Reelin Protects against Colon Pathology via p53 and May Be a Biomarker for Colon Cancer Progression

Abstract

Previous observations made in human and mouse colons suggest that reelin protects the colon from pathology. In this study, we evaluated reelin expression during the transition from either colitis or precancerous lesions to colon cancer and tried to elucidate reelin regulation under these transition processes. Samples of healthy and pathological colons from humans and mice treated with either azoxymethane/dextran sulfate sodium (DSS) or azoxymethane alone were used. The relative abundances of reelin, DNMT-1 and ApoER2 mRNAs were determined by PCR in the colon samples cited above and in the tissue adjacent to mouse colon polyps and adenocarcinomas. In both, humans and mice, reelin mRNA abundance increased significantly in ulcerative colitis and slightly in polyps and decreased in adenomas and adenocarcinomas. Reelin expression was higher in the tissue adjacent to the colon adenocarcinoma and lower in the lesion itself. The reelin expression changes may result, at least in part, from those in DNMT-1 and appear to be independent of ApoER2. Lack of reelin downregulated p-Akt and p53 in healthy colon and prevented their increases in the inflamed colon, whereas it increased GSK-3β in DSS-untreated mice. In conclusion, reelin mRNA abundance depends on the severity of the colon pathology, and its upregulation in response to initial injuries might prevent the beginning of colon cancer, whereas reelin repression favors it. Increased p53 expression and activation may be involved in this protection. We also propose that changes in colon reelin abundance could be used to predict colon pathology progression.

Keywords: Akt; ApoER2; DNMT-1; colitis; colon cancer; p53; reelin.

Figure 1

Mouse model of colon cancer progression. Experimental design of the treatments to induce lesions in the colon. Dextran sulfate sodium (DSS) was administered in the drinking water (3%) for 9 days to induce acute colitis. Azoxymethane (AOM) dissolved in PBS was administered by intraperitoneal injection at the point indicated by the arrowheads. DSS was administered in the drinking water (1%) in 4-day periods. Control groups (untreated mice) received PBS intraperitoneally and water orally. Representative photographs of mice colonic mucosa along colon cancer initiation and progression. Macroscopic (fresh tissue or stained with methylene blue solution in the case of aberrant crypt foci) and microscopic images (eosin–hematoxylin-stained sections) of mice colon with colitis, aberrant crypt foci, polyps, sporadic adenocarcinomas, colitis-associated adenocarcinomas and healthy colons (control groups). The lesions are indicated by red arrows. For each mouse model, the number of animals per experimental group was ten. White scale bars represent 1 cm, the green scale bar represents 50 µm and black scale bars represent 200 µm.

Figure 2

Characterization of colon lesions in the mouse models. (A) Number of aberrant crypt foci (ACF) per mouse observed along the colon: 0–2 cm corresponds to proximal colon, 3–4 cm to proximal–distal colon region and 5–6 cm to distal colon. (B) Number of lesions per mouse in the different models and sorted by size. (C) Ratio of colon weight/colon length expressed in g/cm. (D) α-SMA and FAP mRNA abundance in colon sporadic adenocarcinomas and colitis-associated adenocarcinomas expressed as fold change relative to control (untreated) mouse colons. Data are means ± SEM (n = 8–10 animals per group). ANOVA shows the effect (p < 0.0001) of the colon region on the number of ACF (in (A)); the type of pathology on the number and size of lesions (in (B)); and the type of adenocarcinoma on the ratio colon weight/colon length (in (C)) and the α-SMA and FAP mRNA abundances (in (D)). Tukey’s test: * p < 0.001 vs. 1 centimeter (in (A)); * p < 0.001 vs. polyps; # p < 0.001 vs. 0–2 mm lesions (in (B)); ^a p < 0.05, ^aa p < 0.001 vs. control mice and * p < 0.001 vs. colitis-associated adenocarcinomas (in (C)); # p < 0.05, ## p < 0.001 vs control mice and * p < 0.001 vs. colitis-associated adenocarcinomas (in (D)).

Figure 3

Representative photographs of healthy and pathological human colons. Eosin–hematoxylin-stained, embedded paraffin sections (5–10 µm) of healthy colon, ulcerative colitis, polyps, adenomas and adenocarcinomas are shown. The number of subjects used for each condition was eight. Scale bars represent 100 µm.

Figure 4

Reelin, DNMT-1 and ApoER2 mRNA abundance in human and mouse colon cancer development. Histograms represent mRNA abundance expressed as fold changes relative to healthy colons. Data are means ± SEM (n = 8 human or animal samples per condition). ANOVA shows an effect (p < 0.0001) of colonic disease progression on mRNA reelin, DNMT-1 and ApoER2 abundances. (A) Human tissue samples: ulcerative colitis (UC), polyp, adenoma, adenocarcinoma (AC). Tukey’s test: # p < 0.05; ## p < 0.01 vs. healthy colon and * p < 0.01; ** p < 0.001 vs. ulcerative colitis. (B) Mouse tissue samples: colitis, colitis-associated adenocarcinoma (C-AAC), aberrant crypt foci (ACF), polyp, sporadic adenocarcinoma (SAC). Tukey’s test: # p < 0.05; ## p < 0.01 vs. healthy colon and * p < 0.001 vs colitis; ^a p < 0.001 vs. sporadic adenocarcinoma.

Figure 5

Reelin, DNMT-1 and ApoER2 mRNA relative abundance in the tissue adjacent to polyps and adenocarcinomas in mouse colons. (A) Representative photographs of each lesion and its adjacent tissue (inserts). In total, 5 µm embedded paraffin sections were stained with an eosin–hematoxylin procedure. Scale bars represent 200 µm. (B) Graphs represent mRNA abundance of reelin, DNMT-1 and ApoER2 in the healthy colon (control), in the tissue adjacent to each lesion and in the lesion itself (either polyp, colitis-associated adenocarcinoma or sporadic adenocarcinoma). The mRNA abundance value obtained for each gene was calculated considering that of the healthy colon value 1. Data are means ± SEM (n = 10 animals per group). ANOVA shows significant differences (p < 0.01) in reelin, DNMT-1 and ApoER2 mRNA abundance between adjacent tissues and the lesions. Tukey’s test: # p < 0.05, ## p < 0.001 vs. healthy colon; * p < 0.05, ** p < 0.001 vs. adjacent tissue.

Figure 6

p53, phospho-Ser¹⁵ p53, phospho-Ser⁴⁷³ Akt and GSK-3β expression in wild-type and reeler mouse colons. Distal colon of 3-month-old wild-type (WT) and reeler (R) mice either untreated (control) or treated for 9 days with DSS were used for Western blot (A,C) or for real-time PCR assays (B). (A) Total p53 and activated p53 (p-Ser¹⁵) protein expressions. (B) p53 mRNA relative abundance. (C) p-Akt (Ser⁴⁷³) and GSK-3β protein expressions. Histograms represent protein or mRNA relative quantification in arbitrary units as means ± SEM (n = 3–5 animals per group). The protein or mRNA value measured in WT untreated mice was set to 1. ANOVA shows the effect (p < 0.0001) of DSS treatment and reeler mutation on protein and mRNA abundance. Tukey’s test: * p < 0.05, ** p < 0.001 vs control; # p < 0.05, ## p < 0.001 vs. WT.