n... d“‘?t& :£.b-‘o‘;l!‘x’lfil' ‘:'\""i') ' c .. ‘ ' H ' V- .. t. ‘ - ; .' '—v """" ‘ - ~ 4 ~ - . . smmss OF’HE DETECTION OF. , ommvmsmwmm m LAKE MICHIGAN onus ;_ - (Liucrcmms HOY!) . Thesis for the Degree of M. s. MICHIGAN STATE‘UNW‘ERSITY JAMES ANDREW KOPRowsm 1968 '.o' - <-. ................... o ....... --------- THESIS LIBRAR Y Mkhigan State I 3n§vm {11y ABSTRACT STUDIES OF THE DETECTION OF DIMETHYLNITROSAMINE IN LAKE MICHIGAN CHUB (Leucichthy_s hoE') By James Andrew Keprowski Due to outbreaks of botulism.poisoning, the Great Lakes fish industry has developed a processing method utilizing nitrite as a bacteriostatic agent to minimize the growth of any Clostridium.botulinum.Type E organ- isms that may be present in the chub and also to minimize the necessary heat process. Researchers in Norway'found a potent carcinogen (dimethyl- nitrosamine) in herring meal processed from.nitrite treated herring. They advanced the theory that the nitrite reacts with the naturally occurring methylamines in the fish to produce the nitroso derivative. The major portion of this study was concerned with developing a method sensitive enough to detect dimethylnitrosamine in chub at concen- trations less than 1 ppm. A thin-layer chromatographic method was adopted which will detect 0.5 ppm.dimethy1nitrosamine in chub. Prelimp inary research with other methods of analysis are also reported. Combinations of three nitrite concentrations in the brine with two processing temperatures were studied as to their effect on dimethyl- nitrosamine formation in chub. Apparently none of these variables had any effect on the formation of this compound. The results of this study indicate that there was less than 0.5 ppm.dimethy1nitrosamine in the nitrite processed chub. STUDIES OF THE DETECTION OF DIMETHYLNITROSAMINE IN TAKE MICHIGAN CHUB (Leucichthys hoyi) By James Andrew Kbprowski A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER 0F SCIENCE Department of Food Science 1968 ACKNOWLEDGEMENTS The author wishes to express his appreciation to Professor L. J. Bratzler for his assistance and encouragement throughout the course of this study and for his assistance in the preparation of this manuscript. Acknowledgement is given to Dr. E. J. Benne, Professor of Biochemistry, Dr. P. Markakis, Professor of Food Science, and Dr. A. N. Pearson, Professor of Food Science, for their critical review of this manuscript. He also wishes to thank all of the graduate students and staff of the Meat Laboratory for their encouragement and cooperation. The author is especially grateful for the interest and practical suggestions offered by Mrs. Dora Spooner, Dr. Ki Soon Rhee, and Dr. C. B. Johnson. The author desires to acknowledge his parents for their stimulation and encouragement to pursue a higher degree of education. Last, but not least, the author wishes to extend sincere appreciation to his wife, Jeannene, for the sacrifices she made to enable the author to complete this project. ii TABLE OF CONTENTS INTRODUCTION . . . . . . . . . . . . . . . LITERATURE REVIEW 0 O O O O O O O O O O 0 Properties of Dimethylnitrosamine . . Biochemical and Physiological Effects invivo............... of Nitroso Compounds Occurrence of Dimethylnitrosamine in Fish Meal . . . . . . MethOdS 0f AnalySiS o o o o o o o o o Occurrence of Nitroso Compounds in Foodstuffs . . . . . . . EXPERIMENTALPROCEDURE.......... Treatments . . . . . . . . . . . . . Dimethylnitrosamine Determination . . Moisture Determination . . . . . . . Ether Extract Determination . . . . . Nitrite Determination . . . . . . . . pH Determination . . . . . . . . . . Salt Concentration in the Water Phase RESULTS AND DISCUSSION . . .-=. . . . . . . A Report on Preliminary Research . . Discussion of Experimental Results . SUMMARY AND CONCLUSIONS . . . . . . . . . BIBIJImR-AHIY O I O O O O O O O O O O O O 0 iii Determination . . . . Page 12 16 18 18 19 21 21 22 22 23 24 24 33 34 LIST OF TABLES Table Page 1 Results of chemical analyses of chub loin muscle receiving the 80°C process 0 O O O C O O O I O O O O O O O O O O 0 O 31 2 Results of chemical analysis of chub loin muscle receiving the 107°C mocess O O O O O O O O O O O O O O O O O O O O 32 iv LIST OF FIGURES Figure Page 1 Possible pathways for the formation of urinary dime thylanmine O O O O O O O O O O O O O O O O O O O O O 0 ll INTRODUCTION In recent years smoked chub (Leucighthys hgyi) production has been an important part of the Great Lakes fish industry. About 10 to 11 million pounds of chub were taken for smoking in 1962-63 (Patashnik 23 .§;,, 1964). However, late in 1963 bacterial contamination of smoked fish resulted in the overnight collapse of that industry. Between September 30 and October 7, 1963, seven persons died from, and seven.more were hospitalized and treated for, botulism poisoning. All of these cases were associated.with the ingestion of improperly pro- cessed and handled smoked fish (ciscoes, chubs, or whitefish) caught and processed in the Great Lakes area (Anonymous, 1964). The Federal Food and Drug Administration (F.D.A.) immediately appointed a committee to study the problem and make recommendations to prevent further out- breaks. Since the incidence of Clostridium.botulinum.Type E is wide- sPread in the Great Lakes region (Foster 33 31., 1965) and because of the properties of the organism.(it grows well at refrigerator tempera- tures, does not produce offensive odors, and is not destroyed by the cold smoking process used for smoking fish) (Pivnick 33 55., 1967), it was evident that g. botulinum Type E represented a serious hazard in smoked fish products. The advisory committee established certain requirements for the processing and handling of smoked fish based on denaturing the existing toxin and/or providing conditions subsequent to manufacture and before sale to the consumer that would prevent the germination and growth of existing spores with consequent toxin production. Among the specific requirements set up by the committee were: the incorporation of a heat treatment that would insure a temperature of no less than 82°C (180°F) for at least 30 minutes in the coldest part of the fish, followed by subsequent refrigeration until the product was consumed; or the other alternative was to keep the fish in a frozen state until consumed. The National Fisheries Institute agreed to conform with the recomp mendations set forth by F.D.A.'s advisory committee. Similarly, regulations to this effect have been adopted by regulatory agencies in Michigan, Minnesota, and Wisconsin. Although the toxin is denatured by this 82°C internal temperature, the fish industry indicated that an acceptable product could not be produced under these conditions (Bratzler and Robinson, 1967). Because of this, the fish industry would like to develop a processing method whereby this heat treatment could be reduced to the previous level of about 60°C (140°F). It has been observed that nitrite has a bacterio- static effect on the Clostridium.botulinum Type E organism.(Tarr, 1942). A processing schedule was developed.which included the use of sodium nitrite in the brine cure followed by a three hour cooking and smoking treatment. The National Fisheries Institute submitted a petition to the F.D.A. which described a food processing method for smoked fish processing, including the use of nitrites. F.D.A. was hesitant to permit the use of this method because of the recent concern regarding an increase in the general level of nitrites in the American diet and also because -3- European researchers have found a strong carcinogen (dimethylnitrosamine) in herring meal made from herring preserved with sodium nitrite (Ender, 1964). The purpose of this study was to determine if the addition of nitrite to the brining medium results in the formation of dimethylnitrosamine in chub taken from Lake Michigan. LITERATURE REVIEW Properties pf Dimethylnitrosamine Dimethylnitrosamine (DMNA, N;nitrosodimethy1amine, nitrous dimethyl- amide) is a yellow, oily liquid with the following empirical formula: (CH3)2 - N - N = 0. DMNA has a molecular weight of 74.08, a specific gravity of 1.006 (20/4), a refractive index of 1.43743 at 18°C, and a boiling point of 153°C at 744 mm.of mercury (Scheflan and Jacobs, 1953). It is very soluble in water, alcohol, and ether (Sax, 1963) but insoluble in dilute aqueous mineral acids (Morrison and Boyd, 1964). DMNA absorbs in the near ultra~violet region of the spectrum.