyl)cystine residues. These side chains can be estimated by ultraviolet spectroscopy in intact or hydrolyzed proteins. Recently new fluorescent reagents, N-dansylaziridine and NBD-chloride (7-chloro-4-nitrobenzo-2-oxa1,3-diazole) have been described as selectively reacting with protein thiols (Scouten et al., 1974; Price and Cohn, 1975). These reagents may be used to label SH portions of proteins, to differentiate between buried and exposed sulfhydryls, and to determine the nature of the region (hydrophobic, hydrophilic) surrounding a given sulfhydryl group. In certain instances, the possibility of a specific determination of the content of reduced and oxydized glutathione in tissues is of interest. This can be done by means of enzymatic spectrophotometric methods (Klotsch and Bergmeyer, 1962; Lack and Smith, 1964; Srivastava and Beutler, 1968; Tietze, 1969).
28
KLAUS HOFMANN AND REINER HAMM
C.
DETERMINATION OF SS GROUPS
Usually the determination of disulfide groups in protein is carried out in two steps: (1) by reduction or cleavage of the SS groups to SH groups, and (2) determination of the SH groups formed. Since the methods for the determination of SH groups have been discussed, we will only consider the reduction of S S groups here. The reagents frequently used for this purpose are: (a) mercaptans such as P-mercaptoethanol (Anfinsen and Haber, 1961 ; Thompson and O’Donnell, 1961; Christian and Schur, 1965; Habeeb, 1972; Beveridge et al., 1974), thioglycolate (Katchalski et al., 1957; Sela et a/., 1959; White, 1960; Leach and O’Donnell, 1961), dithiothreitol (Cleland, 1964, 1968) or dithioerythritol (Habeeb, 1972); (b) sodium borhydride (Stahl and Sigga, 1957; Moore et al., 1960; Stauff and Duden, 1958; Brown, 1960; Seon et al., 1965; Cavallini ef al., 1966; Glaseretal., 1970); (c) sulfite (Kolthoff e t a l . , 1958, 1965a; Carter, 1959; Christian and Schur. 1965; Rohrlich and Essner, 1966); (d) cyanide (Grote, 1931; Wronski, 1964; Roberts and Rouser, 1958; De Marco ef al., 1966); (e) Tri-n-butylphosphine (Harrap and Gruen, 197I); and (f) electrolytical reduction (Leach ef al., 1965, see also Friedman, 1973). The corresponding reaction mechanisms are as follows: (a) 2 R-SH + Prot.-SS-Prot. + 2 Prot.-SH + R-SS-R Prot.-SS-Prot. 4 2 Prot.-S- H, or* (b) 2 H4 R-SS-R NaBH, 3 H,O + 8 R-SH + NaH,BO, (c) SOa2- Prot.-SS-Prot. 4 Prot.-SSOg Prot.-S(d) CNRot.-SS-Prot. 4 Prot.-SCN + Prot.-S(e) Bu,P Prot.-SS-Prot. H,O 4 2 Pro[.-SH + Bu,PO (9 2e Rot.-SS-Prot. -+ 2 hot.-.!-
+
+ + + + +
+
+
+
+
The reagents most frequently used are types a, b, and c. Mercaptans also are often added to proteins in order to protect SH groups. For a complete reduction of protein SS groups, the presence of a denaturing agent such as 8 M urea, 4 M quanidine, or 0.5% SDS solution and a considerable excess of the reducing agent are generally necessary. Therefore, in most cases, after reduction but before the SH determination, the excess of the reducing agent must be removed. This is an easy process in the case of NaBH,; by acidification of the alkaline reaction mixture, the reagent is completely destroyed. An excess of mercaptans has to be eliminated, e.g., by precipitation of the proteins. Sulfite has often been used for the cleavage of SS groups and subsequent determination of SH groups by amperometric titration with AgN03, which is not disturbed by an excess of sulfite. *The reaction mechanism seems to be unclear. In the first reaction, hydrogen is released; in thc second, no hydrogen appears (see Jocelyn, 1972; Friedman, 1973).
SULFHYDRYL A N D DISULFIDE GROUPS IN MEATS
29
The fact, however, that, after reaction of SS with sulfite, only one SH group is obtained is disadvantageous (see reaction c). Well-proved methods for the SS determination in proteins using NaBH,, P-mercaptoethanol, and dithioerythritol as reducing agents and subsequent SH determination with Ellman's reagent are described by Habeeb (1972). Reduction of SS by NaBH, has been shown to be convenient for the determination of SS in animal tissues (Hamm and Hofmann. 1965, 1966c; Dubi, 1969; Bognar, 1971a; Habeeb, 1972; Dubeet ul., 1972). A description of the SS determination in tissue (after Hamm and Hofmann, 1966c) follows: The sample, containing about 5 mg protein (e.g., 25 mg minced muscle tissue) is weighed into a conical centrifuge tube (10 ml), mixed with 0.5 ml reducing agent (0.12 gm NaBH, in 5 ml of 8 M urea), and kept for 1 hour at room temperature. Foaming can be prevented by the addition of a drop of octanol or by putting traces of a silicon defoamer on the glass wall of the tube. Then the NaBH, is destroyed by a stepwise addition of 0.15 M HNO, until a pH of 6 . 7 4 . 4 is obtained. After keeping the mixture under nitrogen for 30 minutes, 28 ml of water and 5 ml of a tris buffer pH 7.4 are added. In this mixture the SH groups are determined by amperometric titration. The use of hydrogen selenide allows a direct histochemical demonstration of SS groups (Olszewska et al., 1967); the sections on the slides are saturated with water vapor and treated with gaseous H,Se for 2 hours. The areas containing SS groups stain yellow-brown. The test is based on the reaction: R-SS-R
+ H,Se -+
2 R-SH
+ Se
The presence of thiol groups does not influence the results of this reaction. Another method which is appropriate for the determination of SS groups is based on the fluorescence quenching of fluorescein-mercury(I1)-acetate by SS groups (Karush et al., 1964). SH groups, which also react, are blocked beforehand with iodoacetamide. Maeda et ul. (1970) developed a method for visualization of cystinecontaining peptides in peptide maps. The chromatogram is sprayed with an NaBH, solution in ethanol and the excess of NaBH, is decomposed by dipping it in acid. The paper is dried, neutralized by exposure to ammonia vapors, and then sprayed with Ellman's reagent. Yellow spots appear immediately. These can be eluted and analyzed to establish the amino acid composition of the cystinecontaining peptides. A combination of the methods for demonstrating SH groups by means of 2,2'-dihydroxy-6,6'-dinaphthyldisulfideor Fast Blue B (Gabler and Scheuner. 1966) with Mercury Orange permits simultaneous color differentiation between SH groups (blue or purple) and cystine disulfide bonds (red-orange) (Szydlowska and Junikiewicz, 1973).
30
KLAUS HOFMANN A N D REINER HAMM
Ill. SH GROUPS IN MUSCLE PROTEINS AND THEIR ROLE IN THE FUNCTION OF MUSCLE The muscle cell consists of the myofibrils, the sarcoplasmic reticulum, and some cell organelles such as nuclei, mitochrondia, lysosomes, and ribosomes. This structural material is embedded in the fluid matrix of the sarcoplasma. The cell is surrounded by the cell wall, the sarcolemma. The structural cell elements including the membranes are to a large extent built up by proteins. A great variety of other proteins is dissolved in the sarcoplasmic matrix o r loosely attached to cell structures. The bulk of the sulfhydryl groups in muscle is bound to proteins. Nonprotein thiol is rather low and consists largely of glutathion ( I .5 pmole/gm tissue in rabbit skeletal muscle) (Jocelyn, 1972; see also Section IV,A). Most of the numerous proteins of the muscle cell contain sulfhydryl groups of physiological importance. It is not the purpose of this review to discuss in detail the role of SH groups in muscle physiology. For a better understanding of the reactions that occur in muscle as a food, however, some knowledge of the most important facts from the extensive field of muscle research might be necessary, particularly in future meat research, because thus far not much use has been made of the results from this research. With regard to the great number of different protein SH groups in the muscle, to the numerous effects of these sulfhydryls on enzyme activity, protein interactions, membrane transport reactions, etc., it might be very difficult to elucidate the importance of a particular type of sulfhydryl for meat quality or in changes in meat protein. Most of the information on the biochemical and physiological role of SH groups in muscle proteins has been obtained by studying the effect of blocking these groups with specific reagents during enzyme activity or protein interactions. Chemical modification of sulfhydryls has been extensively used for elucidating the structure of the active sites of adenosinetriphosphatase (ATPase) and other enzymes. In this type of study, the chemical modification must cause stoichiometric inactivation, and there must have been specific protection by the substrate or by competitive inhibitors against the inactivation. Even so, the possibility remains that this particular modification occurs other than at the catalytic site and that specific protecting agents induce a change in confinnation at a location other than the catalytic site (Tonomura, 1973). These points have not always been taken into account. Hence, not all the results mentioned in this chapter can be regarded as irrevocable. A.
MYOFIBRILLAR PROTEINS
The proteins which build up the myofibril are shown in Tables I11 and IV. Although all myofibrillar proteins are related to the contractile mechanism of the
SULFHYDRYL A N D DlSULFlDE GROUPS IN MEATS
31
muscle fiber, their functions are distinctly specialized (Table 111). The fundamental process of contraction is known to be carried out by myosin and actin, but these two proteins alone cannot bring about the contraction process of living muscle. The recognition of the physiological role of tropomyosin and troponin has further stimulated the discovery of new myofibrillar proteins. On the whole, the idea of regulatory proteins (proteins that enable myosin and actin to perform the contraction-relaxation cycle under physiological conditions) has been established (Ebashi and Nonomura, 1973).
TABLE 111 MYOFIBRILLAR PROTEINS OF THE SKELETAL MUSCLE"
'
/Myosin
Contractile proteins
\
Actin
I /
Troponin Functional Regulatory proteins
Structural M-Protein 'I
After Ebashi and Nonomura ( 1973)
TABLE IV CONTENTOFCONTRACTION AND REGULATORY PROTEINS IN RABBIT MYOFIBRIL"
Proteins
Percentage of protein by weight
Myosin Actin Tropomyoain Troponin &-Actinin P -Actinin M-Protein
55-60 20
'I
4.5
3 -5 1-2
-0.5 -0,s
After Ebashi and Nonornura (1973)
32
KLAUS HOFMANN A N D REINER HAMM
I , SH and SS Group Content of Myofbrillar Proteins Table V shows the data for SH and SS content of the single myofibrillar proteins. All myofibrillar proteins contain SH groups. Disulfide groups have been found only in tropomyosin and troponin. No reliable data for SH groups could be found in the literature for p-actinin. The results of the estimates of SH groups in myofibrils are listed in Table VI. Some of the values determined in native myofibrils with NEM, PCMB, and DTNB obviously do not represent the total SH content but rather the SH groups available under the conditions used in the reactions. The average SH content of the different myofibrils is 9.1 moles SH/I05gm protein (values lower than 6 were omitted). This value corresponds very well to the sum of SH groups of the single myofibrillar proteins, calculated from the values of Table VII. As the average SH content of whole muscles was 10.2 (pork) and 10.5 (beef) (see comments on Tables IX and X), it is evident that the SH groups in meat are generally about 90% bound to myofibrils. Table VII provides the figures for mg SH and SS content of the major proteins per 100 gm of the total myofibrillar protein. These TABLE V SULFHYDRYL AND DlSULFlDE GROUPS IN MYOFIBRILLAR PROTEINS. EQUIVALENTS OF SH AND SS GROUPS PER lo5 gm PROTEIN
Protein
Molecular weight
SH
SS
8-9!,.c.d.?J
0u.d.~
8.4' 7.3' 8.6' 7 - 10"
0
SH
+ SS"
~~~~~~~~~~~
Myosin
HMM
220.000
320-360,000
8.4'
7.4" 10' 6.4"
Subfragment I Subfragment 2 LMM Actom yosin
4b.P
n.n"
0 0 0'
7 .5' ) I 8.5'
Actin
Actinin
46.OO0-47.000
I2"."
-95,000
13' 30.3" 7.8' 9-1 I " 5.1'
O"."
7.W 9.3'
33
SULFHYDRYL AND DlSULFlDE GROUPS I N MEATS TABLE V
Protein Tropomyosin
Molecular weight
( L ontinued)
SH
SS
SH
+ SS"
3-5" 4.31 4.7"
70.000 ( o r 35.000)
3. I'
Troponin
8 1 .OOO
5" 5.4"'
Troponin I
24,000
13.9.' 12.0"
Troponin T
37.000
Troponin C
1 8 .OOO-20 .OW
M-Protein
165.000
I .6.r 0. I " 0.4'' 8.9' 5.0'' 3.8" 4.4""
I"
3.7' 4" 4.3" 4.6' 4"
10"
12.2'
10.2''
" Usually determined as cysteine acid by amino acid analysis. in some cases also as SH after reduction 0 1 SS. " Ebashi and Nonomura (1973). ' Woods and Hartley (1967). " Bariny P r t i / . (1964). " Tonomura (1973). Robson and Zeece (1973). "Connel (1961). * Buttkus (1971). a Brennock and Read ( 1972). Hofmann (l972a). Ehrlich (1967); Hamm and Hofmann ( 1965). Staib and Turba (1956). ''I Kofman (1963). " Drabikowski and Nowak (1970). " Carsten (1966). Martonosi (1968). Cohen et a / . (1973). Suzuki P I N / . (1973). Yasui et a / . (1968). Hodges and Smillie (1970). ' Hodges and Smillie (1972). ' Ebashi er d.(1968). Arai and Watanabe (1968). Wilkinson et cd. (1972). ' Greaser EI a / . (1973). Schaub et a / . (1972). "" Masaki and Takaiti (1974).
'
'
34
KLAUS HOFMANN AND REINBR HAMM TABLE VI ESTIMATES OF SH GROUPS IN MYOFlBRlLS
SH reagent
Moles SH (per lo5 gm protein)
Original SH data"
AgNOS NEM NEM PCMB PCMB DTNB DTNB DTNB DTNB DTNB CH3HgN03 o-iodobenzoic acid
8.5-9.0 9.5 7.8-12.9" 3.5-4.0 8.9-9.7
85-90 pmoleslgm protein (1) 3.14 mg/gm protein (2) 2.59-4.27 mg/gm protein (3) 1.16-1.31 mglgm protein" (2) 2.95-3.21 mglgm protein (3) 5.4 moles/105 gm protein (4) 3.06-3.09 mg/gm protein (3) 0.59 pmolelmg N ( 5 ) 3.0-3.9 moles/105 gm protein (6) 8.6-9.6 moles/105 gm protein" (6) 60-65 pmoledgm protein ( I ) 85-90 pmoledgrn protein' ( I ) 8.8 mole/105 gm protein (7) 2.75-4.32 mg/gm protein (3)
''
5.4
9.2-9.3 9.4 3.0-3.9 8.6-9.6 6.0-6.5 8.5-9.0 8.8 8.3-13. I
Numbers in parentheses correspond to the following investigations: ( I ) Tinbergen (1970): (2) Hofmann and Hamrn (1966); ( 3 ) Hofmann, Miiller. and Baudisch (unpublished observations); (4) Arai and Watanabe (1968); (5) Khan rt al. (1968); (6) Hay rr a / . (1973); (7) Barany rt ( I / . ( I 964). " Mean value: 10.3 t 1.26 (n = 20). ' Reaction at pH 6.0. " Reaction at pH 7.4. " Myofibrila denatured. "
TABLE VII DISTRIBUTION OF THE SULFHYDRYL AND DISULFIDE GROUPS UPON THE MAJOR MYOFIBRILLAR PROTEINS
Protein Myosin Actin Tropomyosin Troponin
mg SH/100 gm total myofibrillar protein
%
mg SS/lOO gm total myofibrillar protein
%
I56 70 3
65 29
-
-
-
1
10
4
12 11
48 52
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
35
figures were obtained by combining the mean values of Tables Wand V. As Table VII shows, more than 95% of the sulfhydryls of myofibrils are located in the actinmyosin system, whereas all disulfide bonds are built into the tropomyosintroponin system.
a . Myosin. There are about forty-two SH groups in the myosin molecule. Many authors have found that the number of SH groups in myosin which are titratable with PCMB is approximately equal to the halfcystine content found by amino acid analysis. Myosin probably, therefore, contains no SS bridge (Tonomura, 1973). In Table V,the values for the myosin subunit heavy meromyosin (HMM). including its components, subfragments 1 and 2 and the subunit light meromyosin (LMM), are also listed. The amino acid sequence around the SH residues of myosin has been extensively studied. The conclusion that myosin contains at least 16 and probably between 20 and 22 unique thiol sequences indicates that the molecule consists of two chemically equivalent components (Weeds and Hartley, 1968). There is remarkable agreement that purified myosin has three distinguishable light chains (C,, Cqr C3). SH groups seem to participate in the interaction between the chains, because the blocking of SH groups of myosin with 5,5'dithiobis(2-nitrobenzoic acid) (DTNB) tends to release light chain C,. C, and C3 appear to have similar peptide chains. each containing a single cysteine peptide; C2 contains two cysteine residues. The amino acid sequence of these peptides has been clarified (Taylor, 1972). b. Actin. The SH groups of actin, which are titratable with PCMB or 2.3dicarboxy-4-iodoacetamide azobenzene are found to be approximately equal to the number of halfcystines in the molecules as determined by amino acid analysis. Like myosin, actin contains no SS-bridge (Tonomura, 1973). Only 1 mole of SH per mole of G-actin binds rapidly with SH reagents, and subsequently there is a slower reaction. The same holds true for F-actin: Only I mole per mole G-actin unit binds with SH reagents, while the remainder reacts even more slowly than with G-actin (Tonomura, 1973). Several SH-containing peptides have been isolated from the tryptic hydrolysates of actin (Young, 1969). Now the positions of the reactive and not available cysteine SH groups in the peptide chain of the G-actin molecule have been elucidated (Elzinga and Collins, 1973).
c. Tropomyosin. The molecular weight of tropomyosin was found to be about 70,000. In the presence of reducing agents, e.g., of P-rnercaptoethanol, tropomyosin dissociates into two similar subunits, each of the molecular weight of about 34,000. It has been suggested that a disulfide bond is involved in the association of the two subunits (Tonomura, 1973). In the procedures for isolation
36
KLAUS HOFMANN AND REINER HAMM
of tropomyosin, a reducing environment is usually provided in order to prevent oxidation. Therefore, it is not yet clear whether the tropomyosin is present in the myofibril as the SS-lined dimere or as the SH monomeres. In samples of rabbit tropomyosin from different animals, Woods (1968) found, in the absence of SH reagents, molecular weights between 40,000 and 85,700. After treatment with mercaptoethanol, molecular weights around 34,000 were obtained. Therefore, in some preparations, SS bridges seem to occur, while, in other preparations, they do not. d . Troponin. From the 5 moles SH/105 gm troponin, 1.8-2.5 do not react with NEM (Ebashi et al., 1968). In Table V the halfcystine (SH SS) figures of the three troponin subunits TN-I, TN-T, and TN-C are listed. TN-I inhibits ATPase activity of myosin, TN-T binds to tropomyosin, and TN-C binds Ca2+. In the troponin, these subunits are supposed to be present in the molar proportion 1:I : I . If this proportion of the three troponins is correct, the halfcystine figures obtained by Greaser et ul. (1973) (Table V) should result in 3.9 moles SH + SSlmole troponin. Drabikowski and Nowak (1970), however, found 8.1 moles SH + SSlmole. The latter figure seems to be more probable because it corresponds to the sum of SH and SS groups obtained by others (Table V).
+
2 . Sulfhydryl Groups Involved in the Function of Myofibrillar Proteins Several hundred publications deal with the role of SH groups of myofibrillar proteins in enzymatic activities, in the interaction with ions, substrates, and other proteins, and in the process of muscular contraction and relaxation. It is impossible to quote all these papers in this brief review. But in addition to the more recent research work, research papers and review articles will be cited which mediate the access to all the other literature. a . Myosin. SH groups are important for the adenosine triphosphatase (ATPase) activity of myosin because SH reagents modify this enzyme in various ways (Bendall, 1969; Young, 1969; Ebashi and Nonomura, 1973; Seidel and Gergely, 1973; Tonomura, 1973; Taylor, 1972). Fifty percent of the ATPase activity is lost when five groups of 42 SH groups per molecule are blocked, and inactivation is complete if seven per molecule are blocked. These groups are located in the head region of the molecule, probably fairly close to the active site. It is known that there are two kinds of particularly reactive cysteine residues per subunit of myosin, generally referred to as S, and S2 (or SH1 and SH2). These groups are located at or near the active site of the HMM moiety and near the binding site for ATP. Blocking of the S, sulfhydryls results in an increase in the Ca2+- activated
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
37
ATPase activity with a simultaneous loss in the K+ (EDTA) activated ATPase. Blocking of both sulfhydryls, S, and S2. eliminates both types of ATPases. If the S2 groups alone are modified, the enzyme activity is similar to that of the &-modified myosin; namely, Ca-activated ATPase is activated and K+ (EDTA) activity is lost. The requirement of both sulfhydryls for the inhibition of myosin ATPase in the presence of MgATP has also been demonstrated (Seidel, 1972; Reisler et al., 1974a). Measurements of the binding of nucleoside di- and triphosphates by native, S, or S,-blocked myosin revealed that all three forms bind ATP equally well. Thus, although SH blocking does not alter the binding of a substrate. it does interrupt the catalytic process. In the resting muscle, myosin contains bound MgATP which decreases its reactivity with NEM, an effect which is not reproduced by adenosine diphosphate or any other nucleotide (Barany and Barany, 1973). It has been suggested that maximum inhibition and maximum activation of ATP hydrolysis in vivo occurs in the millimolar range of Mg2+ through the formation and disruption of a cyclic MgATP ternary complex with myosin, involving coordination with the S, and S2 sites. This idea could be confirmed by crosslinking the two essential SH groups with a bifunctional dimaleimide reagent (Reisler et al., 1974b). The results of Petuskova (1973) also indicate a paired arrangement of these SH groups. Petaskova suggested that the role of SH groups in myosin probably consists of maintaining the structural integrity of the molecule and not in the direct participation in the hydrolysis of ATP. The S, groups react more rapidly with NEM or PCMB than the S2 groups. There is only one S , and S, group per heavy chain of myosin. The S, and S2 groups are generally presumed to be in subfragment 1 of the HMM. Phenolic SH reagents were found to activate Ca2+-ATPase and to inhibit K+(EDTA)-ATPase of both HMM and subfragment I , which make up the head of the myosin molecule. The S,-blocking of subfragment 1 proceeds much faster than that of HMM. ATP-induced conformational changes around the active site of myosin and HMM which were caused by the modification of the S2 group were preserved in subfragment 1. Therefore, the effect of phenolic SH reagents on myosin ATPase and that of ATP on the conformation around S2 cannot be interpreted in terms of subunit-subunit interaction (Kameyama et al., 1974). The reaction of myosin SH groups with DTNB liberates a single class of light chains of 18,000 daltons from HMM without a significant loss of ATPase activity. The other light chains. however, cannot be removed without a reduction in such activity (Weeds and Frank, 1973). Kakol (1971) suggested that the SH groups of the light chains of myosin are essential in preserving ATPase activity. The sequences reported for peptides containing the S, and S2 groups, however, are different from the sequences around the SH groups of the light chains (Taylor, 1972). Myosin contains about 2 moles of “intrinsic” Ca2+which cannot be removed
38
KLAUS HOFMANN A N D REINER HAMM
by the usual purification methods but which can be irreversibly removed by treatment with PCMB followed by P-mercaptoethanol. The latter reagent completely removes the PCMB from the myosin so that ATPase activity is completely recovered. The ATPase activity of actomyosin reconstituted from this myosin and actin is very different from that containing Ca2+ (Tonomura, 1973). In addition to ATPase activity, the ability of myosin to combine with actin is essential for the process of muscular contraction. SH groups are involved in this interaction between myosin and actin (Needham, 1973; Tonomura, 1973). Since myosin masked with certain SH reagents can bind with actin, even though its ATPase activity has been completely eliminated, the active sites for two functions must be different. Other results, however, are not consistent with this idea. Out of the 15 SH groups contained in 200,000 gm myosin, two are necessary for ATPase activity, while three are essential for actin combination; one of these groups is supposed to take part in both phenomena. The effect of F-actin on S,-blocked HMM (with NEM) as shown by Kameyama and Sekine (1973) suggests that the S2 region may be involved in the successive and cyclic conformation changes in the contractile protein system thought to occur in the sliding process of filaments. Other experiments have revealed an involvement of the S, group in the actin binding of myosin, because the binding of actin to myosin causes a conformational change of myosin by influencing the S, groups (Seidel, 1973). Therefore, it is not surprising that, at physiological ionic strength, the binding of actin to myosin has the same effect as the activation of myosin ATPase by the modification of the S, group (Burke et al., 1974). Treatment of actomyosin in the absence of salt, i.e., in the gel state, revealed that actin protects the S, group from reaction with NEM. That the Mg-activated ATPase also remains unaffected implies that the same sulfhydryls are necessary for the functioning of the Mg-activated reaction when the ionic forces involved in the overall interactions are in effect. In the presence of salt and Ca2+,however, treatment with increasing NEM concentrations produces stepwise inactivation of both Ca- and Mg-activated ATPases. The protein unit of one myosin and two actins may react first with about 7-8 NEM molecules without effect (Bkiny and Merrifield, 1973; Schaub and Watterson, 1973). 6 . Acfin. Sulfhydryl groups are also involved in the functional properties of actin (Drabikowski and Bitny-Szlachto, 1964). It was shown that mercurials inhibit the polymerization of G-actin to F-actin and also cause the release of bound nucleotides. By the use of various SH reagents, it was possible to separate these processes and to show that the actin molecule contains at least three kinds of SH groups: The first, which apparently is not connected with any specific property of actin, reacts directly with NEM; the second kind is involved in the polymerization of this protein; and the third is more or less directly connected with the nucleotide binding. It was suggested that two SH groups per mole are required for binding ATP.
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
39
The different effects obtained by blocking the same SH groups of G-actin with various SH reagents on polymerization, however, lead to two questions: Can different SH groups be assigned to polymerization and nucleotide binding sites and are the SH reagents specific (Taylor, 1972). From spin labeling experiments, it has been concluded that, in addition to tyrosine, lysine, and histidine, sulfhydryls are involved in the configurational changes during polymerization of actin. Laki and Alving (1973) found that one SH group of the G-actin moiety of F-actin participates in the dephosphorylation of ATP by actin. Modification of actin by treatment with SH reagents (PCMB, iodoacetamide, NEM) did not interfere with actin-myosin combination. Modification of HMM with PCMB, however, inhibited combination. These results of Heazlitt et al. (1973) confirmed the observation of Kuschinsky and Turba (195 I ) that while SH groups from myosin are necessary for the symplex formation of actin and myosin, those from actin are not.
c. “Natural Actomyosin, Troponin, Tropomyosin. “Natural actomyosin,” extracted from muscle fibers or myofibrils, is a complex of the proteins actin, myosin, tropomyosin, troponin, and some minor protein components. This complex represents the contractile system of the myofibril. A mixture of tropomyosin and troponin (“natural tropomyosin”), when added to a purified system of actin and myosin, confers upon the ATPase activity of the latter system the extreme sensitivity against Ca2+ ions which is a marked feature of the intact fiber and fibrillar preparations. Pure tropomyosin, even with SH groups carefully protected against oxidation, fails to do so (Bendall, 1969). SH groups are involved in the Ca-sensitizing effect of the tropomyosintroponin system on myosin (Daniel and Hartshorne, 1972). SH reagents can remove the Ca-sensitive response of “natural actomyosin. This effect has been shown both by the measurement of ATPase activity and by superprecipitation. In either case the inhibitory effect of the troponin-tropomyosin complex in the absence of Ca2+ is blocked. A similar effect has also been shown with muscle fiber in that the reaction with NEM caused tension development under Ca2+-free conditions which normally favored relaxation (Kuriyama et al., I97 1). Troponin was originally considered as a likely site of the critical SH group, and experiments with PCMB suggested that this was the case. However, it has been shown that PCMB can be transferred between different muscle proteins; one therefore cannot be certain of this hypothesis. From labeling experiments with NEM it was concluded that the critical SH site might be on the myosin molecule, and PCMB or NEM treatment of troponin did not produce a marked effect on either its Ca-sensitizing activity or Ca-binding activity (Ebashi and Nonomura, 1973). Hartshorne and Daniel (1970) also came to the conclusion that SH groups of troponin are not essential for its function. So the postulation of Fuchs (1971) that Ca-sensitive SH groups exist at a site of troponin, which is essential for its regulatory function, may be incorrect. I’
”
40
KLAUS HOFMANN A N D REINER H A M M
The integrity of certain SH sites on myosin seems to be essential for the normal Ca-sensitizing effect of “natural actomyosin.” Daniel and Hartshorne (1972) demonstrated that the SH groups, which are essential for Ca sensitivity of the normal “natural actomyosin,” are located in the heavy chains of the myosin molecule and that the critical SH groups are not identical with the S, sulfhydryl groups of myosin. Seidel and Gergely ( 1 973) also concluded from spin labeling experiments that the SH groups, whose blocking has been shown to abolish Ca-sensitivity, are not the S, groups. SH groups appear to play no essential role in the attachment of troponintropomyosin to myosin. As to tropomyosin, carboxymethylation of the SH groups reduced the inhibitory effect of this protein on the Ca-stimulated ATPase of desensitized actomyosin but did not effect the Mg-stimulated ATPase (Cummins and Perry, 1973). The supposition that SH groups of tropomyosin may be involved in the polymerization of this protein could nor be confirmed by Drabikowski and Nowak (1 965). d. Actinins. The substitution of about half of the SH groups of myosin by 2-aminoethyl isothiuronium makes the resulting actomyosin unresponsive for a-actinin (no activation of ATPase activity or of the turbidity response by a-actinin); the same substitution of actin does not affect the response to a-actinin. Either substitution diminishes the binding of c-w-actinin to actomyosin; neither substitution abolishes contractility measured in terms of gel syneresis (Seraydarian et al., 1968). PCMB did not affect the shortening in length of F-actin particles by p-actinin (Maruyama, 1971).
e . Muscle Fibers. In the presence of Ca and Mg ions, ATP is split rapidly by the actomyosin filaments. The free energy from this process is used for contraction and for the development of power through the mediation of the sliding of actin and myosin filaments over one another. By modifying the enzyme sites on the actin and myosin filaments (e.g., by the addition of the SH reagent salyrgan to a fiber which develops tension in the presence of ATP, Mg’+ and Ca2+), the ATPase activity and the tension immediately drop to zero and the fiber releases. This effect of salyrgan can be reversed by the addition of cysteine, restoring the SH groups on the enzyme center once more and allowing ATP to split and tension to be redeveloped (Bendall, 1969).
B. PROTEINS OF THE SARCOPLASMIC RETICULUM (SR) Several different protein SH groups are present in the SR. Some of them are clearly involved in the ATPase activity and the Ca-accumulation function of this
SULFHYDRYL A N D DISULFIDE GROUPS IN MEATS
41
system. It is believed that seven proteins are components of the Ca-transport system of SR, such as ATPase, proteolipid, calsequestrin, 54,000 dalton protein, and "acidic proteins." The proteolipid contains 24 halfcystine equivalents per lo5 gni protein. the 54,000 dalton protein 7.4/105 gm (detected as cysteic acid) (MacLennan p t al., 1973). Hasselbach and Seraydarian (1966) demonstrated that lo5 gm of the vesicular protein contained seven equivalents. of which three reacted readily with NEM without the impairment of Ca transport or of ATPase activity. Loss of the extra ATP splitting associated with Ca uptake, as well as loss of Ca transport and storage. followed blockage of the other four SH equivalents. The SH groups are located on the outer surface of the transporting membranes. According to Panet and Selinger (1 970). ten SH equivalents per 1 O5 gm membrane protein of the SR are titratable by DTNB in the absence of sodium dodecyl sulfate (SDS); however, in the presence of SDS, 14 SH groups were found. ATP protected SH groups essential for the ATPase activity of the SR. The asymmetric distribution of proteins in the SR membranes, containing 10-12 mole SH groups per lo5 gm protein. became symmetric if more than four SH groups were blocked by 2-chloromercuri-4-dinitrophenol: At this point, the Ca-dependent ATPase of the SR was completely inhibited, but Mg-dependent ATPase was slightly activated and Ca transport was inhibited (Dupont and Hasselbach, 1973). PCMB and p-chloromercuribenzene sulfonic acid (CMBS) increased the rate of Ca efflux from the whole frog muscle. While PCMB appeared to inhibit SH groups in the terminal cysternae of SR (causing a fractionating of the muscle twitch), CMBS seemed to act primarily at the surface sites with limited access to the cysternae (Kirsten and Kuperman. 19704. The muscle showed increased rigor tension irz v i m when incubated with 1 mM NEM, and Ca efflux from the whole muscle was increased. NEM apparently produces rigor by inhibition of Ca uptake through the SR (Kirsten and Kuperman, 1970b). So the effect of NEM o n muscle physiology seems to be detemiined not only by its reaction with myosin SH groups as mentioned above but also by the reaction with the SH groups of SR. C.
PROTEINS OF THE SARCOLEMMA
Modification of the SH groups of the sarcolenima with DTNB did not affect the Ca-dependent ATPase, but it decreased the Mg-dependent ATPase of these membranes. When the reagent was added in the absence of ATP, both enzymes were inhibited whether the divalent cations were present or not. Cysteine or dithiothreitol reversed this enzyme inhibition. Modification of the sarcolemma SH groups by NEM strongly inhibited the activity of both ATPases in the presence of ATP, and fully inhibited them in the absence of ATP. This inhibition was
42
KLAUS HOFMANN A N D REINER HAMM
not reversed by cysteine or dithiothreitol (Gimmelreikh and Koval, 1973). The activity of the Na+/K+-stimulatedATPase of the sarcolemma is inhibited by such SH reagents as PCMB or NEM (Matsushima, 1974). D.
PROTEINS OF MITOCHONDRIA
A substantial number of SH groups are present in the mitochondrial membranes (Jocelyn, 1972). Protein SH groups are involved in the electron transport. The flavoprotein and cytochrome components of the electron transport chain possess SH groups. The SH groups of NADH dehydrogenase appear to be mainly structural and not to have a catalytic function in accepting electrons. The SH groups of succinic dehydrogenase may be both structural and catalytic. The transfer of electrons from NADH dehydrogenase to coenzyme Q may require SH groups; the further transfer to cytochrome c does not, although the enzyme complex concerned contains several SH groups. The cytochrome itself contains SH and SS groups. Cytochrome oxydase contains some SH groups, but it is not known whether they are required for activity. Various observations have implicated SH groups in the yet unresolved mechanism of oxidative phosphorylation. Oxidative phosphorylation is inhibited not only by NEM and DTNB but also by arsenicals, suggesting that SH groups are required. The passive membrane transport of inorganic phosphate by a special mechanism, unlike the other mitochondrial exchange systems, is inhibited by SH-combining agents such as DTNB, mercurials, or NEM. As to the active transport, Ca2+ accumulation is partly inhibited by SH reagents such as mercurials, but the Na+/K+-exchange is stimulated for unknown reasons. The effect of mercurials on the ATPase of mitochondria varies from a stimulation at a low concentration to an inhibition at a higher concentration. E.
PROTEINS OF THE SARCOPLASMIC MATRIX
The supernatant, which is obtained after centrifugation of a muscle homogenate at about 100,000 X g, contains a great number of dissolved albumins and globulins. Not all of these contain SH groups. So it is known that no SH groups are present in the myoglobin from the mammalian skeletal muscles. SH groups are, however, involved in the activity of enzymes of the glycolytic chain. Glycolysis is inhibited by the SH reagent iodoacetate at a concentration which does not affect contraction (Jocelyn, 1972). Some of the SH dependent enzymes have been isolated. They include phosphofructokinase, glyceraldehyde-3phosphate dehydrogenase, lactate dehydrogenase, and phosphorylase. In addi-
43
SULFHYDRYL A N D DISULFIDE GROUPS I N MEATS
tion to these glycolytic enzymes, creatine phosphotransferase requires SH groups for its activity.
IV. SH AND SS CONTENT OF MEATS AND MEAT FRACTIONS The SS content of tissues is generally rather low. Therefore, most investigations have dealt with the estimation of the SH content only. Much confusion exists in literature concerning the term “SH content in tissue,” because there are different kinds of SH groups which are not always clearly differentiated from each other. Ellman (1959). who introduced DTNB for the assay of SH groups in biological material, entitled his publication, “Tissue Sulfhydryl Groups.” However, only the SH content in the tissue extracts was estimated; no similar estimate was made for whole tissues. Other authors (e.g., Khan) have used the term “SH groups in meat” quite generally, although only the nonprotein SH content of extracts was estimated. Therefore, it is absolutely necessary to explain what type of SH groups were estimated in each case. The total SH content of meat is distributed on protein and nonprotein substances as well. The different possible fractions are shown in the scheme of Table VIII. Boyne and Ellman (1972) used the term “total soluble SH” to refer to the sum of the SH content in the soluble proteins and in the nonprotein fraction. In the
TABLE VIII SCHEME OF THE DISTRIBUTION OF SH GROUPS IN MUSCLE TISSUE
Water insoluble proteins (e.g.. myofibrils)
e
-reactive SH slowly reacting SH
‘masked SH
,reactive
Water soluble proteins (e.g., sarcoplasmic proteins)
k
\
.Protein SH
SH
- total slowly reacting SH
masked SH cysteine SH
Nonprotein glutathione S H
SH
44
KLAUS HOFMANN A N D REINER HAMM
following tables, the total SH content of meat is generally given. In some cases, one cannot be sure if all SH groups or only a part of them were estimated by the reagent used. Therefore, the kind of reagent-whenever indicated in the literature-has been listed as well. Very different dimensions are used in literature for expressing the SH content in meat. Therefore, the original SH data were transformed to milligrams of SH/100 gm meat and/or to moles of SH/1OS gm protein for better comparison. Values given in parentheses correspond only approximately to the original data, because in these cases the protein contents were not indicated and the mean protein content of lean meat, which is about 18%, was taken as a basis for the calculation. In the following scheme, the factors used for the transformation of the original data are given.
[
mgSH 100 grn meat
X
1'
3.024
percent protein ~
[
1-
mole~H lo5 gm protein
3.024 x
[
mgSH
1
gm protein
097
The question of how the SH content in whole meat can best be expressed has no easy answer. It is not completely correct to relate the total SH content in whole meat to the protein content because a part of the SH groups in meat is not bound to protein. On the other hand, when the whole tissue (as wet weight) is used as a reference standard, the SH values are influenced by differences in concentrations of water, fat, and connective tissues (which are virtually free of SH groups), and the variations in the SH content of different meats can be attributed to changes in these factors. The answer to the question of what reference unit is to be employed depends on the kind and the aim of the investigations involved. There are cases in which statements can be contradictory depending on whether the results refer to meat or to protein. For instance, the SH content in cow and bull muscle was found to be 87.3 and 79.3 mg SH/l00 gm tissue, respectively, or 11.8 and 12.8 moles SH/105 gm protein, respectively. In the first case, the conclusion would be that the SH content in cow muscle was higher than in bull muscle; in the second case, it would seem to be lower. The reason for this apparent contradiction is the different protein content of both meats. This exam-
45
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
ple clearly shows that conclusions and comparisons of SH results must be drawn very carefully. There are many factors that influence the SH content of raw meat. It is not surprising, therefore, to find a considerable variation in the results of different literature in this field. These influences are discussed separately in Section IV,
D. A.
SH AND SS CONTENT OF MUSCLES
Values are listed for total SH content in pork and beef (Tables 1X and X), for muscles of various meat and test animals (Table XI), for nonprotein SH contents in muscles (Table XII) and for SS groups in muscles (Table XIII). Comrnenrs on Table IX: The SH figures in porcine muscles estimated with different SH reagents vary from 8 . 3 to 12.9 moles SH/I05 gm protein (49.6 to
TABLE IX SH GROUPS 1N PORCINE MUSCLES
Muscle
long. dorsib long. dorsi long. dorsi" long. dorsi" long. dorsi long. dorsi long. dorsi long. dorsi semimembr .'' semimembr. semimembr. psoas
SH reagent
&NO,
AgNO, DTNB NEM" NEMJ DTNB NEM' PNSS.*
SH (mg) per 100 gm tissue
Original SH data" 1 .SO pmolesll00 mg meat 1 I .6 moles/105 g prot. (2) 73.6 mg/lOO gm meat (3) 74.0 mg/100 gin meat (3) 3.14 mg/gm protein (4) 8.48 moles/lP gni protein 12.89 moles/105 gni protein 2.89 mglgm protein (4) 3.44 mg/gm protein (4) 1 1 .OX moles/105 gm protein 3.05 mg/gm protein (4) 283 pg/gm tissue (5)
(I)
(2) (2)
(2)
49.6 71 .O 73.6 74.0 (56.5) 50.5 76.7 (52.0) (61.9) 64.5 (54.9) 28.3
SH (moles) per lo5 gm protein (8.3) 11.6 11.2 11.3 9.5 8.5 12.9 8.7 10.4 11.1
9.2 (4.8)
Numbers in parentheses correspond to the following investigations: ( I ) Krylova and Kusnezowa (1964); (2) Hofmann. Bliichel. Miiller. Baudisch, and Hiinim (unpublished observations): (3) Hofmann CI ul. (1974): (4) Fischer and Hamm (1975); (5) Motoc and Banu (1968). 'I
M . /ongis.simits dorsi. Vacuum packaged after slaughter. " Air packaged after slaughter. In presence of dodecylsulfate. Reaction at pH 7.4. 'I M . semimembr-uriuc.eus. 'I Sodium nitroprusside. I'
46
KLAUS HOFMANN AND REINER HAMM
76.7 mg SHll00 gm tissue). The value of 4.8 moles SH, which was estimated with sodium nitroprusside, obviously does not represent the entire SH content. The average value is 10.2 moles SH/105 gm protein (62.2 mg SH/lOO gm tissue) when the low value is excluded. Figures estimated with AgNO, vary from 8.3 to 11.6, those estimated with NEM under different conditions from 8.7 to 12.9 and those estimated with Ellman’s reagent (DTNB) from 8.5 to 11.1 moles SH/105 gm protein. This shows that the values found with AgNO, are in the same range as the values found with NEM (at pH 7.4) and DTNB. No significant differences seem to exist between the SH contents of longissimus dorsi and semimembranaceus muscles. Comments on Table X : The values lower than 8 moles SH/105 gm protein do not seem to represent the normal total SH content in bovine muscle (even in some cases when AgNO, is used as a reagent). It is well known that sodium nitroprusside and NEM at pH 6.0 react only with one part of all SH groups in proteins (see Section 11, B,3 and 4). From the other values it can be concluded that the values for beef muscles vary from 8.5 to 12.1 moles SH/105gm protein (50.5-87.3 mg SH/l00 gm tissue); the mean value is 10.5 (65.9 respectively). These figures are virtually in the same range as those for pork. Therefore, contrary to the statement of Krylova and Kusnezowa (1 964), the SH content of pork is not generally higher than the SH content of beef. In addition there are no considerable differences in SH content between cow and bull muscles, between longissirnus dorsi and supra spinam muscle, or between vacuum-packed and air-packed muscles shortly after slaughter. The number of SH groups in calf muscles varying between 8.4 and 10.4 moles SH/1O5 gm protein (52.7-75.6 mg SH/lOO grn tissues) seems to be somewhat lower than those in beef. However, Krylova and Kusnezowa (1964) found that beef from animals at age 9-10 months gave higher SH values than that from animals at age 16-18 months. The differences in the SH content of longissimus dorsi and diaphragm muscles of calves were found to be significant (P < 0.01) (Fischer, Hofmann, and Hamm, unpublished observations). Comments on Table X I : All SH values estimated with PCMB, iodoacetate, and sodium nitroprusside (3.2-6.7 moles) are lower than those found with Ellman’s reagent or AgN03 (9.0-1 1.9 moles). The latter, which probably represent the total SH content in muscles, are within the range of the values estimated for pork and beef (see Tables IX and X). Therefore, the skeletal muscles of different animals (pigs, cattle, chickens, rats) contain approximately equal amounts of SH groups. The SH values of rabbit, rat, and frog muscles estimated with iodoacetate were also found to be comparable. The figure of 21.4 moles SH found by Caldwell and Lineweaver (1969) is certainly too high and is therefore not included in this discussion. Comments on Table XII: The content of nonprotein SH [respectively of glutathione (GSH)] in muscles varies in a wide range from 1.1 to 7.6 mg SH/100 gm meat (average 3.7). This may be due to the fact that the GSH content of
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
47
TABLE X SH GROUPS IN BOVINE MUSCLES
Sample. Muscle Beef, long. dorsi" Beef, long. dorsi" Beef, long. dorsi" Beef, ('?) muscle Beef, long. dorsi" Beef, long. dorsi' Beef, long. dorsi Beef, long. dorsi Beef, long. dorsi Beef, long. dorsi Beef, long. dorsi Beef, long. dorsi Beef, supra spinam Beef, psoas Bull, long. dorsi Cow, long. dorsi Calf, long. dorsi Calf, semimembr.' Calf. diaphragma Calf, long. dorsi Calf, semimembr. Calf, diaphragrna
SH reagent
Original SH data"
SH (mg) per 100 gm tissue
SH (moles) per 105 gm protein
8.8 moles SH/105 gm protein ( I ) 1.36 pmolesl100 mg meat (2) 1.03 pmolesll00 mg meat (2) 1.1 . moleslmg meat (3) 69.0 mg/IOO gm meat (4) 68.0 mgl100 gm meat (4) 11.25 moles/l05 gm protein (5) 61 pg/IOO mg tissue (6) 3.1-3.6 moles/105 gm protein (6) 106.0-1 15.8 pmoles/gm protein (7) 84.9-94.0 pmoles/gm protein (7) 9.11 moleslgm protein (5) 79.8 mgllOO gm meat (8) 273 pglgm tissue (9) 79.3 mgllOO gm meat (5) 87.3 mgllOO gm meat ( 5 ) 75.61 mg/IOO gm meat (10) 69.55 mg/lOO gm meat (10) 56.59 mgl100 gm meat (10) 9.08 moles/105gm protein (10) 8.99 moles/l05 gm protein (10) 8.42 molesil05 gm protein (10)
(52.4) 45.0 34.1 36.7 69.0 68.0 (67.0) 61.0 17.0-19.2 (63.1 -68.9) (50.5-55.9) 54.2 79.8 27.3 79.3 87.3 75.6 69.6 56.6 66.1 67.1 52.7
8.8 (7.6) (5.7) (6.2) 11.2 10.9 11.3 10.3 3.1-3.6 10.6-1 1.6 8.5-9.4 9.11 ( I 1.20) (4.54) 12.1 11.8 10.4 9.3 9.1 9.1 9.0 8.4
Numbers in parentheses correspond to the following investigations: ( I ) Hamm and Hofmann (1965); (2) Krylova and Kusnezowa (1964); (3) Bolshakov and Mitrofanov ( I 970); (4) Hofmann e r al. (1974): ( 5 ) Hofmann. Bliichel, Muller, Baudisch, and Hamm (unpublished observations); (6) Hamm and Hofmann (1966b); (7) Dzinleski er al. (1969); (8) Bognar (1971a); (9) Motoc and Banu (1968); (10) Fischer, Hofmann, and Hamm (unpublished observations). 'I M. longissimus dorsi . From 9-10-month-old animals. ' From 16-18-month-old animals. After slaughter. vacuum packaged. After slaughter, air packaged. " Reaction at pH 6.8. Reaction at pH 6.0; the meat samples were minced in different ways. i Sodium nitroprusside. j M . semimembrunaceus. ' Reaction at pH 7.4. 'I
('
'
TABLE XI SH GROUPS IN MUSCLES OF VARIOUS ANIMALS, INCLUDING TEST ANIMALS
Sample. Muscle"
SH reagent
Chicken, bright m. Chicken, dark m. Chicken, breast Chicken, breast Chicken, breast Chicken, pect. major Turkey, breast Turkey, breast Rabbit, unknown Halibut, whole m. Rat, gastrocn . Rat, unknown Frog, unknown
AgN03 AgNO, DTNB DTNB PCMB PrUSS.C DTNB AgN4 iodoac ." iodoac . DTNB DTNB iodoac.
Original SH data"
10.1 x moledmg protein (1) 11.1 x moleslmg protein ( 1 ) 0.4144.564 pmole SH/mg N (8) 1.34pmoles SH/mg N (9) 4.45moles x 10-8/mg protein (2) moles x IO-Vgm tissue (3) 5.8d-9.9r 0.491mmole/gm muscle (4) 2.73mmoles X 10-4/gm muscle (4) 0.79% cysteine in protein (5) 0.81 B cysteine in protein (5) 21.5pmoles/gm tissue (6) 1.63 mmoles/100 gm tissue (7) 0.724% cysteine in protein (5)
SH (mg) per 100 gm tissue
(60.1) (66.1) 39.3-53.7 (26.5) 19.2-32.7 1624' 0.W (38.8) (39.8) 71.1 53.9 (35.5)
per
SH (moles) lo5gm protein
10.1 11.1
6.6-9.0 21.4 4.5 3.2-5.5 6.5 6.7 (11.9) (9.0) 6.0
~~~~~
Abbreviations: m. = muscle; pect. = M . pectoralis; gastrocn. = M. gasrrocnemius. investigations: (1) Bolshakov and Mitrofanov (1970); (2)Bolshakov er af. (1972); (3)Chajuss and Spencer (1962b); (4) Bowers (1972); (5) Mirsky (1936);(6)Boyne and Ellman (1972);(7)Sedlak and Lindsay (1968);(8) Miller and Spencer (1975);(9)Caidwell and Lineweaver ( 1969). Sodium nitroprusside. " 0 hours postmortem. " 72 hours postmortem. The original data are obviously not correct. iodoacetate.
* Numbers in parentheses correspond to the following
'
49
SULFHYDRYL AND DlSULFlDE GROUPS IN MLATS TABLE XI1 NONPROTEIN SH A N D GSH CONTENT OF SKELETAL MUSCLES 01. MEAT A N D TEST ANIMALS
Sample
SH reagent
Pork Beef Beef Lamb Chicken" Chicken" Chic ken Rat Rat
ASNOS AgNOz DTNB HgC12 Prusside' Prusside Prusside DTNB DTNB
Original data"
2.6 mg SH/100 gm meat (1) 1.9 mg SHil00 gm meat ( I ) 3.94 mg GSH/gm protein (2) 15 rng CySH/100 gm tissue (3) 238" pg GSH/gm muscle (4) 395 eg GSH/gm muscle (4) 0.33"'pmoIc SH/gin muscle (5) 0.07 mmole SH/100 gni tissue (6) I .96 pmole SH/gm tissue (7)
SH (mg) per 100 gm meat
2.6 1.9 (7.6) (4.1)
2.6 4.3 1.1
2.3 6.5
Numbers in parentheses correspond to the following investigations: ( 1 ) Hofmann ef ul. (1974); ( 2 ) Dub6 ef ul. (1972); (3) Oganessjan and Dschani-
I'
bekova (1958); (4) Khan and van den Berg (1965); (5) Khan and Nakamura (1971); (6) Sedlak and Lindsay (1968); (7) Boyne and Ellman (1972). " Breast muscle. Sodium nitroprusside. " These data obviously represent nonprotein SH, although not characterized as such. " Leg muscle. This value was taken from a graph I'
'
animal tissues is influenced by several factors (see Section IV, D). Nonprotein SH related to total SH content of muscles (about 65 mg/100 gm tissue) varies from 2 to 12 %. Oganesjan and Dschanibekova (1958) found 12-16% nonprotein SH. Using the corresponding data given in Tables IX-XI, the following values for the nonprotein SH content of skeletal muscles were calculated: pork, 3%. beef, 4% (Hofmann et al., 1974), and rat, 4% (Sedlak and Lindsay, 1968). Hornsey (1959) found in the filtrate of the heat-cleared "emulsions" of pork leg muscle in water only 0 . 6 4 . 7 mg SH/100 gm meat. These values are low in comparison to those given in Table XII. The nonprotein SH groups were probably partly oxidized to SS groups during the heat treatment. The nonprotein SS was estimated to be 0.8-1.1 mg SS/l00 gm meat, and the sum of the SH and SS values in meat ( I .6-1.7 mg/lOO gm) corresponds quite reasonably t o some of the SH values listed in Table XII. Comnients on Table X I U : The SS content in muscles found by several investigators varies from 6.1 to 22.9 mg SS/loO gm meat (0.5-2.0 moles SS/105 gm protein). This significant variation may be due to the different degrees of oxidation of the SH groups. On the other hand, the SS content in muscle tissue can
50
KLAUS HOFMANN AND REINER HAMM TABLE XI11 SS GROUPS" IN MUSCLE
Sample, Muscle
Original data"
SS (mg) per 100 gm meat
SS (moles) per lo5 gm protein
Pork, long. dorsi" Pork, semimembr." Pork, long. dorsi Pork. long. dorsi Beef, long. dorsi Beef, long. dorsi Beef, supra spinam Calf, long. dorsi Calf, semimembr. Calf, diaphragm
I .27 mg SS/gm protein (1) 0.95 mg SS/gm protein (1) 6. I mg SS/IOO gm meat (2) 0.83 mg SS/gm protein (3) 1 . 1 1 mg SS/gm protein (3) 7.4 mg SS/lOO gm meat (2) I .23 mg SS/gm protein (4) 0.78 mg SSlgm protein ( 5 ) 0.67 mg SS/gm protein ( 5 ) 1.19 mg SS/gm protein ( 5 )
(22.9) (17. I ) 6 .I 14.9 20.0 7.4 (22. I ) (14.0) (12.1) (2 1.4)
2.0 I .5 (0.5) I .3 1.7 (0.6) I .9 I .2 I .o I .9
" For the assay of the SS values listed, NaBH, was used to reduce SS to SH and AgNO, was used to titrate the SH groups formed. " Numbers in parentheses correspond to the following investigators: ( I ) Fischer and Hamm (1975); (2) Hofmann, Baudisch. and Hamm (unpublished observations); (3) Hofmann, Muller. and Hamm (unpublished observations); (4) Bognar (1971a); ( 5 ) Fischer. Hofmann, and Hamm (unpublished observations). '' M. longissirnus dursi. M. semimembruniireus.
''
depend on the age of the animals tested. For example Oeriu (1962, 1964) reported that the aging process in animals (guinea pig, dog, rabbit, rat) leads to an increase in the SS content in blood and in several tissues. However, the variation in the SS content in beef and veal is so great that a similar tendency cannot be recognized. It should be emphasized that in any case the SS content is only a small proportion of the cysteine plus cystine content in muscles (see Table XIV).
