' lDI~¥>*‘1: . "If r?” @3171} P I 1? II‘IHF‘N " P “:1, ; ;é‘:_l:‘ I ‘ "PEER. “O >181 W4, _., , ~-._.__ Haj-v:— 1g :5 ; Lu}. '5! .. ;_-3.: 2' INS” tr;- it; 1543;? g, f g! I PP! Pig! )1} “1%“??? ”I“ v-u. w ...:.-—" _-_ I“ EI- '«fo' 4-: J. ,h “a; I ...,. _. ”:1 .1} -wm—J #W h v- ~ijévé’rf. _ ‘- -u‘gw , , n 'N - 5.5" @5555 vmy - I ...__. 'r .I‘ Wu - 1“ _ . L... III-I‘m 3:}: I r 3- - L! . .Ir‘lfifll‘ :z‘fiu an 3'- IIIIIFIfi‘EE‘fié .. I-II I ITHtI4"3"ggg1II~~I4£'i¢l] Ieii‘ffzfgt‘x, ": v-'1Im;;si;ygz.zo.g f ' 3.}; ‘3’. 7:52;”! ”[4 , {Pugh ‘P '1 ‘3' 1. .I':‘: ....... .I ‘11}:I‘3555-F’IL 23f: “I .[C'PF \"‘~ 0" ‘r- 5 ‘P P 1T; ”A 41;": A: ‘1‘“. I“ ha ll-.':‘.’ Pg‘ul LL‘ :, "I?” :9‘ER “(I W P”! I‘h‘k {A K and ‘25 . film-1" "P " HIP“ PM ‘I'II'LUP‘H‘ 3W" ‘ Pr ”pvl‘rn ‘15! 5.5}: .II' I 7,31]! m ILPJIPPPP 4:1" PIIW'Ifin 'Wui‘ffnfl Wflé‘q Was-m H Mi" min '!{I 0!)“ 1‘51; :PP‘P’I‘PPPI' Ill“ W Ilflililiiilillliilliiflilfliliflilii 31293 01064 3538 i i I v . 541“ ?Q‘ "77“ “W _ :9!" a H' ’5' ; / ‘s :5 sell-remn- q "w “z.- '5' This is to certify that the dissertation entitled Microbial Degradation of Sorbic Acid presented by Ibrahim Saad Ai-Mohizea has been accepted towards fulfillment of the requirements for Ph.D. degree in Food Science 8 Human Nutrition 011’ finer Mfior profs/so:- Date §//7/9I MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 OVERDUE FINES: 25¢ per day per item RETUMING LIBRARY MATERIALS: Place in book return to remove charge from circulation records JAN 0 7 Lou: 12.20 02; MICROBIAL DEGRADATION OF SORBIC ACID BY Ibrahim Saad Al-Mohizea A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department ofiFood Science and Human Nutrition 1981 . :7 I, I. (1’ / f ' r (1/ L_ // ABSTRACT MICROBIAL DEGRADATION OF SORBIC ACID By Ibrahim Saad Al-Mohizea The main objective of the study was to investigate the microbial decomposition of sorbic acid in cultural media. Initial screening tests on selected microorganisms were conducted to ascertain their resistance to sorbate, and to select some for further studies. Of the fungi tested, the molds Aspergillus flavus, Asgerflus gigs; Penicillium Ea, Fusarium fir Trichoderma viride, and Ceotrichum candidum and the yeasts Candida ljpolxtica, Hansenula saturnus, Brettanomyces claussenii and Schwa nniomyces gccidentalis exhibited relatively high resistance to sorbate. These fungi were not inhibited by 3000 mg/kg of potassium sorbate at pH 5. 5 on YPG medium. Of the bacteria tested, Lactobacillus bulgaricus, Lactobacillus plantarum, Pediococcus cerevisiae and a strain of Pseudomonas fluorescens were the most resistant. Lactic acid bacteria were inhibited by 3000 mg/kg of potassium sorbate on YTC medium at a pH of 5. 5. All of the sorbate-resistant molds. apparently were able to utilize sorbate as a carbon source. The yeasts Candida Iipolytica and Hansenula californica and the bacteria Pseudomonas vulgaris, Staphylococcus aureus and Salmonella senftenbeég also used sorbate as a carbon source. Growth of some microbes which utilized sorbate as a carbon source was retarded by the presence of glucose beside sorbate. This phenomenon was designated as the "glucose inhibitory effect." The fate of sorbate in media inoculated with some selected microbes was followed both qualitatively and quantitatively by chromatographic Ibrahim Saad Al-Mohizea procedures. Data indicate that the extent of depletion of sorbate varied, ranging from no significant loss with the lactic acid bacteria to total loss with molds such as Penicillium sp. , Trichoderma viride, Rhizopus stolon‘ifer and yeasts such as Candida Iipolytica. All Penicillium and Aspergillus species and Trichoderma viride depleted sorbate when present in sub- lethal Concentrations. 1,3-Pentadiene has been shown to be associated with growth of the above molds on sorbate-containing me'dia. Sorbaldehyde, sorbyl alcohol, and ll—hexenol were produced by species of Rhizopus, Mucor and some species of Fusarium. Sorbaldehyde was detected only while sorbate was still in the system. Sorbyl alcohol was produced in the early stages of incubation. ll-Hexenol was detected in high amounts at late stages of growth. Therefore, it appears that sorbic acid is reduced to sorbaldehyde which is in turn reduced to sorbyl alcohol. This alcohol is further reduced to u-hexenol. ll-Hexenoic acid and trace amounts of ethyl sorbate were produced by Ceotrichum candidum, a yeast-like mold, and yeasts such as Candida limlytica. Clostridium perfringens, Clostridium sporogenes and Clostridium tertium were grown on a prereduced medium containing 1000 mg/kg of sorbate with a pH of 6.0. All of these Clostridia grew well on this medium. Clostridium sporg‘oenes and Clostridium tertium metabolized sorbate prin- cipally to li-hexenoic acid, ll-hexenol and sorbyl alcohol. Clostridium Egrfringens apparently did not metabolize sorbate. In order to confirm that the above metabolites originated from sorbic acid, [U-mCI-labelled sorbic acid was introduced into the media. Data obtained indicate that sorbic acid was the precursor of these metabolic products. To my late father, Saad Al-Mohizea ACKNOWLEDGEMENTS I wish to express my sincere gratitude and appreciation to my major professor, Dr. I.J. Cray, for his encouragement, guidance, and helpful reviews and constructive criticism throughout the course of the laboratory work and during the preparation of this manuscript. Thanks and appreciation go to the other members of my guidance committee; Dr. E. Beneke, Dr. L. Bieber, Dr. H. Momose and Dr. A. Pearson for their careful proof reading of the manuscript and helpful suggestions during the course of the study. I would like to express a special appreciation to Dr. E. Beneke for his generousity in providing some pure cultures, to Dr. P. Markakis for his encouragement and his technical assistance in monitoring the radioactivity, to Y. El-Mabsout for his continuous cooperation, and to S. Cuppett and A. Mandagere for their technical assistance. 1 I owe a special debt of gratitude to my mother and to my brother, Mohammed, for giving their complete support during my academic career which was a source of much needed moral support. Lastly, and not the least, I would like to pay a tribute to my wife for her encouragement, advice, devotion and patience; to my son, Saad, and to my daughter, Norah, for their patience and devotion. TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION REVIEW OF LITERATURE Physical and Chemical Properties of Sorbic Acid Sorbic Acid as an Antimicrobial Agent Nitrites and Sorbates in Food Systems Effect of pH on Antimicrobial Properties of Sorbic Acid Mechanism of Inhibition Physiological Properties of Sorbic Acid Decomposition of Sorbic Acid 0' 1. Chemical degradation of sorbic acid t 2. Microbial degradation of sorbic acid MATERIALS AND MET HODS Microorganisms Screening Tests on Molds, Yeasts and Bacteria Media preparation Incubation conditions Growth evaluation Utilization of Potassium Sorbate as a Carbon Source Growing Cells Studies Non-Growing (Resting) Cells Studies Sample Preparation for GC Analysis Extraction of metabolites Preliminary concentration Esterification and final concentration Headspace Analysis Entrapment of Pentadiene Gas Liquid Chromatographic Analysis of Sorbic Acid Metabolites Apparatus and operation conditions Gas Liquid Chromatography-Mass Spectrometry (CC-MS) Quantitative Analysis of Residual Sorbate by High Performance Liquid Chromatography Preparation of sample High performance liquid chromatography conditions Fate of Labelled Sorbic Acid Thin layer chromatography of neutral metabolites of sorbic acid Further confirmation by GC Analysis iv vi vii 26 26 26 26 27 28 28 28 29 31 31 3Q 34 3“ 35 35 35 36 36 36 37 38 39 RESULTS AND DISCUSSION Microbial Resistance to Sorbic Acid hmflds - Yeasts Bacteria Microbial Degradation of Sorbic Acid Sorbic Acid as a Carbon Source Metabolites of Sorbic Acid Sorbic acid metabolism by molds Sorbic acid metabolusm by yeasts Sorbic acid metabolism by bacteria Fate of Labelled Sorbic Acid SUMMARY AND CONCLUSIONS RECOMMENDATIONS FOR FURTHER STUDIES APPENDICES BIBLIOGRAPHY Page 00 40 HO 03 In 50 55 61 61 81 85 88 8'9 92 94 98 LIST OF TABLES Table . Page 1. Inhibitory concentrations of potassium sorbate for some Ill mold cultures on YPG plates at pH 5. 5, 25°C 2. Inhibitory concentrations of potassium sorbate for some 05 yeasts on YPG plates at pH 5. 5, 25°C 3. Inhibitory concentrations of potassium sorbate for some 08 bacterial cultures on YTG broth at pH 5.5, 25°C II. The effect of the presence or absence of glucose (two 56 percent) on the growth of different mold cultures on yeast extract-peptone-agar media (pH 6. 0) containing 3000 mg/kg of potassium sorbate and incubated at 25°C for six days 5. The effect of the presence or absence of glucose (two 58 percent) on the growth of yeast cultures on yeast extract- peptone-agar media (pH 6. 0) containing 3000 mg/kg of potassium sorbate and incubated at 25°C for six days 6. The effect of the presence or absence of glucose (two 60 percent) on the growth of bacterial cultures on yeast extract-peptone-agar media (pH 6.0) containing 3000 mg/kg of potassium sorbate and incubated at 25°C for six days A. I. Mold cultures used for screening tests 94 A. 2. Yeast cultures used for screening tests 95 A. 3. Bacterial cultures used for screening tests 97 vi LIST OF FIGURES Figure 1. 10. 11. Wieland's scheme for fatty acid oxidation (Melnick et aI. , 19511) Flow chart diagram for the isolation and purification of acidic and neutral metabolites of sorbic acid Utilization of sorbate by growing cells of different fungi cultures cultivated on YPG containing 1000 mg/kg of potassium sorbate, at pH 5.5, and an incubation temperature of 25°C Utilization of sorbate by resting cells of different fungi cultures suspended in a phosphate buffer containing 2.0 percent glucose and 1000 mg/kg of potassium sorbate, at pH 5. 5 and an incubation temperature of 25°C A partial gas chromatogram of the neutral metabolites of sorbate-grown Rhizopus oligosporus Mass spectrum of ll-hexenol (retention time, 13.0 minutes) isolated from a medium containing sorbate and inoculated with Rhizoeus olioosmrus Mass spectrum of ll-hexenol (retention time, 13.2 minutes) isolated from a medium containing sorbate and inoculated with Rhizopus oligoserus Mass spectrum of 2,11—hexadienol (retention time, 17. 5 minutes) isolated from a medium containing sorbate and inoculated with Rhizggus oligospgrus Mass spectrum of 2,0-hexadienol (retention time, 18.1 minutes) isolated from a medium containing sorbate and inoculated with Rhizopus oligoseorus Mass spectrum of 2,0—hexadienol isolated from a medium containing sorbate and inoculated with Rhizggus oligosporus Mass spectrum of methyl ester of ll-hexenoic acid isolated from a medium containing sorbyl alcohol and inoculated with Rhizopus oligosporus vii Page 22 33 53 62 66 67 68 .-e .- __‘\_!A a' ' F. M. -3 Figure 12. A partial gas chromatogram of both neutral and acidic metabolites of sorbaldehyde-grown Mucor 59. 13. Mass spectrum of methyl ester of sorbic acid isolated from a medium containing sorbaldehyde and inoculated with Mucor se. 10. Mass spectrum of 1,3-pentadiene isolated from a medium containing sorbate and inoculated with Penicillium SE. 15. CC chromatogram of both neutral and acidic metabolites of sorbate-grown Geotrichum candidum 16. Mass spectrum of ethyl sorbate isolated from a medium containing sorbate and inoculated with Geotrichum candidum 17. A partial gas chromatogram of both neutral and acidic metabolites of sorbyl alcohol-grown Candida Iipolytica 18. A partial gas chromatogram of both neutral and acidic metabolites of sorbate—grown Clostridium serggenes viii Page 72 73 77 79 80 82 86 INTRODUCTION Sorbic acid and its potassium salt are very popular food additives and are used routinely by certain elements of the food industry. Sor- bates have been cleared for use and were approved as food preservatives by the United States Food and Drug Administration almost a quarter of a century ago (Anonymous, 1978). Sorbates are utilized in more than 70 food products designated by standards of identity (Anonymous, 1978) . Sorbates are officially listed as GRAS food additives by the Food and Drug Administration, which means that they are "generally recognized as safe." Consequently, sorbates may be used freely, provided that the quantities do not exceed the amount reasonably required to achieve the desired function (inhibition of microorganisms), and that the application is also in conformance with state or local regulations. With the growing concern over the use of nitrites in cured meat systems, and with the possibility that nitrite may be partially replaced by sorbate in meat systems, sorbates are attracting considerable atten- tion. As a result, sorbates are undergoing intensive studies to justify the usefulness of their new applications. However, there are unexplained problems associated with the application of sorbic acid and its potassium salt in the food industry. One major problem is the loss or degradation of sorbic acid over a relatively short period of time. It is well established that there are some sorbic acid-resistant microorganisms which can grow in the presence of this acid, and, in some cases, metabolize it. As a 2 consequence of this partial or total degradation of sorbic acid, the protection of foods against food spoilage microorganisms is decreased. Hence, microbial degradation of sorbic acid has been cited in many cases for the loss of this preservation activity. It is the scope of the present study to investigate some aspects of the microbial degradation of sorbic acid, using some representative microbes. In order to accomplish the main goal, the following objec- tives were considered: 1) To investigate microbial resistance to sorbic acid. 2) To ascertain which microorganisms utilize sorbic acid as a sole source of carbon. 3) To study the rate of sorbic acid decomposition, in cultural media or buffer systems by growing and [or resting cells and to identify the metabolites thus formed. II) To follow the fate of labelled sorbic acid to confirm that the metabolites originate from sorbic acid. 5) Finally, to attempt to categorize microorganisms according to their behavior towards sorbic acid. REVIEW OF LITERATURE Physical and Chemical Properties of Sorbic Acid Sorbic acid is a naturally occurring food acid and was discovered in 1859 by Hoffman, a German chemist (Luck, 1976) . He treated rowan berry oil, an acid distillation product from the juice of unripe rowan berries with strong alkali, and named the new substance "sorbic acid" after the scientific name of the mountain ash, Sorbus aucuparia Linne, the parent plant. It is now prepared synthetically by many methods (Luck, 1976). Sorbic acid (trans, trans-2, ll—hexadienoic acid, CH -CH=CH-CH=CH-COOH) 3 is an cc, B-unsaturated, aliphatic, straight-chain monocarboxylic acid (Anonymous, 1978). It is a white, crystalline compound and has a very low solubility in water, especially at room temperature where the solu- bility is only about 0.16 percent. However, it dissolves more easily in hot water, ethanol, or diethyl ether. In boiling water, a solution of 3.8 percent sorbic acid can be obtained (Gershenfeld, 1968). Fortunately, solutions of more than 50 percent can be prepared from its potassium salt (potassium sorbate) in cold water. It has been shown that the pH of the medium greatly affects the solubility (Pfizer, as cited by Bradley, 1960) , especially potassium sorbate which tends to hydrolyze to sorbic acid at low pH. Other factors such as high concentrations of glucose, sucrose or sodium chloride, have also been shown to affect the solubility of sorbic acid in water (Gershenfeld, 1968) . u Sorbic acid has a melting point of 1311. 5°C and starts to decompose at 228°C. However, above 60°C , the acid begins to sublime (Luck, 1976) . Muller (1939) and Gooding (19110) independently discovered that sorbic acid has an antimicrobial effect (Luck, 1976). Later, Gooding (19115) was granted the original patent on the use of sorbic acid as a food preservative. This finding stimulated further studies on the inhibitory effect of sorbic acid against microorganisms, physiological and chemical properties of this acid, and the possible mechanism of this inhibition. Sorbic Acid as an Antimicrobial Agent Phillips and Mundt (1950) observed that sorbic acid at a concentration of 1000 mg [kg effectively inhibited film yeasts in cucumber fermentations without noticeable inhibition of the desirable lactic acid fermentation. Vaughn and Emard (1951) reported that sorbic acid selectively favored the growth of Lactobacillus and Leuconostoc strains and inhibited all of__the other test cultures of bacteria, molds, and yeasts. Emard and Vaughn (1952) tested 229 different cultures of bacteria, molds, yeasts, and actinomycetes on two cultural media containing sorbic acid. They found that all of the catalase positive cultures were inhibited to such a degree that growth was not detected in the test media after seven days, while the. catalase negative cultures grew without noticeable inhibition. The selective inhibitory effect of sorbic acid was thus suggested. The pH's of the media were not specified. However, when potassium dihydrogen phosphate was added to the media, the selectivity of the media was reduced, i.e. both catalase negative as well as catalase positive cultures were inhibited. In an attempt to determine the effect of pH on the 5 selectivity of sorbic acid, the same authors concluded that the pH of the media containing sorbic acid should be adjusted so that, after sterilization, the pH is between 5 and 5.5. In an attempt to develop a selective medium for the isolation of the catalase negative bacteria, York and Vaughn (1954a) reported that, at a concentration of 1200 mg [kg of sorbic acid, the ability of the spores of the different cultures of Clostridia to germinate in seven days was markedly decreased as the pH value of the medium was reduced from 5. 8 to 5.0. Consequently, they recommended that, if the sorbic acid medium is to be used for enrichment and isolation of Clostridia, the pH range must be increased to at least six or above. However, at this higher range, sorbic acid is not as effective as in the recommended range of pH 5 to 5. 5. In another study, the same investigators (York and Vaughn, 1950b) noticed a marked resistance to sorbic acid by Clostridium parabotulinum type A in the acid range (pH 4.8- 5.0) . They observed that Clostridium parabotulinum type A was not appreciably affected by the presence of 1200 mg/kg of sorbic acid in liver infusion broth at this pH. Hansen and Appleman (1955) initiated a study to determine whether sorbic acid, caproic acid and propionic acid are growth stimulants for Clostridia. They found that, at pH 6.7 and on two media, liver-infusion broth and glucose-yeast extract broth, these acids do not inhibit or stimulate the growth of Clostridium parabotulinum type A and B, and Clostridium sporogenes. They recommended sorbic acid for its anti- mycotic activity, but concluded it was not suitable to prevent growth of'catalase negative organisms. 