Mice deficient in Muc4 are resistant to experimental colitis and colitis-associated colorectal cancer

Abstract

MUC4, a large transmembrane mucin normally expressed in the small and large intestine, is differentially expressed during inflammatory and malignant conditions of the colon. However, the expression pattern and the role of MUC4 in colitis and colorectal cancer (CRC) are inconclusive. Therefore, the aim of this study was to understand the role of Muc4 during inflammatory and malignant conditions of the colon. Here, we generated Muc4^−/− mice and addressed its role in colitis and colitis-associated CRC using dextran sodium sulfate (DSS) and azoxymethane (AOM)-DSS experimental models, respectively. Muc4^−/− mice were viable, fertile with no apparent defects. Muc4^−/− mice displayed increased resistance to DSS-induced colitis compared with wild-type (WT) littermates that was evaluated by survival rate, body weight loss, diarrhea and fecal blood score, and histological score. Reduced infiltration of inflammatory cells, that is, CD3⁺ lymphocytes and F4/80⁺ macrophages was observed in the inflamed mucosa along with reduction in the mRNA levels of inflammatory cytokines interleukin (IL)-1β and tumor necrosis factor (TNF)-α and anti-microbial genes Lysozyme M and SLPI in the colon of Muc4^−/− mice compared with WT littermates. Compensatory upregulation of Muc2 and Muc3 mucins under basal and DSS treatment conditions partly explains the resistance observed in Muc4^−/− mice. Accordingly, Muc4^−/− mice exhibited significantly reduced tumor burden compared with WT mice assessed in a colitis-induced tumor model using AOM/DSS. An increased percentage of Ki67⁺ nuclei was observed in the tumors from WT compared with Muc4^−/− mice suggesting Muc4 to be critical in intestinal cell proliferation during tumorigenesis. Taken together, we conclusively demonstrate for the first time the role of Muc4 in driving intestinal inflammation and inflammation-associated tumorigenesis using a novel Muc4^−/− mouse model.

INTRODUCTION

Colorectal cancer (CRC) is the third most commonly diagnosed and lethal cancer in the United States.¹ Inflammatory conditions in the gastrointestinal tract referred to as inflammatory bowel disease (IBD) that includes ulcerative colitis (UC) and Crohn’s disease (CD) pose a significant risk for the development of CRC.² Although the etiology of IBD is complex with multiple genetic and environmental factors, dysregulation in intestinal mucus barrier constitutes an important etiological factor.²

The gastrointestinal system is remarkable with respect to the organization of protective mucus barrier, where the secretory mucins form the outer loose and inner dense layer and the membrane-bound mucins form the glycocalyx that covers and protects the surface epithelial cells.³ MUC2 is the principal secretory mucin in both the small and large intestine.³ The transmembrane mucins that comprise the glycocalyx on the surface of intestinal epithelial cells include MUC1, MUC3, MUC4, MUC12, MUC13 and MUC17.³ In humans, MUC1 is expressed at low levels in both small and large intestine, MUC3 predominantly expressed in small intestine⁴ and MUC4 is found in equivalent amounts in both small and large intestine.⁵ This normal distribution of mucins in the mucus layer changes during infection and intestinal inflammatory conditions such as UC and CD.² Several studies have looked into the expression pattern of different mucins in patient samples with inconclusive findings.^2,6–9 The discrepancies observed could be due to several reasons: stage of the disease tissues collected, antibodies used, glycosylation status of the mucins and so on. To address the role of mucins in intestinal health and disease, knockout mouse models of Muc1,¹⁰ Muc2^{(ref. 11)} and Muc13^{(ref. 12)} have been used. Studies using Muc1^−/− mice showed that it forms an important mucosal barrier against Campylobacter jejuni infections, a major cause of gastroenteritis.¹³ On the contrary, Muc1^−/− mice exhibited increased mucosal thickness and decreased disease activity index in response to dextran sodium sulfate (DSS) treatment.¹⁴ Spontaneous development of CRC in Muc2^−/− mice is shown to be due to breach in the inflammatory barrier and Muc2^−/− mice are more susceptible toward DDS-induced colitis.^11,15 Similarly, a recent study using Muc13^−/− mice demonstrated that Muc13 is critical in imparting epithelial protection against the cytotoxic agents.¹²

MUC4, a large type-I transmembrane glycoprotein, is normally expressed in the intestinal, reproductive and airway epithelial cells. It is composed of an N-terminal MUC4α subunit that harbors the hallmark tandem repeat domain, nidogen and adhesion-associated MUC4 and other proteins domains, and a C-terminal MUC4β subunit with von Willebrand factor type D, three epidermal growth factor like domains, a transmembrane domain and a short cytoplasmic tail.¹⁶ Although initially identified as a tracheobronchial mucin from human tracheobronchial mucosal library, it is de novo expressed in pancreatic cancer patients imparting tumorigenic and metastatic functions to pancreatic cancer cells.¹⁷ In addition, aberrant expression of MUC4 has been reported in ovarian,¹⁸ breast,¹⁹ lung,²⁰ gall bladder²¹ and biliary tract²² malignancies. Further, MUC4 is associated with poor clinical outcome in pancreatic ducal adenocarcinoma,²³ intrahepatic chilangiocarcinoma²⁴ and extrahepatic bile duct carcinoma.²²