B . CYSTEINE PLUS CYSTINE CONTENT OF MUSCLES Cysteine is the only amino acid in proteins which contains the SH group. Therefore, the cystine content can be estimated by determination of the SH content. As cystine can be reduced (e.g., with sodium borhydride according to the reaction CySSCy 2 H + 2 CySH), the content of cysteine plus cystine (total cystine) can be determined after the reduction of the protein. Table XIV shows the values for the (total) cyctine content of muscles. These estimations were carried out either by means of amino acid analysis after hydrolysis or by using the assay of SH (SH + SS/2) after reducing the SS groups. The factors for the transformation of the different terms are given in the following schema:
+
51
SULFHYDRYL AND DISULFIDE CROUPS I N MEATS mg SH red. [gm protein]
x 0 363 -+
1
gm cystine 0 120x mole (SH + SSI2) gm protein] + 105 gm protein
[
TABLE XIV CYSTEINE PLUS CYSTINE CONTENT (CALCULATED AS CYSTINE) OF MUSCLES
Sample
Assay
Beef Beef Beef Beef Beef Beef" Beef" Pork Pork Pork Lamb
Amino acid Amino acid Amino acid SH + SSIZ" SH + SSl2" SH + SS/2b SH + SW2" SH + SSI2" SH + SSI2" Amino acid Amino acid
Original data'' 1.0-1.3 gm cystindl6 gm N (1) 1.2-1.5 gm cystine/l6 gm N (2) 1.35 % protein (3) 4.93 mg SHlgm protein (4) 12.1 moles SH/IOs gm protein (5) 10.9-12.1 moles SH/105 gm protein ( 5 ) 10.7-10.8 moles SH/IOs gm protein (5) 12.4 moles SH/105 gm protein ( 5 ) 4.424.45 mg SH red.'/gm protein (6) 1.31 70protein ( 3 ) I .35 9% protein (3)
Cystine (gm) per 100 gm protein
I .o-I .3 I .2-1.5 1.35 I .79 1.45 1.31 - I .46 I .29-1.30 1.49 1.60-1.62 1.31 1.35
Numbers in parentheses correspond to the following investigators: ( I ) Bigwood (1960, cited in Bigwood. 1972); (2) Greenwood et ul. (1951); (3) Schweigert and Payne (1956); (4)Bognar (1971a); (5) Hofmann, Miiller, and Hamm (unpublished observations); (6) Fischer and Hamm (1975). Using AgN03. I ' Freeze-dried meat. " Using NEM. 1.03 x SS [mg]. SH red. = SH
+
The sum of SH and reduced SS is called SH red (calculation see footnotee, Table XIV). The values of the total cystine content in muscles estimated by amino acid analysis (Table XIV) vary from 1 .O to 1.35 gm/l00 gm protein, and the values obtained by determination of the SH groups vary from 1.29 to 1.79. The higher values for SH groups in the nonhydrolyzed protein may be explained by the fact that during hydrolysis cystine is partly destroyed (see Section 11, C ) , and in this respect the results obtained with myofibrils may be interesting. After reduction of the myofibrils with NaBH4 at different temperatures (20"-70°C), 4.26-4.43 mg SH/gm protein was determined, corresponding to I .55-1.60 gm total cystine/lO gm protein (Hofmann, 1964). On the other hand, the amino acid analysis after acid hydrolysis of the same material with 6 N HCl led to a value of 1.2 gm cystine/l00 gm protein, again demonstrating the detrimental effect of protein hydrolysis. Consequently, the most probable mean value of the cysteine plus
52
KLAUS HOFMANN AND REINER HAMM
cystine content of beef, pork, and lamb muscle seems to be 1.5 gm cystinef100 gm protein. This value corresponds to 12.5 moles SH/105 gm protein or 4.1 mg SH/gm protein. C. SH CONTENT OF ORGANS There is a great confusion in the literature in presenting SH values. In the earlier papers, very often the symbols pg, 7 , and u were used which should mean gm. However, the corresponding values would be too low by a factor of 103 in comparison to the values given in the more recent literature or to all the values for muscle tissue (Table IX-XI). Therefore the mentioned symbols were interpreted as gm. The total SH groups content in liver is listed in Table XV, and in several other inner organs in Table XVI. The nonprotein SH content in these organs are listed in Table XVII. The following comments are provided for these tables. Comments on Table XV: The SH content of the liver of pig, rabbit, mouse, and rat varies from 23.4 to 98.6 mg SH/lOO gm tissue. It should be mentioned that the average SH content of liver (47.2 mg SH) is lower than that of muscles (65 mg SH). This is somewhat surprising, but liver is rich only in nonprotein SH TABLE XV ESTIMATES OF SH GROUPS IN LIVER OF MEAT A N D TEST ANIMALS
Sample
SH reagent
Original SH data","
Rabbit Rabbit Mouse Mouse Mouse Mouse Rat Rat Rat Rat Rat
Ferricyanide CMPAN' FCMB PCMB Ferricyanide CMPAN
46.5 ?/I00 mg (I) 45.2 ? / I 0 0 mg ( I ) 23.4 ull00 mg (2) 30-35 $100 mg (3) 31.3-52.9-y/lOO mg (1) 35.6-59.4y/IOO mg (1) 55.4 mg/i00 gm (4) 2.03 mmole/100 gm (5) 29.8fimoleslgm (6) 35.648.4?/I00 gm ( I ) 36.7-53.8y/IOO gm (I)
DTNB DTNB Ferricyanide CMPAN
SH (mg) per 100 gm tissue
46.5 45.2 23.4 30-35 31.3-52.9 35.6-59.4 55.4 67.0 98.6 35.6-48.4 36.7-53.8
Numbers in parentheses correspond to the following investigators: ( I ) Flesch and Kun (1950);(2)Calcutt and Doxey (1959);(3) Calcutt (1961);(4)Bhattacharya (1959);(5)Sedlak and Lindsay (1968);(6)Boyne and Ellman (1972). * Related to tissue wet weight. However, some of the given dimensions are obviously not correct (see preliminary remark to Section 1V.C). CMPAN = I -(4-chloroniercuriphenylazo)-naphthol-2. a
53
SULFHYDRYL AND DlSULFlDE GROUPS I N MEATS TABLE XVI ESTIMATES OF SH GROUPS IN INNER ORGANS" OF TEST ANIMALS ~~
Sample, Organ
SH reagent
Original SH data".'
Rat. heat Rat. heart Rat. heart Rat, h e m Mouse. heart Mouse. heart Rat. kidney Rat, kidney Rat, kidney Mouse. kidney Mouse. kidney Mouse. kidney Rat. brain Rat. brain Rat, brain Mouse. brain Mouse, brain
DTNB Ferricyan ide CMPANd Ferricyanide CMPAN &NO, Ferric yanide CMPAN Ferricyanidc CMPAN PCMB &NO, Ferricyanide CMPAN Ferricyanide CMPAN
31.2 nig/100 grn (I) 1.54 mmoles/lOo grn (2) 20.1 y/lW mg (3) 48.9y/100 rng (3) 17.2 yll00 mg (3) 41.9 y/100 rng (3) 48.3 rng/100 gm (I) 25.2 $100 mg (3) 56.4y/IOO rng (3) 20.4y/lOO mg (3) 54.3 y/l00 rng (3) 9.9 v/lOO rng (4) 4.9prnoleslgrn ( 5 ) 15.6y/lOO mg (3) 59.2y/lOO mg (3) 14.0y/100 mg (3) 52.7 y/lW rng (3)
~
SH (nig) per 100 gni tissue 31.2 50.9 20.I 48.9
17.2 41.9 48.3 25.2 56.4 20.4 54.3 9.9 16.2 15.6 59.2 14.0 52.7
For liver. see Table X V . correspond to the following investigators: ( I ) Bhattacharya (1959);(2) Sedlak and Lindsay (1968);(3) Flesch and Kun (1950);(4)Calcutt and Doxey ( 1959);(5)Gabay et ul. ( 1 968). See Note b to Table XV. CMPAN = 1 -(4-chloromercuriphenylazo)-naphthol-2.
* Numbers in parentheses
compounds (see comments on Table XVlI). The SH contents in liver of the different animals, as far as they were cstimated by the same author, do not diffcr considerably. Commmfs on Table X V I : The SH values listed in Table XVI are very different from each other due to the use of different SH reagents. The following ranges were found for rat and mouse: heart 17.2-50.9, kidney 9.9-56.4, and brain 14.0-59.2 mg SH/100 gm tissue. It is obvious that the SH content of these inner organs is generally lower than that of muscles. In fact, even the maximum value for organs is lower than the average value for muscles (65 mg SH/lOO gm tissue). Comments on Table XVII: The highest nonprotein SH contents for all organs were found in liver (average value 22.7 mg SH/lOO gm tissue) and the lowest in heart, stomach, diaphragm, and lung (4.9-8.6 mg SH/100 gm tissue). The values for kidney lie in between. It is noteworthy that the nonprotein SH content of the organs of trained animals is generally higher than that of untrained ani-
54
KLAUS HOFMANN AND REINER HAMM TABLE XVII NONPROTEIN SH AND GSH CONTENT IN ORGANS OF TEST ANIMALS
SH (mg) per ~
Animal -
Pig Rabbit Mouse Rat Rat Rat Rat Rabbit Rat Rat Rat Rat Dogd Dog' Rat Rat Dogd Dog' Dogd Dog' Dogd Dogp Dogd Dog"
Organ
SH reagent
Liver Liver Liver Liver Liver Liver Liver Liverb Kidney Kidney Spleen Brain Brain Brain Heart Heart Heart Heart Stomach Stomach Diaphragm Diaphragm Lung Lung
PCMB PCMB PCMB PCMB AgNO 3 DTNB DTNB DTNB &NO, PCMB PCMB DTNB lodine Iodine DTNB Iodine Iodine Iodine Iodine Iodine Iodine Iodine Iodine
Original data" 21.4 pg/IOO mg ( I ) 25.2 pg/lOO mg ( I ) 32.4 pg/IOO mg ( I ) 26.4 pg/lOO mg (1) 26.4 mg SH/100 gm tissue (2) 0.10 mmole SH/100 gm tissue (3) 6.00 pmoles SH/gm tissue (4) 8.1 pmolesc/gm tissue ( 5 ) 1 I .3 mg SH/100 grn tissue (2) 18.2 pgll00 mg ( I ) 9.6 pg/l00 mg (1) 2.6 pmoles SH/gm tissue (6) 45.6 mg % GSH (7) 56.6 mg % GSH (7) 0.22 mmol SH/lOO gm tissue (3) 7.2 mg SH/100 gm tissue (2) 54.0 mg % GSH (7) 69.3 mg % GSH (7) 49.5 mg % GSH (7) 53.9 mg % GSH (7) 56.8 mg % GSH (7) 65.1 mg % GSH (7) 4.57 mg % GSH (7) 55.2 mg % GSH (7)
100 gm tissue
21.4 25.2 32.4 26.4 26.4 3.3 19.8 26.8 11.3 18.2 9.6 8.6 4.9 6. I 7.3 7.2 5.8 7.5 5.3 5.8 6. I 7.0 4.9 5.9
Concerning the dimensions of the original datas see Note b to Table XV. Numbers in parentheses correspond to the following investigators: ( I ) Calcutt and Doxey (1962); (2) Bhattacharya (1959); (3) Sedlak and Lindsay (1968); (4) Boyne and Ellman (1972);( 5 ) Ellman (1959);(6) Gabay et al. (1968);(7) Wachholder and Uhlenbrook (1935). * Extracted with 5% TCA. The values for water and alcohol extracts were much lower. The original dimension (mmoles) is obviously not correct. Untrained. ' Trained.
mals. Since the nonprotein SH content in liver, kidney, and heart is markedly higher than the average nonprotein SH content in muscles, the relation of nonprotein SH to total SH is higher in these organs than in muscles. Bhattacharya (1959) showed that rat liver contains 48% and that both kidney and heart contain 23% nonprotein SH (related to the total SH content). By summarizing the SH contents given in Tables IX-XVII, we can conclude that in most cases the total SH as well as the nonprotein SH contents of the same
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
55
organs of different species vary less than the SH contents of different organs of the same species.
D. FACTORS INFLUENCING THE SH CONTENT OF RAW MEAT In this section, some natural factors that can possibly influence the SH content of meat will be discussed. The influences of processing, freezing, and other treatments are discussed separately. 1. Postmortem Aging
Chajuss and Spencer (l962b) reported a rapid decrease in the SH group content of chicken muscle during the development of rigor mortis. Gawronski et al. ( 1 967) also observed a decrease in the SH content and in the ratio of SH to SS in excised chicken breast muscle during the first hours postmortem. After Pavlovskij and Grigoreva (1966) and Golovkin and Korzhemanova (l973), the number of SH groups in muscle decreases during rigor and then increases again during relaxation. Motoc and Banu (1968) reported that the SH content of muscles would decrease due to a denaturation of the myosin. On the other hand, Kolodziejczyk (1965) found an increase of free SH groups in pork and beef during aging. These apparent contradictions may be due to differences in the availability of the SH groups based on the use of different types of reagents in the methods of measurement. (More work has to be done in order to elucidate the role of SH groups in the presence of rigor mortis and aging.) In order to find out whether the preparation of tissue was responsible for this variation in results, Miller and Spencer (1975) investigated aged samples of chicken muscle prepared in the different ways described by the other investigators. Analysis of variance showed that the SH concentration did not change significantly with aging time within any of the homogenate preparation methods. Caldwell and Lineweaver (1969) have found that no change in the SH content of chicken breast muscle occurs during postmortem aging. Furthermore, determination of the SH content with AgNO, has indicated that the postmortem storage at +2"C of beef and pork for up to 1 1 days does not change the SH content of muscle tissue (Hofmann and Schael, 1966; Hofmann et al., 1969). Further studies with different skeletal muscles of pork and beef from numerous animals have confirmed this result (Hofmann, 1971d). Finally, Hay et al. (1972, 1973) found that the aging of chicken muscle fibrils was not accompanied by a decrease in the total SH content; this indicates that there has been no oxidation of SH groups throughout the aging process. It should be mentioned that significant changes in the SH content were never found until at least 5 days after preparation in numerous cases when the SH content of myofibrils of pork and beef muscles was controlled during cold storage (Hofmann and Hamm, unpublished observations).
56
K L A U S H O F M A N N A N D REINER H A M M
2. Other Factors a. Meat Quality. Fischer and Hamm (1975) investigated the influence of the quality of pork on the number of SH groups which react with NEM (easily available SH groups) and with AgNO, (total SH groups content). The so-called pH, value, 45 minutes postmortem, and the water-holding capacity of the meat were measured as criteria of meat quality. The number of easily available SH groups decreased significantly with increasing PSE (pale, soft, exudative) conditions, i.e., with an increasing rate of glycolysis postmortem, but no correlation between total SH or SS groups content of the tissue and PSE properties was observed. In accordance with these results, Usunov and Zolova (1976) reported that rapidly reacting SH groups were drastically reduced in PSE muscle. It seems possible that the decrease in these easily available SH groups is attributable to the masking of the SH groups of myofibrillar proteins by sarcoplasmatic proteins, precipitated at low pH values on the myofibrils (Fischer and Hamm, 1975). Bendall and Wismer-Pedersen ( I 962) give a similar explanation for the finding that in PSE muscle a smaller amount of charged groups were titratable than in normal muscle.
b. Variation in rhr GSH Content. The content of SH glutathion (GSH) in muscles and organs is not constant but is, in vivo, depending on numerous metabolic factors which have been comprehensively discussed in the review of Santavy (1965). Therefore, it may be sufficient to summarize only a few of the most important influences on the GSH content and, consequently, the total SH content of tissues. GSH activates several enzyme systems which participate in the metabolism of carbohydrates. It is also important for aerobic glycolysis and is a cofactor of some dehydrogenases. The ATPase activity of myosin in the muscle is increased by GSH. Furthermore, GSH may play an important role in maintaining the activity or in the reactivation of several SH enzymes. Thus it is not surprising that the GSH level in muscles and organs is increased by training. During stress the GSH content in muscles decreases (more with untrained animals than with trained animals). The GSH level in muscle and liver is influenced by the vitamins C and BP,which are involved in the carbohydrate metabolism. c. fnfluencr ofAge. With the increasing age of animals, an accumulation of disulfide groups in tissues has been observed (Oeriu, 1962; Harisch and Schole, 1974). Lastovskaya (1969) reported that the SH content of rat liver decreased slightly during the first 12 months of age and significantly during 1-2 years of age. However, Harisch and Schole (1974) found that the GSH content of rat liver increased continuously with age, causing a rise in SH and SS groups as well.
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
57
The reactivity and the SH content of myogen A of rats were reported to increase with age, especially in young animals. This protein showed the highest enzyme activity in older rats (Goldshtein and Khilko. 1969). A significant age associated decline in the SH content of serum albumin has been demonstrated in both men and rats (Leto er a l . , 1970). I t seems that the effect of age on the SH and SS content depends on the kind of organs and proteins investigated as well. Some results concerning beef are mentioned in Section IV,A. An increased synthesis of GSH has been shown to be associated with growth hormones (Shacter and Law, 1956). SH groups are involved in cell division and play an important role in carcinogenesis (for review, see Harington, 1967). d. Deboned Meat. The production of volatilc sulfur compounds during storage in mechanically deboned poultry meat has been investigated by O’Palka (1973). In this study he found that cysteine destruction was closely correlated with methyl mercaptane production. This was an unexpected result, because Cascy et al. (1965) and Grill et u f . (1967) have shown methionine to be a precursor of methylmercaptane. The formation of volatile S-containing compounds (I6 were estimated) during storage of deboned meat is probably due to bacterial activities, because many types of microorganism are present in poultry products as a result of contamination (Kraft, 1971).
e . Unsaturated Fatty Acids. Robinson (1966) reported that SH groups are able to react with unsaturated fatty acids, the SH group probably being added to the double bond. Furthermore, SH groups can be oxidized by fatty peroxides, which may be formed during the storage of meat. f. T r u c ~ of ~ sHeavy Metals. Finally the possibility of traces of heavy metals (which are absorbed by meat animals from the environment) decreasing the SH content of meat will be examined. This question may be considered for lead, one of the widest spread trace metals, which can react with SH groups: 2 R - SH
+ Pb’+
-2H’
Pb (RS),
The maximum Pb levels in beef, veal. and pork were reported (Holm, 1976) to be 0.53 I , 0.372. and 0.158 ppm, respectively. The highest values for beef and calf liver were 0.40 and 0.26 ppm Pb, respectively. As a comparison with the SH content of meat shows, these Pb levels are very low. For instance. 0.5 ppni Pb is equal to 0.05 mg Pb in 100 gm meat, which contains about 65 mg SH. Consequently, such Pb2+ levels cannot bind more than 0.024% of the SH groups in muscle (calculated in moles). Therefore, traces of lead which can occur in meats are neglegible in respect to the determination of SH groups. The same may be valid for traces of other heavy metals. An exception is the presence of Cu2+ ions,
58
KLAUS HOFMANN A N D REINER HAMM
which are known to catalyze the autoxidation of SH groups, and which, therefore, can have a strong influence on the results of SH determination.
V. INFLUENCE OF PROCESSING ON THE SH AND SS GROUPS OF MEAT A . INFLUENCE OF HEATING Heating of meat is accompanied by changes in appearance, smell, taste, texture, and nutritive value. The cysteine and cystine moieties of proteins are particularly involved in these alterations. Therefore, the changes in and possible reactions of SH and SS groups in meat proteins during heating must be considered. The thermal formation of volatile sulfur compounds, which are important flavor components, will also be discussed.
I.
Effect of Heat Denaturation on the Availability of SH Groups
Biochemists define denaturation as a change in the specific steric conformation of a protein, i.e., a change in the secondary and tertiary structure without a chemical modification of the amino acids (Fasold and Turba, 1959). Thus, denaturation is a physical process, not a chemical one. One has to be careful in using the term “denaturation” instead of “heating,” because stronger heating results not only in denaturation but also in chemical modifications of the proteins such as reactions of the functional groups and the cleavage of covalent linkages. According to the definition of denaturation mentioned above, the oxidation of SH to SS groups and the reduction of SS to SH groups should not be called “denaturation” because these reactions are chemical modifications; this is also true for the inactivation of SH enzymes caused by auto-oxidation. Therefore, the oxidation of SH-groups by heating will be discussed in the next chapter. Dry heating of myofibrils for 30 minutes at temperatures from 30” to 70°C under both nitrogen and air resulted in an irreversible increase in the SH groups reacting with NEM at pH 6.0, as shown by Fig. 4 (Hamm and Hofmann, 1965). Using the same method for measuring the SH-groups, Schrott (1974) also observed a substantial increase in the available SH groups in actomyosin (from 1.41 to 2.41 mg SH/gm protein) when it was heated to 65°C. The SH increase was explained by the fact that NEM reacts only with a part of the SH groups in the native meat protein and that heating causes an unfolding of the protein molecule during heat denaturation (Haurowitz, 1950), making more SH groups available for NEM (Hamm and Hofmann, 1965). The release of reactive protein SH groups, which were hidden within the native folded protein
SULFHYDRYL A N D DISULFIDE GROUPS I N MEATS
59
FIG. 4. Effect of heating (30 minutes) myofibrils on the amount of SH groups reacting with NEM at pH 6.0 (Hamm and Hofmann, 1965).
structure by denaturation. was first discussed in the case of ovalbumine (Anson, 1945). An increase in the heating temperature to 120°C did not cause a further increase in the available SH groups, even though only about 70% of the SH groups are available to NEM after heating to 70°C. The total number of SH groups is obtained by reaction with AgNO, (see Fig. 4 and Table XVIII). As Fig. 4 shows, oxygen has no remarkable effect on the number of SH groups available for NEM in raw or mildly heated meat. The role of SH groups in the temperature-induced denaturation of muscle proteins was investigated by Jacobsen and Henderson ( 1 973). With actomyosin they found, after heating the protein to 60°C. a marked irreversible increase in the sulfhydryl groups which were titratable with PCMB. These authors did not give any explanation for this increase; but there is no doubt that it may also be due to an unfolding of the protein molecules induced by heating. TABLE XVIII SH LEVELS IN MYOFIBRILS AFTER 30 MINUTES HEATING"."
SH reagent
30°C
50°C
70°C
AgNOa DTNB NEM
8.5-9.0
8.4-9.0
8.5-9.0
6.04.5 3.5
6.5 5.0
8.5 7.0
"
In moles SH/105 gm protein
* After Tinbergen (1970).
60
KLAUS HOFMANN A N D REINER H A M M
Tinbergen (1970) studied the influence of heating myofibrils (from 30" to 70°C) on the number of SH groups reacting with different SH reagents. His results are in good agreement with those obtained by Hamm and Hofmann (1965); they demonstrate that the number of SH groups which react easily with NEM and DTNB increases with increasing temperature, whereas the total number of SH groups (reacting with AgNO,) remains unchanged (see Table XVIII). The heating of whole meat in a superhigh frequency electromagnetic field to an internal temperature of 65°C also led to an increase in the number of SH groups (Malyutin, 1969). Frying meat decreased the total SH content but increased the number of easily reacting SH groups titratable with CH3HgN03. using nitroprusside as an indicator (Dworschak, 1969). Increases in the available SH groups were also found after the mild heating of pork (Randall and Bratzler, 1970), turkey breast muscle (Bowers, 1972), and frog sartorius muscle (Kovaleva, 1967). Contrary to the results reported here, Dub6 (1969) found a decrease in the number of available SH groups using Ellman's reagent after cooking myofibrillar extracts of beef at temperatures between 60' and 90°C. He suggested that this disagreement might be due to the fact that in these experiments the proteins were already denatured by urea, and thus heating did not allow any further conformational change that would have made more SH groups of the molecules available to the reagent. However, not only was there no increase, but rather a decrease in the SH groups (about 6% at 60°C and 16% at 70"), which was explained by the assumption that the SH groups could be oxidized into SS groups. In the case of heating the meat to 70"C, this explanation does not agree with the results of Hofmann (1964) (see Fig. 5), Samejima et al. (1969), Tinbergen (1 970) and Bognar (1 97 I a), who showed that there is no decrease in the total SH content in meat proteins after heating to that temperature. Samejima ef at. (1969) found that neither disulfide reducing nor SH-blocking reagents prevented the heat coagulation of myosin. This supports the conclusion of Hamm and Hofmann (1965) that the heat coagulation of myofibrillar proteins is not due to an oxidation of SH to SS groups but to an intermolecular association of other sidechains of the protein molecules. However, we have to concede that the protein SH groups in different meats and myofibrils may not always react in the same way. It is well known that traces of certain heavy metals are able to catalyze the autoxidation of SH groups. Hence, differing contents of such trace elements in meat could result in an oxidation of SH groups to a different extent (for further discussion of the previous results see next section). Malyutin ( 1 969) found a decrease in the SH groups in meat after heating to 85"C, suggesting an aggregation of proteins. SH groups have indeed been shown to be involved in this heat aggregation of proteins (Connell, 1960a; Jaenicke, 1965a,b). The reaction
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
61
mechanism is suggested to be a SH-SS exchange according to the following equation (Lumper and Zahn, 1965):
HS-Prot/i
f
' S
HS-Prot
__
SH HS-Prot-SS-Prot /
However, as this reaction demonstrates, the SH-SS exchange would not lead to a change in the number of SH and SS groups. The decrease of SH groups established after heating does not necessarily indicate protein aggregation caused by SH-SS exchange. But it may be that the SH groups formed are less reactive than those originally present. No investigations have been attempted as yet to test this hypothesis. It should be pointed out that denaturation alone is not able to increase the number of S S groups. The formation of SS groups is, in any case, the result of an oxidation of SH groups. Neither can the reverse reaction, namely, the formation of SH from SS, be caused by denaturation itself alone, as has been postulated previously (Grau, 1968). It may be concluded that heat denaturation of meat proteins leads to an increase of available SH groups, depending on the temperature applied. There obviously exists a wide temperature range rather than one special temperature in which SH groups, previously masked in the native state, are exposed. This finding is in accordance with denaturation therrnoprofiles of beef muscle tissue which show that molecular changes occur between 49" and 94°C (Karmas and De Marco, 1970). The total amount of SH groups in meat with AgNO, is not effected by heating up to 70°C (Hofmann, 1964; Hamm and Hofmann, 1965; Kovaleva, 1967;
Lob 5 0 5 0 7 0 9 0
1X)OC
FIG. S. Effect of heating myofibrils (30 minutes) on the amount of SH groups reacting with AgNO, at pH 7.4 (Hofmann, 1964).
62
KLAUS HOFMANN AND REINER HAMM
Tinbergen, 1970; Bowers, 1972), whereas at higher temperatures the SH content falls (Fig. 5). The curve in Fig. 5 is quite similar to the temperature-SH curve for actomyosin presented by Hamm and Hofmann (1965; in Fig. 2 of the publication the word “myofibrils” should be corrected to read “actomyosin”). The decrease of SH at higher temperatures is due to oxidation and will, therefore, be discussed in the next section.
2 . Influence of Heating on the Total SH and SS Content Dube (1969) observed that the SH content of myofibrillar extracts of beef decreased on cooking at temperatures up to 90°C. The SH values of the heated and then reduced samples obtained after treatment with NaBH, were lower than the SH values of the raw samples. It is not clear whether the reduction of the disulfide groups was incomplete or if other oxidation products which cannot be reduced by NaBH, were formed during heating. Krylova and Kusnezowa (1 964) heated pork and beef for 30 minutes at 75” to 80°C. Using an amperometric titration with AgNO,, they found a decrease in the SH groups ranging from 27 to 29%. Figure 5 also shows that heating to temperatures higher than 70°C decreases the number of SH groups in muscle proteins. In the presence of air, the SH content dropped more than when exposed to pure nitrogen. At 120”C, the decrease under air was 40%, under nitrogen 25%;at the same time, the SS content rose by 36% and 20% respectively. These changes evidenced that the SH groups were mainly oxidized into SS groups (Hofmann, 1964; Hamm and Hofmann, 1965). Since the oxidation also occurred with the exclusion of atmospheric oxygen, it was suggested that oxidation in the presence of nitrogen is due to residual molecular oxygen included in or bound to the sample, from which it can hardly be removed. In experiments of Schweigertet al. (1949), the total cystine content of pork and lamb has proved stable during normal cooking, but not during heating at 120°C. As is evident from percentages mentioned above, there is a difference between the decrease in SH and the increase in SS (about 5%). which means a loss of the total cystine (CySH CySSCy) content. After heating the samples for 5 hours, the deficit reached 26%and 13% under air and nitrogen, respectively (Hamm and Hofmann, 1965). Bognar (1971a) found a loss of as much as 33% total cystine after the heating of beef for 1 hour at 120°C. He was able to show that this loss was due to the formation of cysteine acid. That the decrease stated by Bognar was more pronounced than the decrease in the case of myofibrils may be attributed to the different experimental conditions: The beef was heated in water in which the oxidation was probably more effective than in the dry-heated myofibrils. The formation of lanthionine, which is also discussed as a possible product of
+
SULFHYDRYL AND DlSULFlDE GROUPS IN MEATS
63
cystine destruction, seems to be unlikely under the conditions used because it occurs only in a basic medium (Hupf and Springer, 1971). Marchenko (1968) found that sterilization of beef and lamb at 120°C lowered the SH content by 62% and 50%, respectively, whereas the SS content increased for 72% and 62%, respectively. However, this hardly explains why in this case the increase in SS was higher than the decrease in SH. Bem rt al. ( I 970) and SusjC rt al. ( 1 974) extensively investigated the influence of the canning of meat which contained various amounts of nitrite, nitrate, and ascorbic acid. It was to be expected that nitrite affects the SH content because of the possible formation of nitrosothiols (Mirna and Hofmann, 1969). Canned meats with the highest addition of nitrite (0.35%) exhibited both the lowest SH content and highest SS content. Nitrate and ascorbic acid did not influence the SH content. The SH content of canned meats which were sterilized at 1 10”-115”C was much lower than that of meat pasteurized at 76°C. The total cystine content of the high-temperature canned meat products was also reduced; consequently, the nutritive value of the meat protein was lowered (see also Section V, A, 3). During the storage of both the sterilized and the pasteurized meats, the SH content decreased continuously, demonstrating that the reaction between the SH groups and nitrite advanced. Finally, a paper published by Khan and van den Berg (1965) in which the use of incorrect terms caused some misunderstanding will be briefly discussed. The authors talk about “sulfhydrylgroup content of muscle proteins,” but it is evident from the procedure described in a previous paper (Khan et a l . , 1963) that in all cases the SH group assay was carried out in an extract of muscle obtained by using 2.25% metaphosphoric acid. Such an extract. free of protein, contains only about 3% (see Section IV) of the total meat SH, mainly as GSH. Therefore, the statement of these authors that the cooking of chicken muscle decreased the SH content of “muscle protein” by about 50% is not correct. 3 . Influence clf the Thermal Destruction of Cysteine Plus Cystine (Total Cysrine) on the Nutririw Value of Meat Protein As was shown in the previous section. heating induces losses of total cystine when meat is heated to high temperatures. This has already been demonstrated in previous work by means of amino acid analysis (Beuk et ul., 1948; Donoso et ul., 1962). However. hydrolysis of the proteins causes losses of total cystine as well. It was later shown (Bognar, 1971a, b) that heating beef for I hour at 120°C caused an average loss of 28.3% methionine and 15.9% total cystine. The decrease in the other amino acids ranged from 5.0 to 8.9% (the amino acids of the meat broth were included). The relatively high losses of the sulfur amino acid content clearly demonstrates the sensitivity of these amino acids to wet heating.
64
KLAUS HOFMANN A N D REINER HAMM
Bjarnason and Carpenter (1970) studied the mechanism of heat damage in proteins using bovine plasma albumin as a model. The protein, which contained 14% moisture, lost 50% cystine when heated for 27 hours at I 15°C. Miller er al. (1965) found a similar loss (60%) for vacuum-dried cod after heating for 27 hours at 1 16°C. Methionine and cysteine plus cystine are involved in many essential functions of every living cell. This should be noted with respect to the fact that the deficiency of the sulfur-containing amino acids in human diets is a critical problem of worldwide importance (Allaway and Thompson, 1966). The nutritive value of meat protein is limited by its content of methionine and cysteine plus cystine (Donoso et a l . , 1962; Hofmann, 1966b). Therefore, any damage of total cystine diminishes irrevocably the nutritive value of meat protein (see Section I). In the heat treatment of meat during cooking and processing, the thermal sensitivity of the sulfur-containing amino acids must therefore be taken into consideration. Time and temperature of heating should not exceed certain limits (Hofmann 1966b, 1 9 7 2 ~ )There . is a formula which states that the time necessary for killing microorganisms can be shortened to a tenth when the temperature is increased by 10°C (Beuk et al., 1948). On the other hand, the rate of chemical reactions (including the damage of amino acids) is only increased about 3-fold by the same increase in temperature (van't Hoff's rule). Thus, heating the meat at a higher temperature for a shorter time (high-short heating) is preferable to heating it longer at a lower temperature. Moreover, in order to reduce the time necessary for sterilization, the cans should not be too large and should be as flat as possible. In the presence of fluid constituents (as in cans containing goulash, beef and pork in their own juices, sausages in brine, etc.), the time of heating can be shortened by rotating the cans (rotation sterilization; Rievel and Reuter, 1955; Heidtmann, 1966; Wirth, 1967; Christiansen, 1968). The quality of canned products can be improved by estimating and using the so-called F-values (Takacs ef al., 1969; Heidtmann, 1970). A loss of total cystine may be compensated by the addition of cysteine (N. N., 1970). It may also be of interest to note that it is possible to introduce SH groups into proteins and, therefore, to increase their SH content (Schoberl, 1948). Perhaps in this way the nutritive value of proteins may be improved. 4. Release of Hydrogen Sulfide During Heating
During the heating of meat numerous volatile compounds which contribute to the formation of meat flavor are split off (Hornstein et a l . , 1960; Brennan and Bernhard, 1964). Some of those volatiles are sulfur-containing compounds; the simplest and most investigated one is hydrogen sulfide. In low concentrations,
SULFHYDRYL A N D DlSULFlDE GROUPS IN MEATS
65
H,S is usually associated with high mean food acceptance scores (Olson rt d., 1959). For the determination of H,S which is released during heating, Hamm and Hofmann (1965; for details, see Hofmann, 1964, 1967) used a modified method of Marbach and Doty ( I 956). The H,S passes into a trap containing NaOH and is determined by photometric measurement (670 nm) of the blue color which develops after reaction with p-aminodimethylaniline and FeCI, in HCI. Optimum conditions for the formation of the dye and its measurement are realized by a moderate excess of FeCl, and by an HC1-concentration of I-3%. Under these conditions, the solution of the dye formed is stable (Hofmann and Hamm, 1967a). For estimation of H,S formed during the production of canned meat, test tubes, called Drager-Riihrchen, have proved useful (Bloeck Y t ul.. 1970). The odor threshold of this compound in water is as low as 10 ppb (Pippen and Mecchi. 1969). In low concentration, H,S has a favorable effect on the meat aroma and probably contributes to the flavor of all heated proteinaceous foods such as chicken, beef, fish, eggs, and milk. At higher concentrations. the objectionable odor of H,S is detrimental to the flavor (Johnson and Vickery, 1964). Fraczak and Pajdowski ( 1 955) first studied the development of H,S during the thermal processing of meat. Their results suggested that 80°C might be the temperature at which the formation of H2S in remarkable amounts begins. The aniount of H2S produced increases drastically with increasing temperature (Fraczak and Pajdowski, 1955; Parr and Levett, 1969). Hamni and Hofmann (1965) showed that the release of H,S induced by heating myofibrils increases exponentically with rising temperature (Fig. 6). The amount ofH,S formed by heating the myofibrils at 120°C for 30 minutes ranged from 16.8 to 18.7 p g H,S/gm protein. This is almost the same amount as was formed during the heating of total muscle tissue (18.4 to 19.3 pg). Therefore, at least 90% of H2S released during heating originates from myofibrillar proteins. This result agrees with the findings of Mecchi ef a / . (1964). With adipose tissue, the reverse was found; i.e., approximately 2% times as much H,S was evolved from the water soluble as from the water insoluble fractions (Pepper and Pearson, 1969). It was demonstrated that, after blocking the SH groups by Ag+ or NEM, myofibrils did not release detectable amounts of H,S during heating (Hamni and Hofmann. 1965).This evidenced that hydrogen sulfide originates from the protein SH groups rather than from disulfide groups or methionine. The same conclusion was reached by Parr and Levett (1969) in the case of chicken meat, whereas Fraczak and Pajdowski (1955) and Mecchi c’t a / . (1964) postulated that H2S might be split off from cysteine and cystine as well. An additional source of the development of H,S induced by the heating of meat is thiamin (Dwivedi and Arnold. 1971). The reported decrease of the SH content due to the formation H2S is relatively small
66
KLAUS HOFMANN AND REINER HAMM
0
70
90
110 120
O C
FIG. 6 Effect of heating (30 minutes) on the formation of H,S from inyofibrils (Hamm and Hofmann. 1965).
(2 to 3%). Therefore, it is not surprising that maximum H,S production at 120°C was reached only after more than 15 hours (Lendvai et al., 1973). The oxidation of SH groups during the heating of meat does not prevent the development of H,S, because enough residual SH groups are still present. Parr and Levett (1969) reported a disappearance of free H,S from freshly cooked meat left standing, which was mainly due to oxidation by atmospheric oxygen. After Sowa (1968), the amount of H,S does not increase significantly when the temperature is extended above 120°C. During the heating of meat with microwaves (27.4 mHz) the H,S formation was reduced (Sowa, 1968). Bloeck et al. (1970) studied the production of H2S during sterilization of fish (sardines). They found the same curve given in Fig. 6. The transformation of this curve into a half-logarithmic coordinate system resulted in a straight line, enabling the prediction of the amount of H,S formed, whereupon the calculation was brought about analogously to determine the F-value. It was found that using the “high-short heating” procedure limited the amount of H2S formed during sterilization (Bloeck er al., 1970). In general, the release of hydrogen sulfide during the sterilization of meat is a serious problem. Possible disadvantages are corrosion of the cans, discoloration of the cans (marbling) and of the content, and an unfavorable or even offensive smell when the can is opened. The marbling (see Andrae, 1969; Dahlke, 1969; Gruenwedel and Patnik, 1971) is due to the formation of iron sulfide or tin sulfide. In canned meats, which are only pasteurized, the activity of surviving microorganism may contribute to the H2S formation (Baumgartner and Baum, 1960). Cheftel (1958) established that metal ions such as Fe2+, Fe3+, A13+, Sn2+,and Sn4+,which may be exposed by corrosion effects, are able to accellerate catalytically the release of H,S. Furthermore, the pH value has a strong effect on the formation of H,S caused not only by heating but also by
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
67
bacterial action. High pH values support the release of H,S during the heating of meat (Johnson and Vickery, 1964; Krylova and Marchenko, 1969). Figure 7 shows the amount of H2S released from heated meats as influenced by the pH value. Irrespective of the type of meat used (mutton, beef, or pork), or of whether or not the pH was adjusted by artificial or natural means, all H2S values were located on the same curve, suggesting that the same mechanism was acting in each case. Muscles from animals in poor condition sometimes contain only small amounts of glycogen at the time of death; consequently, only a small amount of lactic acid arises postmortem, and, therefore, the ultimate pH of the meat is high. Such meat produces more H,S during heating than does the meat from normal animals. A decrease in carcass grade is also accompanied by an increase of both pH and H,S production (Johnson and Vickery, 1964). The addition of polyphosphate increased the concentration of H,S in canned broiler meat (Rao et al., 1975), a result which could not be attributed to the increase in pH caused by the addition of phosphate. Furthermore, it was found that the heating of meat with a high fat content produced significantly more H,S than the heating of lean meat (Kunsman and Riley, 1975).
PH
FIG. 7 . The effect of variation of pH on H,S content of volatilea produced by heating meat (Johnson and Vickery, 1964). (0. mutton; 0 , beef; W . pork.)
In the H,S assay, the hydrogen sulfide formed by heating of meat usually is transferred into the solution of the reagent. Marchenko and Kosenjasheva (1974) developed a procedure for the separate determination of the levels of hydrogen sulfide and mercaptans which are volatilized during meat cooking and those which remain in cooked meat. They found that, after heating meat to 80°C. the majority of the volatile sulfur compounds remained in the cooked meat. The hydrogen sulfide content in cooked pork was discovered to be 1.2-1.3
68
KLAUS HOFMANN AND REINER HAMM
times as much as that in cooked beef. This finding may be due to the fact that the pH value of the pork proved to be higher than that of the beef. The reaction mechanism of the H,S-formation is not yet fully understood. Schoberl (1941) attributed the development of H,S brought about by the influence of hot water on wool keratin to a hydrolytic splitting of disulfide bonds, which proceeds in two steps (Schoberl and Eck, 1936): R-CHZ-SS-CHZ-R + HZO -+ R-CHZ-SH + HOS-CHZ-R HOS-CHZ-R -,HZS + CHO-R
The same mechanism was discussed by Bjarnason and Carpenter (1970) in regard to the release of H2S during the heating of bovine plasma albumin. This reaction, however, needs an alkaline medium which is normally not present in meat. For the formation of hydrogen sulfide from SH groups, p-elimination (a) and hydrolysis (b) are possible reaction mechanisms. -NH,
-
-NH,
,,CH-CH2-SH
-co
-Cd
C=CH,
+
H,S
(a)
-NH,
,,CH-CH,-SH
-co
+HO -m, CH-W-OH 2 , -co
+
H,S
Fraczak and Pajdowski (1955) suggested a trimolecular reaction as an explanation for hydrogen sulfide formation; but trimolecular reactions are very rare; making this an unlikely possibility. In addition to hydrogen sulfide, small amounts of volatile mercaptans, thioethers, and other sulfur-containing conipounds are formed during heating and contribute to meat flavor. The best preventive way for reducing undesirable amounts of H2S and mercaptans which develop during sterilization is to select suitable meat which has been handled and stored under hygeinic conditions. Additionally thio-acceptors (e.g., zinc oxide) may be used as components of the inner lacquer of tins (Nehring, 1968; Bloeck ef al., 1970). Also the addition of weak acids such as lactic acid (Johnson and Vickery, 1964), ascorbic acid, or citric acid reduces the development of H,S (Hofmann, 1974a). The thermal formation of H,S and the oxidation of SH groups to SS groups in meat mentioned above necessarily result in a shift of the redox-potential, which is remarkably influenced by the SH/SS system (Hofmann, 1974b). The measurement of the redox-potential (Leistner and Wirth, 1965; Hofmann, 1974b) can be used to evaluate the quality of meat products. It is
SULFHYDRYL A N D DISULFIDE GROUPS IN MEATS
69
believed that each type of product has a characteristic redox potential range (Wirth and Leistner, 1970). 5 . Formation of Further S-Containing Flavor Components
Most of the typical flavor in meats is developed during heating. Raw meat. such as beef, pork, lamb or chicken, has little flavor (Crocker, 1948). The tlavnr of heated meat consists of a great number of different chemical components which are not present in raw meat but are developed from “precursors” by the influence of heat. None of the components identified in meat aroma has been described as uniquely “meaty” (Wasserman, 1972). The chemical pathways involved in the formation of meat flavor compounds during heating include Maillard browning reactions, fatty acid oxidation, and the formation of some low molecular volatile compounds, such as ammonia and hydrogen sulfide. But they also include inter- and intramolecular cyclization, as well as numerous mechanisms which are made possible by the reactivity of substances such as rnercaptans, hydrogen sulfide, ammonia. and other intermediates, especially at high temperatures (Wilson er a f . , 1973). Some of the first investigators who demonstrated the importance of several sulfur-containing volatiles for the flavor of meat were Yueh and Strong (1960) and Minor et al. (1965). The latter were able to show that the removal of the sulfur-containing components resulted in an almost total loss of meat flavor. The ability of fat to dissolve S-containing substances during cooking was demonstrated by Pippen et a / . (1969), who found that the fat of cooked poultry contains more sulfur than does the fat of raw poultry. Furthermore, it was shown that a reaction between H,S and acetaldehyde was involved, and that such reactions between H2S and carbonyls in fat could occur quite generally (Pippen and Mecchi, 1969). During the last few years, many authors have observed the formation of sulfur-containing compounds during the heating of meat which may be relevant to meat flavor (Kato et al., 1973; Mulders, 1973; Mussinan and Katz, 1973; Led1 and Severin, 1973; Scanlan et a / . , 1973; Schune, 1974; Garbusov et a / . , 1976; for reviews, see Herz and Chang, 1970; Schwimmer and Friedman, 1972; Wasserman, 1972). Boelens et a / . (1974) identified the reaction products between fatty aldehyds. H2S, thiols, and ammonia, which are normal constituents of meat, using a combination of gas chromatography and mass spectrometry. In addition. the organoleptic aspects of the reaction mixtures were discussed. Among the substances formed there were several heterocyclic S-compounds such as alkylated trithianes (I), oxadithianes (I]), dioxathianes (111), trithiolanes (IV), dithiazines (V). thiadiazines (VI), and some aliphatic thio compounds, particularly mercapto-thio-ethers (VII), dimercapto-thioethers (VIII), thiodisulfides (IX). thioaldehydes (X), and thioalkenes (XI and XII) (in the formulas the alkyl residues are omitted):
70
KLAUS HOFMANN A N D REINER HAMM
I
I
I
0
-CH-S I
-S-
S-
(IX)
-CH-CH,-CHO
I S-
4
-CH-CH=CH-SI
-CH=CH-CH
S-
(X)
(XI)
7\
S-
(XI0
Wilson et al. ( 1 973) identified forty-six sulfur-containing chemical compounds present in the volatiles of pressure-cooked beef. The main components were alkyl sulfides and alkyl disulfides, thiophenes, and sulfur-containing heterocycles (trithianes, trithiolanes, thiadiazines, and thiazoles). In addition to cysteine, cystine, and methionine, thiamine can also be a precursor of the volatile S-containing aroma components of meat: The formation of H,S (Dwivedi and Arnold, 1971), several thiazoles, and thiophenes (Arnold et al., 1969; Dwivedi and Arnold, 1972, 1973; Dwivedi et al., 1972, 1973) have been reported to be a result of the thermal degradation of thiamine. In spite of the fact that the content of carbohydrates in meat is very low, these substances may still be involved in the formation of the sulfur-containing aroma components of meat. Thus, Morton et al. (1960) found that the reactions of cysteine and other amino acids with sugar produced a flavor with a basic meat character. The reaction of cysteine with derivatives of hydroxy-dihydrofuranone (XIII) resulted in the formation of roasted meat flavor (van den Ouweland and Peer, 1975). Dihydrofuranones may be formed by the degradation of ribose-5-phosphate and have been found in a natural beef broth (Tonsbeek et al., 1968). The initial stage in the reaction of cysteine with the dihydrofuranones involves a substitution of the ring oxygen by sulfur giving derivatives of hydroxy-dihydrothiophenone (XIV). In this reaction, cysteine acts as an H,S donor. The odor components finally were identified as being derivatives of thiophene (XV), rnercapthothiophene (XVI), mercaptofurane (XVII), and mercaptodihydrofuranone (XVIII). Mulders (1973) heated mixtures of cysteine, cystine, and ribose for 24 hours
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
71
under reflux at 125°C and also identified volatile sulfur compounds such as thiophenes, thiazoles, trithione, alkylthiols, hydrogen sulfide, and carbon disulfide. Arroyo and Lillard (1970) heated mixtures of glucose and each of the sulfur-containing amino acids (methionine, cysteine, and cystine) for 2 hours at 98" C. However, none of these mixtures emitted an odor associated with meat flavor. On the other hand, a meat aroma for ready-to-eat meals could be produced by the reaction of monosaccharides with cysteine or cystine (N. N., 197 la). It may be concluded that the formation of meat flavor is not only due to Maillard-type reactions, the formation of ketones, aldehydes, amines, and heterocyclic N-compounds; it is also due to reactions in which sulfur is included in several ways. A number of patents exist for simulated meat aroma based on heating mixtures of a sulfur-containing compound, amino acids, and carbonyl compounds (Wasserman, 1972). A few patents have specified heating thiamine with various amino acids (Giacino, 1970; Yamamoto et u l . , 1970). It should be mentioned that volatile aroma components are involved not only in meat smell but also in meat taste, because the volatile compounds are, of course, soluble to some extent in meat juices and, especially, in the melted fats. This may be the reason for the fact that meat which contains a certain degree of fat gives a more aromatic taste after frying than does lean meat. It would be interesting to learn whether foreign metals (e.g., Pb, Hg, Zn, Cd) or other possible residues in meat, which are able to react with sulfur compounds, can influence the development of sulfur-containing aroma components during the heating of meat and thus influence meat flavor. As far as is known. this problem has not yet been investigated. 6 . The Texture of Meat as Influenced by Disuuide Groups Formed During Heating The ways in which meat texture is generally influenced by heating have already been discussed by Laakkonen (1973) in his review. One of the factors
72
KLAUS HOFMANN A N D REINER H A M M
that may influence the tenderness* of meat is the possible formation of disulfide bonds between protein chains induced by heating. Because this aspect is not considered in the review mentioned, it should be discussed in this section. Dub6 (1969) described the change in the texture of meat due to heat as follows: “It has been observed that upon heating meat develops a kind of rigidity that, in some ways, may compare with the resistance due to rigor. This hardening of the muscle tissue during heating is a reaction in which many substructural elements may be involved such as the proteins of the myofibrils with their chemically reactive groups. These elements may during the process contribute to the formation of different kinds of cross-bindings that might tighten the structure and increase the resistance to shear.” There is very little known about the chemical nature of the binding forces between actin and myosin during contraction. Szent-Gyorgyi (1966) did not support the assumption that the sulfhydryldisulfide bonds could be involved in the reaction between the two contractile proteins. [But SH groups of myosin are necessary for the myosin-actin interaction (see Section 111, A , 2, a).] Since the linkage can be broken by pyrophosphate and magnesium, Szent-Gyorgi suggested that there is probably an electrostatic interaction. It also seems that the analogy between muscle contraction and the hardening of muscle during heating is a superficial one. Therefore, the possibility of the tenderness of meat being influenced by the heat-induced formation of disulfide linkages should be considered independently of the process of muscular contraction. There are findings that demonstrate that the formation of disulfide groups does influence the texture of meat. Dodge and Stadelman (1959) observed that carcasses aged in air were less tender than those aged in water because of the greater opportunity for oxidation in the air. Treatment of muscle with potassium iodate, which is able to oxidize SH to SS, resulted in an increase of the shear values. The same result was found by Hird and Yates (1961) using several oxidizing agents. Chajuss and Spencer (1962a,b) exposed chicken muscles to a solution of sodium sulfite which is known to split SS bonds. After cooking these pretreated muscles, the shear values were found to be less than those of muscles stored in water only. However, treatment with hydrogen sulfite, surprisingly, did not significantly influence the shear values compared to treatment with sulfite (Chajuss and Spencer, 1962a). This finding signals caution in our interpretation of results obtained using sulfite. Because the authors did not use any buffer for preparing the solution, the solution of sodium sulfite reacted alkaline as a result of hydrolysis. Thus, this observation might be attributed to a swelling effect of the high pH of the solution. There is almost certainly a hardening effect of the oxidizing agents however. Such an effect has been suggested by Connell (1957). He emphasized the importance of cross-links between the protein peptide chains in *In this connection it may be of interest to note that the toughness of meat may be decreased by the injection of cysteine into the blood vessels of an animal 15 to 30 minutes before slaughtering. This treatment causes an activation of the proteolytic enzymes (N. N.. 1971b).
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
73
relation to texture and demonstrated that the SH content of dried fish is less than that of fresh products, assuming the fornmaticin of disulfide cross-links. In addition. Dube (1969) found that the decreasc in SH and the increase in SS in beef muscle proteins induced by heating were accompanied by an increase in the shear values of the muscle. Model experiments concerning the texture of SS-crosslinked gelatin gels were carried out by Okamoto et ctl. (1973). The relation between sensory properties, physical characteristics. and the forces maintaining gel structure were studied for gelatin thiolated t o 6.6 moles SS/I05gm protein. It was found that hardness was affected mainly by temperature, brittleness by disulfide bond content. The hardness was generally attributed t o hydrogen bonding, and the brittleness was attributed largely to disulfide bonding. Transferred to meat, this result would mean that an increase in the number of disulfide groups does not render the meat tougher but more crisp; in other words, the niasticatability of the meat would be influenced favorably rather than unfavorably by the formation of disulfide groups during heating. Indeed, it was observed that heating meat paste for a few minutes led to a product of gum-like consistency, whereas heating for a longer period of time so that the number of disulfide groups was drastically increased, resulted in the paste’s becoming increasingly crisper and more brittle (Hofmann and Hamm, unpublished observations). With whole meat, of course. the connective tissue also plays an important role for the tenderness or toughness. The significance of thiol and disulfide groups in the determination of the rheological properties of dough has been recognized by several workers (Frater et al.. 1960; Blocksma, 1972; Ewart, 1972; Jones et d.,1974). It may be that the thiol-disulfide system is also relevant to the rheological behavior of meat emulsions in the production of sausages. but no corresponding experiments have been carried out as yet.
B. FREEZING AND FROZEN STORAGE Husaini and Alm (1955) investigated the influence of the frozen storage (-4” to -20°C for 130 days) of egg white and cod fillets on the number of “masked” SH groups. Amperometric titration with AgN03 was used for the egg white and ferricyanide and o-iodosobenzoic acid for the fish muscle. The difference between the total SH content of proteins determined after the addition of a denaturing agent (dodecyl sulfate) and the amount of SH groups available in the proteins’ native state represented the number of masked SH groups. In the case of egg white, the values of masked groups showed a decrease during the first 28 days; they then began to rise and after 48 days became more or less constant. In cod muscle, the masked SH groups reached a minimum value after 50 days of frozen storage. Again, the masked SH groups then began to rise. After 90 days, the values were not constant but still varied. However. the number of masked SH
74
KLAUS H O F M A N N A N D REINER HAMM
groups finally decreased in all cases, or, in other words, the number of available SH groups increased during frozen storage. This result seems to show that the frozen storage of protein leads to protein denaturation, resulting in a release of reactable SH groups. Grau (1968) also reported that freezing meat would cause an increase in the SH groups, but no literature reference was given. According to Dzinleski et al. (1969). the amount of SH groups of proteins in both frozen beef muscle and drip increased during frozen storage (3 months at -18°C). The SH groups were determined with Ellman’s reagent. The authors postulated that frozen storage caused physical changes in the muscle tissue which resulted in a release of the reactive SH groups upon defrosting. The increases ranged from 41% to 103%. The maximum values found were 16-17 moles SH/105 gm protein. These values are certainly too high, because the total SH content of all investigated meats was found to be 10-12 moles SH/105gm protein, the maximum value being 14 moles SH (see Section IV, A). Nevertheless, it might be possible that the denaturation caused by freezing or frozen storage increases the number of SH groups which react with Ellman’s reagent, because this reagent does not seem to react with all SH groups of the native meat proteins (see Section 11, B, I ) . Connell (1960b), however, was not able to detect changes in either the easily reactable or the total SH groups in cod flesh during frozen storage for up to 3 years at - 14”, -22”, and -29°C. In an investigation by Hofmann et al. (1974) with lean beef and pork (musculus long. dorsi), one part of the samples was frozen in vacuum-sealed plastic bags, while the remainder was packaged in presence of air; both groups of samples were stored at -19°C for up to 24 months. While no significant changes in total SH content were found, the nonprotein SH content decreased (this will be discussed later in connection with results of Khan et al., 1963). Rahelid et al. (1974) also found no change in the total SH content of beef which was stored at - 18°C for I year. The SH content of pork, however, decreased considerably after 6 months. In these experiments, the stored meat samples were not protected against the influence of atmospheric oxygen by sealing them in plastic bags as was done in the experiments of Khan rt al. (1963) and Hofmann et al. (1974). Hence, the decrease in the SH content of pork might have been induced by the formation of fatty acid peroxides (less probable for beef). The oxidation of cysteine to cystine by autoxidizing lipids is well known (Wedemeyer and Dollar, 1964). A decrease of free SH groups in the protein of fish muscle (Sacramento blackfish) by freezing and to a greater extent during frozen storage was found by Mao and Sterling (1970). Khan et al. (1963) stated that a decrease in the “sulfhydryl group content of muscles” takes place during frozen storage of raw chicken muscle. This term is entirely misleading: In this investigation, “sulfhydryl group content of musc1es”refers to the SH content of a metaphosphoric acid extract of the muscle (containing only the nonprotein SH groups) rather than the total SH content of the
SULFHYDRYL A N D DISULFIDE GROUPS IN MEATS
75
muscle. Unfortunately this fact was ignored during the discussion of the results both in this and in later publications (Khan, 1965, 1966; Khan and van den Berg, 1965; Khan and Nakamura, 1971). At the end of 2 years the remaining SH content was found to be about 40% of the initial (nonprotein SH) value at - 10°C storage temperature and about 70% at - 18°C storage temperature. It was suggested “that the destruction of sulfhydryl groups can be used as an index of protein damage during frozen storage’’ (Khan e t a / . , 1963). However, in light of the fact that the SH groups of the protein were not actually estimated, this conclusion does not seem to be justified. During frozen storage, proteolytic enzymes are still active, causing an increase in the products of protein breakdown (Partmann, 1972). These products can be estimated by the Fohn-Ciocalteu reagent (sodium p-naphthoquinone 4-sulfonate) (Khan et d . , 1963; Khan and van den Berg, 1964). The ratio of SH groups and the Fohn-Ciocalteu reagent-positive nitrogen compounds of both breast and leg muscle tissue decreased progressively with time of frozen storage. Khan (1965), therefore, proposed the use of this ratio as a “quality index” for frozen stored poultry. This SH/N ratio decreased after 2 years of frozen storage at - 18”C, from about 2.8 to 1.9 for leg muscle and from about 1.8 to 0.9 for breast muscle.* The SH content was expressed as pg glutathionigm muscle, the N content as pg tyrosin-N/gm muscle. Davidkova and Davidek (1971) applied the index on pork, observing a decrease in the ratio from about 2.3 to 0.6* after 60 weeks of frozen storage at - 18°C. When the meat was stored at -S”C, the value of 0.6 was reached after only 2 months. In most cases, the meat was no more acceptable when the index fell below 0.8. Hofmann et a/. (1974) found a decrease of nonprotein SH and an increase of nonprotein N in both beef and pork after 2 years frozen storage at -19°C. The mean values ranged from 2.2 to 1.6 mg SH/lOO gm meat and from 3.4 to 5.4 gm N x 6.0/1OO gm meat. Consequently the SH/N ratio decreased by about 5070,corresponding to the values given by Khan et al. (1963). However, it seems to be questionable whether this change in the SH/N ratio is reliable for evaluating a loss in meat quality under practical conditions, because the ratio can vary in fresh meat in a wide range (1.2 to 2.2) (after Davidkova and Davidek, 1971). Furthermore, the initial SH content of frozen stored meat will normally not be known (in our opinion it should be) so that it would most often be impractical to judge meat quality by means of the SH/N ratio. Khan and van den Berg (1965) also stated that the frozen storage of cooked meat results in a gradual decrease in the SH content. Because the tenderness of the meat decreased simultaneously, it was concluded “that loss in this sulfhydryl-group content during storage may serve as an index of tenderness.” It should be pointed out that this statement was also based on the estimation of the nonprotein SH content. Since tenderness of meat is related to the properties of the *These values were taken from Fig. I in Davidkovi and Davidek (1971).