6 Costilow (1957) reported that sorbic acid-treatments of cucumber fermentations reduced the occurrence of bloater-type spoilage; it was significantly reduced by the addition of sorbic acid to the brine in con- centrations as low as 200 mglkg. It has been found that lactic acid fermentation was inhibited by 1000 mg [kg of sorbic acid when the initial brine strengths ranged from 20° to 00° salometer. Concentrations of sorbic acid as low as 500 mglkg were sufficient to inhibit the fermentation process when the initial brine strength was 00° salometer (Costilow et al., 1957). Klis et al. (158) examined the inhibitory effect of sorbic acid on cultural media using 16 common food-spoilage fungi. They found that all of the fungi tested except Aspergillus niger and Zygosaccharomyces barkeri, were completely inhibited by 500 mglkg of sorbic acid. These two fungi were not inhibited by 1000 mglkg of sorbic acid. A study was conducted by Weaver et al. (1957) to determine whether sorbate would be as effective in preserving commercial apple cider as benzoate which imparts an undesirable flavor to cider. They found that sodium sorbate (500 mglkg) controlled molds and yeasts, but was not effective against bacteria at high storage temperatures. Similar results were obtained by Ferguson and Powrie (1957) who pointed out that the spoilage was due to Acetobacter. Bradley et al. (1962) demonstrated that the presence of potassium sorbate usually caused a "definite Iag'l in the growth curve of organisms commonly associated with cottage cheese spoilage, resulting in the extension of the shelf life of the cheese by one to eight days dependent upon the concentration of the sorbate. 7 From the previous studies on the antimicrobial activity of sorbic acid, the following conclusions can be made: 1. Sorbic acid is effective only against catalase positive bacteria and fungi. 2. The inhibitory effect of sorbic acid is very pH dependent. 3. Sorbic acid is considered ideal for the pickling industry since it inhibits only undesirable microorganisms. Nitrites and Sorbates in Food Systems Over the past two decades, there has been much concern over the use of nitrite in cured meats. This concern has originated from reports implicating nitrite as a precursor of carcinogenic N-nitrosamines in cured meats, and in particular, fried bacon. Excellent papers and comprehensive reviews on this subject have been published (Fiddler, 1975; Mirvish, 1975; Shank, 1975; Wogan and Tannenbaum, 1975; Crosby and Sawyer, 1976; Gray, 1976; Mirvish, 1977; Gray and Randall, 1979; Sofos et al., 1979a). It is worthy to mention that nitrite itself has been incriminated as a possible direct carcinogen in recent reports (USDA, 1978; Newberne, 1979). It is for this reason that considerable effort has been directed towards finding a potential substitute for nitrite in meat systems. Sorbic acid and its potassium salt have attracted the most attention as a possible alternative to nitrite, or adjunct to nitrite, i.e. reducing the ingoing nitrite to the lowest level which permits the develop- ment of a satisfactory color and flavor, while adding sorbate at a concen- tration of 2600 mglkg to provide adequate protection against Clostridium botulinum . 8 As a result of studies prior to 1974 (Phillips and Mundt, 1950; Vaughn and Emard, 1951; Emard and Vaughn, 1952; York and Vaughn, 1950a; 1955; Costilow et al., 1955; Hansen and Appleman, 1955; Costilow, 1957; Costilow et al. , 1957) , there was no attempt to utilize potassium sorbate in meat systems, except when the presence of fungi were expected. For this reason, potassium sorbate was approved for use in dry stuffed sausages to retard mold growth on the surface of the sausages by dripping the casings in a 2. 5 percent solution of potassium sorbate (Sofos and Busta, 1979). More recently, Tompkin et al. (19711) conducted a study where they added 1000 mglkg of potassium sorbate to skinless precooked, uncured sausage links which were then inoculated with Staphllococcus aureus, Salmonellae and Clostridium botulinum. They found that potassium sorbate not only delayed the growth of the normal spoilage flora, but also the growth of Staphllococcus aureus and Salmonellae was remarkedly retarded. The growth of Clostridium botulinum was considerably reduced and the production of toxin was delayed for six days. The pH of the system was high enough to exclude the effect of hydrogen ion concentration. These results were contrary to those of previous investigations with regard to the inhibitory effect of sorbic acid on Clostridia. This new finding stimulated further studies. Consequently, several reports have been published recently which reveal that the combination of reduced levels of nitrite (e.g. 00 mglkg) and 2600 mglkg of potassium sorbate gave the same level of protection against Clostridium botulinum and the same desired organoleptic characteristics as that achieved by nitrite in the regularly permitted amounts. Tanaka et al. (1977) demonstrated that in frank- furters, potassium sorbate (2700 mglkg) produced an antibotulinal effect 9 similar to that of 100 mglkg of nitrite. Robach (1980) reported that the effectiveness of potassium sorbate in preventing or delaying the outgrowth of Clostridium sporogenes PA3679 was increased by adding one to five percent sodium chloride. Robach et al. (1980c) also studied the effect of sorbates on microbiological growth in cooked turkey pro- ducts. They reported that 1000 to 2000 mg] kg of potassium sorbate depressed microbiological growth and delayed the formation of the pink color and off odors in uncured, cooked, vacuum packed poultry products, without affecting the organoleptic quality. Robach et al. (1980b) demon- strated that bacon processed with 40 mglkg of sodium nitrite and 2600 mglkg of potassium sorbate contained an average of 8. 7 ug/kg of N-nitrosopyrrolidine (NPYR) compared to 11.11 ug/kg and 28.1 ug/kg of NPYR for bacon made without and with 120 mglkg of nitrite, respectively. Similar results were obtained by Ivey et al. (1978) , Pierson et al. (1979) and Price and Stevenson (1979). It has been shown by many workers that nitrite-sorbate combinations are more effective in retarding Clostridium botulinum growth and toxin production than either nitrite or sorbate when used individually (Tanaka et al., 1977; Ivey and Robach, 1978; Ivey et al., 1978; Robach et al., 1978) . However, the mechanism by which nitrite-sorbate combination interacts to inhibit the growth and toxin production by Clostridium botulinum is not known. Sofos et al. “97%) suggested the possibility of an additive nitrite-sorbic acid effect and the production of a more inhibitory' substance(s) through reaction of nitrite and sorbic acid. It has been reported that the addition of potassium sorbate to cured meat delayed the disappearance of nitrite (Sofos et al. , 1979a,b) . 10 Tompkin et al. (1978) pointed out that the lower the residual nitrite at the time of incubation, the faster the rate of swelling indicating the involvement of nitrite in the inhibition of growth of Clostridium botulinum. Sofos et al. (1979a) suggested that sorbic acid depresses spore germination to an insignificant level, while nitrite retards the growth of germinated spores. Namiki and Kada (1975) reported the isolation of ethylnitrolic acid from the reaction of sorbic acid with sodium nitrite. This compound was considered by the authors to be a strong antibacterial agent. Some controversies have been created when Kada (1970) reported that a mixture of sorbic acid and nitrite (130 g/kg) in aqueous medium, when heated, gave a positive result in the Bacillus subtilis recombination assay. This indicates the possible formation of DNA-damaging substance(s) in this reaction. Hayatsu et al. (1975) and Namiki and Kada (1975) con- firmed the findings of Kada (1974). Ethylnitrolic acid was identified as the mutagenic principle arising from the reaction of nitrite and sorbic acid (Namiki and Kada, 1975). More recent studies have failed to demonstrate the mutagenicity of ethylnitrolic acid in both model system and food system (Difate, 1978; Robach et al., 1980a). Difate (1978) attributed the apparent mutagenicity of ethylnitrolic acid to possible free'nitrite contamination. Tanaka et al. (1978) reported that, when lower levels of nitrite (13,000 mglkg) reacted with sorbic acid (2000 mglkg) at different pH values (1-5) , reaction product(s) were produced which were negative in the Ames Salmonella assay and Bacillus subtilis recombination assay. Furthermore, they found that the presence of sorbic acid inhibited the formation of N-nitrosamines in i_n_ vitro tests under a wide variety of conditions . 11 Effect of pH on Antimicrobial Properties of Sorbic Acid Sorbic acid is one of a group of short chain organic acids which, together with their salts, have been shown to exhibit antimicrobial pro- perties (Bell et al., 1959). Hoffman et al. (139) studied the antimicrobial properties of a number of normal saturated and unsaturated fatty acids containing from one to 10 carbon atoms, and used as test organisms, a mixed culture of common food spoilage molds. Their study revealed that the effectiveness of the acids as fungistatic agents increased with chain length and concentration, and varied with the pH. Rahn and Conn (191111) postulated that preservative acids such as benzoic acid are effec- tive only when present in the undissociated state, i.e. when the pH is low enough to prevent dissociation. Samson et al. (1955) observed that the effectiveness of fatty acids as fungistats was dependent on the pH of the system. The inhibitory effect increased with decreasing pH because the cell was permeable to the acid in the undissociated form only. Beneke and Fabian (1955) observed the influence of both sorbic acid concentration and pH of the media on the inhibition of a number of fungi isolated from tomato and strawberry fruits. Bell et al. (1959) reported that the toxic action of sorbic acid was found to be directly related to the concentration of undissociated acid. They concluded that sorbic acid would be more effective as an antimicrobial agent when used at a low pH which would permit more of the undissociated form to be present in solution. Raevori (1976) stated that the free acid enters the bacterial cell and inhibits several enzyme systems. As previously discussed, it is apparent that the undissociated form of the acid plays the most important role as the antimicrobial agent. The 12 reason is that the microbial cell carries a net negative charge, so the uncharged form (undissociated) will enter the cell, while the charged form (dissociated) which carries a negative charge will be repelled. A—+H+«———-—+AH4—————+AH «————..A’+H+ Outside Cell Inside Cell Membrane Cell Once the acid is inside the cell, the mechanism of inhibition is not quite clear, although many studies have been conducted to elucidate the nature of this inhibition. Mechanism of Inhibition The inhibition of various enzymatic reactions by sorbic acid has been suggested as the possible mechanism by which the growth of microorganism is inhibited (Woodford and Adams, 1970). Melnick et al. (1950) suggested that the accumulation of c, B-unsaturated fatty acids can inhibit the dehydrogenase system in molds. Sorbic acid already exists in this form, so the dehydrogenase enzymes which are vital for cell metabolism are inhibited by sorbic acid. Furthermore, Melnick et al. (1950) reported that when molds are present in high numbers, sorbic acid is metabolized without inhibition of the dehydrogenase enzymes. However, when molds are present in low numbers, sorbic acid remains in the original effective form. Azukas et al. (1961) demonstrated that the fermentation of glucose by resting cells of baker's yeast was greatly inhibited by sorbic acid and that the inhibition was very much pH dependent. However, when the same workers used a cell—free extract of the same yeast, they found no significant difference in the percent inhibition between pH five and six. 13 York and Vaughn (1955b) reported that fumarase activity is suppressed by sorbic acid. Whitaker (1959) demonstrated that the sulfhydryl-con- taining enzymes, ficin and alcohol dehydrogenase, were inhibited by sorbic acid, and suggested that sorbic acid and the other or, B-unsaturated acids were non-specifically effective against sulfhydryl containing enzymes, with which the acids form complexes. York and Vaughn (1964) found that the sulfhydryl enzymes, fumarase, aspartase, and succinic dehydrogenase were inhibited by sorbic acid. They also observed the loss of activity of sorbic atid after reacting with cysteine which led them to conclude that a thiol addition occurred, the mechanism which was believed to be involved in the action against sulfhydryl enzymes. Rehm (1967) reported that sorbic acid strongly inhibited a number of dehydrogenases including malate dehydrogenase, a-ketoglutarate dehydrogenase and succinate dehydrogenase. Pellaroni and Pritz (1960) suggested that the reduction in respiration rate of Saccharomyces cerevisiae var ellipsoideus was the result of the formation of sorbyl coenzyme A. In an attempt to elucidate the mechanism by which sorbate inhibits baker's yeast, Harada et al. (1968) suggested that sorbate competitively combines with coenzyme A and acetate and con- sequently inhibits the enzymatic reaction involving coenzyme A. It has been reported that lipophilic acids, including sorbic acid, uncouple both substrate transport and oxidative phosphorylation from the electron transport system (Freese et al. , I973) . More recently, Przybylski and Bullerman (1980) observed the depletion of the ATP level of the conidia of Aspergillus parasiticus upon exposure to sorbic acid. They suggested that the depletion of ATP levels of conidia may be a possible mechanism for the action of sorbic acid in fungal inhibition. Ill Physiolggical Properties of Sorbic Acid Luck (1976) stated that "as an aliphatic carboxylic acid, the con- sideration of the structure of sorbic acid leads to the expectation that sorbic acid is utilized in the body similarly to other fatty acids." Physiological studies on sorbic acid using laboratory animals proved that the above hypothesis is correct. Witter et al. (1950) reported that sorbic acid, as well as caproic acid, is quantitatively transformed to acetone bodies, since both, on a molar basis, yield 1. 5 moles of acetoacetate when treated with rat liver homogenate or washed liver particulate. A comprehensive study on the metabolism of sorbic acid has been carried out by Deuel et al. (1950a) using rats and dogs in their study. Their findings can be summarized as follows: 1. The gain—in-weight of rats receiving diets containing up to eight percent of sorbic acid was similar to that of rats on a basal diet free of sorbic acid. 2. There was no evidence of abnormality in liver or kidney weights or pathological conditions in other tissues of rats receiving four percent of sorbate. Rats on the eight percent sorbic acid diet exhibited a slight but "statistically significant" increase in liver weight. 3. Sorbic acid did not act as an antimetabolite for essential fatty acids in rats. 11. Sorbic acid is utilized by the animals as a source of calories. 5. Sorbic acid is much less toxic than sodium benzoate as indicated by the ratios of LD 50 (sorbate) :LD50 (benzoate) , 1. 7—1. 9. Ill PhysioLogical Properties of Sorbic Acid Luck (1976) stated that "as an aliphatic carboxylic acid, the con— sideration of the structure of sorbic acid leads to the expectation that sorbic acid is utilized in the body similarly to other fatty acids." Physiological studies on sorbic acid using laboratory animals proved that the above hypothesis is correct. Witter et al. (1950) reported that sorbic acid, as well as caproic acid, is quantitatively transformed to acetone bodies, since both, on a molar basis, yield 1.5 moles of acetoacetate when treated with rat liver homogenate or washed liver particulate. A comprehensive study on the metabolism of sorbic acid has been carried out by Deuel et al. (1950a) using rats and dogs in their study. Their findings can be summarized as follows: 1. The gain-in—weight of rats receiving diets containing up to eight percent of sorbic acid was similar to that of rats on a basal diet free of sorbic acid. 2. There was no evidence of abnormality in liver or kidney weights or pathological conditions in other tissues of rats receiving four percent of sorbate. Rats on the eight percent sorbic acid diet exhibited a slight but "statistically significant“ increase in liver weight. 3. Sorbic acid did not act as an antimetabolite for essential fatty acids in rats. ll. Sorbic acid is utilized by the animals as a source of calories. 5. Sorbic acid is much less toxic than sodium benzoate as indicated by the ratios of LD 50 (sorbate) :LDso (benzoate) , 1.7—1.9. 15 Deuel et al. (1955b) conducted another study with rats to elucidate how sorbic acid is metabolized in the presence and absence of glucose (a glucogenic substance). They found that the intermediate metabolism of sorbic acid is identical to that of the normally occurring fatty acids, caproic and butyric. They concluded that, under normal conditions of alimentation, sorbic acid is completely oxidized to CO2 and H20 and thus yields its potential energy as calories. Stavkuv and Petrova (1960) reported that When sorbic acid was administered daily at levels of 0. 5 to 1. 0 percent over a prolonged period, there was no sign of toxicity to the test animals. However, doses of five to ten percent of sorbic acid when taken daily for four months produced some signs of toxicity. Westos (1960) followed the fate of labelled sorbic acid during metabolism by mice. He found that, of the sorbic acid given to mice during a four day period, 81 percent of the radioactivity was released as 1"C02 and about four percent in the urine, mainly as muconic acid (HOOC-CH=CH-CH=CH-COOH) . Gaunt et al. (1975) conducted a long term toxicity study of sorbic acid and found no signs of toxicity or carcinogenicity when sorbic acid was fed to rats at levels up to ten percent of the diet. This represented an intake of approximately five g/kg/day. They concluded that, on the basis of the 100x safety factor, the daily intake would be 50 mglkg/day, a value in excess of the unconditional level established by the joint FAO/WHO Expert Committee on Food Additives (FAO, 1967). In a similar study on the toxicity of sorbic acid on mice, Hindy et al. (1976) reported that when sorbic acid was fed at levels of up to ten percent of the diet for 80 weeks, there was no evidence of carcinogenicity. They observed no adverse effects although a slight enlargement of the kidney and a 16 reduction in body weight gain were detected in mice fed with 5.0 to 10. 0 percent sorbic acid. These changes were not observed when sorbic acid was given at a concentration of one percent. Lishund (1969) , cited by Gaunt et al. (1975) calculated that if sorbic acid were used at the "technically acceptable levels" in cheese, preserves, dried fruits, nuts, bread, cakes and pastries, pickles and sauces, wine, soft drinks, and fruit, the average daily intake would be 293 mg. This is well below the unconditional acceptable daily intake of 12. 