Case-control study of colorectal cancer and fecal mutagenicity.

Fecal mutagenicity was measured in 68 patients with colorectal cancer and in 114 controls, using Salmonella tester strains TA98 and TA100 with and without S9 activation. Samples were also tested for fecapentaenes by high-performance liquid chromatography, to permit the separation of fecapentaene and non-fecapentaene mutagenicity. Overall, no significant case-control differences in fecal mutagenicity were observed. However, when samples containing high concentrations of fecapentaenes were excluded, non-fecapentaene TA98 mutagenicity was observed in eight cases (12%) and only four controls (4%), resulting in an estimated relative risk of 4.4 (95% confidence interval = 1.0-21.1). The association of colorectal cancer risk with non-fecapentaene TA98 mutagenicity could not be explained as an artifact of diagnostic workup or gastrointestinal bleeding among the cases. Smoking could also be excluded as a source of the TA98 mutagenicity seen, but possible dietary origins are still being explored.

Case-control study of colorectal cancer and fecapentaene excretion.

The fecapentaenes are potent mutagens found in high concentrations in the stools of some individuals. These compounds are produced in vivo by common species of the colonic microflora, from precursors of unknown origin. The fecapentaenes have been postulated to increase the risk of colorectal cancer. To test this hypothesis, we measured fecapentaene excretion in 69 patients with adenocarcinoma of the colon or rectum, newly diagnosed at three Washington, DC area hospitals. The cases were compared with 114 surgical controls, frequency matched to the cases on age, sex, and hospital. We attempted to measure fecapentaene excretion 4 times for each subject: before surgery; and at 1 mo; 3 mo; and 6 mo following surgery. Contrary to our study hypothesis, we found fecapentaene excretion during the four study periods to be similar or even lower in cases compared to controls. An indirect measurement of fecapentaene precursors also tended to be lower in cases. The case-control differences could not be explained as effects of bleeding or of the colorectal diagnostic workup, which was assessed in a separate group of 86 patients. We conclude from these data that the excretion of fecapentaenes does not increase the risk of colorectal cancer, at least when measured near the time of diagnosis.

Characterization of a mutagenic bacterial product in human feces.

Mutagens detectable with the Ames assay have been found in the feces of apparently healthy individuals and the incidence of this mutagenic activity was found to be greater in a population at high risk for colon cancer than in a population at low risk. A compound accounting for the mutagenic activity has been isolated by high performance liquid chromatography. Two closely related forms which behave identically chemically could be resolved. The compound was active on Salmonella typhimurium TA98 and TA100, had a characteristic ultraviolet absorption spectrum with maxima at about 320, 340, and 365 nm, fluoresced green in long wavelength ultraviolet light, and had the same mobility on the thin-layer chromatography as the mutagenic activity in a direct ether extract of feces. The compound was unstable in air but could be stabilized in the presence of butylated hydroxytoluene. Upon oxidation the compound lost its mutagenicity and its ultraviolet absorption spectrum underwent a blue shift so that the absorption maxima were at 295, 310, and 325 nm. Determination of the structure of the mutagen has been difficult since the compound was not volatile and production of a volatile derivative has not been successful. On thin-layer chromatography plates the compound reacted with reagents that detect chlorinated compounds. By thermal energy analysis it did not appear to contain a nitroso group. The compound increased in concentration upon anaerobic incubation of feces at 37 C and this increase was prevented by cold, air, and antimicrobial agents. This suggests to us that the fecal flora produces the compound.

On distribution of different fecapentaenes, the fecal mutagens, in the human population.

It has been shown by an HPLC analysis using a quarternary solvent mixture in an isocratic mode that human excretors of these fecal mutagens excrete both fecapentaene -12 and -14 but the ratios vary greatly between individuals. Since these mutagens are produced by the bacterial flora of the colon, this may indicate differences in the flora between these individuals or differences in the availability of different precursor molecules in their colons. Any relationship of these findings to the etiology of colonic cancer is not clear.

Carcinogenicity tests of fecapentaene-12 in mice and rats.