Although MUC4 constitutes an important component of glycocalyx in the intestine, our understanding of its role in intestinal pathologies including inflammatory and neoplastic conditions is limited. Immunohistochemical and/or mRNA expression analyses of UC and CD patient samples resulted in inconclusive findings with some studies suggesting increased or decreased expression, whereas others showing no change in MUC4 expression. However, MUC4 expression was increased in UC patients when associated with neoplastic conditions⁸ and in mice treated with DSS to induce acute and chronic colitis.²⁵ Similarly, few studies have reported the expression pattern of MUC4 during CRC progression suggesting increase, decrease and/or no correlation of MUC4 expression with different stages of CRC.^26–28 However, a recent study by Shanmugam et al.²⁹ showed that early-stage CRC patients with high MUC4 expression exhibit poorer survival compared to those with low MUC4. Besides, no studies have shown any functional role of MUC4 using colon cancer cells. Therefore, to decipher the biological significance of MUC4 in physiological and pathological conditions, we generated Muc4^−/− mice and explored its role in inflammation-induced colitis and CRC.

RESULTS

Generation and characterization of Muc4^−/− mice

MUC4, a large type-I transmembrane mucin, is normally expressed by the epithelial cells of small and large intestine. Its murine homolog is encoded by 27 exons and is located on chromosome 16. However, exact role of Muc4 in intestinal pathologies is not known. To understand the biological functions of Muc4, we generated a Muc4^−/− mouse using a targeting construct obtained from Knockout Mouse Project Repository targeting exons 2–6 (Figure 1a). This is a deletion design that incorporates a splice acceptor and a LacZ reporter by replacing the exons 2–6 of Muc4. The insertion of the LacZ reporter serves two purposes. First, it indicates loss of Muc4. Second, LacZ expression indicates the promoter activity of Muc4 in tissues and organs. Mouse embryonic stem cells (129/Sv) electroporated with the targeting construct to obtain the appropriate homologous recombination were validated using long-range PCR (Supplementary Figure 1a) and Southern blot (Supplementary Figure 1b) analyses. Two positive clones micro-injected into C57BL/6J blastocysts were implanted into surrogate females to obtain the chimeras that were backcrossed with C57BL/6J for germ line transmission. Further intercrossing between the F1 progenies was performed to obtain heterozygotes and homozygous Muc4^−/− animals. All genotypes were determined by PCR using tail DNA (Figure 1b). Reverse transcriptase–PCR analysis confirmed loss of Muc4 expression in organs that normally express Muc4, that is, colon and lungs (Figure 1c). Pancreas was used as a Muc4 non-expressing organ (Figure 1c) control. With intact Muc4 promoter, a successful deletion of Muc4 would result in LacZ expression in tissues that normally express Muc4. LacZ staining of different Muc4^−/− tissues further confirmed the deletion of Muc4 in colon and testes (Figure 1d). Pancreas was used as a non-expressing organ control (Figure 1d). Muc4^−/− mice are viable, fertile and developed normally with normal sized litters when intercrossed between the heterozygotes. No obvious phenotypic abnormalities were observed in Muc4^−/− mice up to 1 year of age compared with wild-type (WT) control littermates. To assess whether deletion of Muc4 influences the carbohydrate content in the colon, we performed periodic acid–Schiff and Alcian blue staining that recognize the neutral and acidic carbohydrates, respectively. However, no significant differences in the glycan contents of the colon were observed upon deletion of Muc4 (Supplementary Figure 2).

Figure 1.

Targeted disruption of mouse Muc4 gene. (a) Schematic representation of the targeting construct used for generation of Muc4^−/− mice. The targeting construct containing IRES-LacZ-hbactP-Neo cassette was used to replace exons 2–6 of Muc4 (represented as the black boxes) through homologous recombination. The blue ( ) and red ( ) half arrows indicate the position of the primers used for long-range PCR analysis for homologous recombination at 5′ and 3′ regions respectively. (b) PCR analysis using the genomic DNA extracted from the tail biopsies for genotyping. Each mouse was genotyped for WT and KO primer pairs for both 5′ and 3′ regions. (c) Reverse transcriptase (RT)–PCR analysis using primers for the 3′ region (exon-22 forward-primer: exon-27 reverse-primer and exon-26 forward-primer: exon-27 reverse-primer) of Muc4 confirmed the deletion of Muc4. Normal colon and lungs express Muc4, whereas normal pancreas does not. (d) LacZ staining of pancreas, testes and lungs from WT and Muc4^−/− mice. LacZ expression indicates successful deletion of Muc4. As normal pancreas does not express Muc4, LacZ expression is not seen in Muc4^−/−pancreas. Scale bar, 500 μm.

Muc4^−/− mice display increased resistance to DSS-induced colitis Administration of DSS in the drinking water is one of the most widely used and reliable experimental model to induce colonic inflammation.³⁰ To assess the role of Muc4 in acute colitis, 6- to 8- week-old Muc4^−/− and WT mice were challenged with 3% DSS in drinking water for 3 weeks and lethality rate was determined. In case of both males (Figure 2a) and females (Figure 2b), Muc4^−/−mice displayed better survival compared with the WT. However, male mice of either group were more susceptible and succumbed earlier compared with females. Therefore, the survival rate was analyzed separately for males and females (Figures 2a and b).

