' EFFECT OF PROTEIN .souRcr-ss on THE f enema 0F CLosmmIUM PERFRINGENS . ~ '—'" w _——-Uv-- .._____v_ _ - Thesis for the Degree of M. S. ' MICHEGAN STATE UNIVERSITY PATRICIA JEAN KOKOCZKA 1975 ”I? LIBRAF‘: Y m Michigan State Universiiry IMESls aluom..ey M HOAF'& ‘SUNS' 800K BW'KRY INC. "“9AR BINDERS n {III-I‘ch.‘ ABSTRACT EFFECT OF PROTEIN SOURCES ON THE GROWTH OF CLOSTRIDIUM PERFRINGENS BY Patricia Jean Kokoczka The effects of various vegetable protein sources on the growth of Clostridium perfringens were studied. Two strains of Q. perfringens, one with heat-sensitive spores (ATCC 3624) and one with heat-resistant spores (NCTC 8238) were used in this investigation. Growth media containing 2.0% trypticase (or other protein source on a percent protein basis), 0.25% sodium chloride, 0.15% dipotassium phosphate, 0.006% sodium thioglycollate, 0.004% L-cystine, 0.002% sodium sulfite, and 0.005% agar were prepared. The growth media were inoculated with g. perfringens organisms at concentrations of approximately 102 cells per ml. Flasks containing the inoculated growth media were placed in a constant-temperature water bath maintained at 45 i l C and samples were taken at appro- priate intervals to determine the extent of the lag and log growth phases. Nitrogen gas was introduced into the headspace of the growth vessels during sampling to main- tain anaerobic conditions in the growth media. Patricia Jean Kokoczka Numbers of E. perfringens were determined by plating in a freshly prepared tryptose-sulfite-cycloserine (TSC) agar and incubating at 37 C in a nitrogen atmosphere for 24 hours. Generation times were calculated by using linear regression and measuring the slopes of the growth curves during exponential growth. The results of this investigation indicated that the addition of vegetable protein sources as soy flour, texturized soy flour (Bontrae), and cottonseed flour (Proflo and Pharmamedia) had neither a stimulatory nor an inhibitory effect on the growth of the organism as compared to the addition of trypticase. A difference between strains was shown, however, as strain ATCC 3624 had sig- nificantly longer generation times in these media than did strain NCTC 8238. These results suggest a difference between strains in ability to utilize the substrate. When strain ATCC 3624 was grown in media containing meat and vegetable protein sources, results showed that the addition of cottonseed protein in the form of Proflo and Pharmamedia did have an inhibitory effect on the growth of the organism as compared to the growth rate in media containing meat (beef, chicken, or turkey) or meat and soy flour. The addition of soy flour did not affect the growth of the organism as compared to the growth in media contain- ing meat as the sole protein source. Patricia Jean Kokoczka Clostridium perfringens strain ATCC 3624 was grown in media containing meat and meat with cottonseed flour protein or carbohydrate fractions. The addition of Proflo or Pharmamedia carbohydrate fractions to media containing meat exerted an inhibitory effect on the growth of the organism. One or more factors inhibitory to the growth of Q. perfringens may be present in the carbohydrate fraction of the cottonseed products used in this investi- gation. Clostridium perfringens strain ATCC 3624 was also grown in media containing turkey and Pharmamedia carbo- hydrate fraction which had been treated with glucose oxidase. Results of these studies suggested that glucose was not responsible for the inhibitory effect of the carbohydrate fraction of Pharmamedia. EFFECT OF PROTEIN SOURCES ON THE GROWTH OF CLOSTRIDIUM PERFRINGENS BY Patricia Jean Kokoczka A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1975 ACKNOWLEDGMENTS The author would like to express appreciation to Dr. K. E. Stevenson for his guidance and interest through- out her graduate studies and preparation of this thesis. Appreciation is also expressed to the other mem- bers of the graduate guidance committee: Dr. L. G. Harmon and Dr. E. S. Benecke. The excellent technical assistance of Ms. Marguerite Dynnik is also gratefully acknowledged. Appreciation is given to the Department of Food Science and Human Nutrition for providing facilities and funds to make this study possible. Finally, the author is especially grateful to her husband, John, for his moral support and understanding throughout her graduate studies. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . REVIEW OF LITERATURE . . . . . . . . . . . Clostridium perfringens . . . . . . . Growth of the Organism . . . . . . Temperature . . . . . . . . . Nutritional Requirements . . . Hydrogen Ion Concentration . . Oxidation—Reduction Potential Growth in Meat . . . . . . . . Growth in Vegetable Proteins . Incidence of the Organism . . . . Foodborne Illness . . . . . . . . Media and Methods for Isolation . Selective Media . . . . . . . Diluents . . . . . . . . . . . Method of Plating . . . . . . Vegetable Proteins . . . . . . . . . . Forms of Soybean Proteins . . . . Flours and Grits . . . . . . . Concentrates . . . . . . . . . Isolates . . . . . . . . . . . Chemical and Physical Properties of Soybean Proteins . . . . . . . . Cottonseed Protein . . . . . . . . Page vii b ooooommb 12 12 12 15 15 15 16 16 l7 17 17 19 Forms of Cottonseed Protein . . . . . . . Flour . . . . . . . . . . . . . . . . Concentrates . . . . . . . . . . . . . Isolates . . . . . . . . . . . . . . . Physical and Chemical Properties of Cottonseed Proteins . . . . . . . . . . METHODS AND MATERIALS . . . . . . . . . . . . . . Source and Preparation of Cultures . . . . . . Preparation of Growth Media . . . . . . . . . Series I . . . . . . . . . . . . . . . . . . . Series II . . . . . . . . . . . . . . . . . . Series III . . . . . . . . . . . . . . . . . . Extraction of the Cottonseed Flour Fractions . Apparatus . . . . . . . . . . . . . . . . . . Enumeration of C. perfringens . . . . . . . . Analysis of Data . . . . . . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . . . . . Media with Vegetable Protein Sources . . . . . Media with Meat and Vegetable Protein Sources Media with Meat and Cottonseed Flour Protein and Carbohydrate Fraction . . . . . . . . . SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . LIST OF REFERENCES . . . . . . . . . . . . . . . . APPENDIX I . . . . . . . . . . . . . . . . . . . . APPENDIX II . . . . . . . . . . . . . . . . . . . iv Page 19 19 20 21 21 24 24 25 27 27 28 28 31 31 33 34 34 38 50 59 62 69 70 10. 11. LIST OF TABLES Page The amino acid and growth factor require- ments of Clostridium perfringens . . . . . . 7 Essential amino acid content of soybean proteins . . . . . . . . . . . . . . . . . . 18 Amino acid composition of cottonseed flour and isolates . . . . . . . . . . . . . 22 Basal medium for Clostridium perfringens . . . 26 Percent protein in protein constituents and amounts used in basal media for series I O O O O O O O O O O O O O O O O O O 27 Amounts of protein constituents used in basal media containing both meat and vegetable protein sources for Series II . . 29 Amounts of protein and carbohydrate con- stituents used in basal media for Series III . . . . . . . . . . . . . . . . . 30 Generation times of two strains of C. perfringens in media supplemented with four protein additives and trypticase . . . 35 Analysis of variance of generation times of two strains of C. perfringens in media supplemented with four protein additives and trypticase . . . . . . . . . . 38 Generation times of C. perfringens ATCC 3624 in media containing meat or meat and vegetable protein . . . . . . . . . . . 39 Analysis of variance of generation times of C. perfringens ATCC 3624 in media containing turkey or turkey and vege- table protein . . . . . . . . . . . . . . . 44 Table 12. 13. 14. 15. 16. 17. Page Analysis of variance of generation times of C. perfringens ATCC 3624 in media containing beef'or beef and vegetable protein . . . . . . . . . . . . . . . . . . 44 Analysis of variance of generation times of C. perfringens ATCC 3624 in media containing chicken or chicken and vegetable protein . . . . . . . . . . . . . 45 Multiple range test for data of Tables 11, 12, and 13: Generation times of C. perfringens ATCC 3624 in media containing meat or meat and vegetable protein . . . . . 46 Generation times of C. perfringens ATCC 3624 in media containing meat or meat and cottonseed flour protein or carbo- hydrate fractions . . . . . . . . . . . . . 51 Analysis of variance of generation times of C. perfringens ATCC 3624 in media containing meat or meat and cottonseed flour protein or carbohydrate fractions . . 55 Multiple range test for data of Table 15: Generation times of C. perfringens ATCC 3624 in media contaiHing meat or meat and cottonseed flour protein or carbo- hydrate fractions . . . . . . . . . . . . . 56 vi LIST OF FIGURES Figure Page 1. Typical growth curves of C. perfringens ATCC 3624 in media containing trypticase, Bontrae, and Pharmamedia . . . . . . . . . . 36 2. Typical growth curves of C. perfringens NCTC 8238 in media containing trypticase, soy flour, or Proflo . . . . . . . . . . . . 37 3. Typical growth curves of C. perfringens ATCC 3624 in media containing chicken or chicken and vegetable protein . . . . . . 40 4. Typical growth curves of C. perfringens ATCC 3624 in media containing turkey or turkey and vegetable protein . . . . . . 41 5. Typical growth curves of C. perfringens ATCC 3624 in media containing beef or beef and vegetable protein . . . . . . . . . 42 6. Typical growth curves of C. perfringens ATCC 3624 in media contEining turkey or turkey and cottonseed flour protein or carbohydrate fractions . . . . . . . . . . . 52 7. Typical growth curves of C. perfringens ATCC 3624 in media containing chicken or chicken and cottonseed flour protein or carbohydrate fractions . . . . . . . . . 53 8. Typical growth curves of C. perfringens ATCC 3624 in media containing beef or beef and cottonseed flour protein or carbohydrate fractions . . . . . . . . . . . 