Susceptibility of lean and obese Zucker rats to tumorigenesis induced by N-methyl-N-nitrosourea.

To address the possible involvement of hyperinsulinemia in breast cancer development, we have examined the susceptibility of lean and obese Zucker rats to N-methyl-N-nitrosourea (MNU)-induced mammary cancer. Fifty-day-old female lean or obese Zucker rats received intraperitoneal (i.p.) injections of 37.5 or 20 mg/kg MNU, respectively. We showed in separate experiments that these doses produce similar levels of DNA methylation in the mammary epithelial cells of the lean and obese animals. Over the course of 29 weeks following MNU treatment, half of the lean rats developed carcinomas of the mammary gland, demonstrating that they are of intermediate susceptibility to mammary tumorigenesis. During this period, the obese rats developed hyperinsulinemia and insulin resistance as expected. Although palpable tumors developed at a similar rate in the lean and obese rats, only 10% of the obese animals developed mammary carcinomas. The obese rats, however, developed a high incidence (63.3%) of epidermal cysts that occurred mainly in the region of the mammary glands. A 13.3% incidence of colon carcinomas was also found in the obese rats. These results suggest that the development of hyperinsulinemia does not render the obese Zucker rats more susceptible to mammary gland carcinogenesis. Our observation of colon carcinomas in obese, but not lean rats, however, is consistent with evidence that hyperinsulinemia promotes colon cancer in rodents and humans.

Effects of dietary folate on intestinal tumorigenesis in the apcMin mouse.

Dietary folate appears to be inversely related to colorectal cancer risk. This study investigated the effects of dietary intervention with folate or the development of intestinal polyps in Min (Apc +/-) mice. Weanling Mil mice were fed diets containing 0, 2 (basal requirement), 8, or 20 mg folate/kg diet. At 3 and 6 months of dietary intervention, 50% of the mice from each group were sacrificed, and the small intestine and colon were analyzed for polyps and aberrant crypt foci (ACF). Serum folate concentrations accurately reflected dietary folate levels (P < 0.001). At 3 months no significant difference in the average number of total small intestinal polyps was observed among the four groups. However, increasing dietary folate levels significantly reduced the number of ileal, but not duodenal or jejunal, polyps in a dose-dependent manner (P-trend = 0.001); folate supplementation at 20 mg/kg diet was associated with a 68-78% reduction in the number of ileal polyps compared with the other three diets (P < 0.007). The number of ileal polyps was inversely correlated with serum folate concentrations (P = 0.03). At 3 months, increasing dietary folate levels significantly decreased the number of colonic ACF in a dose-dependent manner (P = 0.05); the control and two folate supplemented diets significantly reduced the number of colonic ACF by 75 100% compared with the folate-deficient diet (P < 0.04). The number of colonic ACF was inversely correlated with serum folate concentration (P = 0.05). No significant difference in the number of colonic adenoma was observed among the four groups at 3 months. At 6 months, no significant differences in the average number of total small intestinal, duodenal, and jejunal polyps, colonic adenomas, and colonic ACF wer observed among the four groups. However, the folate-deficient diet had 62-76% lower number of ileal polyps compared with the control and two folate-supplemented diets (P < 0.003). Serum folate concentrations, but not dietary folate levels, were directly correlated with the number of ilea polyps (P = 0.006). These data suggest that dietary folate supplementation suppresses the development of ileal polyps and colonic ACF in this model However, at later time points, folate supplementation appears to have an opposite effect on ileal polyps. These data generally support the role of folate in intestinal tumorigenesis suggested in epidemiological studies and chemical carcinogen animal models. Notwithstanding the limitations associated with this model, these data suggest that the optimal timing and dose of folate intervention need to be determined for safe and effective folate chemoprevention.

Chemopreventive effects of dietary folate on intestinal polyps in Apc+/-Msh2-/- mice.