(Smales and Wilson, 1948) having absorption peaks in aqueous solutions at 226 mu and 330 nu. (Ender _e_t_ 31;, 1964). Very little was known of the toxic properties of DMNA before the mid 1950's. About this time several research teams reported cases of human poisonings associated.with exposure to DMNA (Hamilton and Hardy, 1949; Barnes and Magee, 1954; Jacobson.gtwal., 1955). After two cases of liver cirrhosis developed in men working with DMNA as a solvent in an industrial laboratory, the research team.of Barnes and Magee (1954) initiated work to study the toxic properties of DMNA in greater detail. They found that DMNA acts primarily as a liver poison producing severe liver necrosis in rats, mice, rabbits, guinea pigs, and dogs when administered in the range of 20 to 50 mg/kg of body weight. The species examined seemed to be equally susceptible to DMNA, -4- -5- although differences in symptoms were noticed. Other signs associated with the centrilobular necrosis of the liver were the formation of anti- plastic tumors, cirrhosis and hemorrhages into the injured liver, the peritoneal cavity and lumen of the gut. Also, when DMNA was added to the diet of rats at levels of about 50 ppm, a high incidence of malig- nant hepatic tumors were observed (Magee and Barnes, 1956). Many researchers have also observed that DMNA is a potent carcinogen producing cancers in the liver, kidney and lung (Magee and Barnes, 1956, 1962; Schmahl and Preussmann, 1959; Zak st 31., 1960; Terracini and Magee, 1964). Although Barnes and Magee (1954) suggested that DMNA vapor was not toxic, Freund (1937) and Jacobson 33 El. (1955) described cases of human poisoning associated with the inhalation of DMNA vapor. Jacobson.g£ El. (1955) conducted studies on the toxicity of DMNA vapor and concluded that it was also toxic and resulted in lesions in the prementioned organs. In animals that survive DMNA poisoning, the liver regenerates and appears to recover completely (Magee, 1966). Biochemical and Physiological Effects of Nitroso Compounds in vivo .-— —____—__—-=_=_ Along with many nitrosamines certain nitrosamides such as N-nitroso- methylurea (Druckery st 21., 1961) and N;nitrosomethylurethane (Schoental, 1960) are very potent carcinogenic substances. The nitrosamides, being less stable than most nitrosamines induce lesions at the site of appli- cation and also in distant organs (Druckery 33 31., 1961; Schoental, H“ -6- 1961). The nitrosamines are rapidly metabolized after injection into animals (Dutton and Heath, 1956; Magee, 1956; Heath, 1962) mainly by the liver but also to a lesser degree in the kidney and other organs. Following the administration of DMNA to rats, there is a drastic reduction in the ability of the liver to incorporate amino acids into proteins (Magee, 1958; Bouwers and Emmelot, 1960). The major site of inhibition seems to be at the polyribosomes involving the amino acyl transfer step to the ribosomes (Bouwers and Emmelot, 1960; Mizrahi and Emmelot, 1962). Through the use of nitrosamines labeled with radioactive tracers in their alkyl groups, evidence has been obtained that alkylation of cellular components does occur in animals injected with carcinogenic nitroso compounds (Magee and Hultin, 1962; Magee and Farber, 1962). Much of this work has been done with DMNA. Several research teams have found that DMNA is transformed by a demethylating microsomal enzyme into a very toxic and carcinogenic metabolite presumably 0113" (Hultin 33 31_., 1960; Emmelot 31 21., 1962; Mizrahi and Emmelot, 1962; Kriek and Emmelot, 1963). Model experiments have shown that CH + esterifies the secondary phosphate 3 groups of RNA to yield phosphate triesters which hydrolyze spontaneously (Kriek and Emmelot, 1963) resulting in the loss of its functionability. Magee and Farber (1962) found that DMNA is capable of methylating both RNA and DNA igflzizg to produce 7-methy1guanine. Magee (1966) reviewed previous work done on the significance of 7-methylguanine. This base (7-methylguanine) is normally absent or present in minute quantities in -7- the nucleic acids of untreated animals but is accepted as the major pro- duct of methylation of nucleic acids (Brookes and Lawley, 1964). Magee (1966) reported on several experiments where increases in the 7-methyl- guanine content of rat urine after the injection of DMNA were found. By the incorporation of radioactive labels, it was shown that this urinary 7-methylguanine came from.the cellular nucleic acids (DNA in this experi- ment). Whether this 7—methylguanine was released by cellular injury or by some sort of excision mechanism was not determined. Brookes and Lawley (1964) have suggested two mechanisms, at the molecular level, by which alkylating agents may induce genetic mutations. They hypothesized that alkylation on the 7-position of guanine may cause anomalous base-pairing (thymine pairing with the ionized.7-methylguanine) or that mutations may be caused by deletion by depurination with splitting out of 7-methylguanine from.the DNA chain. Reviews of the metabolic, carcinogenic and mutagenic action of the nitroso compounds have been written by several authors (Magee and Schoental, 1964; Magee, 1966; Miller and Miller, 1966; Shvemberger, 1967). Druckrey et al. (1963) and Jacobsohn (1966) have reported on the relation of chemical structure to carcinogenic activity and localization. Magee (1966) summarized the present situation when he stated that the alkyla- tion hypothesis of nitroso carcinogenesis is attractive but it is far from.being established. -8- Occurrence 23 Dimethylnitrosamine 22 Fish Meal During the years 1961 and 1962 outbreaks of "toxic hepatosis" in ruminants and fur-bearing animals were identified in Norway (KOppang, 1964; Hansen, 1964). Koppang (1964) described the liver injury as char- acterized by extensive centrolobular liver necrosis and vascular damage in the branches of the hepatic veins. He also observed that this dis- order occurred in animals fed a special type of herring meal produced from.nitrite-preserved herring. Ender 32 El‘ (1964) and Ender (1966) found that the toxic factor in the nitrite preserved herring meal was DMNA. Sakshaug 33 El. (1965) tested the toxicity of DMNA on sheep. Their results were in agreement with the previous findings. They observed that the feeding of DMNA and toxic herring meal to sheep resulted in very similar cases of hepatic injury. Furthermore, they detected DMNA by thin-layer chromatography (TLC) in two batches of toxic herring meal. Ender (1964) isolated and identified di- and trimethylamines in herring and suggested that the presence of DMNA in the fish meal resulted from.the reaction of the added nitrite with the naturally occurring methylamines in the fish. It is known that secondary amines combine with nitrous acid by elimr ination of water between two molecules and consequent replacement of the lone amino hydrogen atom by a nitroso group (~N20); the products being nitrosamines (Fieser and Fieser, 1957). They propose the following reaction: R\ R\ N - H + HON = O ----- > - N = O + H 0 Secondary amine + Nitrous acid ----- > Nitrosamine + Water In model experiments, Ender 32 El. (1967) have shown that not only di- and trimethylamines (commonly found in marine fishes) but also