5 mglkg (750 mg/day for a 60-kg adult male) established by the joint FAO/WHO Expert Committee on Food Additives (FAO, 1967). Decomposition of Sorbic Acid Thechcomposition of sorbic acid has been reported earlier by several workers (Melnick and Luckmann, 195113 and b; Melnick et al. , 195ll; Costilow et al. , 1957; Alderton and Lewis, 1958). In spite of the fact that the loss may become severe in some cases, considerably little is known about the mechanism of the decomposition of sorbic acid in biological systems. Losses may be classified into two types: 1) loss due to chemical degradation, and 2) loss due to biological degradation (mainly microbial). 1. Chemical'degadation of sorbic acid Although sorbic acid and its salts are stable in the dry pure form, their solutions are known .to be relatively unstable (McCarthy and Eagles, 1976) . It has been reported that 35 to 65 percent of the original sorbic acid in the pickle brine solution was lost during storage (Costilow et al., 1957). Similar losses have also been reported by Alderton and Lewis 17 (1958). However, Costilow (personal communication with Alderton and Lewis, 1958) strongly suggested the possibility of chemical degradation of sorbic acid during cucumber fermentation. McCarthy et al. (1973) investigated the stability of a sorbic acid solution (pH 11.2) when stored in a variety of plastic and brown glass bottles which had previously been found to transmit no light between the wavelengths of 220 to 1150 nm and only 5. 5 percent transmission at 500 nm. They reported that up to 50 percent of the original sorbic acid (1000 mglkg) was lost in a period of twelve weeks. They pointed out that the temperature at which the bottles were stored greatly affected the rate of sorbic acid degradation. Furthermore, they found that in most cases, the decay was first-order, whereas in a few cases, the decay was first-order for a period of time followed by a rapid drop suggesting catalytic decomposition. The latter was associated with high temperature (50°C). Recently, Arya et al. (1978) demonstrated the effect of some food ingredients on the rate of sorbic acid degradation in aqueous solutions. They reported that sugar, salt, glycerol, and some metallic ions (CuH, Fe++ and Ni”) retarded the rate of degradation of sorbic acid. Further- more, they,"stated that all of the amino acids tested, other than histidine, accelerated the degradation. Thus, they concluded that the autoxidative degradation of sorbic acid differs from the autoxidation of unsaturated fatty acids. More recently, Saxby et al. (1979) conducted a comprehensive study on the chemical degradation of sorbic acid in a model system where some food ingredients were examined for their effects on the rate of degradation. 18 They demonstrated that citric acid and ascorbic acid catalyzed to some extent the degradation of sorbic acid. Their study also demonstrated that light strongly influences the rate of degradation at pH 2.6, but not at pH 1.2 or higher values of pH (11.3 or 5.11). Furthermore, Saxby et al. (1979) detected the presence of or-angelicalactone, albeit in rela- tively low concentrations, i.e. , 0.1 percent of the total decomposed sorbic acid. The fate of the remaining sorbic acid was not determined. /"\ H H low pH I l - SO2 CH COOH CH O CH 3 3 Sorbic acid oc-angelicalactone McCarthy and Eagles (1976) reported that sorbic acid in solutions is oxygen labile. However, no further explanation as to how oxygen affects the stability of sorbic acid was given. In an attempt to explain the loss of sorbic acid from sorbic acid-pre- served high moisture prunes, Bolin et al. (1980) demonstrated that the extent of sorbic acid loss was a function of moisture and temperature. They also concluded that microorganisms did not play any role in the loss. However, no attempts were made to isolate and identify the degradation products. 2. Microbial (Bladation of sorbic acid In spite of the fact that sorbic acid is commonly used as an antimicrobial agent in food systems, there are certain microorganisms which can grow in 19 the presence of relatively high concentration of sorbic acid or its salts. Degradation of sorbic acid by microorganisms has been reported by many investigators (Melnick et al., 1950; York and Vaughn, 1950b; Luckas, 1960; Rehm et al., 1960). Melnick and Luckmann (1950a) noticed that sorbic acid disappeared partially or totally from sorbic acid-preserved cheese when stored at 05°F for six weeks. Neither autoxidation nor sublimation were deemed responsible for this loss (Melnick and Luckmann, 1950b). Melnick et al. (1950) further demonstrated that the loss from the sorbic acid-treated cheese wrapper was due to metabolic oxidation of sorbic acid by molds. Perry and Lawrance (1959) stated that sorbic acid would not suppress heavy contaminations, but would effectively retard small contaminations of microorganisms. Sorbic acid may be used by microorganisms as a source for carbon and energy. Clostridium parabotulinum type A, Clostridium sporoggnes, Clostridium acetobutylicum and Clostridium thermosaccharolyticum have been shown to utilize sorbic acid as a carbon source (York and Vaughn, 1950a). Furthermore, York and Vaughn (1950b) reported that all the cultures of Clostridium parabotulinum type A and B tested utilized 1200 mg/kg to 10000 mglkg of sorbic acid in a basal medium at pH 5.6. However, not all of the microorganisms which can tolerate sorbic acid can use it as a carbon source. Many of the data reported regarding the loss of sorbic acid have been associated with products with high initial microbial counts or with cases where microbial action is involved in the processing, e.g. pickles (Deak and Novak, 1970; Deak et al. , 1972) . However, to rule out the possibility of the utilization of sorbic acid in sorbic acid-treated pickle brine by microorganisms commonly present during the fermentation process, Costilow 20 et al. (1955) tested a number of lactic acid bacteria and a sorbic acid-tolerant yeast, Torulopsis holmii. They found that neither the lactic acid bacteria nor the heavy inoculum of the sorbic acid-tolerant yeast, could grow on a basal medium containing sorbic acid as the sole source of carbon. Rehm et al. (1960) found that Aspergillus niger and Pseudomonas fluorescens destroy and metabolize sorbic acid in sublethal concentrations. Luckas (1963) reported that Aspergillus nigger can decompose sorbic acid and that mycelia growth was inhibited by 1500 mglkg of sorbic acid. Little is known about the mechanisms by which microorganisms degrade sorbic acid. However, because sorbic acid shares a lot of physical and chemical characteristics with fatty acids, it has been thought that it is metabolized in a similar way to that of fatty acids. The degradation of fatty acids by microorganisms have prompted many investigations over the past four decades. Many of these studies have been devoted to the elucidation of the mechanism of microbial degradation of fatty acids. Thaler and Geist (1939a, b), Mukherjee (1951, 19523, b), and Melnick et al. (1950) have contributed to our knowledge in this field. Mukherjee, in a series of papers, established a "definite reaction mechanism" for the degradation of fatty acids by molds, using representative species of Penicillium and Aspergillus. Melnick et al. (1950) investigated Wieland's scheme, i.e. , fatty acid oxidation of saturated fatty acids proceeds through beta-oxidation involving dehydrogenation at the 0:, B-position, hydration and dehydrogenation. The authors suggested the scheme, outlined in Figure 1, as a possible mechanism for the degradation of sorbic acid by molds. 21 Figure 1. Wieland's scheme for fatty acid oxidation (Melnick et al. , 1950) 2 2 0N: + N8 :oou-Nzo-ou-Nzo a :ooo-Nzu-xu-xoxu-Nzu :o: + a .N: + N8 + :ooo-IUnzu-Nzu ~c+ m :ooo-qu-ou-xouzu- :u :N- # :oou-Nzu-zozu-zouzu-Nzu :o... + :coo-zonzu-szu-Nzu .22 29.3 303.998. new moucnumgm u_comou:_m Co eczema-S. or: E coszxo Eon 30v. catacomoLUan Eon >xoLp>sum 88$ 8:23: 3:92.28 coszchca-m .8 cczncomocn>noc 30200.3 pco can Bum 3am“. cotmExO Eon Sov— catacomeaicoo Eon >xcLU>L-m c0396.»: cczmeaummcaam .8 cozmcomoLD>soo .22 39.88 22 .58”. fig 0N: + N8 :cou-Nzu-ou- 1N-» :oou-Nzu-zczu- :0: L zoom-:unzu- IN- 0N: + N8 + :oou-Nzu-Nzo- Na + a N N N :00? :u-cu- :o- :u- :N- a N N N :08. :o-zczo- :o- :u- :o: + a Nzu mIU Nzu Nzu Nzu Nzo 1000-:9: o-Nzo-NIU-Nzo Nr-a :ooo-Nzu-Nzu-Nzo-Nzo- Nzo .22 3983. 22 is“. 23 This scheme results in a fatty acid containing two carbon atoms less than the original fatty acid, ketones, carbon dioxide and water as end products of one cycle. Thaler and Geist (1939a) noticed that the presence of glucose or any glucogenic material in the test medium prevents ketone formation. Melnick et al. (1950) observed that ketone formation was associated with sorbic acid degradation. The same authors reported that in the presence of added carbohydrates such as glucose, metabolic degrada- tion of fatty acids, as illustrated by sorbic acid, proceeds at a more rapid rate, but without an accumulation of ketones. Consequently, the fatty acids are rapidly and completely metabolized to carbon dioxide and water as end products. This mechanism appears to be similar to that of the mammalian system (Deuel et al. , 1950a, b). Luckas (1963) demonstrated that sorbic acid was taken up by the cells and metabolized mainly to carbon dioxide and water and partly to methyl ketones. There was no mention as to whether sorbic acid was used as the sole source of carbon or with a glucogenic substance such as glucose. Rehm et al. (1960) stated that the destruction of sorbic acid is similar to that of fatty acids. More recently, it has been shown that a M s2; can promote a biological hydrogenation of sorbic acid at the second carbon, before or after the reduction of the carboxyl group, resulting in the formation of 0-hexenol (Kurogochi et al., 1970). Furthermore, the same authors reported that a Geotrichum s3; metabolized sorbic acid partly into 0—hexenoic acid and ethyl sorbate. No explanation was given for the mechaniSm of formation or the possible intermediate metabolites. However, in another study, Kurogochiet al. (1975) extensively studied the fungal metabolites of sorbic acid and analogous compounds in an attempt to elucidate the mechanism of sorbic acid degradation. They identified a new 20 intermediate metabolite, sorbyl alcohol (2,0-hexadienol) but only in the early stages (one to two days) of growth of the fungus. However, when the mold, M §&, was grown in the presence of sorbyl alcohol or 0-hexenoic acid, the two substrates were rapidly reduced to 0-hexenol. Hence, it was concluded that the fungus, m 322! converted sorbic acid into 0-hexenol in two independent steps: 1) Reduction of the carboxylic group CH3-CH=CH-CH=CH-COOH ————+ CHB—CH=CH-CH=CH-CHZOH Sorbic acid Sorbyl alcohol 2) Hydrogenation of one olefinic bond in 2-alkenol CH3-CH=CH-CH=CH-CHZOH + CH3-CH=CH-CH2-CH2-CHZOH Sorbyl alcohol 0-Hexenol Tahara et al. (1977) identified an intermediate metabolite, sorbaldehyde (2,0-hexadienal) in the early stages of growth of Mucor s2; Tahara et al. (1979) used the cell free system of Mucor griseo-ganus to clarify the reduction of sorbic acid to 0-hexenol. They suggested the following scheme: OOH Sorbic acid ‘ Sorbaldehyde So rbyl alcohol WHO WHZOH A "f 4-Hexenal 4-Hexenol 25 Tahara et al. (1977) examined 26 species of molds including twelve species of Mucor for their ability to reduce sorbic acid to 0-hexenol. They found that most of the Mucor and Rhizopus species tested were capable of reducing sorbic acid to hexenol. Marth et al. (1966) indicated that a hydrocarbon-like odor appeared in sorbate-fortified mold cultures isolated from cheeses fortified with potassium sorbate. They identified that volatile substance as 1,3-pentadiene. They postulated the fbllowing mechanism: Potassium sorbate+tartaric acid —+Potassium tartrate+sorbic acid decarboxylation Sorbic acid ; 1, 3-pentadiene + CO2 Furthermore, the same authors pointed out t'rat potassium sorbate was degraded by some species of Penicillium, especially if it was in a highly nutritious substrate. More recently, Crowell and Guymon (1975) reported the isolation of ethyl sorbate, 2, 0-hexadien-1-ol , 1-ethoxyhexadiene, 3,5—hexadien—2—ol and 2-ethoxyhexa-3,5—diene from a sorbic acid-containing wine which was spoiled by lactic acid bacteria. They pointed out that the latter compound has an intense odor and is responsible for the geranium-like off—odor associated with spoilage in sorbate-containing wines. MATERIALS AND METHODS Mic roorgani sms All of the microorganisms tested, with the exception of a few molds, were obtained from the culture collection of the Food Microbiology Laboratory, Department of Food Science and Human Nutrition, Michigan State University. Some of the molds were obtained from the Department of Botany and Plant Pathology, Michigan State University. Some food and air-borne molds were isolated from different sources and were included in the preliminary studies. Only those molds which showed a marked resistance to sorbic acid were identified as to genus using the scheme described by Beneke and Stevenson (1978) . ScreenirllTests on Molds, Yeasts and Bacteria Media preparation Screening tests were carried out by growing selected molds (Table A. 1) and yeasts (Table A. 2) on yeast extract—peptone-glucose (YPG) (yeast extract, Difco Laboratories, Detroit, MI, 0.05 percent; peptone, Difco, 0. 5 percent; and glucose, 2. 0 percent) as the basal medium for mblds and yeasts. Sorbic acid was added as its potassium salt (Sigma Chemical Co., St. Louis, MO) in the appropriate concentration (500-5000 mglkg). For the solid medium, two percent of Bacto agar (Difco) was added. Yeasts were tested on both solid and liquid media while only the solid medium was used for molds. 26 27 Two media, yeast extract-tryptone-glucose (yeast extract, Difco, 0. 2 percent; tryptone, Difco, 0. 5 percent; and glucose, 2.0 percent) and trypticase soy broth (TSB) , which was prepared according to the Difco manual (Difco, 1963) with the exception that the pH was adjusted to 5. 5 were used for growing bacteria. The latter medium was mainly used for the fastidious bacteria. Sorbic acid, as its potassium salt, was added in the appropriate concentration (500- 5000 mglkg) . Both solid and liquid media were used for cultivating selected bacteria (Table A. 3). The initial pH was adjusted by adding 1N HCI or 1N NaOH to give the required pH (5. 5) after sterilization and addition of sorbic acid. The pH was recorded by a Beckman Research pH meter (Beckman Instruments, Inc. , Fullerton, California). The medium was autoclaved at 15 psi for 15 minutes, cooled to about 60°C, and potassium sorbate was added at the required concentration under aseptic conditions. The medium was poured into petri dishes, left to solidify at room temperature, and then inoculated with the appropriate microorganisms. The liquid media were prepared as previously described, except that agar was not added and that potassium sorbate was added before steriliza- tion. The complete media were dispensed (five ml) into screw cap culture tubes (15 x 150 mm) and the tubes sterilized as previously described. Incubation conditions Unless otherwise stated, molds and yeasts (both plates and tubes) were incubated at room temperature (25°C). For bacteria, the tubes and plates were incubated at 37°C for the pathogenic bacteria and coliform group members, and at 32°C for the remainder of the bacteria. 28 Growth evaluation Both plates and tubes were visually examined for growth at pre- determined time intervals and compared with the growth of the control cultures (grown on basal media only). At the end of the incubation period, growth on sorbate-treated media were reported as no growth (-) , weak growth (1‘) , moderate growth (+), good growth (++), and vigorous growth (+++). Utilizajfition of Potassium Sorbate as a Carbon Source In all of the utilization studies for molds and yeasts, the media used were the same as for the screening tests, except that glucose was not present in the medium. Instead, potassium sorbate was added as the sole source of carbon. For bacteria, the medium used was yeast extract and tryptone, to which potassium sorbate was added in order to test for their ability to utilize sorbate as the sole source of carbon. Potassium sorbate was used at concentrations ranging from 3000 to 5000 mglkg. The initial pH was adjusted to 6.0 in all utilization experiments. The remainder of the media preparations, inoculation, incubation, and growth evaluation were carried out in the same manner as in the screening tests. Growing Cells Studies Molds and yeasts were precultured for 20 to 72 hours at 25°C on YPG agar plates fortified with 500 mglkg of potassium sorbate. The cultures were suspended in sterilized, distilled, deionized water and the suspen— sions were used as inocula. 29 In all growth studies, unless otherwise stated, the medium was the same as that used in the screening tests except that the potassium sorbate concentration and pH were varied in some cases. All growth experiments were carried out in 500 ml Erlenmeyer flasks containing 200 ml of medium. The flasks were inoculated with the cell suspension and incubated in a gyratory shaker (Model G-25, New Brunswick Scientific Co. , New Brunswick, NJ) at the same temperature as used in the screening tests. The lactic acid bacteria were not shaken, but instead were incubated by standing at the appropriate temperature. In some cases, very slow shaking was provided shortly before the ter- mination of the incubation period. Representative samples wereaseptically drawn at predetermined time intervals and analyzed for residual potassium sorbate according to the method described by Robach et al. (1978). These samples were also analyzed for metabolites using gas liquid chromatography (GC) and mass spectrometry (MS). The Clostridia were cultivated on thioglycol late broth without dextrose (Difco). Potassium sorbate (1000 mglkg) and glucose (2.0 percent) were added. The pH was adjusted to 5. 5. The medium was inoculated with vegetative cell suspensions (suspension in which vegetative cells were predominant). This suspension was prepared by growing the culture under study in thioglycollate broth for 72 hours. Aliquots of one to two ml of these cultures were used as inocula. Non-G rowigg (Restim) Cells Studies The resting cell suspensions of molds and yeasts were prepared according to the method described by Tahara et al. (1977) with some 30 modifications. Aliquots (200 ml) of the basal medium plus 500 mg/kg of potassium sorbate in 500 ml Erlenmeyer flasks were inoculated from fresh stock cultures and incubated for 20 to 72 hours, depending on the strain and the inoculum. For molds, mycelia and spores were separated from the broth by filtration on a Buchner filter with Whatman filter paper No. 02 and washed several times with distilled water to yield a high moisture paste ready for inoculation. For yeast and bacteria, cells were separated from the broth by centrifugation at 10000 x 9 using a Sorvall Super speed RC-2 automatic refrigerated centrifuge (Ivan Sorvall lnc., Norwalk, Conn.) for 20 minutes, and washed at least one time with distilled water. The high moisture paste which was obtained was used for inoculating the buffer system. The reaction medium was prepared according to the method described by Tahara et al. (1977), except that glucose was used at a lower concen— tration. This medium contained a phosphate buffer, glucose (two percent) and 1000 mglkg of potassium sorbate. Phosphate buffer was prepared from stock solutions of 0.2 M monobasic sodium phosphate and 0. 