Epidemiological and animal studies (reviewed in Y. I. Kim, J. Nutr. Biochemistry, 10: 66-88, 1999; J. B. Mason and T. Levesque, Oncology, 10: 1727-1743, 1996) suggest that dietary folate intake is inversely related to the risk of colorectal cancer. However, the optimal timing of folate intervention and mechanisms by which folate modulates colorectal carcinogenesis have not been clearly established. A recently developed murine model of intestinal tumorigenesis, which carries a heterozygous mutation in the Apc gene and a null mutation in the Msh2 gene (Apc+/-Msh2-/-), was used to determine the effect of dietary folate on intestinal tumorigenesis. Apc+/- Msh2-/- mice were randomized to receive either 0 or 8 mg of folate/kg diet starting at either 3 or 6 weeks of age. The 3- and 6-week diet starts represent intervention before and after the establishment of neoplastic foci, respectively. At 11 weeks of age, mice were killed, and the small intestines and colons were analyzed for adenomas and aberrant crypt foci (ACF). Serum folate concentrations were determined by a standard microbiological assay. Genomic DNA methylation was assessed by in vitro [3H]methyl incorporation into hepatic DNA and by a methyl-sensitive restriction digestion method. Microsatellite instability was determined in matched normal and polyp DNA from the small intestine and colon at 5 loci. Serum folate concentrations accurately reflected dietary folate levels (P < 0.005). Folate supplementation, started before the establishment of neoplastic foci, significantly decreased the number of small intestinal adenomas (by 2.7-fold; P = 0.004) and colonic ACF (by 2.8-fold; P = 0.028) and colonic adenomas (by 2.8-fold; P = 0.1) compared with a moderate degree of folate deficiency. In contrast, a moderately folate-deficient diet, started after the establishment of neoplastic foci, significantly reduced the number of small intestinal adenomas (by 4.2-fold; P = 0.001) but had no effect on colonic ACF and adenomas compared with folate supplementation. Genomic DNA methylation and microsatellite instability do not seem to play a major role in folate-modulated intestinal and colonic tumorigenesis in this model. In conclusion, in this murine model, dietary folate supplementation significantly protects against small intestinal and colorectal tumorigenesis if it is provided before the establishment of neoplastic foci However, if it is provided after the establishment of neoplastic foci, dietary folate seems to have an opposite effect. These data suggest that the timing of folate intervention is critical in providing an effective and safe chemopreventive effect on intestinal tumorigenesis. Notwithstanding the limitations associated with this model, our data suggest that the optimal timing of folate intervention must be established before folate supplementation can be used as a safe chemopreventive agent against colorectal cancer.

Possible mechanisms relating diet and risk of colon cancer.

Two recent developments in cancer epidemiology and experimental carcinogenesis provide the basis for two possible mechanisms relating diet and colon cancer risk. The first development is the accumulating epidemiological evidence for an association between insulin resistance and colonic adenomas and cancers. This evidence suggests the following mechanism: the consumption of excess dietary energy results in the development of insulin resistance with increased circulating levels of insulin, triglycerides, and non-esterified fatty acids. These circulating factors subject colonic epithelial cells to a proliferative stimulus and also expose them to reactive oxygen intermediates. These long-term exposures result in the promotion of colon cancer. The second development is the continuing identification of agents that significantly inhibit experimental colon carcinogenesis. These observations suggest the following mechanism: focal loss of epithelial barrier function resulting from a failure of terminal differentiation results in the "leak" of a presently undefined toxin and a focal inflammatory response characterized by evidence of the activation of the COX-2 enzyme and an oxidative stress with the release of reactive oxygen intermediates. The resulting focal proliferation and mutagenesis give rise to aberrant crypt foci and adenomas. The process is inhibited by: (a) demulcents confined to the colonic lumen that "repair" the surface; (b) anti-inflammatory agents; or (c) antioxidants. The two mechanisms, i.e., insulin resistance acting throughout the body and focal epithelial barrier failure acting locally, can describe most of the known relationships between diet and colon cancer risk.

Polyethylene glycol 8000 and colon carcinogenesis: inhibition in the F344 rat, promotion in the Min mouse.

It has recently been reported that 5% polyethylene glycol 8000 (PEG 8000; Mr 8000) in the diet markedly inhibits the development of colonic tumors in carcinogen-treated rats. To assess the possible use of this agent as a preventive or treatment agent for patients with familial adenomatous polyposis, we determined the effect of PEG 8000 on spontaneous carcinogenesis in the Min mouse. PEG at a 5% concentration in the diet of Min mice did not affect the number of small intestinal or cecal tumors but did increase the number of colon tumors and the number of animals with colonic tumors (2 of 18 versus 12 of 22 animals; P < 0.001). Although the chemopreventive effect of PEG 8000 in rats is remarkable, we suggest a cautious approach in long-term testing of PEG as a chemopreventive agent for subjects at risk for colonic neoplasia.