monomethyl— amine and trimethylamine—oxide (to a lesser degree) react with nitrite to form.DMNA. To further support the theory of DMNA formation from.the methylamines in the fish, these same authors processed ide (a freshawater fish which are claimed not to contain methylamines) with nitrite exactly as the herring meal was processed. The ide meal showed no detectable signs of toxicity when fed to mink in large daily doses over an extended period of time. The question of whether fresh-water fish contain trimethylamine or trimethylamine-oxide has broad implications as to the possibility of DMNA (or similar compound) formation in these fish after treatment with nitrite. Most of the older literature supports the idea that freshpwater fish contain no trimethylamine-oxide (Beatty, 1939; Tarr, 1941; Castell, 1949) while more recent data show that trimethylamine-oxide is present in the tissues and fluid of marine and freshpwater fish (Ronald and Jakobsen, 1947; Anderson and Fellers, 1952; Groninger, 1959). However, the researchers that do present evidence for trimethylamine-oxide in fresh-water fish show that the concentration of this compound is much lower in fresh-water fish than in marine fish. -10- Dyer (1952) cited three sources of trimethylamine-oxide and/or tri- methylamine in fish. First, it is present in the food supply. Tarr (1941) and Groninger (1959) reported that marine algae and plants con- tain more of these amines than do the comparable fresh-water species. Similar observations can be made on the zooplankton found in marine vs fresh-water habitats (Groninger, 1959). Secondly, Dyer (1952) reported that trimethylamine-oxide or trimethylamine can be formed by the bacter- ial degradation of choline or choline containing substances. Dyer and Wood (1947) and Bilinski (1962) reported similar observations. Dyer and Wood (1947) have listed some of the intestinal bacteria that are capable of decomposing choline to trimethylamine. Included in this list are Aerobacter aerogenes, Proteus rettgeri, P. ichthyosmins, E. rulgpgis, ‘2. mirabilis, and Schigella alkalescens. Asatoor and Simenhoff (1965) have proposed several pathways for the formation of urinary dimethylamine in humans (Figure 1). Whether these reactions take place in fish remains to be seen, however, the possibility is present. Tarr (1939) and Asatoor and Simenhoff (1965) have reported that bacteria can also reduce trimethylamine-oxide to trimethylamine in fish. Tarr (1941) has also proposed that the reverse oxidation occurs in fish liver. The third mechanism proposed by Dyer (1952) for the occurrence of trimethylamine is its synthesis from.ammonia through a detoxification pro- cess. This process is a series of methylations on the ammonia molecule to form.methylamine, dimethylamine, and trimethylamine. Although these mech- anisms have been postulated for the origin of the methylamines in fish, it is -11- ZZFOOCR R'OOO 0 H20 - o - fl _ 0011201123? 2 (CH3)3 I o (CH3)3- N= 0 I,— a - Lecithin Trimethylamine-oxide Lecithinase; Phosphatase + O . (CH3)3 E N - 012 - CH2 - 0H Bacteria > (CH3)3 :- 4N H Trim.thylamine Choline demethylation (0113)2 = N-H CO2 + N H3 Dimethylamine N\\\\ tr hylation CH3-N--H2 T NH2 Methylamine 3 2 2 Methionine CH3NH - CH2 - COOH Sarcosine Figure 1. Possible pathways for the formation of urinary dimethylamine. -12- hard to arrive at a definite conclusion as to the origin of the methyl- amines in fish (Groninger, 1959). The most widely held explanation as to the reason for the tremendous difference in the methylamine content between marine and fresh-water fishes is attributed to osmotic pressure regulation (Dyer, 1952; Groninger, 1959). Dyer (1952) reported that trimethylamine-oxide is necessary for osmoregulation in marine fish while it is not necessary for this process in fresh-water fish and is simply excreted in the urine. Therefore, it would be deducted from all this literature that the formation of DMNA in nitrite treated fish would have a higher incidence and concentration in marine than fresh-water fishes (Ender, 1964). Nevertheless,trimethylamine-oxide is unevenly distributed among marine animals depending upon geographic environment, species, season, size and location within the animal (Groninger, 1959). It is well known that the relative contents of fixed and volatile nitrogen bases furnish indications as to the freshness of fish. In living fish, a certain equilibrium is established between trimethylamine- oxide and trimethylamine. After death, the changes in the tissues are accompanied by a considerable increase in the di— and trimethylamine forms resulting from.the reduction of the oxide to the free bases (Borgstrom, 1961). Methods gf.Agalysis Prior to the discovery of the toxic and carcinogenic properties of DMNA and the other NAnitroso compounds, very few methods were available for their detection and specific identification. Some of the oldest -13— methods of detecting Nenitroso compounds were the unspecific spot tests (Feigl, 1960). Another more specific method is based on the knowledge that DMNA solutions absorb strongly in the near ultra-violet (Smales and Wilson, 1948), thus it is possible to identify DMNA by absorptivity measurements at 226 mu and 330 mu in aqueous solutions (Ender 33 al., 1964). In conjunction with the work of Barnes and Magee (1954) on the toxic and carcinogenic properties of DMNA, Heath and Jarvis (1955) developed a polarographic method for the analysis of DMNA in animal tissues. The method is specific for DMNA (although other nitroso compounds can be detected with slight modifications) and is sensitive enough to detect 1 ug of DMNA. Tiwari and Sharma (1963) reported a semi-micro iodometric method for the determination of the nitroso group. The same authors have also pub- lished another method for the determination of nitroso and azo groups by reduction with excess titanous sulfate and back titration of the excess reducing agent with a standard ferric sulfate solution (Tiwari and Sharma, 1964). Neurath, Pirmann, and Dunger (1964) developed a method whereby nitro- samines are reduced to their hydrazines by lithium aluminum.hydride, transformed to their respective 5-nitro-2-hydroxybenzal derivatives and separated by TLC. The plates are sprayed with a 3% potassium hydroxide ethanol solution resulting in a detection limit of about 0.5 ug. However, if the plates are sprayed with a 1% potassium.hexa cyano ferrate solution -14- 0.05 u gm.can be detected, but this spray reagent is not as specific as the previously mentioned reagent. Pruessmann 33 al. (1964a and b) reported a method by which mixtures of nitro compounds can be separated and individually identified. The compounds are detected on the thin-layer plates by reaction with a palla- dium chloride-diphenylamine solution under ultra-violet light (detection limit 0.5 - 1 ug) or with sulfanilic acidpa-naphthylamine in 30% acetic acid under ultra-violet light (detection limit 0.2-0.5 ug). Daiber and Preussmann (1964) developed a quantitative colorimetric method for the determination of nitrosamines and nitrosamides. The method is based on the photochemical degradation of the nitroso bond to form.nitrite which is measured by color formation with the Griess- Illosvay reagent. The sensitivity of this method is about 1-2 ug/ml. Ender (1966) and Ender 32 El. (1967) briefly mentioned a method by which DMNA is quantitatively reduced to dimethylhydrazine by zinc and hydrochloric acid which in turn is reacted with p-dimethylaminobenzal- dehyde to form a yellow aldazine. The amount of aldazine produced is measured colorimetrically at 458‘mu. The authors stated that less than 1 ppm.DMNA can be accurately measured. Having isolated a toxic factor from.berring meal, Ender 33 a1. (1964), Ender (1966), and Ender 23 21. (1967) used several methods for the quali- tative determination of