54 vii INTRODUCTION Foodborne disease caused by Clostridium perfringens has become a major concern in the food industry. Cooked meat products that have been improperly handled are the major vehicles of transmission. In the United States in 1973, Clostridium perfringens was responsible for about 7% of the outbreaks and about 19% of the cases of confirmed foodborne disease of known etiology. Also, 25% of the total outbreaks of unknown etiology (or 45 outbreaks) were epidemiologically similar to Clostridium perfringens foodborne illness (USDHEW, 1974). Food incriminated in outbreaks invariably has been held at room temperature or improperly refrigerated in large masses for several hours, often overnight or longer. Outbreaks frequently follow banquets or meals prepared at hospitals and schools where large amounts of meat or poultry are involved. Thus, Clostridium perfringens food- borne illness is a disease closely associated with the food service industry (Bryan, 1969). Public acceptance of vegetable protein as a base in fabricated synthetic meats as well as of vegetable protein supplements in meat systems is increasing (Horan, 1974). Institutional markets, including schools and hospitals, are the major markets for synthetic meats. The trend in the United States toward more away-from- home eating also increases the potential use of these pro- ducts (USDA, 1971). This increase in growth of the food service industry also increases the possibilities of food contamination and abuse. Since natural meats are major vehicles of Clostridium perfringens foodborne disease and since vegetable protein supplementation of meat systems is increasing, an evaluation of the behavior of Clostridium perfringens in these new products is appro- priate. The objective of the study was to determine the effects of various animal and vegetable protein sources on the growth of Clostridium perfringens. REVIEW OF LITERATURE Clostridium perfringens Clostridium perfringens (g. welchii) is a non- motile, encapsulated, short and thick bacillus with blunt ends. It occurs singly, in pairs, and occasionally in short chains. Filamentous cells are sometimes produced. Young cultures are gram-positive, but old cultures may appear gram-negative. Subterminal ovoid spores are pro- duced. Vegetative cells sporulate readily in the human intestines, but rarely in cooked meat (Bryan, 1969). Strains of C. perfringens have been divided into five toxicological types (A to E) on the basis of four major toxins (alpha, beta, episilon, iota) which may be produced (Sterne and Warrack, 1964). There is no evidence to associate these toxins with foodborne illness. Type A strains cause the typical mild cases of gastroenteritis associated with foodborne illness and also cause gas gangrene. Type C (formerly known as type F) strains have caused infrequent outbreaks of necrotic enteritis in Germany and in New Guinea (Bryan, 1969). The major toxin produced by C. perfringens type A is alpha toxin, an enzyme (lecithinase C) which attacks lecithin and similar chemical substances to liberate phos- phorylcholine. Opaque zones of precipitate surrounding colonies of C. perfringens on egg yolk agar are evidence of lecithinase activity. Strains of C. perfringens type A may or may not be hemolytic; they do, however, form collagenase, hyaluronidase, and deoxyribonuclease. Type A strains of C. perfringens were originally divided into two subdivisions: gas gangrene strains and food poisoning strains (Hobbs 33 31., 1953). Organisms with heat- sensitive spores (D—values of less than 1 minute when heated in water at 100 C and 3 to 5 minutes at 90 C) that produced large amounts of lecithinase and theta toxin were considered as gas gangrene strains. Food poisoning strains had heat resistant spores (D-values of 4 to 17 minutes at 100 C), produced little lecithinase, and no theta toxin. Today, however, it is believed that it is not necessary for a particular Type A strain to possess a specific biochemical characteristic in order to produce foodborne illness. Thus, there is no such thing as a "food poisoning strain of C. perfringens" as such (Hall and Dowell, 1967; Hobbs, 1969). Growth of the Organism Temperature. Growth is most rapid near 45 C for almost all strains of C. perfringens. Boyd gt 31. (1948) reported an optimum of 45 C for the highly alpha-toxigenic strain BP6K, and Sasarman and Horodniceau (1961) found the same optimum temperature an important aid in devising selective procedures for isolating g. perfringens because it is above the temperature at which most other bacteria readily grow. Growth of Q. perfringens below 20 C is slow but may even extend to 5 C for a few strains. How- ever, growth at any appreciable rate stops at 15 C for most strains. The majority of strains will grow at 50 C even though most of the cells introduced as inocula will perish within the first few hours. The survivors, however, will start multiplying at very nearly their maximum rate and will continue to do so until nutrients are exhausted. This curious phenomenon was first noted by Collee 33 31. (1961) who gave it the name "Phoenix phenomenon." Although the temperature optimum for growth is clearly in the 43 to 46 C range, this is not necessarily the optimum temperature for production of enzymes or toxins by E. perfringens. For example, the temperature optimum for the production of phospholipase C (alpha toxin) was found to vary, among different strains, from 30 to 46 C (Nakamura 3£_31., 1969), while that for epsilon toxin pro- duction was 30 to 33 C. Spore production by E. perfringens has a temperature optimum of 37 C (Kim 33 31., 1967). Several observations have been made in regard to the effect of frozen and refrigerated storage on 9. perfringens. Traci and Duncan (1974) found that at 4 C 96% of an initial population of exponential-phase cells was killed upon cold shock and 95% of the remaining popu- lation was killed within 90 minutes of continued exposure at 4 C. In general, stationary-phase cells were more resistant to cold shock than exponential-phase cells. Canada 33 31. (1964) studied the effects of lowered temperatures on four strains of g. perfringens. After exposure to freezing or refrigeration temperatures (-l7.7 and 7.1 C, respectively), only small numbers of the vege- tative cells were recovered. After similar treatment, 16 to 58% of the spores were recovered. Nutritional Requirements. According to Smith (1971), strains of E. perfringens are rather demanding in their nutritional needs, requiring 13 to 15 amino acids and 5 to 7 additional growth factors (Table 1). These needs can be fulfilled by ordinary bacteriological media. Hydrogen Ion Concentration. Almost all strains of E. perfringens grow most rapidly somewhat below neutrality, generally in the range pH 6.3 to 6.5. No growth was observed below pH 4.9 in unbuffered medium with glucose as the fermentable carbohydrate. On the alkaline side, pH 8.3 marked the limit for the strains of type A and type D that have been studied. Just as the tempera- ture optimum for toxin production differs from that for growth, so the hydrogen ion optimum for toxin production Table 1. The amino acid and growth factor requirements of Clostridium perfringens.1 Arginine Serine2 Leucine Tyrosine Isoleucine Alanine2 Methionine Aspartic acid Phenylalanine Biotin Threonine Pantothenate Valine Pyridoxal Tryptophane Riboflavin2 Histidine Adenine Cystine Uracil2 Glutamic acid Nicotinamide I?ig;$;ed from Boyd 3£_31. (1948) and Fuchs and Bonde 2Variable. differs from that for growth. The pH optimum for pro— duction of phospholipase is slightly below 7.0 and that for production of beta toxin is 7.5. The pH optimum for production of epsilon toxin is 7.0 to 7.2 (Smith, 1971). Oxidation—Reduction Potential. Finding a suitable oxidation-reduction potential for g. perfringens is not generally a problem. However, on minimal media, in which the nutrient supply is just adequate to provide for barely visible colonies, the oxidation-reduction potential can be important. The toxic effect of oxygen on the organisms is nullified if an active reducing agent is added to the medium. The larger the inoculum and the more active its metabolic state, the higher can be the oxidation-reduction potential. The endogenous metabolism of the cells in the inoculum increases the electron density in the neighbor- hood of the inoculum and provides a nidus of low potential and low pH from which active growth can progress (Smith, 1971). Growth in Meat. The effect of various holding temperatures on the growth of Q. perfringens in laboratory- prepared beef cubes in natural gravy was studied by Hall and Angelotti (1965). At 46 C less than 4 hours were required to produce a 100-fold increase over the inoculum. When raw chicken was cooked to obtain a temperature of 85 to 90 C in the breast muscle, a g. perfringens spore inoculum was reduced from 1.0 x 104 spores per gram to 1.5 per gram. Upon incubation of the cooked chicken at 45 C, a lag period was observed for about 4 hours and the growth then became logarithmic. After 14.7 hours, 1.0 x 107 cells were present. When cooked chicken was inoculated with 1.0 x 103 vegetative cells, they multiplied to a total of 1.0 x 107 cells in 6.3 hours at 45 C (Pivnic 3E 31., 1968). The generation time of g. perfringens can be as short as 8.5 minutes (median of 12 minutes for 22 strains) in broth cultures incubated at 46 C and it is about 20 minutes at 37 C (Bryan, 1969). In various poultry and meat stock soups, the generation times ranged from 24-32 minutes at 45 C (Smith, 1962). Vegetative cells of g. pgrfringens survived for over 6 hours when roast turkey slices in broth were held at conditions of steam table storage of 68 C (Strong and Rip, 1967). When boiled lamb, inoculated with 2 x 105 vegetative cells, was stored in gravy on a hot plate at 40 to 50 C for 3 hours, the count increased to 4.6 x 107 cells of g. perfringens per slice. When meat was stored at the same temperature and time without gravy, about half as many organisms were present. Meat stored at room temperature for 3 hours showed no increase of organisms (Hobbs, 1957). A study conducted by Schroder and Busta (1970) showed generation times of 12.5 minutes and 10.8 minutes 10 in beef loaves at 45 C. Busta and Schroder (1970) reported generation times of 14.3 minutes for Q. perfringens in liquid media containing beef and 19.5 minutes in media containing turkey. Growth in Vegetable Proteins. Effects of various soybean proteins used in meat loaf formulations on the growth of two strains of g. perfringens were studied by Schroder and Busta (1970). Their findings indicate that beef products were excellent growth media for the organism and that the addition of soybean proteins had little effect on the growth rates at 45 C. The addition of soy protein concentrate to beef in meat loaf formulations gave average generation times of 11.5 and 9.9 minutes. The