76
KLAUS H O F M A N N A N D REINER H A M M
structural proteins, this conclusion seems to be somewhat questionable. In our opinion, the results obtained with cooked meat indicate that the decrease in SH groups must be caused by an autoxidation (no participation of enzymes). However, Khan et al. (1968) attributed the decrease in the SH groups mainly to the so-called cryodenaturation, first postulated by Levitt (1962, 1966). We feel it is necessary to discuss this hypothesis in detail because it has often been adapted and discussed rather uncritically (Gaff, 1966; Khan et al., 1968; Partmann. 1968; DubC, 1969; Buttkus, 1974). Levitt (1966) points out “that SS groups are formed. and this has been explained by the SH SS hypothesis, according to which intermolecular SS bonds form between protein molecules during injurious freezing, leading to denaturation of proteins.” However, from the chemical point of view, the formation of SS from SH cannot occur without oxidation. This fact has not even been mentioned in the discussion. We read further: “That this postulated chemical change can be induced or at least accelerated by freezing has been shown in the case of the model system thiogel. When this SH-protein is frozen, intermolecular SS formation is greatly accelerated as compared with unfrozen gel. . . .” It should be pointed out that no experimental evidence was cited for an increase in SS groups induced by freezing. For that reason the following statement can be considered only as conjecture: “Thiogel . . . provides a system in which intermolecular SS bonds formation can be readily detected and measured quantitatively by the simple method of determining its melting point” (Levitt, 1965). Khan et al. (1968) also investigated the influence of frozen storage on SH groups content of chicken myofibrillar proteins using Ellman’s reagent. In fresh samples, the total SH content was found to be 0.59 pmole/mg N (equal to 9.44 moles SH/105 gm protein) and 0.15 pmole after storage at -5°C for 10 weeks. This unusually high storage temperature may have supported the autoxidation of SH groups. The decrease of protein solubility and the loss of tenderness in meat which is generally observed after freezing or frozen storage may be explained by an aggregation of protein molecules, whereby the native state may sometimes be maintained. This aggregation may be brought about by the formation of hydrophobic bonds, hydrogen bonds, ion bindings, and disulfide linkages (Jaenicke, 1964, 1965a,b). The last mentioned may be formed by a sulfhydryl-disulfide exchange reaction (Fig. 8). This principle mechanism of aggregation may be valid not only for mildly heated proteins, for which it was studied, but aIso for protein systems stored at lower temperatures. In this case, of course, the reaction velocity will be lower; however, long-time storage may cause the SH-SS-exchange reaction to some extent. The freezing of meat and other tissues, if it is done correctly, does not necessarily cause the denaturation of the proteins.
SULFHYDRYL A N D DISULFIDE GROUPS IN MEATS
77
FIG. 8. Sulfhydryl-disulfide exchange as a possible mechanism for the aggregation of nativc protein molecules.
C.
FREEZE-DRYING
Connell (1957) investigated the influence of dehydration on the texture and the
SH content of fish muscle. In fresh cod, he found 0.38 gm SH/100 gm protein using o-iodoso-benzoic acid, and 0.14 gm SH/100 gm protein using N ethyl-nialeimide (NEM). The corresponding values of the freeze-dried material were 0.30 gm SH and 0.13 gni SH, respectively. Both air-drying and vacuum contact-plate drying of cod yielded nearly the same results. Connell concluded: “The results do suggest in a general way that the sulfphydryl content of the dried products is less than that of fresh fish, a finding which is substantiated by the evidence from solubility measurements that disulphide links are formed on drying.” However, the latter statement has not been proved directly because experiments to determine SS groups had not been carried out. Connell further states that “the number of experiments is not sufficient to give a conclusive answer.” Potthast ( 1 972) studied the influence of freeze drying and the effect of storage of freeze-dried pork on SH content by means of an indirect amperometric titration method using NEM and AgN03. The results showed that the easily available SH groups. which were detected with NEM. clearly decreased, whereas the total SH content did not change significantly. This finding indicates that during the freeze-drying process the protein structure becomes tighter. There seems to be none of the unfolding of protein molecules (which results in an increase of the easily detectable SH groups) that Hamm and Hofmann (1965) found in the case of the heat denaturation of meat protein. The storage of freeze-dried mcat at 4 0 4 5 % relative humidity did not result in any change of available SH groups, although a decrease in the enzyme activity (Hamm, 1964; Yasui and Hashimoto, 1966) of some muscle proteins does take place during storage. Potthast has therefore suggested that this denaturation effect is not caused by unfolding the protein chains but by the blocking or chemical changes of the active centers.
D. CURING The chemical reactions involved in the transformation of niyoglobin (Mb) to nitrosomyoglobin (NOMb) after addition of nitrate or nitrite to meat, giving the meat-stable pink color of cured meat, are extremely complicated and have not yet been elucidated in complete detail. The different steps of the formation of NOMb
78
KLAUS HOFMANN AND REINER HAMM
may be formulated in a very simplified way: reduction
reduction
(bacterial)
(chemical)
NO,-NOC-NO
t Mb
-NOMb
There is no doubt that the SH groups of meat participate in this reaction chain. Theoretically the SH groups can agitate in three ways: (a) reduction of NO,- to NO, (b) protection of Mb against oxidation to metmyoglobin (MetMb), and (c) formation of Mb by reduction of MetMb (Fe3+ +. Fez+). The role of SH groups in changes in meat color and in the curing process has been studied by numerous workers (Watts and Lehmann, 1952; Watts etal., 1955; Kelley and Watts, 1957; Hornsey, 1959; Tarladgis, 1962a,b; Stewart et al., 1965; Szakaly, 1966; Reith and Szakaly, 1967; Fox and Ackerman, 1968; Mirna and Hofmann, 1969; Olsman and Krol, 1972; Kortz, 1973: Kubberad et af.,1974). The formation of cured meat color can be accelerated by ascorbic acid (or ascorbate) which is probably due to its reducing effect and/or to the protection of the SH groups against auto-oxidation. The first investigations on the relation of free SH groups to cured meat color were carried out by Watts et a / . (1955). They heated a model mixture containing 50% egg white (delivering SH groups), 0.4% hemoglobin (so-called meat pigment), and 0.1 % sodium nitrite in phosphate buffer pH 5.8. For comparison they added ascorbic acid to a second sample of the mixture and iodoacetamide (for blocking the SH groups) to a third sample before heating. The development of the color was most intense in the mixture containing ascorbic acid and was least intense in the mixture with iodoacetamide. The authors concluded: “The reduction of metmyoglobin and nitrite necebsary for the formation of the cured meat pigment may be brought about by SH groups of muscle protein rather than by reducing enzyme systems. In the absence of ascorbic acid, formation of the pink cured meat pigment parallels the appearance of free sulfhydryl groups.” However, because of the far-reaching consequences of this statement, it must be noted that the model system used here is very different from actual conditions in meat. Rather than myoglobin, the pigment of meat color, hemoglobin, whose oxygen affinity is very different from that of myoglobin, was used. On the other hand, the reactivity of the SH groups in egg albumin is very different from the reactivity of the SH groups in meat protein: In myofibrils the SH groups are already reactive in the native state (e.g., with sodium nitroprusside), whereas there are no SH groups detectable in native egg albumin (Hofmann, 1964, 1966; Hofmann and Hamm, 1966, 1975). Watts et al. (1955) stated further, “Whereas nitrite protects the cured meat pigment from oxidation in the presence of protein SH groups it accelerates oxidation in their absence. This interesting statement does not seem to be valid for meat because meat proteins always contain free SH groups in the native as well as in the denatured state. Kelley and Watts (1957) found that in addition to ascorbic acid, cysteine and glutathione were capable of catalyzing the production of nitric oxide hemoglobin and of protecting the sur”
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
79
face of cured meat from fading by light. The color of cured meat is also influencecl by SH groups which are not bound to proteins. Hornsey (1959) found a positive correlation between the color intensity of cured pork and its content of cysteine plus cystine and the ratio of cysteinelcystine as well. The following results demonstrate that SH groups can react directly with nitric acid, which may be an intermediate product during the formation of NOMb. It is well known that nitric acid can react with SH compounds of forming nitrosothiols (Beckurts and Frerichs, 1906). Ashworth and Keller (1967) established that while secondary and tertiary thiols would react in this way, thiols, such as cysteine and glutathione, would not. Nevertheless, Saville (1958) found that at pH 2- 3 a quantitative reaction between nitrite and cysteine takes place with the formation of the corresponding nitroso compound (S-nitroso-cysteine). The amount of nitrosothiol formed decreased with increasing pH value. In order to find out whether a formation of nitrosothiols is possible in meat at its natural pH (5-6). Mirna and Hofmann (1969) carried out experiments with meat and SH solutions: The addition of nitrite to minced beef and pork in amounts nearly equimolar to the SH content caused a decrease in both components of about 20-30% during storage for 1-2 weeks at +2"C and pH 5.6-5.8; in meat to which no nitrite was added the SH content proved to be stable during 12 days of storage at 2-3°C (Hofmann, 1971d). Thus, the decrease of SH in the presence of nitrite might be due to a reaction of the SH groups with nitrite or nitric acid. This explanation is supported by the observation of Olsman and Krol (I 972) that a smaller loss of nitrite occurred when the SH groups in meat had been previously b1ock:d by an SH reagent. A study of the reaction of nitrite with glutathione and with b.ysteine in watery solutions at different pH values led to the following resulis (Mirna and Hofmann, 1969): ( 1 ) At pH 7.4, no reaction takes place. At pH 2.3, the reaction is nearly complete after 15 minutes at 23°C. At pH 5.0 and IOO"C, only 16-18% of the SH groups of glutathione had reacted after 15 minutes. The pH dependence of the reaction shows that it is the free nitric acid rather than the nitric ion which reacts with the SH groups: R-SH
+ HO-NO
+
R-S-NO
+ H20
(2) The nitrosothiols are unstable. The half-life periods of S-nitroso cysteine and S-nitroso glutathione have been estimated to be 2 and 3 hours, respectively, at pH 5.5. (3) The nitrosothiol of cysteine was isolated in the form of red crystals. In water this compound is slowly decomposed under the splitting off of a gas (NO), while unsoluble cystine simultaneously precipitates. The decomposition is probably caused by the following reaction: 2 CyS-NO
+
CySSCy
+ 2 NO
80
KLAUS HOFMANN AND REINER HAMM
Because the NO group can be split off easily from the nitrosothiols formed, it was suggested that the role of the SH groups in the formation of the cured meat color may perhaps consist of a transfer of the NO group from the nitrosothiol primarily formed to the myoglobin (Mirna and Hofmann, 1969): 2 Mb
+ 2 Protein-S-NO + 2 NOMb + Protein-S-S-Protein
Kubberpd et al. (1974) studied the reaction of nitrite with the SH groups of myosin. The reaction was found to depend on the pH and the temperature much the same as the reaction with the low molecular SH compounds. The rate of the reaction was low under conditions similar to those in meat. However, curing takes days and therefore the reaction may occur to some extent. The formation of S-nitrosothiols has been also discussed by Gilbert ef ul. (I 975). Kortz (1973) presented a hypothetical mechanism for the role of the different SH fractions of meat in the development of cured meat color. According to this hypothesis, nonprotein SH compounds act as intermediates between myoglobin and the SH groups of the water-soluble protein fraction.
E.
RIPENING OF DRY SAUSAGES
Sandholm ef al. (1972) estimated the number of SH groups in dry sausages during the process of ripening using amperometric titration with AgNO,. The SH groups increased up to the twentieth day and declined by the twenty-ninth day to almost the initial value. The increase was assumed to be attributable to a bacterial reduction of the SS groups of the proteins in the sausages. However. this cannot be the case: The SS content in meat is low in relation to the SH content (see Section VI, A and comments to Table XIII) so that the SH content, after the proposed reduction of SS groups, could not rise to a value many times that of the initial SH content. The increase reported by Sandholm ef al. was higher than tenfold (for example, the values for the initial and maximum SH contents were 46 and 540 pmoles SH/gm wet weight, respectively). Furthermore, it must be noted that all these values were far too high in comparison with similar values given by other investigators (meat contains about 65 mg SH/IOO gm tissue, whereas the values reported here corresponded to 152 and 1786 mg SH/100 gm tissue respectively). Therefore, either the procedure of determination followed or the kind of calculation used may have been incorrect.
F.
SMOKING
There are a few results available concerning the question of how the constituents of smoke influence the SH groups of meat. Smoke contains about 300 different components (review: Mohler and Baumann, 1968), some of them very reactive. Among these are numerous phenols, aldehydes, and ketones (Tilgner,
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
81
1967) which are able to react with SH groups (see Cecil and McPhee, 1959; Friedman, 1973; Stauffer, 1974). The carcinogenic. benzo[a]pyren. can also react with SH groups (Harington, 1967; Reske. 1971). Because these reactive compounds occur in smoke, it is not surprising that smoking decreases the total SH content of meat (Randall, 1969; Randall and Bratzler. 1970). The number of SH groups in heated but unsmoked pork was (converted) 12.0 moles SH/105 gm protein; in the heated-smoked samples, the value decreased to 7.0 moles SH (smokehouse conditions of 60°C. 45% R.H.. 2.25 hours). This decrease was attributed to interactions of the smoke with various reactive groups of the meat proteins (Randall and Bratzler, 1970). Krylova and Kusnezowa ( 1 964) observed a drastic decrease in the SH groups found in smoke-cured meat. After smoking, only about 40% of the initial SH groups remained; the phenolic fraction was more active and the basic fraction was less active.
G . IRRADIATION Barron (1946) postulated that ionizing radiation (X rays. alpha, beta, and gamma rays) would rapidly oxidize the thiol groups of cells. This oxidation was explained by the formation of oxidizing radicals when water is radiated in the presence of oxygen. This hypothesis has been supported by the finding that several thiol enzymes are inhibited by radiation, and that this inhibition was prevented or reversed on the addition of glutathione (Barron. 1951). Both the oxidation of thiols to disulfides and the reduction of disulfides to thiols have been observed by several authors during irradiation with ultraviolet light. Degradation of thiols and disulfides can also occur (for reviews, see Cecil and McPhee, 1959; Friedman, 1973). In aqueous solution, cysteine is oxidized to cystine under the influence of gamma radiation (Whitcher et a / . , 1953; AI-Thannon, 1968; Owen and Brown, 1969). Furthermore, hydrogen, hydrogen sulfide, hydroperoxide. and alanine was formed, depending on the pH, the concentration of oxygen, and the SH content in the cysteine solution radiated (Trumbore, 1967). Radiation of solutions containing cysteine and cystine produced also SO,. alkanes, and dimethyldisulfide (Merritt. 1966). When meat is irradiated with doses of gamma rays at the level required for the destruction of the most resistant microorganisms, one can see the development of irradiation odors and off-flavor (Wick, 1965) which may lower food acceptance and values. It is generally recognized that relatively high doses of this irradiation are accompanied by the denaturation of proteins, splitting of protein molecules, or the association of these molecular fractions (reported by Fujimaki ct al., 1961). By trapping the volatiles from irradiated meat in a solution of lead, zinc, and mercuric salts, Batzer and Doty ( I 955) obtained precipitates containing sulfur
82
K L A U S H O F M A N N A N D REINER H A M M
compounds thought to be the source of some of the undesirable odors in irradiated meat. Fractionation of some of the off-odors of beef revealed that they arise from water soluble compounds, mainly from glutathione (Batzer and Doty, 1955). Methods for estimating the amount of hydrogen sulfide and methyl mercaptane after gamma radiation of meat (based on the color reaction with N,N-dimethylp-phenylene-diamine) have been described by Marbach and Doty ( 1956), Batzer and Doty (1955), and Sliwinski and Doty (1958). Small doses below 0.8 mrad, which are used for radiopasteurization, did not influence the sensory qualities of meat, but the higher doses of I .5-5 mrad necessary for radiosterilization had a considerable detrimental effect. The specific undesirable flavor produced was due to the formation of sulfide, mercaptans, carbonyl compounds, and others (Palmin, 1970). The better method of preservation is, therefore, a combination of low dose radiation and heat processing. Several SH-containing compounds have a radioprotective effect on animal organisms, but the mechanism of this action is not yet completely understood (Modig, 1969). Graevsky er al. ( I 969) hypothesize that the radioprotective influence exerted by SH compounds is determined only by their SH groups. Experiments demonstrating distinct radiosensibilitation of mammalian cells and bacteria after blocking the SH groups by NEM or PCMB are in agreement with this conception. The treatment of meat with gamma rays (1.5-3.0 mrad) caused a moderate decrease in SH-glutathione, which was intensified drastically by atmospheric oxygen (Palmin and Breger, 1963). The freezing of minced pork and beef before irradiation with 5 mrad at various temperatures significantly protected SHglutathione, particularly if very low temperatures were used (Coleby et al., 1961). It seems that the freezing of water rather than the low temperature itself is responsible for the protection effect observed. In the investigations of Griinewald (1969), the cysteine content of freeze-dried beef and pork and of fresh beef did not decrease during radiation with less than 1 mrad; radiation with 5 mrad invariably caused a reduction in cysteine and to some extent an increase of the cystine content. On the other hand, the sum of cysteine plus cystine decreased. This shows that radiation with high doses lead to the oxidation of cysteine and to the decomposition of total cysteine plus cystine. However, in the case of fresh beef, cryogenic temperatures prevented these losses of cysteine by radiation. This finding coincides with the results of sensoric tests obtained by other investigators (Coleby et al., I96 1 ; Harland et al., 1967). In other cases the radiation of meat did not cause drastic changes in the SH content. Kardashev et al. (1970) treated fresh river sheatfish and Baltic cod with gamma rays (0.3 to 2 mrad and 0.5 to 10 mrad, respectively). This irradiation produced no change in the contents of SH groups in either the actomyosin
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
83
fraction or in the watersoluble fraction of muscle proteins. Fujimaki et al. (1961) treated meat (rabbit muscle) with gamma rays (about 4 mrad) before and after rigor mortis. There was no significant difference between the content of the sulfhydryl groups of actin from both irradiated and control meat samples. According to the investigations of Hamm et al. (1975), radiation of vacuum-packed lean pork with doses of 0.2, 1 and 5 mrad at 4°C did not significantly change the total SH content of the whole tissue. In this case, the unchanged SH content may be explained by the exclusion of atmospheric oxygen so that an oxidation of SH groups induced by radiation cannot take place. Metlitskii et al. (1968) stated that the radiation of pork, beef, and turkey with 3 mrad did not cause a significant decomposition of cysteine. Hedin et al. (1961) came to the conclusion that X irradiation decreased the SH content of a gelatinelike glucoprotein fraction of beef which produced “wet dog” odor when irradiated. The SH assay was carried out with mercuric chloride and Ellman’s reagent. It was clearly shown that mercuric chloride did not react with the SH groups of the irradiated protein, whereas Ellman’s reagent did. Using the latter reagent, the SH content decreased only by about 7% (from 1.37 to I .27 pmoleslgm). Furthermore, the amino acid analyses of the protein hydrolysate gave nearly the same cysteic acid values for both the irradiated and the untreated samples. There is no doubt that, generally, irradiation of meat causes formation of volatile hydrogen sulfide and mercaptans at the expense of the SH content. However, these amounts are very low in comparison with the total SH content of meat (less than 1%). This can be explained by the finding mentioned previously that the volatile compounds are formed solely from the water soluble components of meat, which are present in minor amounts. Kraybill et al. (1960) reported that gamma irradiation (to 9.3 mrad) resulted in an increase in the SH and SS contents of raw skim milk. Ultraviolet irradiation effected a similar increase in SS content, but no changes in SS bonds. To explain these findings, several mechanisms were discussed based on a degradation of methionine. However, it seems to us much more likely that the increase of available SH groups is due to the denaturation of the proteins, known to be accompanied by an increase of the availability of masked SH groups (see Section V, A, 1). The results of McArdle and Desrosier (1953, who found a liberation of SH groups in irradiated solutions of casein and egg albumin, may be explained in the same way. Neuwirt et al. (1964) found that, after treating rat liver nuclei with a dose of 750 roentgen, there was a significant decrease in protein disulfide groups but no decrease or increase in protein SH groups. In addition, they point out that, after ray treatment, no decrease in the activity of any SH-enzyme has ever been demonstrated.
84
K L A U S H O F M A N N A N D REINER H A M M
VI.
INFLUENCE OF THE SH GROUPS ON THE SHELF LIFE OF MEAT AND MEAT PRODUCTS
In the search for new compounds for the preservation of meat and meat products, it has been observed that several SH compounds, for example, cysteine and glutathione, are able to potentiate the microbial inhibition effect of known food preservatives such as sorbic acid, benzoic acid, fatty acids, and others. Mixtures with these SH compounds have proved to be effective for the preservation of meat, fish, and other foods (Troller, 1966). It is well known that SH compounds have an antioxidative effect which retards the rancidity of fats (Maloney et al., 1966). This effect may be due to the decrease in the redox potential by SH compounds. A direct inhibition of microorganisms by SH compounds would seem to be unlikely because the growth of microorganisms is usually inhibited by SHblocking reagents (Zsolnai, 1970). Cooked meat becomes rancid more quickly than unheated meat because the SH content in cooked meat is reduced as a consequence of the heating effect (Hofmann, 1964; Bognar, 197 1a). This preservation effect of the SH groups in meat may be of limited importance; nevertheless, it seems to be worthwhile to prevent SH oxidation during meat processing as far as possible.
VII. TOXICOLOGICAL ASPECTS SH groups are able to bind toxic heavy metals [Pb2+, (see V, D,2,f), Hg2+, and others] and may therefore play an important role in detoxification reactions (Clarkson, 1971). On the other hand, experiments with rats have shown that an increase in the protein content of diets resulted in a significant increase in the retention of lead (Milev et af., 1970). In this case the lead was probably also bound to the protein SH groups of the feed. Therefore, it is difficult to predict generally whether under practical conditions the SH groups in foods have a positive or negative effect on the contamination of living organism. Furthermore, SH groups interact with numerous carcinogens such as polycyclic aromatic hydrocarbons (PAH), hormones, hepacarcinogenic substances (e.g., certain aminoazo dyes and amines, carbon tetrachloride, aflatoxins, thioacetamide, and ethionine), alkylating agents, nitroso compounds, 4-nitrochinoline N-oxide, lactones, quinones, metals, metalic derivatives, and arsenic (for review, see Harington, 1967). Therefore, SH groups in organisms and foods are important for anticarcinogenesis (for review, see Reske, 1971). Nemoto el al. (1975) reported direct evidence that conjugation with SH-glutathion is a significant
SULFHYDRYL A N D DISULFIDE GROUPS I N MEATS
85
100
-
5 80
E
x 60 U,
L
ul a
C
-
w
'0 A0 C
g
20
0 0
1
2
3 4 Tirnc(hrsl
5
24
FIG. 9. Reaction of GSH with patulin in equirnolar amounts at different pH values (20°C). (A: pH 7.4; B : pH 6.0; C: pH 5.0. (After Hofmann ct d..1971 .)
mechanism for the detoxification of the epoxides of PAH, which are mutagenic and may be the carcinogenic forms of the PAH. Hofmann et al. (1 97 1 ) observed that patulin (a potent mycotoxin produced under certain conditions by molds found on meat and meat products) reacts with SH-glutathion. In chick embryo and rabbit or mice skin test. the reaction product formed was proved to be nontoxic. The kinetics of the reaction of GSH and Patulin at different pH values are shown in Fig. 9. It was concluded that, at the pH of meat and meat products, most of the patulin produced by molds might be inactivated by the SH groups occurring in meat. It is also suggested that, in bread, the mycotoxin is inactivated by SHcontaining substances after a prolonged incubation period (Reiss, 1973). Another possibility of detoxification, which has not yet been discussed, is the reaction of SH groups with nitrite (Mirna and Hofmann, 1969). This may be a competitive reaction to the formation of carcinogenic nitrosamines from nitrite and amines. Finally it should be mentioned that there exists a protective action of sulfur compounds such as cysteine and glutathione against acetaldehyde toxicity arising from heavy consumption of alcohol and heavy cigarette smoking (Sprince et d., 1975). As these examples demonstrate, SH groups may be of importance in the detoxification mechanisms of the organism. As we have seen in Section V , the amino acids, cysteine, cystine, and methionine, the exclusive sources for SH compounds in the metabolism, are not very stable under meat processing. For this reason, careful preservation may be of importance not only for the nutritive value but also for the protection of the organism against several classes of toxic substances.
86
KLAUS HOFMANN AND REINER HAMM
VIII.
SUMMARY
Methods for the determination of SH and SS groups in meats and their advantages and disadvantages are discussed. The choice of determination method depends on the type of investigation in question. Numerous problems which may arise in the determination of SH groups in proteins are discussed in detail. Protein SH groups may react quickly, slowly or not at all (“masked” SH groups) depending on the type of reagent, on the protein’s state, and on the conditions of reaction. Meat contains both protein SH (soluble and unsoluble) and nonprotein SH groups. Results found in the literature do not always show clearly whether it is the total SH or the SH content only in a soluble fraction that was determined. This differentiation is important because most SH groups are bound to the water insoluble myofibrillar proteins. One of the most convenient and accurate techniques for the determination of SH groups in meat is a “double-indirect’’ amperometric titration method: this enables the application of different SH reagents (AgNO,, NEM, PCMB, etc.). The determination of SS groups is usually based on the determination of SH groups before and after the reduction of SS (preferably with sodium borhydride) to SH. The procedures are briefly described. The sum of SH and SS represents the “total cystine” (cysteine plus cystine) content. Tables are presented listing the amounts of SH and SS groups in meats (muscles and inner organs), in myofibrils, and in isolated muscle proteins. Although the results given in literature vary considerably (depending on the method of assay or on the type of material investigated), the following general conclusions may be drawn: the average SH contents of pork and beef muscle are nearly equal (62 and 66 mg SH/100gm tissue, or 10.2 and 10.5 moles SH/105 gm protein, respectively); the average SH contents of skeletal muscles from different species are very similar to each other; the nonprotein SH content in muscles varies in a wide range; muscles usually contain 3 to 4% nonprotein SH as compared to the total SH content; the average SH content of myofibrils prepared from the muscles of different animals is 9.1 moles SH/105 gni protein (this value corresponding very closely to the sum of SH groups of the single myofibrillar proteins). The SS content of muscles varies from 0.5 to 2.0 moles SS/I05 gm protein. This high variation may be due to a different degree of oxidation of the SH groups. The SS content may also depend on an animal’s age. The average total SH content of liver (47 mg SH/IOO gm tissue) and of other organs (such as kidney, heart, and brain) is lower than the average SH content of muscles; however, the nonprotein SH content (mainly glutathione) is much higher in liver and in other organs than in muscles. In general, the total SH as well as the nonprotein SH contents of the same organs of different species are less variable than the SH contents of different organs of the same species. Factors which influence the available SH content of raw meat may be training
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
87
and stress (influencing the content of glutathione), the age of animals, heavy metals, postmortem aging, and others. The role of the SH groups in muscle proteins (including the proteins of the sarcoplasmatic reticulum, sarcolemma, mitochondria, and sarcoplasmic matrix) in their physiological functions (ATPase activity, contraction, rigor mortis, and interaction of myofibrillar proteins) and in the tenderness of meat are discussed. Heating meat to 70°C increases the availability of SH groups for several SH reagents (DTNB, NEM, PCMB) as a result of denaturation; however, the total SH content does not change. At temperatures higher than 70°C the SH content decreases, chiefly because of an oxidation to SS groups. High temperatures as used for sterilization may also cause a loss of the cysteine plus cystine content, resulting in a decrease in the nutritive value of meat proteins. During heating, hydrogen sulfide and numerous other sulfur-containing volatile compounds contributing to meat flavor, are split off from the sulfur-containing amino acids of meat. The texture of meat is influenced by the formation of SS groups during heating. It seems more probable that SS groups influence brittleness rather than hardness, and that, therefore, an increase in SS groups does not render the meat tougher but crispier. Although the results to be found in literature are not uniform, freezing and frozen storage does not remarkably influence the total SH content of meat; however, there is a decrease in the nonprotein SH content during long-term frozen storage. As the nonprotein N content increases, the SH/N ratio decreases simultaneously. This ratio has been proposed as a quality index for frozen stored meat. The question of its usefulness is discussed. There is no experimental evidence for the statement that the freezing of proteins would cause an increase in SS groups. Freeze drying of meat does not decrease the SH content estimated with AgNO,, but the SH groups which are easily available for NEM decreased, demonstrating that the protein structure becomes tighter during freeze drying. It was demonstrated that SH groups are involved in the curing process. Different ways in which SH groups can react during the formation of the cured meat pigment, nitrosomyoglobin, are discussed. Smoking decreases the SH content of meat products because smoke contains many compounds (phenols, aldehydes, ketones) which are able to react with SH groups. Irradiation of meat causes the formation of hydrogen sulfide and volatile mercaptans at the expense of the SH content, but these amounts are low (< 1%) related to the total SH content. In the presence of atmospheric oxygen, higher doses of gamma radiation leads to an SH decrease. The freezing of meat before irradiation protects the SH groups. Several SH compounds are able to potentiate the microbial inhibition effect of known food preservatives. Because SH groups can bind toxic elements, toxic
88
KLAUS HOFMANN AND REINER HAMM
compounds, and carcinogens, SH groups in organisms and in foods are important for detoxification mechanisms. Therefore, careful preservation in order to maintain the cysteine content may be of importance not only for the nutritive value but also for the protection of the organism against several toxic substances.
IX.
RESEARCH NEEDS
( I ) Are the relation of SH to SS and the total SH plus SS content of muscle tissues of meat animals significantly influenced by the age of these animals? This question should be investigated with a number of younger and older animals sufficient for statistical evidence, or, more preferably, with the same animals at different stages of age using biopsy samples. (2) The binding forces participating in muscle contraction and rigor mortis are not yet understood. Is the SH-SS exchange reaction involved in these processes? (3) Is the SH-SS system relevant for the rheological behavior of minced meats and meat emulsions'? (4) Several hypothetical mechanisms have been discussed regarding the question of the way in which SH groups are involved in the curing process. What is the actual mechanism? (5) Do traces of metals such as Pb. Hg, Cu, Zn, Cd, and other possible residues in meat which are able to react with SH groups influence the development of sulfur containing volatiles, and, therefore, the development of meat flavor? ( 6 ) Does the introduction of SH groups into proteins (thiolation) improve the nutritive value of proteins which are poor in sulfur content?
ACKNOWLEDGMENTS We wish to express our sincere thanks to Mr. Erich Bliichcl and Mrs. E. Hofmann for their assistance during the preparation of this manuscript.
REFERENCES Alexander, N. M . 1958. Spectrophotometric assay for sulfhydryl groups using N-ethylmaleimide. Anal. Chem. 30, 1292. Allaway. W . H.. and Thompson, .I.F. 1966. Sulfur in the nutrition of plants and animals. Soil Sci. 101, 240. Andrae, W. 1969. Zur Marmorierung von Weilblech-Konservendosen. 2. Miu. Reaktion der Schwefelverbindungen. Nuhrung 13, 549.
SULFHYDRYL AND DlSULFlDE GROUPS IN MEATS
89
Al-Thannon. A. 1968. Kinetics of gamma radiolytic decomposition of aqueous cystcine solutinns. PhD Thesis. Univ. of Deleware. Newark. Anfinsen, C. B.. and Haber, E. 1961. Studies on the reduction and re-forrnation of protein disultide bonds. J . B i d . Chem. 236, 1361. Anson. M. L. 1945. Protein denaturation and the properties of protein groups. A h . Prorc4n Chrm. 2 , 361. Arai. K.. and Watanabe. S . 1968. A study of troponin. a myotibrillar protein from rabbit skelctal muscle. J . Biol. Chem. 243, 5670. Arnold. R. G.. Libbey, L. M., and Lindsay, R. C. 1969. Volatile flavor cornpnunds produced by heat degradation of thiamine (vitamin B,). .I. Agric,. Food. Chrm. 17, 390. Arroyo. P. T.. and Lillard, D. A. 1970. Identification of carbonyl and sulfur compounds from nonenzymatic browning reactions of glucose and sulfur-containing amino acids. J . Food S1.i. 35, 769. Ashworth. M. R. F. 1976. “The Determination of Sulphur-Containing Groups.” Vol. 2: Analytical Methods for Thiol Groups. Academic Press. New York. Ashworth. G. W . . and Keller. R. E. 1967. Ultraviolet determination of tertiary mercaptans as thionitrites. A n d . Cheni. 39, 373. Azzone. G. F., Carafoli. E.. and Barbolini. G. 1956. Ammetric detcrminatinn of SH groups of myosin and actomyosin. Ric. G i . 26, 3035. Bachhawat, J . M.. Ramegowda. N . S . , Koui. A. K . . Narang, C. K.. and Mathur. N. K . 1973. Determination of thiols by oxidation with N-bromosuccininiide. fncfinn J . Chem. 11, 614. BLany. M.. and Birany. K . 1959. Studies on “active centers” of L-Myosin. Biochim. Biophys. Actir 35, 293. Barany. M.. and BBrBny. K . 1973. A proposal for the mechanism of the contraction in intact frog muscle. Cold Spring Hnrhor Sym. Qltunr. Biol. 37, 157. Barany, M . , and Merritield. R. B. 1973. An ATP-binding peptide. Cold Spring Hurbor S.vmp. Quunt. Biol. 37, 121, Barany. M.. Gaetjens. E., Barany. K.. and Karp. E. 1964. Comparative studies of rabbit cardiac and skeletal myosin. Arch. Biochem. Biophys. 106, 280. Barnett. R. I.. and Seligman. A . M. 1952. Histochemical demonstration of protein-bound sulfhydryl groups. S & w P 116, 323. Barron. E. S. G . 1946. “Some Aspects of Biological Action of Radiations.” U . S . Declassified Doc. MDDC 484. USAEC, Oak Ridge, Tennessee. Cited in Barron (1951). Barron, E. S. C. 1951. Thiol groups of biological importance. Adv. Enzymol. R e h f . Suhj. Bioc.hem. 11, 201. Batzer. 0. F., and Dory, D. M. 1955. Nature of undesirable odors formed by gamma irradiation of beef. J . Agric. Food Chem. 3, 64. Baumgirtner, H.. and Baum, F. 1960. Beitrag zur Bildung von Schwefelwasserstoff in Fleisch durch Mikroorganismen beim Erhitzen. Ernuhrutib.s~~r.schung 5 , 40. Beckurts. H.. and Frerichs, G . 1906. Beitrage zur Kenntnis der Thiooxyfettsaureanilide. J . Prukf. C h e m 74, 24. Bem. Z . . Hofmann. D.. Malesevic. D.. and Glisin. B. 1970. Die Mengc an Sulthydryl- und Disulfidgruppen in Fleischkonserven. Proc. E u r . M e e f . Mear Res. Workers, / 6 f h , Vurncr p. 785. Bendall, J . R. 1969. “Muscle. Molecules and Movement.” Heinemann. London. Bendall. J . R.. and Wismer-Pedersen, J . 1962. Some properties of the tihrillar proteins of normal and watery pork muscle. J . Food. Sci. 27, 144. Benedict. R. C.. and Stcdman. R. L. 1970. Ellman’s reagent: Interference in mercapto group determination, with special reference to cigarette smoke. Anulysf 95, 296. Benesch, R.. and Benesch, R. E. 1948. Amperometric titration of sulfhydryl groups in amino acids and proteins. Arch. Biochern. 19, 35.
90
KLAUS HOFMANN AND REINER HAMM
Benesch. R., and Benesch, R. E. 1962. Determination of SH groups in proteins. Methods Biochcm. Anal. 10, 43. Benesch, R. E., Lardy, H. A.. and Benesch. R. 1955. The sulfhydryl groups of crystalline proteins. I. Some albumins, enzymes and hemoglobins. J . B i d . Chem. 216, 663. Bennett. H. S., and Watts. R. M. 1958. The cytochemical demonstration and measurement of sulfhydryl groups by azo-aryl mercaptide coupling with special reference to mercury orange. I n “General Cytochemical Methods” (J. F. Danielli. ed.). Vol. I . p. 317. Academic Press, New York. Berg. J . N . , Lebedeva, N . A , , Markina, J. A , . and Ivanov, I . I . 1965. EinfluBdes hohen Druckes auf einige Eigenschaften des Myosins. Biokhimiyu 30, 277. [Abstr.: Chem. Zentralbl. 137, No. 3 1 1428 (1966).] Beuk. J . F.. Chornock, F. W., and Rice, E. E. 1948. The effect of severe heat treatment upon the amino acids of fresh and cured pork. J . B i d . Chem. 180, 1243. Beveridge, T..Toma, S. J . . and Nakai, S. 1974. Determination of SH- and SS-groups in some food proteins using Ellman’s Reagent. J . Food Sci. 39, 49. Bhattacharya, S . K. 1958. Total sulphydryl (SH) content of blood and tissues. Biochem. J . 69, 43. Bhattacharya, S. K. 1959. Amperometric determination of sulphydryl content of blood and tissues. Nature (London) 183, 1327. Bigwood, E. J. 1972. Amino acid patterns of animal and vegetable proteins4ommon features and diversities. In “Protein and Amino Acid Function” (E. J . Bigwood, ed.), p. 238. Pergamon, Oxford. Bitny-Szlachto. S . . Kosinski, J . . and Niedzielska, M. 1963. Determination of sulfhydryl groups with 2,4-dinitrophenyl-2-hydroxyethyI disulfide. Acta Pol. Phurm. 20, 365. Bjarnason, J . . and Carpenter, K . J . 1970. Mechanism of heat damage in proteins. 11. Chemical changes in pure proteins. Br. J . Nurr. 24, 313. Blocksma. A. H. 1972. The relation between the thiol and disulfide contents of dough and its rheological properties. Cereal Chem. 49, 104. Bloeck, S . . Hofling, E., Baur, R . , and Susin, U. 1970. “Verhalten von H,S-abspaltenden Fullgutern in Steralcon.” Rep. No. 1310. AIusuisse Forschungsinst.. Neuhausen, Switzerland. Bocchini. V., Aloito, M. R., and Najjar, V. A. 1967. Sulfhydrylgruppen der Phosphoglucomutase aus Kaninchenmuskel. Biochemistry 6 , 313. [Abstr.: Chem. Zmtralbl. 139, No. 5-1235 ( I 968). ] Boelens, M . . van der Lindr. L. M.. de Valois. P. J . , van Dort, H. M . . and Takken. H. J . 1974. Organic sulfur compounds from fatty aldehyds, hydrogen sulfide. thiols and ammonia as flavor constituents. J . Agric. Food Chem. 22, 1071. Bognar, A. I97 la. Beitrag zur Ermittlung des ernahrungsphysiologischen Wertes von Fleisch in Abhsngigkeit von der thermischen Behandlung. PhD. Thesis, Univ. Hohenheim, Hohenheim, Germany. Bognar, A. 1971b. EinfluB der thermischen Behandlung auf den Gehalt an Aminosauren in Rindfleisch. Erniihr.-Umsch. 18, 200. Bolshakov, A. S . . and Mitrofanov, N . S . 1970. Evaluation of sulphydryl groups in meat by amperometric back titration (in Russ.). Prikl. Biokhim. Mikrobiol. 6, 606. Bolshakov, A . S . . Karpeev. J . J . . Mitrofanov, N. S . , and Khlebnikov, V. J. 1972. Determination of inertly reacting sulfhydryl groups in meat by their reaction with p-chloromercuribenzoate (in Russ.). Prik/. Biokhim. Mikrobiol. 8, 367. [Chem. Abstr. 77, 6 0 1 5 1 ~(1972).] Bowers, J . A. 1972. Eating quality, sulfhydryl content and TBA [2-thiobarbituric acid] values of turkey breast muscle. J . Agric. Food Chem. 20, 706. Boyer, P. D. 1954. Spectrophotometric study of the reaction of protein sulfhydryl groups with organic mercurials. J . Am. Chem. Soc. 76, 4331.
SULFHYDRYL AND DlSULFlDE GROUPS I N MEATS
91
Boyne. A. F . . and Ellman, G . L. 1972. A methodology for analysis of tissue sulphydryl components. A n d . Biochem. 46, 639. Brennan, M. J . . and Bernhard, R. A . 1964. Headspacr constituents of canned beef. Food Twhnol. 18, 149. Brennock. W . E.. and Read. W . V. 1972. Sulfhydryl and amino acid composition of actin and myosin in hereditary muscle dystrophy in chicken. Res. Commun. Chern. Parhol. Pharmucol. 3, 417. Brezina, M.. and Zuman, P. 1956. “Die Polarographie in der Medizin. Biochemie und der Pharniazie.” p. 607. Akad. Verlagsges.. Frankfurt a.M. Brown. D. W. 1960. Reduction of protein disulfidc bonds by sodium borohydride. Biochim. Biophys. Actu 44, 365. Brush. J . S., Jensen. E. V . . and Jacobsen. H. I . 1963. A study of t h r sulfhydryl group in bovine albumin. Biochim. Biophys. Actu 74, 688. Burke. M.. Reisler. E . , Himmelfarb, S . . and Harrington. W. T. 1974. Myosin adrnosintriphosphatase convergence of activation by actin and by SH, modification at physiological ionic strength. J . B i d . Chem. 249, 6361. Burley. R. W. 1954. Sulfhydryl groups in wool. Nurure (London) 174, 1019. Burton, H. 1958. The amperometrie titration of sulfhydryl groups at a rotating platinum electrode. Biochim. Biophys. Actu 29, 193. Butterworth, P. H., Baum. H.. and Porter. J. W. 1967. Arch. Biochem. Biophys. 118, 716. Cited in Jocelyn (1972). p. 149. Buttkus. H. 1971. The sulfhydryl content of rabbit and trout myosins in relation to protein stability. Can. J . Biochem. 49, 97. Buttkus, H. 1974. On the nature of chemical and physical bonds which contribute to somc structural properties of protein foods: A hypothesis. J . Food Sci. 39, 484. Calcutt. G. 1961. Sulphydryl levels of the liver and kidneys from rats fed dl. ethionine. Br. J . Cancer 15, 683. Calcutt. G . , and Doxey. D. 1959. The measurement of tissue-SH. Exp. Cell. Res. 17, 542. Calcutt. G . , and Doxey. D. 1961. The influence of thiamine and pantothenate upon tissue-SH-levels. Br. J . Cancer 15, 157. Calcutt. G . , and Doxey, D. 1962. The apparent glutathione content of some normal tissues and snme animal turnours. Br. J . Cuncer 16, 562. Calcutt. G.. Doxey. D., and Coates, J . 1961. Further studies uf the effects of chemical carcinogenis upon the SH levels of target and nontarget tissues. Br. J . Cancer 15, 149. Caldwell. K. A., and Lineweaver. H. 1969. Sultkydryl content of excised chicken breast muscle during aging. J . Sci. Food Chem. 34, 290. Calvin, M. 1954. In “Glutathione. A Symposium” (S. P. Colowick, A . Lazarow, E. Rackcr. D. R . Schwarz. E. R. Stadtman, and H. Waelsch, eds.). p. 33. Academic Press. New York. Cited in Lumper and Zahn (1965). Carsten. M. E. 1966. Actin. Its thiol groups. Biochemisrry 5, 297. Carter. J. R . 1959. Amperometric titration of disulfide and sulfhydryl in proteins in 8 M urea. J . B i d . Chem. 234, 1705. Casey. J . C.. Self. R.,and Swain, T. 1965. Factors influencing the production of low-boiling volatiles from food. J . Food. Sri. 30, 33. Cavallini. D., Graziani, M. T.. and Dupre, S. 1966. Determination of disulphide groups in proteins. Nuturr (London) 212, 294. Cecil, R. 1963. Intramolecular bonds in proteins. I . The role of sulfur in proteins. In “The Proteins” (H. Neurath, ed.), p. 379. Academic Press, New York. Cecil, R., and McPhee, J. R. 1959. The sulfur chemistry of proteins. Adv. Prorein Chem. 14, 255.
92
KLAUS HOFMANN AND RElNER HAMM
Chajuss. D.. and Spencer, J. V. 1962a. The effect of oxidizing and reducing aging media on the tenderness of excised chicken muscle. J . Food Sci. 27, 303. Cha.juss, D.. and Spencer. J. V. 1962b. Changes in the total sulfhydryl group content and histochemical demonstratibn of sulfonates in excised chicken muscle aged in air. J . F u o ~ Sci. . 27, 41 I . Chang. S. F., and Liener. J. E. 1964. New chromophoric reagents for labelling mercapto-groups of proteins. Narure (London) 203, 1065. Cheftel. H. 1958. Die Korrosion dcr Weibblechdose in der Konscrvenindustrie. W c v h f . Korros. 9, 630. Chibnall, A. C. 1943. Aminc)-acid analysis and the structure of proteins. Proc. R. Sor.. S u . , B 131, 136. Christian. G. D.. and Schur. P. H. 1965. The amperometric titration of total and interchain disulfidc bonds in y-globulin. Biochim. Eiophys. A m 97, 358. Christiansen, K . 1968. Versuche zur Anwendung der Umkehrrotation bei der Sterilisation von Fleischkonserven. F/eischwir/.schoj 48, 1 149. Clarkson, T. W. 1971. Epidemiological and experimental aspects of lead and mercury contamination of food. Food Cosmet. T O J X ~ C 9,O229. ~. Cleland. W. W. 1964. Dithiothreitol, a new protective reagent for SH groups. Biochemisrry 3, 480. Cleland, W. W. 1968. A specific and sensitive determination ofdisulfides. J . B i d . Chem. 243,716. Cohen, C . , Caspar, D. L., Johnson, J. P.. Nauss. K . . Margossian. S. S . . and Parry, D. A. D. 1973. Trapomyosin-troponin assembly. Cold Spring Harbor Symp. Quanr. B i d . 37, 287. Coleby B . , Ingram, M., Shepherd, H. J . . Thomley. M. J . . and Wilson. 0. M. 1961. Treatment of meat with ionising radiations. VI1. Effect of low temperatures during irradiation. J . Sci. Food Agric. 12, 483. Connell, J . J . 1957. Some aspects of the texture of dehydrated fish. J . Sci. F ~ o dAgric. 9, 526. Connell. J. J. 1960a. Studies on the proteins of fish skeletal muscle. VII. Denaturation and aggregation of cod myosin. Biochem. J . 75, 530. Connell. J. J. 1960b. Changes in the adenosinetriphosphatase activity and sulphydryl groups of cod flesh during frozen storage. J . Sci. F o ~ dAgric. 11, 245. Connell, J. J . 1961. The relative stability of skeletal muscle myosins of some animals. Bi<JchtW.J . 80, 503. Crocker. E. C. 1948. Flavor of meat. Food R m . 13, 179. Cummins, P., and Perry. J . V. 1973. Subunits and biological activity of polymeric forms of tropomyosin. Eioc+wm. J . 133, 765. Dahlke. H . 1969. Der EinfluS des Erhitzens verschiedener Eiweibe auf die Marmorierung durch das Freisetzen von Sulfhydrylgruppen und tliichtigen Schwefelverbindungen. F/ei.sch 23, 80. Daniel, J. L.. and Hartshomc, D. J . 1972. Sulfhydryl groups of natural actomyosin essential for the Ca2+ sensitive response: Location and properties. Biochim. Eiophys. A m 278, 567. Davidkova, E., and Davidek. J . 1971. Anwendung des Khanschen Qualitatsindexes zur Verfolgung der in gefrorenem und gekiihlt gelagertem Schweinefleisch eingetretenen Verlnderungen. 2. Lebensm. - Unters, -Forsch. 146, 6. De Marco, C.. Graziani, M. T . , and Mosti. R . 1966. The spectrophotometric determination of think and disulfides with N-ethylmaleimide. Anal. Eiochem. 15, 40. [Abstr.: 2. A n d . Chem. 218, 237 ( I 966). I Diez, M. J . F., Osuga, D. T.. and Feeney, R. E. 1964. The sulfhydryls of avian ovalbumins, bovine P-lactoglobulin, and bovine serum albumin. Arch. Eiochern. Eiophys. 107, 449. Dodge. J. W., and Stadelman. W. J. 1959. Postmortem aging of poultry meat and its effect on the tenderness of the breast muscles. Food Techno/. 13, 81. Donoso, G.. Lewis. 0. A. M.. Miller, D. S . . and Payne. P. R. 1962. Effect of heat treatment on the nutritive value of proteins: Chemical and balance studies. J . S(,i. Food Agric. 13, 192. Doxey, D. 1961. Measurement of the sulfhydryl content of tissue cultures. Er. J . Cancer 15, 146.
SULFHYDRYL AND DlSULFlDE GROUPS I N MEATS
93
Drabikowski. W.. and Bitny-Szlachto, S. 1963. The action of P-hydroxyethyI-2.4-dinitrophL.nyl disulphide o n sulphydryl groups of actin. Bull. Acuti. P O / . Sc,i.. S r r . S(,i. 11, 165. Drabikowski. W . . and Bitny-Szlachto. S . 1964. Studies o n the sulfhydryl groupa of actin. . 4 ~ . l u Bioc.him. Pol. 11, 421. Drabikowski. W.. and Nowak, E. 1965. Studies o n aulfhydryl groups of tropomyosin. A ~ . t Biochim. u Pol. 12,61. Drabikowski. W.. and Nowak. E. 1970. Thiol group content of tropomyosin and troponin. Al’rL7 Biochim. Pol. 17, 221. Dube. G. 1969. Tenderness evaluation of bovine niusclea and changes in sulfhydryl and dihulfidc contents of myofibrillar extracts cooked at different temperatures. PhD. Thesis. Purdue Univ., Lafayette, Indiana. (University Microfilms. Ann Arbor. Michigan. No. 70-3881 .) Dube. G.. Bramblett. V. D., Judge. M. D.. and Harrington. R . 9. 1972. Physical properties and sulfhydryl content of bovine muscles. J . Food Sci. 37, 23. Dupont. Y . , and Hasselbach. W. 1973. Structural changes in the sarcoplasniic reticulum membrane induced by sulfhydryl reagents. Nrtture (London) 246, 41. Dwivcdi. 9. K.. and Arnold, R. G. 1971. Hydrogen sulfide from heat degradation of thiamin. J . Agric.. Food Chem. 19, 923. Dwivedi, B. K.. and Arnold. R. G. 1972. Chemistry of thiamine degradation. Mechanisms of thiamine degradation in a model system. J . Food Sc,i. 37, 886. Dwivcdi, 9. K . , and Arnold. R. G . 1973. Chemistry of thiamine degradation in food products and inridel systems: A review. J . Agric,. Food C h e m 21, 54. Dwivcdi. B. K.. Arnold, R. G . , and Libbey. L. M. 1972. Chemistry of thiamine degradation: 4-methyl-S-(~-hydroxyl) thiazole from thernially degraded thiamine. J . Food Sc.i. 37, 689. Dwivedi. B . K., Arnold. R. G.. and Libbey. L. M. 1973. Some minor volatile components from thermally degraded thiamine. J. Food S i . 38, 450. Dworschak. E. 1969. Comparative chemical study of the proteins of raw and roasted meat from the point of view ofdiatetics. Elelmi.szrrvizsgukrri Kozl. 15, 263. [Abstr.: FSTA 2, 6 S 494 (1970). I Dworschak. E. 1970. Untersuchungrn iiber den biologischen Wert des EiweiRes in nihem und gebratcnrm Fleisch. Z . Lebensm.-Untersuch. -Forsch. 143, 167. Dzinleski. B. E.. Bratzler. L. J., Pepper, F. H.. and Pearson. A. H. 1969. Anderungen dcr SHGruppen in Proteinen von gefrorenem Rindermuskel und Tropfsaft (in Yugosl.). Techno/. Mew 2. 34. Ebashi, S . . and Nonomura. Y . 1973. Proteins of the niyofibril. / ! I “The Structure and the Function of Muscle” (G. H. Bourne. ed.). 2nd Ed., Vol. 3 . p. 28.5. Academic Press. New York. Ebashi. S . . Kodama. A,. and Ebashi. F. 1968. Troponin I . Preparation and physiological function. J . Biochrm. (Tokyo) 64, 465. Ehrlich. E. 1967. Technique de dosage automatique dc groupetnenta sulfhydriles. Application a I‘actomyosine. Bull. Sor. Chim. Biol. 49, X89. Eldjarn. L.. and Jellum, E. 1963. Organomercurialpolysaccharide, a chromatographic material for the separation and isolation of SH-proteins. Acru C h m i . S ~ t n d .17, 2610. Ellman. G. L. 1959. Tissue sulfhydryl groups. Arch. Bioc,hent. Biophys. 82, 70. Elzinga. M., and Collins. S. H. 1973. The amino acid sequence of rabbit skeletal muscle actin. Cold Spring Hurhor Symp. Quunt B i d . 37, I . Erwin. G. V.. and Pedersen. P. L. 1968. A sensitive gel-filtration method for determination of protein sulfhydryl groups with chloromercuribenzcratc -“C. A n d . Biochem. 25, 477. Ewart. J . 1972. A modified hypothesis for the structure and rheology of glutelins. J . S c i . Food Agric,. 23, 687. Ewing. G. W . 1960. “Instrumental Methods of Chemical Analysi McGraw-Hill, New York. Fasold. H.. and Turba, F. 1959. Denaturierung von Proteinen. Wiss. Ver6f. Dtsch. Gus. D n i i h r . 5, I .
94
KLAUS HOFMANN AND REINER HAMM
Feigl, F. 1960. “Tiipfelanalyse,” p. 233. Akad. Verlagsges. Frankfurt a.M. Fischer. C.. and Hamm, R . 1975. Sulfhydryl- und Disulfid-Gruppen in PSE-Fleisch. Fleischwirrschaft 55, 1295. Flesch, P . . and Kun, E. 1950. A colorimetric method for determination of sulfhydryl groups in tissue homogenates by I -(4-chloromercuriphenylazo)-naphthol-2. Proc. Soc. Exp. Eiol. Med. 74, 249. Flesch, P., Golomb, S . , and Satanove, A. 1954. Colorimetric determination of protein-bound waterinsoluble sulfhydryl groups with the Barmett-Seligman Reagent. J . Lab. Clin. Med. 43, 957. Folin. O., and Marenzi. A . D. 1929. An improved colorimetric method for the determination of cystine in proteins. J . Eiol. Chem. 83, 103. Fontana, A , , Veronese, F. M., and Scoffone. E. 1968. Sulfenyl halides as modifying reagents for polypeptides and proteins. 111. Azobenzene-2-sulfenyl bromide, a selective reagent for cysteinyl residues. Biochemistry 7, 3901. Forbes, W. F., and Hamlin, C. R. 1968. Determination of mercapto and disulfide groups in proteins. I . A reassessment of the use of methylmercuric iodide. Can. J . Chem. 46, 3033. Fox. J. B . , Jr., and Ackerman, S. A. 1968. Formation of nitric oxide myoglobin: Mechanisms of the reaction with various reductants. J . Food. Sci. 33, 364. Fraczak, R.,and Pajdowski, Z. 1955. The decomposition of sulfhydryl groups in meat by thermal processing. Przem. Spozyw. 9, 334. Frater, R., Hird, F. J . R.. Moss, H. J . , and Yates, J. R. 1960. A role for thiol and disulphide groups in determining the rheological properties of dough made from wheaten flour. Nature (London) 186, 45 I. Fridovich, I . , and Handler, P. 1957. Colorimetric assay for reaction of sulfhydryl groups with organic mercurials. Anal. Chem. 29, 1219. Friedman. M. 1973. ”The Chemistry and Biochemistry of the Sulfhydryl Group in Amino Acids, Peptides and Proteins.” Pergamon, Oxford. Friedman, M. 1974. Problems in assaying for half-cystine residues in proteins. Absrr. Pup. Am. Chem. S o ( , . , 168 AGFD 10. [Abstr.: FSTA 7, IA16 (1975)l. Friedrnann, E. 1952. Spectrophotometric investigation of the interaction of glutathione with maleimide and N-ethylmaleimide. Eiorhim. Eiophys. Acru 9, 65. Fuchs, F. 1971. The effect of Ca2+ on the sulfhydryl reactivity of troponin. Evidence for a Ca2+ induced confinnational charge. Eiochim. Eiophys. Acra 226, 453. Fujimaki, M., Arakawa, N., and Ogawa, G . 1961. Effects of gamma irradiation on the chemical properties of actin and actomyosin of meats. J . Food Sri. 26, 178. Gabay, S . , Cabral. A . M., and Healy, R. 1968. A chemical method for the estimation of brain sulfhydryl groups. Proc. Soc. Exp. Biol. Med. 127, 1081. Gabler. W.. and Scheuner, G. 1966. Specifity of dihydroxydinaphtyldisulfide (DDD) reaction. Acra Hisrochem. 23, 102. Gaff, D. F. 1966. The sulfhydryl-disulfide hypothesis in relation to desiccation injury of cabbage leaves. Ausr. J . Biol. Sci. 19, 291. Garbusov, V., Rehfeld, G., Wolm, R. V., Golovnaja, and Rothe, M. 1976. Volatile sulfur compounds contributing to meat flavour. I . Components identified in boiled meat. Nahrung 20, 235. Gawronski, T.H., Spencer, J. V., and Pubols, M. H. 1967. Changes in sulfhydryl and disulfide content of chicken muscle and the effect of N-ethylmaleimide. J . Agrir. Food Chem. 15, 781. Giacino, C. 1970. U.S. Patent 3,519,437. Cited in Wasserman (1972). Gilbert, J., Knowles, M. E., and McWeeny, D. J. 1975. Formation of C-and S-nitroso compounds and their further reactions. J . Sri. Food Agric. 26, 1785. Gimrnelreikh, N. G.. and Koval. Z. A. 1973. Effect of sarcolernma sulfhydryl group modification on activity of calcium- and magnesium-ion dependent ATPases. Ukr. Eiokhim. Zh. 45,587. [Chem. Ahsrr. 80, 46277a (19741.1
SULFHYDRYL AND DISULFIDE GROUPS IN MEATS
95
Glaser, C. B.. Maeda, H . , and Meienhofer, J. 1970. Chromatographic detection of thiols. disulfides. and thioesters with 5,5’-dithiobis (2-nitrobenzoic acid). J. Chromafogr. 50, 15 I . Goa, J . 1961. Micro determination of cysteine plus cystine in proteins. Aria Chem. Scand. IS, 853. Goldshtein, B. L.. and Khilko. 0. K. 1969. Change in the reactivity of sulfhydryl groups of proteins during aging of rats. Vesfn.Akad. Med. Nauk SSSR 24,3 I . [Chem. Absrr. 70, 1 1 3 0 0 8 ~(1969). 1 Golovkin, N. A., and Korzhemanova, L. 1973. Change in protein sulfhydryl groups during the storage of meat in a supercooled state (in Russ.). Myasn. Ind. SSSR 5 , 32. [Chem. Absrr. 79, 64718d 1973).) Golovkin, N . A.. and Meluzova, L. A . 1970. Changes of meat protein during long-temi storage in frozen state (in Russ.). Izv. VysS UP. Zav. Techn. 1, 65. Graevsky. E. Y., Konstantinova, M. M., and Tarasenko. A . G. 1969. Sulfhydryl groups and radiosensivity. S f u d . Biophys. 15/16, 163. Grassetti. D. R.. and Murray, J . F., Jr. 1969. Reagents for the spectrophotometric determination of sulfhydryl groups over a wide wavelength range. Anal. Chim. Acfu 46, 139. Grau. R. 1968. Einfliisse auf Beschaffenheit und Zusammensetzung des Fleisches. In “Handbuch der Lebensmittelchemie” (J. Schormiiller, ed.), Vol. 111/2. p. 1 I 10. Springer-Verlag. Berlin and New York. Gredser. M. L.. Yamaguchi. M., Brekke, C.. Potter, J., and Gergely, J . 1973. Troponin subunits and their interactions. Cold Spring Harbor S,yrnp. Quanf Biol. 37, 235. Greenwood. D. A,, Kraybill. H. R.. and Schweigert. B. S. 1951. Amino acid composition of fresh and cooked beef cuts. J . B i d . Chem. 193, 23. Gregory. J. D. 1955. The stability of N-ethylmaleimide and its reaction with sulfhydryl groups. J . Am. Chem. Soc. 77, 3922. Grill, H.. Patton. S., and Cone. J . F. 1967. Degradation of “5S-methionine to methyl mercaptan in surface ripened cheese. J . Agric. Food Chem. 15, 192. GrBschel-Stewart. U . , and Turba, F. 1963. “C-Markierung und Peptidkarten der SH-Regionen von Actomyosin, Myosin, Actin und H-Meromyosin. Biochem. Z . 337, 104. Grote, I . W . 1931. A new color reaction for soluble organic sulfur compounds. J . Biol. Chem. 93, 25. Gruen, L. C., and Harrap. B. S. 1971. Determination of thiols with a specific-ion electrode. Anal. Biochem. 42, 377. Griinewald, T. 1969. Untersuchungen iiber den EinfluD der Temperatur auf Bestrahlungsverluste von Cystin und Cystein in Lebensmitteln. Kalretech.-Klim. 21, 336. Gruenwedel. D. W.. and Patnik, R. K. 1971. Release of hydrogen sulfide and methyl mercaptan from sulfur-containing amino acids. J . Agric. Fund Chem. 19, 775. Gundlach, H. G.. Moore, S . , and Stein, W. H. 1959. The reaction iodoacetate with methionine. J . Biol. Chem. 234, I76 I . Habeeb. A. F. S. A. 1972. Reaction of protein sulphydryl groups with Ellman’s reagent. In “Enzyme Structure,” Pan B (C. H. W. Hirs, ed.), Methods in Enzymology, Vol. 25. p. 457. Academic Press. New York. Hamm, R. 1964. EinfluO der Gefriertrocknung auf die Qualitat des Fleisches. Dtsch. Lebensm: Rundsrh. 60, 97. Hamm, R.. and Hofmann, K . 1965. Changes in the sulfhydryl and disulfide groups in beef niuscle proteins during heating. Nature (London) 207, 1269. Hamm, R.. and Hofmann, K. 1966a. Schwefelhaltige Verbindungen des Fleisches. I. Uber die Problematik der Bestimmung von Sulfhydrylgruppen in FleischeiweiS. Z . Lebensm.-Unters.Forsch. 129, 143. Hamm. R.. and Hofmann, K . 1966b. Schwefelhaltige Verbindungen des Fleisches. 11. Bestimmung von Sulfhydrylgruppen in Muskelgewebe und Myofibrillen mit N-Athylmaleinimid. Z . Lebensm.-Llnters. -Forsch. 130, 85.