2 M dibasic sodium phosphate in the proper prOportions according to the method described by Sorensen (1909a and b). One to two gram portions of the cell pastes were added to 100 ml of medium in 500 ml Erlenmeyer flasks. Incubation was carried out at room temperature for all molds and yeasts in a slow shaker developing about 50 rpm (Eberbach Corp., Ann Arbor, Michigan). Representative samples were drawn every hour, and analyzed for residual sorbate and metabolites of sorbic acid according to the procedure described later herein . 31 Sample Preparation for GC Analysis Extraction of metabolites At the end of incubation period, the flask contents (200 ml) were either filtered through Whatman No. 02 filter paper using a Buchner filter or centrifuged. Molds usually were filtered, whereas yeast and bacteria were centrifuged at 10000 x g for ten to 20 minutes. Generally speaking, the extraction was carried out according to Tahara et al. (1977). The filtrate or the supernatant was saturated with analytical reagent (AR) grade sodium chloride, acidified with 3N HC1 to a pH of 1—2, transferred into a separatory funnel, extracted with two 50 ml portions of redistilled diethyl ether (original analytical reagent grade diethyl ether was obtained from Mallinckrodt Inc. , St. Louis, MO). The combined ether extracts were washed with distilled, deionized water and dried over anhydrous sodium sulfate. This extract was used for the detection of neutral and acidic metabolites at the same time after methylation . To divide the above extract into acidic and neutral metabolites, the ether extract was washed with a 3N solution of sodium carbonate. This extract was designated as "wash #1." The aqueous phase was saved, reacidified, and extracted with redistilled ether. The ether layer was washed, dried over anhydrous sodium sulfate and saved for the analysis of acidic metabolites after methylation. The ether layer from wash #1 was rewashed with water, dried over anhydrous sodium sulfate and saved for analysis for neutral metabolites. The extraction procedure is outlined in Figure 2. 32 Figure 2. Flow chart diagram for the isolation and purification of acidic and neutral metabolites of sorbic acid 33 Micrdbial Culture - Separate cells (filter and/or centrifuge) -Transfer to a separatory funnel -Saturate with NaC1 -Acidify with 3N HC1 -Extract with diethyl ether Ether layer Aqueous layer rcontaining neutral and acidic metabolites wash with 3N NaHCDB i ii Aqueous layer Ether layer containing acidic metabolites containing neutral -acidify with 3N HC1 metamhtes -extract with ether wash‘with . ‘ water Aqueous layer Ether layer containing acidic metabolites [wash with water I a l Ether layer Aqueous layer Aqueous layer Ether layer containing acidic containing metabolites neutral metabolites wash with water -dry over NaZSO4 I 7 Ether layer Aqueous layer -ggnge:§rate containing acidic , metabolites at 40 C under re- .—..- duced -dry over NaZSO4 pressure -concentrate to 2 ml -::2::nt::te at 40°C under reduced pressure tle g -Methylate with ethereal solution - stream of of diazomethane N2 -concentrate to 0.5 ml under a gentle stream of N2 0.5-1.5 pl injected into cc 30 Preliminary concentration The extracts were transferred into 500 ml round bottom flasks and concentrated under reduced pressure at 00°C, using a Rotavapor-R (Buchi) to a volume of two ml. For further concentration, a nitrogen stream was used to concentrate the extract to 0.1 to 0. 5 ml, depending on the concentrations of the metabolites. Esterification and final concentration Esterification of the acidic metabolites was carried out with an ethereal solution of dia zomethane which was prepared from Diazald (Aldrich Chemical Co.) according to the manufacturer's instructions. The esterification was carried out using the method described by Wick et al. (1969). An ethereal solution of diazomethane was added drop by drop to the concentrate until a yellow color persisted. The solution was permitted to stand at room temperature for ten minutes. To remove the excess dia zomethane and to further concentrate the extract, a stream of nitrogen was directed toward the surface at a rate of 300 to 500 ml/minute. Head space Analysis Headspace analysis was carried out specifically for the detection of 1,3-pentadiene since it is very volatile and is very difficult to extract from aqueous solutions, and to concentrate without entailing significant losses. After incubation of the microorganism had been terminated, the flasks were tightly stoppered, heated to about 60°C, and a gas-tight syringe was used to draw a sample (one ml) from the headspace. This sample was injected directly into the gas chromatograph or gas chroma- tograph I ma ss spectrometer . 35 Entrapment of Pentadiene To entrap pentadiene, the method described by Marth et al. (1966) was essentially followed. The flask contents were heated to about 70°C in a water bath and the volatiles were distilled into a small tube con- taining approximately 0. 5 ml of AR grade methanol. The tube was immersed in a dry ice-acetone mixture. A slow stream of air was used to flush the volatiles out of the flask. The tube contents were trans- ferred into a vial kept in a dry ice-acetone mixture. One to two 111 portions of the vial contents were injected into the gas chromatograph on the same day. Gas Liquid Chromatpgraphic Analysis of Sorbic Acid Metabolites Apparatus and operation conditions A Hewlett Packard gas chromatograph (Model 5830A) equipped with a flame ionization detector (F10) and a Hewlett Packard 18850A GC terminal (Hewlett Packard Corp. , Avondale, PA) was used for the analysis of sorbic acid metabolites. The column used was glass (2 m x 2 mm i.d.) and was packed with ten percent Carbowax 20 M coated on 80/100 mesh acid—washed (AW) Chromosorb W (Supelco, lnc., Bellfonte, PA). The instrument was operated under the following conditions: Initial temperature (T1): 60°C (30°C for 1,3-pentadiene analysis) Time at T1 (t1) : one minute Rate : 5°C per minute Final temperature (T2) : 160 Time at T2 : 5—10 minutes Injection port temperature : 200°C 36 Flame ionization detector temperature : 275°C Chart speed : one cm per minute Attenuation : variable Slope sensitivity : .02 Carrier gas : Nitrogen Carrier gas flow rate : 30 ml per minute Hydrogen flow rate : 30 ml per minute Air flow rate : 200 ml per minute For injection, a Hamilton syringe of tlen ul capacity (701N) (Hamilton Co. , Reno, Nevada) was used to inject volumes of between 0. 5 to 1. 5 ul. Gas Liquid ChromatocraphL-Mass Tpectrometiy (GC- The GC IMS system used was a Hewlett Packard 5985 A Gas Chromato- graph/Hewlett Packard Mass Spectrometer (Hewlett Packard Corp.). The column was the same as that used for CC analysis. Helium was the carrier gas with a flow rate of 25 mI/minute. The analyses were carried out using a temperature program from 60°C to 160°C at 5°C lminute, with a one minute hold at 60°C. The ion source and analyzer temperature of the mass spectrometer were maintained at 200°C . The electron multiplier voltage was 2000 V and the ionizing potential was 70 eV. Quantitative Analysis of Residual Sorbate py High Performance rigid ChFomatography Preparation of sanyale For the analy°sis of residual sorbate, five to ten ml samples were drawn aseptically from the growing culture or from the resting cells 37 reaction medium at certain time intervals, filtered through a Whatman No. 02 filter paper and [or centrifuged as previously described, refiltered through a Millipore filter (0.05 pi) and stored in tightly closed amber vials at 0°C for analysis. _l-Ii_g;h performance Iiguid chromatography conditions A Water Associates high performance liquid chromatograph, Model ALC IGPC 202/R001 equipped with a septumless Model U6K universal injector, a Model M-6000 pump, and a Model 000 absorbance detector (Water Associates, Inc. , Milford, MA) was used for the analysis of residual sorbate. The response was recorded on a Model 232 Linear instruments recorder (Linear Instruments Corp.). The method described by Robach et al. (1978) was followed, except that acetate buffer (pH 0.0) was used in place of phosphate buffer. According to this method, a reverse phase column, uBondapak C18 (3.9 mm x 20 cm) (Water Associates), was used. A degassed mixture of acetonitrile and acetate buffer (20:80, (v/v)) was used as the eluent, with a flow rate of 1. 5 ml per minute. Ten to 00 ul portions of the samples were injected into the chromatograph using a 50 ul syringe. For quantitation, 20 ul portions of different concentrations (five to ten mglkg) of a fresh aqueous solution of potassium sorbate were injected , and a standard curve established by plotting peak heights against sorbate concentrations. From this standard curve, the concentration of potassium sorbate was determined. However, in some cases, the sample had to be diluted to fall within the sensitivity range of the instrument. 38 Fate of Labelled Sorbic Acid Labelled sorbic acid was utilized in this study to confirm that the metabolites isolated from media containing sorbic acid inoculated with selected microorganisms were in fact derived from sorbic acid. Since only a limited amount of uniformly 1uC-labelled sorbic acid (supplied by Monsanto Chemical Company, St. Louis and originally obtained from New England Nuclear, Boston, Mass.) was available, the study was limited to the thin layer chromatographic analysis of the neutral metabolites of Rhizopus stolonifer grown on a medium containing labelled and nonlabelled sorbic acid. For the same reason, only one metabolite in the reaction system was analyzed for the presence of radioactivity. Sorbyl alcohol was chosen for this purpose since a standard sample was commercially available and it had been detected in appreciable amounts over a relatively long period of time among the metabolites of Rhizopus stolonifer grown on sorbate containing medium. Thin layer chromatography of neutral metabolites of sorbic acid Precoated 20 x 20 cm silica gel G plates (Fisher Scientific Co., Pittsburgh, PA) were used to separate the neutral metabolites of sorbic acid. The plates were activated at 105°C for one hour. Approximately five ul portions of the neutral metabolite concentrate were spotted on the plates about 1. 5 cm from the lower edge. On the same plant, five (11 portions of an ethereal solution of standard sorbyl alcohol were also spotted. The mobile phase was a mixture of petroleum ether, diethyl ether and acetic acid (90:50:1, v/v) . The plates were developed in one 39 dimension for a distance of about 15 cm (approximately 75 minutes), after which the plates were removed, dried at 60°C for about 20 minutes and sprayed with a 50 percent aqueous solution of sulfuric acid. The average Rf values for borbyl alcohol were recorded. The neutral extract which was to be examined for the presence of radioactivity was spotted on a preparative TLC plate and developed as previously described. The spot having the same Rf value as the standard sorbyl alcohol was located with the aid of the predetermined Rf value and was removed from the TLC plate and placed in scintillation vials. Ten ml of a liquid scintillation cocktail (Aquasol) (New England Nuclear) were added to the sample. The vial was then placed in a Packard Tri Carb Liquid Scintillation Spectrometer Model 3310 (Packard Instruments Co. , Inc. , Downers Grove, Ill.) and counts per minute were recorded. Further confirmation by GC Analysis Another preparative TLC plate was spotted with the same neutral metabolite extract as above. The compound with an Rf value equal to that of sorbyl alcohol was scraped off the plate, extracted from the silica gel with diethyl either, concentrated and injected into the gas chromatograph. The retention time was compared with that of a standard sorbyl alcohol sample. RESULTS AND DISCUSSION Microbial Resistance to Sorbic Acid One of the main objectives of this study was to test microorganisms for their resistance to sorbic acid. For this reason, various species of molds, yeasts and bacteria were examined. tees Molds in general, exhibited a wide variation in their resistance to sorbic acid. Results of the screening tests for molds using concentra- tions of potassium sorbate ranging from 500 to 3000 mglkg are presented in Table 1. Most of the cultures grew in the presence of 500 mglkg of potassium sorbate at pH 5. 5. Only a few molds were inhibited under these conditions and included a strain of Rhizgpus oligosmrusi, two strains of Cladosporium cladosporiodes, and a strain of Alternaria tenuis. As the concentration of potassium sorbate was increased to 1000 mglkg, a number of other molds were also inhibited (Table 1). However, most of the inhibition occurred when the concentration of potassium sorbate was increased above 2000 mglkg (Table 1). Some of the mold cultures, however, were found to be relatively resistant to sorbic acid and included Aspergillus flavus NRRL 2999, Appegillus flavus NRRL 6550, Fusarium sp. , Geotrichum candidum, Penicillium sp. and Trichoderma viride. These mold cultures grew well in the presence of 3000 mglkg of potassium sorbate. Unfortunately, some of the resistant molds are potential food spoilage agents, and 00 01 Table 1. Inhibitory concentrations of potassium sorbate for some mold cultures on YPG plates at pH 5. 5, 25°C 500 mg/ kg potassium sorbate 2000 mglkg potassium sorbate Rhizopus oligosporus ATCC 2295 Cladosporium cladosporiodes QM 089 Cladosporium cladosporiodes OM 9085 Alternaria tenuis NRRL 2169 Alternaria sp. Mucor pusillus Mucor humiculus Mucor sp. Rhizopus oligosporus NRRL 2710 Rhizopus stolonifer Penicillium janthinellum NRRL 2016 1000 mglkg potassium sorbate 3000 mglkg potassium sorbate Fusarium roseum Fusarium oxysporum NRRL 1903 Fusarium sp. Calvatia gigantica Penicillium oxalicum NRRL 790 Aspergillus ory zae Aspergillus niger Penicillium roqueforti 02 in addition, Aspergillus flavus can produce aflatoxins in foods under certain conditions. In this regard, it has been reported that sorbic acid will inhibit aflatoxin B1 production by Aspergillus flavus when used at a concentration of 100 mglkg. However, much higher levels (1700 mglkg) of sorbic acid are necessary for complete inhibition of the growth of the mold. ' Species of both genera, Aspergillus and Penicillium, were found to be, in general, more resistant than the other species of other genera. Most of these cultures required 2000 mg [kg or more for complete inhibition. This observation agrees with the findings of York (1960) who reported that 25 mg of undissociated acid per 100 ml of medium (250 mglkg) was necessary to inhibit all but the most resistant Penicillium. This concen- tration is equivalent to approximately 2000 mglkg of potassium sorbate at pH 5. 5. It was also reported that some species of Penicillium were resis- tant to the highest sorbic acid concentration (300 mglkg) used (York, 1960) . On the other hand, Klis et al. (1959) who studied the effect of several antifungal agents including sorbic acid on the growth of common food spoilage fungi, reported that a number of fungi including Alternaria sp. , Botrytis cinerea, Fusarium sp., Mucor mucedo, Ofidium lactis (Geotrichum candidum), Penicillium gigitatum, Penicillium expansum, Penicillium 233., and Rhizopus fl were completely inhibited for 96 hours by 500 mglkg of sorbic acid at pH about 5. 6. However, Asperflus nicer was not inhibited by 1000 mglkg of sorbic acid. In the present study, Aspergillus niger was found to require 3000 mglkg for complete inhibition. Similarly, two strains of Geotrichum candidum were found to resist a concentration of potassium sorbate as high as 3000 mglkg. A preliminary study on a strain of Penicillium digitatum indicated that this mold can tolerate at least 1500 mglkg of potassium sorbate. 03 When the concentration of potassium sorbate was raised to 5000 mglkg (600 mglkg of undissociated acid), none of the mold cultures tested grew over a period of five days. Likewise, Bell and Borg (1959) tested 66 species of mold in 32 genera for growth in basal media with or without 0.1 percent sorbic acid (1000 mglkg) at pH 0.5 and 7.0. All of the cultures grew in the control basal media at both pH values and in the sorbic acid media at pH 7.0, but none grew in the basal medium containing sorbic acid at pH 0. 5. At this latter pH, the concentration of the undis- sociated acid is about 650 mglkg. Contrary to previous reports, Marth et al. (1966) , found that a number of Penicillium species including Penicillium cypno-fulcum, Penicillium notatum, Penicillium frequentans, Penicillium roqueforti and a mutant of Penicillium roqueforti were able to grow in the presence of up to 1800, 2300, 2800, 5000 and 7100 mglkg of potassium sorbate at pH 5. 5, respectively. These mold cultures were isolated by the authors from potassium sorbate-preserved cheese. According to the results of the present study and the previous literature reports, variations in the susceptibility to sorbic acid exists between different mold species of a single genus, and to a lesser extent between the cultures of the same species. The latter variation is possibly due to the effect of their previous habitats. 1222 Although both solid and liquid media were used for the screening tests, only those of the solid media will be shown here. Slight differences, however, were observed between the two media, i.e. , slightly higher concentrations of potassium sorbate were required to inhibit the growth 00 of yeast cells in liquid media compared to those required to achieve the same effects in solid media. These findings are in agreement with those of York (1960). He ascribed this to the possible syneresis which occurs at the surface of media upon autoclaving and in subsequent drying of the surface, coupled with the slower rate of diffusion of solutes through agar than through a liquid medium. Results of the screening tests on selected yeasts using concentrations of potassium sorbate ranging from 500 to 3000 mglkg at pH 5. 