Figure 2.

Muc4^−/− mice display enhanced resistance to DSS-induced colitis compared with WT mice. (a, b) Kaplan–Meier survival analysis of age-matched WT (5 males and 10 females) and Muc4^−/− (10 males and 5 females) mice following treatment of 3% DSS in drinking water for 21 days; males (a) and females (b). As female mice of both WT and Muc4^−/− categories displayed less susceptibility to colitis-induced mortality compared with male mice of either group, we performed separate analysis for males and females using log-rank test (P-values shown). (c–e) Percentage change in body weight compared with the baseline weight (c), diarrhea score (d) and fecal blood score (e) in age-matched WT (n = 7) and Muc4^−/− (n = 9) male mice after 2% DSS treatment for 7 days followed by 7 days of regular water. Student’s t-test; error bars indicate mean ±s.e.m., P-values shown. (f) WT male mice (n = 7) had shortened and edematous colon compared with Muc4^−/− male mice (n = 7) after 2% DSS treatment for 7 days. Student’s t-test; error bars indicate mean ±s.e.m., P-values shown.

Next, to assess the role of Muc4 in intestinal restitution following tissue injury, 6- to 8-week-old WT and Muc4^−/− mice were given 2% DSS in drinking water for 7 days followed by 7 days of regular water and were assessed for their body weight, stool consistency and fecal blood every day. We again observed that female mice of either group are more resistant to disease severity compared with males and were analyzed separately. With respect to body weight, increased loss was observed in WT compared with Muc4^−/− males, however, it was not statistically significant (Figure 2c). In case of females, the body weight loss was not very different (Supplementary Figure 3a) between WT and Muc4^−/− mice. WT mice exhibited severe diarrhea (Figure 2d) and fecal bleeding (Figure 2e) compared with Muc4^−/− mice demonstrating that loss of Muc4 imparts increased resistance to DSS-induced colitis. However, no significant difference in diarrhea score was observed between female mice of WT and Muc4^−/−groups (Supplementary Figure 3b) and no rectal bleeding were observed until the end of the experiment. Both male and female WT mice had shortened and edematous colon compared with Muc4^−/− mice (Figure 2f and Supplementary Figure 3c).

Histologically, WT mice displayed extensive erosion, ulceration, severe crypt damage and massive infiltration of inflammatory cells in to the colonic mucosa compared with Muc4^−/− mice (treated with 2% DSS in drinking water for 7 days followed by 7 days of regular water) (Figure 3a). The severity and the extent of tissue damage were more pronounced in the distal compared with the proximal colon. As mentioned earlier, the histological score was observed to be statistically significant in male (Figures 3a and b), but not in female mice (Supplementary Figure 3d). Interestingly, an instance of reepithelialization, a process that is prerequisite for restoring normal intestinal mucosal following injury was observed in two Muc4^−/− mice (Supplementary Figure 4) but not in WT mice (treated with 2% DSS in drinking water for 7 days followed by 7 days of regular water) indicating a rapid healing response in absence of Muc4. Although we observed a discrepancy in the susceptibility of male and female mice to DSS-induced colitis, our findings in the male animals conclusively demonstrate that the Muc4^−/− mice displayed increased resistance to DSS-induced colitis compared with their WT counterparts.

Figure 3.

WT mice exhibit severe histopathological alterations compared with Muc4^−/− mice following DSS-induced colitis. (a) Representative hematoxylin and eosin-stained colonic sections from WT and Muc4^−/− mice with or without 2% DSS treatment. Although no histological difference was observed in mice without DSS treatment, WT mice displayed extensive epithelial ulceration, erosions, severe crypt damage and massive infiltration of inflammatory cells in to the colonic mucosa compared with the Muc4^−/− mice. This effect was more pronounced in the distal colon compared with the proximal colon. Scale bar, 500 μm. (b) Quantification of histolopathological abnormalities in male WT (n = 7) and Muc4^−/− (n = 9) mice following 2% DSS-induced colitis based on extent of tissue damage and inflammatory cell infiltration. Student’s t-test; error bars indicate mean ±s.e.m., P-values shown.

Analysis of expression of intestinal mucin genes in WT and Muc4^−/− mice in response to DSS treatment

To investigate whether deletion of Muc4 results in the compensatory upregulation of any of the colonic mucins and/or expressional variation of colonic mucins in response to DSS treatment (treated with 2% DSS in drinking water for 7 days followed by 7 days of regular water), we analyzed the mRNA expression of Muc1, Muc2, Muc3 and Muc13 in WT and Muc4^−/− mice with and without DSS treatment. We observed an increase in Muc4 expression in WT mice treated with DSS compared with untreated, however, it was not statistically significant (Figure 4a). Further, under no DSS treatment condition, loss of Muc4 led to significant upregulation of the secretory mucin Muc2 without any effect on Muc1, Muc3 and Muc13 (Figure 4b). Following DSS treatment, Muc3 expression was significantly upregulated and the level of Muc2 expression was still higher though was not statistically significant (Figure 4c). This demonstrates that Muc4 influences the expression of other intestinal mucins such as Muc2 and Muc3 under basal as well as inflammatory conditions. To understand whether loss of Muc4 (MUC4) from the epithelial cells alone elicits compensatory upregulation of mucins such as MUC2 and MUC3, we down-regulated MUC4 in HCT-8 colon cancer cells using two short hairpin RNA oligos (Supplementary Figure 5a). Downregulation of MUC4 in the epithelial cells led to compensatory upregulation of MUC2 (Supplementary Figure 5b) suggesting this may be an epithelial cell intrinsic phenomenon (MUC3 was not expressed in HCT-8 cells).