The effect of dietary folate on Apc and p53 mutations in the dimethylhydrazine rat model of colorectal cancer.

Dietary inadequacy of folate enhances and folate supplementation suppresses colorectal carcinogenesis in the dimethylhydrazine rat model. Folate is an essential factor for DNA methylation and the de novo biosynthesis of nucleotides, aberrations of which play important roles in mutagenesis. This study investigated whether the mutational hot spots of the Apc and p53 genes for human colorectal cancer are mutated in dimethylhydrazine-induced colorectal neoplasms and whether dietary folate can modulate mutations in these regions. Rats were fed diets containing 0, 2 (basal requirement), 8 or 40 mg folate/kg diet. Five weeks after diet initiation, dimethylhydrazine was injected weekly for 15 weeks. Mutations were determined by direct sequencing in 11 low and seven high grade dysplasias and 13 invasive adenocarcinomas. A total of six Apc mutations were found in four dysplastic and carcinomatous lesions: two in two low grade dysplasias, two in one high grade dysplasia and two in one adenocarcinoma. All mutations were single base substitutions, four of which were A:T-->G:C transitions. Five of the six mutations were located upstream from the region corresponding to the human APC mutation cluster region. Dietary folate had no effect on the frequency and type of Apc mutations. No mutations were detected in exons 5-9 of the p53 gene in neoplastic lesions. These data suggest that in the dimethylhydrazine rat model of colorectal cancer, the Apc gene is mutated in early stages, albeit to a lesser degree than observed in human colorectal cancer, whereas the mutational hot spot of the p53 gene for human colorectal cancer is not commonly mutated. Although the low frequency of Apc mutations and the small number of neoplasms studied in this study might have precluded our ability to observe modulatory effects of folate, dietary folate appears to have no significant effect on Apc and p53 mutations.

Thermally oxidized dietary fat and colon carcinogenesis in rodents.

Thermally oxidized animal fat (beef tallow) was assessed for colon cancer-promoting and -initiating activity in F-344 rats and CF-1 mice with the use of the aberrant crypt focus (ACF) assay. In two promotion studies, extensively oxidized beef tallow (110 degrees C for 144-168 h, peroxide value approx 200 meq/kg, with > 80% loss of allylic and olefinic protons) had relatively little effect on the growth of ACF in F-344 rats. The multiplication constant for treatment/control of ACF size in aberrant crypts per ACF at 100 days was 1.07 (95% confidence interval = 1.01-1.14) and 0.98 (95% confidence interval = 0.91-1.06). ACF size was not affected by less extensively oxidized beef tallow or by a 10-fold reduction of dietary alpha-tocopherol during the growth of the ACF. In initiation studies, extensively oxidized beef tallow administered by gavage increased the number of animals with ACF and the number of ACF per colon (11 of 23 and 5 of 29 animals with ACF; 1.09 +/- 0.29 and 0.21 +/- 0.09 ACF/colon, respectively). Less severely oxidized beef tallow was without effect. Further studies with CF-1 mice confirmed that extensively oxidized beef tallow increased numbers of animals with ACF and average ACF per colon. The unsaturated aldehyde acrolein was without effect in the ACF assay. These data suggest that highly thermolyzed beef tallow contains an uncharacterized initiator or leads to conditions in which spontaneously initiated ACF are increased.

Insulin promotion of colon tumors in rats.

McKeown-Eyssen and Giovannucci have proposed a mechanism for colon carcinogenesis based on the similarity of the risk factors for colorectal cancer and non-insulin-dependent diabetes. They note that diets high in fat and energy and low in complex carbohydrates and a sedentary lifestyle lead to insulin resistance and hyperinsulinemia and propose that the hyperinsulinemia promotes colon carcinogenesis. In this study, we directly tested for a promoting effect of insulin on colon carcinogenesis in F344 rats. After azoxymethane initiation and injections of insulin given 5 times/week for 17 weeks, the fraction of rats with colon tumors was greater in rats receiving insulin than in rats receiving saline (79 versus 50%, respectively; P < 0.05 for tumors with maximum diameters > or = 2 mm), and the average number of tumors/ rat was also greater (2.00 versus 0.73; P < 0.001). There was no effect on body weight. Our results demonstrate that insulin in a colon tumor promoter in this rat model and support the proposed mechanism linking lifestyle factors and colon carcinogenesis.

Natural history of aberrant crypt foci. A surgical approach.