96
KLAUS HOFMANN AND REINER HAMM
Schwefelhaltige Verbindungen des Fleisches. 111. Bestimmung Hamm, R . , and Hofmann, K. 1966~. von Sulfhydryl- und Disulfid-Gruppen in Myofibrillen und Muskelgewebe mil Hilfe der amperometrischen Titration. Z . Lebensm.-Unters. -Forsch. 130, 133. Hamm, R., and Hofmann, K. 1967.Schwefelhaltige Verbindungen des Fleisches. IV. Bestimmung von Sulfhydrylgruppen in Myofibrillen mit Natrium-p-Chlorquecksilber(I1)benznat. Z . Lebensm.-Unrers. -Forsch. 136,7. Hamm, R., Hofmann, K.. Griinewald, T., and Partmann. W. 1975. Verandemngen yon Aminotransferasen und Muskelproteinen bei der Behandlung von Schweinefleisch mit ionisierenden Strahlen. Fleischwirfschuj? 55, 1 105. Harington, J. S. 1967.The sulfhydryl group and carcinogenesis. Adv. Cuncer RPS.10, 247. Harisch, G . , and Schole. J. 1974.Der Glutathionstatus der Rattenleber in Abhangigkeit vom Lebensalter und von akuter Belastung. 2. Nuturforsch.. Teil C 29, 261. Harland, J . W., Kauffmann, F. L., and Heiligman, F. 1967. Effect of irradiation temperature and processing conditions on organoleptic properties of beef and chemical yields in model systems. In “Radiation Preservation of Foods.” Advances in Chemistry Series, No. 65, p. 35. Am. Chem. Soc., Washington. D.C. Harrap, B. S., and Gruen, L. C. 1971,Application of a specific-ion electrode to the determination of disulfide groups in proteins. Anal. Eiochem. 42, 398. Hartshome, D. J., and Daniel, J. L. 1970.Importance of sulfhydryl groups for the calcium-sensitive response of natural actomyosin. Eiochim. Eiophys. Acru 223, 214. Hasselbach, W.,and Seraydarian. K. 1966. The role of sulfhydryl groups in calcium transport through the sarcoplasmic membranes of skeletal muscle. Eiochem. Z . 345, 1.59. Haurowitz,nF. 1950. Internal structure of globular proteins. In “Chemistry and Biology of Proteins,” p. 127. Academic Press, New York. Havir, J., Fidler, A,, and Husak, R. 1966. Nachweis von Schwefelverbindungen mit FluoresceinI .3.6.8.-tetraquecksiIbertetraacetat. Acru Chim. Acud. Sci. Hung. 51, 39.[Abstr.: Chem. Zentrulbl. 139, No. 21-2229(1968).] Hay, J. D., Currie, R. W . , and Wolfe, F. H. 1972.The effect of aging on physiochemical properties of actomyosin from chicken breast and leg muscle. J . Food. Sci. 37, 346. Hay, J. D., Currie, R. W., Wolfe, F. H., and Sanders, E. J. 1973.Effect of postmortem aging on chicken muscle fibrils. J . Food Sci. 38, 981. Heazlitt, R., Conway, G., and Montag, J. 1973. The role of actin sulfhydryls in actin-myosin interactions. Biochirn. Eiophys. Actci 317, 3 16. Hedin, P. A.. Kurtz. G. W.. and Koch. R. B. 1961. The chemical composition of beef protein fractions before and after irradiation. J . Food Sci. 26, 112. Heide, K. 1955. Funktion und Analyse von SH-Gruppen in Proteinen. Eehringwerk-Mirt. 30, 97. Heidtmann, R. H. 1966.Vorteile neuzeitlicher Konservierungsverfahren. Fleischwirrschuj? 46, 761 . Heidtmann, R. H. 1970.F-Werte fur konservierte und haltbare Nahrungsmittel. Fleischwirrschuji 50, 153. Herz, K. 0.. and Chang. S. S. 1970.Meat flavor. Adv. Food Res. 18, I . Hird, F. J. R., and Yates. J. R. 1961. The oxidation of cysteine. glutathionc and thioglycollatc by iodate, bromate, persulphate and air. J . Sci. Food Agric. 12, 89. Hoch, F., and Vallee, B. L. 1960. Sulfhydryl groups of some mono- and dimercaptans and yeast alcohol dehydrogenase. Arch. Eiochem. Eiophys. 91, 1. Hodges, R. S . , and Smillie, L. B. 1970.Chemical evidence for chain heterogenity in rabbit muscle tropomyosin. Eiorhem. Eiophys. Res. Commun. 41, 987. Hodges. R. S . , and Smillie, L. B. 1972.Cysteine sequence of rabbit skeletal tropomyosin. Can. J . Eiochem. 50, 330. Hofmann. K. 1964. Untersuchung des Einflusses der thermischen Behandlung von Fleisch auf die funktionellen Gruppen der strukturellen Muskelproteine. PhD. Thesis, Justus Liebig-Univ., Giessen, Giessen, Germany.
SULFHYDRYL AND DlSULFlDE CROUPS IN MEATS
97
Hofmann. K . 1965a. Uber den Nachweis von Sulfhydrylverbindungen mit Nitroprussidnatrium und Zinkchlorid. Jahresher. Bundesunsr. F1rischfiirsc.h. Kulmbach p. I 10. Hofmann, K . I965b. Neue Farbreaktionen auf Sulfhydrylgruppen. Ntirurivisspiisc.hirj?en 52, 428. Hofmann. K . 1966a. Zur Reaktionsfahigkeit von SH-Gruppen in Proteinen. Naruri~,isrensc.hufic.,t53, 432. Hofmann. K . 1966b. Uber den Nlhrwert des Fleisches und seine Versnderung beim Erhitzen Fleiwhwirrschuj? 46, I I2 I . Hofniann. K . 1967. Methode zur Bestimmung vim Schwefelwasserstoff in Fleisch. Juhresher. Bundrstin.71. F1rischf)rsc.h. Kulmbach p. S I . Hofmann. K . 1969. Anwendung von Kaliumjodid bei der amperometrischen Titration. Jahresher. Bunrlesunsr. Fleischfmch. Kulint~urhp. 66. Hofmann, K . 1970. Beeinflussung der amperometrischen Titration vim Sulfhydrylgruppen mit Silbernitrat durch Salze. 2. Anal. Chrin. 250, 256. Hufniann. K. I97 la. Sulfhydrylgruppen in Lebensmitteln und ihre Bestimmung. Mirrcilung.sh1. GDCh (Ges. Dlsrh. Chetn. I-Fuchgruppe Let?ensnlirr~,l~~hem. Gerii,htl. Chem. 25, 109. Hofniann. K . I97 Ib. Bestimmung von SH-Gruppen im Muskelgewebe durch indirekte amperometrische Titration mit Silbernitrat unter Anwendung von Kaliumjodid. Z . Lrhensm.-Unrcw. -Forsch. 147, 68. Hofmann. K. 1 9 7 1 ~Bestimmung . von Thiolgruppen nlit AgNO,, NAM und PCMB unter Anwcndung der amperometrischen Titration. Z . Anal. Chem. 256, 187. Hofmann. K . I97 Id. Untersuchung schwefelhaltigcr Verbindungen im Muskel und ihre Bedeutung bei der Lagerung und Verdrbeituny von Flcisch. Diskussionstag. For-srhungskr. Erniihrungsind. No. 30, p. 78. Hofmann. K . 1972a. Bestimmung des Sulfhydryl-Cehaltes in den myofibrillaren Muskelproteinen. Juhresher. Bundesansr. Fleisi~hforsch.Kulinbui~hp. 7 I Hofmann. K . 1972b. Ampawnetrisehe Titration der SH-Gruppen in Milch mit Silbernitrat. Nuhrung 16, 197. Hofmann. K . 1 9 7 2 ~ Der . EinfluB der Konservierungsverfahren auf die Qualitat von Fleisch und Fleischerzeugnissen. Flf,;~ch~,irr.~(.h~ij? 52, 1403. Hofmann. K . 1974a. Schwefelwasserstoffbildung in Fleischerzeugnihsen. Fleisc~hwirrschufi54, 1297. Hofmann. K . 1974b. Das Redoxpotential-sein Wesen und seine Bedeutung fur Fleisch und Fleischwaren. Fleischwirrsrhuj? 54, 465. Hofmann. K . 1 9 7 4 ~Eine . neue Methode zur Vitamin-B,-Bestimmung i n Fleischwaren. Juhresher. Bundesanst. Fleischfmch. Kulmbuch p. 48. Hofmann, K . , and Hamm, R. 1966. Sulfhydryl- und Disulfidgruppen des Fleisches und ihre Bestimmung. Flei.rc,h~virrsr,hu~ 46, I 125. Hofmann. K . . and Hamm, R. 1967a. Zur Bestimmung von Schwefelwasserstoff mit N,NDimethyl-p-phenylendiaminund Eisenchlorid. Z . Anal. Chem. 232, 167. Hofmann, K . , and Hamm. R. 1967b. Methode zur Bestimmung von Sulfhydryl- und Disulfidgruppen in Proteinen unter Anwendung verschiedener Reagentien. Z . A n d . Chem. 231, 199. Hofmann, K . . and Hamm. R. 1974a. Bestimmung von Sulfhydryl- und Disulfid-Gruppen in Proteinen mit Hilfe der amperometrischen Titration. I , Uber die Reaktionsfahigkeit von ProteinSH-Gruppen und die SH-Spezifitat von Silber-lonen. Z . Lrbensm.-Unters. -Forsch. 156, 100. Hofmann, K . . and Hamm. R. 1974b. Bestimmung von Sulfhydryl- und Disulfidgruppen in Proteinen mit Hilfe der amperometrischen Titration. 11. Kritische Betrachtung der Methoden zur amperometrischen Titration von Protein-SH-Gruppen mit Silbernitrat. Z . Lehensm-Unrers. -Forsch. 156, 139. Hofmann. K . , and Hamm, R. 1975. Bestimmung von Sulfhydryl- und Disulfid-Gruppen in Proteinen n i t Hilfe der amperometrischen Titration 111. Untersuchung der Spezifitat von Ag+-Ionen fur Protein-SH-Gruppen, 2. Lebensm.-Unters. -Forsch. 159, 205.
98
KLAUS HOFMANN AND REINER HAMM
Hofmann, K.. and Schael, U. 1966. EinfluB der Lagerdauer auf die Sulfhydrylgruppen des Rindfleisches. Jahresber. Bundrsansr. Flrischforsch. Kulrnbuch p. 50. Hofmann, K.. Schael. U . . and Hamm. R. 1969. EinfluR der Lagerung von Rind- u. Schweineflcisch auf den Gehalt des Gewebes an Sulfhydryl gruppen. Fleischwirrschufr 49, 1502. Hofmann. K., Mintzlaff, H.-J., Alperden. I., and Leistncr, L. 1971. Untersuchung iiber die Inaktivierung des Mykotoxins Patulin durch Sulfhydrylgruppen. Fleischwirtschufr 51, 1534. Hofmann. K.. Bliichel. E.. and Baudisch. K. 1974. lnvestigation of chemical changes during storage of frozen meat. Proc. Eur. M w r . Meat Rcs. Workers, 20rh, Dublin p. 262. Holbrook, J. J., Pfleiderer. G., Schnetgcr, J., and Diemar, S. 1966. The importance of SH-groups for enzymic activity. Biochrm. Z . 344, I . Holm, J. 1976. Untersuchungen auf den Gehalt an Blei und Cadmium in Fleisch- und Organproben bei Schlachttieren. Fleiwhwirrsc.hqfr 56, 413. Homsey, H. C. 1959. The colour of cooked cured pork. 111.-Distribution and relationship of pigments. pH and cysteine. J . Sci. Food Agric. 10, 114. Hornstein, 1.. Crowe, P. F . , and Sulzbacher, W. L. 1960. Constituents of meat flavor: Beef. J . Agric. Food Chem. 8, 65. Hupf, H . . and Springer, R. 1971. Bildung von Lanthionin in Nahrungsproteinen. Z . Lebensm: Linters. -Forsch. 146, 138. Husaini, S. A.. and A h . F. 1955. Denaturation of proteins of egg white and of fish and its relation to the hberdtion of sulfhydryl groups on frozen storage. Food Res. 20, 264. Jacobsen, A. L., and Henderson, J . 1973. Temperature sensitivity of myosin and actomyosin. Can. J . Biochem. 51, 71. Jaenicke. R. 1964. Zum Mechanismus der Warmeaggregation globdarer Proteine. Ber. Bunsenges. Phys. Chem. 68, 857. Jaenicke, R. 1965a. lntermolekulare Wechselwirkungen bei der Warmeaggregation globulkrer Proteine. Z . Naturforsch., Teil B 20, 2 I . Jaenicke. R. 1965b. Warmeaggregation und Warmedenaturierung von Proteinen. In “Wlrmebehandlung van Lebensmitteln,” DECHEMA-Monographie. No. 56. p. 207. Verlag Chemie, Weinheim. Jamieson, D., Ladner, K . , and van der Brenk, H. A. S . 1963. Pulmonary damage due to high pressure oxygen breathing in rats. 4. Quantitative analysis of sulphydryl and disulfide groups in rat lungs. Ausr. J . Exp. B i d . Med. Sci. 41, 491. Jeckel. D.. and Pfleiderer. U. G. 1969. Anderung der optischen Rotationsdispersion von LactatDehydrogenase aus Schweineherzmuskel nach Zugabe von Hamstoff, Coenzym und SHblockierenden Reagenzien. Hoppe-Seylrr’s Z . Physiol. Chem. 350, 903. Jermyn, M. A. 1966. lodoalkanesulphonates as thiol reagents. Aust. J . Chem. 19, 1293. Jocelyn, P. C. 1972. “Biochemistry of the SH group.” Academic Press, New York. Johnson, A. R., and Vickery, J. R. 1964. Factors influencing the production of hydrogen sulphide from meat during heating. J . Sci. Food Agric. 15, 695. Jones, 1. A., Phillips. J. W., and H i d , F. J. R . 1974. Theestimation of rheologically important thiol and disulfide groups in dough. J . Sci. Food Agric. 25, I . Kakat, B., and Vejdtlek, 2 . J. 1974. Thiole und strukturverwandte Schwefelverbindungen. In “Handbuch der photometrischen Analyse organischer Verbindungen,” Vol. I , p. 175. Verlag Chemie, Weinheim. Kakol, I. 1971. The location of the thiol groups of myosin that are protected against reaction with 2,4-dinitrophenyl P-hydroxyethyl disulphide. Biochem. J . 125, 261. Kakol, I . , Gruda, J . , and Bitny-Szlachto, S . 1964. A study on the role of SH groups of myosin by means of P-hydroxyethyl-2,4-dinitrophenyI disulphide. Acra Biochim. Pol. 11, 41 I . Kameyama, T.. and Sekine. T. 1973. The effect of F-actin on the reactivity of a specific sulfhydryl group (S,) in heavy meromyosin. J . Bbchem. (Tokyo) 74, 1283. Kameyama, T . , Ayakawa, A , , and Sekine, T. 1974. Effect of phenol derivatives and chemical
SULFHYDRYL AND DISULFIDE GROUPS I N MEATS
99
modification on the ATPase activities of heavy meromyosin and subfragment I . J . Biochem. (Tokyo) 75, 381.
Kanac.ka, Y.. Machida. M.. Ban, Y., and Sekine. T. 1967. Fluorescence and structure of proteins as measured by incorporation of fluorophore. 11. Synthesis of maleimide derivatives a. fluorescencelabeled protein-sulfhydryl reagents. Chem. Phurm. Bull. 15, 1738. Kanacika, Y.. Machida. M.. Kokubun. H.. and Sekine, T. 1968. Fluorescence and structure of proteins as measured by incorporation of fluorophore. 111. Fluorescence characteristics of N-[p-(2-Benzoxazolyl)phenyl]maleiniide and the derivatives. Chem. Pharm. Bull. 16, 1747. Kanaoka. Y.. Machida. and Sekine. T. 1973. Fluorescent thiol reagents. VI. N( 1 -Anilinonaphthyl-4)1naleImide;a fluorescent hydrophobic probe directed to thiol groups in protein. Biwhim. Biophys. Actu 317, 563. Kardashev. A. V.. Bobrovskaya. N. D.. Klyashtorin, L. B.. and Maslennikova. N . V. 1970. Biochemical changes in y-irradiated fish. Tr. Vses. Nuucho-Isslrd. Inst. Morsk. Rybn. Khoz. Okranogr. 73, 53. [Abstr.: FSTA 3, 5 R 215 (1971).1 Karmas. E., and DeMarco. G. R. 1970. Denaturation thermoprofiles of some proteins. J . Food Sci. 35, 725. Karush. F., Klinman. N . R.. and Marks. R. 1964. An assay method for disulfide groups by fluorescence quenching. A n d . Biochem. 9, 100. Katchalski. E.. Benjamin. G . S . . and Gross, V. 1957. The availability of the disulfide bonds of human and bovine serum albumin and of bovine y-globulin to reduction by thioglycolic acid. J . A m . Chetn. S o c 79, 4096. Kato. S . . Kurata. T.. and Fujinraki. M . 1973. Volatile compounds produced by the reaction of L-cysteinr or L-cystine with carbonyl compounds. Agric. B i d . Chem. 37, 539. Katz. M.. and Mommaerts. W. F. H. M. 1962. The sulphydryl groups in actin. Biochim. Biuphys. Actu 65, 82. [Abstr.: C h e x . Zentrulhl. 136, No. 9-1580 (1965).1 Kelley. G . G.. and Wans. B. M. 1957. Effect of reducing agents on curcd meat color. Food Techno/. 11, 114. Khan, A. W. 1965. Objective evaluation of quality in poultry meat. Nuture (London) 208, 204. Khan. A. W. 1966. Cryochemistry of animal tissue. Cryobiolugy 3, 224. Khan. A. W., and Nakamura, R. 1971. Quality and biochemical changes during frozen storage of meat from epinephrine-treated and untreated chickens. J . Food Sci. 37, 145. Khan. A. W.. and van den Berg, L. 1964. Some protein changes during post-mortem tenderization in poultry meat. J . Food Sci. 29, 597. Khan. A. W., and van den Berg. L. 1965. Changes in chicken muscle proteins during cooking and subsequent frozen storage. and their significance in quality. J . Food Sci. 30, 151. Khan. A. W.. van den Berg. L., and Lentz, C . P. 1963. Effects of frozen storage on chicken muscle proteins. J . Food Sci. 28, 425. Khan. A. W.. Davidkova. E.. and van den Berg. L. 1968. On cryodenaturation ofchicken myofibrillar proteins. Cryobiology 4, 184. Kielley. W. W., and Barnen. L. M. 1961. The identity of the myosin subunits. Biochim. Biophys. A C ~ U51, 589. Kiermeier. F.. and Hamed, M. G. E. 1962. Zur Bestimmung der Sulfhydrylgruppen in Milch und Milchprodukten. Nahrung 6 , 639. King, T. E., and Morris, R. 0. 1967. Determination of acid-labile sulfide and sulfhydryl groups. In “Oxidation and Phosphorylation” (R. W. Estabrook and M. E. Pullman, eds.), Methods in Enzymology, Vol. 10, p. 634. Academic Press, New York. Kirsten. E., and Kuperman, A. S. 1970a. Effect of sulfhydryl inhibitors on frog sartorius muscle: p-chloromercuribenzoic acid and p-chloromercuribenzene sulfonic acid. Br. J . Pharmacol. 40, 814. Kirsten, E., and Kuperman, A. S. 1970b. Effect of sulfhydryl inhibitors on frog sartorius muscle:
1.00
KLAUS HOFMANN AND REINER HAMM
N-ethylmaleimide. Br. J . Pharmacol. 40, 827. Klotsch, H., and Bergmeyer. H. U. 1962. Glutathion. In "Methoden der enzymatischen Analyse" (H. U. Bergmeyer. ed.). p. 363. Verlag Chemie, Weinheim. Klotz, 1. M.. and Carver, 8 . R. 1961. A spectrophotometric titration for the determination of sulfhydryl groups. Arch. Eiochem. Biophys. 95, 540. Kofman. E. B. 1963. With silver titratable sulfhydryl groups in actomyosin gel at syneresis. Biokhimiya 28, 774. [Chrm. Abstr. 60, 4364a (1964).] Kohno, K. 1965. Interaction of thiamin or related compounds with proteins. XI. Micro-determination of protein SH and low molecular SH groups by means of different thiamin disulfides as reagents. Viramins 31, 470. [Abstr.: Chem. Zentralbl. 137, No. 20- 1489 (1966).] Kohno, K . 1966. Microdetermination of protein-SH and low molecular SH-groups using various thiamine disulfides as reagent. J . Vitaminol. 12, 137. Kolodziejczyk, J. 1965. Entstehung und Verschwinden von Sulfhydrylgruppen bei der Gefriertrocknung von Fleisch (in Pol.). Przem. Spozyw. 19, 31. [Abstr.: Chem. Zenrralbl. No. 20, 3131 (1966).] Kolthoff, 1. M., and Harris, W. E. 1946. Amperometric titration of mercaptans with silver nitrate. Ind. Eng. Chem., Anal. Ed. 18, 161. Kolthoff, 1. M., and Lingane, J . J . 1952. "Polarography." Vols. 1 and 2. Wiley (Interscience), New York. Kolthoff, I . M.. and Shore, W. S. 1964. The effect of excess mercury chloride on serum albumin sulfhydryl stability at pH 2. Pro 18535 >~ I
I55 C ~ p ) r ~ g h(c.8 t I Y ~ Hh) Ac.idmiiL P w , . Inc \II right, c,f ~cpnrdu~imn in form ruaewctl
ISBN O-I?-illh4?4-8
156
WALTER M . URBAIN
radiation was the Atomic Energy Commission (A.E.C.), which had a program starting in 1950. The U.S. Army, having had difficulties with the troop acceptance of canned meats in two wars (corned ‘‘Willy” in World War I and “Spam” in World War 11) began its program in 1953. The objective of the Army work was to obtain shelf-stable meats of eating quality superior to what could be obtained by thermal processing. The A.E.C. program was discontinued after a few years, but was reinstated in 1960. Starting in 1960, by agreement, the Army program was limited to high-dose irradiation aimed at sterilized foods with unlimited shelf stability, whereas the A.E.C. program was concerned only with low-dose applications . Initially, the government program was carried out largely by contract work done in universities, commercial organizations, and existing government laboratories. While the A.E.C. program continued in this way, in 1963 the Army established a research facility at Natick, Massachusetts. About 1971, the A. E. C. terminated its program and the Army became essentially the sole United States Agency with a food irradiation activity. Since the Army program is largely an “in-house” activity, and with the termination of the A.E.C. program, funds to support university and other outside laboratory work in food irradiation have not been available. As a consequence, what had been a very large academic activity virtually has disappeared. Over the years, the American food industry had participated in the development of food irradiation to varying degrees. Initially the interest was strong, but, as problems were uncovered and it became apparent that there would be little opportunity for early commercial use of the process, this participation diminished and today industrial activity in the development essentially is nonexistent. Other countries joined in the work. England undertook an extensive program starting in the late 1940s. Programs were started also in Canada and Japan in 1956, in Argentina and the U.S.S.R. in 1957, in Poland in 1958, in India in 1959, and in Israel in 1960. In 1968, the U. S . Department of Commerce (Anonymous, 1968a) listed 76 countries which had food irradiation programs. Quite naturally groups of countries joined together. In Europe, the Commission of European Communities and the Organization for Economic Cooperation and Development both sponsored research and assisted in information transfer. In 1964, a joint activity in food irradiation of the Food and Agriculture Organization and the International Atomic Energy Agency was established. It has greatly assisted in making information on food irradiation known on a world basis, especially among the less developed countries. Among other activities, this joint FAOlIAEA division has set up regional projects of research in Asia and South America. The Council for Mutual Economic Assistance brought together the Eastern bloc countries of Europe. In 1970, approximately 25 nations banded together under the “International Project in the Field of Food Irradiation” to carry out work to establish the safety of irradiated foods for human consumption.
FOOD IRRADIATION
157
A number of countries have given approval for the irradiation of certain foods. Table I lists the approvals by food and by country which now exist or which had been given at one time. Despite these approvals, there has been little commercial use of food irradiation. Actually, only the irradiation of white potatoes to prevent sprouting has been practiced. In 1965, nearly one million pounds of potatoes were irradiated in Canada. Since 1973, potatoes grown in Japan have been irradiated. While other countries appear to be taking steps also to irradiate potatoes (e.g., Chile), other kinds of applications, despite regulatory approval, have not occurred. Unsatisfactory economics as, for example, with the case of wheat, have been a major factor in most instances. Many countries have established requirements that there must be satisfactory evidence of safety for human consumption of irradiated foods before they can be made available to the public. This need was recognized early and for over 25 years the obtaining of such evidence has occupied a very great proportion of the toial research effort. Other food processes have gained acceptance largely as a result of long-term usage which pragmatically has demonstrated safety for use with human foods. In general, they have not been subjected to definitive scientific studies to evaluate their safety. From almost the start, food irradiation, however, was considered by many as “suspect,” probably due to an unavoidable, but mistaken, association with the atom bomb and with the lethal effects of ionizing radiation on living organisms. Food irradiation also has become involved in the current general concern for the safety of processed foods and of food additives. As a consequence, the process of food irradiation and irradiated foods have been studied in terms of possible health hazards for consumers in a manner and to a degree that has not occurred with any other food process. In the United States radiation is classified as a food additive and thereby subject to regulations of the Food and Drug Administration. The United States has had a position of leadership in establishing the requirements for acceptable evidence of safety of irradiated foods. The issue of safety came to a climax in 1968 when a petition of the Army for the radiation sterilization of ham was regarded as not providing sufficient evidence of safety. A previously issued regulation permitting the irradiation of sliced bacon was withdrawn at that time. The effects of these actions by the FDA were manifold both in this country and elsewhere. Commercial organizations in the United States lost interest in food irradiation. Undoubtedly this situation was a major factor in the termination of the A.E.C. program in 1971. Other countries retrenched in their activities. Some continued, but with increased difficulty. Some attacked the actions of the FDA and, in particular, opposed the designation of radiation as a food additive and the requirements for the separate evaluation of each and every irradiated food. They sought to treat food irradiation just as any other food process.
TABLE I GENERAL SURVEY OF IRRADIATED FOOD PRODUCTS CLEARED FOR HUMAN CONSUMPTION IN DIFFERENT COUNTRIES"
Type and source of radiation Country (organization) Bulgaria
Canada
Product Potatoesh Potatoesh Onionsh Garlich Grainh Dry food concentrates') Dried fruitsh Fresh fruits" (tomatoes, peaches. apricot. cherry. raspberry, grapes) Potatoes' Onions Wheat, flour, whole wheat flour Poultry"
Cod and haddock fillets" Chile Pot atoes"." Denmark Potutoes France PotatoesY Federal Republic Deep-frozen mealsbJ of Germany Potatoesb
Purpose of irradiation Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition Insect disinfestation Insect disinfestation Insect disinfestation
finCo
ixcS
Electrons
Dose (bad)
Date of approval
+
+ +
10
100 100
30 April 9 November 14 June 25 March
1972 1960
25 February
1969
20 June 2 October 3 1 October 27 February 8 November 24 March 26 September
1973 1973 1974 1970 1972 1972 1974
10 10
+ + + +
30
Radurization Sprout inhibition
+
Sprout inhibition
+
250 10 max. IS max. 15 max.
Insect disinfestation Radicidation (Salmonella) Radurization Sprout inhibition Sprout inhibition Sprout inhibition Radappertization Sprout inhibition
+
75 max.
+ +
700 max. I50 max.
+
+ + +
+
10 MeV
1971 1972 1972 1972 1972 1972 1972
30April 30 April 30 April 30 April 30 April 30April
15 max. 7.5-15 25004500 15 max.
1963
1965
Hungary
Israel
Potatoes'' Potatoes" Potatoes" Onions" Onions" Strawberries" Mixed spices" (black pepper, cumin paprika, dried garlic: for use in sausages) Poraroes Onions
Italy
Japan Netherlands
Poruroes Onions Garlic, Poraroes Asparagus" Cocoabeans" Strawberries" Mushrooms Deep-frozen meals' Potaroes Shrimps* Onions" Onion.\ Spices and condiments" Poultry. eviscerated (in plastic bags) Chicken
Fresh, tinned and liquid foodstuffs' \o
Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition Radurization
Radicidation Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition Radurization Insect diainfestation Radurization Growth inhibition Radappertization Sprout inhibition Radurization Sprout inhibit ion Sprout inhibition Radicidation Radurization Radurization, Radicidation Radappenization
+
10
+
15 max. 15max.
+
+ +
6
+
+
500
+
15 max. 10 max.
+
+
+
+ + + +
+
+
4 MeV 4 MeV 4 MeV 4 MeV 4 MeV
4 MeV
7.5-15 7.5-1s 7.5-15 15 max. 200 max. 70 max. 250 max. 250 max. 2500 min. 15 max. 50-100 15 mdX. 5 max. 800-1000
+ +
300 max. 300max.
+
2500 min.
23 December 10 January 5 March 5 March 6 August 5 March
1969 1972 1973 1973 1975 1973
2 April 5 July 25 July 30 August 30 August 30 August 30 August 7 May 7 May 7 May 23 October 27 November 23 March I3 November 5 February 9 June 13 September
1974 1967 1963 1973 1973 1973 1972 1969 I %9 1969 1969 1%9 I970 1970 1971 1975 1971
31 December 1971 10 May 1976
8 March
1972
(continued)
s
TABLE I - (continued)
Type and source of radiation Country (organization)
Product
Spices'.P Vegetable fillingh," Powdered batter-
Philippines South Africa Spain
Endive"," (prepared, cut) Potatoes" Mangoes" Potatoes Potatoes
Onions Thailand Union of Soviet Socialist Republics
Onions
Potatoes
Poraroes Grain Fresh fruits and vegetables" Semiprepared raw beef. pork and rabbit products (in plastic bags)*
Purpose of irradiation Radicidation Radicidation Radicidation Radicidation Radurization
'To
'"Cs
+ +
Electrons 4 MeV 3 MeV
+
+ +
Sprout inhibition Control of ripening Sprout inhibition Sprout inhibition Sprout inhibition Sprout inhibition
+ + + +
Sprout inhibition Sprout inhibition Insect disinfestation
+
+
Radurization
+
Radurization
+
Dose (krad)
loo0 75
4 October 26 June 4 October
1974 1975 1974
I50 100
4 October 14 January
1974 1975
lo00
15 max. 75-125 12-24
t
5-15 8 max. 10max.
+
+
10 I MeV
Date of approval
13 September 1972 6 September 1976 19 January 1977 4 November 1969 1971 20March 1973
14 March 17 July
1958 1973 1959
200400
1 I July
1964
600-800
I 1 July
I964
30 30
Dried fruits Dry food concenrrures (buckwheat mush, gruel, rice pudding) Poultry, eviscerated (in plastic bags) Culinary prepared meat products (fried meat. entrecote) (in plastic bags)* Onions" Oniuns United Kingdom
United States of America
Uruguay
-
0'
Any food for consumption by patient5 who require a atenle dietSI . an essential factor in their treatment
Insect disinfestation
100
Insect disinfestation
70
6 June
I966
600
4 July
I966
Radurization
Radurization Sprout inhibition Sprout inhibition
+
800
+
6 6
+
Radappenization
I February 25 February 17 July
I966
1967 1967 I973
1 December I969
Wheui mid urhetii ,flour (changed on 4 March 1966 from wheat and wheat Insect disinfestation product)
+
Whire poturoes
Sprout inhibition
+
Sprout inhibition
+ +
Potntoes
IS February
+ 5 MeV
+ +
20-so 20-50 20-so 5-10 5-10 5-15
21 August 2 October 26 February 30 June 2 October I November 23 June
I963 I964 I966 I964 I964 I965 1970
(continued)
TABLE I - (continued) Type and source of radiation Country (organization) World Health Organization
Purpose of irradiation
Product
Potatoes' Potatoes Onions" Papaya Stra wberries Wheat and ground wheat productsp Wheat and ground wheat products Riceg Chicken Cod and redfish'
Sprout inhibition Sprout inhibition Sprout inhibition Insect disinfestation Radurization
+
l"Cs
Electrons
+
Dose (had)
15 max.
Date of approval
12April 7 September 7 September 7 September 7 September
1%9 1976 1976 1976 1976
12 April
1969
+ +
+ +
+ +
10 MeV 10 MeV 10 MeV 10 MeV
Insect disinfestation
+
+
10 MeV max.
Insect disinfestation Insect disinfestation Radurization Radicidation Radurization Radicidation
+ +
+ +
10 MeV max. 10 MeV max.
15-100 10-100
7 September 1976 7 September 1976
+
+
10 MeV max.
200-700
7 September 1976
+
+
10 MeV max.
200-220
7 September 1976
" Compiled by K. Vas, International Atomic Energy Agency, Vienna Experimental batches. Italicized products indicate unlimited clearance. Test-marketing. Temporary acceptance. Hospital patients. Provisional.
'
T o
+ +
maxy max max. max.
3-15 2-15 50-100 100-300 75 max.
FOOD IRRADIATION
I63
The safety evaluation requirements relied heavily o n animal feeding studies and had become large and complex. As timc progressed, a knowledge of much of the radiation chemistry of irradiated foods had been acquired, and, with this knowledge, the claim was advanced that animal feeding studies were of less importance than was generally regarded and that similar foods responded t o radiation in similar ways. In fact it was stated that the radiolytic changes in foods were so small as to be undetectable by the required animal feeding studies. These viewpoints were advanced in both the United States and elsewhere and reached a focus when, in 1976, the question was referred t o a joint expert committee of the Food and Agriculture Organization, the Intcrnational Atomic Energy Agency, and the World Health Organization. As proposed by the International Project in the Field of Food Irradiation, this committee accepted the concept that food irradiation is a process (as opposed to the concept that radiation is a food additive). The same FAO/IAEA/WHO expert committee approved as “unconditionally safe” for human consumption irradiated potatoes, wheat, chicken. papaya, and strawberries. “Provisional” approval was given to rice, fish, and onions (see Table I). In the United States at least two foods, papaya and beef, shortly will be the subjects of petitions to the FDA. In the Netherlands and Denmark as a result of an animal feeding study on a totally irradiated diet, determination of acceptance o f radiation as a food process will be made. All the recent and impending actions are likely to result in a resolution of the question of the safety of irradiated foods for human consumption. The favorable FAO/IAEA/WHO action, both in accepting food irradiation as a process and in approving the indicated foods, encourages hope that other favorable actions will follow. It appears probable, therefore, that the major hold-up to the use of food irradiation will be overcome in the next few years. Once the process is available to commercial interests, its use will be determined by the conventional factors appropriate to food processes generally, namely, utility in fulfilling needs and opportunities. and economics.
II. RADIATION AND RADIATION SOURCES Chemical change in vital parts of living organisms such as food spoilage bacteria may result in their death. It is the capability of ionizing radiation to accomplish chemical changc that is the key t o most of thc particular applications of irradiation to foods. In order to break chcmical bonds, the energy level of the radiation must be sufficiently great. Typical covalent chcniical bond energies lic in the range of I to 8 electron volts.*: As thcse energies are less than the energy ‘:011rclrctron volt. e V . equals I .6 x 1 0 ~
L ‘ I ~
164
WALTER M . URBAIN
of ionization of an orbital electron, all types of ionizing radiation can break covalent bonds. Particles such as electrons can be accelerated to energies sufficient to break bonds and can be used alternatively to electromagnetic radiation. Since only chemical change is desired, it is necessary to limit the energy level of the radiation employed so as to be less than that which will accomplish nuclear change in the elements in the food and cause it to become radioactive. This limitation of energy level is accomplished by setting limits of 5 MeV for electromagnetic (gamma o r X-ray) radiation and 10 MeV for electrons (Anonymous, 1973). For this reason, also, other particles such as alpha particles or neutrons are not employed in food irradiation. Avoidance of induced radioactivity in irradiated foods is a basic requirement of the process. An important characteristic of the radiation used is its penetrability into the food. If it is to accomplish its purpose, it must reach those molecules which need to be affected. Thus for a sterilized food, all parts of it must absorb sufficient radiation to kill all spoilage microorganisms present. The penetration of X-rays and of gamma rays is a function of their energy levels. The maximum energy limit of 5 MeV for these rays provides adequate penetration for practical applications. The maximum of 10 MeV for electrons, however, restricts applications to foods less than about 5 cm thick (two-sided irradiation). For gamma ray sources, two radionuclides have been used in food irradiation: (1) Cobalt-60 (“Co) produced by neutron irradiation of 59C0and (2) Cesium-] 37 (137Cs)produced by separation from fission products. ‘j0Cohas a half-life of 5.27 years and gives gamma rays of 1.17 and 1.33 MeV. 13’Cs has a half-life of 30 years and gives a gamma ray of 0.66 MeV. For electron beam sources, the linear accelerator has proved most useful. Energies up to the 10-MeV limit can be obtained without difficulty. Irradiation facilities provide for treatment of foods under controlled conditions. In any given application, the amount of radiation is controlled by knowing the rate of energy output of the source, by controlling the physical relationship (mainly distance) between the source and target material, and by controlling the time of treatment. The amount of energy absorbed is termed the “dose” and usually has been measured in rads.* Dosimeters are devices to measure the dose. A number of kinds have been devised. The basic one is the Fricke dosimeter. It utilizes the oxidation of Fe 2+ to Fe3+ in a standardized aqueous solution. There are other chemical dosimeters, some of which are available as convenient plastic materials which change color on exposure. Some dosimeters are electrical devices. A calorimeter of appropriate design may be used, especially in cases where the energy output rate is large, as with the linear accelerator. *One rad equals 100 ergs absorbed per gram of absorber. The International System of Units (SI) replaces the rad with the Gray (Gy). One Gy equals I 0 0 rad equals one joule/kg.
FOOD IRRADIATION
165
The irradiation facility generally has three essential components: ( 1 ) the radiation source, (2) a cave or similar structure to confine the radiation within a given space in order to afford protection to personnel during irradiation, and (3) equipment to take the material to bk irradiated to the source. In the case of radionuclide sources, some arrangement usually is provided to permit storage of the source in a safe manner when it is not in use. This usually takes the form of a “pool” of water into which the radionuclide source is lowered and which has sufficient depth to reduce the radiation at the surface to a safe level. Since machine sources can be turned off, a comparable storage facility is not needed for them. The equipment to carry the food to the source usually consists of a conveyor on which the food may be placed outside the cave, and which carries the food to the source area for exposure, and then outside the cave where it is removed from the conveyor. Suitable controls are provided outside the cave for the needed oper-
FIG. I, Plan of Food Irradiation Research Laboratory, U . S. Army Natick Research and Development Command, Natick. Massachusetts.
166
WALTER M . URBAlN
FIG. 2. Irradiation area of linear accelerator. Food Irradiation Research Laboratory, U. S . Army Natick Research and Development Command, Natick. Massachusetts.
FIG. 3 . A typical food irradiator utilizing "Co. Courtesy o f Atomic Energy of Canada, Limited.
FOOD IRRADlAT10N
167
FIG. 4. Mobile irradiator. The boxed food niovcs through 3 positions on each side of the T o plaque. The entire unit is self-contained. Courtesy of Atomic Energy of Canada, Limited.
FIG. 5 . Plan of commercial potato irradiation facility in Hokkaido, Japan. Courtesy Kawasaki Heavy Industries, Ltd.
I68
WALTER M . URBAIN
FIG. 6. Wobalt potato irradiation facility in Hokkaido, Japan. Capacity 15,000 tons per month. Courtesy of Kawasaki Heavy Industries, Ltd.
ations, which can be made essentially automatic. In the area of the radiation source, the target material may be conveyed in a complex manner in order to obtain a relatively uniform dose distribution throughout it and to improve source efficiency. This kind of movement is appropriate to radionuclide sources since gamma rays cannot be directed or focused. Electron beams are easily directed. Because of the ability to direct electron beams, these sources are more efficient in terms of usefully absorbing their output of ionizing radiation than are radionuclide sources. An efficiency of 50% is considered extremely good for a radionuclide source. In order to gain a practical level of source efficiency with gamma rays especially, it is necessary to accept some latitude in dose distribution within a target material. It is virtually impossible to obtain a narrow range (e.g., 5%) between minimum and maximum dose and have source efficiencies that are realistic in terms of commercial requirements. In at least some cases, a 50% variation of dose should be considered acceptable. Costs, especially capital costs, are directly related to source efficiency. Figures 1 through 6 illustrate experimental and production irradiation facilities .
111.
GENERAL EFFECTS OF RADIATION A.
FOODS
As indicated earlier, ionizing radiation can cause chemical change. If a chemical change occurs in molecules important to the life processes of an organism,
FOOD IRRADIATION
169
there may be biological consequences that will manifest themselves in various ways. depending upon the nature of the organism, the degree and location of the damage, and environmental factors. For example, if the chemical changes involve DNA molecules in the cells of the organism, normal functioning may not be possible. Thus, ionizing radiation affects all forms of life. In food irradiation, advantage is taken of this action. Foods contain various organisms as contaminants: bacteria, yeasts, molds, helminths. and insects. These organisms can change a food, and in many cases, we term these changes as “spoilage.” Many preservation processes have as their objective the control of spoilage microorganisms. Some foods, for example, fresh fruits and vegetables, are themselves living organisms. Radiation can also affect their life processes. In some cases, the changes are useful in extending the period before senescence deteriorates the food. The general ways in which irradiation can be useful in treating foods may be listed as follows: Control of spoilage microorganisms Complete sterilization for unlimited product life Reduction of numbers to delay microbial spoilage Control of food-borne pathogenic microorganisms Control of helminths and other food-borne parasites Control of insects Delay of senescence Product improvement Of these, only the last does not involve affecting a life process, and it is not normally involved in the preservation of food. It generally is concerned with improving a functional property of a food or foodstuff. There is a large variation in the dose needed in connection with the above individual effects and part of the development of particular applications is the determination of dose requirement.
B. ACTION ON MAJOR FOOD COMPONENTS 1. Proteins
Reviews on the radiation chemistry of proteins and related compounds are available (Garrison, 1972; Urbain, 1977). In general, the effect of radiation on proteins is not great at the doses employed in food irradiation. Regardless of origin, protein molecules tend to respond to radiation similarly. To some degree, the nature of the change is related to the particular structure of a protein (namely, fibrous, globular), whether native or denatured, its composition, to the presence of other substances, and its state (wet, dry, in solution, or whether liquid or frozen).
170
WALTER M . URBAIN
Because they are large molecules, proteins generally provide a number of loci for action by radiation. Although the energy may be absorbed at one location, it can transfer to another within the molecule, where, at what might be termed a “sensitive site,” bond breaking occurs. Atoms, or groups of atoms, may split off and form free radicals. In this way, radiation acts upon large complex molecules, such as proteins, in a characteristic rather than random manner. The free radicals formed ultimately disappear. At lower temperatures, diffusion is limited and recombination of radicals is more likely. At higher temperatures, reaction with different species is more probable. The condition of a protein before irradiation affects the end results. Irradiation of a denatured protein leads to a higher level of free radical formation since its disrupted structure has less capability for recombination. Indirect action of radiation plays a very important role when water is present, unless it is “bound” or frozen. In the presence of free liquid water, there may be rupture of hydrogen bonds with consequent unfolding of the molecule, or there may be aggregation or dissociation into smaller units, or there may be fragmentation. In certain of these changes, chemically active groups may be made more available or they may be altered so as to be essentially nonexistent. Changes such as have been indicated above may alter the normal properties of a protein. It may, for example, become denatured. Enzymes no longer may be active. Chromoproteins may be changed in color. Functional properties (as in foods) may be altered. Nucleoproteins may lose their function in biological processes, which may affect the organism of which they are a part. As nutrients, proteins serve primarily as sources of amino acids. Radiation can cause amino acid destruction. At doses employed in food irradiation, however, amino acid values are virtually unchanged, and, as a consequence, proteins suffer no measurable nutritional losses.
2. Lipids Changes resulting from the irradiation of lipids may be grouped as (a) gross changes in physical and chemical properties, (b) autoxidative changes, and (c) nonoxidative radiolytic changes. a. Gross Changes. Below 5 Mrad, there are only very slight changes in the usual indexes for fat quality. At doses between 10 and 100 Mrad, there are significant increases in acid number, trans-fatty acid content, peroxide values, melting point, refractometric and dielectric constants, viscosity and density. Shifts in double bond position occur. Flavor changes in meat fat occur with doses as low as 2 Mrad. In milk fat, a “chalky” or “candle-like” flavor develops. An off odor in fish lipids has been ascribed to oxidative rancidity of unsaturated fatty acids.
171
FOOD I R R A D I A T I O N
b. Autoxidative Changes. As evidenced by electron-spin resonance (ESR) measurements, irradiation produces free radicals in fats. The types of free radicals formed and their decay rates are influenced by temperature. They are more stable at lower temperatures. If exposed to 02, they can react and form new free radicals, such as peroxide radicals (Farmer et al., 1942). The reaction with O2 can occur over an extended period of time after irradiation. The radiation-induced autoxidation process follows the same path as the usual one for fats and, through a free radical chain mechanism, yields hydroperoxides which decompose into a variety of products such as aldehydes, aldehyde esters, oxoacids, hydrocarbons, alcohols, ketones, hydroxy and ketoacids, lactones, and dimeric compounds. Irradiation accelerates the autoxidative process.
c . Nonoxidative Rudiolytic Chunges. The major compounds formed when a saturated fat is irradiated in the absence of 0, are H,, CO,, CO, a series of hydrocarbons (n-alkanes and alkenes) and an aldehyde. In general a similar pattern is obtained with unsaturated fatty acids. However, the presence of one or more double bonds causes the formation of other radiolytic unsaturated compounds. Also, some hydrogenation occurs and produces a saturated fatty acid. Significant amounts of dimers are formed. A general radiolytic mechanism has been given by Nawar (1972, 1977). In a triglyceride molecule a
L O C
d I
I
I
I
c-0-0-c,
I
c-0-0-c,,
where n is the number of carbon atoms in the component fatty acid, cleavage occurs preferentially at five locations (a, b, c, d, and e) and randomly at all the remaining carbon-carbon bonds of the component fatty acids. The resulting free radicals are terminated principally by hydrogen abstraction (from other molecules) and, to a lesser extent, by hydrogen loss or by combination with other free radicals. As a consequence, a number of radiolytic compounds are formed. With due allowance for initial composition, natural fats yield essentially the same compounds as do model systems (e.g., fatty acids, pure triglycerides. and esters). The volatile substances identified in beef fat irradiated with various doses at 25°C are given in Table 11. The same substances are formed at all doses, but the amounts are in proportion to dose. Whether a fat is irradiated in the solid or liquid state affects the relative amounts of radiolytic products formed. The compounds produced by heating a fat are quite similar to those obtained by irradiation. There are, however, both qualitative and quantitative differences.
172
WALTER M . URBAIN
TABLE
n
QUANTITATIVE ANALYSES (mg/kg) OF THE VOLATILES FORMED IN BEEF FAT BY
IRRADIATION"."
Concentration of volatiles (mglkg) at Mrad Compound n-Propane I-Propene n-Butane I-Butene n-Pentane I-Pentene n-Hexane I-Hexene n -Heptane I-Heptene n-Octane I-Octene n-Nonane I-Nonene n-Decane I-Decene n-Undecane I-Undecene n-Dodecane I-Dodecene n-Tridecane I-Tridecene Tridecadiene n -Tetradecane I-Tetradecene Tetradecadiene n-Pentadecane Pentadecene (int.)c I -Pentadecene Pentadecadiene n-Hexadecane Hexadecene (int.)r I-Hexadecene Hexadecadiene n -Heptadecane Heptadecene (int.)' I-Heptadecene Heptadecadiene "
0.5
0.40 0.07 0.34 0.18 0.50 0.02 0.06 0.01 0.26 0.07 0.30 0.05 0.81 0.12 0.49 0.22 0.40 0.16 0.37 0.67 0.95 0.67 trace 0.55
4.23 0.26 2.85 0.48
1
2
3
0.82 0.22 0.88 0.I6 0.89 0.04 0.28 0.05 0.58 0.22 0.71 0.29 0.54 0.28 0.69 0.38 0.62 0.23 0.53 1.08 1.60 0.87
1.04
I .72
0.38 I .72 0.50 0.96 0.39 1.27 0.70 1.08 0.54 0.87 1.74 2.93 1.41
1.76 0.56 1.48 0.49 1.66 0.91 1.32 0.67 1.22 2.49 3.79
trace
trace
0.90 6.43
1.31 14.00 0.86 10.40
0.50
4 1.68 0.21 I .40 0.16 1.14 0.07 0.86 0.29 2.20 0.65 2.44 0.75 1.99 0.74 2.30 1.24 1.97 0.88 1.67 3.61 6.02 3.12
5
6
2.35 4.62 8.27 3.24
2.98 0.26 2.52 0.17 2.23 0.15 1.74 0.36 4.02 1.14 5.30 1.54 3.74 1.01 4.39 I .92 3.26 I .30 2.96 6.18 1.31 4.38
2.84 0.68 3.51 0.81 2.76 0.83 3.11 I .60 2.57 1
.oo
1.51 trace
trace
trace
trace
1.58 15.40 1.22 14.80 I .60 2.42 1.06 0.96
2.53 24.20 I .66 24.50 2.11 3.50 1.68 1.82
3.37 31.10 2.02 29.40 2.59 4.28 1.93 2.30
4.75 37.80 3.07 36.25 3.88 7.29 2.67 4.50
0.26 0.60
5.04 0.48 0.97 0.41 0.78
trace
trace
trace
trace
trace
trice
trace
4.18
5.84 7.39 5.00 5.44 1.12 1.33
7.78 12.60 8.22 5.94 1.72 2.63
14.36 18.54 12.55 15.62 2.22 3.10
23.50 29.10 21.40 25.90 1.98 4.07
27.70 35.20 21.90 27.00 4.32 6.35
31.05 40.40 26.90 32.90
0.64
5.64
4.18 3.78 trace
0.96
0.5 to 6.0Mrad at 25°C. From Nawar (1977). in1 = Internally unsaturated.
1.01
1.95 0.73 1.02
5.10
8.20
FOOD IRRADIATION
173
For example, in the irradiation of triolein, hexadecadiene is formed. Heating triolein results in ethyl-, propyl-, pentyl-, and hexylcyclohexenes, which are absent in irradiated triolein. The amount of pentane produced by irradiation of tricaproin is nearly twice that formed by heat. 3. Carbohydrates
Carbohydrates are components of many foods. They also are available as isolates. The response to radiation varies with the circumstances in which the carbohydrate exists. Pure carbohydrates are very sensitive to radiation when in the crystalline state, and give a response which is dependent upon the particular crystalline form irradiated. Imperfections of the crystalline lattice, as for example, produced by freeze drying, reduce the effectiveness of energy transport in a crystal and probably account for variations related to crystalline form. Depending upon the carbohydrate irradiated, a great many substances have been identified including H,, COB,aldehydes, ketones, acids, and other carbohydrates (Dauphin and SaintLebe, 1977). In aqueous solution, irradiation of carbohydrates causes oxidative degradation. The changes are due both to direct action of the radiation and to indirect action mainly by OH. radicals produced by radiolysis of the water. For the lower saccharides, oxidation at the ends of the molecule produces acids. Ring scission forms aldehydes. D-mannose, for example, in the absence of oxygen forms D-mannonic acid, D-glucose, and two, three, and four carbon aldehyde fragments. In the presence of oxygen, secondary reactions occur leading to D-erythrose, glyoxal, oxalic acid, mannuronic acid, D-xylose, mannonic acid, D-arabinose, and formaldehyde. For aldohexoses in solution, it is clear that the effect of radiation is not confined to any particular part of the molecule and all bonds are affected. For higher saccharides, cleavage of the glycosidic link is part of the radiation effect. This results in fragmentation into smaller molecules. Corn starch, for example, yields glucose, maltose, erythrose, ribose, and mannose (Berger et af., 1973). All carbohydrate solutions produce malonaldehyde and deoxycompounds. In the normal neutral pH of foods, the yields of malonaldehyde are minimal (Phillips, 1972). The amount of deoxysugars produced in starch at 100 krad is less than 0.3 pg/gm (Diehl et al., 1978). Many substances provide protection against radiation degradation of carbohydrates (Phillips, 1972). Among these are amino acids and proteins (Diehl et al., 1978). These observations point to the influence that compounds associated with carbohydrates in a food can exert on the end results, and care must be exercised in extrapolation of findings obtained for pure substances to the complex systems that exist in foods.
