Mutagens from the cooking of food. II. Survey by Ames/Salmonella test of mutagen formation in the major protein-rich foods of the American diet

The formation of mutagens in the major cooked protein-rich foods in the US diet was studied in the Ames Salmonella typhimurium test. The nine protein-rich foods most commonly eaten in the USA—ground beef, beef steak, eggs, pork chops, fried chicken, pot-roasted beef, ham, roast beef and bacon—were examined for their mutagenicity towards S. typhimurium TA1538 after normal ‘household’ cooking (deep frying, griddle/pan frying, baking/roasting, broiling, stewing, braising or boiling at 100–475°C). Well-done fried ground beef, beef steak, ham, pork chops and bacon showed significant mutagen formation. For chicken and beef steak high-temperature broiling produced the most mutagenicity, followed by baking/roasting and frying. Stewing, braising and deep frying produced little mutagen. Eggs and egg products produced mutagens only after cooking at high temperatures (the yolk to a greater extent than the white). Commercially cooked hamburgers showed a wide range of mutagenic activity. We conclude that mutagen formation following cooking of protein-containing foods is a complex function of food type, cooking time and cooking temperature. It seems clear that all the major protein-rich foods if cooked to a well-done state on the griddle (eggs only at temperatures above 225°C) or by broiling will contain mutagens detectable by the Ames/Salmonella assay. This survey is a step towards determining whether any human health hazard results from cooking protein-rich foods. Further testing in both short- and long-term genotoxicity bioassays and carcinogenesis assays are needed before any human risk extrapolations can be made.     

Effect of cooking methods on the mutagenicity of food and on urinary mutagenicity of human consumers

The effects of cooking methods on the in vitro mutagenicity of individual foods, the in vitro mutagenicity of meals containing those foods, and the mutagenic exposure of human volunteers following consumption of the meals were examined using Ames bacterial strain TA98 with S-9 metabolic activation. Three methods of food preparation--boiling, baking and frying/flame-broiling--were compared. With meats, frying or broiling resulted in higher in vitro mutagenicity (10- to 50-fold) than did baking or boiling, whereas for carbohydrates, eggs or vegetables mutagenicity did not vary markedly with cooking method. The observed (experimental) mutagenic activity of the meals was quite similar to their calculated (predicted) mutagenicity, obtained by summing the mutagenicity of the individual foods in the meal. The close agreement between experimental and predicted mutagenicity indicated that components of the meal did not interact in either a synergistic or inhibitory manner. The mutagenicity of fried flame-broiled meals was approximately 10-fold greater than the mutagenicity of baked or broiled meals, which were similar in mutagenicity. The mutagenicity of human urine following consumption of the meals was related to the in vitro mutagenicity of the meals themselves. The in vitro mutagenicity of meals is markedly affected by the cooking method used to prepare them and the mutagenicity of the diet may be reflected in the mutagenicity of body fluids.     

Mutagens from the cooking of food. III. Survey by Ames/Salmonella test of mutagen formation in secondary sources of cooked dietary protein

The formation of mutagens in the major cooked protein-rich foods in the US diet was studied in the Ames Salmonella typhimurium test. The nine protein-rich foods most commonly eaten in the USA—ground beef, beef steak, eggs, pork chops, fried chicken, pot-roasted beef, ham, roast beef and bacon—were examined for their mutagenicity towards S. typhimurium TA1538 after normal ‘household’ cooking (deep frying, griddle/pan frying, baking/roasting, broiling, stewing, braising or boiling at 100–475°C). Well-done fried ground beef, beef steak, ham, pork chops and bacon showed significant mutagen formation. For chicken and beef steak high-temperature broiling produced the most mutagenicity, followed by baking/roasting and frying. Stewing, braising and deep frying produced little mutagen. Eggs and egg products produced mutagens only after cooking at high temperatures (the yolk to a greater extent than the white). Commercially cooked hamburgers showed a wide range of mutagenic activity. We conclude that mutagen formation following cooking of protein-containing foods is a complex function of food type, cooking time and cooking temperature. It seems clear that all the major protein-rich foods if cooked to a well-done state on the griddle (eggs only at temperatures above 225°C) or by broiling will contain mutagens detectable by the Ames/Salmonella assay. This survey is a step towards determining whether any human health hazard results from cooking protein-rich foods. Further testing in both short- and long-term genotoxicity bioassays and carcinogenesis assays are needed before any human risk extrapolations can be made.     

