Dietary Methyl Donor Depletion Suppresses Intestinal Adenoma Development

Abstract

The role of folate one-carbon metabolism in colorectal cancer development is controversial, with nutritional intervention studies producing conflicting results. It has been reported that Apc^Min/+ mice maintained on a diet deficient in the methyl donors folic acid, methionine, choline, and vitamin B12, and supplemented with homocysteine, show a greater than 95% reduction in intestinal tumor development. The present study extends these findings and shows that tumor protection afforded by dietary methyl donor deficiency (MDD) is long-lasting. After 11 weeks of MDD, tumor protection persisted for at least an additional 7 weeks of methyl donor repletion (22.2 ± 3.5 vs. 70.2 ± 4.6 tumors per mouse; P < 0.01). Sustained tumor protection was associated with a reduction in intestinal crypt length (26%, P < 0.01), crypt cell division and crypt fission, and an increase in apoptosis of both normal crypts and tumors (4.9- and 3.2-fold, respectively, P < 0.01). MDD also caused a significant reduction in the number of Dclk1-positive cells in the intestine (62%, P < 0.01), a long-lived crypt cell with cancer stem cell potential. Several undesirable effects associated with methyl donor restriction (e.g., reduced body weight gain) were shown to be transient and readily reversible following methyl donor repletion. Taken together, these results indicate that even temporary dietary methyl donor restriction in adenoma-prone mice can induce persistent changes to the intestinal epithelium and provide long-lasting tumor protection. These data also suggest that transient reductions in dietary methyl donor consumption should be considered when studying the impact of folate on colon cancer risk in humans. Cancer Prev Res; 9(10); 812-20. ©2016 AACR.

Figure 1

Study design and intestinal tumor burden. (A) Study design. At 4 weeks of age, a total of 102 mice were randomized and placed into 4 experimental groups. Group 1 (MDS ad lib., n=28) received MDS diet ad libitum, group 2 (MDS pair fed, n=27) received MDS diet pair fed with group 3 (MDD, n=19) which received the MDD diet; all mice in groups 1, 2, and 3 received their respective diets for 18 weeks, starting at 4 weeks of age. In addition, mice in group 4 (MDD:MDS, n=28) were placed on the MDD diet for 11 weeks, starting at 4 weeks of age, before being transferred to the MDS diet for an additional 7 weeks. All mice were sacrificed at 22 weeks of age. (B) Total number of tumors per mouse in the small intestine. Caloric restriction (MDS pair fed) causes a modest reduction (37 ± 4.2%, P<0.05) in tumor multiplicity, while MDD causes a larger reduction (78 ± 5%, P<0.01). Note, there is no significant difference in tumor number in the MDD and MDD:MDS groups, indicating that protection from tumor formation induced by 11 weeks of MDD persists through at least 7 weeks of methyl donor repletion. (C) Distribution of small intestine tumor sizes. Caloric restriction and MDD both cause a downward shift in the distribution of intestinal tumor sizes. (D) Total number of tumors per mouse in the colon. MDD causes a reduction (51 ± 14%, P<0.01) in the average number of colon tumors; however, caloric restriction does not have an effect on colon tumor multiplicity. (E) Distribution of colon tumor sizes. MDD causes a downward shift in colonic tumor size distribution; however, caloric restriction does not. Thus, even when tumors are able to form, their growth is inhibited by MDD. Error bars indicate means ± SEM. Statistically significant differences (P<0.05) between groups, measured by the Kruskal-Wallis test with Dunn’s Multiple Comparisons post-test, are indicated within dot plots by differences in the letter placed above each group.

Figure 2

Intestinal and colonic crypt length and fission rates. (A) Representative images of normal small intestine crypts with quantification. MDD caused an approximately 25% reduction (P<0.01) in the average length of normal small intestine crypts. Average crypt length is also reduced in MDD:MDS mice (16% reduction, P<0.01), suggesting long-term changes to crypt homeostasis. (B) Representative images of normal colon crypts with quantification; MDD causes an approximately 20% reduction (P<0.01) in average crypt length. (C) Representative image of a small intestine crypt undergoing crypt fission, evidenced by crypt branching. Small intestine crypt branching index (CBI) was reduced in the MDD mice (73%, P<0.01), indicating a reduced rate of crypt fission under conditions of dietary methyl donor restriction. Error bars indicate means ± SEM. Statistically significant differences (P<0.05) between groups, measured by ANOVA with Bonferroni’s post-test, are indicated within bar graphs by differences in the letter placed above each group. Scale bars, 80 and 120 μm.

Figure 3

Immunohistochemistry for CC3 and PHH3 in normal crypts of the small intestine and colon. (A) Representative images of CC3 staining in normal tissue, with quantification. Apoptotic index was determined by calculating the average number of apoptotic cells in 10 40x fields of view per mouse. MDD caused an approximately 5-fold increase (P<0.01) in apoptotic index in normal tissue. Apoptotic cells were primarily located in the villi and tended to cluster towards the villous tip. (B and C) Representative images of PHH3 staining in normal small intestine and colon crypts, with quantification. MDD induced a large reduction in the average number of PHH3⁺ cells per crypt in the small intestine and colon (53%, P<0.01 and 48%, P<0.01, respectively), suggesting an inhibition of mitosis. The average number of PHH3⁺ cells per crypt was also reduced in the MDD:MDS mice (35% reduction, P<0.01), suggesting that mitotic inhibition in normal crypts persists through at least 7 weeks of methyl donor repletion. Error bars indicate means ± SEM. Statistically significant differences (P<0.05) between groups, measured by ANOVA with Bonferroni’s post-test, are indicated within bar graphs by differences in the letter placed above each group. Scale bars, 80 and 200 μm.

Figure 4

Immunohistochemistry and quantification of Dclk1-expressing cells in the small intestine (A) and colon (B). The average number of Dclk1⁺ cells per crypt is reduced in the small intestine (62% reduction, P<0.01) and colon (41% reduction, P<0.01) in MDD mice. Importantly, Dclk1⁺ cells are also reduced in the MDD:MDS mice (32% reduction, P<0.01), suggesting that this reduction is long-lasting. Dclk1 is a putative marker for intestinal tumor stem cells, and so the reduction of Dclk1⁺ cell populations by dietary MDD may underlie its protective effects. Error bars indicate means ± SEM. Statistically significant differences (P<0.05) between groups, measured by ANOVA with Bonferroni’s post-test, are indicated within bar graphs by differences in the letter placed above each group. Scale bars, 80 μm.