DMNA. Included in these are gas-chromatographic studies; infra-red spectroscopy studies; maximum.of spectrophotometric absorption in water at 226 and 330 mu, and in carbon tetrachloride at 350 mu; and toxicological studies with mink. -15- Serfontein and Hurter (1966) reported a method for nitrosamine de- tection involving the formation of the hydrazines which are reacted with 4-nitrozaobenzene-4'-carboxylic acid chloride to form the respective hydrazides which are measured on TLC. The sensitivity of this method was not reported although the authors stated that 2.5 mg was easily detected. Mayrhofer and Mahler (1967) reported a gas-liquid chromatographic (GLC) method for the analysis of dimethyl- and diethylnitrosamine. The sensitivity as stated by the authors is 5 ug in 0.05 ml of methylene chloride. They have also devised a procedure whereby the nitrosamines are collected on TLC plates as they are expelled from.the gas chromato- graph. The value in this method arises from the fact that the nitrosamines from.several samples can be separated and collected as individual spots on the TLijlate. Once the nitrosamines are on the plate they are chromatographed and sprayed according to Preussmann.gtl§l. (1964 a and b). Lydersen and Nagy (1967) developed a polarographic method for the detection of DMNA in fish products. DMNA was separated from the sample by distillation from.a 3M sodium hydroxide solution at 40°C under a vacuum. The distillate was titrated with 0.01M sulfuric acid to a methyl red-methylene blue mixed indicator end point. Next 10 ml of supporting electrolyte (2Miammonium.sulfate, 2.5M sulfuric acid, and 1M calcium bromide) was added and the solution flushed for 30 minutes at 50°C with nitrogen gas. After cooling to room.temperature, the distillate is diluted to 50 ml with water and the DMNA content is determined polaro- graphically. The authors state that approximately 0.5 mg DMNA per kg of sample can be detected. -15- Mbhler and Mayrhofer (1967) have reviewed several of the known methods for the detection and identification of nitrosamines and commented on each method's potential as a means of analyzing foodstuffs for the nitrosamines at the predetermined leve1 set by the authors. Methods evaluated include polarography, spectroscopy (ultraAviolet, infrared, and fluorescence) plus calorimetry and gas-liquid chromatography incor- porated with thin-layer chromatography. Occurrence 2f Nitroso Compounds ig_Foodstuffs During recent years, organic Nenitroso compounds have been found to be potent carcinogenic and/or mutagenic substances. Their occurrence in foodstuffs is receiving considerable attention, especially in foods directly treated with nitrite or where the product comes into contact with smoke, exhaust gases or dehydration gases (Mbhler and Mayrhofer, 1967). Nitrosamines are easily formed from.secondary and tertiary amines after exposure to nitrous acid or nitrous gases (Preussmann, 1964a). DMNA has been detected in nitrite-treated herring meal (Ender £1 21., 1964). The possibility of nitrosamine occurrence in nitrite-treated meat and cheese products is now being studied (Mbhler and Mayrhofer, 1967). Kroeller (1967) found sporadic traces of Nenitrosodipropylamine or N; nitrosodiisopropylamine in samples of cheese. The occurrence of nitro- samines in the cheese was attributed to an abnormal ripening process. Nitrosamines are also found in some semi-hard cheeses that require a bacterial coating of its surface for maturing. Therefore, nitrosamine formation in this instance is apparently furthered by certain microorgan- isms (Anonymous, 1967). -17- Druckrey and Preussmann (1962) commented on the occurrence of nitro- samines and also implied their presence in tobacco smoke. Neurath (1967) found mixtures of Nenitrosodimethylamine and Nenitrosopyrrolidine in tobacco smoke. Kroeller (1967) demonstrated the presence of NLnitroso- methylbutylamine and N-nitrosopiperidine (0.02 ppm) in tobacco smoke. Several researchers have reported the occurrence of nitrosamines in unprocessed foodstuffs. A typical nitrosamine, p-methylnitrosamino- benzaldehyde, has been found and isolated from.an eatable mushroom, Clitocybe suaveolens (Preussmann, 1964a). Marquardt and Hedler (1966) reported that diethylnitrosamine could be detected in wheat flour by thin-layer chromatography. Mbhler and Mayrhofer (1967) reported finding dimethylnitrosamine in raw, unprocessed beef and in wheat meal. It has been reported that nitrosamines may occur in plants, especially in ger- minating plants (Sander, 1967). Serfontein and Smit (1967) presented evidence for the occurrence of NLnitrosamines in tobacco leaves. Sander (1967) investigated the possibility of the reaction between ingested secondary amines and nitrous acid (nitrite), in the stomach, to form.carcinogenic nitrosamines. ‘22 12252 tests have shown that nitro- samines can be formed in the gastric juice from.secondary amines and nitrite. He further stated that assuming the behavior of nitrosamines in cancerogenesis in man is similar to its behavior in animal tests, only part of the nitrite taken up with human food would be required to cause cancer by nitrosamine formation. EXPERIMENTAL PROCEDURE The Lake Michigan chub used in this study were supplied in the frozen state (about 8 kg per box) by the Bureau of Commercial Fisheries, U. S. Department of the Interior, Ann Arbor, Michigan. To facilitate future handling, the fish were thawed in running cold tap water after which they were vacuum.packed in plastic bags (four chub per bag), re-frozen and stored at -29°C until needed. Prior to utilization, the fish were re-thawed in running cold tap water, the scales removed, and groups of four (400-500 gm) were weighed and placed in 1500 m1 glass beakers in which they were brined. Treatments The variables studied included three levels of sodium.nitrite in the brine combined with two cooking temperatures. All fish were brined in a 50° salinometer brine (68.5 gm sodium chloride per 454 gm.cf water) for 16 to 18 hours at a temperature of 4 to 6°C. The brine to fish ratio was 2:1 (w/w). The chub were kept submerged in the brine by placing glass funnels over them. Batches 1, 6, and 9 received this treatment. Other groups were similarly cured but two levels of sodium.nitrite were added to the brining medium. Batches 2, 4, and 10 received 3 gm of sodium nitrite per 2.3 kg of fish while batches 3, 5, and 11 received 5 gm of sodium.nitrite per 2.3 kg of fish. The nitrite content of the latter two brine solutions was approximately 614 ppm and 1,023 ppm, respectively. -]_8- -19- After curing, the chub were removed from.the brine, placed (belly down) on a broiling pan and cooked in an electric oven. Two different processing temperatures were employed, 107°C and 149°C. However, the cooking times varied between 2 1/2 to 3 hours depending upon the time it took for the internal fish loin temperature to reach the desired temper- ature of 82°C or 107°C (180°F or 225°F) and be maintained there for 30 minutes. The oven and internal fish loin temperatures were measured on a 12 point recording potentiometer (copper constantan thermocouples). Following processing, the chub were placed in the cooler (4-6°C) for several hours. The skin and fins were removed and the loin muscles were separated from.the rest of the fish. Loin muscles from similarly cured and.processed chub were collected together, mixed and ground three times through a 6.3 mm.plate, placed in glass sample bottles, the lids tightly screwed on, and stored in the cooler until the chemical analyses were determined. Dimethylnitrosamine Determination Fifty grams of the finely ground fish sample were weighed into a 500 ml round bottom.boiling flask to which 15 gm of calcium.hydroxide and 150 ml of distilled deionized water were added and mixed thoroughly. Several glass beads were added