I74
WALTER M . URBAIN
IV. APPLICATIONS OF FOOD IRRADIATION A.
INTRODUCTION
Applications of food irradiation can be divided into two categories according to the dose employed, high and low dose. This division is somewhat arbitrary but is useful. Applications requiring less than I Mrad are considered to be low dose uses; above 1 Mrad they are high dose (Anonymous, 1970a). High-dose applications generally are concerned with sterilization. All other uses fall into the low-dose category. A nomenclature has been devised to identify the objective of a particular radiation treatment in terms of its affect o n microorganisms (Goresline rt ul., 1964): Ruduppertizution-to produce a condition of “commercial sterility,” the same as with thermal processing (appertization). Rudicidutinn-to reduce the number of viable specific nonspore forming pathogenic microorganisms (other than viruses) so that none is detectable in the treated food by any standard method. Radurizution-to enhance the keeping quality of a food by means of radiation. This is interpreted to mean reduction o f t h e initial population of viable specific spoilage microorganisms to such a level that outgrowth to a spoiled condition is delayed. In addition, the term disinfestation is applied to radiation treatments whose objective is to kill o r inactivate insects or parasites contaminating a food. Other more specific terms are used for particular treatments. “Delayed senescence” applies to the irradiation of raw (living) fruits and vegetables whose usual ultimate spoilage is a kind of overripening, which can be delayed by radiation. “Sprout inhibition” refers to treatments to prevent or delay sprouting of foods such as potatoes or onions.
B.
HIGH-DOSE APPLICATIONS
I.
Meats and Seafood
As stated. the objective of high-dose applications is to achieve sterility, or perhaps, more precisely, to obtain indefinite shelf stability without refrigeration. Radiation alone cannot achieve this, but it does provide a key part of the requirements, namely destruction of spoilage microorganisms, including any that would affect the safety of the food. Irradiation must be combined with (a) suitable packaging to prevent microbial recontamination and also to isolate the product from the atmosphere and (b) inactivation of enzymes native to the food. whose action could cause undesirable changes, such as alteration of flavor or
I75
FOOD IRRADIATION
texture. Metal containers, such as presently used in thermal canning have been found satisfactory (Killoran et al., 1974). Glass containers are functionally suitable but are discolored by the radiation. Flexible film containers have been developed which meet performance requirements. A laminated flexible package consisting of chemically bonded Mylar and medium density polyethylene as the food contactant layer, aluminum foil (middle layer), and Nylon 6 (outside layer) was found to be satisfactory (Killoran, 1972; Wierbicki e t a / . , 1975). Containers are filled with product and closed prior to irradiation. The dose requirement for radappertization is determined by the microorganism associated with the food that has the greatest radiation resistance. For nonacid low-salt foods. not containing critical minimum levels of nitrite, such as many meats, this organism is the spore of Clastridium botulinum. The radiation resistance of this organism is different for the different strains and varies with the food. Table 111 shows the comparative resistance of representative strains of CI. botulinum Types A and B. The determination of the sterilization dose is not simple. While the knowledge gained with thermal sterilization is useful, radappertization is not exactly a parallel process. Most importantly, CI. botulinurn spores are not the most heat-resistant organisms found in foods. As a consequence, while there is greatest concern for the destruction of this organism, due to the potential hazard of botulism
TABLE I l l RESISTANCE TO GAMMA RAYS OF REPRESENTATIVE STRAINS OF Chslridiurn hotdinurn TYPES A A N D B 1N
PHOSPHATE BUFFER. pH 7"
Type
Strain number
D-value" (Mrads)
A A B B B A A A B B
33 36 40 41 53 62 77 1288s 9 51
0.334 0.336 0.317 0.318 0.329 0.224 0.253 0.241 0.227 0.129
From Anellis and Koch (1962). D-value is the dose necessary to accomplish a 90% destruction of the organisms present. (I
176
WALTER M . URBAIN
should outgrowth and toxin formation occur, thermal processes usually are set for organisms other than Cl. botulinum which are more heat resistant and whose outgrowth could cause spoilage. This circumstance provides an insurance factor for the safety of thermally sterilized foods. In the case of irradiation, the most radiation-resistant organism of concern is Cl. botulinum. There are other organisms whose radiation resistance is greater but they are not a factor in the production of radappertized foods (Welch and Maxcy, 1975; Maxcy et al., 1976; Anellis et al., 1977). Although the asporogenous Acinetobacter and Moraxella bacteria have a high radiation resistance, they are easily killed by heat. The preirradiation heat treatment (67"-75"C) for autolytic enzyme inactivation of radappertized meats is sufficient to be lethal to them. The safety of the thermal sterilization process is based on a 12-D reduction in the count of the most heat-resistant strain of Cl. botulinum. Early work (Hannan, 1955) had suggested that about 2 Mrad was a sufficient dose for radappertization, but Hannan (1955, p. 67) considering the 12-D concept applied to thermal processing and using the data of Morgan and Reed (1954) for the radiation resistance of CI. botulinum spores, suggested that a dose approaching 5 X lo6 rep* or 4.65 Mrad. This concept was taken up by Schmidt (1961) and, after much debate, was accepted in principle. Thus, both thermal and radiation processing are placed on the same basis as far as safety with respect to botulism is concerned. Since no other food spoilage microorganism has a greater radiation resistance, irradiation lacks the added safety factor that organisms more heat resistant than Cl. botulinum provide the thermal sterilization process. Partly because of this, the minimum radiation sterilization dose must be known accurately and the process must be carefully designed to assure its delivery. The conventional practice of estimating the 12-D dose (Schmidt and Nank, 1960) is based on the assumption that the rate of spore kill in an inoculated pack is exponential (no initial shoulder). An improved method which replaces the conventional one has been developed (Anellis et al., 1975, 1977; Ross, 1974). This method employs two interrelated functions which operate simultaneously in foods undergoing irradiation: (1) A spore inactivation rate that is not necessarily a simple exponential one; and (2) a can sterilization rate which is dependent upon inactivation of the most resistant spore in the can. The conventional method uses only the first function and assumes it to be a simple exponential function. The second function arises as the extreme or largest value derived from the first. The new method yields the largest dose complying with inoculated pack data and which also is a 12-D dose. It is, therefore, a conservative dose in the sense that it may be greater than is actually needed (Ross, in press). The inoculated pack data are obtained by inoculating the specific product in *One rep (old unit) equals 93 ergs per gram, or 0.93 rad.
177
FOOD IRRADIATION
TABLE IV INOCULATED PACK EXPERIMENTAL DESIGN FOR BEEFn
Prototype food CI. hotdinurn strains Spore inoculum Containers Foodicontainer Cansidose Vacuum seal Radiation source Radiation doses (Mrad) Radiation temperature ("C) Incubation Analysis
~~~~~~
Beef formulated with 0.75% NaCI, 0.38% TPP A mixture of 33A, 36A, 62A, 77A 12885A. 9B,40B, 419, 53B. 679 !@/strain; 107/can 21 1 x 101.5 (epoxy enamel) metal cans 40 t 5 gm 100 replicate 16 kPa T o gamma rays 1.4, 1.8, 2.2, 2.6, 3.0. 3.4, 3.8,4.2. 4.6, 5.0
-30 t 10 6 months at 30 t 2°C Swelling: daily-1st month weekly-2nd thru 6th month Botulinal toxin: 7th month Recoverable CI. hotulinum: 7th month ~
~~
~
~
From Anellis el ol. (1976). First Int. Congr. Eng. Food, as sponsored by the American Society of Agricultural Engineers. "
question with a spore level of about lo7 per unit (can). Determination of the minimum radiation dose (MRD) is based upon (a) the presence or absence of viable botulinal cells in the cans, regardless of their ability to outgrow and produce toxin and/or can swelling; (b) a single most resistant strain of Cl. botulinum, and (c) a shifted exponential (an initial shoulder followed by a semilog decline) rate of spore death. The irradiation conditions employed are identical (particularly with respect to irradiation temperature) as those of the commercial process. The experimental design of the inoculated pack of a beef prototype food is shown in Table IV. A sophisticated statistical treatment (Ross, 1976) provides a margin of safety and yields the maximum (most conservative) dose. Determination of the MRD on the basis of surviving organisms rather than on swelling or toxin formation provides an additional margin of safety. The MRDs for a number of radappertized meats and codfish cakes are given in Table V. The variation in values reflects the differences in composition and in irradiation temperatures and the impact of these differences on the resistance of CI. botulinum. Irradiation at lower temperatures reduces the lethal effect of radiation and increases the dose requirement for radappertization (Rowley et ul., 1968; Grecz ef a / . , 1971; Maxcy et af., 1976). The presence of NaCl and NaNO, lowers the dose requirement. The dose requirement for radappertization of low-acid low-salt foods has now been established. The problems related to the high MRD values have been the subject of much research. Virtually all the recent research and development in the high dose category has
178
WALTER M . URBAIN TABLE V MINIMUM RADIATION DOSE (MRD) FOR RADAPPERTIZED MEATS AND CODFISH" ____
~~
~
MRD
Food
Irradiation temperature ("C)
(Mrad)
Bacon Beef" Ham" Ham" Pork Codfish cakes Corned beef Pork sausage
5 to 25 -30 10 5 to 25 -30 2 10 5 to 25 -30 5 10 -30 2 10 - 3 0 2 10
2.5 4.1' 3.1 3.3 4.3 3.2 2.4 2.7
"
*
From Wierbicki et a / . (1975).
* With additives: 0.75%NaCI, 0.375% Na tripolyphosphate. Anellis et a / . (1977). " High NaN02/NaN03( I 56/700 mg/kg)--regular. "Reduced NaN02/NaN03(25/100 mglkg).
been concerned with meats and seafood and has been done only at one place, the U. S . Army Natick Research and Development Command. The Army program began in 1953 and, as noted, has the objective to provide foods of greater consumer acceptability, improved nutritive quality, and better storage characteristics to be used as military rations. The military program is expected to have a spin-off of benefits to the civilian sector. Among the products that have been developed are radappertized bacon, ham, pork, chicken, beef, hamburger, corned beef, pork sausage, codfish cakes, and shrimp. In the raw state, all of these foods contain indigenous enzymes, which must be inactivated for long-term preservation. At the doses employed in radappertization, radiation does not effect sufficient enzyme inactivation, as shown by the data of Table VI. In order to obtain the degree of enzyme inactivation needed for product stability, the use of heat has been found to be the only practical and effective method. Heating of meats to 7Oo-75"C prior to irradiation is sufficient (Shults and Wierbicki, 1974a). The high-dose requirement for radappertized food results in some undesired side effects, namely the formation of unpleasant and characteristic odor and flavor, texture changes, and, in meats and seafood containing myoglobin pigments, color changes. Of these, from the standpoint of consumer acceptance, the flavor change is the most important. A considerable effort over many years has been expended to find the cause of the off-flavor. Little real progress was made until gas chromatography and mass
179
FOOD IRRADIATION TABLE VI
EFFECTS OF IRRADIATION DOSE AND TEMPERATURE ON THE PROTLOLYTIC ENZYME ACTIVITY OF BEEF MUSCLE"
Irradiation temperature ("C)
Dose (Mrad)
+21 7c reduction
0 0 reduction
2 4 6 8
57
44
33
0
65 19 86
65 12 82
45 40
40
I'
- 30
- 80
%' reduction
% reduction
73
18 60
From Shults et cd. (1975)
spectrometry provided the analytical techniques needed to identify the radiolytic products present. No positive identification of the substances responsible for the flavor, however, has been secured. Wick rt af. (1967) concluded that methonal, 1-nonanal, and phenylacetaldehyde are the principal substances responsible for the flavor. The sensitivity of protein foods to off-flavor development by radiation varies with the species of animal from which the food is derived, as shown by the data of Table VII (Sudarmadji and Urbain, 1972). It was observed that irradiation in the frozen state significantly lessened the off-flavor in meat (Brasch and Huber, 1948; Coleby et al., 1961). Lean meat is of the order of two-thirds water. Irradiation of water can produce a variety of substances including free radicals such as OH., the aqueous electron e&, hydrogen atoms, and active compounds such as H,O, (Draganic and Draganic, 1971; Hart, 1972; Swallow, 1977). These radiolytic products from the water present in meat can cause an indirect action of radiation. Freezing prior to irradiation lessens the effect o f this indirect action and results in less off-flavor development. Table VIII shows data on the change of flavor and textural characteristics of beef with irradiation temperature. Lowering the temperature clearly leads to improvement of sensory properties. Harlan rt af. (1967) found a similar relationship. The beneficial effect of low-temperature irradiation, as measured by subjective criteria, has been confirmed by measurement of the amount of radiolytic products formed. Figure 7 shows the change with irradiation temperature of flavor intensity score and amounts of detected volatile radiolytic products. Correlation of flavor intensity with amount of volatiles formed seems clear. The simultaneous lowering of both with lowered irradiation temperatures points to their origin with the indirect action of radiation through the radiolytic substances produced in the water. As noted earlier. the lethal effect of radiation o n microorganisms also is reduced by lowering the irradiation temperature. In this manner the dose required
180
WALTER M. URBAIN TABLE VII THRESHOLD DOSE FOR DETECTABLE OFF-FLAVOR FOR PROTEIN FOODS FROM VARIOUS ANIMALS IRRADIATED AT 5" TO 10"C'
Threshold dose (krad)
Animal food
I50
Turkey Pork Beef Chicken Lobster Shrimp Rabbit Frog Whale Trout Turtle Halibut Opossum Hippopotamus Beaver Lamb Venison Elephant Horse Bear
175 250 250 250 250 350 400 450 450 450 500 500 525 550 625 625 650 650 875
* From Sudarmadji and Urbain (1972). Reprinted from Food TechnologylJournal of Food Science 37, 671-672, 1972. Copyright @ by Institute of Food Technologists. TABLE VlIl EFFECT OF IRRADIATION TEMPERATURES ON FLAVOR AND TEXTURAL CHARACTERISTICS OF UNITED STATES COMMERCIAL BEEF LOIN",b
Irradiation temperature ("C)
Irradiation flavorr
Mushiness"
Friability'
4.1 3.3 2.9 2. I 1.5
5.3 3.4 2.5 2.0 2.0
5.0 3.0 1.9 1.8 1.9
+60 t21 +40 - 80 - I85 "
From Shults and Wierbicki (1974b).
* Dose 4.5 to 5.6 Mrad. I'
Intensity scale of 1-9 ( I denoting "none" and 9 "extreme").
FOOD IRRADIATION
181
*Oo0[ 1500
TEMPERATURE (“C)
FIG. 7. Change with irradiation temperature of flavor intensity score and amounts of detected volatile radiolytic products of beef irradiated at 5.6 Mrad. Arrow denotes value for nonirradiated control. Flavor intensity scale of 1-9 (1 denoting “none” and 9 “extreme”) (Memtt et al., 1975). Reprinted with permission from Journal qfAgriculrural and Food Chemisrry 23, 1037-1041. Nov.1 Dec. 1975. Copyright by the American Chemical Society.
for sterility is increased. The amount of radiolytic products is a function of dose, as may be seen from the graph of Fig. 8. As a consequence, the value of low-temperature irradiation may be questioned. Is the gain in sensory quality improvement offset by higher dose requirements to gain sterility? It appears not. For its products, the Natick laboratory has selected an irradiation temperature of -30” 2 10°C. This temperature was selected as producing an adequate acceptance improvement for beef which was not significantly bettered by using lower temperatures (Shults and Wierbicki, I974b). While lower irradiation temperatures have been considered, it was concluded that -30” 2 10°C provides the
IRRAOLATION DOSE (rnrads)
FIG. 8. Graph showing relative amounts of detected volatile radiolytic products produced as a function of dose in beef irradiated at - 185°C (Merritt F I a / . . 1975). Reprinted with permission from Jourriul of Agricultural and Foad Chemisrn 23. 1037-1041, Nov./Dec. 1975. Copyright by the American Chemical Society.
182
WALTER M . URBAIN
most favorable balance of quality, cost, and required irradiation dose (Wierbicki er ul., 1975). While the use of subfreezing irradiation temperatures unquestionably yields improvement in the sensory properties of radappertized foods, trained expert evaluators can note some “irradiation flavor” in the products listed on page 178. Such persons have expressed reservations about the acceptance of these foods by ordinary consumers. The Army has compared the acceptance by volunteer troops of radappertized meats, poultry, and seafoods with nonirradiated controls. In these tests literally thousands of testers were employed. Table IX shows such consumer acceptance data obtained with Army and Air Force personnel. Army experience indicates that products scoring 5 or higher on the 9-point hedonic rating scale are acceptable as rations. Because the ratings for radappertized foods developed by the Army have exceeded the value of 5 , the views of the product experts relating to a detectable irradiation flavor have been set aside. It appears that the final test for acceptability can be made only when the products become available to consumers, both military and civilian, on a basis that allows the open competition of the market place. In this connection it should be noted that radappertized meats have obviously superior texture characteristics and do not undergo moisture release, as is the case with thermal sterilization. It is also significant that many common foods undergo flavor changes as a result of processing and yet obtain a high degree of consumer acceptance. The technological development of the products considered by Natick and listed earlier largely has been recorded in a number of publications. Bibliographies on this work are available (Wierbicki, 1974; Cohen and Mason, 1976).
TABLE IX ACCEPTANCE OF RADAPPERTIZED MEATS, POULTRY. AND SEAFOODS“
Irradiated
Nonirradiated control
Item
Number of evaluators
Rating”
Number of evaluators
Rating”
Ham Chicken Pork Beef Bacon Shrimp Codtish cakes
1,651 583 39 1 589 25,656 539 53 1
5.87 6.07 5.71 5.99 6. I6 6.09 5.40
1.437 548 458
6.66 6.36 6x5 6.61
“
644 849 578
-
6.43 6.30
From Urbain (1970).
’ Based on a 9-point hedonic scale, 9 like extremely. 5 neither like nor dislike.
1 dislikeextremely.
FOOD IRRADIATION
I83
2 . Fruits and Vegetubles The success with the development of radappertized meats and seafood must be correlated with the sustained intensive research effort of the United States Army. There has been no comparable program with other radappertized foods. In the early years of research on food irradiation many foods were examined. Apparently sensory changes discouraged further work with fruits and vegetables. Irradiation in the frozen state had a protective effect on flavor and color, but did not prevent texture damage (Hannan, 1955). Dipping fruit in a calcium chloride solution reduces the softening caused by irradiation (Al-Jasim et al., 1968). The preparation of shelf-stable fruit juices was attempted in Europe, but without definitive results (Kaindl, 1966; Anonymous, 1967; Kiss and Farkas, 1968). Just as with animal products, enzyme inactivation is required for long storage of fruits and vegetables and is best done by heating (blanching).
3 . Spices There was early interest in the radiation sterilization of spices (Proctor et al., 1950) and it has continued into the present. In this application, the objective is not preservation but rather the reduction or elimination of a bacterial population normally indigenous to these materials as used and whose presence constitutes a problem when incorporated in foods (Hansen. 1966; Farkas, 1973; Inal el al., 1975; Farkas and Beczner, 1973). Because preservation is not the objective. the minimum dose requirements, such as those for radappertized meats, are not necessary and adequate effects can be obtained with 1 to 2 Mrad. Vajdi and Pereira (1973) showed that gamma radiation was more effective than ethylene oxide in reducing the bacterial population of spices. While ethylene oxide reduced the oil content of certain spices and affected the color of paprika, gamma irradiation caused insignificant changes. Tables X and XI provide data on bacterial reduction and oil content of selected spices. Bachman and Gieszczynska (1973) obtained similar results. 4 . Diets for Speciul Patients Similar to the treatment of spices, irradiation with high doses is used to remove microbial populations of other products without having preservation as a purpose. In a few countries (United States, United Kingdom. the Netherlands, and West Germany), government approvals have been granted to permit irradiation of diets of hospital patients whose circumstances require extraordinary protection from infection. Generally these patients, as part of their therapy, have been treated to reduce their immuno response. Patients receiving organ transplants and
TABLE X THE COMPARATTVEEFFECT OF ETHYLENE OXIDE AND GAMMA IRRADIATTON ON THE BACTERlAL FLORA OF SELECTED RAW SPICES"
Treatments (Number of organisms per gram) Raw
Spices
Total count
B. pepper Paprika Oregano Allspice Celery seeds Garlic
4 . 0 X lo6 9.86 X lo6 3.26 x 104 1.74 x I@ 3.7 X lo" 4.65 x 104
Ethylene oxide
Thermophilic 1.58 X 3.24 X 1.8 x 1.5 x 1.3 X 9.0 x
lo6
lo5 I@ loR lo" 18
Aerobic spores
6.34 X 3.0 X 1.0 x 1.05 x 3.94 X 0.0
104 10' 18 I@ 10'
Total count
1.48 X 16 0.0 0.0 4.25 x 10 0.8 x 10 1.45 X I 0 4
Thermophilic
4.3
X I 8 0.0 0.0
0.0 0.0
3.5
X
l@
Gamma irradiation Aerobic spores
Total count
Thermophilic
Aerobic spores
0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
0.0 0.0 0.0 0.0 0.0 0.0
" From Vajdi and Pereira (1973). Reprinted from Food TechnologylJournul of Food Science 38, 8 9 3 4 9 5 . 1973. Copyright @ by Institute of Food Technologists.
185
FOOD IRRADIATION
TABLE XI THE COMPARATIVE EFFECT OF ETHYLENE OXIDE AND GAMMA IRRADIATION ON THE VOLATILE A N D NONVOLATILt OIL CONTENT OF SIX RAW SPICES"
Treatments (Volatile and nonvolatile oils. a) Raw
Spices
B. pepper Paprika Oregano Allspice Celery seeds Garlic
Ethylene oxide
Volatile oil
Nonvolatile oil
Volatile
3.6 3.3 6. I6 1.73
10.436 14.345 10.401 10.910 23.420 1.340
I .6 3.2 I.6 I .73 -
-
oil
Gamma irradiation
Nonvolatile oil
Volatile 011
9.210
3.6
II ,380
-
8.060 6.130
3.33 6. I 6 I .73 -
21.360
0.607
Nonvolatile oil
9.363 15.010 10.248 9.650 23.100 1.040
" From Vajdi and Pereira (1973). Reprinted from Food TechnologylJournd of Food Science 38, 893-895, 1973. Copyright @ by Institute of Food Te indicate rng/100 grn dry wt; the value$ in parentheses indicate percentage of the total carotenoids. Data of Venkatakrishna ct ul. (1977).
level was also determined by the genetic structure of the plant, rainfall. altitude of growth, and soil conditions. The fluoride content of tea is also unusually high, being about 150 ppm for black tea and more than twice this level for green tea (Singer et al., 1967). The
248
ROBERT L. WICKREMASINGHE
major mineral component of tea is, however, potassium, which accounts for about 50% of the total mineral content (Sanderson et al., 1976), and deficiency of this element leads to defoliation and death of the tea bush (Portsmouth, 1953). Deficiency of zinc has been found (Tolhurst, 1962) to lead to decline of yield, and the spraying of tea with zinc is now a standard practice in tea-growing countries. Symptoms produced in tea leaves of deficiencies of nitrogen, potassium, calcium, magnesium, and sulfur have been described (Pethiyagoda and Krishnapillai, 1970) and this study was extended to include the effect on leaves of deficiencies of the minor elements, manganese and boron (Pethiyagoda and Krishnapillai, 197 1). 10. Volatile Compounds The mixture of volatile compounds present in tea constitutes only about 0.01% of its dry weight but plays an important role in determining the overall aroma of the brew. More than 300 compounds have been identified in this mixture, and the characteristic aroma of a tea depends on the correct balance between the proportions of certain key compounds (Wickremasinghe et al., 1973; Kozhin and Treiger, 1973), the nature of which is described in Section IV. Subsequent to the compilation of a list of the compounds identified as constituents of tea aroma (Yamanishi, 1975; Sanderson, 1 9 7 3 , several other compounds (Table VIII) have recently been found to be also present as constituents of the volatile fraction of black tea (Renold et al., 1974; Vitzthum et al., 1975). However, in spite of detailed analysis of the tea aroma complex, there are still a number of constituents which remain unidentified. Furthermore, tea aroma is not one single entity, but a complex characteristic which is dependent on location of growth of the tea bush (Yamanishi et al., 1968a), climate (Wickremasinghe, 1974), genetic constitution of the tea clone, rate of fertilizer application, conditions of processing, and a number of other factors. Green tea flavor is distinct from black tea flavor, and the flavor of a Darjeeling black tea is distinct from that of a black tea from the Dimbula District of Sri Lanka. Nevertheless all of these teas have a qualitatively similar profile of volatile constituents (Yamanishi et al., 1968b), and it would appear that it is the balance between volatile constituents that determines the flavor of tea. B . FACTORS AFFECTING CHEMICAL COMPOSITION 1.
Climate
Ramaswamy (1964) reported that the composition of black tea liquors was related to the weather conditions (wet or dry) prevailing at the time the tea was produced, as well as to the elevation of growth of the tea bush. In an analysis of a limited number of liquors, the level of soluble solids and nitrogenous substances was found to be higher during the dry season than the wet season; mineral
TEA
249
constituents showed the opposite trend. With respect to the effect of elevation, the level of theaflavins and oxidizable matter was higher in liquors obtained from high- than low-grown tea, whereas nitrogenous constituents and mineral content were lower. These changes in liquor composition with climate are related to the composition of the fresh tea flush, and Evans (1930) reported that there were marked seasonal variations of total soluble solids, tannin, total soluble nitrogen, and total nitrogen in fresh tea leaf; these findings were confirmed and extended by subsequent workers (Sanderson, I964b; Sanderson and Kanapathipillai, 1964; Wickremasinghe ct al., 1966) who studied the variations with season of minerals, pectin, polyphenols, oxidase activity, amino acids, and chlorophyll. Gianturco et a/. (1 974) studied the seasonal variations of the volatile constituents of black teas from three different locations over a period of fifteen months, and found a correlation between their composition and the quality of tea aroma. In a study of the effect of shade on the chemical composition of tea flush, Hilton ( I 974) found that 25% shading of clonal tea bushes increased the proportion of gallated to nongallated flavanols and also increased polypbenol oxidase activity. Anan and Nagakawa (1 974) observed that shading of developing tea shoots with several sheets of black net for 13 to 25 days had no effect on the content of (-)epicatechin and (-)epigallocatechin, which however, showed a gradual increase in the unshaded shoots. They also observed that total amino acids and caffeine increased on shading. Extensive studies have been carried out in Tanzania on the effect of irrigation on the yield of tea (Carr, 1970, 1974) and it was found that irrigation during the long dry season doubled the annual yield of tea and, to some extent, evened the yield over the year. In Malawi experiments, Ellis (1976) obtained a yield increase of 30% over unirrigated tea, but considered that this return was uneconomic in relation to the high cost of irrigation. Devanathan (1975) found a correlation between yield of tea and the product of rainfall and average daily hours of sunshine of the previous month. This correlation was subsequently refined by inclusion of factors for variations in soil and the temperature coefficient of photosynthesis (Devanathan, 1976).
2 . Clone The differing genetic constitution of the various clones of tea would predictably influence the chemical composition of the tea flush produced. Studies have been made of the clonal variations in respect of polyphenol oxidase activity, individual polyphenols, individual amino acids, and chlorophyll of green leaf (Tocklai Exp. Sm., 1974). Analyses of flush from 11 Sri Lanka clones for ash, total flavanols, caffeine, total nitrogen, and polyphenol oxidase activity (Sanderson, 1964b) indicated that there was some intercloncal variation, and that there appeared to be a relationship between the quality potential of a clone, polyphenol oxidase activity, and ash content-this relationship was not, however, perfect.
250
ROBERT L. WICKREMASINGHE T A B L E VIlI NEWLY IDENTIFIED VOLATILE CONSTITUENTS OF BLACK TEA“
1. ALDEHYDES cis-3-pentenal trans-2-heptenal trans-2-nonenal trans-2-decenal rrans-2-undecenal trans-2, cis-4-hexadienal trans-2, trans-4-octadienal trans-2, cis-4-octadienal trans-2, rrans-4-nonadienal trans-2. cis-4-nonadienal trans-2, cis-6-nonadienal rrans-2, cis-4-decadienal neral P-cyclocitral safranal 2-methylbenzaldehyde 4-methoxybenzaldehyde
4-methyl-2-phenyl-2-pentenal 5-methyl-2-phenyl-2-hexanal 4-ethyl-7, 1 1 -dimethyl-trans-2, trans-6, 10-dodecatrienal 4-ethyl-7, I I -dimethylrrans-2. cis-6, 10-dodecatrienal 2. KETONES 2-heptanone 5-isopropyl-2-heptanone 2-octanone 3-octanone trans-3, cis-5-octadien-2-one 2-nonanone 6, 10-dimethyl-2-undecanone benzyl ethyl ketone 2. 6, 6-trimethylcyclohex-2-enI -one 2, 6, 6-trimethylcyclohex-2-en1, 4-dione P-damascenone a-damascone P-damascone I , 5, 5, 9-tetramethylbicyclo [r. 3. 01 non-8-en-7-one 3. ESTERS hexyl formate rrans-2-hexenyl formate cis-3-hexenyl formate
trans-2-hexenyl acetate ethyl phenylacetate hexyl phenylacetate trans-2-hexenyl propionate trans-3-hexenyl propionate cis-3-hexenyl propionate hexyl butyrate trans-2-hexenyl butyrate trans-3-hexenyl butyrate benzyl butyrate cis-3-hexenyl 2-methylbutyrate rrans-2-hexenyl hexanoate cis-3-hexenyl trans-2-hexenoate trans-3-hexenyl cis-3-hexenoate methyl octanoate ethyl octanoate methyl trans-dihydrojasmonate 4. MISCELLANEOUS 2-ethyl- 1-hexanol 4-terpineol 4-methyl-5-hexen-4-olide carvacrol thymol 2-acetylfuran safrole 2, 6, 10, 10-tetramethyl-I-oxa-spiro [4, 51 dec-6-ene (“theasprane”) 6, 7-epoxy-2, 6, 10, 10-tetramethylI-oxa-spiro [4, 51 decane (“6, 7epoxy -dihydrotheaspirane ”) 6-hydroxyl-2, 6, 10, I O-tetramethyl1-oxa-spiro [4, 51 decane (‘ ‘6-hydroxy-dihydrotheaspirane”)
phenylacetic acid trans-geranic acid trans-2-octenoic acid 5 . PYRIDINES Pyridine 2-Methylpyridine 3-Methylpyridine 4-Methylpyridine 2-Ethy lpyridine 3-Ethy lpyridine 2.6-Dimethylpyridine (continued)
25 1
TEA
TABLE VIII-(continued) 2-Methy lbenzothiazole 8. QUINOLINES 2-Methy lquinoline 6-(or 7-)Methylquinoline 2,6-Dimethylquinoline 2,4-Dimethylquinoline 4.8-Dimethylquinoline 3-n-Propylquinoline 4-n-Butylquinoline 9. AROMATIC AMINES Aniline N-Methylaniline N-Ethylaniline o-Toluidine N.N-Dimethy lhenzylamine 10. AMIDES N-Ethy lacetamide N-Ethylpropionamide 1 1, MISCELLANEOUS Benzoxazole I .4-Diacetylhenzene I .3-Diacetylhenzene 2,4-Dimethylacetophenone p-Ethy lacetophenone 2.4-Dimethy lpropiophenone p-Ethylpropiophenone 3,4-Dimethoxyacetophenone
2SDimethylpyridine 2-Methyl-6-ethy lpyridine 2-Methyl-Sethylpyridine 3-Methoxypyridine 4-Vinylpyridine 2-Acetylpyridine 2-n-Butylpyridine 2-Phenylpyridine 3-Phenylpyridine 6. PYRAZINES Methylpyrazine 2.6-Dimethylpyrazine 2.5-Dimethylpyrazine 2.3-Dimethylpyrazine Ethylpyrazine 2-Ethyl-6-methylpyrazine 2-Ethyl-5-methylpyrazine Trimethylpyrazine Tetramethy lpyrazine 2-Ethyl-3.6-dimethylpyrazine 7. THIAZOLES 2,4-Dimethylthiazole 2,5-Dimethylthiazole 5-Methylthiazole 2,4,5-Trimethylthiazole 2,5-Dimethyl-4-ethylthiazole Benzothiazole ~
Groups 1-4 identified by Renold
cf
ti/.
(1974) and 5-1 I by Vitzthum
el a1
(1975).
Ill. CHANGES DURING THE PROCESSING OF TEA A.
BLACKTEA
The starting material of black tea processing is the young tea shoot, consisting ideally of the terminal bud and two adjacent leaves, which are generally handpicked by teams of experienced pluckers. In some countries the high cost of manual labor has led to the development of devices for mechanical plucking, but this practice suffers from the drawback that more mature leaves are included in the harvest, with a resulting loss in quality of the processed black tea. One of the reasons for this loss of quality is the progressive decline with leaf maturity of polyphenol oxidase activity, polyphenols, caffeine, and amino acids (Wickremasinghe and Perera, 1973).
252
ROBERT L. WICKREMASINGHE
The processing of tea flush to black tea comprises the following stages: withering, preconditioning, rolling, fermentation, firing, and grading.
I. Withering Withering is generally accomplished by thinly spreading the flush on tats, or the more recent practice of loading it into troughs. The loss of moisture may be hastened by blowing hot air through the “withering loft,” and withering is allowed to proceed for a period of 8 to 18 hours, during which time the moisture content of the leaf drops to between 60% (“soft” wither) and 50% (“hard” wither), and the leaf acquires a “kid glove” feel. It was believed for many years that physical conditioning of the leaf was the only purpose of withering, but it is now known that several chemical changes (Table IX) occur during this stage of processing. All of these changes with the exception of the increase in cell wall permeability are independent of moisture loss during withering (Sanderson, 1968) and play an important part in improving the character of the finished product.
2 . Preconditioning Preconditioning of withered tea leaf consists of rolling the leaf for 10 to 15 minutes without the application of any pressure-the action is similar to lightly rubbing the leaf between the palms of one’s hands, which was the traditionally used method, still employed in a few countries. The purpose of preconditioning is to impart the desired “twist” and compactness to the leaf (Keegel, 1958) and also to make available the maximal amount of polyphenol oxidase for fermentation (Wickremasinghe, 1978). This enzyme was found to be located in discrete cells of the epidermis of the tea leaf (Wickremasinghe et af., 1967) and light rolling results in disruption of the separating walls and the production of a homogenous layer of enzyme. Rupture of this layer at even one point during the subsequent state of rolling under pressure would, therefore insure full utilization of the leaf complement of polyphenol oxidase, which plays a central role in the following stages of processing.
3 . Rolling The purpose of rolling is to macerate the leaf in order that the contents of the leaf (such as enzymes and their substrates) may be intimately mixed. There are several machines which have been developed for this purpose, but the so-called conventional or orthodox roller (Keegel, 1958), the rotorvane, and the C.T.C. (Crushing, Tearing, Curling) machines, or the use of these machines in various combinations, are the most popular in present-day black tea processing. During rolling some of the changes initiated during withering and preconditioning proceed at an accelerated rate, consequent to the rise in temperature caused by frictional forces. Additionally, new reactions that are of importance
253
TEA TABLE IX CHANGES DURING THE WITHERING STAGE OF BLACK TEA PROCESSING
Importance to black tea characteristics
Nature of change Formation of amino acids
Precursors of compounds determining flavor and extent of nonenzymic browning reactions
Formation of keto acids Formation of mevaionic acid
Precursor of compounds determining flavor Precursor of compounds determining flavor
Formation of caffeine
Pharmacological activity and taste of tea Improves fermentation
Increase of polyphenol oxidase activity Breakdown of chlorophyll
Affects appearance
Increased cell wall permeability
Enhances efficient mixing of reactants during fermentation Unknown
Increase of organic acids Breakdown of polysaccharides
Unknown
Reference Roberts and Wood (1951); Bhatia and Deb (1965); Roberts and Sanderson (1966); Wickremasinghe ( I 978) Wickremasinghe (1964) Wickremasinghe and Sivapalan ( 1966) Wood and Chanda (1955); Stagg and Millin (1975) Takeo ( 1966a) Wickremasinghe and Perera ( 1966a) Sanderson ( 1968)
Sanderson and Selvendran (1965) Sanderson and Perera ( 1965)
for black tea character are themselves initiated because of more intimate mixing of the leaf constituents. These reactions continue during rolling and are allowed to proceed to the desired stage in the fermentation stage of tea processing. The nature of these reactions is described in the following section. 4.
Fermentation
The designation of this stage of black tea processing as fermentation is a misnomer, stemming from the erroneous view, held at one time, that the changes occurring at this stage were mediated by microorganisms. The procedure for fermentation is to pile the leaf, which has already been macerated by rolling, in a layer 5 to 7.5 cm thick, and let it stand at room temperature for periods of time varying from 45 minutes to 3 hours, depending on the qualities being sought for in the processed black tea. Some of the reactions occurring during fermentation are enzyme-catalyzed, and of these the more important are the oxidation of tea
254
ROBERT L. WICKREMASINGHE
flavanols by polyphenol oxidase, which leads to the development of color, strength, and quality in tea brews, and the occurrence of reactions responsible for the characteristic aroma of black tea. a . Tea Flavanol Oxidation and Development of Color, Strength, and Quality during Fermentation. Following the pioneering studies of Roberts and his coworkers (Roberts, 1962), it became clear that the formation of substances known as theaflavins and thearubigins was one of the central reactions which occurred during the fermentation stage of black tea processing. Roberts (1958) using a model tea fermentation system demonstrated that theaflavins, gallic acid, and a number of unidentified substances were formed on incubating an acetone-dried powder of fresh tea flush (Roberts and Wood, 1951) with an extract of dried flush containing a mixture of (-) epigallocatechin, (+) gallocatechin, (-) epicatechin, (+) catechin, (-) epigallocatechin gallate, and (-) epicatechin gallate together with small quantities of flavonol glycosides, leucoanthocyanins, chlorogenic acids, andp-coumarylquinic acids, but no caffeine, sugars, or amino acids. In this model system there was no detectable formation of thearubigins (as occurred during fermentation of tea flush), and this lack of thearubigin formation was ascribed to the comparatively low concentration of reagents in the system studied. Sanderson et al. (1972) also studied the oxidation, in model systems, of the major flavanols found in tea leaves by partially purified soluble polyphenol oxidase preparations (Co and Sanderson, 1970). Their results confirmed the earlier finding of Roberts and Wood (1950) that epigallocatechin and its gallate were oxidized more readily than the epicatechins, and also showed that theaflavins and thearubigins are formed only in those fermentation systems which contained the appropriate combinations of flavanols. Based on the results of this investigation, the authors proposed that theaflavins are formed according to Eqs. (1) to (4), which are consistent with the reaction proposed by Takino et al. (1964) for the formation of theaflavins. ( I ) epicatechin + epigallocatechin + i)O, + theaflavin (XVI) + CO, ( 2 ) epicatechin + epigallocatechin gallate + 0, + theaflavin gallate A (XVII) + COP (3) epicatechin gallate + epigallocatechin + +02+ theaflavin gallate B (XVIII) + COP (4) epicatechin gallate + epigallocatechin gallate + 0, theaflavin digallate (XIX) +
+
co,
+
--f
In earlier studies Bryce et a f . (1970) and Coxon et a f . (1970a) had isolated and fully characterized the three theaflavin gallates proposed in these reactions. Of the compounds formed during tea fermentation, it may be considered that theaflavins play a premier role in determining the characteristic cup quality of black tea brews (Roberts, 1962; Hilton and Ellis, 1972) and considerable attention has therefore been paid to determining the structure of theaflavins. Roberts et al. (1957) were the first to isolate theaflavin from black tea liquors, and suggested that it was derived from one molecule of (-)epigallocatechin plus one molecule of ( - ) epigallocatechin gallate. However, Dzemukhadze et al. (1957, 1964) found that all catechins decreased during processing of tea flush to black
255
TEA
OH XVI
Theofloum.
R , = R, = H
XVll
Theoflovln gallale A , R , = 3 . 4 . 5 i r i h y d r a x y b e n z a y l , R,: H
Xvlll
Theaflovin pollate B ,
XIX
R, =H
0
,
OH
R , = 3 . 4 , 5 - lrihydrorybenzoyl Theaflavin digallole. R , = R 2 = 3.4.5trihydroiybenzoyl
XX
OH
Irotheoflovin
tea, and Vuataz and Brandenberger (1961) and Bhatia and Ullah (1961, 1965) confirmed that epicatechin gallate decreased during tea fermentation. Takino et al. (1964) next demonstrated that a mixture of (-) epicatechin and (-) epigallocatechin was oxidized with either a polyphenol oxidase preparation, or potassium ferricyanide in a bicarbonate solution, to give a compound which had the same properties as theaflavin. Shortly afterward, Takino and Imagawa (1964) established that the compound formed was identical to that isolated from black tea by Roberts ( 1 958). Takino et al. ( I 964) proposed structure XVI for theaflavin, the configuration of which was determined by Takino et al. (1965) and by Brown ez al. (1966). More intensive studies on separation and characterisation of theaflavins have led to the identification of isotheaflavin ( X X ) (Coxon er al., 1970b) as well as of epitheaflavic acid (XXI) (Coxon et al., 1970c; Bryce et al., 1972) and epitheaflavic acid-3-gallate (XXII) (Bryce et al., 1972). It is evident, therefore, that the theaflavin fraction of black tea consists of a number of benzotropolone derivatives, the approximate relative proportions of which were determined (Coxon et al., 1970a) as being 8% theaflavin gallate A, 20% theaflavin gallate B, 40% theaflavin digallate, and 4% isotheaflavin, together with epitheaflavic acids. The commonly used photometric method for the quantitative estimation of total theaflavins in tea liquors (Roberts and Smith, 1961, 1963) gives values in the range 0.3 to 1.8% of the dry weight of black tea, amounting to about 1 .O to 5.9% of tea solids in a cup of brewed tea. More recently individual theaflavins have been separated from theaflavin gallate by gel filtration using Sephadex L H 20 (Lea and Crispin, 1971), and by gas-liquid chromatography of their trimethyl silyl ethers (Collier and Mallows, 1971b). Apart from the theaflavins which impart the properties of quality and brightness of color to tea liquors, (Roberts, 1962), the group of thearubigins (Roberts, 1958) also makes an important contribution to the color, strength (Roberts and
256
ROBERT L. WICKREMASINGHE
OH XXI
Epitheaflavic a c i d , R = H
XXll
Epitheaflavic acid - 3 ' - g a l l a t e ,
R = 3 , 4 5 - trihydroxybenzoyl
Smith, 1963) and mouthfeel (Millin et a l . , 1969a) of tea liquors. Thearubigins are a heterogenous group of compounds which are estimated according to the method described by Roberts and Smith (1963). They constitute about 10 to 20% of the dry weight of black tea (Roberts, 1962) and comprise 30 to 60% of the solids in brewed tea liquors. Roberts (1962) obtained thearubigin fractions (Sl, S1,, Sl,) by solvent extraction and found that their absorption spectra differed markedly from those of the theaflavins. He proposed that they were derived from theaflavins by oxidation leading to destruction of the benzotropolone nucleus. Vuataz and Brandenberger (1961), however, detected the presence of nitrogen in the thearubigins extracted with 80% ethanol-water and obtained fourteen amino acids on acid hydrolysis. Brown e f al. (1969a,b) isolated thearubigins by successive extraction of aqueous extracts of black tea with different solvents, followed by acidification and further extraction with n-butanol. This procedure yielded five main fractions, which yielded cyanidin and delphinidin, or flavan-3-01s and flavan-3-01 gallate, or gallic acid under different conditions of hydrolysis, and it was concluded on the basis of these findings that theambigins are mixtures of polymeric proanthocyanidins containing flavanoid residues. Berkowitz et al. (1971) investigated the oxidation of mixtures of epicatechin and gallic acid, as well as epicatechin gallate and gallic acid in a model tea fermentation system containing a crude soluble tea enzyme preparation, and obtained bright red phenolic compounds, which were identified as epitheaflavic acid (XXI) and its gallate (XXII). The epitheatlavic acids were found to be present in very low levels in black tea, probably because of further oxidation to thearubigins through coupled oxidation with oxidizing tea catechins (Berkowitz et a f . , 1971), as depicted in Fig. 3. From these diverse studies it is apparent that the mode of formation and nature of the theambigins has not, as yet, been satisfactorily resolved, and this is undoubtedly due to the complexity of this group of compounds.
6 . Development of Tea Aroma during Fermentation. The aroma of tea is an important parameter in the commercial valuation of tea, as is evident by the price
257
TEA Catechol oxidase
/I
Epicatechin
(02)
4
Epicatechin
Oxidized epicatechlne
7Thearubigins Epitheaflavic
acid-3'-gallate Gallic acid
Oxidized callie acid
FIG. 3 . Formation of epitheaflavic acid and epitheaflavir-3'-gallate and their transformation to theambigins. (From Berkowitz el a / . , 1971 .)
of tea with flavor being 2 to 3 times higher than that of a tea which is devoid of flavor. The available evidence suggests that black tea aroma develops during fermentation, and Yamanishi et al. (1966a) studied the changes in flavor constituents during the various stages of black tea manufacture and found an increase in almost all constituents during fermentation, especially in the contents of I-penten-3-01, cis-2-penteno1, benzylalcohol, rruns-2-hexena1, benzaldehyde, n-caproic, cis-3-hexenoic and salicylic acids. Saijo and Kuwabara (1967) also observed increases during fermentation of rruns-2-hexen- 1 -a1 and cis-3-hexenoic acid, as well as of n-capronaldehyde, together with a decrease in the amounts of n-hexyl alcohol, cis-3-hexen-1-01, and methyl salicylate. Two primary mechanisms have been suggested for the formation of volatile compounds, the first being dependent on the polyphenol oxidase mediated oxidation of tea flavanols (Sanderson, 1975) and the second on direct biosynthetic reactions (Wickremasinghe, 1974). The first of these is based on the evidence that oxidized flavanols cause oxidative degradation of compounds, notably amino acids, carotenes, and linolenic acid, during fermentation. It has been established (Bokuchava and Popov, 1954; Popov, 1956; Nakabayashi, 1958; Wickremasinghe and Swain, 1964; Saijo and Takeo, 1970a; Co and Sanderson, 1970; Saijo, 1973) that amino acids are transformed to carbonyl compounds in the presence of oxidized flavanols by Strecker degradation according to Eq. 5 . (5) RCHNH,COOH
+0
oxidized
tlavanols
RCHO
+ COO+ NH3
The bouquet of freshly brewed tea may be greatly influenced by the production of carbonyl compounds in this manner, e.g., the formation of phenylacetaldehyde from phenylalanine (Finot et ul., 1967; Saijo and Takeo, 1970b). c . Carotenes. It was suggested that the decrease in carotenes during black tea processing (Tirimanna and Wickremasinghe, 1965; Miiggler-Chavan er al., 1969) may have been due to their conversion to volatile compounds which contribute to tea flavor. This suggestion was examined by Sanderson et al.
258
ROBERT L. WICKREMASINGHE TABLE X BLACK TEA AROMA CONSTITUENTS SUPPOSED TO BE DERIVED FROM CAROTENOID COMPOUNDS',!' ~
Carotenoids found in tea leaves (Tirimanna and Wickremasinghe, 1965) p-Carotene
Secondary oxidation products
Primary oxidation products
Dih ydroactinidiole
+p-Ionone +TAKr
f 2.2-6-Trimethyl
cyclohexanone \r 5,6-Epoxy ionone 2,2,6-Trimethyl-6hydroxycyclohexanone
+
a -Carotene
Lutein
Phytoene
Theaspirone
+ +
Lycopene y-Carotene
Cryptoxanthin
+
Violaxanthin Zeaxanthin ~~
ja-lonone +p-lonone + TAK 3 (3-Hydroxy-/3 ionone)+ + (3-Hydroxya-ionone) + TAK jLinalool TAK jLinalool TAK 3 @-lonone +Linalool +TAK j/3-lonone +(3-Hydroxy-P-ionone) +TAK (3-Hydroxy-5.6-epoxy ionone + TAK 3 (3-Hydroxy-/3-ionone) + TAK
~
~
(Known reaction based on results of this investigation, + ; Highly probable reaction based on results of this investigation, 3 ; Probable reactions, + ; Supposed reactions, j; Compounds shown in brackets have not yet been identified in tea). Data of Sanderson er al. (1 97 1). TAK = Terpenoid-like aldehydes and ketones. Oxidation products of all the carotenoids listed.
(1971) and Kawashima and Yamanishi ( I 973) who obtained evidence that @carotene was converted during black tea processing to p-ionone and other compounds, which included (Reymond, 1976) damascenones, a-ionone, theas-
259
TEA
pirone, and dihydroactinidiolide. Using a model system, Sanderson et al. (1971) found that three basic ingredients, an active tea enzyme preparation, tea flavanols, and P-carotene, were necessary for the production of p-ionone, and the results further indicated that many of the important black tea aroma constituents are probably formed during black tea processing by oxidative degradation of the carotenoid compounds present in the system (Table X). d. Fatty Acids. The formation during black tea processing of trans-2hexenal from linolenic acid has been demonstrated by several workers (Saijo and Takeo, 1972; Gonzales et al., 1972; Hatanaka and Harada, 1973; Hatanaka et al., 1976b; Sekiya et al., 1976). This reaction assumes importance in view of the finding (Gianturco et al., 1974), that the proportion of trans-2-hexenal in the aroma complex is one of the factors which determine the flavor of tea. Hatanaka and his associates have proposed the mechanism indicated in Fig. 4 for the formation of trans-2-hexenal from linolenic acid. The second mechanism for the formation of volatile compounds during fermentation is direct biosynthesis. Wickremasinghe and Swain (1 965) observed that flavory teas contained less leucine than nonflavory teas and this, together with the correlation between flavor and the formation of a-ketoisocaproic acid during withering (Wickremasinghe, 1964) led to the suggestion (Wickremasinghe, 1967) that L-leucine could be a precursor of compounds responsible for tea flavor. This suggestion was supported by the identification of leucine-a-ketoglutarate transaminase in tea flush (Wickremasinghe et al., 1969). and the conversion of 14C-leucine to 14C-mevalonic acid and unidentified 14Cvolatile compounds during tea processing (Wickremasinghe and Sivapalan, 1966). In this study, it was found that I4C-acetate too could act as precursor of mevalonic acid and of volatile compounds, among which nerolidol has been
enzyme
Phospholipid
Linolenic acid
1
enzyme*/O,
IPeroxide)
t
ADH'
ris-3-Heienal
1
GZ=
cis-3-Herenol (leaf alcohol)
enzyme *
It-wi.+2-Hexenal (leaf aldehyde)
ADH' ZZ=
tmns-2-Hexenol
*Located I n chloroplast +Alcohol dehydrogenase
FIG. 4. Mechanism of formation of trans-2-hexenal from linolenic acid. [Based on Hatanaka cr a!. (1976b) and Sekiyaer al. (1976).]
260
ROBERT L. WICKREMASINGHE
identified (Saijo and Uritani, 1971). It was proposed (Wickremasinghe, 1974) that the well-established fact that flavor develops in conditions of climatic stress was related to whether leucine or acetate acted as precursor of volatile compounds. In conditions that are favorable for plant growth, normal intrachloroplastidic reactions are dominant and lead to the formation of acetate which acts as precursor of the volatile compounds; whereas in conditions of climatic stress, extrachloroplastidic biogenesis of terpenoid compounds from leucine, rather than from acetate, is operative. It was proposed that the acetate pathway favored the formation of linolenic acid which was converted to trans-2-hexenal during processing; whereas formation of this aldehyde, an excess of which detracts from tea flavor, is minimal when the leucine pathway comes into operation. Additionally, the qualitative and quantitative changes in the tea leaf carotenoid composition under conditions of climatic stress lead to the formation of compounds (such as p-ionone, theaspirone, and dihydroactinidiolide) in those proportions which were necessary for the organoleptic perception of flavor. 5.
Firing
Firing is the final stage of tea processing when the rolled and fermented leaf having a moisture content of about 45-50% is dried to produce a black tea containing 3% moisture. This is accomplished by blowing hot air through the fermented leaf as it is conveyed on an endless chain. The temperature of the hot air at the inlet is 87"-93"C, and that of the outlet is 56"-57"C, and the drying process normally takes about 20 minutes. More recently the technology of fluid bed drying has been applied to the firing of tea (Kirtisinghe, 1974), using a temperature of 125°C for 20 minutes. Important changes occur during the firing stages of tea processing and some of these are: a. loss of moisture to a level (3%) which makes the product suitable for storage; b. arrest of fermentation reactions due to destruction of polyphenol oxidase and other enzymes. There is, however, some acceleration of enzyme-mediated reactions during the initial stages of firing, and 10-15% of the theaflavin content of black tea is formed during the first 10 minutes of firing (Wickremasinghe, unpublished); c. conversion of chlorophyll to pheophytin, which imparts to black tea the desired black appearance (Wickremasinghe and Perera, 1966a). This transformation occurs at the elevated temperature of firing in the acidic conditions of the tea leaf. It has been observed, in this connection (Wickremasinghe, unpublished), that freeze drying, or raising the pH of the fermented leaf, yields a brown product which is not acceptable to the tea trade; d. reduction of astringency of fermented leaf due to combination of polyphenols with tea leaf proteins at the elevated firing temperature (Wickremasinghe and Swain, 1965). Prior to firing the taste is harsh and metallic, but this mellows on firing;
TEA
26 1
e. firing is essential for the development of black tea aroma (Bhatia and Ullah, 1965) because the loss of low boiling volatile compounds (Yamanishi et al., 1966a) is accompanied by the formation of other compounds which are considered to be important constituents of black tea aroma, e.g., p-ionone (Sanderson et a / . , 1971; Kawashima and Yamanishi, 1973), theaspirone, and dihydroactinidiolide (Ina et af., 1968). The resultant change in the relative proportions of the different volatile constituents probably has an important effect on the overall flavor of the tea (Yamanishi et al., 1968a). Additionally, the pyrazines, pyridines, and quinolines detected in the basic fraction of tea (Vitzthum et al., 1975) are probably formed during firing as a result of interaction between free sugars and amino acids (Reymond, 1976). 6 . Grading Grading is based entirely on the physical separation of the different sizes of particles of the fired tea. This is achieved by the use of mechanically oscillated sieves fitted with mesh of varying sizes, and the different grades of tea are defined by the mesh size of the sieves. Some of the tea grades which are commonly produced are known as Broken Orange Pekoe (B.O.P.), Broken Pekoe (B.P.), Broken Orange Pekoe Fannings (B.O.P.F.), Orange Pekoe (O.P.), Flowery Broken Orange Pekoe (F.B.O.P.), Fannings, and Dust. The tea is finally cleaned by winnowing to remove the fine dust and fiber and passing through a stalk extractor, which works on the principle of electrostatic attraction. Each grade of tea is then packed in plywood boxes and sealed with similar material. Further particulars regarding the grading of tea are described by Harler (1 963). B.
GREENTEA
The world production of green tea is quantitatively less than that of black tea, but it is the principal form in which tea is drunk in several countries, e.g., China, Japan, Taiwan, and Indonesia. Small-leaved tea varieties (so-called China or low jat types) are generally more suitable than the large-leaved varieties (so-called Aassam or high jat types) for green tea production, although Indonesian green tea is, in fact, produced from the large-leaved varieties. In Japan, green tea is manufactured from particular clones of tea (e.g., Yabukita, Natsumidori, Tamamidori), whereas black tea is made from other clones (e.g., Benihomare, Benifuji); the principal chemical differences between green and black teas appear to be the relatively higher content of amino acids and the lower content of polyphenols in the former (Table XI). Recognition of these differences is reflected in agricultural practices where every effort is made to increase the content of nitrogenous material in the leaf by the liberal application of fertilizer, while reduction of polyphenolic material is achieved by shading.
262
ROBERT L. WICKREMASINGHE
TABLE XI COMPARISON OF VARIOUS TYPES OF GREEN TEA AND BLACK TEA“
Tea
Polyphenols
Amino acids
Finest grade (Japan) Popular grade (Japan) Popular grade (China) Black tea High grown (Sri Lanka) Low grown (Sri Lanka)
132 229 258
48 21 18
280 302
16 17
Green tea
“ Values expressed as mg/grn dty weight
Processing of Green Tea
As mentioned earlier, the main difference in processing of green tea and black tea is the prevention of fermentation in the former by initial heat inactivation of the enzymes present in the tea flush. This inactivation is achieved in Japan by steaming the flush, and in China, Taiwan, and Indonesia by a process known as “panning” where the flush is fed into a hot rotating drum. In Japan the steps of manufacture are described as follows in an undated publication of the Ministry of Agriculture and Forestry, Japan: “SteamingPrimary rolling and firing-Rolling-Secondary drying-Final rolling-Final drying. ” Steaming is the first step of green tea manufacture in which the polyphenol oxidase and other enzymes are inactivated, and the green color of the leaf maintained. Steam is introduced from a boiler into a rotating cylindrical drum containing the tea leaves and the period of exposure to steam is 15-20 seconds. The heated leaves are cooled by a fan as soon as possible after steaming; and in the next stage of primary rolling and firing, the leaves are introduced into a wooden box containing a rotating shaft fitted with sweepers and forks, through which hot air is blown. The leaves are subjected to this treatment for about 55 minutes, when the moisture content is reduced to about 50%. They are then rolled under pressure for 10 minutes, and dried again in a rotating drum for about 20 minutes, until the moisture content drops to about 30%. Final rolling is for 35 minutes in heated machines which impart a twist to the leaf, and the final drying is by hot air at about 65°C until the moisture content is reduced to 3 4 % . In the panning process of green tea manufacture, the steps followed are similar to those for the steaming process, except that the initial inactivation is effected by dry heat (Wu, 1976). An important component of green tea flavor is dimethylsulfide, the precursor of which was identified as methylmethionine sulfonium salt (Kiribuchi and
TEA
263
Yamanishi, 1963). The other volatile constituents of green tea have been extensively studied by Yamanishietul. (1956, 1957, 1963, 1965, 1966a,b. 1970) and Nose er ul. (197 1). A list of the numerous compounds identified, together with a comparison of their occurrence in unprocessed tea leaf and black tea, has been compiled by Yamanishi ( I 975). A further step in the processing of green tea commonly employed in Indonesia, and which is gaining popularity in Japan. is further roasting at 200°C. This roasted green tea (“Hoji-cha”) has an aroma which is quite distinct from green tea, and was found to contain a total of 66 compounds, including 21 pyrazines (Yarnanishi et ul., 1973). In a study of the changes in aroma components during the roasting of green tea, Hara and Kubota ( 1 973a) described the production of pyrazines, furans, and pyrroles, and it was also found (Hara and Kubota, 1973b) that the amount of carbonyl compounds doubled during roasting, presumably due to Strecker degradation of the free amino acids in green tea. In a study of the polyphenols in roasted green tea, Nakagawa ( 1 967) found that these had undergone marked epimerization, polymerization, and thermal decomposition with an accompanying decrease in the amounts of flavanol gallates.