BACKGROUND: The aberrant crypt focus (ACF) appears to be an important early step in colorectal carcinogenesis. Our objectives were to determine the natural history of ACF in a surgical model. METHODS: The natural history of ACF was followed by marking the lesions in vivo with tattoos. Rats were given four weekly injections of azoxymethand (AOM; 20 mg/kg). One hundred days after the first injection of AOM, rats were anesthetized, and individual aberrant crypt focus was identified by staining with methylene blue. A 3 x 3 mm area, identifying one large (4-8 crypts) ACF was marked with a tattoo dye in each colon. Control animals received saline or AOM injections and were tattooed in areas without ACF. At 200 days, colons were examined for the presence of macroscopic lesions. RESULTS: A total of 54 tumors were found in the study group of 38 animals, and 21 of these were in the transverse and proximal descending colon. The marked areas (all in transverse and proximal descending colon) yielded 6 tumors and 2 ACF, but in 30 instances no abnormality was noted. Probability of observing a tumor in the 3 x 3 mm area of the colon that was identified as containing ACFs was 17 times greater than expected from the observed tumor rate in approximately the same zone (16 vs. 1.7 percent; 95 confidence interval, 10 to 22 and 0.5 to 1.3 percent). Twenty control animals receiving saline had no tumors of epithelial origin. Nine control animals that were carcinogen-treated and tattooed in areas without ACF had no tumors in the marked areas. CONCLUSION: Results thus show regression of many ACF identified early in the carcinogenesis process. Results also support the hypothesis that some ACF are precursor lesions for adenomas and cancers.

Colonic protein fermentation and promotion of colon carcinogenesis by thermolyzed casein.

Thermolyzed casein is known to promote the growth of aberrant crypt foci (ACF) and colon cancer when it is fed to rats that have been initiated with azoxymethane. We speculated that the promotion was a consequence of increased colonic protein fermentation (i.e., that the thermolysis of the casein decreases its digestibility, increases the amount of protein reaching the colon, and increases colonic protein fermentation and that the potentially toxic products of this fermentation promote colon carcinogenesis). We found that the thermolysis of casein reduces its digestibility and increases colonic protein fermentation, as assessed by fecal ammonium and urinary phenol, cresol, and indol-3-ol. Thermolysis of two other proteins, soy and egg white protein, also increases colonic protein fermentation with increased fecal ammonia and urinary phenols, and thermolysis of all three proteins increases the levels of ammonia and butyric, valeric, and i-valeric acids in the cecal contents. We found, however, that the increased protein fermentation observed with thermolysis is not associated with promotion of colon carcinogenesis. With casein, the kinetics of protein fermentation with increasing thermolysis time are clearly different from the kinetics of promotion of ACF growth. The formation of the fermentation products was highest when the protein was thermolyzed for one hour, whereas promotion was highest for protein that had been thermolyzed for two or more hours. With soy and egg white, thermolysis increased colonic protein fermentation but did not promote colon carcinogenesis. Thus, although thermolysis of dietary casein increases colonic protein fermentation, products of this fermentation do not appear to be responsible for the promotion of colon carcinogenesis. Indeed, the results suggest that protein fermentation products do not play an important role in colon cancer promotion.

Diet, aberrant crypt foci and colorectal cancer.

We have used the aberrant crypt focus (ACF) assay to test and develop hypotheses linking diet and colon cancer. The hypotheses were suggested by epidemiological studies that identified possible dietary factors associated with colorectal cancer risk. The ACF assay was used to quantitate the effect of the dietary factors on the initiation and growth of these putative precursors of colon cancers in experimental animals. Using this approach we have developed 3 new hypotheses for the role of diet in colorectal cancer. These are (1) a risk associated with 5-hydroxymethyl-2-furaldehyde in caramelized sugar, (2) a risk associated with some factor in thermolyzed casein, and (3) a risk associated with single nutrient boluses of sucrose and fructose. The importance of these hypotheses has still to be tested in long term carcinogenesis experiments, in analytic epidemiology studies and then, perhaps, in intervention trials.

K-ras mutations in aberrant crypt foci, adenomas and adenocarcinomas during azoxymethane-induced colon carcinogenesis.