Fecapentaenes, a class of direct-acting bacterial mutagens, have been isolated from the feces and intestinal tract of humans on a Western meat-containing diet. Two bioassays to test pure fecapentaene-12 (FP-12) for carcinogenicity were performed. FP-12 in dimethylsulfoxide (DMSO) solution was injected i.p. into newborn ICR/MA mice on days 1, 3, 7, 10, 14 and 21. The mice killed after 21 months had neoplasms in liver, lung, glandular stomach and subcutaneous fibrosarcoma. Intrarectal (i.r.) infusion of FP-12 in an aqueous vehicle into male F344 rats for 71 weeks, and killing the rats after 21 weeks more, displayed no evidence of neoplasia associated with FP-12 exposure. The positive control, N-nitrosomethylurea (NMU), given i.r. as 4 2-mg doses in 2 weeks, as expected, yielded multiple colonic neoplasms in less than 11 months. Fecapentaene may exert its effect in bacteria and in newborn mice through the generation of hydroxy radicals. However, adult rodent and human colon may have adequate biochemical defense mechanisms against low level, even continuous exposures to chemicals like FP-12, and thus be at low risk of neoplasia, as was found.

Metabolism of 1,4-dinitro-2-methylpyrrole, a mutagen formed by a sorbic acid-nitrite reaction, by intestinal bacteria.

1,4-Dinitro-2-methylpyrrole (DNMP), a mutagenic product formed by the interaction of two common food additives, sorbic acid and sodium nitrite, was transformed to 1-nitro-2-methyl-4-aminopyrrole (NMAP) by human fecal mixtures and various intestinal bacterial strains. Under anaerobic conditions the cell suspensions of Actinomyces, Bacteroides, Clostridium, Eubacterium, Fusobacterium, and Peptostreptococcus spp. demonstrated the nitroreduction activity. Under aerobic conditions, only Actinomyces and Bacteroides spp. showed activity, and this was at a decreased level. In cell suspensions of Bacteroides thetaiotaomicron VPI 5482, NAD(P)H and glucose accelerated the reduction rate, whereas dicoumarol and heat significantly inhibited the rate, and flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN) did not affect the rate. With cell-free preparations of the same strain, reduction required NAD(P)H as a cofactor in a dose-dependent fashion and was inactivated by air and heat.

Production of iota toxin by Clostridium spiroforme: a requirement for divalent cations.

The effects of divalent cations (Ca2+, Co2+ and Zn2+) on the production of iota toxin by Clostridium spiroforme were studied. Toxin production had an absolute requirement for one or more cations in the range 1-5 mM. Using bispecific antisera, we showed that production of both the components of the toxin (ia and ib) were enhanced by divalent cations added to brain-heart infusion supplemented with peptone and glucose.

Metabolism of dietary genotoxins by the human colonic microflora; the fecapentaenes and heterocyclic amines.

The microflora of the human colon is a complex ecosystem of anaerobic bacteria which have the capability of enzymatically transforming a variety of dietary (or biliary) compounds to genotoxic metabolites. In the past, most investigators studying the interplay between diet and colonic flora and its role in the etiology of cancers focused on the reductive and glycosidic potential of the bacterial enzymes--many of which reverse the oxidative and conjugative reactions performed by the liver. Recent work in our laboratory has focused on the metabolism of two relatively new classes of genotoxins, the fecapentaenes and the heterocyclic amines (pyrolysis carcinogens). The fecapentaenes (conjugated ether lipids) are produced in the colon by Bacteroides spp. from polyunsaturated ether phospholipids (plasmalogens) whose natural origin and function are unknown. The fecapentaenes are potent direct-acting genotoxins that are detected in the feces of most individuals on normal western diets. The heterocyclic amines, which originate from fried or broiled proteinaceous foods, normally require activation by the liver before being potent mutagens or carcinogens. However, the "IQ" subclass (e.g. IQ and MeIQ) can be activated in the colon by Eubacterium and Clostridium species to a 7-hydroxy form which is directly mutagenic in Salmonella. Although there is no direct evidence that the fecapentaenes or the 7-hydroxy "IQ" compounds influence risk for colon cancer, the potency and prevalence of these bacterial metabolites is cause for concern.

Conversion of IQ, a dietary pyrolysis carcinogen to a direct-acting mutagen by normal intestinal bacteria of humans.