Figure 4.

Expression of colonic mucin genes in WT and Muc4^−/− male mice in response to DSS-induced tissue damage. (a) Expression analysis of Muc4 was determined using mRNA extracted from the frozen colon tissues from WT and Muc4^−/−male mice with (n = 4) and without (n = 6) 2% DSS treatment. (b, c) Expression analysis of other colonic mucins Muc1, Muc2, Muc3 and Muc13 was determined using mRNA extracted from the frozen colon tissues from WT and Muc4^−/− male mice without (n = 6) (b) and with (n = 4) (c) 2% DSS treatment. Real-time PCR was performed to quantitate the mRNA expression. Expressed relative to Gapdh and the fold change in expression was calculated with respect to WT mice. The statistical analysis was done with data transformed to the log2 scale using t-tests, and the means and 95% confidence intervals (CIs) were back transformed to the original scale, P-values shown.

Reduced infiltration of CD3⁺ T-lymphocytes and F4/80⁺ histiocytes along with reduced expression of inflammatory cytokines in Muc4^−/− mice following DSS treatment

We assessed the infiltration of inflammatory cells associated with DSS-induced colitis of WT and Muc4^−/− mice (treated with 2% DSS in drinking water for 7 days followed by 7 days of regular water) by immunohistochemical analysis of colonic tissues using anti-CD3 antibody for T-lymphocytes and anti-F4/80 antibody for histiocytes. Compared with WT mice, Muc4^−/− mice had significantly reduced histiocytes and T-lymphocytes in the inflamed colonic mucosa (Figures 5a and b). Consequently, mRNA expression analysis of inflammatory cytokines such as tumor necrosis factor (TNF)-α showed a significant reduction and a trend in the downregulation of interleukin (IL)-1β in the colonic tissues from Muc4^−/−compared with WT mice (Figure 5c). In addition to the inflammatory cytokines, there were significantly lower mRNA levels of anti-microbial enzymes lysozyme M and SLPI in the colonic mucosa of Muc4^−/− mice treated with DSS compared with WT mice (Figure 5d).

Figure 5.

Reduced infiltration of histiocytes and T-lymphocytes and expression of inflammatory cytokines in the colon of Muc4^−/− male mice compared with WT male mice following DSS-induced colitis. (a) Representative colon sections from WT and Muc4^−/− male mice stained with F4/80⁺ histiocytes and CD3⁺ T-lymphocytes after 2% DSS treatment for 7 days followed by 7 days of recovery. Scale bar, 500 μm. (b) Quantitative representation of average number of F4/80⁺ histiocytes and CD3⁺ T-lymphocytes per high-powered field (HPF) with 10 HPFs per mouse from male WT (n = 6) and male Muc4^−/− mice (n = 6). Wilcoxon rank sum test was used to compare the average number of CD3⁺ and F4/80⁺ cells between WT and Muc4^−/− mice groups, P-values shown. (c, d) Expression analysis of inflammatory cytokines (c) and antimicrobial factors (d) was determined using mRNA extracted from the frozen colon tissues from WT and Muc4^−/− male mice with 2% DSS treatment. Real-time PCR was performed to quantitate the mRNA expression (n = 4). Expressed relative to Gapdh and the fold change in expression was calculated with respect to WT mice. The statistical analysis was done with data transformed to the log2 scale using t-tests, and the means and 95% confidence intervals (CIs) were back transformed to the original scale, P-values shown.

Muc4^−/− mice displayed increased resistance to colitis-associated colorectal tumors in AOM/DSS model

Chronic inflammatory conditions of the colon such as UC and CD are major risk factors for CRC.³¹ Having established that Muc4 exacerbates the inflammatory condition of the colon in response to DSS, we investigated the role of Muc4 in colitis-associated CRC using the well-established azoxymethane (AOM)/DSS model. Twenty-four weeks following administration of AOM/DSS (Figure 6a), WT and Muc4^−/− mice were analyzed for tumors in the colorectal region. Primarily, the malignant tumors were observed in the distal colon and rectum with occasional small adenomas in the proximal colon. Approximately fourfold higher number of tumor nodules were observed in WT mice compared with Muc4^−/− mice (Figures 6b and c) and the size of the tumor nodules were larger in WT compared with Muc4^−/− mice (Figures 6b and c). Histologically, most of the lesions were adenocarcinomas with different degrees of differentiation in both WT and Muc4^−/− mice. In the Muc4^−/− group, there was a prevalence of carcinoma in situ and low-grade adenocarcinomas with invasion into lamina propria (Figure 7a). Occasional invasion to submucosa was also observed. On the other hand, in WT group there was a prevalence of high-grade adenocarcinomas with frequent invasion into submucosa and occasional invasion into muscularis propria (Figure 7a).