addition of soy protein isolate gave generation times of 11.5 and 9.3 minutes. Textured soy flour, when added to the beef loaf formulation, resulted in generation times for g. perfringens of 11.7 and 9.6 minutes. The addition of soy grits gave average generation times of 11.0 minutes and 9.0 minutes. In another study by Busta and Schroder (1970), liquid media containing meat and meat plus vegetable pro- tein supplements were evaluated for stimulatory or inhibitory effects on some strains of g. perfringens. When compared to trypticase, some soy proteins had stimulative effects on the growth of the organism, whereas sodium caseinate and some soy proteins were inhibitory. In 11 liquid media in which meat or soy meat was the source of protein, there was a marked stimulation by beef, chicken, and soy beef. Soy chicken supported growth at a rate less than observed with trypticase. Incidence of the Organism Clostridium perfringens is a normal inhabitant of the intestinal tract of man and animals. The organisms exist in sewage and soil in both spore and vegetative states. Taylor and Gordon (1940) reported that 190 of 196 samples of soil examined contained 9. perfringens, mostly type A. Eighty-five percent of dust samples from a com- mercial kitchen environment contained C. perfringens iso- lates (McKillop, 1959). Strong 3E 31. (1963) isolated E. perfringens from 16% of 122 samples of raw meat, fish, and poultry. Hall and Angelotti (1965) examined 262 samples of retail cuts of meat for the presence of g. perfringens. Of this total, 167 samples were raw unprocessed beef, veal, lamb, pork, and chicken and 101 were processed raw samples of meat and meat dishes. Clostridium perfringens was isolated from 58% of the raw samples and from 20% of the cooked samples. The difference in values reported by these studies could be due to the use of enrichment methods by Hall and Angelotti (1965). 12 Foodborne Illness When several million viable cells of g. perfringens are ingested, diarrhea and abdominal pain accompanied by large volumes of gas in the intestine occur after an incubation period of from 4 to 22 hours, with an average of about 12 hours. Nausea, vomiting, and fever are rare (Hobbs 33 31., 1953). The vegetative cells multiply and sporulate in the intestine and produce an enterotoxin 13 3133. The enterotoxin is released with lysis of the sporangium upon release of the mature spore and causes increased capillary permeability, vasodilation, and excess fluid movement into the intestinal lumen resulting in diarrhea (Hauschild, 1971). Purified C, perfringens enterotoxin has a UV absorption spectrum characteric for proteins, and is essentially free of nucleic acids, lipids, and reducing sugars. Its apparent molecular weight is 36,000 : 4,000 and its isoelectric point is pH 4.3 (Hauschild, 1971). The toxin is sensitive to heat (Duncan and Strong, 1969) and its specific toxicity is about 2,000 mouse MLD/mg N (Hauschild and Hilsheimer, 1971). Media and Methods for Isolation Selective Media. Several selective media have been developed for isolating g. perfringens from material containing other bacteria. Almost all of the media 13 contain sulfite, since it has been demonstrated that this substance is not only inhibitory to many other bacteria, but does not inhibit the Clostridia (Mossel, 1959; Mossel 33 31., 1956). The most common solid media also contain iron and allow the sulfite-reducing Clostridia to produce black colonies. Of these media, the following are cur- rently being used: SPS (sulfite-polymyxin-sulfadiazine) (Angelotti 33 31., 1962), TSN (tryptose-sulfite-neomycin) (Marshall 33 31., 1965), SFP (Shahidi-Ferguson perfringens) (Shahidi and Ferguson, 1971), and TSC (tryptose-sulfite- cycloserine) agars (Harmon 33 31., 1971b). SPS agar selectively inhibits growth or inter- feres with the formation of black colonies by the sulfite- reducing Enterobacteriaceae and Achromobacteriaceae. It also inhibits growth of most other facultative anaerobes and members of the genera Pseudomonas, Bacillus, and Lactobacillus (Angelotti 33 31., 1962; Harmon 33 31., 1971a). However, low recoveries of E. perfringens in commercial SPS agar have been reported (Shahidi and Ferguson, 1971). In one preparation, the cause was traced to a particular lot of yeast extract. It appears that the selective ingredients of this agar are at a level where a slight adverse change in the medium may result in a drastic inhibition of g. perfringens (Hauschild and Hi1- sheimer, 1974). l4 TSN agar has not been used as extensively as SPS agar but reports indicate that TSN is inhibitory to a number of g. perfringens strains (Handford and Cavett, 1973; Harmon 33 31., 1971a). While SFP agar allows quantitative recovery of E. perfringens (Handford and Cavett, 1973; Harmon 33 31., 1971a; Harmon 33 31., 1971b), it does not prevent growth of a large number of facultative anaerobes, some of which are sulfite—reducing (Harmon 33 31., 1971b; Shahidi and Ferguson, 1971). The elaborate preparation involved in the use of SFP agar is another disadvantage since it requires addition of fresh egg yolk, surface plating, and pouring of an agar overlay. Harmon 33 31. (1971a) evaluated SFP, TSN, and SP8 agars to determine which medium was most selective for g. perfringens and most inhibitory to other microorganisms. It was found that while SFP gave significantly higher recoveries of spores and vegetative cells than the other two, it was the least selective. TSN was found to be most selective. SFP agar was later modified (Harmon 33 31., 1971b) by replacing polymyxin B and kanamycin with 0.04% D-cycloserine. This antibiotic had been shown to selectively inhibit growth of many of the common faculta— tive anaerobes (Fuzi and Csukas, 1969). The use of this new modified medium (TSC) was further improved by 15 elimination of the egg yolk and by using a pour-plate technique (Harmon 33 31., 1971b). Diluents. Ordinary buffered salt solutions often contain trace amounts of heavy metals that are lethal or bacteriostatic to g. perfringens when the concentration of bacteria is relatively low. Viable counts were three or four times higher when 0.1% peptone solution was used as the diluting fluid instead of phosphate buffer (Hauschild 33 31., 1967). Method of Plating. The difference between the pour plate count and the roll tube count is significant only when a minimal medium is used and is insignificant with complete media. Viable counts are about the same with both methods. Similarly, there is little difference in viable counts when the plates are surface-inoculated or when the inoculum is added to the agar before pouring (Smith, 1971). Vegetable Proteins At the present time, soybeans are the major source of edible oil in the United States and soybean meal pro- vides an important source of protein for animal feeds. Although food use of soybean protein in the form of defatted flour began in the 19305, the market for these products has developed slowly. With the exception of certain textured vegetable protein products (such as l6 meat-flavored dice, chunk, and crumble products), the current uses of these soy proteins are mainly as ingredi— ents to give desired changes in structure, texture, and composition at a reasonable price. Soy protein products are, in many cases, used to supplement, extend, or replace more costly ingredients derived from animal protein sources, such as meat, eggs, and milk (Wolf, 1970a; Wolf 1970b). Forms of Soybean Proteins Flours and Grits. Forms of soybean proteins pres- ently used as raw materials in the food industry are classified into three groups based on protein content. The least refined forms are flours and grits which have vary- ing fat contents, particle sizes, textures, and degrees of heat treatment. Flours are prepared by grinding soy- bean flakes to lOO-mesh or finer, whereas grits are coarser than loo-mesh. Minimum protein contents of these materials range from 40 to 50%, depending on the fat con- tent. Defatted flours usually contain close to 50% pro- tein (Wolf, 1970b). Proteins, carbohydrates, and ash are the major components of the defatted flour with about one- half of the flour carbohydrates being oligosaccharides-- sucrose, stachyose, and raffinose--while the other half is made up of polysaccharides, which are insoluble in water or alcohol (Aspinall 33 31., 1967). l7 Concentrates. Soybean protein concentrates are more refined than flours and grits and contain 70% or more protein on a dry basis. The protein concentrates are generally prepared from defatted flakes or flour by removing the oligosaccharides, part of the ash, and some of the minor components by alcohol extraction (Mustakas 33 31., 1962), acid leach, or moist heat followed by a water wash (Wolf, 1970b). The physical properties of the concentrates vary according to the method of their prepara- tion (Meyer, 1966). Isolates. The isolates are the most refined forms of soybean proteins. They contain 90% or more protein and are prepared by removing the water-insoluble polysaccharides, as well as the oligosaccharides and other low molecular weight components (Wolf, 1970b). Chemical and Physical Prqp3rties of Soybean Proteins Amino acid composition of the soybean proteins determines the nutritional value. Table 2 lists the essential amino acid compositions for the three major forms of soybean proteins. The proteins are high in lysine and so are useful as supplements for cereals, which tend to be low in this amino acid. Methionine, however, is the first limiting amino acid in soy proteins and this should be considered when the proteins are added for nutritional 18 Table 2. Essential amino acid content of soybean proteins.l Amino Acid Meal Concentrate Isolate 9/169 N 9/169 N 9/169 N Lysine 6.9 6.6 5.7 Methionine 1.6 1.3 1.3 Cystine 1.6 1.6 1.0 Tryptophan 1.3 1.4 1.0 Threonine 4.3 4.3 3.8 Isoleucine 5.1 4.9 5.0 Leucine 7.7 8.0 7.9 Phenylalanine 5.0 5.3 5.9 Valine 5.4 5.0 5.2 1 Table from Wolf (1970b). purposes rather than merely for functional purposes (Wolf, 1970). The major soybean proteins are globulins, which are insoluble at their isoelectric points. The proteins are, however, soluble in water or dilute salt solutions at pH values above or below the isoelectric point (Wolf, 1970b). Soy proteins bind fat, are hydrophylic, and have film-forming and adhesive qualities. Although the functional properties of soy products are usually attributed to the protein fraction, when