to prevent bumping and half the solution (75 ml) was distilled through asplash—head and Liebig condenser. The distillate was collected in a 100 ml graduated cylinder. -20- The DMNA, if present, was separated from.the distillate by extract- ing 4 times with 20 m1 of methylene chloride in a 250 ml separatory funnel. The DMNA containing methylene chloride solution was collected in a 125 ml Erlenmeyer flask and evaporated (under a stream of nitrogen gas in a warm water bath) to a volume between 20 and 25 ml. The solvent was then transferred to a 25 ml graduated test tube. The Erlenmeyer flask was washed three times with methylene chloride and the washings were added to the remaining sample. The solvent was further evaporated until 2 m1 of the methylene chloride extract remained. Two hundred ul of this solution were spotted on a thin-layer plate (Silica gel G Merck, 250 u, activated at 100°C for one hour), chromato- graphed in a mixture of hexane, di-ethyl ether, methylene chloride (4:3:2), and irradiated for 3 minutes under an unfiltered ultra-violet lamp (Mineralight R-51). To develop the DMNA spots, the plates were sprayed lightly with a 1:1 (V/V) solution of 1.0% sulfanilic acid in 30% acetic acid and 0.1% a-naphthylamine hydrochloride in 30% acetic acid. The TLC techniques employed were first described by Preussmann st al. (1964a and b). The only modification was the substitution of'a-naphthylamine hydrochloride in lieu of a-naphthylamine. This change was necessary since the plates turned pinkish-red following spraying with the solution containing a-naphthylamine. Thus, the pink DMNA spots could not be differentiated from.the background interference. This undesirable effect was almost completely eliminated by incorporation of the hydrochloride salt. - 21- Preussmann gt al. (1964b) reported two spray reagents for the devel- opment of nitrosamine spots. In order to make a positive identification of a nitroso compound, both sprays must give positive results. However, the sulfanilic acidau-naphthylamine hydrochloride spray was used since the other reported spray (palladium.chloride-diphenylamine) brought out some interfering compound (blue spot, Rf 0.25) close to the blue DMNA spot (Rf 0.22). Because this unknown spot was not detectable with the sulfanilic acidsa-naphthylamine hydrochloride spray and since the Rf value was different, it was assumed that DMNA was not responsible for the colored area developed by the palladium.chloride-diphenylamine spray. Prior to the adoption of this method several other methods were studied. These procedures and their undesirable results will be dis- cussed in the Results and Discussion section. Moisture Determination The moisture content of the samples was determined according to the method described in A.O.A.C. (1965). Five grams of fish were placed in an aluminum foil dish and dried in a 100°C oven for 24 hours. Ether Extract Determination Ether extract was determined from.the same samples used in the moisture analysis. The fat was extracted with anhydrous ether for 3 1/2 hours in a Goldfisch.Fat Extractor. - 22- Nitrite Determination Nitrite was determined by a modification of the procedures outlined in A.O.A.C. (1965). The modification employed a gravimetric procedure in lieu of the volumetric procedure as described in A.O.A.C. (1965). Three grams of the fish sample were weighed into a 500 ml Erlenmeyer flask. Approximately 300 gm.cf 80°C water (distilled deionized) were added. The flask was placed in a steam.bath for 2 hours after which 5 ml of a saturated mercuric chloride solution was added and mixed thoroughly. The solution was allowed to cool to room.temperature and the weight was increased to 500 gm by the addition of distilled deionized.water. The solution was filtered and added to 2 m1 of Griess reagent in a 50 ml volumetric flask. The Griess reagent consisted of 1:1 (E/V) solution of 1% sulfanilic acid in 30% acetic acid and 0.1%Ia-naphthylamine in 30% acetic acid. The nitrite content was determined colorimetrically by comparison with a standard nitrite curve. p§>Determination Five grams of the fish sample were mixed with five ml of distilled deionized water and the pH of the resulting slurry was determined by a Corning Model 15 pH meter. -23- Salt Concentration 2g :23 water Phase Determination The percent sodium.chloride was determined by a modification of the Vohlard method as outlined in the A.O.A.C. (1965). One and a half grams of fish were weighed into a 250 m1 Erlenmeyer flask. Twenty five m1 of a 0.1N silver nitrate solution were added to precipitate all the chloride and provide an excess of silver nitrate. The solution was mixed and 15 ml of concentrated nitric acid were added. Next, the mixture was boiled (5-10 minutes) until all the fish disintegrated. A 5% potassium.permanganate solution was added in successive small portions until the permanganate color almost disappeared after each addi- tion and continued until the solution was colorless or nearly so. Twenty five ml of distilled deionized water were added and the solution boiled for 5 minutes to eliminate the lower oxides of nitrogen. The solution was diluted to approximately 150 ml and 25 ml of diethyl ether were added to dissolve the fat. Five ml of ferric indicator and 5 ml of a 1:1 solution of nitric acid and.water were added. Next, the excess silver was titrated with potassium thiocynate until a permanent light brown color appeared. Then from.the volume (ml) of silver nitrate used, the quantity of chloride was calculated as follows: (ml of O.1N.Ag_NO3 used) X (0.58% NaC1) I(sample wt:) % NaCI = The salt concentration in the water phase was calculated as follows: ( NaCl) x (100) % NaCl 3 gm N301) 4. (gm H20 in sample) RESULTS AND DISCUSSION A Report 23 Preliminary Researgh The major portion of this study was devoted to the development of a method to detect DMNA in chub at concentrations of less than 1 ppm. At the onset of this work very few methods for the analysis of DMNA could be found in the literature and even less work had been done with DMNA in biological systems. Heath and Jarvis (1955) reported a polarographic method for the determination of DMNA in animal tissues. The procedure was tried on model systems of aqueous DMNA solutions. However, inconsistent results were obtained. Consultation with a staff member from.the Chemistry Department who has had considerable experience with polarographic tech- niques, led to the idea that a change of the buffer might result in more consistent results. After using other buffers which did not improve repeatability, the method was discontinued. Next, the colorimetric method described by Daiber and Preussmann (1964) was examined. One ml of a standard DMNA solution (1 to 10 ppm) was combined with 0.5 ml of a 5% sodium.carbonate solution and irradiated for 15 minutes under UV light (Mineralight R951). To this 3 m1 of Griess reagent were added and the color development was measured spectrophoto- metrically at 528 mu after a 15 minute development period. Theoretically, the UV light breaks the nitroso bond and the cleaved product reacts to form.nitrite in the solution. This nitrite reacts with the Griess reagent -24— -25- to form a red azo dye. The photochemical degradation of the nitrosamine linkage was very inefficient and extremely erratic. In an effort to increase the efficiency of the irradiation process, the length of ex- posure and the type of irradiation vessel were studied. Irradiation beyond 15 to 20 minutes did not result in any benefit. Daiber and Preussmann (1964) stated that the nitrosamine cleavage is maximum at 15 minutes and that pure nitrite solutions change their nitrite content with irradiation. Nitrite is oxidized to nitrate under UV radiation. The writer substantiated these results. Samples were placed in 10 ml beakers, watch glasses, and flame photometer