Figure 6.

Colorectal tumorigenesis was delayed and less severe in Muc4^−/− mice after AOM/DSS treatment. (a) Schematic for AOM/DSS treatment protocol for induction of colorectal tumor in WT and Muc4^−/− mice (both males and females). (b) Representative longitudinal luminal views of colons from WT and Muc4^−/− mice at the end of the AOM/DSS treatment protocol. WT mice display increased number and larger tumor nodules compared with the Muc4^−/− mice. We used Alcian blue (the blue colored dye) as a smear to demarcate the boundaries of the tumor nodules that was very helpful in counting and measuring the numbers and size of the tumor nodules respectively. (c) Tumor nodules were compared between WT (n = 9) and Muc4^−/− mice (n = 9), for <2 mm and >2 mm separately using Wilcoxon rank sum test, P-values shown. Both male and female mice were included in this study.

Figure 7.

Muc4-mediated cellular proliferation contributes to enhanced colitis-associated tumorigenesis. (a) Representative hematoxylin and eosin-stained colonic sections from the WT and Muc4^−/− mice after AOM/DSS treatment. Tumors in the WT mice were found with frequent invasion into submucosa and sometimes to muscularis propria (yellow arrow), whereas tumors in the Muc4^−/− mice were of carcinoma in situ (CIS, red arrow) type with invasion into lamina propria. Scale bars, 500 μm. (b) Representative colon tumor sections from WT and Muc4^−/− mice with Ki67 staining. Scale bars, 500 μm. (c) Percentage of Ki67⁺ nuclei was determined as a ratio of Ki67⁺ nuclei to total nuclei per high-powered field (HPF) with 10 HPFs per mouse from the WT (n = 6) and Muc4^−/− mice (n = 7) group. Wilcoxon rank sum test was used to calculate the statistical significance, P-value shown. Both male and female mice were included in this study. (d) Representative immunohistochemical staining of colon tumor sections from WT and Muc4^−/− mice with β-catenin staining. Nuclear β-catenin staining is seen in the neoplastic tissues (red arrow) with the cytoplasmic β-catenin staining in the adjacent non-neoplastic crypts (red arrow heads). Scale bars, 500 μm.

To investigate the molecular mechanism by which loss of Muc4 imparts resistance to colorectal tumorigenesis, we performed immunohistochemical analysis of cellular proliferation (Ki67) and apoptosis (cleaved caspase-3) of the colonic tumors from WT and Muc4^−/− mice. We observed a significant reduction in the percentage of Ki67-positive cells in the tumors from Muc4^−/− mice group compared with WT mice (Figures 7b and c) suggesting presence of Muc4 influences the colonic epithelial cell proliferation that was abrogated in Muc4^−/− mice. There was no difference in the cleaved caspase-3 staining between the WT and Muc4^−/− mice tumors (data not shown). Altered β-catenin signaling is frequently observed and is the major molecular pathway in human CRC development and progression.^32,33 Immunohisto-chemical analysis of tumors from the WT and Muc4^−/− mice displayed a strong nuclear β-catenin signal. In both groups cytoplasmic β-catenin signal was present in the adjacent benign mucosa (Figure 7d).

DISCUSSION

IBD that includes UC and CD is a major health problem in the western world.² Although the etiology and the mechanism(s) of IBD pathogenesis is multi-factorial and incompletely understood, a dysfunctional mucus barrier on the surface of the luminal epithelial cells is considered to be a major factor.² In addition, patients with IBD are at a higher risk of developing CRC.³¹ MUC4 is an important component of the glycocalyx on the surface of colonic epithelial cells and is believed to be critical for mucosal protection in addition to other secretory (MUC2) and cell surface mucins (MUC1, MUC3 and MUC13).³ However, the expression pattern and the role of MUC4 during inflammatory and neoplastic conditions of colon are not well understood. Here, we generated mice that is genetically deficient in Muc4 and addressed the in vivo functional relevance of Muc4 during DSS-induced colitis and AOM/DSS-induced CRC. Mice lacking Muc4 were more resistant to colitis and colitis-associated symptoms as well as colitis-associated CRC induced by AOM/DSS. These data demonstrate that Muc4 exacerbates the effects of colitis and colitis-associated CRC.

Upon deletion of Muc4, we did not observe any histological abnormality in the colon of Muc4^−/− mice compared with WT under steady-state conditions. In addition, the compositions of the neutral and acidic carbohydrates were not significantly altered. However, the expression of Muc2 was found to be significantly upregulated in Muc4^−/− mice compared to WT suggestive of compensatory upregulation. Following DSS treatment, a significant upregulation of Muc3 (orthologue of human MUC17) was observed with Muc2 expression still on the higher side (although not statistically significant). As Muc2^−/− mice are prone to spontaneous and DSS-induced colitis,¹⁵ upregulation of Muc2 in Muc4^−/− mice could be one of the reasons for increased resistance of Muc4^−/− mice to DSS-induced colitis. Previous studies using recombinant protein fragments of Muc3/MUC17 and MUC17 knockdown have demonstrated that Muc3/MUC17 has a critical role in maintaining intestinal epithelial barrier to enteroinvasive Escherichia coli infection and accelerated healing in response to chemical-induced colitis.^34–36 Therefore, upregulation of Muc3, an orthologue of human MUC17 that helps in intestinal restitution and healing of experimental colitis, in Muc4^−/− mice treated with DSS also partly explains the resistance displayed by Muc4^−/− mice to DSS-induced colitis. Accordingly, we observed that the process of reepithelialization that is required for the healing following mucosal injury in Muc4^−/− mice, but not in WT mice. However, this needs to be validated in a larger cohort to attribute any role of Muc4 in preventing the process of reepithelialization.