5 are presented in Table 2. Most of the yeasts tested possessed some resis- tance to sorbic acid. The majority of these yeasts are resistant to concentrations of potassium sorbate ranging from 500 to 1000 mglkg (60-120 mg [kg of undissociated acid). Rhodotorula rubra, Endomycopsis selenospora, Schizosaccharomgces embe and Sporobolomjces coralliformis were very susceptible to as low as 500 mglkg of potassium sorbate (Table 2). These cultures either did not grow at all or grew very weakly. Thus, this concentration was considered inhibitory to this group. The growth of another group of yeasts was found to be severely retarded or completely inhibited by 1000 mglkg of potassium sorbate (Table 2). The members of the last two groups required concentrations of potassium sor- bate ranging from 2000 to 3000 mglkg for complete inhibition of growth. This range corresponds to 237 to 355 mglkg of the undissociated acid. It is apparent that the members of the four groups of yeasts do not have many things in common, except the way they behave toward sorbic acid, i.e., the four groups include cultures of different criteria. York (1960) pointed out that yeasts with a predominantly oxidative metabolism, e.g. Pichia, Debaryomyces, Rhodotorula, and Cryptococcus are more 05 Table 2. Inhibitory concentrations of potassium sorbate for some yeasts on YPG plates at pH 5. 5, 25°C 500 mglkg potassium sorbate 2000 mg/kg potassium sorbate Rhodotorula rubra Saccharomyces carlsbergensis ATCC 9080 Endomycopsis selenospora Hansenula anumala NRRL Y-1357 Schizosaccharomyces pombe Hansenula anumala Sporobolomyces coralliformis Hansenula californica ATCC 16039 NRRL Y-1027 Saccharomyces cerevisiae Saccharomyces cerevisiae Saccharomyces rouxii Saccharomyces cerevisiae var elliposoideus Torulopsis sphaerica Hansenula wingei ATCC 10355 Hansenula wingei ATCC 10355 Endomycopsis bispora Saccharomyces kluyveri Saccharomyces kluyveri (C-26) 1000 mglkg potassium sorbate 3000 mglkg potassium sorbate Endomycopsis bispora ATCC 10628 Candida krusei Torulopsis sake NRRL Y-1622 Schizosaccharomyces octosporus Nadsonia fulvescens NRRL Y-991 Debaryomyces membranaefaciens . NRRL Y-089 Rhodotorula sp. Saccharomyces oleaginosus Rhodotorula glutinis var glutinis Pichia fermentans Candida utilis NRRL Y-900 Saccharomyces sp. Schizosaccharomyces aponicus var virsatilis Trigonopsis variabilis ATCC 10679 Candida steatolytica Trichosporon fermentans NRRL Y-1092 Trichosporon cutaneus ATCC 13005 06 sensitive to sorbic acid than many of the yeasts which have fermentative properties. In the present study, to the contrary, several yeasts with strong fermentative activities were shown to be very susceptible to sorbic acid (Table 2) . It appears that there are factors other than the fermentative properties that influence yeast behavior towards sorbic acid. In contrast to the results of this study, Emard and Vaughn (1950) tested ten cultures representing ten species of yeasts and concluded that the cultures tested did not exhibit marked resistance to sorbic acid. The same authors tested Debaryomyces membranaefaciens on a glucose broth containing a 0. 07 percent sorbic acid at different pH values. They found that the growth of this yeast was severely retarded at pH 6.0, while growth was completely inhibited when the pH was lowered to 5.0 (170 mg/kg of undissociated acid). Results of the present study indicate that the same yeast species required 355 mglkg of undissociated acid. Some of the yeast cultures investigated in the screening tests were found to possess relatively high resistance to sorbic acid and included Candida lipolytica, Hansenula saturnus, Schwanniomyces occidentalis, and Brettanomyces claussenii. These yeast cultures could grow well in the presence of 3000 mglkg of potassium sorbate at pH 5. 5 (355 mglkg of undissociated acid). This concentration is well above the concentration (236 mglkg undissociated acid) which Costilow, et al. (1955) reported was necessary to inhibit the most tolerant strains of yeasts, Torulopsis holmii and Candida M. In the present study, 355 mglkg of undissociated acid was necessary to inhibit the latter yeast, Candida krusei. When the concentration of potassium sorbate was raised to 5000 mglkg, none of the tested yeast cultures grew. lowever, Splittstoesser et al. tn (1978) reported that a strain of Saccharomyces bispgrus var bisporus exhibited exceptional resistance to sorbic acid. This yeast, which was isolated from grape juice, grew well in a grape juice broth containing over 800 mglkg of the undissociated acid. This concentration is equivalent to about 6700 mglkg of potassium sorbate at pH 5. 5. From the results obtained in present study and from previous litera— ture reports, it can be concluded that there is a variation within members of a single species in their behavior toward sorbic acid. One explanation for this variation could be the type of media used in the experiments. For example, Marth et al. (1966) and Emard and Vaughn (1952) have shown that the media greatly affect the microbial response to sorbic acid. Another explanation is the previous habitat of the yeast strain. Bacteria Both liquid and solid media were used for the preliminary screening tests on bacteria. Essentially, no Significant difference was found between liquid and solid media as far as the inhibitory action of sorbic acid is concerned. Thus, the liquid medium was chosen for this study and only those observations will be presented here. Results of the screening tests on bacterial cultures when grown in different concentrations of potassium sorbate (500-3000 mglkg) at pH 5. 5 are presented in Table 3. As was the case with molds and yeasts, the resistance of bacteria to sorbic acid varies widely. Most of the bacterial cultures tested in the study were not inhibited in culture media fortified with 500 mglkg of potassium sorbate at pH 5. 5. Bacillus c_e_r_et£ can cause food borne illnesses and is readily isolated from nearly all plant foods and dehydrated foods (Ayres et al. , 1980) . Table 3. Inhibitory concentrations of potassium sorbate for some bacterial cultures on YTG broth at pH 5. 5, 25°C 500 mglkg potassium sorbate 2000 mg [kg potassium sorbate Bacillus cereus Bacillus coagulans Bacillus polymyxa Sarcina Iutea Salmonella typhimurium Serratia marcescens Pseudomonas fluorescens Pseudomonas schuylkilliensis ATCC 15916 Pseudomonas schuylkilliensis ATCC 15917 Pseudomonas vulgaris Staphylococcus aureus Streptococcus zymogenes Salmonella senftenberg Micrococcus flavus 1000 mglkg potassium sorbate 3000 mg/kg potassium sorbate Aerobacter aerogenes Bacillus subtilis E. coli Pseudomonas fragi Staphylococcus epidermidis Salmonella pollorum Lactobacillus plantarum Lactobacillus bulgaricus Leuconostoc mesenteroides Pedicoccus cerevisiae Pseudomonas fluorescens Lactobacillus brevis 09 This bacterium was shown to be very sensitive to sorbic acid in this study. Potassium sorbate at a concentration of 500 mglkg was very effective in inhibiting a strain of this bacterium in culture medium at pH 5. 5. When the concentration of potassium sorbate was increased to 1000 mglkg, the growth of a number of bacterial species were markedly retarded or completely inhibited (Table 3). This group included some enteric bacteria such as Aerobacter aelgggnes, Escherichia coli and Salmonella pollorum. As the concentration of potassium sorbate was increased to 2000 mglkg, more bacteria were inhibited (Table 3). Two cultures of Salmonellae LSaImonella senftenberg and Salmonella typhimurium), which are well-known food illness agents, were effectively inhibited as was Staphylococcus aureus which is also known to cause food poisoning. Emard and Vaughn (1952) reported that strains of Staphylococcus are considerably more resistant to sorbic acid than other catalase positive microorganisms, including some species of Salmonella. According to their data, Staphylococci grew in 1000 mglkg of sorbic acid in liver infusion media but were inhibited by 2000 mg/kg of sorbic acid in liver infusion media and by 700 mglkg of sorbic acid in glucose yeast extract media. Recently, Tompkin et al. (1970) reported that growth of the Salmonellae (Salmonella m, Salmonella infantis, Salmonella senftenberg, Salmonella choleraesuis, and Salmonella newport) in cooked, uncured sausage with a pH of 6. 2 were remarkedly retarded by 1000 mglkg (<50 mglkg of undissociated acid) of potassium sorbate. This concentration is much lower than the concen- tration (120-100 mglkg) which was found to be necessary to inhibit some of these bacterial cultures in the present study. Tompkin et al. 50 (1970) reported that Staphylococcus aureus was not inhibited by 1000 mglkg of potassium sorbate, which agreed with the results obtained in this study for the same microorganism. Some of the food spoilage bacteria such as strains of Pseudomonas fluorescens and Pseudomonas vuljpris were also effectively inhibited by this concentration (2000 mglkg) (Table 3). The members of the third group (Table 3) were found to possess considerable resistance to sorbic acid. They required 3000 mglkg of potassium sorbate for marked retardation or complete inhibition of growth. Of the most resistant bacteria among the test bacteria were lactic acid bacteria. It is interesting to note that when these bacteria were cultivated in a rich medium such as trypticase soy broth, the degree of inhibition was less, i.e. , more potassium sorbate (5000 mglkg) was needed to accomplish the same degree of inhibition. These results confirmed those of Emard and Vaughn (1952), Costilow et al. (1955), and York (1960) with respect to the tolerance of the catalase negative lactic acid bacteria to sorbic acid, especially in highly nutritious media. Microbial D_egradation of Sorbic Acid‘ Melnick et al. (1955) were the first to report that mold enzymes accom- plish metabolic degradation of sorbic acid by reducing it to carbon dioxide and water. In spite of the importance of sorbic acid as a food preserva- tive, few studies have been carried out on the metabolism of sorbic acid, and especially the identification of the metabolites of this additive. One of the objectives of the present study was to qualititatively follow the fate 1The terms sorbic acid and potassium sorbate will be used interchangeably throughout the text. 51 of sorbic acid in the presence of microorganisms in terms of the metabolites produced, and quantitatively in terms of how much sorbic acid was con— sumed. In order to follow the fate of sorbic acid in the presence of different microbes, both resting and growing cells studies were performed using some selected microbes. Observations on only some of these microbes will be presented herein. Data regarding the utilization of sorbate by growing cells of different fungi cultivated on YPG media containing 1000 mglkg of potassium sorbate are presented graphically in Figure 3. As indicated, some microbes such as Trichorderma viride and Penicilliun sp. degrade sorbic acid completely in a relatively short time. Other molds such as Rhizopus stolonifer were relatively slower in degrading sorbic acid, and thus took a longer time to bring about a total destruction of sorbic acid. The reason for this delay was that mold growth was initially retarded for at least one day, after which growth proceeded normally. Geotrichum candidum and Candida limytica were though to be very effective in degrading sorbic acid, especially Geotrichum candidum, since they started to grow vigorously in a relatively short time. Results indi- cated that the growing cells of Geotrichum candidum required two to three days to bring about total degradation of sorbic acid, while Candida limlytica required three days to achieve the same effect (Figure 3). The resting cells of both organisms were also much slower than the other fungi in bringing about the total destruction of sorbic acid (Figure 0). This behavior may be ascribed to the catabolic repression or the "glucose effect." This phenomenon is usually encountered when a certain microbial culture is grown in a medium containing pairs of compounds as carbon sources, 52 100 801- ' ~0- Candida lipolytica E -o- Geotrichum candidum LE“ 4- MC—IILLUID .59; 8 60— «a- LIZIELQdQLma liclde i‘—" o- Rhizouus stglgnifer s c: g 40-— as 000 1 ‘ 2 INCUBATION TIME, Days Figure 3. Utilization of sorbate by growing cells of different fungi cultures cultivated on YPG containing 1000 mglkg of potassium sorbate, at pH 5. 5, and an incubation temperature of 25°C %SORBATE CONTENT 53 . -0- Candida lipglflica '0' Geotrichum candidum "3- Penicillium _s_p._ +Trichoderma viride a'Rhizopus stolonifer .\ \. \ \ . \, \e \' \ e \. \ '\\. \. '\,\. \. \'\\'\ \. '\'.\. \ \.\. \ ‘0‘ \3 \ L l l t. [A 0' °° 1 2 3 4 5 a 1 INCUBAI'I'ON TIME. Hours Figure 0. Utilization of sorbate by resting cells of different fungi cultures suspended in a phosphate buffer containing 2.0 percent glucose and 1000 mglkg of potassium sorbate, at pH 5. 5 and an incubation temperature of 25°C 50 one of which is more easily metabolized than the other (Stanier et al. , 1976). It appears that these two microorganisms depleted the glucose after which growth ceased. The microorganisms enter a short lag phase where they start to synthesize the inducible enzymes necessary for sorbic acid metabolism. In order to exclude the complications of growing cells, and to follow the fate of sorbic acid in a relatively short period, experiments were also carried out with resting cells (Figure 0). Since the experiments were initiated with heavy inocula, total destruction of sorbic acid was accom- plished in a relatively short time. Resting cells of Rhizopus stolonifer depleted sorbic acid in six hours. This was in contrast to the growing cells of the same mold which took a relatively long time (three days). These observations appear to be consistent with those of Kurogochi et al. (1975) and Tahara et al. (1977) on Mg; Species of Aspergillus and Mucor behaved toward sorbic acid in a similar way to that of Penicillium and Rhizopus, respectively. Thus, results with only the latter genera were included in this study. Results, in general, indicate that the species of the genera Penicillium, Aspergillus, Mucor, and Rhizopus were very effective in degrading sorbic acid when present in sublethal concentrations. Trichoderma M3 also very effective in degrading sorbic acid, but since only one species was tested, one cannot make any generalizations about the entire genus. These findings are in general agreement with those of Lukas (1963) who reported that sorbic acid was decomposed by Agpegillus pige_r, Aspergillus oryzae, Penicillium sp., Penicillium expansum, Alternaria §p., Trichoderma legnorum, Mucor racemorus and Rhizopus pigricans (Rhizopus stolonifer). 55 To rule out the possibility of loss due to chemical and /or physical means, uninoculated samples treated the same way, were analyzed for residual sorbate. These samples, however, did not exhibit any loss of sorbate during the incubation period of the inoculated samples. Hence, the microbial degradation of sorbic acid was confirmed. The bacteria, Lactobacillus bulgaricus, Lactobacillus plantarum, Pediococcus cerevisiae and Pseudomonas fluorescens were grown on trypticase soy broth fortified with 1000 mglkg of potassium sorbate. Sorbate concentration were monitored over a period of six days. All of the above bacteria, except Pseudomonas fluorescens, did not exhibit any significant loss. Pseudomonas fluorescens showed about 28 percent loss after three days. Sorbic Acid as a Carbon Source Having established that sorbic acid could be depleted by some of the microorganisms tested, further studies were conducted to establish which microorganisms could use sorbic acid as a sole source of energy and carbon. Results of this study help in understanding the pathway of degradation and the mechanism by which microorganisms detoxify sorbic acid. For this reason, most of the microbial cultures were grown on basal media containing 3000 mglkg of potassium sorbate as the only carbon source. Growth on this medium was the criterion for utilization of sorbic acid as a carbon source. The state of growth of selected mold cultures on the medium containing both sorbic acid and glucose and the basal medium containing only sorbic acid is shown in Table 0. Trichoderma viride grew vigorously on both media as did Penicillium sp., Geotrichum candidum, Aspgiggillus niger and 56 Table 0. The effect of the presence or absence of glucose (two percent) on the growth of different mold cultures on yeast extract-peptone— agar media (pH 6.0) containing 3000 mglkg of potassium sorbate and incubated at 25°C for six days Growth on yeast Growth on yeast extract- extract-peptone- peptone-agar+ agar+3000 mglkg 3000 mglkg potassium Mold potassium sorbate sorbate + glucose Aspergillus flavus NRRL 2999 ++ + Aspergillus flavus NRRL 6550 ++ + Penicillium roqueforti + ++ Aspergillus niger + + Trichoderma viride +++ +++ Penicillium sp. ++ ++ Geotrichum candidum ++ ++ Aspergillus oryzae + + Fusarium sp. (2) ++ + + Moderate growth ++ Good growth +++ Vigorous growth 57 Aspergillus was. Only Penicillium roqueforti exhibited more vigorous growth on the basal medium containing both sorbic acid and glucose than it did on the medium containing only sorbic acid. Marth et al. (1966) also reported that the depletion of sorbic acid was enhanced by a highly nutritious substrate and retarded by a minimal growth medium. Unexpectedly, two strains of Aspergillus flavus and one strain of Fusarium 5p. exhibited better growth on the basal medium containing only sorbic acid. This behavior was designated as the "glucose inhibitory effect." This may be explained by the assumption that these organisms, once grown on media containing both sorbic acid and glucose, can produce glucose metabolites in the early stages of growth. These metabolites may enhance the inhibitory action of sorbic acid. In this regard, Phaff et al. (1966) stated that metabolites of glucose may reduce the pH, thus accentuating the inhibitory effect of sorbic acid. When the concentration of potassium sorbate was raised to 5000 mglkg in both media, none of the mold cultures could grow. The yeasts which could grow when sorbic acid was used as the sole source of carbon are listed in Table 5. As indicated, Candida lipolytica grew vigorously on media containing sorbic acid as the sole source of carbon. Unexpectedly, sorbate (3000 mg/ kg) in the presence of glucose (2.0 percent) markedly inhibited growth of Candida lipolytica. An almost similar result was observed with Hansenula californica, although the inhibition here was more apparent (Table 5). Contrary to the observations in this study with regard to the inhibitory effect of glucose, Pitt (1970) reported that increasing the glucose concentration from 0. 5 to 1.0 percent in a yeast nitrogen base broth increased the maximal levels of sorbate per- mitting growth of Sacchanomyces bailii from 380 mglkg to 600 mglkg. This 58 Table 5. The effect of the presence or absence of glucose (two percent) on the growth. of yeast cultures on yeast extract-peptone-agar media (pH 6.0) containing 3000 mglkg of potassium sorbate and incubated at 25°C for six days Growth on yeast Growth on yeast extract- extract-peptone- peptone-agar + agar + 3000 mglkg 3000 mglkg potassium Yeast potassium sorbate sorbate + glucose Candida lipolytica +++ i Hansenula californica ++ - NRRL Y-1025 Candida steatolytica + - Trichosporon cutaneus + - Saccharomyces occidentalis + + ATCC 2320 Brettanomyces claussenii + + NRRL Y-109 Candida krusei i i Debaryomyces membranaefaciens i : NRRL Y-109 Hansenula wingei ATCC 10355 i i Hansenula wingei ATCC 10356 i i Schizosaccharomyces octosporus i - i i Saccharomyces cerevisiae - No growth 1‘ Weak growth + Moderate growth ++ Good growth +++ Vigorous growth 59 case, in particular, could be explained by the fact that this yeast is an osmophilic one (Ayres et al. , 1980). Interestingly, the yeasts which grew vigorously in the presence of 3000 mg/kg of potassium sorbate at pH 5. 