The dietary carcinogen, 2-amino-3-methyl-3H-imidazo[4,5-f]quinoline (IQ) is mutagenic in the Salmonella/microsomal mutagenicity assay when activated by microsomal enzymes. IQ is found in many cooked foods, notably fried beef and pork. In laboratory rodents IQ is carcinogenic. We showed that mixed and pure cultures of human intestinal anaerobes, notably Eubacterium spp., metabolized IQ to 2-amino-3,6-dihydro-3-methyl-7H-imidazo[4,5-f]quinoline-7-one (HOIQ). Unlike IQ, both the synthetic and bacterially produced HOIQ were direct-acting mutagens, i.e. active without microsomal activation. This new direct-acting mutagen, from the bacterial metabolism of a dietary pyrolysis carcinogen, raises new concerns about the possible role of this class of genotoxins in the etiology of human cancer.

Anaerobic metabolism of 2-amino-3-methyl-3H-imidazo[4,5-f]quinoline (IQ) by human fecal flora.

Incubation of the heterocyclic cooked food mutagen IQ with mixed human fecal microflora under anaerobic conditions yielded 2-amino-3,6-dihydro-3-methyl-7H-imidazo[4,5-f]quinoline-7-one (2) as the major detectable metabolite.

Fecapentaenes and their precursors throughout the bowel--results of an autopsy study.

The fecapentaenes are potent mutagens found in the stool of some humans and pigs. These compounds are produced by Bacteroides species in the gut from an uncharacterized family of precursor compounds, and have been postulated to pose a risk of human colorectal cancer. To better understand fecapentaene production in vivo, and to determine if excreted levels measured in epidemiologic studies are representative of the entire colon, fecapentaenes were assayed from multiple sites in the bowel in an autopsy study of 16 humans and 2 pigs. An indirect measurement of fecapentaene precursors was also made. Colonic concentrations of fecapentaenes and precursors varied widely between individuals, but were consistent for each individual throughout the colon. In addition, the measurements of rectal contents, assumed to approximate values in excreted stool, were equivalent to measurements from the colon.

Fecapentaene concentration and mutagenicity in 718 North American stool samples.

The fecapentaenes (FP) are the predominant fecal mutagens identified to date, but they have not been shown to be carcinogenic. Epidemiologists looking for other fecal mutagens that may be related to colorectal cancer must disentangle from their investigations the pervasive mutagenic effect of the fecapentaenes. As a first step to studying the epidemiology of fecal mutagenicity independent of fecapentaenes, we compared FP measurements and Salmonella mutagenicity assay results for 718 acetone-extracted stool samples collected from a variety of subjects in the Washington DC metropolitan areas. In this large group, 50% of mutagenic samples contained elevated fecapentaenes. Specifically, three-quarters of the samples mutagenic in TA100 contained high FP levels. In contrast, mutagenicity in TA98 was not generally explainable by fecapentaenes, suggesting that non-fecapentaene TA98 mutagenicity should be one focus of future efforts to uncover colorectal carcinogens of etiologic importance.

Mutagens in the feces of 3 South-African populations at different levels of risk for colon cancer.

The incidence of mutagens in the feces of 3 South-African populations at different risk levels for colon cancer has been determined. Lyophilized fecal samples were extracted with ether and the mutagenicity of the extracts determined using the Salmonella/mammalian microsome mutagenicity test. 19% of the samples from urban white South-Africans, a population at a high risk for colon cancer, were mutagenic using Salmonella typhimurium strain TA100. This incidence was significantly greater (p less than 0.001) than the incidence of mutagen excretion in the low-risk populations of urban blacks (2%) and rural blacks (0%). This pattern was also obtained using Salmonella typhimurium strain TA98. The incidence of mutagen excretion for urban whites was 10%, as compared to 5% and 2% for urban and rural blacks, respectively.

The precursors of fecapentaenes: purification and properties of a novel plasmalogen.