the less refined products are used, other constituents may also contribute to the effect. For example, in flours, grits, and concentrates 19 the polysaccharides as well as the proteins absorb water (Wolf, 1970b). Cottonseed Protein It was recognized early in the history of the pro- cessing of cottonseed that the cottonseed could be a valu- able source of both protein and oil (Richardson, 1917). However, the presence of the chemically reactive anti- nutritional factor--gossypol (Berardi and Goldblatt, 1969)-- prevented the rapid development of edible protein products from the cottonseed. Gossypol is a highly reactive yellow polyphenolic binaphthaldehyde. In the metabolically active or "free" form, it has adverse physiological effects when ingested by monogastric animals (Berardi and Goldblatt, 1969). It also may impart an undesirable color to the cottonseed protein product. Gossypol is contained in relatively large intercellular structures called pigment glands. These glands and, hence, the gossypol, have been reduced through breeding (McNichael, 1959; Miravalle, 1969). This development plus the development of the Liquid Cyclone method (Gastrock 33 31., 1969) for processing glanded cottonseed have created new interest in the potential of edible cottonseed products. Forms of Cottonseed Proteins Flour. Cottonseed flour is the finely ground material produced from dehulled and defatted seed. 20 Cottonseed, when essentially completely dehulled and defatted to a lipid content of 1% or less, provides a product which contains 60% protein (N x 6.25) and 3% or less fiber. Solvent extraction operations using heat and mechanical pressure are commonly used commercially. How- ever, these means must be used judicially or the protein constituent will be affected. Denaturation effects due to the defatting operation can be adequately estimated by the nitrogen solubility of the flour. The standard pro- cedures developed for soy protein products are not appli- cable to cottonseed protein products. The cottonseed proteins, unlike soy proteins, are not readily dispersible in water. Alkaline conditions are needed to solubilize the major proteins of the cottonseed (Martinez 33 31., 1970). Concentrates. Further concentration of the pro- teins of the cottonseed can be accomplished by either wet or dry processing of the defatted flour. Ninety percent ethanol gave optimum extraction of residual lipid and sugars with minimum removal of nitrogen (Berardi 33 31., 1968). Ethanol extraction increased the protein content from 66 to 72%. Aqueous extraction at essentially neutral pH (pH 6.3 to 6.8) is another wet procedure for the preparation of cottonseed protein concentrates. Water or dilute divalent cationic salt solutions can also be used. 21 Isolates. Further concentration of the cotton- seed protein requires extraction of the protein from the defatted seed. A two-step extraction procedure has been developed by Berardi 33 31. (1968) that provides the selective separation of low and high molecular weight pro- teins from the nondenatured flours. The flour is extracted with water (Step I) and then with 0.01N NaOH (Step II). Physical and Chemical Properties of Cottonseed Proteins Amino acid composition of cottonseed flour and isolates is listed in Table 3. Glandless cottonseed flours, when properly processed, were equal in nutritive value to the casein control when tested in three separate trials using the standard AOAC procedure for protein efficiency ratio (PER) evaluation (Hopkins, 1967). Lysine is the first limiting amino acid in cottonseed. There- fore, the air-classified concentrates, which were slightly lower in lysine than the parent flours, were significantly lower in PER, 2.3 rather than 2.5 (casein set at 2.5). However, the residues from the air classification process were equal in PER to the parent flours (Hopkins, 1967). On the basis of extractability characteristics, the proteins of the cottonseed can be divided into two groups, water-soluble and water-insoluble. The water— soluble proteins are low in molecular weight and many in number. They have minimum solubility at pH 4. The 22 Table 3. Amino acid composition of cottonseed flour and isolates. Amino Acid Glang13zg glour I33123eNI 1:3133eNII Lysine 4.3 6.0 3.0 Histidine 2.8 2.6 3.0 Arginine 12.1 10.4 11.3 Aspartic 8.8 6.7 8.4 Threonine 3.0 2.9 2.7 Serine 4.0 3.4 4.5 Glutamic 19.8 21.8 18.9 Proline 3.4 3.1 3.1 Glycine 4.0 3.2 3.7 Alanine 3.6 3.2 3.5 Valine 4.5 3.3 4.1 1/2 Cystine --- 2.6 0.3 Methionine 1.2 1.7 1.0 Isoleucine 3.3 2.6 3.1 Leucine 5.6 5.1 5.8 Tyrosine 3.0 3.3 2.6 Phenylalanine 5.3 3.7 6.3 1Martinez 33 31., 1970. 23 water-insoluble proteins are high in molecular weight and are few in number. They have a minimum solubility at pH 7 (Martinez 33 31,, 1970). METHODS AND MATERIALS Source and Preparation of Cultures Two strains of 3. perfringens, one with heat- sensitive spores (ATCC 3624) and one with heat-resistant spores (NCTC 8238) were obtained from the culture col- lection of the Department of Food Science and Human Nutrition, Michigan State University. Spores were produced using the procedure of Duncan and Strong (1967). Active cultures were produced by inoculating 0.1 ml of the original cultures into 10 ml of fluid thioglycollate medium (Difco) which was then incubated at 37 C for 16 hours. Two subsequent transfers were made into 10 ml of fresh fluid thioglycollate medium, using 10% inocula and 4-hour incubation periods at 37 C. The entire contents of the last tube were inoculated into 100 m1 of sporulation medium which had been freshly steamed and cooled to 37 C in a 250-ml screw-cap Erlenmeyer flask. The sporulation medium contained 0.4% yeast extract (Difco), 1.5% proteose peptone (Difco), 0.4% soluble starch (BBL), 0.1% sodium HPO '7H 0 (Merck). A 2 4 2 Petroff-Hausser Bacteria Counter (C. A. Hausser & Son, thioglycollate (Difco), and 1.0% Na Philadelphia, Pa.) was used in conjunction with a Series 10 24 25 Microstar phase microscope (American Optical Corporation, Buffalo, N.Y. 14215) to estimate the percent sporulation in cultures after incubation for 40 hours in the sporu- lation medium. Aliquots of the cultures containing spores and vegetative cells were frozen and stored at -12 C and they served as stock cultures. As needed, cultures were removed from the freezer and thawed at room temperature. Active cultures were produced by inoculating 0.1 m1 of the stock culture into 10 m1 of fluid thioglycollate medium without dextrose (Difco) which was then incubated at 37 C for 24 hours. After four such transfers, less than 10 spores per ml were present. These cultures were used as inocula and were diluted in 0.1% peptone to achieve a con- centration of approximately 102 cells per ml in the growth media. Preparation of Growth Media Various test protein sources were substitutedfor trypticase (Difco) in thioglycollate medium without dextrose (Busta and Schroder, 1970). Table 4 lists the ingredients of the basal growth medium and the protein sources used as variables. All of the meat used in these experiments was obtained from Food Stores, Michigan State University. The ground chicken and ground turkey were mixtures of 50% (w/w) light meat and 50% dark meat. The ground beef was lean ground round steak. United States Department of Agriculture (Anonymous, 1972) statistics were used to calculate 26 Table 4. Basal medium for Clostridium perfringens. Constituent Amount 9/1 Trypticase or substitution by . . . 20.0 proteins on a % protein ba51s Sodium chloride 2.5 Dipotassium phosphate 1.5 Sodium thioglycollate 0.6 L-cystine 0.4 Sodium sulfite 0.2 Agar 0.5 1Proteins: Soy flour (Fern brand Low Fat Soya Powder), texturized soy flour (Bontrae), cottonseed flour (Proflo), cottonseed flour (Pharmamedia), ground beef, ground turkey, and ground chicken. amounts of meat added to the growth media to obtain the same protein concentration (w/w) as the trypticase control (Table 5). The soy flour product used was FernR brand Low Fat Soya Powder (Richards Food Corp., Melrose Park, Illinois 60160). The texturized soy product was BontraeR (General Mills, Minneapolis, Minn. 55453) and the two cottonseed products used were ProfloR and PharmamediaR (Traders Protein Div., Traders Oil Mill Co., Fort Worth, Texas). Manufacturers' specifications were used to cal- culate the amounts of vegetable protein substituted in the growth media to obtain the same protein concentration as the control (Table 5). 27 Table 5. Percent protein in protein constituents and amounts used in basal media for Series I. Constituent Percent Protein Agognt Soy flour 50.0 40.0 Bontrae 20.0 100.0 Proflo 60.0 33.3 Pharmamedia 56.8 35.2 Ground beef 19.0 106.0 Ground turkey 23.0 86.0 Ground chicken 22.0 90.0 Series I. A series of experiments was conducted using the basal growth medium (Table 4) containing a 2% level of protein. The protein sources tested in this series were: a. trypticase (control) d. Proflo b. soy flour e. Pharmamedia c. Bontrae Percent protein in the protein sources and the amounts used in the growth media are listed in Table 5. Series II. A series of experiments was also con- ducted using the basal growth medium containing a 2% pro— tein level from meat alone or meat plus vegetable protein. Combinations of meat (beef, chicken and turkey) plus vege- table protein used in this series were: 28 a. meat + soy flour c. meat + Proflo b. meat + Pharmamedia d. meat Table 6 lists the amounts of the protein sources used in the growth media. Series III. A series of experiments was also con- ducted using the basal growth medium containing a 1% level of meat protein plus protein and carbohydrate fractions prepared from cottonseed products, Proflo and Pharmamedia. The protein and carbohydrate fractions were substituted into the basal growth medium with each of the three types of meat at a level to give the same concen- tration (w/w) as a 1% protein level of the whole cotton- seed flour. Combinations of meat (beef, chicken, and turkey) and vegetable protein sources used in this series were: a. meat + Proflo protein fraction b. meat _ Proflo carbohydrate fraction c. meat + Pharmamedia protein fraction d. meat + Pharmamedia carbohydrate fraction e. meat Table 7 lists amounts of protein and carbohydrate consti- tuents used in the basal growth media. Extraction of the Cottonseed Flour Fractions. Separation of the cottonseed flour into protein and carbohydrate fractions was accomplished by a modification 29 Table 6. Amounts of protein constituents used in basal media containing both meat and vegetable protein sources for Series II. Amount Used Constituent Meat Veg. Protein 9/1 g/l Beef 106.0 Beef and Soy Flour 53.0 20.0 Beef and Pharmamedia 53.0 17.0 Beef and Proflo 53.0 17.5 Chicken 90.0 Chicken and Soy Flour 45.0 20.0 Chicken and Pharmamedia 45.0 17.0 Chicken and Proflo 45.0 17.5 Turkey 86.0 Turkey and Soy Flour 43.0 20.0 Turkey and Pharmamedia 43.0 17.0 Turkey and Proflo 43.0 17.5 30 Table 7. Amounts of protein and carbohydrate constituents used in basal media for Series III. Constituents Amount USGd 9/1 Beef 45.0 Chicken 53.0 Turkey 43.0 Pharmamedia protein fraction1 12.1 Proflo protein fraction2 12.5 Pharmamedia carbohydrate fraction 1.8 Proflo carbohydrate fraction 1.8 lContains 82.8% protein (N x 6.25). 