IV. ORGANOLEPTIC PROPERTIES A.
BLACKTEA
Evaluation of tea is carried out by skilled and experienced tasters who determine the market value of a tea on a purely subjective basis. Apart from the appearance (black or brownish) of the dry tea and the color (coppery or greenish) of the leaf after infusion, it is the tea taster’s palate which assesses the characteristics of the brewed liquor. Some of the important characteristics are the “color.” “strength,” “quality,” and “briskness” of the tea liquor, as well as the formation of “cream” on cooling. These, however, are only a selected few of the terms employed in the tea trade and a fuller description is given by Harler (1963) and Eden (1976). The relationship between color and strength of tea liquors in terms of chemical compounds was studied by Roberts and his co-workers (Roberts, 1962) who concluded that apart from small contributions by flavanotropolones, triacetidein, and possibly products of nonenzymic browning, the color or a liquor is due to theaflavins and thearubigins. Strength of tea liquors was found to be directly related to oxidase activity and polyphenol content in the green leaf, and considered to be determined by the oxidation products produced during fermentation. In subsequent studies, Roberts and Smith (1963), and Nakagawa (1969), found a positive correlation between theaflavin content and the theaflavidthearubigin ratio and the tea tasters’ assessment of quality, color, and strength of tea brews. Similar results were obtained by Takeo (1974b) who observed that the optical
264
ROBERT L. WICKREMASINGHE
density values of theaflavins, thearubigins, and also theaflavins plus thearubigins showed a high positive correlation with the quality of tea infusions. In a statistical evaluation of North-East Indian Plains teas, Biswas et al. (1 97 1) found that total oxygen uptake of unprocessed tea shoots, and the theaflavin, epicatechin gallate, and theogallin contents of the processed black teas determined the cash valuation, which was itself found to be dependent on quality and/or briskness (Biswas and Biswas, 1971). From a study of the characteristics of black teas processed from the tender stem, and tea leaves of differing maturity, Wickremasinghe and Perera (1 973) concluded that the factors affecting quality, strength and color of black tea liquors were the proportions of different polyphenols, polyphenol oxidase activity, and contents of caffeine, theanine, theogallin, and the unidentified compound G 36 detected by Forrest and Bendall ( I 969). The ability of black tea infusion to “cream” and the color of the cream formed (bright or dull) is one of the yardsticks employed by tea tasters to judge the quality of tea. “Cream” is the haze or precipitate that is formed when a strong infusion of tea is allowed to cool down; the relationship between cream formation and quality was first demonstrated by Bradfield and Penny (1944). It was found by Roberts (1963) and Bhatia (1964) that the theaflavins, thearubigins, and caffeine were the main constituents of cream, and later studies (Wickremasinghe and Perera, 1966b; Smith, 1968) showed that theobromine, theaflavin gallate, epigallocatechin gallate, epicatechin gallate, triacetidin, caffeic acid, gallic acid, ellagic acid, chlorophyll, bisflavanols A and B, flavonol glycosides, and mineral matter were also components of the cream complex. A relatively high proportion of theaflavins may be expected to yield a cream which has the desirable bright color, whereas a high proportion of thearubigins results in a dull cream (Wickremasinghe and Perera, 1966b). In this empirical practice, the tea taster is therefore assessing the black tea infusion for its content of theaflavin gallate. A study of the contribution of the nonvolatile compounds of black tea to the character of the beverage was made (Millin et al., 1969a) by tasting pure compounds and various fractions isolated from black tea liquors. It was found that, with the exception of caffeine, none of the monomeric nonvolatile substances examined (flavanols, flavonols, theogallin, chlorogenic acid, p-coumarylquinic acid, caffeic acid, theanine) contributed significantly to the taste of the beverage. Among the oxidation products of flavanols, theaflavin and other oxidation products of intermediate molecular weight were astringent, and it was suggested that together with caffeine, and in the absence of protein, these compounds could influence briskness and strength of tea liquors. Oxidation products of high molecular weight were thought to be responsible for “soft,” “flat,” “thin” liquors, and overall quality depended on the correct balance of a number of substances. Sanderson et al. (1976) confirmed that the astringency of a tea beverage is largely dependent on the amount of polyphenolic compounds present, the degree of oxidation (polymerization) of the tea flavanols, and particu-
TEA
265
larly by the amount of galloyl groups present on the flavanols and their oxidation products. These workers also confirmed the beneficial effect of caffeine on the briskness of tea brews (Roberts, 1962; Wood and Roberts, 1964; Millin et ul., 1969a) and demonstrated that the addition of milk or lemon juice modified the taste of tea polyphenolics. The effect of flavor constituents on the character of black tea has been evaluated by a number of workers; and, in gas chromatographic separations of the aroma constituents, it was observed (Yamanishi et al.. 1968b) that the proportion of compounds having a R, value greater than linalool to those of lower R, value was related to the country of origin of the black tea. In an investigation of Ceylon tea, Yamanishi et al. (1968a) found that the proportion of compounds of relatively high boiling point to more volatile compounds was several times higher in black teas with quality than in those devoid of this characteristic. Similar results were obtained on comparison of flavory and nonflavory Ceylon black teas (Wickremasinghe et a / . , 1973) and in a study of the seasonal variations in the composition of volatile constituents of black teas (Gianturco et al., 1974). Using a mathematical approach to the evaluation of tea quality, Vuataz and Reymond (1 970) identified three regions, namely ionones, linalool. and dimethylsulfide of gas chromatograms as being positively related to quality. It is evident, therefore, that several parameters exist for the evaluation of the characteristics of black teas, but in this plurality lies their disadvantage. An experienced tea taster assesses several hundred of samples in a working day and it is still impractical to replace him by chemical procedures that would need several weeks of work by an experienced chemist.
B.
GREEN TEA
Green tea brews, unlike those of black tea, contain no highly colored products formed by the oxidation of polyphenolic compounds, and the desired color is greenish or lemon yellow without any trace of red or brown color. The liquor should remain clear on cooling without any turbidity, and the infused leaf should be green with no sign of discoloration due to damage. The desired greenish yellow color of the liquor is believed to be dependent on the composition and content of flavonols and their glycosides. as well as flavones (Sakamoto, 1967, 1970). Nakagawa ( 1970) studied the correlation between chemical composition and organoleptic properties by analysis and sensory evaluation of various grades of tea, that is, Sencha (common green tea) of high, medium, and low grades and Gyokuro (the finest grade of green tea). His results indicated that the multiple correlation coefficient between sensory evaluation and catechins, amino acids, caffeine, and other soluble substances was highly significant. In a subsequent study, Nakagawa and Ishima (1971) reported that the taste of green tea brews was affected by the contents of aspartic acid, glutamic acid, theanine, epigal-
266
ROBERT L. WICKREMASINGHE
locatechin gallate, epigallocatechin, other catechins, and other soluble residues. In a more detailed study, Nakagawa (1975a,b) concluded that defined compounds contributed to the four main taste sensations of tea which he described as bitter, astringent, brothy , and sweet. In this investigation, tea infusions were separated to five fractions by gel filtration and the results of chemical and organoleptic tests indicated that the bitterness and astringency of green tea brews was determined by catechins and other phenolic compounds, the brothy taste by amino acids (particularly theanine), and sweetness by sugars. It was considered that the balance between these different tastes was of importance, and that individual taste elements could be accentuated by the conditions of brewing.
V. STORAGE OF TEA A.
TEALEAVES
Changes of the chemical constituents of tea leaves during their storage for 10-21 days at 5" or 10°C in atmospheres containing varying proportions of nitrogen, oxygen, and carbon dioxide was investigated by Tsushida et al. (1976) who found that total ascorbic acid content decreased markedly after storage for two weeks in a manner which was positively related to the content of oxygen in atmosphere. The amount of tannin and caffeine were not greatly affected by storage and in the conditions used; glutamic acid, aspartic acid, and theanine decreased, while glutamine and asparagine increased. Ethylene was produced by the stored leaves, and the extent of production was directly related to the oxygen concentration in the storage atmosphere. Zarnadze (1971) found that storage of fresh leaf at O"-l"C and 96-100% relative humidity for up to 10 days did not affect its suitability for processing. In these storage conditions, there was an increase in polyphenol oxidase, peroxidase, and invertase activities, a decrease in chlorophyll and protopectin contents, but no significant change in the level of catechins. In a study of the essential oils in tea leaves stored at 5°C for 10-12 days, Takeo (1956b) reported the occurrence of quantitative changes in some of the volatile compounds, accompanied by a deterioration in the flavor of green tea and black tea made from the stored leaves.
B. TEABREWS Storage of black tea brews was observed to lead to darkening of the liquors (Roberts, 1959), and was accompanied by an increase in the amount of "nondialyzable theambigins" (Roberts, 1961). The occurrence in tea brews of substantial amounts of nondialyzable material, containing polysaccharide as the major component together with protein, nucleic acid, and polyphenols, was described by Millin et al. (1969b); it was also found (Millin el al., 1969c) that
TEA
267
darkening of tea brews was accelerated by heating and accompanied by a large increase in the quantity of nondialyzable polyphenolic material. C.
GREENTEA
Storage of processed green tea was found (Furuya, 1970) to lead to deterioration of aroma, color, liquor characteristics, taste, and ascorbic acid content. The factors responsible for this loss of quality were considered to be the moisture and oxygen content in the atmosphere and the temperature of storage.
D.
BLACKTEA
Brews of freshly fired black tea have a “raw” or “green” taste but after storage of the fired tea for some weeks this rawness is replaced by a balanced astringency and flavor. However, prolonged storage for several months, especially under unfavorable conditions of exposure to light, elevated temperatures, and high humidity, causes considerable loss of astringency and flavor and the brew is considered, in tea taster’s parlance. to be “flat” or “soft.” Hearne and Lee (19%) observed that the amount of carbon dioxide produced during the storage of black tea was in excess of that expected from “browning” reactions, and suggested that breakdown of constituents occurred during the aging of tea. It was also observed that the extent of carbon dioxide production was dependent mainly on moisture content and temperature of storage of the tea. These results were confirmed and extended by Roberts and Smith ( I 963), who found that carbon dioxide production was associated with a loss of theaflavin. Studies of the changes occurring under different conditions of storage (Wickremasinghe and Perera, 1972b) showed that, during a period of 22 weeks, moisture content of the tea rose from 4.2 to only 5% in an airtight container, whereas it increased to 9.9% in conditions when air was not excluded. Theaflavin and epicatechin gallate content decreased during storage, and levels of thearubigins, amino acids, and total polyphenols showed an undulating pattern. Cash valuations of the samples increased during the first few weeks after storage, after which they declined. In a more detailed study, Stagg (1974) found a decrease of theaflavin and creaming index and an increase of nondialyzable material with time; these trends were accelerated by moisture uptake. There was also a reduction in content of free amino acids (particularly theanine), glucose, volatile compounds (particularly aliphatic aldehydes and aocohols), whereas the levels of total lipids and total fatty acids (particularly palmitic and other free fatty acids of chain length above C 12) increased. The absorption of moisture was considered to be the most important single parameter operative during storage because all the changes described (with the exception of lipid oxidation) were accelerated by moisture contents in excess of 6.5 to 7.5%, and it was suggested that optimal conditions for storage of black tea would include maintenance of its moisture content in the range of 3 to 5% at a temperature below 30°C.
268
ROBERT L. WICKREMASINGHE
VI. POTENTIAL BY-PRODUCTS Black, green, and instant teas together with smaller amounts of oolong and paochong tea are the only products obtained from the tea bush, and, at the present time, none of the other products obtainable is being exploited on even a small scale. A list of these potential by-products has been compiled (Wickremasinghe, 1972) and it is possible that some of them may become useful in the future. Considering each of the possible by-products individually, the extraction of caffeine from waste tea (such as the residue of stalk and fiber remaining after black tea processing) was at one time a profitable industry. Tea waste contains 1.5 to 3% caffeine and there are numerous patented processes for its extraction but these have diminished in importance due to competition from synthetic caffeine. A method for the extraction of vitamin-containing food dyes is the subject of a patent (Bokuchava and Pruidze, 1970), where the starting material is any tea forming material or nonstandard tea leaf. This raw material is heated in hot air or water vapor at about 75°C for 1-10 minutes, after which it is ground, and extracted with aqueous alcohol at 50°-55"C to produce a green food dye, or with water at 65"-80"C to produce a yellow dye. For production of a brown food dye the yellow aqueous extract was heated at about 90°C at atmospheric pressure or at 130" to 200°C at 1.5-1.6 atmospheres. After extraction all dyes were filtered and either spray or freeze-dried. The use of tocopherols extracted from tea flush as antioxidants has been suggested by Tirimannaet al. (1967), who found that a-tocopherol accounted for an appreciable portion of the total tocopherol content of 0.387 mg/gm (dry weight basis). It has also been found (Lea and Svoboda, 1957) that the unoxidized polyphenols extractable from tea may be used as antioxidants, but neither of these possibilities has been placed on a commercial basis. The possibility of using waste products of the tea industry as a source of protein for cattle feed and for composting has been considered by Croyle et al. (1974) who suggested that spent (extracted) tea leaves from instant tea processing plants may be used for this purpose. It was found that this material contained about 25% crude protein, the availability of which would need to be improved by acid-heat or fermentation treatments, prior to its utilization as a feed material. Seeds of Camellia sinensis contain about 20% of oil which is remarkably similar to olive oil in composition (Chakrabarty and Chakrabarty, 1954; Roberts and de Silva, 1972). The content of oil in seeds of Camellia sasanqua is higher (about 35%) and an unconfirmed report (Reddy, 1958) states that tea seed oil production was, at one time, a commercial enterprise in China. The main difficulty at present, would appear to be the difficulty in obtaining the tea seeds themselves because the current practice of vegetative propagation of tea has reduced the availability of seed. Seed residues after extraction of oil contains about 15% (dry weight basis) triterpenoid saponins (de Silva and Roberts, 1972)
TEA
269
which may be used as detergents, or foaming agents, but the extraction of tea seeds for these products is not, at the present time, considered to be a commercially feasible proposition.
VII. CLINICAL EFFECTS A comprehensive account of the nutritional and therapeutic value of tea has been recently published by Stagg and Millin (1975), and this together with the reviews of Das et a1. (1964, 1965) cover much of the published literature on this subject. The more important constituents of tea, from a clinical point of view, are considered to be caffeine and the polyphenol fraction, because many of the beneficial effects claimed for tea may be traced to these constituents. However, as indicated by Stagg and Millin (1975), these effects of tea may be due to interaction between a number of compounds rather than to any single component or group of components. Caffeine is a vasodilator having diuretic and stimulant properties, and tea drinking has been recommended for the treatment of a variety of disorders (Krantz, 1955; Das et al., 1965) since an average cup of tea contains about 40 mg caffeine (Sanderson ef al., 1976). Furthermore, it seems that harmful effects of caffeine, such as a rise in fatty acids in the blood, are not apparent when caffeine is administered in the form of tea (Akinyaju and Yudkin, 1967), and it has been suggested (Stagg and Millin, 1975) that a possible reason for the ameliorating effect could be due to cream formation (see Section IV, A) which may influence the rate of assimilation of caffeine from tea. The other group of medically important compounds of tea are the polyphenols which constitute 48.5% of the total solids in a cup of tea (Sanderson et al., 1976). These polyphenols include the flavanols and flavanol gallates, flavonol glycosides, theaflavins, thearubigins, bisflavanols, epitheaflavic acid, gallic acid, and chlorogenic acid; a variety of pharmacological activities have been ascribed to this group of compounds. It has, for instance, been claimed for many years that tea polyphenols possess the property of strengthening the walls of blood vessels and regulating their permeability (Ul’yanova and Erofeyeva, 1966) and the substances responsible for this action were designated vitamin P. Preparations derived by extraction of green tea leaf are marketed in the U.S.S.R. for their content of this vitamin, although U.S. Food and Drug Administration in 1968 sought withdrawal of such bioflavonoid drugs from the market on the grounds that they were ineffective. Other effects of tea flavonoids discussed in the review by Stagg and Millin (1975) are their effect in increasing levels of catecholamines, the capillary strengthening action, the anti-inflammatory action, normalization of thyroid hyperfunction causing thyrotoxicosis, protection against the harmful effects of exposure to radiation, bacteriostatic effect on a number of microorganisms, and stimulation of folic acid biosynthesis. In this connection, it has been pointed out (Stagg and Millin, 1975) that the common practice of referring to tea polyphenols as “tannins” is misleading, because the strong, irreversible
270
ROBERT L. WICKREMASINGHE
protein-binding effects associated with true tannins are not exhibited by the tea pol yphenols. Several reports have appeared on the antiatherosclerotic effect of tea (Little et al., 1966; Akinyaju and Yudkin, 1967; Young et al., 1967; Naismith et ul., 1969; Mahendra et al., 1972). The results (Table XII) of Mahendra et al. (1972) on feeding mice with an atherogenic diet containing respectively, whole and fractionated black tea extract, green tea extract, and coffee indicated that tea, but not coffee, had a beneficial effect in combating the rise of serum cholesterol, triglyceride, and total esterified fatty acids. Black tea was not very different from green tea in its effect, but the black tea fractions separated by gel filtration (Wickremasinghe, 1977) were more effective than whole black tea, and the fraction containing the polyphenols was found to be especially effective in reducing triglyceride levels. Black tea has a relatively high content of fluoride (Cheng and Chou, 1940; Zimmerman et al., 1957; Singer et al., 1967; Okada and Furuya, 1969) and it was found that a black tea brew provided 1-2 ppm fluoride (Karunanayake et al., 1972) indicating that tea drinking could make a significant contribution to the fluoride intake, which is required for the prevention of dental caries (Schwerp, 1971). All of the properties of tea mentioned above are beneficial to human health, but there are also reports that tea may be injurious. Among these is the claim (Fedrick, 1974) that there is some correlation between tea drinking among expectant mothers and the subsequent incidence of anencephalic births. However several factors cast doubt on the validity of this claim as the experimental procedure, interpretation of the statistics, incorrect assumptions, and lack of a doserelated effect do not support the conclusion drawn (Stagg and Millin, 1975). It has also been reported that the phenols of tea have the effect of promoting the TABLE XI1 EFFECT OF TEA AND COFFEE ON SERUM LIPID LEVELS OF RATS ON AN ATHEROGENIC DIET
Treatment
Total esterified fatty acids
None Coffee Green tea Black tea Black tea Fraction 1" Black tea Fraction 2"
Triglycerides
Cholesterol
8
156
8 6
102 70 64 90 48
108 111 91 92 64 67
5.4 6.5 5.8
" Contains non-polyphenolic material of high molecular weight.
* Contains polyphenols and other material of low moleculer weight (e.g., caffeine and amino acids).
TEA
27 1
carcinogenic effect of benzpyrene painted on the necks of mice (Kaiser, 1967), but the lack of adequate controls in the experimentation procedure nullifies the inference drawn. It has also been claimed (Morton, 1972), on the basis of epidemiological study, that there may be a correlation between the incidence of esophageal cancer and the consumption of tannin rich plants, among which tea is included, but this conclusion rests entirely on evidence of a purely speculative nature.
VIII.
HOST PLANT-PEST RELATIONSHIPS
Studies of host plant-pest relationships in tea are few and comparatively recent, but there are already indications that the results obtained may afford practical means of pest control. In a study of the factors influencing the infestation of Ceylon tea by the scolytid beetle, Xyleborus fornicutus, it was found (Wickremasinghe et ul., 1976) that ambient temperature determined the distribution of this pest, while moisture content and the availability of a-spinasterol (a possible precursor of insect moulting hormones) determined the degree of infestation by the beetle. It was found that (1) the beetle preferentially attacked that portion of the tea stem which had a moisture content of 61-63%, as well as a relatively low level of saponins. the mid-portion of the stem (Table VI); (2) those clones of tea which were tolerant to infestation by X . fornicatus contained more saponin than those which were susceptible; and (3) tea saponins had the property of binding sterol, and so reducing its availability. On the basis of these findings, it has been suggested that control of X . .fornicutus could be effected by measures which increased the amount of saponin in the host plant. Another use of saponins is the improvement of laboratory methods for the recovery of nematodes from tea roots (Sivapalan, 1976) where it was found that incorporation of saponin to the extraction medium lead to a fourfold increase in the number of nematodes recoverable from the roots. In a study of the relationship between polyphenol content of tea roots and degree of susceptibility to nematode infestation, Sivapalan and Shivanandarajah (1974) reported that there was a significant increase in the total free polyphenol content in the feeder roots of nematode-tolerant clones following infestation with the root lesion nematode, Prutylenchus loosi, whereas infestation of nematode susceptible clones lead to a decrease of polyphenol content. On the basis of this finding. it was suggested that polyphenols played an important role in determining infestation of tea roots by P. loosi. A correlation between the carotenoid pigment, rhodoxanthin, and the degree of attack of tea by the red spider mite Oligonychus coffeue Nietn., was proposed by Fernando (1967) who found that the severity of mite infestation was related to the rhodoxanthin content of the tea leaf, and it was suggested that this carotenoid probably acted as a phagostimulant or reproductive stimulant.
272
ROBERT L . WICKREMASINGHE
IX. INSTANT TEA The commercial production of instant teas began in the 194Os, and this form of tea has grown in popularity in the United States to the extent that it now comprises 42% of tea sales in that country, although on a worldwide basis instant tea accounts for only a very small proportion (less than 5%) of consumption. The demand in the United States is for instant teas soluble in cold water, because it is iced tea which is the real basis for the success of instant tea in that country. In other countries, hot tea is the preferred beverage, and it would appear that here the popularity of tea bags has been the factor responsible for the limited consumption of instant teas. The methods used for instant tea production have been protected by patents, and the patents published up to 1969 have been reviewed by Pintauro (1977). The basic steps in the preparation of instant teas are extraction of tea solids from fermented but unfired tea leaf, black tea, or green tea, followed by concentration of the extract, and drying of the concentrate to a powder. Extraction may be effected by a variety of methods among which counter current extraction and percolation methods have been widely used. Concentration of the extract is effected by evaporation of the water under reduced pressure at a moderately elevated temperature, and during this process various methods for trapping the escaping volatile compounds have been devised. These trapped volatiles are concentrated and retained for incorporation into the final dried product. The concentrated extract is turbid due to the formation of cream (see Section IV, A), and solubilization of this cream is a fundamental problem in the production of instant teas soluble in cold water. Methods for solubilization include treatment with tannase (Takino, 1971), and the use of sulfites and of oxidation as outlined by Sanderson (1 972b), who also discusses the additional problem of “dehazing” for imparting hard water stability to instant tea products. The final step of drying the concentrated tea extract is commonly achieved by spray drying, but other methods, such as freeze-drying or drum-drying are the subject of published patents. The importance of instant teas to the World Tea Industry may be gauged by the intense activity which is current in the field, and this activity is itself an indication that a truly acceptable instant tea has not been yet produced.
X.
ADDITIONAL RESEARCH NEEDS
A great deal of fundamental work has been done in research into various aspects of tea cultivation and tea processing during the last few decades. There are, however, several fields of investigation which remain open, one of which is the development of methods for the rapid selection of planting material which
TEA
273
will provide bushes with a capacity for higher yields, preferably by reason of a genetic constitution which makes for more efficient absorption and utilization of the fertilizer applied. Yield increase may also be achieved by a systematic study of host plant-pest relationship in tea, which could afford a means for a biological and integrated program of pest control without placing undue reliance on the necessity for chemical pesticides and their attendant drawbacks. In the field of biochemistry, a more thorough understanding of the mechanism of biosynthesis of compounds contributing to the unique characteristics of tea may, in the future, provide a means of tailoring conditions during the cultivation and processing of tea to realize the full potential of the tea leaf. At the same time, the numerous clinical effects need to be studied in greater detail, as these show promise of providing a pleasant, inexpensive and easy method for promoting the general well-being and better health of tea consumers. Better packaging and improvements in technology for prolonging the storage life of tea are also areas which need further study, particularly in view of the rather long time lag between production and consumption, and finally, innovations leading to the presentation of tea in a convenient form will contribute to the continuance of tea as being the most popular beverage in the world.
REFERENCES Akinyaju. P., and Yudkin, J . 1967. Effect of coffee and tea on the serum lipids in the rat. Nature (London) 214, 426-427. Anan. T.. and Nagakawa, M. 1974. Effect of light on the chemical constituents in the tea leaves. Nippon Nogei Kagaku Kaishi 48, 9 1-96, Association of Official Agricultural Chemists. 1960. 1n “Methods of Analysis,” 9th Ed., p. 185. Assoc. Off. Agric. Chem., Washington, D.C. Baker. J . E.. and Takeo. T. 1974. Acid phosphatases in plant tissues: Changes in activity and multiple forms in tea leaves and tomato fruit during maturation and senescence. Srudy Tea 46, 63-75. Bendall, D. S . . and Gregory. R. P. F. 1963. Purification of phenol oxidases. In “Enzyme Chemistry of Phenolic Compounds” (J. B. Pridham. ed.). pp. 7-24. Pergainon, Oxford. Berkowitz, J . E., Coggon. P., and Sanderson, G.W 1971, Formation of epitheaflavic acid and its transformation to thearubigins during tea fermentation. P hyrochemisrry 10, 227 1-2278. Bezbaruah. H . P. 1974. Tea breeding-a review. Indian J . Genet. Planf Breed. M A , 89-100. Bhatia. I. S. 1964. The role of chemistry in tea manufacture. Two Bud 2, 109-1 18. Bhatia. I . S . , and Deb, S. B. 1965. Nitrogen metabolism of detached tea shoots. I . Changes in amino acids and amides of tea shoots during withering. J . Sci. Food Agric. 16, 759-769. Bhatia. I . S., and Ullah, M. R. 1961. Oxidation of I-epicatechin gallate during the processing of Assam tea leaf. Chem. Ind. (London) p. 1169. Bhatia, 1. S., and Ullah, M. R . 1965. Quantitative changes in the polyphenols during the processing of tea leaf and their relation to liquor characters of made tea. J . Sci. Food Agric. 16, 408-416.
214
ROBERT L. WICKREMASINGHE
Bhatia, I. S., and Ullah, M. R. 1968. Polyphenols of tea. IV. Qualitative and quantitative study of the polyphenols of different organs of some cultivated varieties of tea plant. J . Sci. Food Agric. 19, 535-542. Bhattacharya. A. K., and Ghosh, J. J. 1968. Studies on the ribonucleic acids of fresh and processed leaves. Biochem. J . 108, 121-124. Biswas, A . K., and Biswas. A . K . 1971. Biological and chemical factors affecting the valuations of North-East Indian plains teas. I. Statistical association of liquor characteristics with cash valuations of black teas. J . Sci. Food. Agric. 22, 191-195. Biswas, A. K., Biswas, A. K . , and Sarkar, A. K. 1971. Biological and chemical factors affecting the valuation of North East Indian plains teas. 11. Statistical evaluation of the biochemical constituents and their effects on briskness, quality and cash valuations of black teas. J . Sci. Food Agric. 22, 196-204. Blanc, M. 1972. Chlorophyllic pigments of tea. Distribution and behaviour during black tea manufacture. Lebensm.-Wiss. Technol. 5 , 115-1 17. Bokuchava, M. A. 1950. Transformation of tannins under the action of polyphenol oxidase and peroxidase. Biokhimiya 6, 100-109. Bokuchava, M. A., and Popov, V. R. 1954. The significance of amino acids in the formation of tea aroma by interactions with tannin substances under conditions of elevated temperature. Dokl. Akad. Nauk SSSR 99, 145-148. (Chem. Abstr. 29, 2329.) Bokuchava. M. A., and Pruidze, G. N. 1970. Production of vitamin containing food dyes from a non-standard tea leaf and tea forming materials. Otkrytina Izobret, Prom. Obraztsy. Tovaruye Zuaki 47, 6 3 6 4 . (Chem. Absrr. 73, 129748.) Bokuchava, M. A . , and Skoboleva, N . I. 1969. The chemistry and biochemistry of tea and tea manufacture. Adv. Food Res. 17, 215-292. Bokuchava, M. A., Shalamberidze, T. K., and Soboleva, G. A. 1970. Localization of polyphenol oxidase in a tea leaf. Dokl. Akad. Nauk SSSR 192, 1374-1375. Bradfield, A. E., and Bate-Smith, E. C. (1950). Chromatographic behaviour and chemical structure. 11. The tea catechins. Biochim. Biophys. Acta 4, 441-444. Bradfield, A. E., and Penny, M. 1944. The chemical composition of tea. The proximate composition of an infusion of black tea and its relation to quality. J . Soc. Chem. Ind., London 63, 306-310. Brown, A. G.,Falshaw, C. P.,Haslam, E., Holmes. A . , and Ollis, W. D. 1966. The constitution of theaflavin. Tetrahedron Lett. pp. 1193-1204. Brown, A. G., Eyton, W. B., Holmes, A., and Ollis, W. D. 1969a. Identification of the thearubigins as polymeric proanthocyanidins. Nature (London) 221, 742-744. Brown, A. G., Eyton, W. B., Holmes, A., and Ollis, W. D. 1969b. The identification of thearubigins as polymeric proanthocyanidins. Phytochemisfry 8, 2333-2340. Bryce, T . , Collier, P. D., Fowlis. I., Thomas, P. E., Frost, D., and Wilkins, C. K. 1970. The structure of the theaflavins of black tea. Tetrahedron Lett. pp. 2789-2792. Bryce, T., Collier, P. D., Mallows, R., and Thomas, P. E. 1972. Three new theaflavins from black tea. Tetrahedron Lett. pp. 463466. Buzun, G. A., Dzhemukhadze, K. M., and Mileshko, L. F. 1970. Properties of tea polyphenol oxidase. Biokhimiya 35, 1002-1006. Carr, M. K. V. 1970. The role of water in the growth of the tea crop. In “Physiology of Tree Crops” (L. C. Luckwill and C. V. Cutting, eds.), pp. 287-305. Academic Press, New York. Can, M. K. V. 1974. Irrigating seedling tea in Southern Tanzania: Effects on total yield, distribution of yield and water use. J . Agric. Sci. 83, 363-378. Cartwright. R. A . , and Roberts, E. A. H. 1954. Theogallin, a polyphenol occurring in tea. J. Sci. Food Agric. 5 , 593-597. Cxtwright, R. A., and Roberts. E. A . H. 1955. Theogallin as a galloyl ester of quinic acid. Chem. Ind. (London) pp. 230-23 1.
TEA
275
Chakrabany, S. R., and Chakrabarty. M. M. 1954. The composition of Indian tea seed oils. Sci. Cult. 20, 186-187. Chalamberidze, T. K., Soboleva, G. A . , Pristupa, N . A,, Petrova, R. K.. and Bokuchava. M. A. 1969. Localization of catechins in tea leaf. Soohshch. Akad. Nuuk Gruz. SSR 54, 697-700. Chenery, E. M. 1955. A preliminary study of aluminium in the tea bush. Plant Soil 6 , 174-200. Cheng, L. T., and Chou, T. P. 1940. The fluorine content of foodstuffs in Szechman. Chung-Kuo Sheng Li Hsueh Tsa Chih 15, 263-268. Co, H.. and Sanderson, G. W. 1970. Biochemistry of tea fermentation: Conversion of amino acids to black tea aroma constituents. J. Food Sci. 35, 160-164. Coggon. P., Moss. G. A . , and Sanderson, G. W. 1973. Tea catechol oxidase. Isolation, purification and kinetic characterization. Phytochemisrry 12, 1947-1955. Coggon, P., Romanczyk. L. J., and Snaderson, G. W. 1977. Extraction, purification and partial characterization of a tea metalloprotein and its role in the formation of black tea aronia constituents. J. Agr. Food Chem. 25, 278-283. Collier, P. D., and Mallows, R. 1971a. Estimation of flavanols in tea by gas chromatography of their trimethylsilyl derivatives. J. Chromatogr. 57, 2 9 4 5 . Collier, P. D., and Mallows, R. 1971b. Estimation of theaflavins in tea by gas-liquid chromatography of their trimethylsilyl ethers. J. Chromafogr. 57, 19-27. Coxon, D. T . , Holmes, A., Ollis, W. D., and Vora, V. C. 1970a. The constitution and configuration of the theaflavin pigments of black tea. Tefrahedron Lett. pp. 5237-5240. Coxon, D. T., Holmes, A., and Ollis, W. D. 1970b. Isotheaflavin. A new black tea pigment. Tetrahedron Leri. pp. 5241-5246. Coxon, D. T.. Holmes, A.. and Ollis, W. D. 1 9 7 0 ~Theaflavic . andepitheaflavic acids. Tetrahedrnn Left. pp. 5247-5250. Croyle, R. D., Wilson, L. L.. and Long, T. A. 1974. Potential of spent tea leaves for animal feeds and composting. Compost Sci. 15, 28-30. Das, D. N., Ghosh, J. J., and Guha. B. C. 1964. Studies on tea. I. Nutritional aspects. Indian J. Appl. Chem. 27, 199-214. Das, D. N., Ghosh, J. J.. Bhattacharya, K . C., and Guha, B. C. 1965. Tea. 11. Pharmacological aspects. Indian J. Appl. Chem. 28, 15-40, de Silva, U. L. L., and Roberts, G. R. 1972. Products from tea seeds. 2. Extraction and properties of saponins. Tea Q . 43, 91-94. Devanathan, M. A. V. 1975. Weather and the yield of crop. Exp. Agric. 11, 183-186. Devanathan, M. A. V. 1976. The quantification of climatic constraints on plant growth. Tea Q . 45, 49-78. Dzemukhadze, K . M., Shalneva, G. A., and Mileshko. L. F. 1957. Changes in catechins during tea fermentation. Biokhimiya 22, 836-840. Dzemukhadze, K. M., Buzun. G. A., and Mileshko, L. F. 1964. Enzyme oxidation of catechols. Biokhimiyn 29, 882-888. Eden, T . 1976. “Tea,” 236 pp. Longmans, London. Ellis. R. T. 1976. The tea industry in Malawi and the Tea Research Foundation of Central Africa. Symp. Teh I , Bandung, Indonesia. Evans. D. I. 1930. “Report ofthe Biochemist,” Bull. No. 4. pp. 20-28. Rep. Tea Res. Inst. Ceylon for 1929. Fedrick, J. 1974. Anencephalus and maternal tea drinking. Evidence for a possible association. /‘roc,. R . Soc. Med. 61, 356-360. Fernando, E. F. W . 1967. Studies on the aetiology and biochemistry of Oligonvchus cnfear Nietner (Acarina: Tetranychidae), the red spider mite of tea in Ceylon. M.S. Thesis. Univ. of Sri Lanka Library, Peradeniya.
276
ROBERT L. WICKREMASINGHE
Finot, P. A,. Muggler-Chavan, F., and Vuataz, L. 1967. La phtnylalanine prtcurseur de la phCnylacetaldehyde dans I’ar6me de thC noir. Chimia 21, 26-27. Food and Agricultural Organization of the United Nations. 1976. “The Longer Term Outlook for Tea,” Rep. CCPTe 76/4. FAD, Rome. Forrest, G. I., and Bendall, D. S . 1969. The distribution of polyphenols in the tea plant. Biochem. J . 113, 741-755. Fujita, Y . , Fujita, S . , and Yoshikawa, H. 1973. Essential oils of Camellia sesanqua Thunb., C. japonica Linn. and Thea sinensis L. (Comparative biochemical and chemo-taxonomical studies of the plants of the Theaceae. Part 1.) Nippon Nogei Kagaku Kaishi 47, 645-650. Furuya, K. 1970. Inert gas packaging of tea. Jpn. Agric. Res. Q . 5 , 4 5 4 9 . Gianturco, M. A , , Biggers, R. E., and Ridley, B. H. 1974. Seasonal variations in the composition of the volatile constituents of black tea. A numerical approach to the correlation between composition and the quality of tea aroma. J . Agric. Food Chem. 22, 758-764. Gonzales, J. G . , Coggon, P., and Sanderson, G. W. 1972. Biochemistry of tea fermentation: formation of trans-2-hexenal from linolenic acid. J . Food Sci. 37, 797-798. Gregory, R. P. F., and Bendall, D. S . 1966. The purification and some properties of the polyphenol oxidase from tea (Camellia sinensis, L.). Biochem. J . 101, 569-581. Hara, T . , and Kubota, E. 1973a. Changes in aroma components during roasting of green Tea. Nippon Shokuhin Kogyo Gakkai-Shi 20, 283-286. Hara, T., and Kubota, E. 1973b. Volatile carbonyl compounds of heated green tea (Hiire-Cha). Nippon Shokuhin Kogyo Gakkai-Shi 20, 31 1-315. Harler, C. R. 1963. “Tea Manufacture,” 126 pp. Oxford Univ. Press, London and New York. Hashizume, A. 1967. Saponin from tea leaf. Isolation and properties. Nippon Nogei Kagaku Kaishi 40, 8-12. Hashizume, A. 1969.Saponin from the leaf of Thea sinensis. 11. Component sapogenins and organic acids from the leaf of Thea sinensis. Nippon Nogei Kagaku Kaishi 43, 750-757. Hashizume, A. 1970. Tea saponins. Chagyo Kenkyu Hokuku Shinyu 2, 52-61. Hasselo, H. N. 1965. The nitrogen, potassium, phosphorus, calcium magnesium, sodium, manganese, iron, copper, boron, zinc, molybdenum and aluminium contents of tea leaves of increasing age. Tea Q 36, 122-136. Hatanaka, A. 1976. Biosynthesis of trans-2-hexenal in chloroplasts of Thea sinensis. Phytochemistry 15, 1125-1 126. Hatanaka, A., and Harada, T. 1973. Formation of cis-3-hexenal rrans-2-hexena1, and cis-3-hexenol in macerated Thea sinensis leaves. Phytochemistry 12, 2341 -2346. Hatanaka, A,, Sekiya, J . , and Kajiwara, T. 1976a. Subunit composition of alcohol dehydrogenase from Thea sinensis seeds and its substrate specificity for monoterpenes. Phytochemistry 15, 487-488. Katanaka, A,, Kajiwara. T., and Sekiya, J. 1976b. Biosynthesis of trans-2-hexenal in chloroplasts from Thea sinensis leaves. Phytochemistry 15, 1125-1 126. Hearne, J. F., and Lee, H. N. 1955. The evolution of carbon dioxide from tea. Chem. Ind. (London) p. 1633. Hillis, W. E., and Ishikura. N. 1970. The biosynthesis of polyphenols in tissues with low phenylalanine lyase activity. Phytorhmistry 9, 15. 17-1528. Hilton, P. J. 1974. The effect of shade upon the chemical constitution of the flush of tea (Camellia sinensis L.). Trop. Sci. 16, 15-22. Hilton, P. J., and Ellis, R. T. 1972. Estimation of the market value of Central African tea by theaflavin analysis. J . Sci. Food Agric. 25, 227-232. Hoefler, A. C., and Coggon, P. 1976. Reverse phase high-pressure liquid chromatography of tea constituents. J . Chromatogr. 129, 460-463.
TEA
277
Imagawa, H.. Takino, Y., and Shimizu, M. 1976. Studies on the nucleotides of tea. Part 111. Degradation of RNA by tea leaf nucleases. Nippon Shokuhin Kogyo Gakkai-Shi 23, 138-144. h a . K.. Sakato, Y., and Fukami, H. 1968. Isolation and structure elucidation of theaspirone. a component of tea essential oil. Tetrahedron Lett. pp. 2777-2780. Itoh, T.. Tamura, T., and Matsumoto, T. 1974. Sterols, methylsterols, and triterpene alcohols in three Theaceae and some other vegetable oils. Lipids 9, 173-184. Iwasa, K. 1974. Changes in activity of phenylalanine ammonia-lyase in tea leaves. Nippon Nogei Kagaku Kaishi 48, 445450. Iwasa, K. 1976. Physiological aspects of catechin biosynthesis in tea plants. Jpn. Agric. Res. Q . 10, 89-93. Kaiser, H. E. 1967. Cancer-promoting effects of phenols in tea. Cancer 20, 614-616. Karunanayake, E. H., Mahadeva, K . , Weerakoon, S . N.. and Wickremasinghe, R. L. 1972. Fluoride in black tea. Tea Q. 43, I 1 1-1 13. Kato, C.. Uritani, I . , Saijo, R., and Takeo, T. 1976. Cellular localization of particulate-bound polyphenol oxidase in tea leaves. Plant Cell Physiol. 17, 104-1052. Kawashima, K., and Yamanishi, T. 1973. Thermal degradation of Betaxarotene (Note). Nippon Nogei Kagaku Kaishi 47, 79-8 I . Keegel, E. L. 1958. “Tea Manufacture in Ceylon,” Mongr. No. 4. Tea Res. Inst., Talawakele, Sri Lanka. Kingdon-Ward, F. 1950. Does wild tea exist? Nature (London) 165, 297-299. Kiribuchi. T., and Yamanishi. T. 1963. Studies on the flavor of green tea. Part IV. Dimethyl sulphide and its precursor. Agric. Biol. Chem. 27, 56-59. Kirtisinghe, D. 1974. The TRI-CCC fluid bed drier developed in Sri Lanka. Tea Q. 44, 151-153. Kito, M., Kokura. H.,Izaki, J.. and Sasoka, K. 1968. Theanine, a precursor of the phloroglucinol nucleus of catechins in tea plants. Phytochemistry 7 , 599-603. Konishi, S. 1969. Studies on the metabolism of theanine in tea plant using radioactive carbon. Proc. Conf. Radioisor., 9th, Jpn. At. Ind. Forum, pp. 423425. Konishi, S . , and Takahashi, E. 1966. Existence and synthesis of L-glutamic acid-y-methylamide in tea plants. Plant Cell Physiol. 7 , 171-175. and its metabolic distribution in the Konishi, S . , and Takahashi, E. 1969. Metabolism of N-e~hyt-’~C tea plant, VI. Metabolism and regulation of theanine and related compounds in the tea plant. Nippon Dojo-Hiryogaku Zasshi 40, 479-484. Konishi, S., Ozasa, M., and Takahashi, E. 1972. Metabolic conversion of N-methyl carbon of y-glutamylmethylamide to caffeine in tea plants. Plant Cell Physiol. 13, 365-375. Kozhin, S. A,. and Treiger, N. D. 1973. Gas liquid chromatography and computer aided evaluation of the influence of components on the essential oil of Georgian black tea on its aroma. Prikl. Biokhim. Mikrobiol. 9, 895-900. Krantz, J . C. 1955. In “Tea: A Symposium on the Pharmacological and the Physiological and Psychological Effects of Tea” (H. J . Klauberg, ed.). p. 4. Biol. Sci. Found., Washington, D.C. Krishnamurthy. K., Venkatasubramanism, T. A,, and Giri, K. V. 1952. Circular paper chromatographic analysis of the amino acids of tea and coffee infusions. Curr. Sci. 21, 133. Kursanov. A. L. 1956. Tannins of the tea plant. Kulrurpflanze Beih. 1, 2 9 4 8 . Lamb. J . . and Ramaswarny, M. S. 1958. Fermentation of Ceylon tea. XI. Relations between polyphenol oxidase activity and pectin methyleaterase activity. J . Sci. Food Agric. 9, 51-56. Lea, A. G. H., and Crispin, D. J . 1971. The separation of theaflavins on Sephadex LH 20. J . Chromafogr. 54, 133-135. Lea, C. H.. and Svobodd, P. A. T. 1957. The anti-oxidant action of some polyphenolic constituents of tea. Chem. Ind. (London) pp. 1073-1074. Li, L., and Bonner. J. 1947. Experiments on the localization and nature of tea oxidasc. Biuc.hern. J . 41, 105-110.
278
ROBERT L. WICKREMASINGHE
Little, J . A,, Shanoff, H. M., Csima, A., and Yone, R. 1966. Coffee and serum lipids in coronary heart disease. Lancer 1, 732. Machida, S . 1938. Chemical studies on green tea. 11. Saponin of green tea. Nippon Nogei Kagaku Kuishi 14, 301-308. Mahendra, C. C., Sentheshanmuganathan, S . , Perera, W. D. A,, and Wickremasinghe, R. L. 1972. The effect of tea and coffee on the serum lipid levels of rats. Proc. Ceylon Assoc. Adv. Sci. 1, 17-18. Matsumoto, T., Wainai, T., and Miyaka, Y . 1955. On sterol of tea leaf. Nippon Kugaku Zasshi 76, 1057-1061. Memedov, M. A. 1961. Tea selection in Azerbaijan (Russia). Agrobiologiya 1, 62-67. (Horfic. Abstr. 1961. p. 7219.) Millin, D. J . , and Rustridge, D. W. 1967. Tea manufacture. Process Biochem. 2, 9-13. Millin, D. J., Crispin, D. J., and Swaine, D. 1969a. Nonvolatile components of black tea and their contribution to the character of the beverage. J . Angric. Food Chem. 17, 717-722. Millin. D. J . , Sinclair, D. S . , and Swaine. D. 1969a. Some effects of ageing on pigments of tea extracts. J . Sci. Food Agric. 20, 303-306. . and classification of the brown pigments Millin, D. J . , Swaine, D., and Dix, P. L. 1 9 6 9 ~Separation of aqueous infusions of black tea. J . Sci. Food Agric. 20, 296-302. Mizuno, T. 1968. Studies on the carbohydrates of tea plants. Part XII. The component sugars of theasaponin from the tea seed. Nippon Nogei Kugaku Kuishi 42, 491-501. Morchiladze, Z. N.. Tkemaladze, G. S., Soseliya. M. F.. and Dzhamapishvili, T. S . 1972. Isolation and purification of malate dehydrogenase from the tea plant. Soobshch. Akad. Nauk Gruz. SSR 65, 181-184. (Chem. Abstr. 76, 96142.) Morton, J . 1972. Further associations of plant tannins and human cancer. Q. J . Crude Drug Res. 12, 1829-1841, Miiggler-Chavan, F., Viani, R., Bricout, H., Marion, J. P.. Mechtler, H., Reymond. D.. and Egli, R. H. 1969. Sur la composition de I’arbme de the. 111. Identification de deux cktones apparanties aux ionones. Helv. Chim. Acra 52, 549-550. Naismith, D. J . , Akinyaju, P. A,, and Yudkin, J . 1969. Influence of caffeine-containing beverages on the growth, food utilization and plasma lipids of the rat. J . Nutr. 97, 375-38 I . Nakabayashi, T. 1958. Studies on the formation mechanism of black tea aroma. Part V. On the precursor of the volatile carbonyl compounds of black tea. Nippon Nogei Kagaku Kaisshi 32, 941 -945. (Chem. Abstr. 55, 15772.) Nakagawa, M. 1967. The nature and origin of polyphenols in Hoji-cha (roasted green tea). Agric. Biol. Chem. 31, 1283-1287. Nakagawa, M. 1969. Correlation of theaflavin and thearubigin contents with tea tasters’ evaluation. Nippon Shokuhir Kogyo Gakkai-Shi 16, 266-271. (Chem. Absrr. 73, 127912.) Nakagawa, M . 1970. Constituents in tea leaf and their contribution to the taste of green tea liquor. Jpn. Agric. Res. Q . 5 , 4 3 4 7 . Nakagawa, M. 1975a. Contribution of green tea constituents to the intensity of taste element of brew. Nippon Shokuhin Kogyo Gukkai-Shi 22, 59-64. Nakagawa, M. 1975b. Chemical components and taste of green tea. Jpn. Agric. Res. Q. 9, 156-160. Nakagawa, M., and Ishima, N. 1971. Correlation of the chemical constituents with the organoleptic evaluation of green tea liquors (continued). Statistical analysis. Chagyo Gijursu Kcnkyu 41, 41-44. (Chem. Absrr. 74, 139695.) Nakagawa, M . . and Torii. H . 1964. Studies on the flavonols of tea. 11. Variation in the flavanolic constituents during the development of tea leaves. Agric. Biol. Chem. 28, 497-504. Neish, A. C. 1960. Biosynthetic pathways of aromatic compounds. Annu. Rev. Plunr Physiof. 11, 55-80.
TEA
279
Nikolaishvili, D. K.. and Adeishvili, N . 1. 1966. Chromatographic study of quantitative changes of pigments of the tea leaf upon its processing. Byull. Vses. Nuuchno-Issled. Inst. Chain. Promsti. 2, 57-60. Nose. M.. Nakatani. Y., and Yamanishi, T. 1971. Studies o n the flavour of green tea. Part IX. lo-ntification and composition of intermediate and high boiling constituents in green tea flavour. Agric. B i d . Chem. 35, 261-271. Ogotuga, D. B. A., and Northcote, D. H. 1970a. Caffeine formation in tea callus tissue. J . Exp. Eor. 21, 258-273. Ogotupa, D. B. A.. and Northcote. D. H. 1970b. Biosynthesis of caffeine in tea callus tissue. Binc,hem. J . 117, 715-720. Ogura. N. 1969. Chlorophyllase of tea leaves. 11. Seasonal change of a soluble chlorophyllase. Shokubutsugaku Zusshi 82, 392-396. Ogura. N . , and Takamiya, A. 1966. Studies on chlorophyllase of tea leaves. Shokubutsugaku Zusshi 79, 588-594. Okada. F., and Furuya. K. 1969. Studies of fluorine content in tea. Study Ttw pp. 32-38. Oparir.. A. 1.. and Shuben, T. A. 1950. On oxidative respiratory systems in the tea leaf. Biokhirn. Chain. Proisvod. Sb. 6, 82-89. (Chem. Abstr. 46, 263 1 . ) Ozawa, M.,Sato. N., Inatomi. H., Suyama, Y.. and Inukai, F. 1969. Free amino acids in plants. X . L-pipecolic acid in tea plant (Theu sintwsis) seeds. Neiji Daigaku Nogakubu Krnkyu Hokoku 24, 3 ; -35. Perera. K. P. W. C. 1972. Aspects ofthe chemistry of tca. M.S. Thesis. Univ. of Sri Ldnka Library. Colombo. Perera. K. P. W. C.. and Wickremasinghe, R. L. 1972. Propcnies of tea polyphenol oxidase. T r u Q . 43, 153-163. Pethiyagoda, U., and Krishnapillai. S. 1970. Studies o n the mineral nutrition of tea. 2. Experimentally induced major nutrient deficiency symptoms. Trtr Q. 41, 107-120. Pethiyagoda, U., and Krishnapillai, S . 1971. Studies oii the mineral nutrition of tea. 3 . Experimentally induced minor nutrient deficiency symptoms. Tea Q . 42, 19-29. Pintauro, N . 1977. “Soluble Tea Production Processes.” Noyes Data Corp.. Park Ridge. New Jersey. Pochet. P.. Fleural, L., Stainier. F.. and Detroz. J . 1974. Une relation cntre I’anatomie fnliare at la productivitk chez le theier (note prelirninaire). Cyfe, CcccarJ,The 18, 97-106. Popov. V . R. 1956. Oxidation of amino acids in the presence of tannins and polyphenol oxidase of tea. Biokhimiya 21, 383-386. Portsmouth, G. B. 1953. Potash deficiency in tea. Teu Q. 24, 79-81, Proisier, E., and Serenkov. G. P. 1963. Biosynthesis of caffeine in tea leaves. Biokhirniva 28, 857-861. Ramaswamy, M. S. 1964. Chemical basis of liquoring charactcristics of Ceylon tea. Part 3. The effect of elevation and climatic conditions on the composition of tea liquors. Teu Q. 35, 164- 167. Ramaswamy. M. S . . and Lamb, J. 1958. Fermentation of Ceylon tea. X . Pectic enzymes in tea leaf. J . Sci. Food Agric. 9, 46-5 I . Reddy. H. A. 1958. “ECAFE Report on Oil Seed Extraction in Ceylon.’’ Renold. W., Naf-Muller. R., Keller, U., Willhalm. B., and Ohlaff. G. 1974. An investigation of the tea aroma. I . New volatile black tea constituents. Hc4t. Chirn. Actu 57, 1301-1308. Reymond. D. 1976. Flavour chemistry of tea, cocoa and coffee. A m . Chem. Soc. Meet., Neiz York. Roberts. E. A. H. 1952. The chemistry of tea fermentation. J . Sci. Food Agric. 3, 193-198. Roberts. E. A. H. 1958. The phenolic substances of manufactured tea. 11. Their origin as enzymic oxidation products in fermentation. J . Sci. Food Agric. 9, 212-216.
280
ROBERT L. WICKREMASINGHE
Roberts, E. A. H. 1959. Tocklai E.X. Stn.. Rep. p. 388. Roberts, E. A. H. 1961. Tocklai Exp. Stn., Rep. p. 1 10. Roberts, E. A . H. 1962. Economic importance of food substances: Tea fermentation. lti “The Chemistry of Flavonoid Substances” (T. A. Geissmann, ed.) pp. 468-512. Pergamon, Oxford. Roberts, E. A. H. 1963. The phenolic substances of manufactured tea. X . The creaming down of tea liquors. J . Sci. Food Agric. 14, 700-705. Roberts, G. R. 1974. Polar lipid composition of the leaves and seeds from the tea plant (Camellia sinensis L.) J . Sci. Food Agric. 25, 473475. Roberts, C. R., and de Silva, U. L. L. 1972. Products from tea seeds. I . Extraction and properties of tea seed oil. Tea Q. 43, 88-90. Roberts, G . R., and Fernando, V. 1975. Glutamyl transferase activity in roots of the tea plant. Proc. Sri Lanku Assoc. Adv. Sci. 1, 58. Roberts, E. A. H.. and Myers, M. 1958. Theogallin, a polyphenol occurring in tea. 11. Identification as a galloyl quinic acid. J . Sci. Food Agric. 9 , 701-705. Roberts, G. R.. and Sanderson. G. W. 1966. Changes undergone by free amino acids during the manufacture of black tea. J . Sci. Food Agric. 17, 182-188. Roberts. E. A. H., and Smith, R. F. 1961. Spectrophotometric measurements of theaflavins and thearubigins in black tea liquors in assessments of quality in teas. Analyst (London) 86, 94-98. Roberts, E. A. H., and Smith, R. F. 1963. The phenolic substances of manufactured tea. IX. The spectrophotometric evaluation of tea liquors. J . Sci. Food Agric. 14, 689-699. Roberts, E. A . H., and Wood, D. J . 1950. The fermentation process in tea manufacture. 11. Oxidation of substrates by tea oxidase. Eiochem. J . 47, 175-186. Roberts. E. A. H., and Wood, D. J . 1951. The amino acids and amides of fresh and withered tea leaf. Curr. Sci. 20, 151-153. Roberts. E. A . H., Cartwright, R. A , . and Oldschool. M. 1957. The phenolic substances of manufactured tea. I . Fractionation and paper chromatography of water soluble substances. J . Sci.. Food Agric. 8, 72-80. Saijo, R. 1970. Degradation of chlorophylls in tea leaves by fermentation and heating. Chagyo Gfjutsu Kenkyu 40,66-68. Saijo. R. 1973. Study on black tea aroma in special reference to the formation mechanism of volatile carbonyl compounds. Jpn. Agric. RPS. Q. 7 , 202-207. Saijo, R., and Kuwabara. Y . 1967. Volatile flavour of black tea. Part I . Formation of volatile components during black tea manufacture. Agric. Eiol. Chem. 31, 389-396. Saijo, R., and Takeo. T. 1970a. The formation of aldehydes from amino acids by tea leaf extracts. Agric,. Biol. Chem. 34, 227-233. Saijo, R., and Takeo. T. 1970b. The production of phenylacetaldehyde from phenylalanine in tea fermentation. Agric. Eiol. Chem. 34, 222-226. Saijo, R.. and Takeo. T. 1972. The importance of linoleic acid and linolenic acid as percursors of hexanal and trans-2-hexenal in black tea. Plunt Cell Physid. 13, 991 -998. Saijo. R.. and Takeo, T. 1974. Induction of peroxidase activity by ethylene and indole-3-acetic acid in tea shoots. Agric. B i d . Chem. 38, 2283-2284. Saijo, R . , and Uritani. I. 1971. Biosynthesis of terpenoids in tea plant. Incorporation of acetate-2-14C into nerolidol in excised tea shoots. Agric. B i d . Chem. 35, 2132-2134. Sakamoto, Y. 1967. Flavones in green tea. I. Isolation and structures of flavones occurring in grcen tea infusion. Agric,. B i d . Chem. 31, 1029-1034. Sakamoto. Y . 1970. Chemical constituents related to the color of green tea infusion. Isolation and structure of C-glycosyl flavones. Chagyo Shikenjo Kenkyu HIkoku 6 , 1-63. (Chem. Ahstr. 73, 97543.) Sakato, Y . 1950. Studies on the chemical constituents of tea. 111. On a new amide-theanine. Nippon Nogei Kaguku Kuishi 23, 262-264.