5 grew weakly when the pH was raised to 6.0 (Table 5). One possible reason for this behavior is that these yeasts were maintained on a medium with a pH of approximately 5.0. Thus, it appears that they adapted to this low pH. The other possibility is that these yeast cultures possess pH optima close to 5. 5. It is apparent from the present study that only yeasts with high tolerances to sorbic acid can metabolize sorbic acid as the sole source of carbon. However, one cannot make any positive conclusions about these observations since other yeasts which are sensitive to sorbic acid may utilize it, if present in sublethal concentrations. Hansenula wengg, when grown in a liquid medium fortified with 1500 mglkg of potassium sorbate as the only sole source of carbon, was effective in metabolizing sorbic acid in a very short time. It appears, however, that at high concentrations of sorbate (3000 mglkg) , sorbic acid no longer serves as a substrate, but instead becomes an inhibitory substrate. In this regard, Deak and Novak (1972) reported that sorbic acid, at high concentration, exerted an inhibitory effect on yeast species capable of metabolizing it, while at low concentration it acted as a substrate. Bacteria, which grow in the presence of 3000 mglkg of potassium sorbate, both in the presence or absence of glucose are listed in Table 6. Staphjlococcus aureus, Salmonella senftenberg and Pseudomonas vulgaris grew well in the presence of sorbic acid when used as the sole source of carbon. However, growth of these bacteria was retarded by the presence of glucose and sorbic acid together. Aerobacter aer_ggenes grew when 60 Table 6. The effect of the presence or absence of glucose (two percent) on the growth of bacterial cultures on yeast extract-peptone- agar media (pH 6. 0) containing 3000 mglkg of potassium sorbate and incubated at 25°C for six days Growth on yeast Growth on yeast extract- extract-tryptone- tryptone-agar + agar + 3000 mglkg 3000 mglkg potassium Bacteria potassium sorbate sorbate + glucose Staphylococcus aureus H .t Staphylococcus epidermidis i - Lactobaccilus plantarum - + Lactobacillus bulgaricus - + Leuconostoc mesenteroides : : Enterobacter aerogenes + - Stretococcus lactis i i Salmonella senftenberg ++ - Pseudomonas vulgaris ++ - Pediococcus cerevisiae - i Serratia marcescens + — - No growth i Weak growth + Moderate growth ++ Good growth +++ Vigorous growth 61 sorbic acid was the only source of carbon, but in the presence of both glucose and sorbic acid, this bacterium was completely inhibited. The previous explanation to this phenomenon, i.e. , the glucose inhibitory effect, in connection with molds may also apply here. The lactic acid bacteria did not grow when sorbic acid was used as the sole source of carbon (Table 6). Leuconostoc mesenteroides, however, grew weakly under similar conditions. Costilow et al. (1955) tested a number of lactic acid bacteria for their abilities to deplete sorbic acid and found that none could metabolize sorbic acid. Rehm et al. (1960) reported that Lactobacillus buchneri and Lactobacillus arabinosus (Lactobacillus plantarum) can grow in high con— centrations of sorbic acid but no metabolites of sorbic acid were observed. Results of this study are in agreement with those of previous reports. In contrast to these observations, Crowell and Guymon (1975) reported the isolation of several sorbic acid metabolites from sorbic acid-preserved red wine spoiled by lactic acid bacteria. These authors did not mention which lactic acid bacterium or bacteria was/were responsible for the metabolites. Metabolites of Sorbic Acid Sorbic acid metabolism by molds Selected molds were grown in liquid media containing sublethal con- centrations of sorbate, and the media analyzed for the presence of metabolites of sorbic acid. Figure 5 shows a partial GC chromatogram of the neutral metabolites of sorbic acid produced by the mold, Rhizgpus oligosporus. Four major peaks were observed on this chromatogram. Two of these metabolites (retention time, 13 and 13.2 minutes) gave mass spectral fragmentation patterns (Figures 6 and 7) which were found to be characteristic 62 QfHEXENOL 2,4-HexA0iEN0L RECORDER RESPONSE _ l J 5 10 15 RETENTION TIME, Min. Figure 5. A partial gas chromatogram of the neutral metabolites of sorbate-grown Rhizopus oligosErus 20 63 of trans-0-hexenol (Wick et al. , 1969; Kurogochi et al., 1975). This product was detected only in the late stages of growth of this mold. The other two major metabolites (retention time, 17. 5 and 18.1 minutes) exhibited similar fragmentation patterns (Figures 8 and 9) when subjected to mass spectrometric analysis. These fragmentation patterns are typical of 2, 0-hexadienol (sorbyl alcohol) (Kurogochi et al., 1975). One minor metabolite, however, was detected by injecting two ul of the final neutral extract of a 36 hour culture into the GC [MS system and scanned for the mass 96. The mass spectrometric fragmentation pattern of this minor metabolite was similar to that of 2, 0-hexadienal (sorbaldehyde) (Figure 10). A possible explanation for having two gas chromatographic peaks for both 0-hexenol and sorbyl alcohol is that sorbic acid could have originally existed in two or more forms, i.e. , the sorbic acid used in this study was not 100 percent trans-2-trans—0-hexadienoic acid. Instead, it may be a mixture of two or more isomers. However, it is also possible that microbial enzymes produce isomers of. sorbic acid as a means of detoxification. Tahara et al. (1975) reported that four species of Rhizopus accumulated sorbyl alcohol in their broth after 17 to 18 days of cultivation. Data from the present study indicate that selected species of Rhizopus did produce appreciable amounts of sorbyl alcohol in the early stages of growth. However, it was very difficult to detect any trace of sorbyl alcohol after five days of growth as judged by successive GC analyses. The apparent discrepancy between the two studies may have been the result of the differences in the mode of incubation, i.e, shaked versus still cultures. Sorbaldehyde, on the other hand, was detected only in the early stages of growth but disappeared quickly. From preliminary studies with 60 “my-cameo“ :o msmoNEm 5; “8323005 ocn ouanOm 9:53:00 $3505 a Eat 25203 33:55 9.: .08: acreage... _ocoxo.._-s Co .5583 $3.4 .o mean-3E 3:. o 0 cm. co. co av cw. -—®P i—VP . — u -% b -— - _— co 2. _ z. __ .2 low + Na . no. ice Ice 8132 T .8982-.. 8 . Ice. Ne BALLV'IEH ALISN 3.I.Nl 6'5 weaned. __o mamoNEm cuts noun—3005 och 0523 9:53:00 5:50.: m Eat couflofl 333:2: ~.m— .223 203.363 .ocoxon-s Co Eabuogm mum“... .N. 0.59...» 2E c3 62 as cm 8 8 _ m l. T ; 8 ca N ___ + I. ON Na 12. 8. law at n v No loc— emu-139% AllSNSlNl 66 . wagonmom=o mamoNEm 0:; 252385 new 0593 mEEmacou 5350:. n 50.: “63202 33356 m .5 .08: c2330.: _oco_omxo£-a.~ 06 Sateen—m 3.22 .a 0.59“. 2E 8N. . 8. 8 8 8 . . __ - , - _ w r. co _ _ _ 18 12. mm loo 2 B + IOQ was: . mm .8252; N s. -8. ALISNBLNI BAllV'lSH 67 macaw—momma nomads”. 5.3 noun—309.: can 059.3 9.33500 5352: m Eat “82209 . 3335:. 7»— .oEB c3330.; _oco__uuxo.._-=.~ Co E33009.“ mum—2 .m ocami . or. . 8. M 8. . 8 8 2. — - p , —- . _ , - , — _ l - L- - OD __ __ “a _ TON 3 .- w. 8 M. 2 I l + B 8 m . :8 a 3:2. . w. .8252; N 8 u. z. 18. A 68 magoamogfi named—E 5:3 noun—3005 new 3228 mEEmucou 53508 a 59: “.3203 _mcw_nmx0£ua.~ mo Eabuoam mums“ .3 0.59“. 3.: cc. o ow. _ .3 _ cm. _s :6“. 4.. _ _ — _ _ new 3 . 13 2 mm. + 59 lac . :8 _Eewfiqvu . E :8. ALISNSLNI BALLV'ISH 69 the same mold, it appears that sorbaldehyde is present as long as sorbic acid is in the reaction medium, after which it disappears. Similarly, species of Mucor produced sorbyl alcohol in appreciable amounts. The production of this alcohol started in the early stages of growth and it continued to be present until the fifth day of growth, after which it disappeared. During the later stages of growth, u—hexenol was detected. The production of this alcohol .was associated with a strong aroma which was reminiscent of an over-ripe banana. It appears that these molds produce relatively very high amounts of ll-hexenol. Sorbal- dehyde, also, has been detected as a minor metabolite in the early stages of growth. Tahara et al. (1977) tested 15 species of Mucor for their behavior on media containing sorbic acid. They reported that all but one of them grew in the presence of 500 mg/ kg of potassium sorbate and produced ll-hexenol in appreciable amounts. One species, however, failed to grow under the same conditions. ' Small amounts of u-hexenol were also detected by Tahara et al. (1977) in the cultures of Penicilliun crys_oge_num and Trichoderma viride. In spite of an intensive search for such metabolites in other Penicillium species and Trichoderma viride, no other traces of this compound have been detected. In an attempt to follow the fate of sorbic acid qualitatively in the presence of Rhizopus and Mucor species, sorbyl alcohol and sorbaldehyde were introduced into the media, individually, instead of sorbic acid. Sorbaldehyde was found to be very toxic to molds, even in small amounts and for this reason, only 100 mglkg of this compound was added to the media. Sorbyl alcohol was not as toxic as sorbaldehyde since both Mucor sp. and Rhizopus oliggsgrus grew well in the presence of 70 750 mglkg of the alcohol. This may explain why sorbyl alcohol was produced in appreciable amounts by the species of this genus. Troller and Olsen (1967) examined the antimicrobial activities of both sorbaldehyde and sorbyl alcohol and reported that sorbaldehyde possesses a fungi static activity much less than that of sorbic acid. They also reported that sorbyl alcohol had a fungistatic activity less than that of sorbic acid, especially at low pH. Tahara et al. (1977) reported tht sorbyl alcohol was more toxic than sorbic acid to the cells of M22159; A73. When sorbyl alcohol was introduced into the medium, a very high amount of ll-hexenol was produced, while only traces of sorbyl alcohol could be detected after two days of growth. A minor acidic metabolite was also shown to be present. The mass spectral fragmentation pattern of this compound was similar to that of the methyl ester of ll-hexenoic acid (Figure 11). Tahara et al. (1975) did not detect this compound in a medium containing either sorbic acid or sorbyl alcohol and inoculated with M gfi When sorbaldehyde, instead of sorbic acid, was added to a medium inoculated with either M2291 ER; or Rhizopus oligosporus, two compounds, a major and a minor metabolite, were present in the acidic fraction. Two other metabolites were present in the neutral fraction in addition to residual sorbaldehyde (Figure 12). The major acidic metabolite produced a mass spectral fragmentation pattern similar to that of the methyl ester of sorbic acid (Figure 13); hence, the original compound was thought to be sorbic acid. The minor acidic metabolite was identified as ll-hexenoic acid. The major neutral metabolite was identified as sorbyl alcohol, while the minor neutral metabolite was identified as ll-hexenol. 71 macommomzo msmouzm £33 252309: new .0383 39.5w. mficfiucou 82.99: m Eot beam—om. Eon 03:38:... do Loam... 350:. no Estevan mmmfi. .: ocamfi 0?. 0o. 03 ON— cop a1 co .9. ON _ _ _ - w i _ - _ 8 _ __ _ _ _ la ow. lav .2 + :8 2. LL , $132 3 :8 28 32.32% 22 ALISN 3.LN I SAILVWSH .Mm .8032 czocmnoo>£opfincom mo 333930.: 320... ocm .9339. soon mo Emcmoumsoccu mom 33ng < .2 0.59m .52 3.2:. zofizmpmm a E 2 _ _ lug), aw axvouaxaH—aw _ - Il‘ SSNOdSBH uaoeooau 7 axvaaos-aw I 1onalovxaH-v‘z 1VN3IOVX3H-t‘z 73 .Mm c833. 5:3 noun—zoos: ocm oo>coEmnc8 GEES—coo £352: a Eot U323; Bum 05.3w mo .353 350:. o Estevan 33.. .3 0.39m 2:. . o: om. . :9 8 8 3 cm _ p . . . . .- . . _ cc H [ON 3 .1 W. Tee M ON— 3 .2 8 18 m .. a. Tom We . so .1 £252 :8. u. 28 2&8 22 E 711 Based on the data gathered in this study on the degradation of sorbic acid by some species of Mucor and Rhizopus and from the data presented by Tahara et al. (1975) , the following scheme is suggested as a possible mechanism for the degradation of sorbic acid. 7 CH3-CH=CH=CH=CH-COOH ' r CH3-CH=CH-CH2-CH2-COOH Sorbic acid ll—hexenoic ac id " i CH -C H=CH-CH z-CHz-CHZOH 3 ll-hexenol CH3-CH=CH-CH=CH—CHO + CH3-CH=CH-CH=CH—CH20H Sorbaldehyde Sorbyl alcohol This scheme can be explained as follows: 1. The carboxyl group of sorbic acid is reduced to aldehyde which functions as a transitory compound, i.e. , it is reduced further to alcohol as soon as it is formed. 2. Sorbyl alcohol is reduced further to ll-hexenol. 3. It is believed that, at least part of the ll-hexenol is oxidized to the corresponding acid. ll. It is not known at the present if sorbic acid, in part, is reduced to ll-hexenoic acid. The reduction of the a, 8 unsaturated double bond of sorbyl alcohol appears to be a slow process which indicates that the enzymes responsible for this process are inducible. Tahara et al. (1977) concluded from a study on resting cells of a M _s‘p_._ previously grown in the presence of several compounds, such as sorbic acid and hexanoic acid, that the enzymes responsible for this conversion tend to be inducible ones. 75 It is not known, however, whether ll-hexenoic acid can be formed directly from sorbic acid. Therefore, the possibility exists that this acid is an intermediate in the reduction of sorbic acid to ll-hexenol, or it is formed only upon the oxidation of lI-hexenol as was shown when sorbyl alcohol was introduced into the media. Tahara et al. (1977) investigated the metabolic pathway of the conversion of sorbic acid to ll—hexenol by M9504 s2; and concluded that the reduction process pro- ceeded via two independent reactions: CH3-CH=CH-CH=CH—COOH %CH3-CH=CH-CH=CH-CH20H Sorbic acid Sorbyl alcohol CH3-CH=CH—CH=CH—CH20H ;CH3-CH=CH-CH2-CH2-CHZOH Sorbyl alcohol ll-hexenol They also suggested that sorbic acid is first converted to sorbaldehyde, which in turn is converted to the corresponding alcohol. The data in the present study indicate that sorbaldehyde is a transitory metabolite since it is reduced as soon as it is formed. Sorbaldehyde, however, was not produced when sorbyl alcohol was used as a substrate, supporting the contention that this compound is an intermediate in the conversion of sorbic acid to sorbyl alcohol. Philip et al. (1963) have demonstrated the reduction of n-butyric acid to n-butanol by cultures of Aspergflus niger growing in a medium con- taining glucose and butyric acid. Similarly, Walker et al. (1963), reported the reduction of isobutyric and isovaleric acids to the corresponding alcohols. In both cases, the presence of glucose was necessary for the reduction pro— cess. More recently, the reduction of sorbic acid by resting cells of Mucor sp. was 76 investigated by Tahara et al. (1977). Their results indicated that the reduction process required an energy source such as glucose. Results in the present study on growing cells of Rhizopus oligosporus and M E& indicate that both a carbon and energy source are necessary for the reduction of sorbic acid, because both molds do not grow at all in the absence of a carbon source other than sorbic acid. Penicillium sp. , Penicillium roqueforti, AsErgillus sp., and Trichoderma 1i_ri_d_e_ were found to effectively degrade sorbic acid in culture media. It was surprising initially when the chromatograms of both the neutral and the acidic fractions showed no signs of any metabolite. Hence, it was thought that sorbic acid was rapidly metabolized to CO2 and H20. The hydrocarbon, 1,3-pentadiene, however, has been shown to result from the decarboxylation of sorbic acid by Penicillium rgueforti (Marth et al., 1966). The above mold was observed to produce a strong aroma in the presence of sorbic acid. It was assumed that this aroma may have resulted from the production of 1,3-pentadiene. For this reason, a special procedure to entrap 1,3-pentadiene was developed in order to confirm the identity of compound(s) responsible for the strong aroma. This substance, when analyzed by mass spectrometry produced a fragmentation pattern similar to that of 1,3-pentadiene (Figure M). It appears that the reason for the failure to detect the 1,3-pentadiene in the early attempts is that all or most of it is lost during the evaporation of the organic solvents, since 1,3-pentadiene has a boiling po'nt of 112°C. It was not determined whether all of the metabolized sorbic acid was decarboxylated to 1,3-pentadiene, or whether part of it was metabolized through a different pathway, such as B-oxidation, to CO2 and H20. It is .Mm 53:23.3.” 5.: US$309: _ucm 3.33m mc_c_3cou £2.35 M So...— Uouflofl ocflomucoaum; so Eacuuodm mmmé .3 95.9.... 2:. .m. .w. 4. ._. a . , I , Do _ __ ___ Tow _ :2. 766 we 8 .8 r8 megs. . .2 23:5: Im~ +8 18. BALLV'IEH ALISN aim 78 also not known whether 1,3-pentadiene itself is metabolized by the molds or whether it is lost to the atmosphere. In an attempt to gain more insight into the decarboxylation process, sorbic acid was used as the sole source of carbon. This system supported the growth of the previously mentioned molds and 1,3-pentadiene was stil produced. Geotrichum candidum was found to grow vigorously on the YPG medium containing sorbic acid. Sorbic acid slowly disappeared from the medium with the simultaneous formation of the ll—hexenoic acid as a major metabolite (Figure 15). However, it is not known whether ll-hexenoic acid is metabolized by the mold or not. The first possibility is supported by the fact that this acid was no longer detected after four to five days of growth. A precise search for sorbaldehyde, sorbyl alcohol, u-hexenol and 1,3-pentadiene revealed that such compounds are not produced by Geotrichum candidum. It is also not known whether all or just part of the depleted sorbic acid is reduced to ll-hexenoic acid. A minor compound was detected and was shown by mass spectrometry to be the ethyl ester of sorbic acid (Figures 15 and 16). Kurogochi et al. (19711), reported that Geotrichum sp. partially converted sorbic acid into ethyl sorbate and ll-hexenoic acid. Crowell and Guymon (1975) isolated and characterized ethyl sorbate as one of the compounds derived from sorbic acid which had been added to a red table wine subsequently spoiled by lactic acid bacteria. They believed that it was formed by simple ethyl esterification of the sorbic acid in the wine. An acid medium with a pH less than 4.0 accounts for ethyl sorbate formation. It is not known whether this hypothesis would be applicable in our case, especially since the final pH of the media was 8:26ch Eacutuoou czocmnoumngom . mo 3.3.0939: 823 new .9530: Lyon co Emcmoumsocco 00 .m— 0.59“. .52 £2: zozzwhwm 79 9 b w. I m p o L. l l o.