2Contains 79.8% protein (N x 6.25). of the extraction procedure described by Lillevik (1970). A 10-gram sample of the cottonseed product was placed in a 250-ml Erlenmeyer flask along with 150 m1 of 50% ethanol and stirred for 3 hours with a Stir—Plate magnetic stirring apparatus (Thermolyne Corp., Dubuque, Iowa). The mixture was made to 250 ml with 95% ethanol, stirred, and filtered through No. 4 Whatman filter paper. The filtrate was then dried in a Thelco Precision Model vacuum oven (Precision Scientific Products, Chicago, Illinois) at 40 C-and 25 inches of Hg for 24 hours. The protein content (N x 6.25) of the residue was determined by the micro-Kjeldahl method (Joslyn, 1970). The dried residue (referred to as 31 Pharmamedia protein fraction or Proflo protein fraction) was substituted into the basal growth media as previously described. The extract produced by the extraction procedure was evaporated to dryness on a steam bath and the residue (referred to as Pharmamedia carbohydrate fraction or Pro- flo carbohydrate fraction) was substituted into the basal growth media as previously described. The growth media were adjusted to pH 7.0 i 0.1 with a glass electrode Beckman research pH meter and auto- claved at 121 C for 15 minutes. The media were steamed for 20 minutes before inoculation to insure a low oxidation- reduction potential. Apparatus. The growth vessels consisted of 1000- ml screw-cap Erlenmeyer flasks for Series I and II and 250-ml'screw-cap Erlenmeyer flasks for Series III. The flasks containing the growth media were placed in a constant-temperature water bath maintained at 45 i l C. Samples were taken at appropriate intervals to determine the extent of the lag and log growth phases. Nitrogen gas was introduced into the headspace of the growth vessels during sampling to maintain anaerobic conditions in the growth media. Enumeration of C. perfringens. Samples were serially diluted in 0.1% peptone dilution blanks (APHA, 32 1967) and numbers of g. perfringens were determined by plating in duplicate or triplicate in a freshly prepared plating medium, tryptose-sulfite-cycloserine (TSC) agar, consisting of the basal ingredients of sulfite-polymyxin- sulfadiazine (SPS) agar (Angelotti 33 31., 1962), with D-cycloserine (Sigma Chemical Co., St. Louis, Mo.) sub- stituted for polymyxin B sulfate and sulfadiazine (Harmon 33 31., 1971b). The basal SPS medium contained 1.5% tryptone (Difco), 1.5% yeast extract (Difco), 0.05% ferric citrate (Baker Chemical Co., Phillipsburg, N.J.), 0.1% sodium metabisulfite (Mallincrodt Chemical Works, St. Louis, Mo.), and 20% agar (Difco). These ingredients were dissolved in 1 liter of distilled water and auto- claved at 121 C for 15 minutes. The D-cycloserine was dissolved in distilled water (0.1 gram per 25 m1), dis- pensed in screw-cap dilution bottles, and steamed for 20 minutes. The solution was stored at 4 C. A final con- centration of 400 ug of the antibiotic per ml of SPS basal medium was obtained by adding 10 m1 of D-cycloserine solution to a liter of the plating medium after auto- claving and just prior to plating. The agar plates were inverted and incubated in a nitrogen atmosphere in an anaerobic incubator (National Appliance Co., Portland, Oregon) at 37 C for 24 hours. The plates were placed in the incubator and a vacuum of 24-27 inches of Hg was produced with a Cenco Hyvac vacuum 33 pump (Cenco Instruments Corp., Chicago, I11.). The chamber of the incubator was filled with nitrogen gas, re-evacuated, and again filled with nitrogen. Ana1ysis of Data. Generation times were cal- culated by using linear regression and measuring the slope of the growth curves during the exponential growth phase. The growth curves were obtained by plotting the specific plate counts obtained vs time. Generation times obtained in Series I experiments were analyzed for significant differences between growth rates in the presence of the protein sources by analysis of variance. Results from Series II were also analyzed for significant differences among the trials using meat proteins alone and trials using meat and vegetable protein combinations. Results from Series III experiments were analyzed for significant differences among the trials using the meat proteins at the 1% level and the meat proteins plus cottonseed fractions. Duncan's multiple range test was also applied (Duncan, 1955). RESULTS AND DISCUSS ION Media with Vegetable Protein Sources. Table 8 indicates the generation times of 3. perfringens strains ATCC 3624 and NCTC 8238 in the basal growth medium supplemented with trypticase and four vegetable protein additives (Series I experiments). Figures 1 and 2 show typical growth curves for one of the organisms in these media. An analysis of variance of the results is pre- sented in Table 9. There were no apparent significant differences, at the 5% level, in generation times of these two strains of g. perfringens as the result of the addition of the various proteins to the basal medium. However, a significant difference was revealed between the results obtained with strain ATCC 3624 and strain NCTC 8238. The average generation times obtained with strain NCTC 8238 were significantly shorter, at the 0.1% level, than those obtained with strain ATCC 3624 in media containing the vegetable protein variables. In contrast, the average generation times obtained in media containing trypticase were significantly longer for strain NCTC 8238 than for strain ATCC 3624 (0.1% level). This significant interaction between strain and media indicates 34 35 .Av manna wmmv Edficoa Amman mnu pmcamucoo memEmm Hamm v.ma m.ma m.va m.ma m.vH mmmum>< m.mH m.na m.va m.¢a m.ma m H.0m m.~a m.vH v.5H m.ma m mmmm oeoz m.mH o.mH m.vH H.5H m.¢a a m.ma m.ma k.om m.~a m.mH mmmum>< m.ma m.ma m.Hm o.ma m.ma m m.mH o.mH w.ma m.>a m.wa m vmmm ousfi 3.0H m.H~ ¢.Hm a.ma m.om H I I I I I I I I I I I I I I I I I I I bamev I I I I I I I I I I I I I mmmoflummua MHUmEmEumcm oamoum mmuuaom usoam mom mumoflammm samuum .mmmflmummuu Ufim mm>aufiopm aflmuoum HDOM nufl3 cmucme Imammsm maome ca mammcflumuwm .0 mo mcflmuum 03» mo mwswu coaumnmcmo .m manme m Clostridium perfringens. calls/g Fig. I09 l07 l05 IO3 1. 36 I Trypticase / o Bontrae A Pharmamedia I Time (hr) Typical growth curves of 3. perfringens ATCC 3624 in media containing trypticase, Bontrae, and Pharmamedia. 3L Clostridium porfrimn , cells/q Ioa Io6 Io2 Fig. 2. 37 A I / /. L Ir )- ./ I Trypticase I A/ A Proflo 0 Soy flour L l l l l I g I 5 7 Time (hr) Typical growth curves of 3. perfringens NCTC 8238 in media containing trypticase, soy flour, or Proflo. 38 Table 9. Analysis of variance of generation times of two strains of g. perfringens in media supplemented with four protein additives and trypticase. Source of Variation Dggzzgszf Sgggrgg Sgizge F value Total 29 174.4 Strains l 48.38 48.38 20.2* Protein Sources 4 4.32 1.08 0.45 Interaction 4 73.9 18.48 7.73** Error 20 47.8 2.39 *Significant at the 0.1% level. **Significant at the 5% level. differences between strains in their ability to utilize the substrates. There appeared to be no difference, however, within a strain in the ability to utilize the various vegetable protein sources as compared to the trypticase control. Apparently these vegetable protein sources are all comparatively similar in terms of protein availability to the organisms under the conditions used in this investi- gation. Media with Meat and Vegetable Protein Sources. Table 10 summarizes data on the growth of 3. perfringens ATCC 3624 in growth media containing meat and vegetable proteins. Typical growth curves for the organism in these media are shown in Figures 3, 4, and 5. Analyses of variance of these results are presented in Table 10. Generation times of 3. perfringens ATCC 3624 in media containing meat or meat and vegetable protein. a Generation times, min Sample 1 2 3 Average Beef control 11.3 10.9 11.1 Beef + Soy Flour 12.0 12.2 11.5 11.9 Beef + Proflo 13.9 13.6 12.9 13.5b Beef + Pharmamedia 17.1 16.8 17.1 17.0b Chicken control 12.1 12.6 12.4 Chicken + Soy Flour 12.1 12.7 12.8 12.5 Chicken + Proflo 17.0 16.2 16.5 16.6C Chicken + Pharmamedia 14.9 14.0 14.9 14.6C Turkey control 12.0 12.0 12.0 Turkey + Soy Flour 11.3 12.9 12.6 12.3 Turkey + Proflo 14.5 14.2 13.6 14.1d Turkey + Pharmamedia 14.4 15.0 15.6 15.0d aAll samples contained the basal growth medium (see Table 4). bSignificantly different from beef control at 0.1% level. CSignificantly different from chicken control at 0.1% level. dSignificantly different from turkey control at 0.1% level. e, cells/g Clostridium perfrin Fig. 40 Io'° I- 8 _ ,‘ IO fl/z IOBF- . I Chicken alone A Chicken 4» Proflo IO4 0 Chicken + Pharmamedia e Chicken + Soy Flour I02 I 1 A l l l 1 l l 3 5 Time (hr) 3. Typical growth curves of C. perfringens ATCC 3624 in media containing chicken or chicken and vegetable protein. Clostridium perfringens . cel le/g Fig. Io'° IOa IO4 4. 41 e Turkey + Soy Flour I Turkey alone 0 Turkey + Proflo A Turkey + Pharmamedia l 3 5 Time (hr) Typical growth curves of C. perfringens ATCC 3624 in media containing turkey or turkey and vegetable protein. 42 [olO _ l0a . I—--—I A / / . / IOB . Clostridium perfringens, cells /9 8 e Beef + Soy Flour 0 Beef + Proflo a Beef + Pharmamedia l0: .z” ” l 1 I l l l 3 Time (hr) Fig. 5. Typical growth curves of g. perfringens ATCC 3624 in media containing beef or beef and vegetable protein. 