Claudin-1 overexpression in intestinal epithelial cells enhances susceptibility to adenamatous polyposis coli-mediated colon tumorigenesis

Abstract

Background: The tight junction protein Claudin-1, a claudin family member, has been implicated in several gastro-intestinal pathologies including inflammatory bowel disease (IBD) and colorectal cancer (CRC). In this regard, we have demonstrated that claudin-1 expression in colon cancer cells potentiates their tumorigenic ability while in vivo expression of claudin-1 in the intestinal epithelial cells (IECs) promotes Notch-activation, inhibits goblet cell differentiation and renders susceptibility to mucosal inflammation. Notably, a key role of inflammation in colon cancer progression is being appreciated. Therefore, we examined whether inflammation plays an important role in claudin-1-dependent upregulation of colon carcinogenesis.

Methods: APCmin mice were crossed with Villin-claudin-1 transgenic mice to generate APC-Cldn1 mice. H&E stained colon tissues were assessed for tumor number, size and histological grade. Additionally, microarray and qPCR analyses of colonic tumors were performed to assess molecular changes due to claudin-1 expression. APC-Cldn1 and APCmin controls were assessed for colonic permeability via rectal administration of FITC-dextran, and bacterial translocation via qPCR analysis of 16S rDNA.

Results: Claudin-1 overexpression in APCmin mice significantly increased (~4-fold) colonic tumor growth and size, and decreased survival. Furthermore, transcriptome analysis supported upregulated proliferation, and increased Wnt and Notch-signaling in APC-Cldn1 mice. APC-Cldn1 mice also demonstrated inhibition of mucosal defense genes while expression of pro-inflammatory genes was sharply upregulated, especially the IL-23/IL-17 signaling. We predict that increased Notch/Wnt-signaling underlie the early onset of adenoma formation in APC-Cldn1 mice. An increase in mucosal permeability due to the adenomas and the inherent barrier defect in these mice further facilitate bacterial translocation into the mucosa to induce inflammation, which in turn promote the tumorigenesis.

Conclusion: Taken together, these results confirm the role of claudin-1 as a promoter of colon tumorigenesis and further identify the role of the dysregulated antigen-tumor interaction and inflammation in claudin-1-dependent upregulation of colon tumorigenesis.

Figure 1

IEC-specific constitutive expression of claudin-1 in APC^Min mice increases colon tumorigenesis and decreases survival. (A) APC^Min mice were crossed with Cld-1Tg mice and colon tumors were quantitated from littermate of APC^Min and APC-Cldn1 mice (n = 18 APC mice, n = 17 APC-Cldn1 mice; p = 0.0003). (B) Representative images of the colon from APC^Min and APC-Cldn1 mice showing increased sporadic colon tumors. (C) Tumor area was measured in APC^Min and APC-Cldn1 mice (p = 0.0178). (D) Representative H&E staining. (E) Invading carcinoma in APC-Cldn1 mouse. (F) Kaplan Meir survival curve demonstrating survival of APC^Min (n = 43) and APC-Cldn1 (n = 40) mice (p = 0.0027). Values are mean ± S.E.M.

Figure 2

Claudin-1 overexpression promotes inflammation driven colon tumorigenesis. (A) Endoscopic images were obtained from APC^Min and APC-Cldn1 mice under conditions of regular drinking water (control) (n = 3) and 2% DSS w/v in drinking water (treated) (n = 5) mice at the indicated time points, following 7 days of DSS administration. (B) Tumor number/mice in DSS-treated APC^Min versus APC-Cldn1 mice (p = 0.022) (C) Representative H&E images. Values are mean ± S.E.M. *p < 0.05.

Figure 3

Increased proliferation and upregulated Wnt- and Notch-Signaling contributes to the increased tumorigenesis in APC-Cldn1 mice. (A) Tumors from APC^Min and APC-Cldn1 mice were immunostained using anti-Ki67 antibody to quantitate the proliferation, (n = 9 APC, n = 10 APC-Cld-1mice) p = 0.0125. (B) Activation of Wnt signaling was determined by immunostaining for β-catenin using anti-β-catenin antibody. The nuclear/cytoplasmic β-catenin staining is increased in APC-Cldn1 mice tumors compared to APC^Min mice. (C) qRT-PCR analysis of Wnt target genes and (D) Notch target genes in tumors of APC^Min and APC-Cldn1 mice (n = 3 mice each group). Values are mean ± S.E.M. *p < 0.05.

Figure 4

APC-Cldn1 tumors have decreased mucosal defense and increased inflammation. RNA was isolated from tumors of APC^Min and APC-Cldn1 mice and was utilized for qRT-PCR analysis of the (A) mucosal defense genes Muc2 (p = 0.0022), Klf4 (p = 0.0031), and Tff3 (p = 0.0653) and (B) inflammatory genes (n = 3 mice each group). Values are mean ± S.E.M.

Figure 5

APC-Cldn1 mice demonstrate increased colonic permeability and bacterial translocation into the mucosa. Permeability was determined to FITC-dextran via rectal administration (A) at 6 - 8 weeks (n = 6) and (B) 14–16 weeks of age in APC and APC-Cldn1 mice (n = 7). (C) qRT-PCR analysis of 16S rDNA in the distal colon of APC and APC-Cldn1 mice at 16 weeks of age (n = 3 mice in each group). Values are mean ± S.E.M. *p < 0.05.

Figure 6

IL-23 signaling is upregulated in colon tumors of APC-Cldn1 mice. RNA was isolated from tumors of APC^Min and APC-Cldn1 mice and was utilized for qRT-PCR analysis of the IL23 pathway associated genes (IL-23, IL-12p40, IL-17A and IL-6). Values are mean ± S.E.M. (n = 3 mice in each group).