Fecapentaene-12 and fecapentaene-14 are genotoxic poly-unsaturated ether-lipids produced by the colonic microflora in humans and pigs. Although the fecapentaenes have been extensively characterized, little is known about the nature of the precursors from which they are produced. We purified one form of these precursors from feces of an individual who excreted high levels of fecapentaene-12 and its precursors. Purification was carried out by a series of extractions and precipitation in organic solvents followed by silica and amine high performance liquid chromatography. The purified precursor had identical UV spectral characteristics as the fecapentaenes indicating that it contained the same ether-linked pentaenyl functional group. However, it was not mutagenic. The precursor was amphiphilic in nature, behaving like a synthetic "model" ether-phospholipid on silica and C18 thin layer chromatography. When incorporated into phosphatidylcholine micelles it could be hydrolyzed in vitro by a combination of lipase and phospholipase C to fecapentaene-12. Our findings indicate that the general structure of this precursor is that of a phospholipid, specifically a plasmalogen--the exact structures of which remain to be determined.

Isolation of auxotrophs of Bacteroides fragilis.

Techniques employing ethane methanesulfonate and modified penicillin enrichment were developed whereby auxotrophs of Bacteroides fragilis could be readily isolated. Several auxotrophic phenotypes are described.

Metabolism of levamisole, an anti-colon cancer drug, by human intestinal bacteria.

1. Anaerobic incubation of levamisole with human intestinal flora resulted in the formation of three thiazole ring-opened metabolites, namely, levametabol-I, II and III. These new hydroxamic lactam-type metabolites were isolated and characterized by various spectroscopic methods. 2. Various pure cultures of human intestinal bacterial strains were shown, by quantitative h.p.l.c. analysis, to have ring-opening ability. Strong metabolizers include Bacteroides and Clostridium spp. Bacterial mixtures prepared from human faeces showed much greater ability to transform levamisole (74% in 48 h) than any pure strain culture. 3. Greatly decreased levamisole-transforming activity was observed with autoclaved bacterial cultures, and no activity was found with broth medium alone. This indicates that metabolism requires the presence of anaerobic bacteria and involves, at least in part, a non-enzymic process.

Purification and characterization of an enterotoxin from Bacteroides fragilis.

An enterotoxin produced by Bacteroides fragilis was purified to homogeneity and characterized as to its biological activity and basic molecular properties. Toxin preparations were prepared by growing B. fragilis VPI 13784 in brain heart infusion broth to early stationary phase, immediately precipitating the culture supernatant fluid with 70% ammonium sulfate, and stabilizing the precipitate with the protease inhibitor TPCK (tolylsulfonyl phenylalanyl chloromethyl ketone). The toxin was sequentially purified by anion-exchange chromatography on Q-Sepharose, hydrophobic interaction chromatography on phenyl-agarose, and high-resolution ion-exchange chromatography on Mono Q. The toxin appeared homogeneous as judged by polyacrylamide gel electrophoresis. The estimated molecular weight of the highly purified toxin as determined by gel filtration chromatography on Superose-12 and sodium dodecyl sulfate-polyacrylamide gel electrophoresis is 19,000. It has an isoelectric point of approximately 4.5 and is stable at pHs 5 to 10. The purified toxin is stable at -20 and 4 degrees C and upon freeze-drying, but it is unstable at temperatures above 55 degrees C. It is sensitive to proteinase K and Streptomyces protease but is resistant to trypsin and chymotrypsin. The activity of the purified toxin is neutralized by antiserum to a toxigenic strain of B. fragilis but not by antiserum to nontoxigenic strains. N-terminal amino acid analysis reveal an unambiguous sequence of Ala-Val-Pro-Ser-Glu-Pro-Lys-Thr-Val-Tyr-Val-Ile-Xxx-Leu-Arg-Glu-Asn-Gly- Ser-Thr . The highly purified toxin induced a strong fluid accumulation response in the lamb ileal-loop assay as well as a cytotoxic response (cell rounding) on HT-29 colon carcinoma cells. Thus, the purified toxin can cause both enterotoxic and cytotoxic activities.

The normal intestinal microflora: ecology, variability and stability.

The composition and functions of the intestinal floras from humans and other animals is compared and contrasted. Intrinsic and extrinsic factors, such as anatomy, age, and diet, help define the nature of the flora typical to each species of animal. Stable differences, such as the ability to reduce cholesterol or not, between the flora of individuals of the same species are described. The difficulties of characterizing a flora by cultural taxonomy are discussed and alternatives are suggested. These alternatives, collectively known as microflora associated characteristics (MACs), are measures of the biochemical activity of the bacterial flora. How MACs can be used to measure the effects of subtherapeutic levels of antibiotics on a flora is described.