43 Table 11 for the trials with turkey, in Table 12 for the trials with beef, and in Table 13 for the trials with chicken. These analyses showed apparent significant differences, at the 0.1% level, among the average genera- tion times obtained in the trials using the various vege- table protein sources in combination with meat. Duncan's multiple range test was applied to the data to compare dif- ferences among the average generation times obtained in this set of trials (Series II). Table 14 shows the multiple range test for the data of Tables 11 through 13. Comparisons showed that the average generation time obtained with media containing turkey plus soy flour was not significantly different from the average generation time obtained with the turkey control (meat alone). Average generation times obtained with the cottonseed pro- ducts, Proflo and Pharmamedia, plus turkey were signifi- cantly longer than those obtained with the turkey control. Generation times obtained with turkey and Pharmamedia were not significantly different from those obtained with turkey and Proflo. In the case of beef, comparisons showed that generation times obtained with media containing beef plus soy flour were not significantly different from the average generation time obtained with the beef control (meat alone). Average generation times obtained with beef plus the cottonseed products were significantly longer 44 Table 11. Analysis of variance of generation times of 3. perfringens ATCC 3624 in media containing turkey or turkey and vegetable protein. Source of Degrees of Sum of Mean . . F value Variation Freedom Squares Square Total 10 19.34 Proteins 3 16.74 5.58 14.99* Error 7 2.61 0.372 *Significant at the 0.1% level. Table 12. Analysis of variance of generation times of 3. perfringens ATCC 3624 in media containing beef or beef and vegetable protein. Source of Degrees of Sum of Mean F value Variation Freedom Squares Square Total 10 56.95 Proteins 3 55.97 18.66 134.42* Error 7 0.97 0.14 *Significant at the 0.1% level. 45 Table 13. Analysis of variance of generation times of g. perfringens ATCC 3624 in media containing chicken or chicken and vegetable protein. Source of Degrees of Sum of Mean . . F value Variation Freedom Squares Square Total 10 33.60 Proteins 3 32.44 10.81 65.50* Error 7 1.16 0.17 *Significant at the 0.1% level. than those obtained with the beef control. Generation times obtained with beef plus Proflo were significantly shorter than those obtained with beef plus Pharmamedia. The multiple range test also showed that, in the case of chicken, the average generation time obtained with media containing chicken plus soy flour was not signifi- cantly different from the average generation time obtained with the chicken control (meat alone). In these trials the generation times obtained with chicken plus Pharmamedia were significantly shorter than those with chicken plus Proflo. Generation times observed with meat plus soy flour in this investigation are consistent with findings pub- lished by Busta and Schroder (1970). Natural meat systems appear to be excellent growth media for 3. perfringens and the addition of soybean proteins in the form of soy flour 46 .ummfi can mflmemEumcm u Eumcm .ummE cam oamoum n oumv .ummE can “soda mOm n womm d .mwmcmu ucmoflwflcmflm ummuuonm H mm .Aamma .umuumm “coma a .uwpumm «mmma .cmocso Eoum mabmev Eopmmnm mo mmmnmmc h MOM mmsHm> “Madame H OH . . . . . . . m 0 ma m ma m HH a HH mma H mma a mad H m d Eumnm one wow mmmm om~.m mva.m mem.e o m o.mH H.va m.ma o.mH Hem.a mmm.a mmm.a m _ m shone one mom smkuse oe~.m mea.m aem.e o . . . . . . . a o ma m ea m NH v NH emm a mom H mam H mm a meme sheen msom sexuaeo oem.m mea.m mem.e 0H fiuc mug NHC mQOmHHmmEOU Hm>mH ma momcmu ucmoHMHcmHm ummuuocm .cflmuoum wanmummm> can ume no name mcHCHmucoo manna ca «mom OOB< mammcflnmuum .0 mo mmEHu cofiumumcmo ”ma cam .NH .HH meanme mo mumc MOM umwu mmcmu mamfluasz .vH manna 47 does not appear to affect the potential for rapid growth of the organism. The addition of cottonseed proteins in the form of cottonseed flour did have an inhibitory effect on the growth rate of 3. perfringens strain ATCC 3624. In each set of trials the generation time of the organism was sig— nificantly longer, at the 1% level, in the media containing meat plus cottonseed flour than in the media containing protein from meat alone and meat plus soy flour. Manu- facturer's processing of the cottonseed flour may have caused a change in solubility or some modification of the cottonseed protein which may affect the growth of 3. perfringens. Excessive heat during processing of the cottonseed flour can affect the protein solubility and amino acid availability of the final product. With cotton— seed, only the barest minimum of dry heat in excess of that needed for removing the solvent can be tolerated without causing browning and coagulation of water—soluble proteins (Martinez 33 31,, 1970). Another possibility is that the presence of one or more inhibitors affected the growth of g. perfringens. Manufacturer's specifications state that the typical gossypol content of Proflo and Pharmamedia is 0.029%. Even this small amount, however, may have been responsible for all or part of the inhibition observed. Gossypol is highly reactive chemically and has undesirable physiological effects. Gossypol reacts with 48 other seed constituents during processing of cottonseed and is responsible, in part, for reduction of the nutri- tive quality of the protein of cottonseed meal and for other adverse physiological effects in nonruminants (Berardi and Goldblatt, 1969). According to manufacturer's literature, Pharmamedia contains approximately 5.5% glucose while Proflo contains approximately 6.1%. A study of the effects of food com- ponents on 3. perfringens conducted by Schroder and Busta (1973) indicated that the addition of D-glucose to sodium caseinate or to isolated soy protein greatly extended the lag phase of growth. Other sugars tested did not have this inhibitory effect. These findings offer another possible explanation of the inhibitory effect of cottonseed flour as observed in these experiments. Comparison of the generation times obtained with media containing meat plus Proflo and meat plus Pharma- media revealed that in the case of turkey, the generation times obtained with Proflo additive were not significantly different from those obtained with Pharmamedia additive. In the case of beef, the trials using Proflo additive resulted in generation times that were significantly shorter than those obtained with Pharmamedia additive. With chicken, however, the trials using Pharmamedia resulted in significantly shorter, at the 1% level, generation times than did the trials using Proflo. These 49 results suggest that there is an interaction between the rate of growth of the organism and the composition of the protein sources. For example, different amino acid com- positions of the different meat and vegetable protein sources may result in varying levels of the nutrients necessary for the growth of 3. p3rfringens. Thus, beef plus Pharmamedia may have a different level of some nutrient or factor necessary for the growth of 3, p337 fringens than the combination beef plus Proflo. The results obtained in the trials using turkey suggest that turkey contains some factor that is lacking (or present in a different concentration) in beef and chicken. Alter- natively, turkey may lack, or contain a lesser concentra- tion of, some factor that is contained in both Proflo and Pharmamedia, since the generation times obtained with turkey plus these cottonseed products were not signifi- cantly different from each other. In the trials with chicken, since the generation times with Proflo additive were longer than with Pharmamedia, chicken may lack, or contain a lesser concentration of, some factor that is in Pharmamedia and absent, or at a lower concentration, in Proflo. A study conducted by Clifford 33 31. (1974) revealed that at a 1.0% level of substitution, the inclusion of xylose or ribose in the plating medium inhibited the growth of 3. perfringens spores while the 50 inclusion of glucose enhanced growth (the greatest increase being with 0.01 and 0.1% concentrations). Xylose is present in Proflo and Pharmamedia at levels of approximately 1.4% and 1.3%, respectively, according to manufacturer's literature. The combination of mannose, ribose, and rhamnose is present at a concentration of 2.3% in both Proflo and Pharmamedia. While Clifford 33 31. (1974) did not identify the point of inhibition (i.e., spore acti- vation, germination, or outgrowth) the presence of these carbohydrate components in the cottonseed products sug- gests another possible explanation for the inhibitory effects of cottonseed flour observed in these experiments. Media with Meat and Cottonseed Flour Protein and Carbohydrate Fraction. Table 15 lists the generation times of E. perfringens ATCC 3624 in growth media containing meat or meat and cottonseed flour protein and carbohydrate fractions. Figures 6, 7, and 8 show typical growth curves of the organism in these media. Analysis of variance of the data in Table 15 is summarized in Table 16. Signifi- cant differences were revealed and a multiple range test was applied to the data to compare the differences (Table 17). The analyses revealed no apparent significant dif- ference, at the 5% level, between generation times obtained with media containing meat (beef, chicken, or turkey at the 1% substitution level) and meat with Pharmamedia protein Table 15. Generation times of C. perfringens ATCC 3624 in media containing meaE or meat and cottonseed flour protein or carbohydrate fractions. a Generation Sample times, min Average 1 2 Beef control (1%) 14.9 14.8 14.9 Beef + Pharmamedia protein 15.0 14.3 14.7 fraction Beef + Pharmamedia carbohydrate fraction 16'9 15°1 16'0 Beef + Proflo protein 15.3 14.9 15.1 fraction Beef + Proflo carbohydrate fraction 15.5 16.0 15.8 Chicken control (1%) 15.3 13.2 14.3 Chicken + Pharmamedia 16.2 14.0 15.1 protein fraction Chicken + Pharmamedia carbohydrate fraction 16'6 15'9 16'3 Chicken + Proflo protein fraction 15.2 16.9 16.1 Chicken + Proflo ‘ carbohydrate fraction 15'4 15°? 15’6 Turkey control (1%) 14.6 15.4 15.0 Turkey T Pharmamedla 15.2 14.7 15.0 protein fraction Turkey + Pharmamedia carbohydrate fraction 16'9 16‘2 16'6 Tugkey f Proflo protein 16.7 16.9 16.8 raction Turkey + Pr°f1° 16.5 16.3 16.4 carbohydrate fraction aAll samples contained the basal growth medium (see Table 4). 