Ras mutations are an important early event in a number of carcinogen-induced rodent tumors. Colon carcinogenesis induced in rats by azoxymethane is a useful model as it mimics the adenoma-carcinoma sequence observed in humans. In addition, aberrant crypt foci develop in the rat and these lesions appear to be potentially important precursors to adenomas in colorectal cancer. Recent studies have shown that specific K-ras codon 12 and 13 mutations are present in up to 66% of carcinogen-induced rat colon adenocarcinomas. We studied the frequency of these mutations during the aberrant crypt focus-adenoma-carcinoma sequence in azoxymethane-induced Fisher F344 rats. K-ras codon 12 GAT and codon 13 GAC mutations were detected with a sensitive assay based on the amplification of DNA using the polymerase chain reaction. No mutations were present in normal mucosa. Of 27 aberrant crypt foci, K-ras mutations were identified in 2 lesions containing 5 and 10 aberrant crypts, respectively. Mutations were present in 1 of 23 and 10 of 27 adenomas and adenocarcinomas, respectively. These data suggest that K-ras mutations play a role during the stages of carcinogenesis in azoxymethane-induced rat colon cancer. The demonstration of a genetic mutation in aberrant crypt foci provides further evidence for the significance of these lesions as precursor markers of malignant potential during colorectal tumorigenesis.

Sucrose enhancement of the early steps of colon carcinogenesis in mice.

The association of refined sugars and colorectal cancers and polyps in three recent case-control studies led us to investigate the effects of sucrose, fructose and glucose on colonic epithelial proliferation and sensitivity to carcinogenesis. CF1 and C57BL/6J mice were used; proliferation was assessed as vincristine-accumulated mitotic figures per crypt section; sensitivity to carcinogenesis was assessed as the number of aberrant crypt foci (ACF) per colon observed following the colon carcinogen, azoxymethane (AOM, 3 mg/kg and 5 mg/kg). Oral gavages of sucrose and fructose in CF1 mice (10 g/kg) increased colonic proliferation 16 h later (2.8 +/- 0.6 and 4.1 +/- 0.7 (mean +/- SEM) accumulated mitotic figures/crypt section), compared with glucose and water (1.0 +/- 0.2 and 0.4 +/- 0.1). Sucrose and fructose given 14 h prior to the AOM (5 mg/kg) increased the sensitivity of the colon to carcinogenesis (18.4 +/- 1.5 and 13.1 +/- 1.8 ACF/colon), compared with glucose and water (11.4 +/- 2.0 and 8.6 +/- 1.1). Similar results were observed with C57BL/6J mice. We conclude that dietary sucrose and fructose may represent risk factors for colorectal cancer through a direct effect of the sugars on colonic epithelial proliferation.

Aberrant crypt foci and microadenoma as markers for colon cancer.

Foci of aberrant crypts similar to those seen in experimental animals exposed to colon carcinogens have been identified and quantified on the mucosal surface of fixed resections of human colon after methylene blue staining. Many of the foci in humans showed dysplasia on histologic examination and were considered to be microadenoma (MA). These lesions may be precursors for adenomatous polyps and colorectal cancer. Rats and mice initiated with azoxymethane, then fed diets containing sucrose or casein heated at 180 degrees C to stimulate normal cooking conditions, had three to five times more large MA after 100 days than controls. Thus, cooked sugar and protein contain promoters of the growth of colonic MA. 5-Hydroxymethylfuraldehyde was identified as a promoter in cooked sugar.

Promotion of aberrant crypt foci and cancer in rat colon by thermolyzed protein.

BACKGROUND: We have previously shown that thermolyzed protein (casein) cooked with fat in the diet of the rat promotes the growth of aberrant crypt foci (putative precursors of colon cancer) assessed at 100 days. PURPOSE: To determine how thermolysis affects this promotion, we examined thermolysis conditions, quantity of thermolyzed protein in the diet, and duration of thermolysis. To determine whether the previous finding of promotion of aberrant crypt foci corresponds to promotion of cancers assessed much later, we carried out promotion studies until colon cancers appeared. METHODS: F344 rats were given an initiating dose of azoxymethane and were then randomly allocated to groups receiving diets differing in their quantity and quality of casein. The groups were examined for aberrant crypt foci and tumors in the colon. RESULTS: Aberrant crypt foci were promoted by diets containing thermolyzed casein (180 degrees C, 2 hours). Promotion increased with increasing level of thermolyzed casein in the diet (to 20%) and with increasing thermolysis time (to 4 hours). The number of animals with polyps and cancers was higher in the animals receiving thermolyzed protein (2 hours), 16/23 versus 9/26 (P less than .05) and 10/26 versus 3/27 (P less than .05), respectively. The number of aberrant crypts per focus and the number of large aberrant crypt foci were higher in the tumor-bearing animals. CONCLUSIONS: Thermolyzed casein promotes early colonic precursor lesions in a dose-dependent and thermolysis time-dependent manner; thermolyzed casein also promotes colon cancer. IMPLICATIONS: The promoter formed on thermolysis could be involved in colon cancers associated with diets cooked at elevated temperatures, such as can occur with high-fat diets.