TEA
28 I
Sakato. Y . 1957. Recent advances in tea chemistry in Japan. Proc.. Symp. Phyrochem., Kualr Lampur pp. 168-172. Sakato, Y . , Hashizume. T.. and Kishinioto, Y. 1950. Studies on the chemical constituents of tea. Part V. Synthesis of theanine. Nippon Nogei Kugaku Kaishi 23, 269-27 I . Sanderson. G . W. 1964a. Extraction of soluble catechol oxidase from tea shoot tips. Biochim. Biophvs. Acra 92, 622424. Sanderson. G. W. 1964b. The chemical composition of fresh tea flush as affected by clone and climate. Tea Q . 35, 101-109. Sandcrson, G. W. 1965. On the nature of enzyme catechol oxidase in tea plants. Tea Q. 36, 103-1 1 I . Sanderson, G. W. 1966. 5-Dehydroshikimate reductase in the tea plant (Camellia sinensis L.). Biochem. J . 98, 248-252. Sanderson. G. W. 1968. Change in cell membrane permeability in tea flush on storage after plucking and its effect on fermentation during manufacture. J . Sci. Food Agric. 19, 637439. Sanderson, G. W. 1972a. The chemistry of tea and tea manufacturing. In "Structural and Functional Aspects of Phytochemistry" (V. C. Runeckles. ed.). pp. 247-316. Academic Press, New York. Sanderson. G. W. 1972b. The practice of instant tea manufacture. World Coffee Tea April, pp. 54-57. Sanderson. G. W. 1975. Black tea aroma and its formation. In "Geruch und Geschmackstoff" (E. Drawert, ed.). pp. 65-97. Verlag Hans Carl, Nurenberg. Sanderson. G. W.. and Gonzales. J. G. 1971. Biochemistry of tea fermentation. The role of carotenes in black tea aroma formation. J . Food Sci. 36, 231-236. Sanderson, G. W.. and Kanapathipillai. p. 1964. Further studies on the effect of climate on the chemical composition of fresh tea flush. Teu 0 . 35, 222-229. Sanderson. G. W., and Perera, B. P. M. 1965. Carbohydrates in tea plants. I . The carohydrates oftea shoot tips. Tea Q . 36, 6-13. Sanderson. G . W.. and Roberts, G . R. 1964. Peptidase activity in shoot tips of the tea plant (Camellia sinensis L.). Biochem. J . 93, 419423. Sanderson. G. W.. and Selvendran. R. R. 1965. The organic acids in tea plants. A study of the non-volatile organic acids separated on silica gel. J . Sc.i. Food Agric. 16, 251-258. Sanderson, G. W.. Co, H.. and Gonzales. J . G. 1971. Biochemistry of tea fermentation: The role of carotenes in black tea aroma formation. J . Food Sci. 36, 23 1-236. Sanderson. G. W . , Berkowitz, J. E., and Co, H . 1972. Biochemistry of tea fermentation: Products of the oxidation of tea flavanols in a model tea fermentation system. J . Food Sci. 37, 399404. Sanderson. G. W., Ranadive. A. S . . Eisenberg. L. S . . Farrell. F. J . , Sirnons, R., Manley, C. H., and Coggon. P. 1976. Contribution of polyphenolic compounds to the taste of tea. In "Sulfur and Nitrogen Compounds in Food Flavors" (G. Charalambous and 1. Katz. eds.), ACS Symp. Ser. No. 26, pp. 14-16. Am. Chem. Soc.. Washington, D.C. Sasaoka. K., and Kito, M. 1964. Synthesis of theanine by tea seedling homgenate. Agric. B i d . Chem. 28, 313-317. Schwerp, H. W. 1971. Dental caries: Prospects for prevention. Science 173, 1199-1205. Sekiya, J.. Numa, S . , Kajiwara, T . , and Hatanaka, A. 1976. Biosynthesis of leaf alcohol. Formation of 3Z-hexenal from linolenic acid in chloroplasts of Thea sinensis leaves. Agric. B i d . Chem. 40, 185-190. Selvendran, R. R. 1969. Metabolism of nucleotides and phosphate esters in tea shoots during black tea manufacture. Tea Q. 40, 93-98. Selvendran. R. R., and King. N. R. 1976. Ultrastructural changes in tea leaves during 'orthodox' black tea manufacture. Ann. Appl. Biol. 83, 463473. Selvendran. R. R., and Perera, B. P. M. 1971. Chemical composition of tea leaf cell wall. Chem. Ind. (London) pp. 577-578.
282
ROBERT L. WICKREMASINGHE
Selvendran. R. R.. and Selvendran, S. 1972. Changes in the polysaccharides of the tea plant during post-prune growth. Phyrorhemistry 11, 3167-3171. Selvendran, R. R.. Perera, B. P. M., and Selvendran, S . 1972. Changes in the ethanol-insoluble material of tea leaves (Camellia sinensis L.) during maturation. J. Sci. Food Agric. 23, 11 191123. Singer, R., Armstrong, W . D., and Vatassery, G . T. 1967. Fluoride in commercial tea and related plants. Econ. Bo!. 21, 285-287. Sivapalan, P. 1976. Personal communication. Sivapalan, P., and Shivanandarajah, V. 1974. Polyphenol content in the feeder roots of nematode tolerant and susceptible tea clones in relation to infestation by Prarylenrhus loosi Loof. Tea Q . 44, 173-176. Smith, R. F. 1968. Studies on the formation and composition of "cream" in tea infusions. J. Sci. Food Agric. 19, 530-534. Sreerangachar, H. B. 1943a. Studies on the fermentation of Ceylon tea. 6. The nature of the tea oxidase system. Biochem. J . 37, 661-667. Sreerangachar. H. B. 1943b. Studies on the fermentation of Ceylon tea. 7. The prosthetic group of tea oxidase. Biochem. J. 37, 667-674. Stafford, H. A. 1974. The metabolism of aromatic compounds. Annu. Rev. Plant Physiol. 25, 459-486. Stagg, G . V. 1974. Chemical changes occurring during the storage of black tea. J. Sci. Food Agric. 25, 1015-1034. Stagg, G.V., and Millin, D. J. 1975. The nutritional and therapeutic value of tea. J . Sri. Food Agric. 26, 1439-1459. Stagg, G. V., and Swaine, D. 1971. The identification of theogallin as 3-gallolylquinic acid. Phytochemistry 10, 1671-1673. Stahl, W. H. 1962. The chemistry of tea and tea manufacturing. Adv. Food Res. 11, 201-262. Suzuki, T., and Takahashi, E. 1975a. Metabolism of xanthine and hypoxanthine in the tea plant. Biochem. J . 146, 79-85. Suzuki, T., and Takahashi, E. 1975b. Biosynthesis of caffeine by tea leaf extracts. Enzymic formation of theobromine From 7-methyl xanthine and of caffeine from theobromine. Biochern. J . 146, 87-96. Suzuki, T., and Takahashi, E. 1976. Caffeine biosynthesis in Camellia sinensis. Phyrochemisrry 15, 1235-1 240. Swain, T., and Hillis, W.E. 1959. Phenolic constituents of Prunus domcsrira. Quantitative analysis of phenolic constituents. J . Sci. Food Agric. 10, 63-68. Swain, T., and Williams, C. A. 1970. The role of phenylalanine in flavonoid biosynthesis. Phytorhemistry 9, 21 15-2122. Takeo, T. 1965a. Tea leaf polyphenol oxidase. Part I. Solubilizationand properties of the structurally bound polyphenol oxidase in tea leaves. Agrir. Biol. Chem. 29, 558-563. Takeo, T.1965b. Studies on the flavour of tea made from stored tea leaves at low temperature. Part I. Examination of the essential oil of the stored tea leaves by gas liquid chromatography. Agric. Biol. Chem. 29, 269-274. Takeo, T. 1966a. Tea leaf polyphenol oxidase. Part HI. Studies of the changes of polyphenol oxidase during black tea manufacture. Agric. Biol. Chem. 30, 529-535. Takeo, T. 1966b. Tea leaf polyphenol oxidase. Part IV. The localization of polyphenol oxidase in tea leaf cell. Agric. Biol. Chem. 30, 931-934. Takeo, T. 1974a. L-alanine as a precursor of ethylamine in Camellia sinensis. Phytochemistry 13, I 4 0 1- 1406. Takeo, T. 1974b. Photometric evaluation and statistical analysis of tea infusion. Jpn. Agric. Res. Q . 8, 159-164.
TEA
283
Takeo. T.. and Baker, J. E. 1973. Changes in multiple forms of polyphenol oxidase during maturation of tea leaves. Phyfochemistry 12, 21-24. Takeo, T., and Kato. Y. 1971. Tea leaf peroxidase-isolation and partial purification. Plant Cell Physiol. 12, 2 17-223. Takeo, T.. and Uritani, I. 1966. Tea leaf polyphenol oxidase. 11. Purification and properties of solubilized polyphenol oxidase in tea leaves. Agric. Biol. Chem. 30, 155-163. Takino, Y. 1971. Enzymic solubilization of tea cream. Br. Patent No. 1,249,932. Takino, Y., and Imagawa, H. 1964. Crystalline reddish orange pigment of manufactured tea. Agric. Biol. Chem. 28, 255-256. Takino, Y., and Imagawa. H. 1973. Studies on the nucleotides of tea. Part 11. Nucleotides in black tea. Nippon Shokuhin Kogyo Gakkai-Shi 20, 143-150. Takino. Y..Imagawa, H.. Horikawa, H., and Tanaka. A. 1964. Studies on the mechanism of the oxidation of tea leaf catechins. Part III. Formation of a reddish orange pigment and its spectral relationship to some benzotropolone derivatives. Agric. B i d . Chem. 28, 64-71, Takino. Y., Ferretti, A,, Flanagan. M., Gianturco, M., and Vogel, M. 1965. The structure of theaflavin. a polyphenol of black tea. Terrahedron Lett. pp. 40194025. Takino, Y., Imagawa, H.. and Shishido, K. 1972. Studies on the nucleotides of tea. Part I. Nucleotides in green tea. Nippon Shokuhin Kogyo Gakkai-Shi 19, 213-218. Tambiah, M. S., Nandadasa, H. G., and Amarasuriya, M . J . C. 1966. The localization anddistribution of catechins, oils and other ergastic substances in the leaves of Camellia sinensis. Proc. Ceylon Assoc. Adv. Sci. 1, 22. Tirimattna, A. S. L. 1967. Acid phosphatases of the tea leaf. Tea Q . 38, 331-334. Tirimanna, A. S. L. 1972. Starch gel electrophoresis of the peroxidase isozymes of the tea leaf. J . Chromatogr. 65, 587-588. Tirimanna, A. S. L., and Wickremasinghe, R. L. 1965. Studies on the quality and flavour of tea. 2. The carotenoids. Tea Q. 36, 115-121. Tirimanna, A. S. L., Wickremasinghe,R. L., and Perera, K. P. W. C. 1967. An anti-oxidant in tea. Tea Q. 38, 3 6 4 0 . Tocklai Experimental Station (1974). Ann. Sci. Rep. 1973-1974, pp. 57-59. Tolhurst, J. A. H. 1962. Zinc deficiency of tea in Ceylon. Tea Q . 33, 134-137. Toyao, T. 1966. Studies on Koro tea. I . The inheritance of Koro type characters and the estimation of the degree of selfing of the tea plant. Sfudy Tea 32, 18-22. Toyao, T. 1975. Early selection for black tea quality. Jpn. Agric. Res. 9, 107-1 10. Tsushida. T., and Takeo, T. 1976. Partial purification and properties of tea leaf ribonuclease. Agric. Biol. Chem. 40, 1279-1285. Tsushida, T., Saijo, R., and Takeo, T. 1976. Changes in chemical constituents in tea leaves during controlled atmospheric storage. Nippon Shokuhin Kogyo Gakkai-Shi 23, 1-5. Tubbs, F. R. 1936. Drought and pruning. Tea Q. 9, 25-29. Ul’yanova, M. S. 1963. Flavones in tea leaves. USSR Congr. Biochem., 1st Abstr. No. 31. Ul’yanova. M. S., and Erofeyeva, N. N . 1966. Isolation of flavonols from black tea and evaluation of their biologic activity. Biokhim. P rogr. Tekhnol. Chain. Proizvod. pp. 14-20. (Chem. Abstr. 66, 462.) Venkatakrishna, S . , Premachandra, B. R . , and Cama, H. R. 1977. Distribution of carotenoid pigments in tea leaves. Tea Q. 47, 28-3 I . Venkataramani, K . S., and Padmanabhan, T. S . 1964. Third progress repon on tea selection. 11. A preliminary assessment of the relationship between certain leaf characteristics and cup quality. Rep. United Planters’ Assoc. South India, 1963-1964 pp. 50-63. Visser. T., and Kehl, F. H. 1958. Selection and vegetative propagation of tea. Tea Q. 29, 76-86. Vitzthum, 0. G., Werkhoff, P., and Hubert, P. 1975. New volatile constituents of black tea. J . Agric. Food Chem. 23, 999-1003.
284
ROBERT L. WICKREMASINGHE
Vogel, G., Marak, M. L., and Oertner, R. 1968. Pharmacology of the tea seed saponin, a saponin mixture from Thea sinensis. Arzneim.-Forsch. 18, 1466-1467. Vuataz, L., and Brandenberger, H. 1961. Plant phenols. 111. Separation of fermented and black tea polyphenols by cellulose column chromatography. J. Chromatogr. 5 , 17-3 1. Vuataz, L., and Reymond, D. 1970. Mathematical treatment of gas chromatographic data. Application to tea quality evaluation. Proc. In?. Symp. Chromatogr., Miami, Fla. pp. 69-74. Vuataz, L., Brandenberger, H., and Egli, R. H. 1959. Plant phenols I. Separation of the tea leaf polyphenols by cellulose column chromatography. J. Chromatogr. 2, 173-187. Wickremasinghe, R. L. 1964. Report of the Biochemist (Manufacture). Tea Res. Inst. Ceylon Annu. Rep. 1964 pp. 106-107. Wickremasinghe, R. L. 1967. Fact and speculation in the chemistry and biochemistry of black tea manufacture. Tea Q. 38,205-209. Wickremasinghe, R. L. 1972. By-products of tea. Tea Q.43, 85-87. Wickremasinghe, R. L. 1974. The mechanism of operation of climatic factors in the biogenesis of tea flavour. Phytochemistry 13, 2057-2063. Wickremasinghe, R. L. 1978. “Tea Biochemistry in Practice.” (in press), Cave & Co. Ltd., Sri Lanka. Wickremasinghe, R. L. 1977. Process of making cold water soluble tea concentrates and powders. U.S, Patent No. 4,004,038. Wickremasinghe, R. L., and Perera, V. H. 1966a. The blackness of tea and the colour of tip. Tea Q. 37, 75-79. Wickremasinghe, R. L., and Perera, K. P. W. C. 1966b. Analysis of ‘cream’ of tea. Tea Q. 37, 13 1-133. Wickremasinghe, R. L., and Perera, K. P. W. C. 1972a. Site of biosynthesis and translocation of theanine in the tea plant. Tea Q. 43, 175-177. Wickremasinghe, R. L., and Perera, K. P. W. C. 1972b. Chemical changes during the storage of black tea. Tea Q. 43, 147-152. Wickremasinghe, R. L., and Perera, K. P. W. C. 1973. Factors affecting quality, strength and colour of black tea liquors. J . Natl. Sci. Counc. Sri Lanka 1, 1 1 1-121. Wickremasinghe, R. L., and Sivapalan, K. 1966. The role of Ieucine in tea flavour. Proc. Ceylon Assoc. Adv. Sci. 1, 47. Wickremasinghe, R. L., and Swain, T. 1964. The flavour of black tea. Chem. Ind. (London) pp. 1574-1575. Wickremasinghe, R. L., and Swain, T. 1965. Studies on the quality and flavour of black tea. J. Sci. Food A&. 16,5 7 4 4 . Wickremasinghe, R. L., Perera, K. P. W. C., Perera, V. H., and Kanapathipillai, p. 1966. Analyses of polyphenols, amino acids and chlorophyll levels in tea flush at different seasons. Tea Q. 37, 232-235. Wickremasinghe, R. L., Roberts, G. R., and Perera, B. P. M. 1967. The localization of the polyphenol oxidase of tea leaf. Tea Q. 38, 309-310. Wickremasinghe, R. L., Perera, B. P. M., and de Silva, U. L. L. 1969. Studies on the quality and flavour of tea. 4. Observations on the biosynthesis of volatile compounds. Tea Q. 40,26-30. Wickremasinghe. R. L., Wick, E. L., and Yamanishi, T. 1973. Gas chromatographic-mass spectrometric analysis of “flavory” and “non-flavoly” Ceylong black tea roma concentrates prepared by two different methods. J. Chromatogr. 79, 75-80. Wickremasinghe, R. L., Perera, B. P. M., and Perera, K. P. W. C. 1976. a-Spinasterol, temperature, and moisture as determining factors in the infestation of Camellia sinensis by Xyleborus jorniratus. Biochem. Syst. Ecol. 4, 103-1 10.
TEA
285
Wight, N., and Barua, P. K. 1954. Morphological basis of quality in tea. Narure (London) 173, 63043I . Wood, D . J.. and Chanda, N. B. 1955. Tocklui Exp. Stn. Annu. Rep. 1954, Rep. Biochem. Branch pp. 4 5 4 4 . Wood, D. J., and Roberts, E. A. H. 1964. Chemical basis of quality in tea. 111. Correlation of analytical results with tea tasters' reports and valuations. J . Sci. Food Agric. 15, 19-25. Wood, D. J . , Bhatia, I. S., Chakraborty, S., Choudhury, M. N . D., Deb, S. B.. Roberts, E. A. H . , and Ullah, M. R. 1964. The chemical basis of quality in tea. I. Analyses of freshly plucked shoots. J . Sci. Food Agric. 15, 8-14. Wu, C. T. 1968. The anatomy of tea leaves and its relation with tea leaf quality. J . Agric. Assoc. China 64, 2 9 4 . (Plant Breed. Abstr. 40, 3556.) Wu, C. T. 1976. Tea in Taiwan. Symp. Teh I , Bandung, Indonesia. Wu, C. T.. Wu, H. K . , and Fong, C. H. 1958. The distribution of plucking leaf hair of tea varieties and its correlation with yield and quality. J . Agric. Assoc. China 24, 78-82. Yamanishi. T. 1975. Tea aroma. Nippon Nogei Kugaku Kuishi 49, 1-9. Yamanishi, T., Takagaki, J., and Tsujimura, M. 1956. Studies on the flavor of green tea. Part 11. Changes in components of essential oil of tea leaves. Bull. Agric. Chem. Soc. Jpn. M ,127-130. Yamanishi, T., Takagaki, J . , Kurita, H., and Tsujimura. M. 1957. Studies on the flavor of green tea. Part nI. Fatty acids in essential oils of fresh tea leaves and green tea. Bull. Agric. Chem. Soc. Jpn. 21, 55-57. Yamanishi. T.. Kiribuchi, T., Sakai, M.. Fujita, N.. Ikeda. Y . . and Sasa, K. 1963. Studies on the flavor of green tea. Part V. Examination of the essential oil of the tea leaves by gas liquid chromatography. Agric. Biol. Chem. 27, 193-198. Yamanishi, T . , Kiribuchi. T., Mikumo, Y., Sato, H.. Ohmura, A., Mine, A., and Kurata, T. 1965. Studies on the flavor of green tea. Part VI. Neutral fraction of essential oil of tea leaves. Agric. Biol. Chem. 29, 300-306. Yamanishi. T.. Kobayashi, A , , Sato. H..Nakamura, H . . Osawa. K..Uchida, A., Mon, S., and Saijo, R. 1966a. Flavor of black tea. Part IV. Changes in flavour constituents during the manufacture of black tea. Agric. Biol. Chem. 30, 784-792. Yamanishi. T., Kobayashi, A,, Uchida, A., and Kawashima, Y. 1966b. Studies on the flavor of green tea. Part VU. Flavor components of manufactured green tea. Agric. Biol. Chem. 30, 1102-1105. Yamanishi. T., Wickremasinghe. R. L., and Perera, K . P. W. C . 1968a. Studies on the quality and flavour of tea. 3. Gas chromatographic analyses of the aroma complex. Tea Q. 39, 75-80. Yamanishi. T . , Kobayashi, A , , Nakamura, H.. Uchida. A., Mori, S., Ohsawa, K., and Sasakura, S . 1968b. Flavor of black tea. Part V. Comparison of aroma of various types of black tea. Agric. Biol. Chem. 32, 379-386. Yamanishi. T.. Nose, M.. and Nakatani, Y. 1970. Studies on the flavor of green tea. Part VIII. Further investigations of flavor constituents in manufactured green tea. Agric. Biol. Chem. 34, 599-608. Yamanishi, T.. Shimojo, S., Ukita. M., Kawashima, K . . and Nakatani, Y. 1973. Aroma of roasted tea (Hoji-cha). Agric. Biol. Chem. 37, 2147-2153. Yosioka. I., Matsuda, A , . Nishimura. T.. and Kitagawa, I . 1966. Structure of theasapogenol E. Chrrn. lnd. (London) p. 2202. Young. W.. Hotovec. R. L., and Romero, A. G . 1967. Tca and atherosclerosis. Mature (London) 216, 1015. Zaprometov, M. N. 1961. lsolation of quinic and shikimic acids from the shoots of the tea plant. Biokhimiyu 26, 373-384. Zaprometov. M. N . 1962. The mechanism of biosynthesis of catechins Biokhirniya 27, 366-367.
286
ROBERT L. WICKREMASINGHE
Zaprometov, M . N. 1963. Catechin biosynthesis in tea shoots. Fiziol. Rust. 10, 73-78. Zaprometov, M. N . , and Bukhlaeva, V. Y. 1971. Efficiency of various carbon-I4 precursors for the biosynthesis of flavonoids in the tea plant. Biokhirniya 36, 270-276. Zarnadze, D. N . 1971. Biochemical study of the preservation of a tea leaf at a low temperature. Subtrop. Kul’t. 50, 148-153. (Chem. Abstr. 76, 44839.) Zimmerman, P. W., Hitchcock, A. E., and Gwirtsman, A. 1957. Fluoride in food with special reference to tea. Contrib. Boyce. Thompson Inst. 19, 44-53.
.
ADVANCES IN F M D RESEARCH VOL .
24
JONATHAN W . WHITE. JR . Agricultural Reseurt.h Service.
U S. Department of Agriculture Philudelphiu. Pennsdvaniti
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Production and Processing . . . . . . . . . . . . . . . . . . . A . Principal Areas and Types . . . . . . . . . . . . . . . . B . Production Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . Harvesting . . . . . . . ............................... D . Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Market Forms of Honey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Retail Products . . . . . . . . . . . . . . . . . . . . . . . . . B . Product for Manufacturing Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV . Analysis and Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Moisture Content . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Carbohydrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................. C . Acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .................. D . Mineral Content . . . . . . . . . . . . . . . . . . . . . . . . E . Proteins and Amino Acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......... F . Enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G . Flavor and Color . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........................... H . Vitamins . . . . . . . . . . . . . . . . . . . . . .................. I . Toxic Constituents . . . . . . . . . . . . . V . Physical Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Rheology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Thermal Propetties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . Hygroscopicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D . Crystallization . . . . . . . . . . . . . . . . . ........................... Vl . Storage of Honey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . Effects of Time and Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B . Fermentation . . . . . . . . . . . . . . . . . . . . . . C . Recommended Storage for Honey . . . . . . . . . . . . . . VII . Nutritive Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A . As a Carbohydrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . €3 . Minerals and Vitamins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C . Folklore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
288 289 289 292 293 293 295 295 296 297
304 305 305 312 331 332 333 333 333 335 338 339 344 344 351 352 352 352 354 354 287
.
Copyright 0 1978 by Academic Press Inc . All rights of reproduction in any form reserved .
ISBN 0-12-016124-8
288
JONATHAN W . WHITE. J R
VIII. Uses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Food . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Nonfood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IX. Standards, Specifications, and Quality Control A. United States Standards . . . . . . . . . . . . . . . . B. Codex Alimentarius . . . . . . . . . . . . . . . . . . . C. Specifications . . . . . . . . . . . . . . . . . . . . . . . . D. Quality control . . . . . . . . . . . . . . . . . . . . . . . X. Research Needs.. . . . . . . . . . . . . . . . . . . . . . . . . References ..............................
1.
357
INTRODUCTION
Honey is the only sweetening material that can be stored and used exactly as produced in nature. No refining or processing is necessary before enjoying this unique material, which can be traced through the entire span of recorded history. Honey, which was man’s first sweet, was used earliest as a ceremonial material and a medicinal ingredient. Not until the era of the Greeks and Romans did honey come to be regarded as a food also. It so remained until relatively recently displaced by cane and beet sugar during the past 100 years. Honey is the sweet, viscous substance elaborated by the honeybee from the nectar of plants. This simple definition excludes honeydew honey, which is produced by the bee from honeydew excreted by various plant-sucking insects. The bee harvests, transports, and processes the nectar to honey, and packages and stores it in the comb. Processing consists of simultaneously reducing the moisture content from the 30-60% common to nectars to the self-preserving range of 15-19%, inverting the considerable proportion of sucrose by the addition of invertase, preserving it meanwhile by adding a glucose oxidase which produces small amounts of acidity and hydrogen peroxide. Ripening takes place in open cells of the comb, which are sealed when the honey reaches full density. The combs, of course, are constructed by the bees from wax they secrete, the production of which requires about 8-10 times its weight in honey. As a unique natural product, honey produces an interesting link to earlier times, and a wealth of observations such as: a bee colony flies about 75,000 miles (12 1,000 km) to produce a pound (454 gm) of honey, but the “fuel” consumption in this (at 1 million miles per gallon or 426,000 km per liter) is only about three ounces (85 gm). The purpose of this review is to provide scientists and technologists with information on the composition, properties, processing, and uses of honey necessary to making informed decisions about its use and value in their operations.
HONEY
289
II. PRODUCTION AND PROCESSING A.
PRINCIPAL AREAS AND TYPES
Honey, even when processed for comniercial use, is essentially a natural product. As produced, it is highly variable, particularly in color, flavor, moisture content, and sugar composition, indeed in nearly every constituent. These attributes depend upon climate, the floral type, and individual beekeeping practices. While bees are kept in all 50 states of the United States and in every country of the world, conditions favorable to commercial beekeeping (honey production) are not as widely available. Further, as agricultural practices and crops change, the value of areas for beekeeping or the quality, type, and amount of honey produced will be influenced. Table I shows honey production, imports, and exports for the major honey-producing and consuming countries, providing a summary of world trade in the commodity.* I.
United States
About one-third of the United States honey crop is sold by the producer directly to the consumer, the remainder to packers. Nearly half of the crop is produced by about 1200 fulltime beekeepers (400 or more hives), about twofifths by parttime beekeepers (25-400 colonies), and the remainder by hobbyists (.hington.New York. and other fruit-producing areas.
C.
HARVESTING
Management of honeybee colonies for maximum honey production is a blend of art and science and is beyond the scope of this review. Details are described by Cale et al. (1975). The hive bodies (”supers”) containing combs of ripened honey, largely capped over, are removed from the colony. freed of bees, and taken to a central location for extraction. Therc the cappings of the cells are removed mechanically and the honey is extracted by centrifugation. I t may be run directly into 55-gallon drums for shipment to processors or storage or may be cleaned (to a greater or lesser extent) by allowing it to stand to permit extraneous material to rise to the surface (ripe honey has a density of around 1.42) for removal. It may also be strained through coarse (23 mesh) or fine (100 mesh) screens depending upon the needs of the immediate customer. The frequency of removal of supers will depend upon the honey tlow and the need to prevent mixture of different floral types.
D.
PROCESSING
I.
Why Process Y
Honey immediately after extraction is at its best in terms of flavor and color. It is not suitable for large-scale marketing without further treatment, however, unless the producer has carried out the required processing (which qualifies him as a “producer-packer”). Most producers sell most of their honey to processors who prepare it for marketing and package it. As extracted, “raw” honey contains extraneous matter such as pollen, bits of wax, variable amounts of sugar-tolerant yeasts, and probably crystals of dextrose
294
JONATHAN W . WHITE. JR
hydrate.* It is thus prone to fermentation unless the moisture content is below 17%; most honey will crystallize in time unless action is taken to prevent it. Processing of honey thus includes controlled heating to destroy yeasts and dissolve dextrose crystals, combined with fine straining or pressure filtration. 2.
Processing Methods
a. For a Liquid Product. Even though supersaturated with respect to glucose, honey will not granulate for months if correctly processed, handled, and stored. Pressure filtration, introduced by Lothrop and Paine (1 934) for honey, improved shelf life of liquid honey by eliminating seed crystals and fine particles of crystallization-inducing substances. Heat exposure, because of the great sensitivity of honey to heat resulting from its acidity, fructose content, and high viscosity, should be limited only to that necessary to accomplish the functions: ‘‘melting” (dissolution of dextrose granulation), pumping, filtration, pasteurization, and filling. Figure 1 diagrams a plant packing 12 million pounds (5.44 million kg) of honey per year. Honey is received from producers in 55-gallon (208 liter) drums (660 Ib, 300 kg), classified for color, floral type, flavor, and moisture and held for use. The melter is designed to liquify 24 drums in about 4 hours without exposure to excessive heating. Most of the liquefaction occurs in the tank beneath the oven from which honey is pumped to batch storage. From this point, it is raised to 150°F (65.6”C) by a heat exchanger, passed through a plate-type filter, and cooled to 120°F (49°C) in the heat exchanger before going to a series of holding tanks in the packing area. Total time at 150°F (65.6”C)is about 30 seconds in this operation. The filling lines for liquid pack honey are conventional; care is needed to avoid reseeding the liquid from holdout residues of honey in lines and equipment. b. For a Solid Product. A semisolid honey product results from the controlled crystallization of some of the dextrose in very fine grain, producing a fondant-like texture. The line for this product is shown on the bottom of the diagram. Not shown is the addition of about 10% of finely crystallized “seed” honey previously prepared by grinding crystallized honey and storing it at 57°F (14°C) for 5-7 days. This is added in the creamer after the batch temperature is reduced to 80°F (27°C). After thorough mixing, the material is filled into retail containers and held at 57°F (14°C) for a week to complete the fine-textured crystallization. The process as described was patented by Dyce (1935); most production is based on it. Details of equipment may be found in the article by Geddes (1964). Townsend (1975) has described several sizes of honey packing lines. *The terms “glucose” and “dextrose” are equivalent, as are “fructose” and “levulose.”
295
HONEY
AT EXCHANGE
COIL
GRILL DRUMS REST ON
HEATER FOR53-GALOPEN END DRUMS CAPACITY 24 DRUMS
I T I FEEDFROM 5 ORPlGE DOUBLE HEAT EXCHbNGER BbTCH STORAGE TANKS 111)
1II
SIOUX HONEY’S NEW PROCESSING-FACKAGING OPERATION
I I
I
I
FILLER SEAYER
I
I
PLbSTlC CAPPER 12OO.LB BATCH STORAGE
I
TWIN PLSTCU FILLER
ICE-TANK WATER CIRCULATES THROUGH COIL
CREAMED HONEV LINE-DIXIE CUP FILLING
FIG. I . Layout and flow diagram for commercial honey packing plant with an annual capacity of 12 million pounds. (From Geddes. 1964.)
Ill. MARKET FORMS OF HONEY A. RETAIL PRODUCTS 1 . Liquid
The United States retail market appears to favor liquid honey, while in many other countries a solid form is preferred. Supermarket exposure seems to require the clear, nonturbid product which results from filtration. A considerable amount of honey is sold in alternate markets such as health food stores, roadside stands, or department or specialty stores. This may not be filtered or processed for clarity and may also be partially granulated, a natural state for honey. Unfortunately, if the honey has been pasteurized, granulation may be coarse and gritty, reducing
296
JONATHAN W . WHITE. JR
the appeal of the product. If the honey has not been heated, however, fermentation may take place unless moisture content is 17% or less. Containers include clear glass, translucent or opaque plastic, and, in larger sizes, metal. Dispensers such as squeeze bottles or drip-cut servers are sometimes available. 2 . Comb Honey Forms Honey in the comb is the ultimate in natural flavor and unprocessed nature. It has virtually disappeared from most urban markets, being difficult and expensive to produce and ship. A sealed comb is a real guarantee of a natural product, exactly as prepared by the honeybee. Market forms include: section comb, a 43 inch (1 1.4 cm) square frame which the bees have filled with honey; cutcomb, which is a piece cut by the beekeeper from a larger comb; and chunkcomb or bulk-comb honey, which is a piece of sealed honey comb in a container filled with liquid honey. 3 . Solidified Honey
Since most honeys are supersaturated in dextrose, the most stable form would seem to be biphasic. The truth appears to be that for nearly all honey there is actually no completely stable form, although for most marketing requirements the liquid form is sufficiently stable. As briefly noted above, honey is also sufficiently shelf-stable for sale in a semisolid form known as “creamed,” “spun,” “churned,” recrystallized, or “honey spread.” This is a fondant of fine dextrose hydrate crystals in the honey matrix. It has a “short” consistency and can be spread or handled without the difficulties of a thick syrup. Nothing extraneous is added in manufacture; the product is a result of a controlled crystallization process which follows the normal pasteurization. This is necessary because the liquid portion of the product is somewhat higher in moisture content than before crystallization and hence more liable to fermentation. Storage at temperatures over about 81°F (27°C) will lead to softening and eventually partial liquefaction, since the equilibrium between soIid and solution is temperature dependent.
B.
PRODUCT FOR MANUFACTURING USE
Honey for use in food, confectionery, and pharmaceutical products is currently available in the liquid form. Generally darker honey types or blends of more pronounced flavor are required to ensure that an identifiable flavor contribution is made to the product. Such types are also somewhat less costly. A mild, light clover type will provide honey attributes other than flavor in a product (hygro-
297
HONEY
scopicity, browning of baked goods, "doctoring" or fondants, etc.) but may not contribute its flavor significantly to that of a product with a flavor of its own. Industrial honey is generally purchased in 55-gallon drums, but tank trucks can be available. It should be purchased on sample and on specification if possible; specification should be concerned only with attributes pertinent to use. Honey must contain, at most, 18.6% moisture, and be of acceptable color and flavor for the intended use, be filtered to assure cleanliness, be processed to remain liquid, and, if needed, to inactivate enzymes (see later). Since the semisolid honey spread retail form is more easily handled at the table without drip and stickiness, it is conceivable that a similar form would be of use in certain food manufacturing operations; it could be handled similarly to solid shortenings. Such a product could be made available by honey processors if demand indicated a need.
IV. ANALYSIS AND COMPOSITION Honey as produced by honeybees from plant nectars is rather variable in its composition, reflecting contributions of the plant, climate and environmental conditions, and beekeeper skills. Table 111 summarizes the general composition of United States honey. Data available from similar studies in other countries (White, 197Sa) provide similar values. TABLE III AVERAGE COMPOSITION OF 490 SAMPLES OF HONEY AND RANGE
OF VALUES"
Characteristics measured
Average
Moisture, percentage Levulose, percentage Dextrose, percentage Sucroqe. percentage Maltose. percentage Higher sugars. percentage Undetermined, percentage PH Free acid, meq/kg Lactone, meq/kg Total acid, meqlkg Lactone/free acid Ash, percentage Nitrogen, percentage Diastase value
17.2 38. I9 31.28 1.31 7.3 1 1.50 3. I 3.91 22.03 7.11 29.12 0.335 0. I69 0.041 20.8
"
From White et ul. (1962)
Standard deviation
1.46 2.07 3.03 0.95 2.09 I .03 I .97 -
8.22 3.52 10.33 0.135 0. I5 0.026 9.76
Range
13.4-22.9 27.2544.26 22.0340.7 5 0.25-7.57 2.74- 15.98 0.13-8.49 0.0-13.2 3.42-6.10 6.7547.19 0.00-18.76 8.68-59.49 0.000-,950 0.020-1.028 O.ooO-. 133 2.1-61.2
298
JONATHAN W . WHITE, J R .
A. MOISTURE CONTENT
I.
Analysis
The amount of water in honey is of major importance to its stability against fermentation and granulation. Normally ripened honey has a moisture content below 18.6%;honey of higher content does not qualify for the USDA grading classifications. The determination of honey moisture has been reviewed extensively by White (1969); no significant developments have occurred since that time. In that review, moisture determination by direct drying, Karl Fischer reagent, measurement of viscosity, and density by weighing and hydrometry are critically discussed, as are certain errors and inconsistencies in the literature. The most accurate and convenient procedure uses the refractometer with the conversion table recalculated by Wedmore (1955), which appears as method 3 1.1 12 of the Association of Official Analytical Chemists (Honvitz, 1975). Approximations suitable for many purposes (standard error 2 0.4%) may be obtained with a hand refractometer, providing proper calibration is used, since sucrose solutions and honey of the same refractive index differ in their solids content. Table IV provides a conversion of solids (sucrose) by refractometer to honey solids. 2 . Relation of Moisture Content to Stability The principal short-term instabilities of honey are granulation and fermentation. Liability to each is related to moisture content: fermentation by osmophitic TABLE IV CONVERSION OF REFRACTOMETER CALIBRATION AT 20°C FROM SUCROSE TO HONEY SOLIDS" _____
___
Sucrose
Honey solids
Refractive index
76.00 77.00 78.00 79.00 80.00 81.00 82.00 83.00 84.00
77.56 78.56 79.60 80.64 81.68 82.76 83.76 84.80 85.80
1.4804 1.4829 1.4855 1.488 1 1.4907 1.4934 1.4960 1.4987 1.5014
Calculated from data of Wedmore (1 955) and AOAC table 52.012 (Honvitz, 1975).
299
HONEY
TABLE V LlABlLlTY OF HONEY TO FERMENTATION”
Moisture content (%)
Liability
Below 17.1 17.1-18.0 18.1-19.0 19.1-20.0 Above 20.0
None None if yeast count < 1000/gm None if yeast count < lO/gm None if yeast count < l/gm Always liable
Data of Lochhead (1933) ba\ed on 319 honey samples.
yeasts will ensue if the combination of moisture content, temperature, and yeast count is favorable (Lochhead, 1933); granulation tendency appears to be fairly predictable by the glucose/water ratio (White et a l . , 1962; Hadorn and Zurcher, 1974). Normally ripened honey with a moisture content of 17.5-18%, with a water activity of 0.58, requires a natural inoculum of about 1000/gm to ferment (Lochhead, 1933). Table V shows the general relation between moisture content and yeast count in honey liable to fermentation. USDA (optional) standards (USDA, 1951) require that honey contain no more than 18.6% water to qualify for grading. Retail honey is usually blended to 18% water or less. A survey of composition of United States honey (White et al., 1962) showed that honey from the Western and intermountain areas is lower in moisture than that from the East and West North Central areas, with other United States areas intermediate (average 17.5% for 238 samples). Exact values will be affected by seasonal factors.
B . CARBOHYDRATES The sugars of honey have been intensively reviewed recently by Siddiqui (1970), White (1975b), and Doner (1977). For that reason, the discussion here will be limited, including only a description of our present understanding of the carbohydrate composition of honey and its analyses. I.
Average Composition and Ranges
Table 111 lists the amounts of glucose, fructose, sucrose, “maltose” (reducing disaccharides), and higher sugars found in a survey of nearly 500 samples of United States honey. The variability of honey is illustrated by the ranges shown. A better conception of this is seen in Fig. 2 which illustrates the distribution of individual values within the range for these sugars. Individual analyses for these sugars (and other components) are given by White et al. (1962) for 504 honey
300
fK1
JONATHAN W WHITE, JR.
LT
#
I
40
2 20 0
25
30
35 40 40 LEVULOSE (%)
LEVULOSE / D E X T R O S E
FIG. 2. Distribution of carbohydrate contents among 490 honey samples. Arrows indicate means. (From White et al., 1962.)
and honeydew samples from 47 states, representing 83 single floral types, 93 blends of “known” composition, and 4 honeydew types, all from two crop years. The routine paper chromatography carried out as a control in the fractionation procedure indicated that all samples had the same pattern of sugars present. Also in the publication, the carbohydrate (and other) composition of 74 honey types was compared with the average values. Effects of area of production were examined for alfalfa, cotton, and orange honeys produced in widely different areas. Any differences found were not significant. It is noteworthy that only 3 of the honey samples had a 1evulose:dextrose ratio
30 1
HONEY 46
I
I
I
I
I
I
1
D E X T R O S E I%)
FIG. 3. Dispersion of monosaccharide content of 457 honey samples; line indicates L/D = 1 . (Data of White el al., 1962.)
less than 1.0. Figure 3 shows the individual values of this ratio found for 457 honey samples. The floral source has the strongest influence on carbohydrate composition; area and seasonal influences are minor.
2 . Identity of Sugars Sugars which have been unequivocally identified in honey are listed in Table Vl. Identification requires isolation and identification by sound physical or chemical methods of analyses, not simply by comparison of chromatographic mobility. Details of isolation and identification are included in the reviews of Siddiqui (1970) and Doner (1 977). Table VII provides an approximation of the amounts of the oligosaccharides found in honey by Siddiqui and Furgala (1967, 1968a). 3 . Analytical Problems
In common with other syrups, the carbohydrate analysis of honey remained empirical for many decades. Not until White and Maher (1954) applied class separation on charcoal columns was a reasonably accurate method available for determining dextrose and levulose in honey; three fractions are obtained. This method, accepted by the AOAC, remains the method of choice; it is somewhat
302
JONATHAN W. WHITE, JR. TABLE V1 SUGARS ESTABLISHED AS HONEY CONSTITUENTS'
Trivial name Glucose Fructose Sucrose Maltose
Systematic name
a-D-glucopyranosyl-/3-D-fructofuranoside O-a-D-glUCOpyranOSyl-(1+4)-~-glucopyranose
Reference
Elser (1924); van Voorst (1941)
Isomaltose Maltul ose Nigerose Turanose Kojibiose Laminaribiose a ,P-Trehalose Gentiobiose Melizi tose 3-a-Isomaltosylglucose Maltotriose 1-Kestose
Panose Isomaltotriose Erlose Theanderose Centose Isopanose ISOmdkO-
tetraose Isomaltopentaose
O-a-D-glUCOpyranoSyl-(1+6)-~-glucopyranose ~-a-D-g~ucOpyranosy~-( 1+4)-D-fructose O-a-D-glUCOpyranOSyl-(1+3)-~-glucopyranose O-a-D-glucopyranosyl-(1+3)-~-fructose O-a-D-glUCOpyranOSyl-(I+2)-~-glucopyranose O-P-~-glucopyranosyl-( I+3)-~-glucopyranose a-D-ghcopyranosyl-P-D-ghcopyranoside O-P-D-glUCOpyranOSyl-(1+6)-~-glucopyranose O-a-D-glUCOpyranOSyl-(l-+3)-0-/3-~-fructofuranosyL(2- 1)-a-D-glucopyranoside O-a-D-glUCOpyranOSyl-(1+6)-O-a-o-glucopyranosyl-( 1+3)-D-gh1COpyranOSe O-a-D-glUCOpyranoSyl-(1-+4)-O-Cr-D-ghCOpyranosyl-( 1+4)-D-glUCOpyranOSe O-a-D-giucopyranosyl-(I+2)-P-D-frUCtOfuranosyl-(1+2)-/3-~-fructofuranoside O-cY-D-glucopyranosyI-(1+6)-O-c~-~-glucopyranosyl-( 1-+4)-D-g~ucopyranose O-a-D-glUCOpyranoSyl-(1+6)-U-cY-D-gIUCOpyranosyl-( 1+6)-D-glucopyranose O-a-D-glUCOpyranOSyl-(1+4)-a-D-glUCOpyranosyl-/3-o-fructofuranoside O-a-D-glUCOpyranOSy1-(1 -+6)-a-~-glucopyranosyl-P-D-fructofuranoside O-a-D-glUCOpyranOSy1-(1+4)-O-a-D-ghCOpyranosyl-( 1-+2)-D-glUCOpyranOSe O-a-D-glucopyranosy1-( 1+4)-O-Cy-D-glUCOpyranosyl-( 1+6)-D-glucopyranose O-a-D-glUCOpyranoSy1-(I+6)-[O-a-D-gIUcOpyranosyl-( 1+6)],-~-glucopyranose O-cY-D-glUCOpyranOSyI-(1+6)-[O-a-~-glucopyranosyl-( 1+6],-~-glucopyranose
' From Doner (1977).
White and Hoban (1959) White and Hoban (1959) White and Hoban (1959) White and Hoban (1959) Watanahe and Aso (1959) Siddiqui and Furgala (1967) Siddiqui and Furgala (1967) Siddiqui and Furgala (1967) Siddiqui and Furgala (1968a)
Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968b) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a) Siddiqui and Furgala (1968a)
303
HONEY TABLE VII
YIELDS OF THE PRINCIPAL SUGARS IN THE OLIGOSACCHARIDE FRACTION (3.65%)OF HONEY”
Disaccharide
(%)
Trisaccharide
(7o)
Higher oligosaccharide
(%)
~~
Maltose Kojibiose Turanose lsomaltose Sucrose Maltulose, (and 2 unidentified ketoses) Nigerose a ,P-Trehalose Gentiobiose Laminaribiose
29.4 8.2 4.7 4.4 3.9
3. I 1.7 1.1 0.4 0.09
Erlose Theanderose Panose Maltotriose I-Kestose Isornaltotriose Melizitose lsopanose Centose 3-a-Isomaltosylglucose
56.99 a
4.5 2.1
Isomaltotetraose lsomaltopentaose
2.5
1.9 0.9 0.6 0.3 0.24 0.05
0.33 0.16 0.49
Acidic Fraction
Not investigated
6.51
trace
13.69
Data of Siddiqui and Furgala (1967, 1968a).
lengthy if many analyses are required. Recent development of high-performance liquid chromatographic (HPLC) analyses of sugar mixtures (Conrad and Palmer, 1976) promises a more rapid procedure without loss of accuracy. The several gas chromatographic (GLC) procedures are of intermediate value, suffering as they do from the need to derivatize. A large degree of empiricism remains in the charcoal column procedure, however. Sucrose can be analyzed specifically only if yeast invertase hydrolysis is used; melezitose is eluted in the disaccharide fraction and interferes if acid hydrolysis is used. The remaining mixture of reducing disaccharides is measured by reducing power and reported as “maltose.” It is too complex to allow individual sugars to be quantitated, even by HPLC. All oligosaccharides (other than melezitose) are present in the higher sugars fraction, reported as glucose after hydrolysis. Literature reports of the sugar composition of honey must be examined with a knowledge of the analytical procedures employed. Values obtained by polarimetric or saccharimetric means are only roughly approximate; those using specific methods for either glucose or fructose and calculation of the other by difference are not accurate unless a class separation has first been done. Values from GLC analysis for the monosaccharides are usually acceptable; for sucrose and other sugars they may not be since demonstration of the singular nature of peaks, considering all of the honey sugars, has not been done. Data from HPLC may be quite acceptable for monosaccharides and sucrose, but all other disaccharide peaks should be combined as “maltose,” since the
304
JONATHAN W . WHITE, JR.
chromatograms do not reflect the known complexity of the honey sugars. Improvements in columns and detectors may eventually provide analyses of more components.
C . ACIDS The characteristic flavor of honey (if such a variable commodity can be said to have a characteristic flavor) includes a contribution due to its acidity. The pH of honey (Table 111) averages 3.91, with a range for 490 samples of 3.42 to 6.10. This level of active acidity probably also contributes to the stability of honey against microbiological attack.
I.
Gluconic as the Principal Acid
Gluconic acid, in equilibrium with gluconolactone, is the principal acid of honey (Stinson et a l . , 1960). It is produced by the action of the glucose oxidase normal to honey on the glucose (White et a l . , 1963b). This reaction is extremely slow in full-density honey but rapid when honey is diluted. It has been proposed that this acid is produced from nectar glucose during the ripening of nectar to honey by the bee. The combined effect of acidity and the hydrogen peroxide concurrently produced is thought to assist in preserving nectar from spoilage during the ripening. Burgett (1974) has shown that this also occurs in nine other eusocial Hymenoptera.
2. Other Acids Ten other organic acids have been identified in honey by suitably rigorous procedures, and seven more are probably present. The former group includes acetic, butyric, lactic, and pyroglutamic (Stinson et al., 1960), citric and succinic (Nelson and Mottern, 1931), formic (Vogel, 1882, cited by Farnsteiner, 1908), maleic (Goldschmidt and Burkert, 1955), malic (Hilger, 1904), and oxalic (von Philipsborn, 1952). In the latter group are glycollic, a-ketoglutaric, and pyruvic (Maeda et al., 1962), tartaric (Heiduschka and Kaufmann, 1913), and 2- or 3-phosphoglyceric acid, a- or P-glycerophosphate, and glucose-6phosphate (Subers et a l . , 1966).
3 . Analysis The titration of total acidity of honey had been an empirical procedure because of a fading endpoint. This was shown to be caused by hydrolysis of gluconolactone (White et a l . , 1958); the present procedure measures free acid and lactone. Of 490 samples of United States honey, only two were found not to contain
HONEY
305
gluconolactone; their pH values were uncommonly high: 5.01 and 6.10 (White et al., 1962). The amount of gluconic acid in honey should be a reflection of several contributing factors, the most significant being the time between the collection of the nectar by the bee and the attainment of full density in the comb, since the action of glucose oxidase essentially stops at full density. This is governed by the sugar content of the nectar, the weather, the strength of the colony, and the quality (i.e., density and volume) of the nectar flow. A greater time needed for ripening permits production of more gluconic acid; it also results in more manipulation of the ripening honey by the bees, with addition of more enzyme. A need exists for an analytical procedure to determine total gluconic acid and gluconolactone in honey. Present lactone titration is not satisfactory because the position of the lactone-acid equilibrium, as related to honey pH, is not known.
D. MINERAL CONTENT 1 . Average Amounts of Principal Minerals The wide variability of honey composition is reflected also in the ash content. Table 111 shows an average of 0.17%, with a range from 0.02-1.03%. The predominating mineral element is potassium, which averages about one-third of the ash; sodium content is roughly one-tenth as much. Schuette and his students at Wisconsin (Schuette and Remy, 1932; Schuette and Huenink, 1937; Schuette and Triller, 1938; Schuette and Woessner, 1939) published the data summarized in Table VIII. They found honey in the two lightest color classes to have lower mineral content than the darker honey types. This was confirmed by White et ul. (1962). 2.
Trace Minerals
The literature on the content of these minerals and 13 others in honey from other areas of the world was reviewed by White (1975a). Tong et al. (1975), in an examination of the value of trace analysis of honey as an indicator of pollution. reported ranges for 41 elements in 19 New York State honey samples. Samples collected by bees in the vicinity of the New York Thruway appeared to contain elevated levels of elements known to be emitted by internal combustion engines. E.
PROTEINS AND AMINO ACIDS
The nitrogen content of honey is low and variable. Table 111 shows an average for United States honey of 0.04196, with a standard deviation of 0.026 (63%).
306
JONATHAN W. WHITE. JR. TABLE VlIl MINERAL CONTENT OF HONEY”
As parts per million of honey
Mineral element Potassium (K) Sodium (Na) Calcium (Ca) Magnesium (Mg) Iron (Fe) Copper (Cu) Manganese (Mn) Chlorine (Cl) Phosphorus (P)
Sulfur (S) Silica (Si02)
Honey color
No. samples
light dark light dark light dark light dark light dark light dark light dark light dark light dark light dark light dark
13 18
13 18 14 21 14 21 10 6 10 6 10 10 10 13 14 21 10 13 10
10
Range
Average
100-588 1154,733 6-35 9-400 23-68 5-266 11-56 7-126 1.2O-4.80 0.70-33.50 0.14-0.70 0.35-1.04 0.17-0.44 0.46-9.53 23-75 48-20 I 23-50 27-58 36- I08 56- I26 7-12 5-28
205 1676 18 76 49 51 19 35 2.40 9.40 0.29 0.56 0.30 4.09 52 I13 35 47 58 100 9 14
“ Data of Schuette and Remy (1932), Schuette and Huenink (1937), Schuette and Triller (1938). and Schuette and Woessner (1939). Paine et al. (1934) reported an average of 55% of the nitrogen lost by ultrafiltration (range 26-93%); White and Kushnir (1967b) noted that about 4 0 4 0 % of the nitrogen is lost on dialysis. Bergner and Diemair (1975) more recently reported 3345% to be removed by ultrafiltration (10,000 limit). Most of the nonprotein nitrogen is in free amino acids.
I.
Proteins
Early interest in protein content was in distinguishing honey from artificial mixtures and blends. The volume of precipitates with honey and tannin (Lund, 1909), phosphotungstic acid (Lund, 1910), or alcohol (Laxa, 1923) was used. Immunological tests were studied as early as 1903 (Langer, 1915). Thoni (1913) proposed using an antiserum to royal jelly or “beebread” for this purpose.
307
HONEY
Indeed, Langer (1 915) immunologically differentiated honey protein and proteins of hand-collected pollen, refuting Kiistenmacher’s earlier claim that protein in honey was extracted from the pollen and that Langer was in error in ascribing it to the bee. Studying the colloidal material removed from honey by ultrafiltration, Paine et al. (1934) found it to be more than half protein, isoelectric at pH 4.3, and precipitable by colloidal bentonite. Helvey ( 1 953) found three components in the colloidal material from a buckwheat honey: proteins of molecular weight of 146,000 and 73,000 and a presumed polysaccharide of 5,000 weight. White and Kushnir (1967b), using gel filtration, ion-exchange chromatography, and starch-gel electrophoresis, examined proteins of eleven floral types of honey and sugar-fed stores. From four to seven proteins were found, of which four originate with the bee. The molecular weights of two of the latter were approximately 40,000 and 240,000; those from the plant were about 98,000 and >400,000. Figure 4 shows gel filtration and starch-gel electrophoresis of a preparation from
4.9 A
B
LA) C
C * H H
-
D
1
-
20
ORIG
I
,
4 0 60 80 100 120 140 F R A C T I O N NO. L2rnl)
II I @ ,H,I,
20 4 0 60 80 100 120 140 M l G R A T l O N TOWARD ANODE ( m m l
FlG. 4. (A) Filtration of I ml dialyzed concentrate (= 10 gm honey) goldenrod-aster honey on 2.1 X 60-cm column of Sephadex (3-200 in 0.01 M phosphate pH 6.5; (B) Starch gel electrophoresis of fractions combined in pH 8.9 borate, 4.0 Vlcm as indicated in A. Pattern at bottom is from original material as applied to column. (From White and Kushnir, 1967b.)
308
JONATHAN W . WHITE. JR.
(A)
0.5 0.4
0.3
~
t A
0.2 0.1
u =
'0
05
" 01 40 60 80 F R A C T I O N NUMBER ( 2 m l )
FIG. 5 . DEAE-cellulose chromatography (bed 0.8 x 18 cm) in 0.01 M phosphate (pH 8.0) of concentrated dialyzed protein preparations. Solid lines: apparent protein by optical method. Broken lines: by Lowry method (scales left). Straight solid lines: gradient of concentration (scale right). (A) Goldenrod-aster preparation, 0.25 ml, (B) Lespedeza, from single comb, I .O ml. (From White and Kushnir, 1967b.)
a goldenrod-aster honey. In Fig. 5 is seen the greater resolution produced by ion exchange cellulose chromatography of this preparation. The chromatogram of a preparation from sugar-fed bees (no nectar components) is shown in the same figure to indicate its less complex nature. In Fig. 6 are gel filtrations of protein preparations from another honey and from stores from sugar-feeding. The larger number of components in the former is apparent. Bergner and Diemair (1975) have also examined by gel filtration protein preparations from several types of honey and from sugar-feeding. Their results have generally confirmed those of White and Kushnir. They ascribed three of the five elution peaks to the bee and two to plant components.
2. Amino Acids The formol titration, essentially a measure of total amino acid content, was applied to honey by Tillmans and Kiesgen (1927) who proposed that it be used to
HONEY
309
FIG.6 . Filtration of concentrated dialyzed honey preparations (0.5 ml) on Sephadex G-200. Protein by Lowry method. (A) cotton honey, (B) stores from sugar-fed bees. (From White and Kushnir, 1967b.)
authenticate honey. European limits for this value were shown by Schuette and Templin ( 1 930) to be inapplicable to United States honeys, which were generally lower and more variable. Lothrop and Gertler (1933) described a procedure for amino nitrogen in honey, reporting an average of 0.0033% (range 0.00240.0066%). Schuette and Baldwin (1944) reported averages of 0.0034% for light and 0.0058% amino nitrogen for dark honeys. The introduction of paper chromatography renewed interest in honey amino acids; several investigators identified up to 17 amino acids in various samples. Komamine (1960), quantitating paper chromatography, first noted that proline was the preponderant amino acid. Later the automatic amino acid analyzer was used for honey analyses; a considerable body of analyses is now available (Curti and Riganti, 1966; Mizusawa and Matsumuro, 1968; Michelotti and Margheri, 1969; Hahn, 1970; Biino, 1971; Bergner and Hahn, 1972; Petrov, 1974; Davies, 1975). Table IX shows Davies’ values for free amino acids in 98 honey samples. All agree that proline predominates, representing 50-85% of the total. Davies (1975) has reviewed the sources of honey amino acids. Since pollen contains about 1.5% amino acids, with proline predominating, Komamine proposed this as the source. Nectar contains small amounts of free amino acids but little proline. Davies calculated that far too little pollen is present in honey to account for the proline. Bergner and Hahn (1972), noting proline to comprise 80% or more of the generally lower amino acid content of sugar-fed bee stores, ascribed
W
0
TABLE IX AMINO ACID ANALYSIS OF
No. samples:
Argentina 8 Avg.