52 IOB r. // ’ / Q .2 8 I06 . .5 .12 E 4 Turkey + Pharmamedia protein a '0 fraction e Turkey + Pharmamedia carbohydrate fraction A Turkey + Proflo protein fraction I Turkey + Proflo carbohydrate fraction a Turkey l% IOZ I[ I n l 1 I I 3 5 Time (hr) Fig. 6. Typical growth curves of 3. perfringens ATCC 3624 in media containing turkey or turkey and cottonseed flour protein or carbohydrate fractions. 53 I09 F '08 .. /;D ///.__.I e / //'r . / / ~ / / A E / / ‘8 / ’ 3‘ nos. 4 .g //// 't / a / / .5 - / it: // e Chicken + Pharmamedia protein .3 . // fraction 0 .04 . / e Chicken + Pharmamedia F carbohydrate fraction / A Chicken + Proflo protein fraction I I Chicken + Proflo carbohydrate / fraction ‘ // A Chicken l% / 2 IOZ I i 1 n 1 j 1 I . 5 6 Time (hr) Fig. 7. Typical growth curves of 3. perfringens ATCC 3624 in media containing chicken or chicken and cottonseed flour protein or carbohydrate fractions. 54 IO9 _ P I /e / 7 / IO . - / / / 9’ //’/// 0 e . / / 5 // 8 // / 6' / a l05 /‘ / i * / / / I E // / / g L ’ /‘ I Beef + Pharmamedia protein *3 / ‘ / fraction 2 // / e Beet + Pharmamedia o / / carbohydrate traction IO3 _ y I 0 Beef + Proflo protein fraction / / / A Beet + Proflo carbohydrate / ’ fraction 4/ 4 Beef l% IO2 L l 1 l l l J l 3 Time (hr) Fig. 8. Typical growth curves of C. perfringens ATCC 3624 in media containing beef or beef and cottonseed flour protein or carbohydrate fractions. 55 Table 16. Analysis of variance of generation times of C. perfringens ATCC 3624 in media containing meat or meat and cottonseed flour protein or carbo- hydrate fractions. Source of Degrees of Sum of Mean . . F value Variation Freedom Squares Square Total 29 25.82 0.89 Meat 2 2.43 1.213 3.441* Additives 4 11.91 2.98 8.449** Interaction (M x A) 8 2.31 0.289 0.819 Error 26 9.17 0.353 *Significant at the 5% level. **Significant at the 1% level. fraction. No apparent significant difference was seen, at the 5% level, among the generation times that were observed with media containing meat and Proflo carbohydrate fraction, meat and Proflo protein fraction, and meat and Pharmamedia carbohydrate fraction. Although the generation times observed with media containing meat only were not significantly different from generation times observed with media containing meat and Pharmamedia protein fraction, the generation times obtained with these two types of media were significantly different, at the 1% level, from the generation times obtained with media containing meat and Pharmamedia carbohydrate fraction, Proflo carbohydrate fraction, or Proflo protein fraction. These results 56 .cmMOHno Eoum no: pun .uonuo some Eoum acoHoMMHp MHDGMUAMHcmHm mumz moxuau can mean .mocmofimwcmflm mo Hm>ma wm nae m .mocmu uCMUMMHcmflm ammunosm u mm .Aaema .houamm rowed .ucuumm e ammma .cmocso Eoum mannav Eoommum mo mooummo on How mosam> “washes n 0m coauowum mumnpmconumo oamoum + Dome u OIHm cofluomum cflmuoum Camoum + Home u mInm cofluomnm mumuoasonumo MHUmEmEumnm + Dome u OInm cofiuomum cflmuoum aflomfimfiumcm + poms u mIcm .Hm>oa cflmuonm wH pm wou50m cwmuoum oaom mm Ammxusu no .cmeHso .momnv Dome n ocoH¢ "wmmH O O O O O Q m ma v ma m ma mmh o mom 0 omn.o m emmxnse coxowcu «mmom moa.e mmo.v mmm.m do a om.oa o.oa mm.mH mm.va mn.va omo.a oao.a mmm.o mvm.o an m ouch eIue ouum eInm ecoaa om~.e mea.e emo.e mmm.m 0m mug @"C mug NHC Hmcomflummfioo Hm>oa wa mmmcmu unmeamwcmfim umouuonm .mGOHuomum wumuomconumo no cflmuoum HDOHM comm chwuoo com name no umofi mcwcwmucoo mwooe ca «mom UOB¢ mawmcflnmuwm .0 mo mafia» aoflumuocwm “ma manna no «you now umau amcmn mamfluasz .nH dance 57 suggest that while the addition of Pharmamedia protein fraction had no significant effect on the growth of 3. perfringens ATCC 3624 in media containing beef, chicken, or turkey, the addition of Proflo or Pharmamedia carbohydrate fractions or Proflo protein fraction exerted an inhibitory effect on the growth of the organism. The fact that the inhibitory effect was observed in media containing both Proflo and Pharmamedia carbohydrate fractions suggests that one or more inhibitory factors may be present in the carbohydrate fraction of the cottonseed products used in this investigation. A significant difference was also revealed, at the 5% level, among generation times observed with media con- taining beef, chicken, or turkey at the 1% level of sub— stitution. Generation times observed with media containing beef were significantly shorter than generation times observed with turkey. No significant difference was seen between generation times observed with chicken and those observed with beef or between generation times observed with chicken and those observed with turkey. These results suggest that there may be different amounts of nutrients or growth factors in these types of meat that account for the slightly different rates of growth of 3. perfringens in the media. Since the inhibitory effect of D-glucose on the growth of 3. perfringens had been previously reported 58 (Schroder and Busta, 1973), an attempt was made to elimi— nate this component from the carbohydrate fraction of Pharmamedia by treatment with glucose oxidase. The basal growth medium (Table 4) was prepared using turkey and Pharmamedia carbohydrate fraction that had been treated with glucose oxidase. An experiment was conducted (using appropriate controls) to determine the generation time of 3. perfring3ns ATCC 3624 in this medium. Results of this experiment are listed in Appendix Table I. Typical growth curves of the organism are shown in Appendix Figure II. These data, however, are inconclusive due, in part, to the small number of trials conducted. The use of media con- taining carbohydrate fraction from which the glucose had been enzymatically removed did not result in significantly shorter generation times. The inhibitory effect of the carbohydrate fraction does not appear to be due to its glucose content. SUMMARY AND CONCLUS IONS The effects of various vegetable protein sources on the growth of Clostridium perfringens were studied. Results show that the addition of vegetable protein sources as soy flour, Bontrae, Proflo, and Pharmamedia to growth media for 3. perfringens had neither a stimulatory nor an inhibitory effect on the rate of growth of the organism as compared to the addition of trypticase. A difference between strains was shown since heat sensitive strain ATCC 3624 had significantly longer generation times in media containing vegetable protein sources than did heat resistant strain NCTC 8238. Thus, a difference between strains in ability to utilize the substrate was shown, although the vegetable protein sources used were apparently similar in terms of protein availability to the organism. Clostridium perfringens strain ATCC 3624 was grown in media containing meat and vegetable protein sources. Results show that the addition of cottonseed protein in the form of Proflo and Pharmamedia has an inhibitory effect on the growth rate of the organism when compared to the growth rate in media containing beef, chicken, or turkey 59 60 as the sole protein source or media containing meat and soy flour. The growth rate of the organism in media con- taining soy flour and meat was similar to the growth rate in media containing meat as the sole source of protein. Since generation times of the organism in media containing chicken and Pharmamedia were significantly shorter than in media containing chicken and Proflo, an interaction between protein sources in a given type of media was indicated. Generation times in media containing beef and Proflo were significantly shorter than in media containing beef and Pharmamedia. Generation times of the organism in media containing turkey and Proflo were not significantly different from generation times in media containing turkey and Pharmamedia. Some factor or factors may be present in Pharmamedia and Proflo that exert an inhibitory effect on the growth of g. perfringens. The interaction observed between meat and cottonseed protein sources may be due to differing concentrations of one or more factors in the protein sources that are utilized by the organism for growth. Clostridium p3rfringens ATCC 3624 was grown in media containing meat and meat with cottonseed flour pro- tein and carbohydrate fractions. Results show that the addition of Proflo or Pharmamedia carbohydrate fractions to media containing meat exerted an inhibitory effect on the growth of the organism. One or more inhibitory factors 61 may be present in the carbohydrate fraction of the cotton- seed products. A significant difference was also seen among growth rates observed with media containing various meat protein sources. Generation times observed with media containing beef were significantly shorter than with media containing turkey. No significant difference was seen between gene- ration times observed with chicken and those observed with beef or between generation times observed with chicken and those observed with turkey. Meat is generally an excellent growth medium for g. perfringens and different levels of nutrients or growth factors among these types of meat may account for the slightly different rates of growth observed. Clostridium perfringens ATCC 3624 was grown in media containing turkey and Pharmamedia carbohydrate fraction which had been treated with glucose oxidase. The results were inconclusive. The use of media containing carbohydrate fraction from which the glucose had been enzymatically removed did not result in significantly shorter generation times. Glucose does not appear to be responsible for the inhibitory effect of the carbohydrate fraction of Pharmamedia. LIST OF REFERENCES LIST OF REFERENCES American Public Health Association. 1966. Recommended methods for the microbiological examination of foods. 2nd ed. American Public Health Associ- ation, Inc., Washington, D.C. Angelotti, R., H. E. Hall, M. J. Foter, and K. H. Lewis. 1962. Quantitation of Clostridium perfringens in foods. Appl. Microbiol. 10:193-199. Anonymous. 1972. The merits of meat. Turkey World 47: 21. Aspinall, G. 0., R. Begbie, and J. E. McKay. 1967. Polysaccharide components of soybeans. Cereal Sci. Today 12:223-228ff. Berardi, L. C., C. J. Fernandez, and W. H. Martinez. 1968. Southern Utilization Research and Develop- ment Div., ARS, USDA, New Orleans, La., private communication. In W. H. Martinez, L. C. Berardi, and L. A. Goldblatt. 1970. Cottonseed protein products-—composition and functionality. J. Agr. Food Chem. 18:961-968. Berardi, L. C. and L. A. Goldblatt. 1969. Gossypol, p. 211—266 in I. E. Liener, Toxic Constituents of Plant Foodstuffs. AcademiE Press, New York. Boyd, M. J., M. A. Logan, and A. A. Tytell. 