— axvauos 151-14.: 3 BLVBUOS'GW --=: % 3 SLVONBXSH-OW -—"'-=:— .11 J3 asnoclsau uaoeooaa 1N3 A105 80 .2anch 5350.330 5:3 noun—38$ .ocm San—com 9:53.30 8359.. a ES... .6323. oumbcom 350 mo Eacuuoam mums. .o— 0.39". 0?. oo. 9: cm. 8. 4e 8 e cm P _ P . 66 = 18 Wu 1 r w. on. S. m. a: . . . TOO I + . N a. . co §§2 w m. 22.8 _Em 5 8 B 18. M. 81 above five. It is worth mentioning that some of the yeasts investigated in this study also produced ethyl sorbate in trace amounts. It is believed that other metabolites may have enhanced the esterification of sorbic acid in the presence of ethyl alcohol, which may have been produced from glucose. Sorbic acid metabolism by yeasts Preliminary studies with Candida liglytica, Debaryomyces membranae- faciens, and Saccharom1ces occidentalis indicate that some yeast species of different genera reduce the double bond of sorbic acid in the on, 8 position producing ll—hexenoic acid. Preliminary quantitative data indicate, however, that this acid does not account for all of the consumed sorbic acid. It is possible that part of sorbic acid is oxidized through a B-oxidization pathway to CO2 and H20. In an attempt to understand more about the fate of sorbic acid, some of the commercially available metabolites were also used as substitutes for sorbic acid. Sorbyl alcohol was found to be more toxic to yeast cells (Candida lipolytica) than sorbic acid. When used at a concentration of 1000 mg/kg, sorbyl alcohol completely inhibited the growth of Candida lialytica, even when a heavy inoculum of the microorganism was used. For this reason, the sorbyl alcohol concentration was reduced to 300 mglkg in subsequent experiments. CC analysis of the metabolites of sorbyl alcohol indicated the formation of sorbic acid, ll-hexenol, ll-hexenoic acid, and sorbaldehyde (Figure 17). Sorbaldehyde was also used as a smstrate, but in small amounts (100 mglkg) since it was found to be very toxic to the yeast, Candida lipolytica. Troller and Olsen (1967) who examined a number of sorbic RECORDER RESPONSE 82 III p. ‘ D C O ”I’ . s a s . is 2 g 3 m x = 5?! m l I .- V’ V“ 3 . °' 2 $ III A x Ill 3‘ h 0 2 I l 1 I L l 5 10 15. 20 RETENTION TIME, Min. Figure 17. A partial gas chromatogram of both neutral and acidic metabolites of sorbyl alcohol-grown Candida molytica 83 acid derivatives for their antimicrobial activities reported that sorbal- dehyde possessed a fungistatic activity much superior to that of sorbic acid itself. On the other hand, sorbyl alcohol possessed a fungistatic activity much less than that of sorbic acid, especially when used at low pH. On the other hand, Tahara et al. (1977) reported that sorbyl alcohol was more toxic to My; 39; than was sorbic acid. The preliminary findings in the present study seem to support the conclusions of Troller and Olsen (1967) regarding sorbaldehyde, while agreeing with those of Tahara et al. (1977) with regards to sorbyl alcohol. Qualititatively speaking, the compounds produced upon the metabolism of sorbyl alcohol were also detected when sorbaldehyde was introduced into the medium. The results of these studies on the growth of some yeasts in media containing sorbic acid, sorbyl alcohol or sorbaldehyde are summarized as follows: CH3-CH=CH-CH=CH-COOH 4 CH3-CH=CH-CH2-CH2-COOH Sorbic acid lI-hexenoic acid I , I CH3-CH=CH-CH2-CH2-CH20H u-hexenol CH3-CH=CH-CH=CH-CHO < >CH3-CH=CH-CH=CH-CHZOH Sorbaldehyde Sorbyl alcohol 811 1. Sorbic acid is reduced to lI-hexernoic acid and it is believed that this is only a partial reduction. 2. Sorbyl alcohol is partially oxidized to sorbaldehyde and the latter is in turn oxidized to sorbic acid. 3. Sorbaldehyde is partially reduced to sorbyl alcohol and the latter is further reduced to ll-hexenol. II. The possibility exists that lI-hexenol is oxidized to ll-hexenoic acid. From preliminary growth studies with Debaryomyces membranaefaciens, Saccharomyces occidentalis and Saccharomyces roxii, in addition to Candida Iimlytica, it was found that all of these yeasts shared the same pathway for the degradation of sorbic acid, i.e., they reduce sorbic acid to ll-hexenoic acid. Trace amounts of ethyl sorbate were detected in some instances. This indicates a similarity between the yeast-like mold, Geotrichum candidum, and the fore-mentioned yeasts. It is not known why these yeasts and Geotrichum candidum are unable to reduce sorbic acid to alcohols or aldehydes while possessing the ability to oxidize both the alcohol and aldehyde to sorbic acid. The reduction of aldehydes and ketones to the corresponding alcohols by yeasts in a sugar medium was described many years ago by Nord (1919) and Newberg and Nord (1919), as cited by Walker et al. (1963). Farmer et al. (1959) have reported the conversion of certain aromatic acids to aldehydes and alcohols by the wood fungus, Polystictus versicolor. Walker et al. (1963) and Philip et al. (1963) have also demonstrated the ability of Aspergillus niger to reduce lower fatty acids to the corresponding alcohols in the presence of glucose. Studies with two different cultures of Hansenula wengei indicate that this species produces 1,3-pentadiene from sorbic acid. It is also interesting 85 to note that this yeast is very effective in degrading sorbic acid when present in sublethal levels. It is not known whether all or part of sorbic acid is decarboxylated to 1,3-pentadiene. It is also not known if the production of 1,3—pentadiene from sorbic acid is common to other yeasts. Sorbic acid metabolism bybacteria Studies with aerobic bacteria were limited to Lactobacillus bulgaricus, Lactobacillus plantarum, Pseudomonas fluorescens and Pediococcus cerevisiae. These cultures were grown on trypticase soy broth (pH 6.0) fortified with 1000 mglkg of potassium sorbate. All of the above cultures grew well, but no significant loss of sorbic acid was observed, except with Pseudomonas fluorescens. In this case, up to 28 percent of the sorbate was lost in three days. In another series of experiments, different concentrations of potassium sorbate were used with both growing and resting cells of Lactobacillus bulgaricus, Lactobacillus plantarum, Pediococcus cerevisiae and Pseudomonas fluorescens. GC analysis failed to detect any metabolites in the media under the conditions used in the experiment. These findings agree with those reported by Reha et al. (19611) who reported that two lactic acid bacteria showed no signs of utilization of sorbic acid. These authors also reported that Pseudomonas fluorescens metabolized sorbic acid through respiration. Costilow et al. (1955) reported similar conclusions with respect to the lactic acid bacteria. Three cultures of Clostridia, Clostridia perfrm, Clostridia sporogenes and Clostridia tertium were grown on sorbate-containing media. Clostridia sporogenes and Clostridia tertium were able to convert sorbic acid to ll-hexenoic acid in very high amounts as indicated by GC analysis (Figure 18) . Sorbyl alcohol and lI-hexenol were also produced in moderate 86 mocomocomm 222.536 csochoumncom mo 3509308 I 0.28 Dan .9539. Son mo EmcmoumEoccu mmm .1...th < .3 9.32“. .55 .22: 295mm”. 3 S m A _ . ._. 3 gig I8 no 3 3 0 no G 3 3 ..V no w. G. 8 d . X W 0 W a 8 N a w .V S S g m l u. 3 .7. m m .V a u a u U. D. 0 a D. I. N... p a m. w a u p m. 87 amounts, especially when glucose was added. Trace amounts of sorbal- dehyde was also detected. The exact route(s) through which these metabolites were produced is/are not known, but it is believed that two pathways do exist; one is the reduction of the on, B-double bond of sorbic acid producing ll-hexenoic acid which in turn is reduced to ll-hexenol. Thauer et al. (1968) reported that Clostridia klgveri when grown on media containing crotonate as the sole source of carbon produced butyrate as the major product. Butanol, however, was produced in only negligible amounts. The production of high amounts of ll-hexenol in the present study may be ascribed to the presence of glucose. The other pathway involves the reduction of the carboxyl group of sorbic acid to sorbyl alcohol. Davis (19“) reported the reduction of n-butyrate to l-butanol by Clostridium acetobutylicum in the presence of glucose. Barker (1956) postulated that butyraldehyde is formed through the action of aldehyde dehydrogenase and is subsequently reduced to butanol under the influence of alcohol dehydrogenase. CHB-CH=CH-CH=CH-COOH —; CHB-CH=CH-CH2-CH2-COOH Sorbic acid 11- hexenoic acid I CHB-CH=CH-CH2-CH2-CH20H ll-hexenol CH3-CH=CH-CH=CH-CHO 4CH3-CH=CH-CH=CH=CH20H Sorbaldehyde Sorbyl alcohol Clostridium perfringens, although it grew well on sorbate-containing media, did not produce any of the above metabolites in measurable amounts. 88 Fate of Labelled Sorbic Acid In order to confirm that the metabolites which have been isolated from the media and identified by mass spectrometric analysis were indeed produced from sorbic acid, labelled sorbic acid ([U-mCI-sorbic acid) was introduced into a medium containing cells of Rhizopus stolonifer in order to monitor the radioactivity in the metabolites. However, due to the limited amount of [U-MC1-sorbic acid available, only one metabolite, sorbyl alcohol, was chosen for the study. Thin layer chromatography plates were used to isolate this metabolite from the solvent extract of the medium. Sorbyl alcohol was monitored for the presence of radioactivity and at least 300 counts per minute (cpm) were recorded. Further confirmation was also achieved by extracting this metabolite from the silica gel on the TLC plate and analyzing the extract by CC. This metabolite had a retention time comparable to that of a standard sorbyl alcohol sample, thus confirming that the precursor of sorbyl alcohol was sorbic acid. It was assumed that the other metabolites also originated from sorbic acid. SUMMARY AN D CONCLUSIONS Results of the screening tests on three microbial groups, molds, yeasts, and bacteria, confirm the previous reports regarding the wide spectrum antimicrobial activity of sorbic acid. The microbes which have been tested, however, exhibited a wide variation within species of a single genus and, in some cases such variations were observed within cultures of the same species. 0f the molds tested, most of the cultures belonging to the two genera, Aspergillus and Penicillium, were generally found to' be more resistant than the other cultures. Of these cultures, Minus flavus, Asper- gillus niger, Penicillium _s_&_, Fusarium sp., Geotrichum candidum and Trichoderma 1159.2. were also found to be resistant to sorbic acid. All of the above mold cultures grew well in the presence of 3000 mglkg of potassium sorbate and apparently utilized sorbate as a carbon source. Unexpectedly, the presence of glucose increased the inhibitory action of sorbate against most of the molds which utilized sorbate as a carbon source. This phenomenon was designated as the glucose inhibitory effect. Of the yeasts tested, Candida lipolytica, Hansenula saturnus, Brettanornjces claussenii and Saocharomlces occidentalis exhibited relatively high resistance to sorbate. The above yeasts, especially Candida lipglytica grew vigorously on a medium containing 3000 mglkg of sorbate at pH 5. 5. Some of the yeasts utilized sorbate as a'carbon source. Candida lipglytica and Hansenula californica grew well in the 89 90 presence of 3000 mglkg of potassium sorbate at a pH of 6.0. Once again the "glucose inhibitory effect" was observed. Most of the bacteria tested were found to be resistant to at least 500 mg/ kg of potassium sorbate at pH 5. 5. The majority of them were inhibited by potassium sorbate concentrations ranging from 1000 to 2000 mg/ kg. Some lactic acid bacteria and a strain of Pseudomonas fluorescens were found to require more than 3000 mglkg of potassium sorbate in highly nutritious media. Salmonella senftenberg, Pseudomonas vulgaris and Staphllococcus aureus grew well in media containing sorbate as the only carbon source. Results of the studies on metabolites indicate that sorbic acid is converted to ll—hexenol by the species of the two genera, Rhizopus and Mucor and some species of Fusarium. Observations on these molds suggest that sorbic acid, when present in sublethal concentrations, is reduced to sorbaldehyde Which is immediately reduced to sorbyl alcohol; sorbyl alcohol is reduced further to ll-hexenol. Our observations, also, suggest that the first and second steps are catalyzed by constitutive enzymes while the latter step is catalyzed by inducible enzyme(s). Penicilli'um and Aspergillus species and Trichoderma viride produced l, 3-pentadiene. Some yeasts and a yeast-like mold, Geotrichum candidum, reduced sorbic acid to ll-hexenoic acid. Ethyl sorbate was also detected but in relatively small amounts. Studies on aerobic bacteria revealed that the metabolism of sorbic acid by aerobic bacteria is not as common as it is the case with molds. 91 Only Pseudomonas fluorescens showed some loss of sorbic acid, but no signs of metabolites were observed. Clostridium sporogLenes and Clostridium tertium metabolized sorbate mainly to ll—hexenoic acid, ll—hexenol and sorbyl alcohol. Apparently, metabolism of sorbate did not occur when Clostridium perfringens was used, although they grew well in the presence of sorbate. The use of labelled sorbic acid enabled the confirmation of the assumption that the metabolites come from sorbic acid. RECOMMENDATIONS FOR FURTHER STUDIES The metabolism of sorbic acid by various microorganisms has been studied by GC and CC IMS techniques. However, many questions per- taining to the overall metabolism of sorbic acid remain unanswered. Some aspects requiring further study can be classified under the following headings: 1. Toxicology of sorbic acid and metabolites Since a number of metabolites has been shown to be produced from sorbic acid by certain microorganisms under certain conditions, some attention should be directed toward the toxicological aspects of these metabolites. 2. Metabolism of sorbic acid in human body Early reports indicate that sorbic acid is metabolized in a similar way to that of the fatty acids, i.e. , to H20 and CO2 or to ketone bodies. This study indicates that a wide diversity among microorganisms does exist as far as sorbic acid metabolism is concerned. It would be of interest to see if mammalian enzymes such as liver dehydrogenases would convert sorbic acid to sorbyl alcohol or sorbaldehyde. 3. En zymatic studies Some efforts should be directed toward enzymes which are responsible for the degradation of sorbic acid in the different microbial groups. ‘1. The use of food systems This study was conducted in what might be considered model systems (cultural media). Some work should be done in food systems or in a model 92 93 system which simulates the food system of concern, and inoculating this system with microbe(s) of potential hazard. 5. The fate of the metabolites It would be interesting to follow the fate of the metabolites by using them as substrates to see whether they are metabolized by the microbes, whether they accumulate in the media or whether they are lost to the atmosphere. 6. Glucose inhibitory effect This phenomenon needs to be further investigated in order to ascer- tain what concentrations of glucose or other carbohydrate substance (e.g. , fructose, sucrose) brings about this effect. 7. Sorbate-nitrite interaction It would be of interest to study what effect the presence of nitrite may have on the decomposition of sorbic acid in terms of the rate of decomposition and the type of metabolites which result from this decom- position. 8. Sorbate balance experiments The extraction procedure used for the isolation of sorbate metabolites from the media needs to be assessed in terms of extraction selectivity. The use of labelled sorbic acid would facilitate this study and provide information as to whether the metabolites identified in the present study were the major decomposition products of sorbate. APPENDICES APPENDIX 1 Table A. 1. Mold cultures used for screening tests Mold Code 1. Aspergillus flavus NRRL 2999 1 2. Aspergillus flavus NRRL 6550 3. Aspergillus niger (Beneke) 25 ll. Aspergillus oryzae ATCC 11601 26 5. Aspergillus sp. 7 6. Alternaria sp. 111 7. Alternaria tenuis NRRL 2169 16 8. Calvatia gigantica 21 9. Cladosporium cladosporiodes QM 1189 8 10. Cladosporium cladosporiodes QM 91185 10 11. Fusarium sp. 2 12. Fusarium roseum 9 13. Fusarium oxysporum NRRL 19113 15 111. Fusarium sp. 20 15. Geotrichum candidum (Beneke) 29 16. Geotrichum candidum I 23 17. Mucor humiculus 28 18. Mucor pusillus 27 19. Mucor sp. 24 20. Penicillium janthinellum NRRL 2016 13 21. Penicillium oxalicum NRRL 790 12 22. Penicillium roqueforti 23. Penicillium digitatum (Beneke) 211. Penicillium italicum (Beneke) 25. Penicillium sp. 3 26. Rhizopus oligosporus ATCC 22959 ll 27. Rhizopus oligosporus NRRL 2710 5 28. Rhizopus stolonifer 5 29. Trichoderma viride 11 State University. 9“ 1Code designated by the Foo'd Microbiology Laboratory at Michigan APPENDIX 2 Table A.2. Yeast cultures used for screening tests Yeast Code1 1. Brettanomyces claussenii NRRL-Y-llllll y-29 2. Candida krusei y-2 3. Candida lipolytica y-l ll. Candida steatolytica y—31 5. Candida utilis NRRL-Y-900 y-39 6. Debaryomyces membranaefaciens NRRL-Y-989 y-ll 7. Endomycopsis bispora y- 16 8. Endomycopsis bispora ATCC 111628 y-18 9. Endomycopsis selenospora NRRL Y-1357 y—17 10. Hanseniaspora valbyensis var. guilliermondii ATCC 106 y-30 11. Hansenula anomala y-6 12. Hansenula anomala y—GD 13. Hansenula californica NRRL Y-11125 y-19 Ill. Hansenula saturnus y-7 15. Hansenula wingei ATCC 111355 y-8 16. Hansenula wingei ATCC 10356 y-9 17. Nadsonia fulvescens NRRL Y-991 y-20 18. Pichia fermentans y-5 19. Rhodotorula glutinis var glutinis y-36 20. Rhodotorula rubra y-23 21. Rhodotorula spp. (Beneke Y-9) y-22 22. Saccharomyces carlsbergensis ATCC 9080 y-35 23. Saccharomyces cerevisiae (dry yeast) y-2!l 211. Saccharomyces cerevisiae 9-25 25. Saccharomyces cerevisiae var. ellipsoideus y-1ll 26. Saccharomyces kluyveri y-12 27. Saccharomyces kluyveri (C-26) y-15 28. Saccharomyces oleaginosus y-15 29. Saccharomyces rouxii ATCC 2619 y-26 30. Saccharomyces sp. y-llo 95 96 Table A.2 (continued) 1 Yeast Code 31. Saccharomyces occidentalis ATCC 2320 y-28 32. Schizosaccharomyces japonicus var versatilis y-38 33. Schizosaccharomyces octosporus y-10 311. Schizosaccharomyces pombe y-11 35. Sborobolomyces coralliformis ATCC 16039 y-3ll 36. Torulopsis sake NRRL Y—1622 y-21 37. Torulopsis sphaerica y-3 38. Trichosporon cutaneum ATCC 13005 y- 33 39. Trichosporon fermentans NRRL Y-11192 y-32 110. Trigonopsis variabilis ATCC 10679 y-27 State University. 