Glucosaminic acid Methionine sulphoxide Aspartic acid Unknown A "Amides" Glutamic acid Proline Unknown B Glycine Alanine Cystine Valine Methionine
SD
1.14 1.76 0.64 6.48 2.12 53.10
1.241 0.587 0.218 2.961 0.809 16.589
0.55 1.73
0.176 0.679
1.27 0.07
0.346
Australian eucalypt/clover 16 Avg.
SD
Australia unspecified 15 Avg.
SD
Canadian 16 Avg.
SD
SD
0.955 2.169
2.05
0.245
0.59 4.50
1.463
9.90
1.592
6.68 2.02 51.81
0.561 0.2% 6.954
14.82 5.73 44.96
4.532 1.831 16.170
0.71 1.85 0.30 1.32
0.115 0.482 0.125 0.48 1
30.90 6.00 83.42 13.54 1.48 3.84 0.45 6.16 0.09
4.673 0.643 23.930 3.691 0.208 0.601 0.179
7.57 1.66 93.95
1.120 0.245 16.658
50.49
3.439 1.981 25.424
0.54 2.42 0.66 1.43 0.25
0.169 0.673
0.43 1.21
0.190 0.632
0.43 1.16
0.102 0.121
0.318 0.090
0.92 0.33
0.372 0.223
1.32 0.97
0.248
1.51
Avg.
SD
1.664
1.268 0.577
1.47 0.46 5.09
Avg.
Yucatan 14
3.17 2.38 1.24
1.44
U.S. clover 13
r
1.545
AII samples
98 Avg.
SD
3.21
1.688
1.74 3.44 0.95 11.64 2.94 59.65 21.04 0.68 2.07 0.47 2.00 0.33
1.174 3.212 0.937 9.334 2.163 26.765 20.612 0.407 1.523 0.212 1.854 0.232
Isoleucine Leucine Tyrosine Phenylalanine P-Alanine y-Amino butyric acid Unknown C Unknown D Unknown E Lysine Unknown F Ornithine Histidine Tryptophan Arginine Total:
1.76 0.74
0.317 0.288 0.420 0.204 0.223
0.69 0.54 1.13 3.35 0.66
0.278 0.205 1.280 3.757 0.387
3.82 2.89 7.26 60.49 1.00
1.52 0.41
0.207 0.156
0.26 1.28
0.1 15 0.337
0.131 0.602 1.431 1.238
0.421 0.113 0.148 0.327 0.393 0.139 0.112 1.238 0.986 0.767
4.23 0.87 0.61 0.63 1.97 0.93 0.43
0.600
0.12 5.42 1.91 0.86
1.30 0.37 0.25 0.37 1.06 0.31 0.25 4.83 3.83 1.12
33.976
83.88
7.530
90.46
26.601
0.72 0.69 0.91 2.07 0.98
0.279 0.267 0.391 1.525 0.452
0.73 0.64 1.29 3.17 I .57
0. I70 0.203 0.382 1.241 0.400
0.47 0.88 2.24 5.07 1.29
0.229 0.753 1.689 4.253 0.260
1.34 0.63 0.30 0.33 I .32 0.64 0.24 6.04 2.06 0.82
0.318 0.681 0.068 0.175 0.768
2.76 1.49
0.770 0.242
0.69 0.44 0.34 0.20 2.70
0.082 0.206
1.62 1.19 0.77 0.64 0.34 0.41 0.27 1.85
0.810 0.538 0.176 0.225 0.1 18 0.191 0.191 0.398
0.437
1.30
0.690
1.12
84.10
26.484
127.96
19.857
77.25
0.064
2.305
0.106
0.084 0.728
0.75 0.59 1.04
~~
' From Davies (1975).
* mg amino acid/100 gm honey (dry wt). A blank in the SD column indicates that only one sample contained the amino acid.
0.514 2.176 13.800 0.402
1.12 1.03 2.58 14.75 1.06
1.191 0.898 2.826 24.806 0.510
2.15 0.81 0.66 0.57 0.99 0.50 0.26 3.84 3.84 1.72
1.186 0.513 0.476 0.442
6.17 6.05
0.617 0.275 0.092 0.244 0.486 0.440 0.172 0.838 0.996 3.372
0.359 0.223 1.982 3.393 2.269
252.28
41.846
118.77
69.491
5.64
0.%1
0.666
312
JONATHAN W . WHITE, J R .
05[
03
,
,
I
O""'S
0005 001
003 005 01 ASPARTIC / P R O L I N E
3
FIG. 7. Regional separation of honeys by ratios between concentration of individual amino acids. A Australian, 0 Canadian, 0 United States clover, 0 Yucatan. (From Davies, 1975.)
it to the bee. Petrov related it to the important part proline plays in aerobic muscle exchange products in all insects. Davies, using data for 98 samples of honey, has suggested that certain ratios between contents of various amino acids could be used to determine the geographic source of a honey; Fig. 7 indicates one such approach. Later (Davies, 1976) this approach was refined by using a computer-aided selection of 60 amino acid ratios. Fifteen of 16 samples not used to establish the program were correctly assigned to one of the four locations shown in Fig. 7 , showing that while there are variations in the ratios between samples of the same area, the variation between sources is much greater.
F. ENZYMES That honey contains enzymes has been known for more than a century since Erlenmeyer and Planta (1 874) reported their presence in bees, pollen, beebread, and honey. As the author has noted earlier, The enzymes are among the most interesting materials in honey, possibly have received the greatest amount of research attention over the years, and have supported the greatest burden of nonsense in the lay and even scientific press. The use of enzyme activity in some countries as a test for overheating of honey seems to support by implication the occasional supposition by food faddists that the enzymes of honey have a dietetic or nutritional significance of themselves (White, 1975a).
The greatest volume of literature reports on honey enzymes until most recently dealt with their use as indicators of honey identity and quality, largely heat
313
HONEY TABLE X ENZYME ACTIVITIES OF HONEY
Enzyme a-Glucosidase (invertase, sucrase) Diastae (a- and P-amylase) Glucose Oxidase
Catalase Phosphatase
Average activity
Number of samples
7.5-10
1468
References Duisberg and Hadorn (1966) Duisherg and Hadorn (1966)
16-24 20.8 80.8 167 210 4.97" 86.8" 13.4 5.07
263 90 24 10
2R 10 25
White er ul. (1962) White and Suhers (1963) Dubtmann (1971a) Dustmann ( 197 1h) Schepartz and Suhers (1966a) Du\tmann (1971h) Dzialoszynski and Kuik (1963) Zalewski ( I 965)
units
a-Glucosidase: Diastase: Glucose oxidase:
Catalase:
Phosphatase: "
gm sucrose hydrolyzed per 100 gm honey per hour at 40°C gm starch converted per 100 gm honey per hour at 40°C pg H,O, accumulated per gm honey in I hour under experimental conditions. Because honey contains substances oxidized by H,Oz, this is not a true measure of glucose oxidase Catalatic activity per gram, K , = I h (In x d x ) DIW where x o is initial substrate. x is substrate at time, r , D is dilution. and W is sample in grams (Schepartz and Suhers, 1966a) mg P/100 gm honey/24 hours
Includea nine values of zero
' Includes four values of zero. exposure. Most countries other than the United States require minimum values for amylase activity and proposals for use of other enzyme activities for this purpose still arise. The honey enzymes of most direct interest in food applications are amylase, invertase, and glucose oxidase. Catalase and acid phosphatase are also present. The amounts of these enzymes normally found in (unheated) honey are shown in Table X , to provide an idea of order of magnitude. I.
Invfrtase
A sucrose-splitting enzyme is added to nectar by the honeybee during its harvesting and ripening to honey. It continues its activity in extracted honey unless destroyed by heating. It is an a-glucosidase (White, 1952; White and Maher, 1953a) with inherent transglucosylase action. During its action on su-
3 14
JONATHAN W. WHITE. JR
crose, six oligosaccharides are formed, all eventually hydrolyzed to glucose and fructose by the completion of the reaction. The principal intermediate is a-maltosyl p-D fructofuranoside (White and Maher, 1953b) trivially named erlose (also termed glucosucrose, fructomaltose). It can accumulate to as much as 11% of the original sucrose (White and Maher, 1953b) during the reaction. Maltose is formed in lesser amounts. Echigo and Takenaka (1973) have studied the carbohydrates and a-glucosidase in stores produced by sucrose-fed caged bees; during the ripening they reported erlose to appear in the earlier part of the ripening period and remain throughout the ripening period. Figure 8 shows the progress of ripening of sucrose stores. The optimal conditions for the transferase reaction were found to be pH 6.0, 30°C, and 0.25 M sucrose. Examination of a-glucosidase from several honeys and from stores of sucrose-fed bees (White and Kushnir, I967a) indicates that preparations seemingly homogeneous by Sephadex gel filtration (Fig. 9) show 3-9 components by ion-exchange chromatography. The preparation from sugar-fed bees, however, appeared homogeneous with an approximate molecular weight of 5 1,000 indicated. Figure 10 illustrates DEAE cellulose chromatography of a-glucosidase from honey and of stores from sugar-fed caged bees. A highresolution starch gel electrophoresis procedure (White and Kushnir, 1966) further resolved all preparations into 7-18 isozymes. Figure l l compares the pattern of a-glucosidase isozymes from a bulk honey (i.e., extracted from combs taken from many colonies at several locations) with that of the a-glucosidase from a comb honey (i.e., produced by a single colony) and that of stores from sugar-fed bees.
-
\
50c3
0
\ \
\
24
48
72
96
HOURS
FIG. 8. Changes in sugar content during the process of honey formation. x fructose, A glucose, 0 sucrose, 0 erlose. (From Echigo and Takenaka, 1973.)
315
HONEY
8
r-----l
0
FIG. 9. Sephadex (G-200) filtration (2. I X 31 cm) in 0.014 M phosphate (pH 6.5) of enzyme preparations from honey (solid lines, scale left): a-glucosidase from (1) 0.5 ml preparation (= 1 . 1 g) goldenrod-aster honey; (2) 0.5 ml preparation ( g 5 . 5 g) clover honey; (3) 0.25 ml preparation from (= 8.9 gm) stores from sugar-fed bees. Broken line (scale right) from same clover honey. (From White and Kushnir. 1967a.)
The greater complexity of the preparation from bulk honey is probably the result of blends of honey from many colonies. The single-colony samples have equivalent numbers of isozymes. Noteworthy is the much lower migration rate of the sugar-fed samples, which have no plant components. White and Kushnir suggest that the bands may represent genetic differences among bees. Methods
0 u 2
D
0 FRACTION NUMBER 121111 1
40 60 ao FRACTION NUMBER ( 2 m I l
FIG. 10. (A) Chromatography on DEAE-cellulose of a-glucosidase preparation from cotton honey elution with 0.01 M potassium phosphate (pH 8.0); KCI gradient as shown; solid line: a-glucosidase activity (scale left); broken line: “protein” measured by optical method (scale right). ”Protein” retained and fractionated, 14.9 mg (73%); a-glucosidase retained and fractionated. 73.5 units (27%); (B) a-glucosidase preparation from stores of sugar-fed bees. (From White and Kushnir, 1967a.)
316
JONATHAN W . WHITE, JR. 401
I
I
I
1
I
I
1
I
1
M I G R A T I O N TOWARD ANODE ( m m )
FIG. 1 I . Starch-gel electrophoresis of a-glucosidase preparations from honey, borate (pH 8.9): (A) clover honey. 96 unitslml; (B) Lespcdeza honey, 38 unitslml; (C) stores from sugar-fed bees. 44 unitdml. A at 3.70 V/cm, B and C at 3.52 V/cm. (From White and Kushnir, 1967a.)
used are sufficiently sensitive to examine the a-glucosidase of single bees in this fashion. a . Origin and Kinetics of Honey Invertase. The question of the source of the sucrose-inverting enzyme of honey has intrigued scientists since its discovery in honey. No purpose is served by reviewing the earlier literature; it has been accepted for many years that the major portion is that added by the bee during the collection of nectar and the ripening process. Whether any plant enzyme from nectar or pollen is present has not been definitively shown. Gothe (1914) concluded that both plant and insect were sources, since more enzyme activity was present in honey than in stores from sugar-feeding. Schonfeld (1927), however, found the invertase activity of sugar-fed stores to be inversely related to the concentration of the feed; no information is available on the concentration fed by Gothe. The kinetic study of honey invertase by Nelson and his colleagues (Nelson and Cohn, 1924; Nelson and Sottery, 1924; Papadakis, 1929) remained the definitive work until recently. Differences from yeast invertase were found in pH optima and the initial reaction course, initial yeast invertase rates being practically
317
HONEY
constant, in contrast to a marked rate increase in the honey invertase inversion. Two kinetic studies with the objective of determining the source of honey invertase provide the only recent kinetic data on the reaction. Rinaudo et al. ( I 973) undertook to demonstrate that invertases from the other possible sources (pollen, nectar) differ from that of honey. Invertase from the hypopharyngeal gland of the bee and from honey were shown to have the same pH and temperature sensitivity, substrate and reaction products (glucose and fructose; intermediates were not mentioned), and inhibition by fructose. Comparison of reciprocal rate plots for the enzymes from the bee, honey, two pollens, and nectar showed identical Michaleis constants (0.17 M )only for the first two. However, they reported that none of the preparations showed maltase, contrary to earlier reports (White and Maher, 1953a; Gontarski, 1954; Maurizio, 1961). The definitive study of the a-glucosidase of the hypopharyngeal gland of the honeybee (and hence of honey) is that of Huber. Huber (1975) and Huber and Mathison ( 1 976) have purified two sucrases from honeybees, confirming Gontarski’s ( 1 954) earlier studies. The less soluble of these precipitated between the same values of ammonium sulfate saturation and exhibited kinetics very similar to those of honey sucrase. Final purification was by affinity chromatography. The previously reported transglycosylase activity was confirmed and a kinetic study of the hydrolytic and synthetic reactions was carried out. As seen in Fig. 12, the rate of release of fructose is rectilinear; the rate of glucose release drops at
an FRUCTOSE
40
0
ZOO
400
600
BOO
1000
Yo/ [ SU C R 0sE l
FIG. 12. Kinetic Hofstee plot of the production of glucose and fructose from sucrose. For incubation details see original. Units are micromole product per minute from 0.2 M sucrose at 30°, pH 6.5. From Huber and Mathison, 1976. Reproduced by permission of the National Research Council of Canada from the Canadian Journal of Biochemistry 54, 153-164, 1976.
318
JONATHAN W . WHITE, JR.
FIG. 13. Proposed mechanism for action of the major sucrase of honey bees; S . sucrose, G . glucose, F. fructose. G-S trisaccharide. W , water. From Huber and Mathison, 1976. Reproduced by permission of the National Research Council of Canada from the Canudion Journal ofBiochcrnisrry, 54, 153-164, 1976.
high sucrose concentrations. The scheme in Fig. 13 explains this: the reaction releasing fructose (K3) is rate limiting, and sucrose is not only initial substrate, but an acceptor for the transglucosylation to form the glucose-sucrose trisaccharide, erlose. Huber did not identify the trisaccharide with erlose. The K cat for fructose, glucose, and erlose formation according to the scheme are: Kcat (fructose) = K3
(1)
K for formation of all three products by this mechanism is K 2 / K I . (K cat values which include substrate concentrations are not, in fact, constants, since the values change with substrate concentrations.) The earlier value of 51,000 for the molecular weight (obtained by Sephadex filtration in 0.1 M phosphate pH 6.5) was confirmed by equilibrium ultracentrifugation, but Huber and Mathison found values of 82,500 by Sephadex filtration ( 0 . 2 M citrate pH 6.5) and 78,000 by SDS electrophoresis. The glycoprotein nature of the enzyme was confirmed by amino acid and amino sugar analysis. Honey a-glucosidase has been reported to have optimal maximum activity at pH 6.0 (10% hydrolysis), 5.7 (35-55% hydrolysis) (Nelson and Cohn, 1924), 5.9 (Rinaudo et af., 1973), and 5.5 (Huber and Mathison, 1976). Hadorn and Zurcher ( I 962), comparing several published procedures for determination of sucrase in honey with particular attention to the pH optimum, selected 6.3, commenting that very little difference in activity was found between pH 5.8-6.5. Huber and Mathison, who used the most highly purified enzyme, show a flat maximum between pH 5.5 and 6.0. This discrepancy in K , values reported for the major honeybee enzymes (and for honey sucrase) by Rinaudo et al. (1973) (0.17 M ) and Huber and Mathison (1976) (0.030 M ) may perhaps be resolved by a calculation of K M for honey
HONEY
319
invertase using the data published by Nelson and Cohn ( I 924). A LineweaverBurk plot of the 5-minute data from their Table 10 gives KM = 0.031 M. Considering the crude nature of their preparation, this is excellent agreement. When comparing data among the few recent reports on these a-glucosidases, the sources and extent of purification must be considered. Huber and Mathison prepared their material from whole honeybees, necessitating extensive purification. They also prepared the enzyme from an unpasteurized supermarket honey. The latter was purified by dialysis and ammonium sulfate precipitation. They stated that the kinetic properties and apparent K M values of the honey enzyme were the same as the “major honeybee sucrase.” The sucrase of the head portion of the honeybee is almost entirely of this type. Rinaudo er a / . (1973) prepared honey sucrase by dialysis, and that from excised honeybee food (hypopharyngeal) glands and pollen by extraction and centrifugation only. No estimates of purity were made. The hypopharyngeal glands are known to contain an active glucose oxidase (Gauhe, 1941) which, if not removed, can distort results by preferentially removing glucose. This glucose oxidase activity was probably eliminated in Huber’s preparation from bees, and it was probably present in those of Rinaudo. It is not clear, however, whether the less extensive purification given the honey enzyme by Huber effectively removed the glucose oxidase known to occur in honey. Huber makes no mention of glucose oxidase. There is no question that the a-glucosidase of honey acts upon maltose; no reason is apparent for the contrary report of Rinaudo et a/. (1973). Huber and Mathison (1976) report activity against maltose to be 83% that against sucrose. White and Kushnir (1967a) show exactly parallel activity for 13 isozyme peaks separated by starch-gel electrophoresis; the activity against maltose was only 30% of that against sucrose. Takenaka and Echigo (1975) purified honey a-glucosidase by DEAE cellulose chromatography and obtained K M for maltose = 0.00526M, with optima at pH 6.0 and 30°C. Sucrose was hydrolyzed at three times the rate for maltose. b. Heat Inactivation. The papers of Rinaudo et al. (1 973) and of Huber and Mathison (1976) include data on heat inactivation of the a-glucosidase in buffer. The latter report only that the activity rapidly disappears at 55’45°C; Huber and Mathison state that with 10-minute exposure destruction begins at 40°C and is essentially complete at 60°C. Trade interest in the invertase activity of honey centers about its possible use as an indicator of heating history. Most of the literature on the subject deals, therefore, with inactivation in full-density honey. White et a / . (1964) have shown that the rate constant for inactivation of honey invertase in buffer is 24 times that in full-density honey. Several proposals have been made to establish minimum values for sucrase in honey, to be used together with diastase values for estimating heat exposure
320
JONATHAN W . WHITE, JR.
(Kiermeier and Koberlein, 1954; Duisberg and Gebelein, 1958; Hadorn et a / . , 1962). The Codex Alimentarius standards, however, do not include sucrase activity. No discussion of analytical methods will be included here, beyond the observation that Dustmann (1972) pointed out that existing methods in which the reaction takes place in diluted honey (including those of the preceding three papers, above) underestimate the activity by 10-30%, presumably because of the inhibitory effect of the monosaccharides of honey. He recommended dialysis as a pretreatment. The procedure earlier used by White et al. (1 964) in their study of the effect of storage and processing on honey enzymes had used dialysis for this reason. In that study, a single expression was found adequate to describe the inactivation of invertase in full-density honey by heating and long-term storage. The first-order rate equation is log K = 26.750 -
39730 2.303 RT
(4)
which is plotted in Fig. 14. Figure 15 shows the half-life of honey invertase over the temperature range of 2Oo-80"C. 10,
00
1
I
I
'
-10 L L9
5 -20
-30
-4.0I 280
1
I
I
300
320
340
f
1
x lo5
FIG. 14. Effect of temperature on rate of heat inactivation of diastase and invertase in honey; + Schade et al. (1958). W Lampitt et a / . (1929), 0 Duisberg and Warnecke (1959), A Kiermeier and Koberlein (1954). From White er al., 1964. Reprinted from Food TechnologylJournal of Food Science 18(4), 153-156, 1964. Copyright @ by Institute of Food Technologists.
HONEY
32 I
DAYS
68
104 140 STORAGE TEMPERATURE
FIG. 15. Approximate time required at a given temperature between 20°C (68°F) and 80°C (176°F) for the diastase and invertase activities of a honey sample to be reduced to one-half of the initial value. (From White, 1967.)
2 . Amyluse For an enzyme whose occurrence in honey has been known for 100 years, with hundreds of reports written on methods of quantitation, factors affecting activity in honey, and reporting assays of many thousands of honeys, very little is known of its kinetics, mode of action, and indeed, the significance of its presence in honey. a . tsolation and Pur$cation. The importance attached to amylase assay as a quality factor in honey is indicated by its inclusion in the Codex Alimentxius standards for honey. This reflects the preference of consumers in many countries for honey with relatively minor exposure to heat. Establishment of minimum acceptable values for honey “diastase” provides a control procedure. Since a few honey types are known that are naturally deficient in diastase, special provision is made for exceptions. Kerkvliet and van der Putten (1973) have compared five methods for determining the diastatic activity of honey by measuring the loss of iodine-coloration power.
322
JONATHAN W. WHITE, JR. TABLE XI SUMMARY OF ANALYTICAL DATA ON THE PREPARATION OF AMYLASE FROM HONEY"
Preparation stage Original honey (200 gm) Dialysed solution Acetone precipitate Amm. sulph. precipitate DEAE (cellulose) bands: A B
Total diastase units
Total a-amylase units
Protein (mg)
Diastase
a-Amylase Diastase ___ mg protein mg protein a-amylase
6520 6191 4590 1853
220.0 229.0 148.0 59.0
3180.0 1663.0 731.0 134.0
2.0 4. I 6.3 13.8
0.07 0.14 0.20 0.44
29.6 29.7 31.0 31.4
292 241
31.0 8.2
2.2 66.2
130.0 3.1
13.80 0.12
9.4 30.1
-
~
" From Schepartz and Subers (1966b) Honey has both a-and P-amylase activity; both the increase in reducing sugar and loss of coloration with iodine have been used in assays. The latter is most commonly used. Lampitt er a f . (1929) reported optimal pH for the a-amylase to be about 5.0 in the 22"-30"C temperature range, and 5.3 between 45"-50"C. For P-amylase, a value of 5.3 was reported. Schepartz and Subers (1 966b) attempted to separate the a- and P-amylase activities of honey by several procedures. Used as the final isolation procedure, ion-exchange cellulose chromatography provided two principal fractions. Table XI summarizes the research. Efforts to characterize the pooled fractions were unproductive because of their instability; the presence of a-glucosidase further complicated interpretation of the results. A 200-fold purification of the a-amylase was attained. White and Kushnir (1967a) carried out Sephadex gel filtration of dialyzed honey concentrates and determined amylase activity on the fractions. As seen in Fig. 16 a single peak was obtained using maleate buffer, indicating an approximate molecular weight of 21,600. Interaction with the Sephadex is evident when phosphate buffer is used with three maxima in the elution curve. Bergner and Diemair (1975), however, obtained single elution peaks in phosphate (pH 5.3) from G-200, as seen in Fig. 17, but considerably more retarded than the glucosidase peak in maleate, similar to the highest peak in phosphate in Fig. 15.
b. Heat Inactivation. Diastase has been used for at least 75 years as an indicator of honey heating, Nearly all of the reports are therefore oriented to this aspect, as is true for the invertase. Most of this work has measured a-amylase activity. The few papers reporting P-amylase inactivation in honey must be discounted because the a-glucosidase and glucose oxidase in honey may vitiate the results by their effect on maltose or glucose from the amylolytic reaction.
323
HONEY 1
t L
= w
1
1
SEPHADEX G-200 -M A L E A T E _.. PHOSPHATE .
L
3l
2
F R A C T I O N NUMBER (2111 I
FIG. 16. Sephadex (2.1 x 31 cm) filtration of enzyme preparation from cotton honey in 0.01 M maleate (pH 6.5). and in 0.01 M phosphate (pH 6.5). (From White and Kushnir, 1967a.)
The most complete study of heat inactivation of diastase (a-amylase) in honey remains that of White et al. (1964), who showed that loss of diastase by heating and by extended storage at lower temperatures obey first-order kinetics and can be described by the equation
Iog K
=
22.764
-
35010 2.303 RT
as shown in Fig. 14, which also includes data from other investigators. Figure 15 provides an estimate of the half-life of diastase in honey within the temperature range 20"-8OoC. It should be pointed out that although the relationship between time, temperature, and invertase and diastase activity have been widely quoted by honey
GLUCOSE OXIDASE
i I I
400
V, m l
-
500
V, ml
-
FIG. 17. Elution diagram of 10 ml protein concentrate on Sephadex G-200 (2.3 x 83.5 cm) in 0.03 M phosphate (pH 5.3) after freezing (-20°C) and remelting. (From Bergner and Diemair, 1975.)
324
JONATHAN W. WHITE, JR
scientists and control officials, they are based on a study of only three United States honeys (White et a/., 1964). This work should be extended to include other representative honeys. c . Source of Honey Amylase. In contrast to the a-glucosidase, which has a clear and essential function in the conversion of nectar to honey, no such function has been assigned to the starch-digesting enzymes in honey. Nectar contains no starch or dextrins. The question of its origin, in view of this, has been examined for many years. The presence of definite amounts of diastase in stores from sugar-fed bees (diastase numbers about 10 or less) led Gothe (1914) and others to ascribe it mostly to the bee, with a contribution from pollen. Vansell and Freeborn (1929) later contended that pollen, known to have diastatic activity, was the principal source, and Lothrop and Paine (193 1 ) supported this, citing the great variation in diastase value among honeys of different floral type. On the other hand, Fiehe (1932) considered nectar to be the major source; most honey has diastase numbers considerably in excess of the lower values common to stores from sugarfeeding. Braunsdorf (1932) found diastase numbers of 17.9 in two sugar-fed samples and proposed that it originates largely from the bees, with the variability resulting from the different degree of manipulation by strong or weak colonies upon slow or heavy flows of nectar. Weishaar (1933) ascribed only 1.5-2.5% of the diastase to nectar, 0.25-0.75% to pollen, and the remainder to the bee. Rinaudo et al. (1973) considered honey amylase originating from the bee as the basis of an optimal pH of 5.6-5.9 for preparations of the enzyme from honey and from honeybee hypopharyngeal glands. Nectar amylase pH optimum was 7.2, but that from pollen at pH 5.9 did not differ appreciably from that of honey. Amylase from honey and the bee was activated by chloride ion, in contrast to that of pollen and nectar. If diastase originates largely in the food glands of the honeybee, as does a-glucosidase, it would be expected that the ratios of these two enzymes in honey be relatively constant. Since a-glucosidase is more heat (and storage) labile, values for unheated honey should be used. The writer has calculated the correlations between diastase number and sucrase number of 39 unheated samples from the literature: 30 Swiss honeys from Table 1 of Hadorn et a/. (1962), 4 United States honeys from Table 2 of the same paper (actually supplied to the authors by the writer), and 5 stores from sugar-feeding, described in Table 1 from Hadorn and Zurcher (1963). A correlation coefficient of +0.83 resulted with F = 26.2, significant at less than 0.01%. There appears to be little doubt that the major portion of the diastase in honey originates from the bee, and the variability probably reflects the specific conditions during gathering and ripening of the nectar.
HONEY
325
3 . Glucose Oxidase Honey has been thought from ancient times to have wound-healing and antiseptic properties, and within the past 40 years a distinct heat-labile antibiotic activity has been the subject of considerable interest. The activity was named "inhibine" and a biological test was devised for its measurement in honey (Dold et al., 1937). During analytical studies on honey, the writer and his colleagues found that the drifting end-point common in the determination of acidity in honey, ascribed by Cocker (19%) to an acid-producing enzyme, was actually caused by the hydrolysis of lactone material in honey (White et al., 1958). Further studies indicated gluconic acid, in equilibrium with gluconolactone. to be the principal honey acid (Stinson er af., 1960). With the knowledge that a glucose oxidase had been reported in the hypopharyngeal glands of honeybees (Gauhe, 1941), it was demonstrated that the enzyme was present in honey and its production of gluconic acid and hydrogen peroxide during the standard microbiological test for inhibine was responsible for the major part of the antibiotic effect (White et al., 1963b). A chemical assay was described (White and Subers, 1963) in which the accumulation of hydrogen peroxide in diluted honey during a I-hour incubation was measured colorimetrically. From the results on 45 samples assayed by the Dold et af. (1937) plate assay and by the chemical assay, a relation between the inhibine number and log of peroxide accumulation was found. The effect of heating honey for 10 minutes at 70°C on the inhibine number and on peroxide accumulation was investigated for 29 samples, and for 6 the half-life of the peroxide accumulation system was determined (White and Subers, 1964a). A wide range in stability was found: most samples lost 85-95% of the activity when heated 10 minutes at 70°C, but seven lost less (6-71%) and five lost more (96-100%). Figure 18 shows a comparison of the heat sensitivity of the peroxide accumulation system in six honeys with that of honey diastase and invertase. The great variability, which precludes its use an index of heating exposure, is obvious. In a study of the previously reported instability of inhibine to light, White and Subers ( I 946b) found a wide variation in this effect; some honeys lost 90% of the activity on exposure to normal laboratory fluorescent light for 1 hour, others lost only 10% in full sunlight for 10 minutes. The sensitivity is maximal at 425-525 nm and pH 3, and is negligible at pH 6-7; a heat- and light-stable. nonvolatile sensitizer was postulated. The use of inhibine number or glucose oxidase activity as a measure of honey quality on heat exposure is therefore impractical because of the wide range of activity and the wide range (70-fold) of heat sensitivity shown by authentic honeys.
326
JONATHAN W. WHITE. JR TEMPERATURE
(OF)
I00
10
Y W
-
_I
u.
0. I
A U I
0.0 I
TEMPERATURE
('CI
FIG. 18. Effect of temperature on the half-life o f the peroxide accumulation system in honey. Diastase and invertase shown for comparison. (From White and Subers, 1964a.)
Schepartz and Subers (1964) and Schepartz (1965a,b, 1966a) have examined the glucose oxidase of honey. The enzyme is a true glucose oxidase, aerobically transferring H2 directly to molecular oxygen. Its pH optimum is 6.1 and requires 0.1 M Na+ for maximal activity. Of 32 carbohydrates tested, only glucose (100%) and D-mannose (9%) were oxidized. The optimal temperature is 40°C; it is completely inactivated at 60°C. It is strongly inhibited by NaCN and semicarbazide, somewhat by EDTA, mannose, fructose, and azide. Heavy metals at 0.001 M did not inhibit. The enzyme shows a preference for P-D-glucose over a-D-glucose of about 6:l. A kinetic study (Schepartz, 1965b) revealed the unusually high optimum substrate concentration of about 2.7 M (equilibrium glucose) as indicated in Fig. 19A. The Michaelis constant is 1.49 M, shown in Fig. 19B. In terms of 0-D-glucose, the optimal substrate concentration is 1.8M. The reaction follows zero order kinetics and is stoichiometric. Further investigation of the apparent contradictory action of fructose as an inhibitor or activator under certain conditions was carried out by Schepartz (1966a). Using manometric procedures, the results in Fig. 19 were obtained using varying concentrations of glucose (0.5-2.7 M ) and fructose (0.1-2.2 M).
HONEY
0
10 SUBSTRATE
327
20 30 40 C O N C E N T R A T I O N (MI (A1
(El
FIG. 19. (A) Effect of substrate concentration in velocity of reaction of honey glucose oxidase with glucose. measured manometrically. For details see original. (B) Lineweaver-Burk plot. Points were derived from those in A; v , initial velocity, (S), substrate concentration. K,, Michaelis constant, 0, points included in statistical analysis, A points not included in statistical analysis since beyond optimum concentration, 63, points derived from statistics and used to locate line. (From Schepartz, I965b.)
The plotted data in Fig. 20A suggest a coupling or uncompetitive inhibition by fructose. The plot of the transposed data from the experiments in which both sugars were present by the method of Hunter and Downs (1945) yielded the hyperbolic curve in Fig. 20B, which is found, according to Webb (1 963), in the rare instance of coupling or uncompetitive inhibition wherever the inhibitor combines only with the enzyme-substrate complex, never with the enzyme alone.
328
JONATHAN W . WHITE, JR.
0
200
100
300
400
v /IS1
12,
I
FIG. 20. (A) Plot of velocity-substrate data. In (a) the complete system contained: 0.5to 2.7 ml 3.5 M glucose in 0 . 2 M sodium phosphate (pH 6. I ) , and enough of the same buffer to total 3.4ml in the main space; 0. I ml (419 units) enzyme preparation in a side-arm sac; 0 . 2 ml 10% KOH in the center well. In (b) the system was the same except I .7M fructose was present in the main space at the 1.7 M and 1.OM glucose levels. Blanks were run without enzyme, without substrate; v , initial velocity in p molehin, (S). substrate concentration in M , vgOI. maximum velocity, K,,,Michaelis constant. (B) Plot of transposed velocity-substrate data. Conditions same as in A, except that fructose concentration was varied from 0. I to 2 . 2 M ; the combined sugar concentrations never exceeded 3.4M.(From Schepartz, 1966a.)
HONEY
329
Webb states that this gives rise to circumstances in which the inhibitor can cause activation. There is little doubt that honey glucose oxidase originates in the bee. Gauhe (1941) has shown that the glucose oxidase of the hypopharyngeal gland of the honeybee also has a high optimal substrate (about 2 M ) and a high Michaelis constant (KM0.63), greatly in contrast to those of other reported glucose oxidases such as the KM of 0.0042 for a mold enzyme reported by Keilin and Hartree (1948). Bee and honey glucose oxidases have pH optima at 6.1 and 6-7, respectively, and are equally specific for glucose. Most of the few other glucose oxidases oxidize a number of substrates. The great variability among honeys of peroxide accumulation, related to the antibiotic effect. does not imply a corresponding variability in glucose oxidase content of honey. The peroxide accumulation assay is carried out with diluted honey so that any constituents oxidized by peroxide will depress the value found. Thus data such as those of White and Subers (1963) showing a thousandfold variability in peroxide accumulation cannot be cited as indicating that the honeybee is not the source of the enzyme. To the writer’s knowledge, no true assays of glucose oxidase in honeys have been reported. 4 . Other Enzymes Catalase and an acid phosphatase are the remaining enzymes demonstrated to occur in honey. Gontarski (1948) described a “vitamin C oxidizing enzyme” in the hypopharyngeal glands of bees and observed a similar action in honey. He proposed that it might be identical with the glucose oxidase Gauhe (1941) reported in the bee glands. It is now apparent that this enzyme is in fact glucose oxidase; Schepartz (1966a) showed ascorbic acid to be a powerful activator of honey glucose oxidase by way of product removal and not due to action on the enzyme itself.
a. Caruhse. Schepartz (1966b) has reviewed critically the eight reports of catalase in honey which have appeared since Auzinger first reported it in 1910. Because of the earlier use of inappropriate methods and inconclusive experiments, he rejected them and, using manometric and spectrophotometric procedures, has claimed the first unequivocal evidence for catalase in honey. Using a dialyzed honey solution, he found a pH optimum at 7-8.5, a Michaelis constant of 0.0154M. and an optimum substrate concentration at 0.018M H202, with the reaction being first order. Subsequently Schepartz and Subers (1966a) described a kinetic assay procedure and reported catalase values for 28 honeys. Since diastase and peroxide accumulation values were available for the same samples, correlations were calculated. A direct correlation (r = -0.76, sig. at 0.01 probability level) was found between diastase and catalase, and as expected, an inverse
330
JONATHAN W . WHITE. JR
correlation (r = -0.71, sig. 0.01) between catalatic activity and peroxide accumulation. In the latter calculation, however, 12 samples which showed little or no catalase but also had little or no peroxide accumulation were not included. Inclusion of these samples reduces the correlation to 0.11, significant at less than the 90% probability level. It is evident that catalase activity is but one of the factors contributing to variability in peroxide accumulation. Dustmann ( 1 97 1 b) provides further evidence; he assayed 11 samples for catalase, using Schepartz’s procedure, and also for peroxide accumulation. The four samples with extremely high peroxide values (380-662) were totally devoid of catalatic activity. Dustmann’s other catalase values (46.1-241) are all greatly higher than those of Schepartz (0.5-17.8); the same procedure and units appear to have been used. Calculations by the writer based on the seven samples with catalatic activity showed a correlation coefficient between catalase and peroxide accumulation of 0.023 (not significant). When all samples were included, the correlation coefficient is -0.71 (sig. at .05). Using the assay procedures that Schepartz has since declared inapplicable (Schepartz, 1966b), Gillette (1931 ) reported source of catalase to be pollen. Dustmann (1971b), in the only study using acceptable procedures, has found very high catalase activity for pollen, very little in nectar. No reports could be found of catalase assay of stores of sugar-fed bees on which acceptable assay methods were used. b. Phosphutuse. Giri (1938), on the basis of the production of inorganic phosphorus from P-glycerophosphate during a 24-hour incubation at 35°C with diluted honey, stated that honey contains an acid phosphatase. The activity was maximal at pH 4.5-6.5 and was increased by magnesium ions. His two (of eleven) most active samples were “slightly fermented”; Giri stated that fermented honey samples were characteristically high, and values for unfermented samples were decidedly low, and it is lowered somewhat by pasteurization. He suggested that it is “derived chiefly from fermentation yeast and bees and partly from the plants.” Giinther and Burckhart (1967) described an improved procedure requiring a 3-hour incubation with p-nitrophenylphosphate. Zalewski (1965) assayed honey, pollen, nectar, and bees for acid and alkaline phosphatases using disodium phenyl orthophosphate as substrate, with incubation of 2.25 hours at 37°C for honey and nectar, 18-24 hours for pollen. Chloroform was added to control microbiological action. Acid phosphatase activity in honey and nectar ranged between 30-2140 and 15-2750 pmolell00 gm dry weight, respectively. Stores from caged sugar-fed bees had about ‘/16 of the honey average. The acid phosphatase assay of pollen ranged from 1260-145,500 pmole/l00 gm. It is implied that pollen is the principal source, although it is apparent that nectar contains sufficient to account for the activity in honey. Whether the enzyme can pass through the wall of the intact pollen grain is debatable.
HONEY
33 1
G. FLAVOR AND COLOR 1.
Flavor
Table honey is attractive to the consumer for a variety of reasons, flavor possibly being the most significant. While there seems to be a characteristic “honey flavor,” the wide variety of flowers attractive to bees overlays a great multiplicity of source-specific flavors and aromas. Color is also variable and strongly influenced by source, but more susceptible to environmental factors than is flavor. Flavors are sufficiently distinctive that dozens of different floral types can be identified by flavor alone by the experienced taster. Typical flavors can range from the most delicate and desirable to some that are harsh and objectionable. Generally, though not invariably, the lighter colors are associated with the milder, more pleasant flavors. The flavor complex includes, in addition to the volatile aromatic materials dominating sweetness, contributions from the acids, traces of polyphenolics, amino acids, and in some cases specific bitter or characteristic nonvolatile notes.
2. Aroma Relatively little attention has been given to the volatile aroma constituents. Gas-liquid chromatography has been applied by several groups of investigators (Dorrscheidt and Friedrich, 1962; ten Hoopen, 1963; Cremer and Riedmann, 1964). As is true of most natural products examined in this way, the lower aliphatic aldehydes, ketones, alcohols, and esters make up the bulk of the identified components. Cremer and Riedmann identified over half of 120 compounds separated by a 1 mm X 100 m Golay column and observed, after long storage, increases in pentanol, 2-methyl- I-butanol, 3-methyl-]-butanol, and n-propanol, suggesting that these may arise from the corresponding amino acids. Sixteen of the 22 honeys examined contained phenylethyl alcohol, and 14 also had benzyl alcohol. It is noteworthy that most synthetic honey flavors contain large amounts of lower aliphatic esters of phenylacetic acid and phenylethylsalicylate or phenylacetate; Jacobs (1955) states that nearly all phenylacetic esters are characterized by a honeylike taste and odor. An example of a type-specific aromatic is methyl anthranilate (MA), reported in citrus honey by Lothrop (1932). Lavender honey also contains it (Hadorn, 1964). Knapp (1 967) has elaborated upon White’s (1966) suggestion that MA content may be a useful quality measure for citrus honey, commenting that only 1 of 1000 samples would be expected to contain less than 1.5 ppm. He proposed that additional work be done with known single-source citrus samples as needed. The 80 predominately citrus samples reported by the two investigators averaged 3.8 ppm MA, range 0.844.9. Twelve noncitrus samples analyzed by White
332
JONATHAN W . WHITE, JR
averaged 0.07 ppm (range 0-0.28). Data for Knapp's 14 noncitrus samples were not available. 3.
Color
Little is known of the specific compounds responsible for the color of honey. Of 92 honeys Browne (1908) analyzed, 25 gave a positive test for polyphenolic compounds with FeCI,; the most intense reactions were from the darkest honey. Milum ( 1 939) ascribed the increase in color of honey upon storage to reaction of iron from processing equipment and containers with polyphenols, the browning reaction of reducing sugars and amino acids, and the instability of fructose in acid solution. Von Fellenberg and Rusiecki ( 1938) found water-soluble coloring materials to increase with honey color more than did fat-soluble colors. H.
VITAMINS
Honey has measurable amounts of six vitamins but at such low levels that they have no nutritional significance. Table XI1 summarizes the significant data. Widely conflicting reports of the ascorbic acid content of honey have been ascribed to interfering materials in the chemical determination. Most honeys contain less than 5 mg/100 gm. Some reports of values as high as 390 mg/100 gm by chemical means should be discounted, but Griebel(1938) confirmed chemical values for mint honey of 160-280 mg/l00 gm with bioassays in which 1 gm honey/day protected guinea pigs, corresponding to 100-200 mg/l00 gm. High ascorbic acid content ( I 18-240 mg/l00 gm) of Iranian honey was reported by Rahmanian et al. (1970) by chemical analysis and TLC of derivatives, and also confirmed by bioassay, which indicated levels of 75-150 mg/100 gm. They TABLE XI1 VITAMIN CONTENT IN MICROGRAMS PER 100 GRAMS OF HONEY
Samples
Riboflavin
Pantothenic acid
Niacin
Thiamine
Pyridoxine
Ascorbic acid
61 63 22 26 12-54
105 96 20 54 -
36Oh 32Oh I24 108 442-978
5.5 6.0 3.5 4.4 8 -22
299 320 1.6 10.0 -
2400 2200 2000-3400
~~
29 Minnesota" 38 U.S. and Foreign" 21 U.S. 3-7 years old" 19 U.S. 1-2 years old" 4 India"
Haydak et al. (1942).
* Corrected from original data in publication as later shown (Haydak et a / . , 1943). Kitzes er a / . (1943). Kalimi and Sohonie (1965).
HONEY
333
proposed use of this specific honey type (of unknown floral source) for helping relieve marginal vitamin C deficiency often found in Iran.
I.
TOXIC CONSTITUENTS
Since tremendous numbers of organic compounds are synthesized by various plants, many with substantial physiological activity, it is inevitable that some may, on occasion, be found in honey. The remarkable aspect is that, as widely as bees forage, the instances of toxic reactions are so few. White (1973) has reviewed the subject in some detail. Perhaps the best-known toxins are those of honeys from the Ericuceue (Rhododendron, Azulra, Andromeda, Kulmiu spp.), with literature descriptions reaching back to Xenephon's description of the mass poisoning of the expedition of Cyrus in 401 B.C. in Asia Minor, presumably by honey from Rhododendron; instances still occur in that area. Other areas from which reports of intoxication from Ericucear honeys are USSR, eastern and Pacific Northwest United States, and Japan. Beekeepers are largely aware of the problem and take appropriate steps to avoid it. Other toxic honey types are those from the tree tutu of New Zealand (actually a honeydew), henbane (Darura metel), Datura stramonium and Hyoscyamus niger, yellow jasmine (Jessamine), euphorbia, and arbutus. Details of toxicology, compounds responsible, and other aspects are in the review by White ( I 973).
V.
PHYSICAL CHARACTERISTICS
The physical attributes of honey are largely conferred by the high concentrations of sugars that compose most of the solids. Viscosity, refractive index, and specific gravity are so closely related to solids content that each has been used to measure moisture (solids) content. Refractive index is the most easily used, as implied in Table IV. A complete table of refractive index-moisture content equivalents appears in the Book of Methods, AOAC (Horwitz, 1975). Specific gravity (20/20") varies regularly with moisture content, between 1.4404 at 14.0% moisture through 1.4174 at 18.0% to 1.3550 at 21 .O%. A table at 0.2% intervals is available (Wedmore, 1955). Because of the fairly wide natural range, care must be taken to mix thoroughly when blending honeys of different moisture content to avoid layering. A.
RHEOLOGY
Much early effort was expended in attempts to determine moisture of honey with such instruments as the hydrometer (Chataway, 1933) and the falling-ball viscosimeter (Chataway , 1932; Oppen and Schuette, 1939) with approximately
334
JONATHAN W. WHITE, JR
the accuracy, but without the facility, of the refractometric measurement. Using absolute viscosity values, Lothrop (1939) found a rather wide variation among honeys adjusted to equivalent moisture contents. Munro's (1943) data (Table XIII) are the most extensive available. The high viscosity of honey is most apparent when draining containers or in pumping or processing it. Although Munro stated that most of the decrease of viscosity on warming takes place from room temperature to about 30°C, his observation was based on a linear plot. Pryce-Jones' (1953) plot of Munro's data as log viscosity versus 1/T shows that rate of change is relatively constant; only the extent of heating needed to obtain the required viscosity reduction should be applied to minimize heat-induced damage to color and flavor. MacDonald (I 963) has examined the effect of temperature on the flow of honey through pipes under a constant head. Table XIV shows the results. For the average of all four pipe
TABLE XI11 VISCOSITY OF HONEY
Moisture content
Temperature
(8)
("C)
Sweet clover" (Melilotus)
16.1
Sage" (Sulviu)
18.6
White clover!' (Trifoliurn repens)
13.7 14.2 15.5 17.1 18.2 19.1 20.2 21.5 16.5 16.5 16.5
13.7 20.6 29.0 39.4 48. I 71.1 11.7 20.2 30.7 40.9 50.7 25.0
Type
Sageb Sweet clover" White clover" a
Data of Munro ( 1 943).
* Interpolated from Munro's data.
25 25 25
Viscosity (poise) 600.0 189.6 68.4 21.4 10.7 2.6 729.6 184.8 55.2 19.2 9.5 420 269 138 69.0 48.1 34.9 20.4 13.6 1 I5 87.5 94.0
335
HONEY
TABLE XIV RELATIVE FLOW OF HONEY IN PIPES”
Temperature Pipe diameter (inside)
82°F (28°C)
102°F (39°C)
122°F (50°C)
% in. (19 mm) 1 in. (25 mm) 1% in (31 mm) 1% in. (38 mm)
149 361 129 1263
400 913 1895 2609
1125 2353 5000 6192
Rate of flow (in pounds per hour) through 4-inch (10-cm) length of pipe with 4-inch head. Data of MacDonald (1963).
sizes, the rate of flow increases equally with each of the two temperature increments. The importance of pipe diameter in moving honey at the lower temperatures is shown by the eightfold increase in flow obtained when the cross-section area of the pipe is increased four times. This effect declines as the viscosity decreases with increasing temperature. Most honeys are Newtonian liquids but some have been reported to have non-Newtonian properties. Pryce-Jones (1953) has examined the rheology of heather honey, which is so thixotropic that it cannot be removed from the comb by a centrifugal extractor unless the gel-sol transformation is effected by applied vibrating rods. This property is ascribed to the properties of the proteins; if isolated heather honey protein is added to clover-honey, it exhibits thixotropic behavior. Manuka (Leptospermium scoparium) honey from New Zealand and Karvi (Carvia caflosa)from India (Deodikar et al., 1957) are markedly thixotropic. Pryce-Jones (1952) also reported that Opuntia honey from Nigeria and several Eucalyptus types exhibited dilatancy , which he ascribed to the presence of a high-molecular dextran.
B. THERMAL PROPERTIES Relatively little data are available on the physical properties of honey with respect to heat, even though honey can easily be damaged by its improper application. Processing equipment design has generally been based on data from sugar processing. The specific heat of honey at 17.4% moisture was reported by Helvey (1954) to be 0.54 cal/gm/”C at 20°C with a temperature coefficient of 0.02 caV”C. He also measured specific heat of honey solutions. Townsend (1954b) has described McNaughton’s determination of specific heat. He used a considerably larger sample and obtained somewhat higher results, as seen in Table XV.
336
JONATHAN W. WHITE, JR
TABLE XV SPECIFIC HEAT OF HONEY"
Moisture content (%)
Specific heat
20.4 19.8 18.8 17.6 15.8 14.5 Coarsely granulated Finely granulated
0.60 0.62 0.64 0.62 0.60 0.56 0.64 0.73
" Data of MacNaughton (Townsend, 1954b). Basic data necessary in designing a heat exchanger for honey processing have been obtained by Detroy (1966). Using a concentric-tube exchanger, the surface conductance or film coefficient for honey was determined at flow rates of 700975 lb/hour and two temperature ranges of interest in processing, preheater (65"68°C) and flash heater (85"-88"C). Figure 21 shows the values calculated from these data for a range of honey flow velocities. In this work honey was in laminar flow, water in turbulent flow. Detroy used Helvey's value for specific heat of 0.54. He pointed out the desirability of experimental verification of his values
I
I
I
I
0.18 0.20 0.22 0.24 0.26 V E L O C I T Y OF HONEY FLOW ( f t / s e c )
FIG.21. Change of film coefficient with velocity of honey flow in the honey-to-watertemperature difference range of each heating circuit. Upper line, flash heater water circuit; Y = 10.6 + 272x; Sv,z = 1.68. Lower line, preheater water circuit;y = 26.51 + 53.3~;SZ," = 1.7. (From Detroy, 1966.)
337
HONEY
under accurately controlled conditions, since the possibility for experimental error is considerable. The heat sensitivity and relatively low heat conductance of honey have encouraged examination of high-frequency heating of honey. Lackett and Wilson (1971) used a kitchen-type microwave oven operating at 2450 mHz to heat and liquify completely granulated honey in 1 -Ib jars. With the metallic caps removed, heating was rapid but not uniform; when a temperature of 60°C was reached in the center of the jar, a damaging 98°C was reached at the top. Difficulty was encountered in heating larger jars (2% Ib) without boiling the surface layers. Bergel and Stuwe ( 1 972) have proposed the use of dielectric heating for honey processing. They have estimated from small-scale heating experiments that a 25-kW dielectric heating installation would be required to heat I000 kg per hour from 30" to 55°C. The frequency was not specified. Normal range for dielectric heating is 2-100 mHz. The arrangement and results of the small-scale tests are y
0 5 K W HIGH FREOUENCY HE ATlN G EO U I P M E N T
60
IW
a
ELECTRODE AREA 240 I I B O m m
13
I0: 4
I
20 0
CONTAINER PETRI D I S H
wt-
WEIGHT OF HONEY m:200q
Lg 0
BETWEEN ELECTRODES ON T E F L O N S U P P O R T S
(A1
0
05 10 1 5 2 0 2 5 HEATING T I M E t I m i n )
FIG 22 (A) Dielectric heating of honey in a glass dish (From Bergel and Stuwe, 1972 ) (B) Temperature variation in high-frequency heating of honey in ajar (From Bergel and Stuwe. 1972 )
338
JONATHAN W . WHITE. JR.
shown in Fig. 22A. The authors point out that, as seen in the figure, the power absorption decreases with increasing temperature, i.e., the dielectric loss value for honey decreases with increasing temperature. This automatically provides temperature equalization. In Fig. 22B is seen the results of heating honey in a commercial jar (500 gm) with metal cap. Under the proper conditions the jar of honey is heated to 59'43°C in 2% minutes. Overheating induced boiling in a ring under the cap. Analysis of all honey samples in the latter experiment showed no change in diastase value or HMF content. Samples remained liquid for at least 6 months. By heating I kg in two jars in the 2-kW chamber at full power, a linear heating rate of 32"C/minute was found between 20" and 60°C. From this a high frequency power requirement of 25 kW was calculated for heating 1000 kg/hour from 30" to 55°C. C.
HYGROSCOPICITY
I. Equilibrium Relative Humidity The ripening of nectar to honey by the bee includes its repeated exposure in a thin film to warm air. The solids content reached is a function of the extent of moisture saturation of the air in the hive, which is related to temperature and to the external air conditions. Nearly all honey contains less glucose than fructose, the more hygroscopic carbohydrate, and is remarkably hygroscopic for a natural material. As seen in Table XVI, honey in its normal moisture range of 16.818.3% is in equilibirum with air at 55-60% RH. In general, attention must be given to hygroscopicity in handling and processing, since, as Martin (1958) has TABLE XVI APPROXIMATE EQUILIBRIUM BETWEEN RELATIVE HUMIDITY OF AIR AND THE WATER CONTENT OF A CLOVER HONEY"
Relative humidity
Water content
("/.)
(%)
50 55 60 65 70 75 80
15.9 16.8 18.3 20.9 24.2 28.3 33. I
"
Interpolated from the data of Martin (1958).
339
HONEY
shown, moisture from the air diffuses only slowly into the mass, so that aerobic yeast growth is encouraged at the surface. For example, Martin (1958) showed that a honey sample at 22.5%moisture exposed to air at 86% RH for 7 days had 26% moisture in the surface layer; 2 cm below no change was found.
2 . Comparison with Other Carbohydrates Relatively few data are available to provide for comparison of commercially available carbohydrates. Table XVII summarizes values for honey, invert syrup, fructose syrup, and commercial glucose at 20% moisture. Uncertainty among values of different investigators makes it difficult to determine if real differences exist among honey, invert syrup, and fructose syrup. Conventional corn syrup is definitely less hygroscopic. Data are not available for high fructose corn syrup.
D.
CRYSTALLIZATION 1 . Glucose
a . Cause and Prediction. As noted earlier, the stable form of most extracted honey is a matrix of glucose hydrate crystals in a syrup. This is due to a considerable extent to the lower storage temperature to which the honey is exposed after removal from the bee colony. Nearly all honey is supersaturated with respect to glucose except for a few nongranulating types that are relatively low in glucose, such as tupelo and sage. Crystallization of glucose from honey while in the
TABLE XVlI EQUlLIBRlUM RELATIVE HUMIDITY OF VARIOUS CARBOHYDRATES AT 20% MOISTURE"
Material Honey Invert sirup
Levulose sirup Commercial glucose
'I
tJ
E.R.H.
Reference
63.5%" 63.2 67 67.5 57.5 63.5 61.3 75 72
Lothrop ( 1937) Martin ( 1958) Lothrop (1937) Dittmar (1935) Money and Born (1951) Money and Born (1951) Lothrop (1937) Lothrop (1937) Money and Born (1951)
Most values interpolated from original data Average of five samples.
340
JONATHAN W . WHITE, JR.
comb, though relatively rare, may be encountered with such honey as dandelion, blue curls, and ivy. It is likely that the extraction process encourages subsequent granulation by introducing fine glucose crystals from equipment, from the air of the extracting plant, and from containers. Natural crystallization, before heating, is usually fine grained, reflecting the presence of myriads of fine seed crystals and initiators such as dust, pollen, and fine air bubbles. After honey is heated and/or filtered, seed crystals are no longer present, and when crystallization finally takes place it is usually coarse grained and slow. Proper heating and processing will delay granulation for many months. Two general approaches to predicting granulation tendency of honey have been made: study of model systems and empirical correlation of various parameters with observed behavior. Examples of the former are the work of Jackson and Silsbee of the U.S. Bureau of Standards, Lothrop of the U.S. Department of Agriculture, and Kelly of the University of Tasmania. Jackson and Silsbee (1924) examined several systems at 30°C and discussed the glucose-fructosewater system with reference to honey. In the presence of solid glucose hydrate, solubility of glucose decreased from 54.6% without fructose to 32.5% at 39.4% fructose. Their conclusion that all honey is supersaturated with respect to glucose (even never-granulating tupelo honey) was based on inadequate analytical procedures then used for honey, which overestimated glucose; in addition, their data did not extend to the higher fructose concentrations found in some honey. Lothrop (1943), in an unpublished thesis, extended their data to higher fructose concentrations. He found an abrupt increase in dextrose solubility at a fructose concentration of about 150 gm in 100 gm water. The solid phase in the region of higher solubility was anhydrous glucose. Identification was by crystal form. Lothrop felt that the increased solubility was not related to the a-p equilibrium, but rather to the extent of hydration of the glucose in solution, and concluded that this accounted for the failure of certain honeys to crystallize. Lothrop's data, replotted on a ternary diagram, are shown in Fig. 23, since they were never published. Kelly (1 954), without knowledge of Lothrop's work, published the complete diagram for the system at 30°C. He also noted an area in which anhydrous glucose is the solid phase (Fig. 24) and an invariant point at which both forms are in equilibrium. He suggested that in solutions saturated with fructose, the transition temperature of the monohydrate was reduced from >50"C to