Detection of genotoxicity in fried bacon by the Salmonella/mammalian microsome mutagenicity assay

The potential for mutagen formation in fried bacon and the possible reduction or elimination of this hazard was examined in the Salmonella/mammalian microsome mutagenicity assay using Salmonella typhimurium strain TA98. Alkaline dichloromethane extracts were prepared from green pork bellies, commercial bacon (nitrite-treated and nitrite-free), and pilot-plant bacon (nitrite-free). When fried, all forms of bacon and the green belly samples gave positive mutagenic responses with the plate-incorporation technique. Unfried samples were not mutagenic. Aroclor-activated rat-liver S-9 fractions plus NADPH were essential to demonstrate a mutagenic response. When the frying temperature was held constant (171°C) maximum mutagen formation was observed in samples fried for 6 min; when samples were fried for 6 min a mutagenic response which increased with temperature, in a linear manner, was observed at temperatures above 125°C. Volatile nitrosamines were not detected in the bacon samples. The data indicate the generation of one or more mutagens in fried bacon and green pork belly, the levels of which can be reduced by decreasing heating temperature and/or time.     

Mutagenicity of deep-frying fat, and evaluation of urine mutagenicity after consumption of fried potatoes

Mutagen formation during deep-frying was evaluated using standard frying conditions. Portions of pre-fried, sliced potatoes were fried in a commercial brand of hydrogenated vegetable frying fat, which was used repeatedly and for a prolonged period of time. Concentrations of polar oxidation and degradation products, and of dimeric and polymeric triglycerides, were found to increase in the frying fat as well as in fried potatoes with prolonged use of the fat. Thiobarbituric acid-reactive substances were detectable neither in the frying fat nor in the fried potatoes. Polar fractions of repeatedly used frying fat significantly increased the number of revertants in Salmonella typhimurium strain TA97 without S-9 mix. In the presence of S-9 mix mutagenic activity was reduced. As a consequence of ongoing formation of polar degradation and oxidation products, the mutagenicity of the fat increased after repeated use. Polar fractions of lipids extracted from commercially obtained pre-fried potatoes, as well as from fried potatoes, marginally increased the number of revertants in strain TA97 without S-9 mix. The mutagenicity of the lipid fractions of fried potatoes was not related to the heating time of the fat. Methanol extracts of fat-free residues of fried potatoes significantly increased numbers of revertants in strain TA97 after metabolic activation, which indicated that a different class of mutagens had been isolated. The mutagenicity of methanol extracts was not increased after either prolonged or repeated use of the fat. Urine samples of six healthy, non-smoking volunteers, collected during the 24 hr following consumption of portions of potatoes fried in repeatedly used fat, showed no increase in mutagenicity compared with control samples. Since the exact identity of mutagens formed during deep-frying, as well as their metabolic fate in man, is unclear at present, evaluation of possible adverse biological effects associated with consumption of fried foods will require strictly controlled metabolic studies.     

Effects of temperature, patty thickness and fat content on the production of mutagens in fried ground beef

The high-pressure liquid chromatography (HPLC) profiles of mutagenic components were compared for extracts of ground beef patties fried at 200, 250 and 300°C for 6 min/side. The HPLC profiles of the mutagenic samples were similar, although total mutagenic activity in Salmonella typhimurium TA 1538 was roughly four times as high after the 300°C than after the 200°C frying. Six mutagenic peaks were analysed quantitatively at different temperatures and meat thicknesses. Two major components, 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline and 2-aminotrimethylimidazo[4,5-f]quinoxaline, and a minor component, 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), were all present at the three different temperatures. Thus, in general, cooking temperature seems to have a major effect on the quantities of mutagens produced but not on their HPLC profiles. The thickness of the meat patty did not affect the total yield of mutagens except at longer cooking times (8–10 min/side) and, in general, neither did it affect the HPLC profiles of the mutagenic components. Total mutagenic activity increased with increasing cooking times. Increasing the fat content lowered the total mutagenicity, with 150,000 revertants/kg of fresh beef at 30% fat compared with 230,000 revertants/kg at 15%, but had little effect on the mutagenicity due to IQ.     

Comparison of the cytotoxic and mutagenic potential of liquid smoke food flavourings, cigarette smoke condensate and wood smoke condensate.