flasks and subjected to overhead irradiation. NOne of these variations seemed to increase the nitrosamine cleavage. Mechanical agitation of the sample during irra- diation also failed to improve the efficiency of cleavage. However, when the sample was placed in spectrophotometer cuvettes (silica) and irradiated from.the two transparent surfaces, excellent repeatability and a high degree of cleavage were obtained. Fish muscle was extracted and distilled according to the procedure employed by Heath and Jarvis (1955). However, nitrite was found to be present in the distillate. It became necessary to separate the nitrite from.the DMNA since after UV irradiation the amount of cleaved nitrite was used as an index of the DMNA content. Separating the two compounds presented a problem due to the solubility characteristics of both compounds. DMNA is very soluble in most polar and non polar solvents and most nitrite salts are readily soluble in aqueous solutions. After -26- several unsuccessful attempts to separate the two compounds, a method was devised which eliminated all detectable traces of nitrite with minimal losses of DMNA. This was accompliShed by placing a 60 cm distillation column (packed with glass beads and covered by an electrical heating tape) between the boiling flask and the condenser. During distillation, the heating tape was regulated to keep the column at a temperature of approximately 90°C. Although no nitrite could be detected in the dis- tillate, a positive nitrite reaction was obtained after the UV treatment. The author postulates that some substance(s) could be oxidized during the distillation or irradiation processes and thus give a positive nitrite reaction with the Griess reagent. Possibly nitrite (or similar substance) could be oxidized, traverse the column in the vapor phase and be reduced to a reactable form.by the UV irradiation. Because of this, the system was purged with nitrogen gas before and during the distillation process. In addition, the samples were irradiated under a nitrogen atmosphere in the cuvettes. These procedures eliminated all interfering compounds from aqueous nitrite-nitrate-DMNA solutions. The same procedures were applied to fish extracts and interfering substances were found. After several unsuccessful attempts to eliminate the interference, the method was set aside. Since a very sensitive method was desired, the author directed his attention to fluorescence as a possible solution to the problem. NOthing pertaining to the problem.could be found in the literature. Therefore, the author tested to see if DMNA did fluoresce. From.these studies it was concluded that DMNA does not fluoresce in the tested range. Excita- tion and emission wavelengths were scanned from 200 to 800 mu inclusively on an Aminco-Bowman Spectrophotofluorometer. This involved observing the emission wavelengths from.200 to 800 mu for each excitation wave- length between 200 and 800 mu. Similarly, the azo dye formed by the reaction of nitrite (from.DMNA) with the Griess reagent exhibited maximum.fluorescence at an emission wavelength of 346 mu when excited at 306 mu. However, the sulfanilic acid-acetic acid solution also showed a fluorescent peak at 346 mu when excited at 303 mu. Therefore, it was not known whether the emission was from.the azo dye or from.the excess sulfanilic acid. Research in this area was abandoned because of the inability to differentiate between the fluorescence of each compound. However, in a subsequent review of the literature a report by Mohler and Mayrhofer (1967) was discovered which described their attempts to develop a fluorescence method for nitrosamine determination on TLC. The TLC nitrosamine spots emitted a bluish fluorescence under filtered UV light. The bluish fluorescence was not detected in aqueous, hexane or methylene chloride solutions. The authors reported that concentrations exceeding 0.1% could not be detected. Interference emissions from the solvents were also experienced. The TLC methods described by Preussmann.g£ al. (1964a and b) were investigated. The sensitivity of the method (as reported by the authors) is between 0.5 and 1 pg of nitrosamine. Aqueous solutions of DMNA in the range of 250 ppm.could be detected when 4 pl of the solution were -28- spotted on the plate. Projecting this type of reasoning, theoretically 1 m1 of a l ppm.DMNA solution would have to be spotted on the plate for a positive test. Spotting such a large amount of an aqueous solution is very difficult and time consuming if a reasonably small spot is main- tained. Liquid-liquid extraction of the aqueous solution appeared to be a likely solution. Smales and Wilson (1948) found that chloroform and carbon tetrachloride had very unfavorable distribution coefficients although they could be increased by the presence of salts or alkalis in the aqueous phase. Diethyl ether was tested by the author and found to be inefficient at low DMNA concentrations even when a continuous liquid- liquid extraction was maintained for 24 hours. Mohler and Mayrhofer (1967) successfully extracted DMNA from aqueous solutions with.methylene chloride. Therefore, a methylene chloride extraction was incorporated in the method for DMNA analysis as outlined in the experimental proce- dure. This combination worked very well since the DMNA concentration could be increased by reducing the methylene chloride volume by evapora- tion. Because DMNA is so volatile, Mohler and Mayrhofer (1967) recommended evaporating the methylene chloride solution to a volume of 10 ml. Below this level they encountered rather large losses of DMNA. However, it was the author's experience that good results were obtained when the solution was reduced to a volume of 2 ml. When DMNA was added to ground chub loin muscle to provide a concentration of 0.5 ppm, and following the previously outlined procedures, a positive DMNA test could be repeat- edly obtained when 0.2 ml (200 pl) of the methylene chloride extract were spotted. -29- The distillation procedures of Heath and Jarvis (1955) were modified for the samples to be analyzed by TLC. Heath and Jarvis (1955) distilled an aqueous rat tissue extract whereas the author placed the fish sample directly in the boiling flaSk. Sodium hydroxide was replaced by calcium hydroxide as the alkali for increasing the concentration of DMNA in the vapor phase during distillation. It was found that the replacement of sodium hydroxide with calcium hydroxide resulted in much less foaming during distillation. Emphasis should again be placed on the evaluation of the developed .plates. Preussmann gt_al, (1964b) definitely stated that the colors are only specific if both sprays give positive results. Chub extracts Sprayed with the palladium chloride-diphenylamine reagent developed a blue spot (Rf 0.25) when chromatographed for approximately 13 cm. Standard DMNA spots are also blue but have Rf values of 0.22 when similarly chromatographed. A comparable volume of the same fish extract was chromatographed and sprayed with the sulfanilic acids! naphthylamine hydrochloride reagent. A spot could not be seen in the expected area. However, when DMNA was added to the fish extract before chromatographing, a typical DMNA spot (Rf 0.22) was seen along with the suspect compound (Rf 0.25) when sprayed with the palladium chloride-diphenylamine reagent. Likewise, only the DMNA spot was visible after spraying with the sulfanilic acidsa naphthylamine reagent. Thus, the suspect spot was not DMNA. No attempts were made to determine the identity of this compound. Preuss- nenn.gt al. (1964b) discussed possible interfering compounds in connection with the use of both spray reagents. Discussion 2: Experimental Results Ender §_t_ _a_1. (1964), Ender (1966), and Ender 9:5 31. (1967) reported that DMNA formation has been observed during the processing of nitrite treated herring into herring meal. They hypothesized that the nitrite combines with the natural methylamines in the fish to form.the nitroso derivative. In model experiments, they were able to synthesize DMNA from.monomethylamine, dimethylamine, trimethylamine, and trimethylamine.- oxide (Ender st 21., 1967). Ender (1966) and Ender st 31. (1967) re- ported that by increasing the nitrite content of the fish or by increasing the cooking and drying temperatures of the meal they could increase the amount of DMNA formed. In model experiments (buffered solutions), DMNA could be formed at 4°C if the amines and nitrite were allowed to react for relatively long periods of time as compared to the time necessary for DMNA formation at 100°C. DMNA formation at two different pH's were also studied by Ender 33 El. (1967). They found a greater quantity of DMNA formed at a pH of 6.0 as compared to similar reactions conducted at a pH of 6.5. However, they did not report any work on altering the fish pH during processing to see if pH did have an effect on DMNA formation. To test their hypothesis they processed ide, a freshpwater species (which are thought to contain only traces of the methylamines), in a similar manner as that which produced the toxic herring meal. The ide meal showed no detectable signs of toxicity when fed to mink in large daily doses over an extended period of time. This study was undertaken to determine if DMNA was formed in chub under the specified treatment of nitrite brining and cooking. -31- The results of this study are tabulated in tables 1 and 2. Attention is directed to the different temperature processes, which indicate the internal fish temperature reached and maintained for 30 minutes. Table 1. Results of chemical analyses of chub loin muscle receiving the 82°C process. gm NaNOz WNaCf added.per % % Ether in water ppm Batch 2.3 kg chub pH Moisture extract phase NaN02 DMNA 1 o 6.6 63.96 12.45 5.48 . 4 -1 2 3 6.7 63.30 13.47 4.98 108 -- 3 5 6.7 63.90 11.51 6.36 201 -- 6 0 6.7 61.90 13.44 5.40 5 -- 7 3 6.7 62.10 14.30 5.36 90 -- 8 5 6.7 62.35 14.31 6.73 110 -- 1Less than 0.5 ppm The range in values of the different treatments exemplifies the diversity of biological systems. Neither the level of nitrite in the brine nor the processing temperature appeared to have any effect on DMNA formation since all DMNA tests were negative. These results are in agreement with the findings of Ender g: 31. (1964), Ender (1966), and Ender g: 31. (1967) who also were unable to detect DMNA in freshpwater fish after processing with nitrite. The author, like Ender 33 El. (1967), was able to produce detectable levels of DMNA by refluxing 1.5 gm.cf di- or trimethylamine hydrochloride with 1.0 gm sodium nitrite in 200 m1 of an acetic acid-sodium.acetate Table 2. Results of chemical analyses of chub loin muscle receiving the 107°C process. gm.NaN02 ‘_%9Na01 added per % % Ether in water ppm Batch 2.3 kg chub pH Moisture extract phase NaN02 DMNA 4 3 6.6 59.29 15.04 6.61 96 -1 5 5 6.8 58.71 16.39 6.75 160 -- 9 0 6.7 57.70 18.49 11.12 3 -- 10 3 6.7 54.38 16.95 7.00 126 -- ll 5 6.7 55.69 14.16 8.69 183 -- 1 Less than 0.5 ppm. buffer (pH 6.0) for 1 hour. Several chub, which had been cured in a nitrite brine were injected with an aqueous solution of dimethylamine hydrochloride immediately before the heat process. 82°C a positive DMNA test was obtained. After processing at This also tends to support the theory that the methylamines must be present in rather high concentra- tions before DMNA formation can be detected. SUMMARY.AND CONCLUSIONS A method was developed to detect dimethylnitrosamine in chub loin muscles at a concentration of 0.5 ppm. The variables studied (three levels of nitrite brining combined with two processing temperatures) apparently did not result in the formation of dimethylnitrosamine. All fish loin muscle samples tested gave negative results, indicating DMNA concentrations in the chub, if present, were less than 0.5 ppm. -33- BIBLIOGRAPHY Anderson, D. W., and C. R. Fellers. 1952. The occurrence of trimethyl- amine and trimethylamine-oxide in fresh water fishes. Food Research 17: 4720 Anonymous. 1964. Botulism.outbreak from.smoked Whitefish. Food Technol. 18:71. Anonymous. 1967. Report of the Ninth Session of the Joint FAQ/WHO Committee of Governments Experts on the Code of Principles concern- ing Milk and Milk Products, Section II - Report of Subcommittee I on Food.Additives (p. 19, No. 60 of the Report). A. 0. A. C. 1965. Qiiigig Ethodfi of: Aglxgis of the A, 0.11.0. A. 0. A. C. , waSh. , D. C. , Tenth Edition. Asatoor, A. M., and M. L. Simenhoff. 1965. The origin of urinary dimethylamine. Biochemu and Biophys. Acta. 111:384. Barnes, J. M., and P. N. Magee. 1954. Some toxic properties of dimethyl- nitrosamine. Brit. J. Industr. Med. 11:167. Beatty, S. A. 1939. Studies of fish spoilage. III. The trimethylamine-1 oxide content of the muscles of Neva Scotia fish. J. Fish. Res. Bd. Canada 4:229. Bilinski, E. 1962. Biosynthesis of trimethylammonium.compounds in aquatic animals. III. Choline metabolism.in marine crustacea. J. Fish. Res. Bd. Canada ,l2’505° Borgstrom, G. 1961. Eish as Food I Production, Biochemistgy, and Migrobiology. Academic Press, New York p. 155. Bratzler, L. J., and M. 0. Robinson. 1967. Processing Great Lakes chub (Leucichthyg ho 1). Quarterly Bulletin of the Michigan Agri. Expt. Station, Michigan State University'42(4):440. Brookes, P., and P. D. Lawley. 1964. Reaction of some mutagenic and carcinogenic compounds with nucleic acids. J. Cell Comp. Physiol. 64:111. Brouwers, J. A. J., and P. Emmelot. 1960. Microsomal Nedemethylation and the effect of the hepatic carcinogen dimethylnitrosamine on amino acid incorporation into the proteins of rat livers and hepatomas. Exp. Cell Res. 19:467. -34— -35- Castell, C. H. 1949. Bacteria and the reduction of trimethylamine-oxide. Food in Canada g(10):11. Daiber, D., and.R. Preussmann. 1964. Quantitative determination of organic Nenitroso compounds with the photochemical degradation of the nitrosamine linkage. Z. Anal. Chem. 226:344. Druckrey, H., R. Preussmann, D. Schmahl, and M. Miller. 1961. Pro- duction of a gastric carcinoma by nitrosamines in rats. Natur- wissenschaften 48, 165. Druckrey, H., and R. Preussmann. 1962. The formation of carcinogenic nitrosamines for example, in tobacco smoke. Naturwissenschaften 49:498. Druckrey, H., R. Preussmann, and D. Schmhl. 1963. Carcinogenicity and chemical structure of nitrosamines. Acta Un. int. Cancer. 19:510. Dutton, A. H., and D. F. Heath. 1956. Demethylation of dimethylnitro- samine in rats and mice. Nature 128:644. Dyer, F. E., and A. J. Wood. 1947. Action of enterobaoteriaceae on choline and related compounds. J. Fish. Res. Bd. Canada 2317. Dyer, W. T. 1952. Amines in fish muscle. ‘VI. Trimethylamineroxide content of fish and marine invertebrates. J. Fish. Res. Bd. Canada‘§:3l4. Emmelot, P., I. J. Mizrahi and E. Kriek. 1962. Prevention by cysteamine of the inhibitory effect of carcinogenic Nenitrosodialkylamines on incorporation of amino acids in rat liver. Nature 123:1158. Ender, F., G. Havre, A. Helgebostad, N. Koppang, R. Madsen, and L. Ceh. 1964. Isolation and identification of a hepatotoxic factor in herring meal produced from.sodium.nitrite preserved herring. Naturwissenschaften 51:637. Ender, F. 1966. NLnitrosodimethylamine, the active principle in cases of herring meal poisoning. VIII International Buiatric Congress, Zurich. Ender, F., G. Havre, R. Madsen, L. Ceh, and.A. Helgebostad. 1967. Studies on conditions under which Nenitrosodimethylamine is formed in herring meal produced from.nitrite-preserved herring. The risk of using nitrite uncritically as a preservative agent. Z. Tierphysi01., Tierernaedr. Futtermittelk. 22:181. Feigl, F. 1966. S 0t Tests In Organic Chemistry. Elsevier Publishing Company, New ark. 7th Edition. Fieser, L. F., and M. Fieser. 1957. Introduction To Organic Chemistgy. D. C. Heath and Company, Boston p. 188. Foster, E. M., J. S. Deffner, T. L. Bott, and E. McCoy. 1965. Clostri- dium botulinism.food poisoning. J. Milk and Food Technol. 28:86. Freund, H. A. 1937. Clinical manifestations and studies in parenchya matons hepatitis. Ann. Int. Med. 29:1144. Groninger, H. S. 1959. Occurrence and significance of trimethylamine- oxide in marine animals. United States Fish and Wildlife Service, Special Scientific Report -- Fisheries No. 333. Hamilton, Alice, and Harriet L. Hardy. 1949. Industrial Toxicology. 2nd Edition. P. B. Hoeber, New York. p. 301. Hansen, M. A. 1964. An outbreak of toxic liver injury in ruminants. Clinical observations and results of some hepatic tests in cattle and sheep. Nerd. Vet.-Med. 16:323. Heath, D. F., and J. A. E. Jarvis. 1955. The polarographic determina- tion of dimethylnitrosamine in animal tissue. Analyst 29:613. Heath, D. F. 1962. The decomposition and toxicity of dialkylnitrosamines in rats. BiOChem-o Jo §:720 Hultin, T., E. Arrhenius, H. L'ow, and P. Magee. 1960. Toxic liver injury inhibition by dimethylnitrosamine of incorporation of labelled amino acids into proteins of rat-liver preparations ig vitro. Biochem. J. 76:109. Jacobsohn, K. 1966. Chemical structure and cancer localization. Rev. Port. Quim. §_:192 : Chem. Abstr. 91:19573u (1967). Jacobson, K. H., H. J. Wheelwright, J. H. Clem, and.R. N. Shannon. 1955. Studies on the toxicology of NLnitrosodimethylamine vapor. A.M.A. Arch. Ind. Health 12:617. Koppang, N. 1964. An outbreak of toxic liver injury in ruminants. Case reports pathological-anatomical investigations, and feeding experi- mentS. Nerd. veto-Med. $9.: 305. Kriek, E., and P. Emmelot. 1963. Methylation and breakdown of micro- somal and soluble ribonucleic acid from.rat liver by diazomethane. Biochemistryl2:733. Kroeller, E. 1967. Detection of nitrosamines in tobacco smoke and food. Deut. Lebensm.-Rundsch. 99:303 : Chem. Abstr. 91:115822v (1967). Lydersen, D. L., and K. Nagy. 1967. Polarographic determination of dimethylnitrosamine in fish products. Fresenins Z. Anal. Chem. 229: 277 : Chemp Abstr. 91399021a (1967). Magee, P. N. 1956. Toxic liver injury. The metabolism of dimethyl- nitrosamine. Biochem. J. 93:676. Magee, P. N., and J. M. Barnes. 1956. The production of malignant primary hepatic tumours in the rat by feeding dimethylnitrosamine. Brit. J. Cancer _1_0_:ll4. Magee, P. N. 1958. Toxic liver injury. Inhibition of protein : syn- thesis in rat liver by dimethylnitrosamine $3 vivo. Biochem. J. 70:606. Magee, P. N., and J. M. Barnes. 1962. Induction of kidney tumors in rats with dimethylnitrosamine (Nenitrosodimethylamine). J. Pathol. Bacteriol.‘§é:19. Magee, P. N., and E. Farber. 1962. Toxic liver injury and carcinogenesis Methylation of rat-liver nucleic acids by dimethylnitrosamine i2 vivo. Biochem. J. ‘92:114. Magee, P. N., and T. Hultin. 1962. Toxic liver injury and carcinogenesis Methylation of proteins of rat-liver slices by dimethylnitrosamine ‘23 vitro. Biochem. J. 92:106. Magee, P. N., and R. Schoental. 1964. Carcinogenesis by nitroso com, pounds. Brit. Med. Bull. 29:102. Magee, P. N. 1966. Alkylation of nucleic acids and carcinogenesis. Colloq. Ges. Physiol. Chem. 21579. Marquardt, P., and L. Hedler. 1966. Occurrence of nitrosamines in wheat flour. Arzneimittel.-Forsch 29:778 : Chem. Abstr. 99:9616c (1966). Mayrhofer, O. L., and K. Mbhler. 1967. Integration of gas-chromato- graphic fractions with thin—layer plates. Z. Lebensm.vUnters. Forsch. '22g:246 : Chem. Abstr. 91:1158262 (1967) Miller, E. C., and J. A. Miller. 1966. Mechanisms of chemical carcino- genesis: nature of proximate carcinogens and interactions with micromolecules. Pharmacol. Rev. 99:805. Mizrahi, I. J., and P. Emmelot. 1962. Effect of cysteine on the metabolic changes produced by two carcinogenic Nenitrosodialkylamines in rat liver. Cancer Res. 22:339. —38- Mghler, K., and 0. L. Mayrhofer. 1967. Detection and determination of nitrosamines in foodstuffs. Zeitschrift ffir Lebersmittel-Untersuchung und Forschung (in press). Morrison, R. T., and R. N. Boyd. 1964. Organic Chemistry. Tenth Edition. Allyn and Bacon, Inc., Boston. p. 563. Neurath, G., B. Pirman, and ML Dunger. 1964. Identification of Nenitroso compounds and unsymmetrical hydrazines as 5-nitro-2-hydroxy benzal derivatives and application on a micro scale. Chem. Ber. 91:1631. Neurath, G. 1967. Occurrence of Nenitroso compounds in tobacco smoke. Experientia 22:400 : Chem. Abstr. 92g29937c (1967). Patashnik, M., C. F. Lee, H. L. Seagran, and F. B. Sanford. 1964. Pre- liminary report on experimental smoking of chub (Leucichthys SP.). Commercial Fisheries Review 29(11):1. Pivnik, Rubin, Barnett, N0rdin, Ferguson, and Perrin. 1967. Effect of sodium.nitrite and temperature on toxinogenesis by Clostridium botulinum in perishable cooked meats vacuumrpacked in airéimpermeable pIast1c pouches. Food Technol. 22(2):100. Preussmann, R., D. Daiber, and H. Hengy. 1964a. A sensitive colour reaction for nitrosamines on thin-layer chromatograms. Nature 29;: 502. Preussmann, R., G. Neurath, G. Wulf-Lorentzen, D. Daiber, and H. Hengy. 1964b. Coloring methods and thin layer chromatography of organic N-nitroso compounds. Z. Anal. Chem. 292:187. Ronald, 0. A., and F. Jakobsen. 1947. Trimethylamine-oxide in marine products. J. Soc. Chem. Ind. (London) 99:160. Sakshaug, J., E. Sggnen, M. A. Hansen, and N. Koppang. 1965. Dimethyl- nitrosamine; its hepatoxic effect in sheep and its occurrance in toxic batches of herring meal. Nature 299:1261. Sander, J. 1967. Kann nitrit in der menschlichen nahrung ursache einer Krebsentstehung durch nitrosaminblidung sein? Archiv ffir Hygiene und Bakteriologie 199(1/2):22. Sax, N. I. 1963. Dangerous Propertiggof Industrial Materials. Rein- hold Publishing Corporation, New YOrk. p. 1048. Scheflan, L., and M. B. Jacobs. 1953. The Handbook of Solvents. D. Van Nostrand Company, Inc., New York. p. 561. -39- Schmahl, D., and R. Preussmann. 1959. The carcinogenic effect of nitro- sodimethylamine in rats. Naturwissenschaften 39:175. Schoental, R. 1960. Carcinogenic action of diazomethane and of nitroso- N—methyl urethane. Nature 999:420. Schoental, R. 1961. Interaction of the carcinogenic Nemethyl-nitroso- urethane with sulphydryl groups. Nature 292:670. Serfontein, W. J., and P. Hurter. 1966. A method for identifying small amounts of nitrosamines in biological material. Nature 299:1238. Serfontein, W. J., and J. J. Smit. 1967. Evidence for the occurrence of Nenitrosamines in tobacco. Nature 223:169. Shvemberger, I. N. 1967. Carcinogenic nitrosamines. Vop. Onkol. 29:97 : Chem. Abstr. 99:63430y (1967). Smales, A. A., and H. N. Wilson. 1948. The determination of small amounts of dimethylamine in aqueous solutions containing ammonia, mono- and trimethylamine. J. Soc. Chem. Ind. 913210. Tarr, H. L. 1939. The bacterial reduction of trimethylamine-oxide to trimethylamine. J. Fish. Res. Bd. Canada 5:367. Tarr, H. L. 1941. The fate of trimethylamine-oxide and trimethylamine in man. J. Fish. Res. Bd. Canada 9:211. Tarr, H. L. 1942. The action of nitrites on bacteria: Further 3X? periments. J. Fish. Res. Bd. Canada 9;74. Terracini, B., and P. N. Magee. 1964. Renal tumors in rats following injection of dimethylnitrosamine at birth. Nature 292:502. Tiwari, R. D., and J. P. Sharma. 1963. Semimicro iodometric determina- tion of nitroso group. Proc. Natl. Acad. Sci., India Sect. A33, pt 3, p 379 : Chem. Abstr. 99:1111d (1964). Tiwari, R. D., and J. P. Sharma. 1964. Micro and semimicro determination of nitroso and azo groups by reduction with titanous sulfate. Indian J. Chem. 2:173 : Chem. Abstr. 91510045h (1964). Zak, E. G., J. H. Holzner, E. J. Singer, and H. Popper. 1960. Renal and pulmonary tumors in rats fed dimethylnitrosamine. Cancer Res. 29:96.