1948. The growth requirements of C. perfringens (welchii) BP6K. J. Biochem. l74:I013-1025. Bryan, F. L. 1969. What the sanitarian should know about 3. perfringens foodborne illness. J. Milk and Food Technol. 32:381-389. Busta, F. F. and D. J. Schroder. 1970. Effect of soy proteins on the growth of Clostridium perfringens. Appl. Microbiol. 22:177-1837 62 63 Canada, J. C., D. H. Strong, and L. S. Scott. 1964. Response of Clostridium perfringens spores and vegetative cells to temperature variation. Appl. Microbiol. 12:273-276. Clifford, W. J., A. Anellis, and E. W. Ross. 1974. Evaluation of media, time and temperature of incubation, and method of enumeration of several strains of Clostridium perfringens spores. Appl. Microbiol. 27:784-792. Collee, J. G., A. Knowlden, and B. C. Hobbs. 1961. Studies on the growth, sporulation, and carriage of C. welchii with special reference to food poisoning strains. J. Appl. Bacteriol. 24: 326-339. Duncan, C. L. and D. H. Strong. 1967. Improved medium for sporulation of Clostridium perfringens. Appl. Microbiol. 18:82-89. Duncan, C. L. and D. H. Strong. 1969. Ileal loop fluid accumulation and production of diarrhea in rabbits by cell-free products of Clostridium perfringens. J. Bacteriol. 100:86-94. Duncan, D. B. 1955. Multiple range and multiple F tests. Biometrics 11:1-42. Fuchs, A. R. and G. J. Bonde. 1957. The nutritional requirements of Clostridium perfringens. J. Gen. Microbiol. 16:317-329. Fuzi, M. and Z. Csukas. 1969. New selective medium for the isolation of g, perfringens. Acta Microbiol. Acad. Sci. Hung. 16:273-278. Cited in A. H. W. Hauschild and R. Hilsheimer. 1974. Evaluation and modification of media for enumeration of C. perfringens. Appl. Microbiol. 27:78-82. — Gastrock, E. A., E. L. D'Aquin, P. H. Eaves, and D. E. Cross. 1969. Edible flour from cottonseed by liquid classification using hexane. Cereal Sci. Today 14:8-11. Hall, H. E. and R. Angelotti. 1965. Clostridium perfringens in meat and meat products. Appl. Microbiol. 13:353-357. 64 Hall, H. E. and V. R. Dowell. 1967. Foodborne disease outbreaks caused by C. perfringens. In "Coopera- tive study of C. perfringens isolates" packet. National Communicable Disease Center. Atlanta, Georgia. Handford, P. M. and J. J. Cavett. 1973. A medium for the detection and enumeration of Clostridium perfringens (welchii). J. Sci. Food Agr. 24:487. Harmon, S. M., D. A. Kautter, and J. T. Peeler. 1971a. Comparison of media for the enumeration of Clostridium perfringens. Appl. Microbiol. 21: 922-927. Harmon, S. M., D. A. Kautter, and J. T. Peeler. 1971b. Improved medium for enumeration of Clostridium perfringens. Appl. Microbiol. 22:688-692. Harter, H. L. 1960. Critical values for Duncan's new multiple range test. Biometrics 16:671-685. Harter, H. L. 1961. Corrected error rates for Duncan's new multiple range test. Biometrics 17:321-324. Hauschild, A. H. W. 1971. Clostridium perfringens enterotoxin. J. Milk and Food Technol. 34:596-599. Hauschild, A. H. W., I. E. Erdman, R. Hilsheimer, and F. S. Thatcher. Variations in recovery of Clostridium perfringens on commercial sulfite-polymyxin-’ sulfadiazine (SPS) agar. J. Food Sci. 32:467-473. Hauschild, A. H. W. and R. Hilsheimer. 1971. Purification and characteristics of the enterotoxin of Clostridium perfringens type A. Can. J. Microbiol. 17:1425-1433. Hauschild, A. H. W. and R. Hilsheimer. 1974. Evaluation and modifications of media for enumeration of Clostridium perfringens. Appl. Microbiol. 27: 78-82. Hobbs, B. C. 1957. The growth of heat resistant Clostridium welchii in cooked meat. Bacteriol. Proc. 14:14. Hobbs, B. C. 1969. Clostridium perfringens and Bacillus cereus infections, p. 131-171 in H. Riemann, Foodborne Infections and Intoxications. New York: Academic Press. 65 Hobbs, B. C., M. R. Smith, C. L. Oakley, and G. C. Warrack. 1953. Clostridium welchii food poisoning. J. Hygiene 51:75-101. Hopkins, D. T. 1967. Central Research, Ralston Purina Co., St. Louis, Mo., private communication. In W. H. Martinez, L. C. Berardi, and L. A. Goldblatt. 1970. Cottonseed protein products--composition and functionality. J. Agr. Food Chem. 18:961-968. Horan, F. E. 1974. Meat analogs, p. 366-413, in A. M. Altschul, New Protein Foods. New York: Academic Press. Kim, C. H., R. Cheney, and M. Woodburn. 1967. Sporu- lation of Clostridium perfringens in a modified medium and selected foods. Appl. Microbiol. 15: 871-876. Joslyn, M. A. 1970. Extraction methods and separation processes, p. 141-200, in M. A. Joslyn, Methods in Food Analysis. New York: Academic Press. Lillevik, H. A. 1970. The determination of total organic nitrogen, p. 601-616 in M. A. Joslyn, Methods in Food Analysis. New York: Academic Press. Marshall, R. S., J. F. Steenbergen, and L. S. McClung. 1965. Rapid technique for the enumeration of Clostridium perfringens. Appl. Microbiol. 13: 599-563. Martinez, W. H., L. C. Berardi, and L. A. Goldblatt. 1970. Cottonseed protein products--composition and functionality. J. Agr. Food Chem. 18:961-968. McKillop, E. J. 1957. Bacterial contamination of hospital food with special reference to Clostridium welchii food poisoning. J. Hygiene 57:31-46. McMichael, S. C. 1959. HOpi cotton, a source of cotton- seed free of gossypol pigments. Agron. J. 52:630. Meyer, E. W. 1966. Proc. Int. Conf. on Soybean Protein Foods, Peoria, 111., ARS-7l-35, p. 142-154, U.S. Department of Agriculture, Washington, D.C., in W. J. Wolf, Soybean proteins: their functional, chemical, and physical properties. J. Agr. Food Chem. 18:969-976. 66 Miravelle, R. J. 1969. Conference on Protein-Rich Food Products from Oilseeds, New Orleans, La., ARS-72- 71, p. 76, U.S. Department of Agriculture, Wash- ington, D.C., in W. J. Wolf, Soybean proteins: their functional, chemical, and physical properties. J. Agr. Food Chem. 18:969-976. Mossel, D. A. 1959. Enumeration of sulfite-reducing Clostridium occurring in foods. J. Sci. Food Agr. Mossel, D. A., A. S. DeBruin, H. M. VanDiepen, C. M. Vendrig, and G. Zoutewelle. 1956. The enumeration of anaerobic bacteria, and of Clostridium species in particular, in foods. J. Appl. Bacteriol. 19: 142-154. Mustakas, G. C., L. D. Kirk, and E. L. Griffin. 1962. Flash desolventizing defatted soybean meals washed with aqueous alcohols to yield a high-protein product. J. Amer. Oil Chem. Soc. 39:222—226. Nakamura, M., J. A. Schulze, and W. R. Cross. 1969. Factors affecting lecithinase activity and pro- duction in Clostridium perfringens. J. Hygiene 67:153-162. Pivnic, H., I. E. Erdman, S. Manzatiuk, and E. Pommier. 1968. Growth of food poisoning bacteria on barbe- qued chicken. J. Milk Food Technol. 31:198-201. Richardson, A. E. 1917. Texas University Bulletin No. 1732, p. 2-20, in W. H. Martinez, L. C. Berardi, and L. A. Goldblatt. 1970. Cottonseed protein products--composition and functionality. J. Agr. Food Chem. 18:961-967. Sasarman, A.L and T. Horodnigeau. 1961. Influence de la temperature sur le development et la tougenese_du Clostridium perfringens. Zones optima de tempera- ture des clostridies. Arch. Roum. Pathol. Exper. 20:691-697, in L. 08. Smith. 1972. Factors involved in the isolation of Clostridium perfringens. J. Milk Food Technol. 35:71-76. Schroder, D. J. and F. F. Busta. 1970. Growth of Clostridium perfringens in meat loaf with and without added soybean protein. J. Milk Food Technol. 34:215-217. 67 Schroder, D. J. and F. F. Busta. 1973. Glucose inhibits growth of Clostridium perfringens in food proteins. J. Milk Food Technol. 37:4-10. Shahidi, S. A. and A. R. Ferguson. 1971. New quali- tative, quantitative, and confirmatory media for rapid analysis of food for Clostridium perfringens. Appl. Microbiol. 21:500—506. Smith, L. DS. 1962. Clostridium perfringens food poison- ing, p. 77-83, in L. W. Slanetz, C. O. Chichester, A. R. Gaufin, and Z. J. Ordal. 1963. Microbio- logical Quality of Foods. New York: Academic Press. Smith, L. DS. 1971. Factors involved in the isolation of Clostridium perfringens. J. Milk Food Technol. 35:71-76. Sterne, M., and G. H. Warrack. 1964. The types of Clostridium perfringens. J. Pathol. Bacteriol. 88:270-283. Strong, D. H., J. C. Canada, and B. B. Griffiths. 1963. Incidence of Clostridium perfringens in american foods. Appl. Microbiol. 11:42-44. Strong, D. H. and N. M. Ripp. 1967. Effect of cookery and holding on hams and turkey rolls contaminated with Clostridium perfringens. Appl. Microbiol. 11:1172-1177. Taylor, A. W. and W. S. Gordon. 1940. A survey of the type of Clostridium welchii present in soil and in the intestinal contents of man and animals. J. Pathol. Bacteriol. 82:529-531. Traci, P. A. and C. L. Duncan. 1974. Cold shock lethality and injury in Clostridium perfringens. Appl. Microbiol. 28: - . United States Department of Health, Education, and Welfare. 1974. Foodborne outbreaks annual summary, 1973. Center for Disease Control. Atlanta, Georgia. United States Department of Agriculture. 1971. Syn- thetics and substitutes for agricultural products. Projections for 1980. Marketing Research Report. No. 947. Washington, D.C. 68 Wolf, W. J. 1970a. Purification and properties of the proteins, p. 93-143, in A. K. Smith and S. J. Circle. Soybeans: Chemistry and Technology. New York: AVI Publishing Co., Inc. Wolf, W. J. 1970b. Soybean proteins: their functional, chemical, and physical properties. J. Agr. Food Chem. 18:969-976. APPENDICES APPENDIX I APPENDIX I Generation times of C. perfringens ATCC 3624 in media con- taining turkey, Pharmamedia carbohydrate fraction, and glucose oxidase. 1 Generation Sample times, min Average 1 2 Turkey2 14.5 15.6 15.1 Turkey + Pharmamedia 14.7 16.6 15.7 carbohydrate fraction Turkey + Pharmamedia carbohydrate fraction + 14.8 14.5 14.2 glucose oxidase 1All samples contained the basal medium (see Table 4). 2Glucose oxidase added, but reaction stopped immediately by autoclaving the medium. Analysis of variance of generation times of C. perfringens ATCC 3624 in media containing turkey, Pharmamedia carbohy- drate fraction, and glucose oxidase. Source of Degrees of Sum of Mean F value Variation Freedom Squares Square Total 5 4.97 0.994 Treatment 2 2.45 1.225 1.458 Error 3 2.52 0.84 69 APPENDIX II Clostridium pertringee, cells/g APPENDIX I I IO9 , /I / / I07 . l05 . Turkey + enzyme (no reaction) I Turkey + enzyme + carbohydrate fraction (no reaction) e Turkey + enzyme + carbohydrate fraction (reaction) IO3 ’0 l l I J LO 3.0 5.0 Time (hr) 70 TUTTI”Tim”:gliriliujmflmmmmmn 485