1Code designated by the Food Microbiology Laboratory at Michigan APPENDIX 3 Table A.3. Bacterial cultures used for screening tests Bacteria PSSmFPP“ wNNNNNNNNNN—I-I-I—t—I—a-n—I—o—a Presswswvrsrssswsrvrss Aerobacter aerogenes Acetobacter aceti Bacillus cereus Bacillus coagulans Bacillus polymyxa Bacillus subtilis Escherichia coli Lactobacillus brevis Lactobacillus bulgaricus Lactobacillus casei Lactobacillus plantarum Leuconostoc mesenteroides Microococcus flavus Pediococcus cerevisiae Pseudomonas fluorescens Pseudomonas fluorescens Pseudomonas putida ATCC 15070 Pseudomonas putida ATCC 27212 Pseudomonas schuylkilliensis ATCC 15916 Pseudomonas schuylkilliensis ATCC 15917 Pseudomonas vulgaris Pseudomonas fragi Salmonella typhimurium Salmonella senftenberg Salmonella paratyphi Salmonella pollorum Sarcina lutea Serratia marcescens Staphylococcus aureus Staphylococcus epidermidis 97 BIBLIOGRAPHY BIBLIOGRAPHY Alderton, O. and J.C. Lewis. 1958. Determination of sorbic acid and its disappearance from pickle brines. Food Res. 23:338. Anderson, T.E. and R.N. Costilow. 1963. Site of inhibition of yeast respiration by sorbic acid. Bacteriol. Proc. 14. Anonymous. 1978. Sorbic acid and potassium sorbate for preserving food freshness and marketing quality. Monsanto Co. , St. Louis, MO. Arya, 5.5.; K. Vidyasagar and 0.8. Parihar. 1978. Degradation of sorbic acid in aqueous systems. Proc. First lnd. Conv. Food Sci. Technol., p. 75. Ayres, J.C.; J.O. Mundt and W.E. Sandine. 1980. Microbiology of Foods. W.H. Freeman and Company. San Francisco. Azukas, J.J.; R.F. Costilow and H.L. Sadoff. 1961. Inhibition of alcoholic fermentation by sorbic acid. J. Bacteriol. 81:189. Bell, T.A.; J.L. Etchells and A.F. Borg. 1959. Influence of sorbic acid on the growth of certain species of filamentous fungi. J. Bacteriol. 77:573. Beneke, E.S. and F.W. Fabian. 1955. Sorbic acid as a fungistatic agent at different pH levels for molds isolated from strawberries and tomatoes. Food Technol. 9:086. Beneke, E.S. and K.E. Stevenson. 1978. Classification of food and beverage fungi. In: "Food and Beverage Mycology." L.R. Beuchat (Ed.), Avi Publishing Co., Inc., Westport, Conn. Bolin, H.R.; A.D. King and A.E. Stafford. 1980. Sorbic acid loss from high moisture prunes. J. Food Sci. l15:111311. Bradley, R.L. 1960. The effect of potassium sorbate on some organisms associated with cottage cheese spoilage. M.S. Thesis. Michigan State University, East Lansing, Michigan. Bradley, R.L.; L.G. Harmon and C.M. Stine. 1962. Effect of potassium sorbate on some organisms associated with cottage cheese spoilage. J. Milk Food Technol. 25:318. Brock, T.D. 1970. Biology of Microorganisms. Prentice-Hall, Inc., Englewood Cliffs, NJ. 98 99 Costilow, R.N. 1957. Sorbic acid as a selective agent for cucumber fermentations. lll. Evaluation of salt stock from sorbic acid treated cucumber fermentations. Food Technol. 11:591. Costilow, R.N.; F.M. Coughlin; E.K. Robbins and W.T. Hsu. 1957. Sorbic acid as a selective agent in cucumber fermentations. ll. Effect of sorbic acid on the yeast and lactic fermentations in brined cucumbers. Appl. .‘licrobiol. 5:373. Costilow, R.N.; W.E. Ferguson and 5. Ray. 1955. Sorbic acid as a selective agent in cucumber fermentations. I. Effect of sorbic acid on microorganisms associated with cucumber fermentations. Appl. Microbiol. 3:341. Crosby, N.T. and R. Sawyer. 1976. N-Nitrosamines: A review of chemical and biological properties and their estimation in foodstuffs. Adv. Food Res. 22:1. Davis. R. 1902. Studies on the acetone-butyl alcohol fermentation, 2. Intermediates in the fermentation of glucose by Clostridium acetobutylicum. Biochem. J. 36:582. Deak, T. and E.K. Novak. 1970. Effect of sorbic acid on the growth of yeasts on various carbohydrates. Acta Microbiol. Acad. Sci. Hung. 17:257. Deak, T. and E.K. Novak. 1972. Assimilation of sorbic acid by yeasts. Act. Alimentaria. 1:87. Deak, T.; M. Tuske and E.K. Novak. 1970. Effect of sorbic acid on the growth of some species of yeast. Acta Microbiol. Acad. Sci. Hung. 17:237. Deuel, H.J.; R. Alfin-Slater; C.S. Weil and H.F. Smyth. 1950a. Sorbic acid as a fungistatic agent for foods. I. Harmlessness of sorbic acid as a dietary component. Food Res. 19:1. Deuel, H.J.; C.E. Calbert; L. Anisfeld; H. McKeehan and H.D. Blunden. 1950b. Sorbic acid as a fungistatic agent for foods. ll. Metabolism of or, B-unsaturated fatty acids with emphasis on sorbic acid. Food Res. 19:13. Difate, V.G. 1977. Microbial mutagenicity evaluation of sorbic acid/sodium nitrite reaction products. Report smeitted to the Expert Panel on Nitrites, Nitrates and Nitrosamines, USDA, Washington, D.C. Difco Laboratories. 1953. Difco Manual, llth Ed. Detroit, Michigan. Elbereir,M.;C.Kurg;K.CrunertandC.Franzk.1975.1nfluence of preservatives on aflatoxin formation by Aspergillus flavus. Nahrung. 19: k3-kll. IOO Emard, L.V. and R.H. Vaughn. 1952. Selectivity of sorbic acid media for the catalase negative lactic acid bacteria and clostridia. J. Bacteriol. 63:087. FAO IWHO Expert Committee on Food Additives. 1967. Toxicological evaluation of some antimicrobials, antioxidants, emulsifiers, stabilizers, flour-treatment agents, acids and bases. FAO Nutr. Mtg. Rep. Ser. No. 00 A, B, C, WHO/Food Add./67.29. Farmer, V.C.; M.E.K. Henderson and J.D. Russell. 1959. Reduction of certain aromatic acids to aldehydes and alcohols by Polystictus versicolor. Biochem. Biophys. Acta. 35:202. Fiddler, W. 1975. The occurrence and determination of N—nitroso compounds. Toxicol. Appl. Pharmacol. 31:352. Freese, E.C.; W. Shell and E. Galliers. 1973. Function of lipophilic acids as antimicrobial food additives. Nature 201:321. Gaunt, l.F.; K.R. Butterworth; J. Hardy and 5.0. Grugolli. 1975. Long term toxicity of sorbic acid in the rat. Food Cosmet. Toxicol. 13:31. Gershenfeld, L. 1968. Antimicrobial agents in cosmetic preparation. Am. Perf. Cosmet. 78:55. Gooding, C.M. 1905. Process of inhibiting growth of molds. U.S. Patent, 2, 379, 290. Gooding, C.M.; D. Melnick; R.L. Lawrence and F.H. Luckmann. 1975. Sorbic acid as a fungistatic agent for food. IX. Physiochemical considerations in using sorbic acid to protect foods. Food Res. 20:639. Gray, J.l. 1976. N-Nitrosamines and their precursors in bacon. A review. J. Milk Food Technol. 39:689. Gray, J.l, and C.J. Randall. 1979. ~ The nitrite/N—nitrosamine problem in meats: an update. J. Food Prot. 02:168. Hansen, J.D. and M.D. Appleman. 1955. The effect of sorbic, propionic, and caproic acids on the growth of certain clostridia. Food Res. 20:92. Harada, K.; R. Higuchi and l. Utsumi. 1968. Studies on sorbic acid. IV. Inhibition of the respiration in yeast. Agr. Biol. Chem. 32:900. Hayatsu, J.; K.C. Chung; T. Kada and T. Nakajima. 1975. Generation of mutagenic compound(s) by a reaction between sorbic acid and nitrite. Mut. Res. 30:017. Hendy, R.J.; J. Hardy; l.F. Gaunt; I.S. Kiss and K.R. Butterworth. 1976. Longterm toxicity studies of sorbic acid in mice. Food ‘ Cosmet. Toxicol. 10:381. 101 Hoffman, C.; T.R. Schweitzer and G. Dalby. 1939. Fungistatic properties of the fatty acids and possible biochemical significance. Food Res. 0:539. Ivey, J.F. and M.C. Robach. 1978. Effect of potassium sorbate and sodium nitrite on Clostridium botulinum growth and toxin production in canned comminuted pork. T. Food Sci. 03:1782. Ivey, F.J.; K.H. Shaver; L.N. Christiansen and R.B. Tompkin. 1978. Effect of potassium sorbate on toxinogenesis by Clostridium botulinum in bacon. J. Food Prot. 01:621. Kada, T. 1970. DNA-damaging products from reaction between sodium nitrite and sorbic acid. Ann. Rep. Natl. Inst. Gen. 20:03. Klis, J.B.; L.D. Witter and Z.J. Ordal. 1959. The effect of several antifungal antibiotics on the growth of common food spoilage fungi. Food Technol. 13:120. Kurogochi, S.; S. Tahara and J. Mizutani. 1970. Fungal metabolites of sorbic acid. Agr. Biol. Chem. 38:893. Kurogochi, S.; S. Tahara and J. Mizutani. 1975. Fungal reduction of C5 or, B-unsaturated carboxylic acid. Agr. Biol. Chem. 39:825. Luck, E. 1976. Sorbic acid as a food preservative. lnt. Flavors Food Additive. 7:122. Luckas, E.M. 1963. Zur kenntnis der antimikrobiellen wirkung der sorbinsaur. ll. Die wirkung der sorbinsaure auf Aspergillus niger van tieghem und andere schimmelpilze. Zbl. Bakt.Abt. ll Bd. 117:510. McCarthy, T.J.; C.R. Clarke and J.A. Myburgh. 1973. Antimicrobial effectiveness of stored sorbic acid solutions. Cosmet. Perf. 88:00. McCarthy, T.J. and P.F.K. Eagles. 1976. Further studies on glass-stored sorbic acid solutions. Cosmet. Toilet. 91:33. Marth, E.H.; C.M. CaPP; L. Haenzahl; H.W. Jackson and R.V. Hussong. 1966. Degradation of potassium sorbate by Penicillium species. Melnick, D. and F.H. Luckmann. 1950a. Sorbic acid as a fungistatic agent for foods. I” . Spectrophotometric determination of sorbic acid in cheese and cheese wrappers. Food Res. 19:20. Melnick, D. and F.H. Luckmann. 1950b. Sorbic acid as a fungistatic agent for foods. IV. Migration of sorbic acid from" the wrapper into cheese. Food Res. 19:28. 102 Melnick, D.; F.H. Luckmann and C.M. Gooding. 1950. Sorbic acid as a fungistatic agent for foods. VI. Metabolic degradation of sorbic acid in cheese by molds and the mechanism of mold inhibition. Food Res. 19:00. Mirvish, 5.5. 1975. Formation of N-nitroso compounds: chemistry, kinetics, and in vivo occurrence. Toxicol. Appl. Pharmacol. 31:325. Mirvish, 5.5. 1977. N-Nitroso compounds: their chemical and in vitro formation and possible importance as environmental carcinogens. J. Toxicol. Environ. Health 2:1267. Mukherjee, S. 1951. Studies of the degradation of fats by microorganisms. I. Preliminary investigations on enzyme systems involved in the spoilage of fats. Arch. Biochem. Biophys. 33:360. Mukherjee, S. 1952a. Studies on the degradation of fats by microorganisms. II. The mechanism of formation of ketones from fatty acids. Arch. Biochem. Biophys. 35:23. Mukherjee, S. 1952b. Studies on the degradation of fats by microorganisms. Ill. The metabolism of butyric acid by Aspeqillus giger. Arch. Biochem. Biophys. 35:30. Namiki, M. and T. Kada. 1975. Formation of ethylnitrolic acid by the reaction of sorbic acid with sodium nitrite. Agr. Biol. Chem. 39:1335. Newberne, P.M. 1979. Nitrite promotes lymphoma incidence in rats. Science. 100:1079. Pellaroni, N.J. and M.J.R. DePritz. 1960. Influence of sorbic acid on acetate oxidation by Saccharomjces cerevisiae var. ellipsoideus. Nature 185:688. Phaff, H.J.; M.W. Miller and E.M. Mrak. 1968. The Life of Yeasts. 2nd Ed. MIT Press, Cambridge, Mass. Philip, E.T.; A.N. Hall and T.K. Walker. 1963. Some effects of sodium n-butyrate on the behavior of As r illus niger growing in a glucose medium. Arch. Biochem. lop ys. 102:238. Phillips, G.F. and J.O. Mundt. 1950. Sorbic acid as inhibitor of scum yeast in cucumber fermentation. Food Technol. 0:291. Pierson, M.D.; M.C. Robach; K.R. Van Tassell and L.A. Smoot. 1979. Sodium nitrite and potassium sorbate inhibition of Clostridium botulinum as influenced by storage temperature variation. Presented at 39th Annual Meeting of the Institute of Food Technologists, St. Louis, MO, June 10-13, 1979, Paper 373. Pitt, J.l. 1970. Resistance of some food spoilage yeasts to preservatives. Food Technol. Aust. 26:238. 103 Price, J.R. and K.E. Stevenson. 1979. Effects of sorbate and nitrite in bacon on color, flavor, and Clostridium botulinum toxigenesis. Presented at 39th Annual Meeting of thejnstitute (Y Food Tech- nologists, St. Louis, MO, June 10-13, Paper 372. Przybylski, K.S. and L.B. Bullerman. 1980. Influence of sorbic acid on viability and ATP content of conidia of Ajoergillus parasiticus. J. Food Sci. 05:375. Raevori, M. 1976. Effect of sorbic acid and potassium sorbate on growth of Bacillus cereus and Bacillus subtilis in rice filling of Karelian pastry. European J. Appl. MErobiol. 2:205. Rahn, C. and J.E. Conn. 1900. Effect of increases in acidity on antiseptic-efficiency. Ind. Eng. Chem. 36:185. .. sorbinsaure. 6. Die wirkung von sorbinsaure auf den Kohlenhydratstoffwechsel von Escherichia coli. Abl. Bakt. Abt. II Bd. 121:091. - Rehm, H.J. 1967. Zur kenntnis der antimikobiellen wirkung der fl Rehm, H.J.; C. Nimmermann-Vaupel and P. Wallnofer. 1960. Aurkenntnis der antimikrobiellen wirkung der sorbinsaure. V. Abbau and vertmung von sorbinsaure durch verchiedene mikroorganismen. Abl. Bakt. Abt. II Bd. 118:072. Robach, M.C. 1980. Interaction of salt, potassium sorbate and temperature on the outgrowth of Clostridium sarocenes PA 3679 spores in a pre-reduced medium. 3. Food SCI. 05:702. Robach, M.C.; v.0. Difate; K. Adam and L.D. Kier. 1980a. Evaluation of the mutagenicity of sorbic acid-sodium nitrite reaction products produced in bacon-curing brines. Food Cosmet. Toxicol. 18:237. Robach, M.C.; F.J. Ivey and C.S. Hickey. 1978. System for evaluating Clostridial inhibition in cured meat products. Appl. Environ. Microbiol. 26:210. Robach, M.C.; J.L. Owens; M.W. Paquette; J.N. Sofos and F.F. Busta. 1980b. Effects of various concentrations of sodium nitrite and potassium sorbate on nitrosamine formation in commercially prepared bacon. J. Food Sci. 05:1280. Robach, M.C.; E.C. To; S. Meyday and G.F. Cook. 1980c. Effect of sorbates on microbiological growth in cooked turkey products. J. Food Sci. 05:638. Samson, F.E.; A.M. Katz and D.L. Harris. 1955. Effects of acetate and other short-chained fatty acids on yeast metabolism. Arch. Biochem. Biophys. 50:006. Saxby, M.J.; M.A. Stephens and R.G. Reid. 1979. The degradation of sorbic acid under model conditions. Technical Circular No. 703, Leatherhead Food R.A. 100 Shank, R.C. 1975. Toxicology of N—nitroso compounds. Toxicol. Appl. Pharmacol. 31:361. Smith, D.P. and N.J. Rollin. 1950. Sorbic acid as a fungistatic agent for foods. Vll. Effectiveness of sorbic acid in protecting cheese. Food Res. 19:59. Sofos, J.N. and F.F. Busta. 1980. Alternatives to the use of nitrite as an antibotulinal agent. Food Technol. 30:200. Sofos, J.N.; F.F. Busta and C.E. Allen. 1979a. Botulism control by nitrite and sorbate in cured meats: A review. J. Food Prot. 02:739. Sofos, J.N.; F.F. Busta; K. Bhothipaks and C.E. Allen. 1979b. Sodium nitrite and sorbic acid effects on Clostridium botulinum toxin formation in chicken frankfurter-type emulsions. J. Food Sci. 00:668. Sorensen, S.P.L. 1909a. Enzymstudien ll. Mitteilung Uber die messung und die bedentung der wasserstoffionen konzentration bei enzymatischen prozessen. Biochem. 2.21:131. Sorensen, S.P.L. 1909b. Erganzung zu der abhandlung: Enzymstudien. ll. Uber die messung und die bedentung der wasserstoffion konzentration bei enzymatischen prozessen. Biochem. Z. 22:352. Splittstoesser, D.P.; D.T. Queale and L.R. Mattick. 1978. Growth of Saccharomyces bisporus var bisporus, a yeast resistant to sorbic acid. Am. J. Enol. Vitic. 29:272. Stanier, R.V.; E.A. Adelberg and J. Ingraham. 1976. The Microbial World, 0th Ed. Prentice—Hall, Inc., Englewood Cliffs, NJ. Stavkov, B. and A. Petrova. 1960. The effect of sorbic acid on fat metabolism and certain enzyme functions.Khigiena- 7:52 (Bulg.). [CA(62), 11051, 1960]. Tahara, S.; S. Kurogochi; M. Kudo and J. Mizutani. 1977. Fungal metabolism of sorbic acid. Agr. Biol. Chem. 01:1635. Tahara, S.; M.) Kudo and J. Mizutani. 1979. Enzymatic reduction of sorbic acid to trans-0-hexenol. Agr. Biol. Chem. 03:2627. Tanaka, K.; K.C. Chung; H. Hayatsu and T. Kada. 1978. Inhibition of nitrosamine formation in vitro by sorbic acid. Food Cosmet. Toxicol. 16:209. Tanaka, N.; N.J. Worley; E.W. Sheldon and J.M. Geopfert. 1977. Effect of sorbate and sodium acid pyrophosphate on the toxin production by Clostridium botulinum in pork macerate. Food Res. Inst., University ofT‘lisconsin, Ann. Report., p. 366. 105 Thaler, H. and G. Geist. 1939a. Zur chemie der ketonranzigkeit. l. Mitteilung: Uber den abban gesattigter fettsauren durch Penicillium glfaucum. Biochem. 2. 302:121. Thaler, H. and G. Geist. 1939b. Zur chemie der ketonranzigkeit. ll. Mitteilung: Uber die bidung von methylketonen aus oxyfettsauren durch PenicilliUm glaucum. Biochem. 2. 302:369. Thauer, R.K.; K. Jungermann; H. Henninger; J. Wenning and K. Decker. 1968. The energy metabolism of Clostridium kluyveri. European J. Biochem. 0:173. Troller, J.A. 1965. Catalase inhibition as a possible mechanism of the fungistatic action of sorbic acid. Can. J. Microbiol. 11:611. Troller, J.A. and R.A. Olsen. 1967. Derivatives of sorbic acid as food preservatives. J. Food Sci. 32:228. Tompkin, R.B.; L.N. Christiansen; A.B. Shaparis and H. Bolin. 1970. Effects of potassium sorbate on Salmonallae, Staphylococcus aureus, Clostridium grfrin ens and Clostridium botu1i_num in cooked, uncured sausage. pfl. Microbicfl. 28:262. USDA. 1978. Nitrites. Fed. Regist. 03:36607. Vaughn, R.H. and L.O. Emard. 1951. Selectivity of sorbic acid media for catalase negative bacteria and obligate, sporulating anaerobes. Bact. Proc. 38. Walker, T.K.; A.N. Hall; M.C. Langtons and K.C. Selim. 1963. Biological reduction of fatty acids. Arch. Biochem. Biophys. 103:156. Weaver, E.A.; J.F. Robinson and C.H. Hills. 1957. Preservation of apple cider with sodium sorbate. Food Technol. 11:667. Westos, G. 1960. The metabolism of sorbic acid in the mouse. Acta Chem. Scand. 18:1373. Whitaker, J.R. 1959. Inhibition of sulfhydryl enzymes with sorbic acid. Food Res. 20:37. Wick, E.L.; T. Yamanishi; A. Kobayashi; S. Valenzuela and P. lssenberg. 1969. Volatile constituents of banana. J. Agr. Food Chem. 17:751. Witter, R.F.; E.H. Newcomb and E. Stotz. 1950. The oxidation of hexanoic acid and derivatives by liver tissue in vitro. J. Biol. Chem. 185:537. Wogan, G.N. and S.R. Tannenbaum. 1975. Environmental N-nitroso compounds: Implication for public health. Toxicol. Appl. Pharmacol. 31:375. 106 Woodford, R. and E. Adams. 1970. Sorbic acid. Am. Perf. Cosmet. 85:26. York, G.K. 1960. Studies on the inhibition of microbes by sorbic acid. Ph.D. Thesis, University of California, Davis, California. York, G.K. and R.H. Vaughn. 1950. Use of sorbic acid enrichment media for species of Clostridium. J. Bacteriol. 68:739. York, G.K. and R.H. Vaughn. 19553. Resistance of Clostridium parabotulinum to sorbic acid. Food Res. 20:60. York, G.K. and R.H. Vaughn. 1953). Site of microbial inhibition by sorbic acid. Bacteriol. Proc. 20. York, G.K. and R.H. Vaughn. 1960. Mechanisms in the inhibition of microorganisms by sorbic acid. J. Bacteriol. 88:011. ‘ o ’\T ‘ ‘ ‘ ’ 94)) («2&9 ugh“ ex {£14 c.) “—1“. ‘3. é“); 4‘wa O W": it}; 53“?" (“‘37 P \‘l. ~\ — czar-t3“ €33?) _ MIL/MAJ)" I 0"fo a 9,» J.L./69y} a A‘s-I M £599“ ‘09:? by); F9919 ‘93?» if» z ‘aQQ/gwfi Egaarwuaaaygaflasér .wga~yaq:uc.wul £2,935? a» '3 ea”; firyffwfiflxgéw WI» a» M ’ '01-?) mm av.) aeguuwewéwl?‘ e921 area-wwwvwévrjwewz aware)». V»? .0369}; wflyfiéfio‘i We?) 92592:)" iv/‘x‘rL-‘T‘ ‘4‘ w 09” 9 499309;. wiewmvazg Me?» a}; a; .296. (Wuagéfixiwgiié 02;). week).- ca—r =4: .(m (I) Mai-‘2?) (cl—v 9%) wwjé'z Va 4%de g «’0? in)? r9 v: ay¢éxé v); e‘l.‘ dag. “ .5114 degwaxawzawwvyfiux a?) PM 499E120}: ér’ ybhdbrwwzw’abfiwwpb Fae!“ 'W‘Jvéva'vihwwjlfiawéwpwi»woe-9‘4» (La/(”"9"4"5"”’*fia~£3‘&bj<9‘M/WI?€5%PU ’2‘} Dita» 28'3- -&L:J;.~ " an.» Meagado may! 2.422992%» ,3 , (My. , ' I, seesaw my»... .Mx )CZCZ/"glfijl, so él-fi‘rya-‘Jevi oéwébyLrJvL/wzu—aéywi we $92.: w: are: EJVJM [97/9/l’mfifi, '. ‘ . I 0’91"“?on 0v 4;?"me . M 5 5:5. as». .m by,» 1562'; v9.30 éb/c'J’Q-‘L .9944 - o/,...,-...’ 2’. tin»: eeengw/ngk wet/9.130% emu?" ”- W; (cribs: 1;:ch all? éyyn LI . 023/;5 ways: r 4.; a)» ‘ 9299.2?» cat/Mg 51/4423 ’ Mfr. E4 r31}: VI )3 5:5» Ca Qarfudlé/Cd's‘fiw I diva/1,26 ”Is—an" . JV .yiméwxvxé/wwél Héyerr/z