Dose response and proliferative characteristics of aberrant crypt foci: putative preneoplastic lesions in rat colon.

Foci of aberrant crypts (ACF) have been identified in the unsectioned methylene blue stained rodent colons and hypothesized to represent precursor lesions of colon cancer. In the present study, induction and growth characteristics of ACF were investigated in response to a single injection of varying dosages of 1,2-dimethylhydrazine-2HCl (DMH), a colon carcinogen. Female Sprague-Dawley rats were given a single injection of DMH (5-150 mg/kg). Two and 19 weeks after the injection, animals were killed and their distal 10 cm of colons were enumerated for the number and crypt multiplicity of ACF. Number of ACF increased with increasing dosages of DMH plateauing at 100 mg/kg. However, percentage of ACF exhibiting different crypt multiplicity (1 to greater than 4) were similar among different dose groups. Aberrant crypts and normal crypts were enumerated for total number of cells and number and distribution of S-phase cells along the crypt height 19 weeks after DMH injection after autoradiography. The labeling index (LI) (percentage of S-phase cells) and LI along the crypt height were determined. Compared to the surrounding normal crypts, aberrant crypts exhibited significantly higher (P less than 0.05) number of cells (1122 +/- 81 versus 411 +/- 28) and higher (P less than 0.05) LI (21 +/- 1 versus 12 +/- 1). For the eight ACF analysed in the present study, the distribution of S-phase cells in the aberrant crypts were similar to that of normal crypts in that S-phase cells were restricted to the lower two-thirds of the crypts rather than distributed throughout the height of the crypts as reported for adenomatous epithelium.

Classification of aberrant crypt foci and microadenomas in human colon.

Aberrant crypt foci (ACF) can be observed and quantified on the mucosal surface of formalin-fixed human colon resections after staining with methylene blue. To determine whether these ACF could be identified in fresh tissue, 10 colon resections were collected after surgery for colorectal cancer. Unfixed and fixed flat normal colonic mucosa from each colon were scored for ACF under a dissecting microscope after methylene blue staining. The number of ACF per cm2 and the average number of crypts per foci correlated highly in unfixed and fixed mucosa (r = 0.93 and 0.78, respectively). A significantly higher frequency of lesions was found in left-sided compared to right-sided colon resections. To determine whether the topographic features of the ACF gave an indication of the histological appearance, 68 specimens containing ACF or normal mucosa were examined histologically. The presence of slit-like lumen in the crypts of ACF on the mucosal surface correlated with the presence of dysplasia at histology, thus identifying microadenomas. These two observations suggest that the topographic classification of ACF in vivo could be used to distinguish microadenomas, a putative precursor lesion of colon cancer.

Sequential analyses of the growth and morphological characteristics of aberrant crypt foci: putative preneoplastic lesions.

The main objective of the present study was to sequentially analyze growth and morphological characteristics of aberrant crypt foci (ACF) in the rat colon. Sprague-Dawley rats were given a single injection of a carcinogenic dose of 1,2-dimethylhydrazine-HCl and at varying time points ranging from 2 to 57 weeks, groups of 5 rats were terminated. The number and crypt multiplicity of ACF were determined in the distal 8 cm of the colon. In addition, ACF were processed for histology and then graded for the presence of nuclear atypia using a score of 0-4. The findings of the present study demonstrated that ACF exhibit the characteristics expected for precursor lesions. ACF were present at all time intervals in large numbers in the colons of rats treated with 1,2-dimethylhydrazine-HCl and were present when adenocarcinomas were observed. The number of ACF with 4 or more crypts and those exhibiting a higher grade (grade 4) of nuclear atypia increased significantly at or beyond 19 weeks. These features of ACF, particularly the presence of nuclear atypia indicative of dysplasia, provide strong support for the hypothesis that ACF are precursor lesions of chemically induced colon cancer.