Although inflammatory environment is known to influence mucin expression, our study demonstrates that Muc4 also influences the expression of other mucin genes under basal and inflammatory conditions. Particularly, previous studies have demonstrated that proinflammatory cytokines such as TNF-α, IL-1β, IL-4 positively regulate MUC2 expression in human colon cancer and airway epithelial cells.^37–40 Promoter analysis of mouse Muc3 gene revealed binding sites for NF-κB, SP1, CREB and was responsive for the treatment of IL-4, IL-6, TNF-α and so on.⁴¹ As DSS-treated Muc4^−/− mice show increased expression of proinflammatory cytokines such as TNF-α and IL-1β, these inflammatory mediators may be one of the reasons for upregulation of Muc2 and Muc3 in Muc4^−/− mice. But, how Muc4 influences the inflammatory environment with and without DSS treatment further needs to be investigated. Further, downregulation of MUC4 in HCT-8 colon cancer cells (epithelial origin) was sufficient to induce compensatory upregulation of MUC2, suggesting this compensation may be an epithelial cell intrinsic phenomenon. Therefore, a more holistic approach, including single-cell tran-scriptomics, cytokine array and so on, would be required to further elucidate the exact mechanism of compensatory upregulation of other mucins following deletion of Muc4.

DSS-induced colitis that resembles the phenotypes of UC is one of the most widely used experimental colitis model. It is believed to destabilize the intestinal mucus layer making it more permeable to intestinal luminal content including luminal bacteria.³⁰ Using the DSS-induced colitis model, we observed a significant resistance toward the severity of colitis as assessed by body weight loss, diarrhea and fecal blood score, extent of tissue damage, and inflammatory cell (CD3⁺ T-lymphocytes and F4/80⁺ histiocytes) infiltration in Muc4^−/− mice compared with WT mice. The inflammatory cytokines such as TNF-α and IL-1β have a critical role in the pathogenesis of IBD⁴² and upregulation of these cytokines in WT mice suggests the involvement of Muc4 in IBD pathogenesis. Although the inflammatory cell infiltration and cytokine production could be due to the damage inflicted by DSS, a direct role of Muc4 in immune cell functions cannot be ruled out owing to the resistance displayed by the Muc4^−/− mice in response to DSS treatment compared with WT mice.

Although Muc4 and Muc13 belong to the transmembrane mucins category, deletion of either of them resulted in opposite phenotypes in response to DSS treatment. Although Muc13^−/−mice resulted in aggressive phenotype (increased loss of body weight, diarrhea score, fecal blood score and severe histologic damage) in response to DSS-induced colitis,¹² Muc4^−/− mice resulted in protective phenotype (decreased loss of body weight, diarrhea score, fecal blood score and less severe histologic damage). Loss of Muc13 resulted in increased susceptibility of colonic epithelial cells to apoptosis suggesting Muc13 provides protection to colonic epithelial cells from apoptotic stimuli, thereby protecting it from damage-induced cell death. No compensatory upregulation of other mucins such as Muc1, Muc3 and Muc4 was observed in Muc13^−/− mice, whereas compensatory upregulation of Muc2 and Muc3 was observed in Muc4^−/− mice, which we propose as one of the mechanisms by which lack of Muc4 provides protection to colonic epithelial cells. Muc1, another transmembrane mucin, when deleted resulted in protection of colonic epithelial cells¹⁴ like that of Muc4^−/− mice when challenged with DSS.

As patients with IBD are at greater risk for developing CRC,³¹ we addressed the role of Muc4 in WT and Muc4^−/− mice using a well-established AOM/DSS model. As in DSS-induced colitis model, Muc4^−/− mice were found to be more resistant to AOM/DSS-induced CRC compared with WT mice. Although several studies have asserted the oncogenic role of MUC4 in multiple malignancies including pancreatic, gastric, breast, ovarian cancer and so on,^17,43–49 the pattern of expression and functional role of MUC4 in colon cancer is not well understood. Few studies investigated the expression pattern of MUC4 during CRC progression suggesting increase, decrease and/or no correlation of MUC4 expression with different stages of CRC.^26–28 A recent study by Shanmugam et al.²⁹ showed that early-stage CRC patients who have high MUC4 expression exhibit poorer survival compared to those with low MUC4. Our study in Muc4^−/− mice using AOM/DSS model of CRC suggests that Muc4 is involved in the progression of the CRC, where inflammation has a significant role. MUC4 may also be associated with poor prognosis in the subset of IBD patients who develop CRC. However, this needs to be validated in a large cohort of IBD patients that eventually develop CRC. Our future study will aim to address the role of Muc4 in inflammation-independent CRC using genetically defined CDX2P-NLSCre;Apc^+/loxP and CDX2P-NLSCre;KRas^G12D/+ models of CRC.