Although products of pyrolysis are often cytotoxic and mutagenic, the relationship between the type of material pyrolysed and the toxicity of the resulting pyrolysis products is poorly understood. The objective of this study was to evaluate and compare the cytotoxicity and mutagenicity of several types of common pyrolysis products. The cytotoxicity and mutagenicity of these products were assessed by using neutral red uptake and Ames mutagenicity assays, respectively. The biological activities of four liquid smoke food flavourings (LSF) were compared with two other pyrolysis-derived materials; cigarette smoke condensate (CSC) and a wood smoke condensate (WSC). Results indicated all of the mixtures exhibited a concentration-dependent cytotoxic response. The CSC and WSC were less cytotoxic than three of the LSFs, but more cytotoxic than one of the brands. The CSC was mutagenic in two Salmonella strains; however, none of the LSFs or WSC was mutagenic using TA98, and only three of the LSFs were positive with TA100. The six pyrolysis-derived materials evaluated in this study showed differing patterns and magnitudes of cytotoxicity and mutagenicity. These results indicate that the cytotoxicity and mutagenicity of complex mixtures derived from pyrolysis products are affected by the type of material pyrolysed and/or the method used to prepare the mixture. The cytotoxic potential of some commercial smoke flavourings is greater than cigarette smoke condensate and several of the food flavourings are mutagenic in one Salmonella strain.     

Analysis of commercial bouillons for trace levels of mutagens

A new method, developed specifically for the extraction of heterocyclic aromatic amine (HAA) type mutagens from different food matrices, was applied to various forms of commercially available bouillons. This procedure is based on liquid-liquid extraction of the sample at different pH values. Recovery and reproducibility of the procedure was determined by processing spiked samples using a mutagenicity bioassay technique as an endpoint. The mutagenicity was tested in the Salmonella/microsome assay using strain TA98 with metabolic activation. 22 bouillon samples in liquid, cube or powder forms from seven manufacturers were extracted and tested for potential mutagenicity. The mutagenic activity of these samples varied and ranged from non-detectable to about 1200 induced revertants per gram of solid material, with a median value of approximately 250 revertants/g. The mutagenic response appeared to be dependent on the source rather than the type or form of the product tested. A negative response was obtained from only one chicken bouillon, and the highest positive response was obtained from a beef bouillon in cube form. It appears that the average beef sample, regardless of form, has a higher mutagenic potency than chicken or chicken and turkey samples. Overall, the intake of mutagens from commercial bouillons (obtained as cubes, concentrates or dry mixes) to prepare one serving (as bouillon, soup, casseroles, etc.) is considerably less than that reported in the literature for one serving of fried beef or pork. The extractability and mutagenic characteristics of these samples indicate the presence of HAA-type mutagens. Work is in progress to identify the mutagenic factors in bouillons.     

Effect on cooking time on mutagen formation in smoke, crust and pan residue from pan-broiled pork

The effect of cooking time on mutagenic activity in crust, pan residue and smoke from pan-broiled pork patties was studied in the Ames Salmonella mutagenicity test system. The effect on mutagenicity of reheating the cooked patties and of keeping them warm was also studied. The meat was broiled at 200 degrees C for various times between 2 and 10 min. Broiled meat was reheated up to 5 times at 200 degrees C, each time to a centre temperature of 70 degrees C. Reheating was also performed in a microwave oven for 2 min and in an electric oven at 200 degrees C for 10 min. In addition, broiled patties were kept warm at 60 degrees C in an incubator for up to 9 hr. The mutagenic activity increased rapidly in all fractions except the volatile phase over the first 6 min of cooking, after which time only a slight increase was seen. At cooking times below 4 min no mutagenic activity was detected in the smoke. Reheating or keeping the meat warm for up to 9 hr had very little effect on the mutagenic activity of the meat. Reversed-phase high-performance liquid chromatography mutagenicity profiles of the aerosol, crust and pan-residue extracts showed no major qualitative differences in samples cooked at different times. It is concluded that during pan broiling at 200 degrees C the major part of the mutagenic activity is formed during the first 6 min of cooking. Reheating the meat or keeping it warm does not significantly affect the mutagenic activity. No major additional mutagens are formed during continued heating for up to 25 min.     

Mutagenicity of basic fractions derived from lamb and beef cooked by common household methods

Mutagen production was examined in lamb and beef in relation to certain common household cooking methods. Mutagenicity was assessed, after extraction of the basic fraction of cooked meat samples, using Salmonella typhimurium strain TA1538 with added rat-liver S-9 homogenate. Little or no mutagenicity was found in barbecued lamb chops, in microwave-cooked lamb chops, sirloin steak, leg of lamb, or rolled beef loaf, in roasted leg of lamb or rolled beef loaf, in stewed blade steak or in boiled chuck steak. However, the basic fraction from well-done, edible fried or grilled meat contained mutagenic activity equivalent to approximately 30,000 TA1538 revertants/100 g cooked meat. It was found tht the mutagenic activity of grilled lamb chops, sirloin and rump steaks was directly related to the average surface temperatures attained during cooking. Use of butter as a frying medium was particularly associated with higher mutagenicity in meat samples. Fried meats (rump and fillet steaks) generally yielded higher mutagenic activity than did grilled meats (rump steak, lamb chops) at comparable temperatures of the cooking medium. Using similar cooking procedures, lamb did not differ markedly from beef in mutagenic activity.     

Comparative studies of the mutagenicity of urine from smokers and non-smokers on a controlled non-mutagenic diet

Measuring the mutagenicity of urine is widely viewed as a means of evaluating human exposure to potentially genotoxic materials. Diet and cigarette smoking have both been reported to affect the mutagenicity of human urine, but the relationship between smoking status and the expression of diet-related urinary mutagenicity is unknown. It has been reported that some promutagens are more active in in vitro assays when tested in the presence of urine from smokers than when tested in the presence of urine from non-smokers. We aimed to determine whether the differences in urinary mutagenicity between smokers and non-smokers result from increased urinary mutagenicity from dietary heterocyclic amine mutagens in smokers compared with non-smokers. Groups of smokers and non-smokers (6-12) were given identical diets, previously shown to be low in heterocyclic amines and very low in mutagenicity. The diet consisted exclusively of raw food and of food cooked in boiling water. After a 2-day dietary stabilization period, 24-hour urine samples were collected for three consecutive days. The regimen was repeated in the following week. For comparison, both groups were also placed on a "western" diet, consisting of a variety of foods prepared by several cooking methods, designed to reflect what a typical United States family might consume. Urine was concentrated using XAD-2 resin and then assayed for mutagenic activity in the Ames test. The urine of smokers was significantly more mutagenic than that of non-smokers when on both the raw/boiled and the "western" diets. These results indicate that the increased urinary mutagenicity observed in smokers compared with non-smokers is not due to enhanced mutagenicity of diet-related heterocyclic amine mutagens in the urine of smokers.     

Mutagenicity of pan-fried bovine tissues in relation to their content of creatine, creatinine, monosaccharides and free amino acids

The mutagenicity of pan-fried patties of five bovine tissues (meat, heart, tongue, liver and kidney) containing various concentrations of creatine, monosaccharides and free amino acids were studied. Two experiments were performed, one on single tissues fried at 150, 175 or 200 degrees C for 3 min and the other on mixtures of meat and one of the other four tissues in various proportions, fried at 200 degrees C for 3 min. For both experiments, a double-sided Teflon-coated plate was used. Frying at 150 degrees C induced mutagenicity to Salmonella typhimurium strain TA98 only in the heart sample-6000 revertants/100 gE (grams initial raw weight). Meat, heart and tongue fried at 175 or 200 degrees C showed mutagenicity values between 6000 and 19,600 revertants/100 gE. A linear relationship between mutagenicity and temperature was obtained for each of the three muscles and creatine was converted to creatinine with increasing temperature. Liver or kidney samples fried alone showed insignificant mutagenicity at all three temperatures. The creatine plus creatinine levels of raw meat, heart and tongue samples were between 19 and 33 mumol/g wet tissue. Liver and kidney both showed very low amounts of creatine plus creatinine (about 2 mumol/g wet tissue) in the raw tissue, while free amino acids were high. Glucose levels were high in liver but low in kidney samples. In meat/heart and meat/tongue mixtures the mutagenicity varied between 10,800 and 17,300 revertants/100 gE. The meat/liver and meat/kidney mixtures showed linear relationships between mutagenicity and the proportions of the mixture. The values for the slopes and intercepts of the two lines were almost equal. Among the three groups of precursors (creatine plus creatinine, monosaccharides and free amino acids) the creatine plus creatine in raw tissue seems to be the most important for producing mutagenicity. However, in crusts, the creatinine concentration was the variable with which most of the mutagenicity was associated.     

Mutagenicity and aromatic amine content of fumes from heated cooking oils produced in Taiwan.

According to toxicological studies, there are several unidentified mutagens derived from cooking oil fumes appearing in kitchens of Chinese homes where women daily prepare food. Data are limited to an analysis of aromatic amines from cooking oil fumes, which are known to be carcinogenic for bladder cancer. Fume samples from three different commercial cooking oils frequently used in Taiwan were collected and analysed for mutagenicity in the Salmonella/microsome assay. Aromatic amines were extracted from the samples and identified by HPLC and confirmed by gas chromatography/mass spectrometry (GC/MS). Extracts from three cooking oil fumes were found to be mutagenic in the presence of S-9 mix. All samples contained 2-naphthylamine (2-NA) and 4-aminobiphenyl (4-ABP). Concentrations of 2-NA and 4-ABP were 31.5 and 35.7 microg/m3 in fumes from sunflower oil, 31.9 and 26.4 mg/m3 in vegetable oil, and 48.3 and 23.3 microg/m3 in refined-lard oil, respectively. Mutagenicities of the three cooking oil condensates were significantly reduced (P<0.05) by adding the antioxidant catechin (CAT) into the oils before heating. Significant difference existed between the amounts of aromatic amines with and without adding CAT (P<0.05). These results indicate that exposure to cooking oil fumes in Taiwan might be an important but controllable risk factor in the aetiology of bladder cancer.     

Lack of mutagens in deep-fat-fried foods obtained at the retail level

The basic methylene chloride extract from 20 of 30 samples of foods fried in deep fat failed to elicit any mutagenic response that could be detected in the Salmonella typhimurium/mammalian micro-some assay. The basic extracts of the remaining ten samples (all three chicken samples studied, two of the four potato-chip samples, one of four corn-chip samples, the sample of onion rings, two of six doughnuts, and one of three samples of french-fried potato) showed evidence of weak mutagenic activity. In these samples, amounts of the basic extract equivalent to 28·5–57 g of of the original food sample were required to produce revertants at levels of 2·6–4·8 times the background level. Only two of the acidic methylene chloride extracts from the 30 samples exhibited mutagenic activity greater than 2·5 times the background reversion level, and in both cases (one corn-chip and one shrimp sample) the mutagenic response was quite weak. The basic extract of hamburgers fried in deep fat in a home-style fryer possessed higher levels of mutagenic activity (13 times the background reversion level). However, the mutagenic activity of deep-fried hamburgers is some four times lower than that of pan-fried hamburgers.     

Failure of short-term in vitro mutagenicity tests to predict the animal carcinogenicity of hair dyes

A number of hair-dye chemicals that have given positive results in various short-term mutagenicity tests have shown no clear evidence of carcinogenicity in animal bioassays. Commercial HC Blue No. 1 and its analogues HC Red No. 3 and HC Blue No. 2 are all mutagenic in the Ames test but only the HC Blue No. 1 is carcinogenic in animals. A carcinogenicity study in mice was carried out on both a commercial sample of HC Blue No. 1 and a highly purified sample which was negative in a battery of short-term tests for mutagenic activity. Both samples, administered at 0.3% in the diet for up to 24 months, were carcinogenic to mice, inducing hepatocellular carcinomas in greater than or equal to 89% of the mice examined. Therefore the presence of mutagenic impurities is not responsible for the carcinogenicity of commercial HC Blue No. 1. Discrepancies between the results of mutagenicity and carcinogenicity tests on hair dyes and other chemicals are discussed, and the value of short-term mutagenicity tests for assessing chemical safety is questioned.     

Genetic toxicology in the food industry

In vitro testing of food products for mutagenic activity presents particular problems, especially in connection with the administration of the test material to the assay system, the possible interference of food components with the genetic end-point used for the assay, the presence in foods of various factors that may modify mutagenic activity, the identification of appropriate negative or positive controls and the avoidance of artefactual mutagen formation during the preparation of test samples. Ideally mutagenicity testing requires in vivo studies (although these too have particular problems when applied to foods) but, at present, in vitro tests provide the only practical means of screening large numbers of food samples or modifying factors and of assessing food-processing techniques. The tests can be carried out on model systems (e.g. amino acid/sugar mixtures), on isolated and purified constituents of foods, on fractionated solvent extracts or on whole-food homogenates subjected to digestion procedures. However, to determine genotoxic risk from foodstuffs, quantitative data from mammalian in vivo tests and from human consumption and metabolism studies are also required.     

Enzyme-mediated mutagenicity in Salmonella typhimurium of contaminants of synthetic indigo products

The mutagenic potential of two natural and seven synthetic, commercial indigo dye products was investigated. The natural products showed no mutagenicity in Salmonella typhimurium stains TA98 and TA100. In the presence of rat-liver homogenate from Aroclor 1254 pretreated rats all of the synthetic products were mutagenic towards strain TA98 but not towards strain TA100. The mutagenic effect produced was highly dependent on the amount of rat-liver homogenate added. Because of its high mutagenic potential, one product was further investigated. In the presence of rat-liver homogenate this product was weakly mutagenic towards strain TA1537 and strongly mutagenic towards strain TA1538. No mutagenicity was observed in strain TA1535. Experiments with purified synthetic indigo and natural indigo revealed that the mutagenic activity of the synthetic commercial products can be ascribed to one or more contaminants.     

Ames mutagenicity tests of repeatedly heated edible oils

Six cooking oils with different levels of unsaturation were each heated at 180 +/- 3 degrees C for 6 hr on each of four consecutive days and cut potatoes were fried in the oils at hourly intervals. Samples of the heated and unheated oils were tested in the Ames mutagenicity assay. None of the oil samples (taken before or after 24 hr heating) showed any mutagenicity in the Ames test with or without metabolic activation.     

Mutagen excretion and cytochrome P-450-dependent activity in germfree and conventional rats fed a diet containing fried meat

To evaluate a possible role of the intestinal microflora in the metabolism of the highly mutagenic compounds formed in fried meat, conventional and germfree male AGUS rats were fed a semi-synthetic diet containing fried meat. Changes in mutagen excretion in urine and faeces over time were studied using the Ames Salmonella assay. The faecal and urinary extracts were separated by means of high-performance liquid chromatography (HPLC), and the mutagenicity of the collected fractions was determined. Cytochrome P-450 IA (IA1 and/or IA2) were detected by the use of antibodies with the Western blot technique, and the corresponding enzyme activities were measured in microsomes from the small intestine and the liver. A quantitative as well as qualitative difference in excretion of mutagens between germfree and conventional rats was observed. The total excreted of mutagenicity was significantly higher for the conventional than for the germfree rats, as a result of a higher faecal excretion of mutagens in the conventional animals. The HPLC separations of urinary and faecal extracts showed a different mutagenic metabolite pattern between the germfree and conventional rats. An increased activity of the cytochrome P-450-dependent enzyme ethoxyresorufin-O-deethylase was observed in the small intestine of conventional rats on the fried meat diet, whereas no effect of this diet was observed in the germfree rats. Similar results were obtained in immunoblotting experiments using a P-450 IA antiserum. The present study indicates that the excretion pattern and thus also the metabolism of compounds present in fried meat are affected by the germfree status.     

Mutagenicity and DNA-damaging activity caused by decomposed products of potassium sorbate reacting with ascorbic acid in the presence of Fe salt.

Although potassium sorbate (PS), ascorbic acid and ferric or ferrous salts (Fe-salts) are used widely in combination as food additives, the strong reactivity of PS and oxidative potency of ascorbic acid in the presence of Fe-salts might form toxic compounds in food during its deposit and distribution. In the present paper, the reaction mixture of PS, ascorbic acid and Fe-salts was evaluated for mutagenicity and DNA-damaging activity by means of the Ames test and rec-assay. Effective lethality was observed in the rec-assay. No mutagenicity was induced in either Salmonella typhimurium strains TA98 (with or without S-9 mix) or TA100 (with S-9 mix). In contrast, a dose-dependent mutagenic effect was obtained when applied to strain TA100 without S-9 mix. The mutagenic activity became stronger increasing with the reaction period. Furthermore, the reaction products obtained in a nitrogen atmosphere did not show any mutagenic and DNA-damaging activity. PS, ascorbic acid and Fe-salts were inactive when they were used separately. Omission of one component from the mixture of PS, ascorbic acid and Fe-salt turned the reaction system inactive. These results demonstrate that ascorbic acid and Fe-salt oxidized PS and the oxidative products caused mutagenicity and DNA-damaging activity.