Morphological and cell kinetic effects of dietary manipulation during colorectal carcinogenesis

Are dietary fiber-induced alterations in colonic epithelial cell proliferation predictive of fiber's effect on colon cancer?

Alterations in cell proliferation of the colon have been observed as a result of changes in amount and type of dietary fiber and in relation to risk of developing colon cancer. Although some human observational and intervention studies contribute to the database, most information results from experiments on rodents. Because of numerous contradictory reports linking dietary fiber, cell proliferation, and colon cancer, we undertook a critical review of existing methods in an attempt to explain the inconsistencies. Although there may be some individual types of dietary fiber that protect against chemically induced colon cancer, dietary fiber as a single entity does not appear to afford any consistent protection. Because of significant differences in experimental protocols among laboratories, it is not yet possible to state with certainty that increases in cell proliferation, induced by fiber consumption, are predictive of increased tumorigenesis. Much of what has been observed and interpreted as elevation of risk may simply be normal homeostatic changes in cell proliferation. Even though fermentation to short-chain fatty acids is a mechanistically attractive hypothesis to explain why fiber modulates cytokinetics, data do not consistently support short-chain fatty acids as biological intermediates in risk of colon cancer. The state of the art in this field has not yet progressed to the point where a clear effect of dietary fiber on cytokinetics and colon carcinogenesis can be assessed with any degree of certainty. Additional markers of apoptosis, differentiation, and cell-cell communication may be required for a more accurate analysis of the relation among fiber, cytokinetics, and colon cancer.

Insulin-like growth factor-I and II receptor expression in rat colon mucosa are affected by dietary lipid intake

Epidemiologic data and animal models have demonstrated a correlation between dietary fat composition and colon cancer risk. We have previously found that dietary fat alters cell proliferation in rat colon, which may influence the risk of colon cancer. Growth factors, including insulin-like growth factor (IGF) I and II, regulate the cell cycle in most mammalian tissues. Hence, we measured IGF-I and IGF-II receptor expression in colonocytes from Sprague-Dawley rats fed diets containing either beef tallow (BT) or corn oil (CO) at 12, 30 or 37% of energy for 4 wk. Quantitative reverse transcriptase polymerase chain reaction (RT-PCR) using an internal standard was used to examine the relative expression of both IGF-I and II receptor mRNA in three sections of the colon. The IGF-I receptor protein was also measured by Western immunoblot. In the distal colon, IGF-I receptor gene expression and protein increased significantly as the percentage of CO increased. In both proximal and middle colon, an increased percentage of BT resulted in significantly increased IGF-II receptor expression. In the proximal colon, IGF-II receptor expression decreased with increasing CO concentration, whereas in the middle colon, rats fed 37% CO had significantly higher IGF-II receptor expression than rats fed 12 or 30% CO. IGF-II receptor gene expression in proximal colon decreased with increased fat quantity, independently of fat source, whereas in the middle colon, increased fat quantity resulted in increased IGF-II receptor expression. Thus IGF-I and IGF-II receptor mRNA and IGF-I receptor protein level in colon mucosa were significantly altered by dietary fat source and quantity, thereby suggesting a potential influence of dietary fat on the endocrine regulation of colon cell mitogenesis.

Flow cytometric analysis of DNA synthetic phase fraction of the normal appearing colonic mucosa in patients with colorectal neoplasms

DNA synthetic (S) phase fractions of normal appearing colonic mucosa in Japanese and British patients with colorectal neoplasms were compared with those in patients without colonic neoplasms. Normal crypts were isolated from fresh surgical specimens of the large intestine by the use of EDTA. After fixation with 70% ethanol, isolated crypts were digested with pepsin into single nuclei suspensions. These were stained with propidium iodide and examined by flow cytometry. S phase fraction was calculated from the flow cytometry DNA histogram using Baisch's method. S phase fractions of normal appearing crypts in Japanese and British patients with colorectal tumours were not significantly different and analysed together. S phase fraction of normal appearing colonic crypts in 14 patients with familial adenomatous polyposis (FAP) was 10.23 (2.59%) (mean SD)) ranging from 5.8 to 18.8. S phase fraction of background normal mucosal in patients with large adenomas (over 2 cm) and adenocarcinomas were 9.74 (3.76%) (range, 2.7-16.1) and 8.93 (3.54%) (range, 2.9-18.9) respectively. In normal mucosa of patients without any colorectal neoplasms, S phase fraction was 8.99 (3.94)% (range, 3.9-17.7). There was no statistically significant difference in S phase fractions of normal mucosa in the four groups. Our results show that an increase in proliferative activity of background colonic crypts is not necessary for tumour development.

Dietary calcium does not reduce experimental colorectal carcinogenesis after small bowel resection despite reducing cellular proliferation

It has been proposed that colorectal carcinogenesis is accompanied by increased mucosal cell proliferation and that the converse may also apply. To examine this thesis, the crypt cell production rate (CCPR) was measured in eight groups of rats (n = 187) that had received 1,2 dimethylhydrazine, 70% small bowel resection, supplemental dietary calcium, or a combination of these. Analysis of variance showed the following: (1) the CCPR decreased between the ileum and distal colon; (2) the CCPR decreased between 16 and 32 weeks; (3) 1,2 dimethylhydrazine and small bowel resection increased the CCPR and calcium decreased the CCPR independently of one another; (4) the CCPR interacted with 1,2 dimethylhydrazine x small bowel resection, calcium x 1,2 dimethylhydrazine and interacted between the site of bowel and calcium, 1,2 dimethylhydrazine, small bowel resection, and 1,2 dimethylhydrazine x small bowel resection (p = 0.014 to p < 0.001). The tumour yield was reduced by calcium in 1,2 dimethylhydrazine treated animals (chi 2 = 14.1, df = 3, p < 0.01) but was unaffected by calcium in 1,2 dimethylhydrazine and small bowel resection treated animals despite significant differences in the CCPR. An increase of the CCPR both preceded and accompanied colorectal carcinogenesis but reduction of the CCPR was not invariably accompanied by reduced carcinogenes.

Arachidonic acid and docosahexaenoic acid are increased in human colorectal cancer

Increased arachidonic acid concentrations in experimental rodent colonic cancer have been described recently. In humans, a reduced erythrocyte stearic acid to oleic acid ratio has been reported in patients with colorectal cancer and it has been proposed that similar changes exist in the cancer tissue. The long chain fatty acids in the cancers of 15 patients with colorectal cancer were measured and compared with values in the unaffected mucosa. The values were expressed as mean (SD) mg fatty acid/g tissue and compared by analysis of variance. In the cancer tissue arachidonic acid was increased (0.703 (0.109) mg/g v 0.603 (0.127) mg/g, p less than 0.05) as was docosahexaenoic acid (0.211 (0.066) mg/g v 0.148 (0.039) mg/g, p less than 0.001). In contrast, the stearic acid to oleic acid ratio in the cancer tissue was increased rather than decreased, as previously suggested (0.36 (0.05) v 0.29 (0.7), p less than 0.01). Increased arachidonic acid and docosahexaenoic acid concentrations may be related to reduced lipid peroxidation, which is a feature of rapidly growing cells. Alternatively, the increased arachidonic acid values could be due to enhanced desaturase activity upon linoleic and linolenic acid, leading perhaps to increased formation of prostaglandins and other lipoxygenase products.

Modulation by dietary factors of BHA-induced alterations in cell kinetics of gastro-intestinal tract tissues in rats

To determine the effects of dietary ethanol or fibre on 2(3)-tert-butyl-4-hydroxyanisole (BHA)-induced alterations in cell kinetics in gastro-intestinal tract tissues, groups of six male Wistar rats were fed diets containing 0% (control) or 1.5% BHA for 2 wk. One group fed 1.5% BHA and one pair-fed control group received 10% ethanol in the drinking-water; two similarly fed groups received drinking-water only. Another group fed 1.5% BHA and a pair-fed control group received a diet supplemented with 20% cellulose; two similar groups received no fibre supplementation. Cell kinetics in the forestomach, glandular stomach and oesophageal tissue were determined, after 14 days, by bivariate 5-bromo-deoxyuridine/DNA analysis using immunocytochemistry and flow cytometry. In the fibre experiment, colorectal tissue was also examined. In both experiments the labelling indices in all the gastro-intestinal tract tissues were significantly altered in the BHA-fed groups compared with the corresponding control groups. In the ethanol experiment no statistically significant difference in the labelling indices was observed in the forestomach or glandular stomach between the two control groups or between the two BHA-fed groups. However, intake of ethanol-supplemented drinking-water induced increases in oesophageal labelling indices in rats fed a BHA-free diet. Thus 14 days of simultaneous ethanol administration has no effect on BHA-induced alterations in cell kinetics in the oesophagus, glandular stomach or forestomach of rats. In the forestomach and colorectal tissue, a high-cellulose diet resulted in a significant decrease in the BHA-induced elevation of labelling indices. Thus dietary cellulose provides a partial protection against the proliferation-enhancing effects of BHA in the rat gastro-intestinal tract.

Reduced tumour growth of the human colonic cancer cell lines COLO-320 and HT-29 in vivo by dietary n-3 lipids

Seventy-five nude mice received subcutaneous inoculation with 1 X 10(7) cells of the human colonic cancer cell lines COLO-320 or HT-29. Tumour growth was assessed over 4 weeks in animals given one of three iso-caloric diets; standard diet, high saturated fat (20% coconut) diet and high n-3 fat (20% Maxepa fish oil) diet. The n-3 diet produced significant tumour growth reduction compared to the other diets for COLO-320 at 3 to 4 weeks (P less than 0.05 at least) and similarly for HT-29 at 4 weeks (P less than 0.05). Significant incorporation of n-3 fatty acids occurred in red cell membranes, adipose tissue and both neutral lipid and phospholipid fractions of tumour lipids in animals fed Maxepa (P less than 0.01 at least). This was accompanied by reduction of linoleic acid and arachidonic acid in these tissues (P less than 0.01 at least) but was most marked in the metabolically labile phospholipid fraction. There was high mitotic activity in the tumours from all the groups but there was no difference according to diet.