We found that WT mice frequently had high-grade adenocarcinomas with deeper in growth into the colonic wall than Muc4^−/−mice. These findings are in agreement with the tumorigenic and metastatic role of MUC4 observed in various other epithelial malignancies.^43,47,48 The Muc4-mediated tumorigenicity may be associated with increased cell proliferation as we found significantly higher Ki67-positive tumor cells in WT group compared with Muc4^−/− group. Interestingly, there was no apparent difference in the apoptotic activity between the two groups (data not shown). Altered β-catenin signaling is frequently implicated in human CRC development and progression.^32,33 We found that the levels of nuclear β-catenin expression were similar in two groups. It may indicate a common mechanism of colon cancer in these two mouse models. It also suggests certain comparability with human CRC.

Taken together, we demonstrated for the first time in vivo relevance of Muc4 in colitis and colitis-associated CRC using mice genetically deficient in Muc4. These Muc4^−/− mice are highly resistant to chemical-induced colitis and CRC. These mice will further be useful in understanding the role of Muc4 in intestinal infections as well as colitis-independent CRC.

MATERIALS AND METHODS

Generation of Muc4^−/− mice

The targeting construct for the Muc4 gene knockout was obtained from the Knock Out Mouse Project Repository (UC Davis). The schematic for the WT and targeted allele are depicted in Figure 1a. The targeting construct included an IRES-LacZ cassette in place of exons 2–6 of Muc4 that gets expressed in tissues and cells wherever Muc4 promoter is active. The cells positive for a β-galactosidase staining indicates loss of Muc4. The targeting construct was electroporated into murine embryonic stem cells (129/SvJ) by homologous recombination. The surviving colonies after G418 selection were screened using long-range PCR and few positive clones were confirmed by Southern blotting (Supplementary Figure 1). Two ES cell clones that had undergone correct homologous recombination were microinjected into C57BL/6J blastocysts, the resultant chimeras were bred to C57BL/6J mice for germ line transmission. Heterozygous F1 progenies were bred to generate homozygous Muc4 KO mice (Muc4^−/−). Genotypes of the mice were determined by PCR using the genomic DNA extracted from the tail. Primers used for genotyping are listed in Supplementary Table-1 (A). Six- to 8-week-old mice in the 129/Sv and C57BL/6J mixed background were used for experiments. WT littermates were used as controls. All mouse experiments were performed in compliance with protocols approved by the Institutional Animal Care and Use Committees of the University of Nebraska Medical Center.

LacZ staining

LacZ staining was performed according to previously published protocol.⁵⁰ Briefly, organs collected were washed in phosphate-buffered saline (PBS), fixed for 4 h in lacZ fixative (0.2% glutaraldehyde, 5 mM EGTA and 100 mM MgCl₂ in PBS), washed three times in lacZ wash buffer (2 mM MgCl₂, 0.01% sodium deoxycholate, and 2.5% NP-40 in PBS), cryoprotected in 15% sucrose/PBS for 1 h at 4 °C followed by 30% sucrose/PBS overnight at 4 °C, and cryosections were prepared. In all, 7-μm thick cryosections of different tissues from the WT and Muc4^−/− mice were fixed in 0.2% glutaraldehyde for 10 min and then washed with lacZ wash buffer three times. The slides were stained overnight in lacZ stain solution (0.5 mg/ml X-gal, 5 mM potassium ferrocyanide and 5 mM potassium ferricyanide in lacZ wash buffer) at 37 °C. The slides were rinsed in PBS and were stained with nuclear fast red.

Histology and immunohistochemistry

For histology and immunohistological analyses, tissues collected from WT and Muc4^−/− mice were fixed in 10% buffered formalin, processed, paraffin embedded and sectioned (5 μm). Tissue sections were stained with hematoxylin and eosin (for pathological evaluation), periodic acid–Schiff (Sigma-Aldrich, St Louis, MO, USA; for neutral carbohydrates), Alcian blue (Poly Scientific, Bay Shore, NY, USA; for acidic polysaccharides). For immunohistochemical analysis using antibodies, sections were deparaffinized in xylene (x4) and then rehydrated with graded alcohols. Endogenous peroxidase was blocked using 3% H₂O₂ in methanol for 30 min. Antigen retrieval was performed using 0.01 M preheated citrate buffer (pH – 6.0, 90 °C) for 15 min. The slides were then washed with PBS and blocked using horse serum (ImmPRESS kit; Vector Labs, Burlingame, CA, USA) for 2 h. The sections were incubated overnight at 4 °C with respective primary antibodies diluted in PBS (anti-CD3, 1:300, Abcam, Cambridge, MA, USA; #ab16669; anti-F4/80, 1:75, e-Biosciences, San Diego, CA, USA; #14-4801-82; anti-β-catenin, 1:500, BD-Biosciences, San Jose, CA, USA; #610153; anti-Ki67, 1:300, Abcam, #ab15580). Slides were washed four times with PBS and incubated with appropriate secondary antibody (peroxidase-labeled anti-mouse/anti-rabbit IgG (ImmPRESS kit, Vector Labs) or anti-rat IgG, Thermo Scientific, Grand Island, NY, USA) for 1 h at room temperature. The slides were washed four times in PBS and were developed using DAB as substrate (DAB substrate kit; Vector Labs). The sections were counterstained with hematoxylin (Vectors Lab) and washed in tap water, dehydrated in increasing grades of alcohol (20–100%) and then with xylene and dried. The slides were mounted in paramount mounting medium (Fisher Scientific, Pittsburg, PA, USA) and photographed using Nikon Eclipse E400 light microscope (Kawasaki, Japan). Quantification of CD3⁺ T-lymphocytes and F4/80⁺ histiocytes was performed by counting CD3⁺ and F4/80⁺ cells in 10 high-powered fields from each mouse. Percentage of ki67-positive cells was calculated as a ratio of number of Ki67⁺ nuclei to total number of nuclei per high-powered field with 10 high-powered fields from each mouse.

DSS-induced colitis

To determine the lethality because of DSS, mice aged 6–8 weeks (WT and Muc4^−/− males and females) were given 3% DSS (40 kDa, TdB Consultancy, Uppsala, Sweden) in drinking water for 21 days. Mice were monitored regularly for mortality. To induce colitis, mice were given 2% DSS in drinking water for 7 days followed by 7 days of regular drinking water and killed for histological analysis. During the course of the experiment, mice were monitored daily for their body weight, stool consistency or diarrhea and presence of gross or occult blood in the rectum or stool according to previously published study.⁵¹ Colon length was measured at the end of the experiment. Tissues were then either fixed in 10% formalin and embedded in paraffin for histopathological analysis or were snap frozen in liquid nitrogen for RNA extraction and subsequent mRNA analyses. Histological scoring for the DSS-induced colitis was carried out in a blinded manner with a combined score for extent of infiltration of inflammatory cells and tissue damage separately for proximal and the distal colon.⁵¹

AOM/DSS-induced colorectal tumorigenesis in mice

Six- to 8-week-old male and female WT and Muc4^−/− mice were administered a single intraperitoneal injection of AOM (10 mg/kg body weight; Sigma-Aldrich). One week after injection, mice were fed with 2% DSS (40 kDa, TdB Consultancy) in drinking water for 7 days and were then allowed to recover for 15 days. This cycle was repeated two times followed by regular drinking water till killing. Mice were killed at 24 weeks following AOM injection. The colon was cut open longitudinally and examined for the presence of gross tumors. Number and size of the tumor nodules were recorded. Tissues were then fixed in 10% formalin and embedded in paraffin for histopathological analysis.

Cell culture and transfections

HCT-8 colon cancer cells from ATCC (Manassas, VA, USA) were cultured in alpha-minimum essential medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 μg/ml streptomycin and 4 mM L-glutamine. MUC4 was knocked down in HCT-8 cells using two short hairpin RNA oligos (shMUC4 #1: 5′-GGAGATGGCTATTTCGAAA-3′ and shMUC4 #3: 5′-GCATGAAACTCGACGCGTT-3′) cloned into pSUPER.retro. puro vector between BglII and HindIII resctriction sites. Briefly, phoenix cells were transfected with either control (SCR) or short hairpin RNA vectors using Lipofectamine 2000 (Invitrogen, Grand Island, NY, USA) and the supernatants collected at 48 and 72 h following transfection were used to infect HCT-8 cells. After 48 h, cells were grown in medium containing 5 μg/ml of puromycin. The selected pooled populations of cells were used for further experiments.

Immunoblotting

For immunoblotting, cell lysates were collected in RIPA buffer (50 mM Tris-HCl pH-7.5, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate and 0.1% sodium dodecyl sulfate) supplemented with complete protease inhibitor mixture (Roche, New York, NY, USA), 2 mM Na₃VO₄, 10 mM NaF and 1 mM PMSF on ice. Cell lysates cleared by centrifugation were quantified using bicinchoninic acid method and run in 2% sodium dodecyl sulfate–agarose for MUC4 and 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis for β-actin under reducing conditions and blotted onto polyvinylidene difluoride membrane (Millipore, Massachussets, MA, USA). Membranes were probed with specific antibodies. Blots were washed and probed with respective secondary peroxidase-conjugated antibodies, and the bands visualized by chemiluminescence (Thermo Scientific). Antibodies used were anti-MUC4 (1:1000, 8G7, developed in our lab) and anti-β-actin (1:5000, Sigma, St Louis, MO, USA; #A1978).

RNA extraction from mouse tissues and HCT-8 cells

Tissues collected from the WT and Muc4^−/− mice were snap frozen in liquid nitrogen and were stored at – 80 °C until further use. Total RNA was extracted using miRVana miRNA isolation kit (Ambion/Life Technologies, Grand Island, NY, USA) according to the manufacturer’s instructions. RNA from HCT-8 cells was isolated using RNeasy minikit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. The extracted RNA was further treated with DNase I (Qiagen) to obtain pure mRNA without DNA contamination.

Polymerase chain reaction

A total of 2 μg of DNase I-treated mRNA was converted to complementary DNA using either oligo-dT or random hexamer primers and SuperScript reverse transcriptase II (Life Technologies, Grand Island, NY, USA) was used for the reverse transcription according to the manufacturer’s instructions. Conventional end point PCR was performed to demonstrate the loss of Muc4 expression following deletion of Muc4. Gapdh was used as an internal reference gene. