Inverse regulation of claudin-2 and -7 expression by p53 and hepatocyte nuclear factor 4α in colonic MCE301 cells

Abstract

Colonic epithelial cells move up along the crypt villus axis and are differentiated into absorptive or secretory cells. Claudin-7 (CLDN7), a tight junctional protein, is mainly located at the surface of crypt, whereas CLDN2 is located at the bottom. However, the expression mechanism and function of these CLDNs are not fully understood. The expression levels of CLDN2 and CLDN7 were altered depending on the culture days in MCE301 cells derived from mouse colon. The nuclear levels of transcriptional factors p53 and hepatocyte nuclear factor 4α (HNF4α) at day 21 were higher than those at day 7. Tenovin-1 (TEN), a p53 activator, increased the nuclear levels of p53 and HNF4α. The mRNA level and promoter activity of CLDN7 were increased by TEN, whereas those of CLDN2 were decreased. The changes of CLDNs expression were inhibited by p53 and HNF4α siRNAs. The association between p53 and HNF4α was elevated by TEN. In addition, the binding of p53 and HNF4α to the promoter region of CLDN2 and CLDN7 was enhanced by TEN. Transepithelial electrical resistance was decreased by TEN, but paracellular fluxes of lucifer yellow and dextran were not. In the Ussing chamber assay, TEN increased dilution potential and the ratio of permeability of Cl^- to Na⁺. Both p53 and HNF4α were highly expressed at the surface of mouse colon crypt. We suggest that p53 and HNF4α alter the paracellular permeability of Cl^- to Na⁺ mediated by the inverse regulation of CLDN2 and CLDN7 expression in the colon.

Keywords: Colon; HNF4Α; claudin; p53.

Graphical abstract

Figure 1.

Effects of culture periods on CLDNs expression. (a) MCE301 cells were cultured for 1, 7, 14, 21, 28 days. DIC images were obtained using an EVOS M5000. The confluency was shown as percentage of covered area. Scale bar represents 500 μm. n = 8. (b) Real time PCR was performed using primer pairs of mouse villin-1, Lgr5, CLDN2, CLDN7, and β-actin. The mRNA contents were represented relative to the values in day 1. n = 4

Figure 2.

Effects of culture periods on the cellular localization of CLDNs and transcriptional factors. (a) MCE301 cells cultured for 7 and 21 days were immunostained with CLDN2, CLDN7, and ZO-1 in the presence of DAPI. The scale bar represents 20 μm. The colocalization of CLDNs and ZO-1 in the TJ was represented relative to the total fluorescence values. (b) The cells were immunostained with p53 and HNF4α in the presence of DAPI. The nuclear localization of p53 and HNF4α was represented relative to the fluorescence values in day 7. ** P < .01 significantly different from day 7

Figure 3.

Effect of TEN on the expression of CLDNs and transcriptional factors. (a) MCE301 cells cultured for 7 days were incubated in the absence (control) and presence of 10 μM TEN for 6 h. The cells were immunostained with p53 and HNF4α in the presence of DAPI. The scale bar represents 20 μm. The nuclear localization of these proteins was represented relative to the fluorescence values of control. (b) The cells were incubated with 10 μM TEN for 6 h. Nuclear extracts were immunoblotted with anti-p53, anti-HNF4α, or anti-nucleoporin p62 (p62) antibody. The contents of p53 and HNF4α normalized by p62 were represented relative to the values of control. (c) The cells were incubated with 10 μM TEN for 24 h. Cytoplasmic extracts including membrane and cytoplasmic proteins were immunoblotted with anti-CLDN2, anti-CLDN7, or anti-β-actin antibody. The contents of CLDN2 and CLDN7 normalized by β-actin were represented relative to the values of control. ** P < .01 significantly different from control

Figure 4.

Effect of TEN on reporter activity and mRNA expression of CLDNs. (a) MCE301 cells were incubated in the absence (control) and presence of 10 μM TEN for 6 h. Real time PCR was performed using primer pairs for CLDN2, CLDN7, and β-actin. The contents of CLDNs were represented relative to the values of control. n = 4. (b) Promoter luciferase construct of CLDN2 or CLDN7 was co-transfected with pRL-TK vector into the cells. After 40 h of transfection, the cells were incubated in the absence and presence of 10 μM TEN for 6 h. After normalizing transfection efficiency with Renilla luciferase, the relative promoter activities of CLDNs was represented as the values of control. (c) The cells were transfected with mock or CLDN7 expression vector. After two days of transfection, total RNA was isolated. Real time PCR was performed using primer pairs of CLDN2, CLDN7, and β-actin. The mRNA contents of CLDNs were represented relative to the values of mock. (d) The cells were transfected with negative or CLDN2 siRNA. Real time PCR was performed using primer pairs of CLDN2, CLDN7, and β-actin. After two days of transfection, total RNA was isolated. the mRNA contents of CLDNs were represented relative to the values of negative siRNA (Neg). (e) The cells were transfected with negative (Neg), p53, or HNF4α siRNA. After two days of transfection, the cells were incubated in the absence and presence of TEN for 6 h. Real time PCR was performed using primer pairs of CLDN2, CLDN7, and β-actin. The mRNA contents of CLDNs were represented relative to the values of negative siRNA. n = 4

Figure 5.

Binding of p53 and HNF4α to CLDN2 and CLDN7 promoter by TEN. (a) MCE301 cells were incubated in the absence and presence of 10 μM TEN for 6 h. The cell lysates were immunoprecipitated with anti-p53 or anti-HNF4α antibody and then blotted with anti-HNF4α or anti-p53 antibody. The amount of association of p53 with HNF4α protein was represented relative to the values of control. (b) The cell lysates were immunoprecipitated with rabbit IgG and then blotted with anti-p53 or anti-HNF4α antibody. Input was shown in the right side. (c) Nuclear proteins were prepared from the cells incubated in the absence and presence of 10 μM TEN for 6 h. After immunoprecipitation of genomic DNA by anti-p53 or anti-HNF4α antibody, real time PCR was performed using the primer pairs amplifying the putative p53 and HNF4α binding sites of CLDN2 and CLDN7 promoter. Input chromatin was used as a normalizer. ** P < .01 and * P < .05 significantly different from control. n = 4

Figure 6.

Effects of TEN on paracellular permeability. (a-c) MCE301 cells were cultured on transwell inserts. Then the cells were incubated in the absence and presence of 10 μM TEN for 24 h. (a) TER was measured using a volt ohmmeter. (b) LY (50 μg/ml), FD4 (50 μg/ml), or FD20 (50 μg/ml) were applied to the apical compartment. The buffer in the basal compartment was collected after 30 min, and fluorescence intensity was measured. Papp was calculated as described in the method. (c and d) Transwell inserts were mounted on the Ussing chamber. Dilution potential was measured by exchanging the bathing solution to the solution containing half NaCl concentration. P_Cl/P_Na was calculated by the Goldman-Hodgkin-Katz equation. (e and f) Dilution potential and P_Cl/P_Na were measured using the cells transfected with mock (negative siRNA plus empty vector) or CLDN2 siRNA (si-CLDN2) plus CLDN7-expressing vector (CLDN7OE). n = 4. ** P < .01 significantly different from control or mock. NS P > .05 not significantly different from control

Figure 7.

Effect of CLDN1 overexpression on paracellular permeability. (a) MCE301 cells were incubated with 10 μM TEN for 6 h. Real time PCR was performed using primer pairs for indicated genes. The mRNA contents of CLDNs were represented relative to the values of control. (b) The cells were incubated with 10 μM TEN for 24 h. Cytoplasmic extracts were immunoblotted with anti-CLDN1, anti-CLDN8, or anti-β-actin antibody. The protein contents of CLDN1 were represented relative to the values of control. The protein expression of CLDN8 was under detection limit (ND). (c and d) The cells cultured on transwell inserts were transfected with mock or CLDN1-expressing vector. Then, the cells were mounted on the Ussing chamber. Dilution potential and P_Cl/P_Na were measured. n = 4. ** P < .01 significantly different from control. NS P > .05

Figure 8.

Localization of CLDNs and transcriptional regulatory factors in the mice crypt. (a) The crypt of distal colon was divided into three sections; surface, middle, and bottom. After isolation of total RNA, real time PCR was performed using primer pairs of mouse CLDN2, CLDN7, and β-actin. The mRNA contents of CLDNs were represented relative to the values in the bottom. n = 4. (b) The segments of distal colon were immunostained with p53 and HNF4α in the presence of DAPI, a nuclear marker. The fluorescence intensities of p53 and HNF4α were represented relative to the values in the bottom. n = 3–4. ** P < .01 significantly different from bottom. NS P > .05

Figure 9.

A putative regulatory model of CLDN2 and CLDN7 expression in the colon. The complex of p53 and HNF4α controls the expression of CLDN2 and CLDN7 in the crypt of colon, leading to alteration of paracellular fluxes of Na⁺, K⁺, and Cl