Does dietary fibre stimulate intestinal epithelial cell proliferation in germ free rats?

The aim of the present experiment was to investigate the role of hind gut fermentation in the proliferative response of the intestinal epithelium to dietary fibre. We have previously shown that refeeding starved rats with an elemental diet supplemented with fermentable dietary fibre (but not inert bulk) is capable of stimulating intestinal epithelial cell proliferation throughout the gastrointestinal tract. Three groups of 10 germ free (GF) rats and three groups of 10 conventional (CV) rats, were used. All groups were starved for three days and then refed for two days with either an elemental diet (Flexical); Flexical plus 30% kaolin; or Flexical plus 30% of a fibre mixture. Cell production was determined by the accumulation of vincristine arrested metaphases in microdissected crypts. There was no significant difference between refeeding the rats with an elemental diet alone or with kaolin supplementation, however, the addition of fibre in CV rats was associated with a significant increase in intestinal crypt cell production rate in both the small intestine (p less than 0.01) and the colon (p less than 0.001). This marked proliferative effects of fibre was abolished in the GF rats. It can be concluded that it is the products of hind gut fermentation, not fibre per se that stimulate intestinal epithelial cell proliferation in the colon and small intestine.

Measurement of the proliferative status of colonic epithelium as a risk marker for colon carcinogenesis: effect of bile acid and dietary fiber

The proliferative status of mouse colonic epithelium, as affected by dietary fibers with or without cholic acid (CA), was studied by autoradiography and the metaphase arrest technique. In the first study, groups of mice were fed natural ingredient (laboratory chow) or semisynthetic diets containing 0% (control) or 0.2% (test) CA. After the mice were fed two weeks, the effect of CA was significantly more pronounced in the semisynthetic diet group than in the natural ingredient diet group with respect to labeled cells/crypt section (7.8 +/- 0.8 vs. 2.9 +/- 0.4) and mitotic figure (MF)/crypt section (3.0 +/- 0.5 vs. 1.8 +/- 0.2). In the second study, diets formulated to contain 5 or 10% cellulose (C), pectin (P), or wheat bran (WB) with or without CA (0.2%) were fed to animals for two weeks and colonic proliferative indices were measured. When compared with 5% C group, the 10% WB group exhibited lower labeling index (LI) values (4.2 +/- 0.5 vs. 6.4 +/- 1.0) and the 10% P group exhibited higher LI values (10.0 +/- 1.1 vs. 6.4 +/- 1.0). CA-induced increases in the LI and MF values responded independently in some cases to dietary fiber. Among the CA-treated groups, only the 10% P diet resulted in lower LI when compared with the 5% C group (p less than 0.05) (7.4 +/- 0.8 vs. 12.5 +/- 2.8) but had no effect on MF/crypt section. However, the 5 or 10% WB diet resulted in lower MF values (1.7 +/- 0.2 and 1.8 +/- 0.6 vs. 2.6 +/- 0.3). A long-term feeding study comparing 10% P with 10% C diets also demonstrated that the LI was elevated in the 10% P group without any effect on the mitotic activity of the colonic epithelium. This paradoxical finding suggests that the value of the LI and/or mitotic index as a risk marker of colon carcinogenesis should be further investigated.

Dietary manipulation during experimental colorectal carcinogenesis: a morphological study in the rat

The effect of the dietary manipulation of fat and fibre on the gross pathological, histopathological and scanning electron microscopic appearances of colonic mucosa was studied before and during experimental carcinogenesis in 232 male Albino Swiss rats. Carcinogen treated animals were given 12 consecutive weekly injections of Azoxymethane in a dose of 10 mg/kg per week. The animals were divided between four dietary groups (1: high fat high fibre, 2: low fat high fibre, 3: high fat low fibre and 4: low fat low fibre). The colorectal morphology was assessed at autopsy using standard histopathological techniques. In addition, scanning electron microscopic studies on selected samples were performed. A scoring system was introduced to allow a more accurate comparison of the surface architectural appearances. The high fat low fibre diet was associated with the greatest risk for macroscopic tumour production and the low fat high fibre diet with the lowest risk. Statistically significant differences between each of the dietary groups were noted with the exception of the comparison for tumour induction between the high fat high fibre and low fat high fibre dietary groups. Histopathological analysis confirmed the inter dietary relationships with respect to the extent of neoplastic change. The scanning electron microscopic study revealed that both dietary manipulation and carcinogen treatment influenced the surface characteristics. Control animals fed diet 2 (low fat high fibre) but not treated with carcinogen, showed the greatest deviation from the normal appearances. Animals treated with carcinogen and fed diet 3 (high fat low fibre) consistently showed the greatest surface abnormalities.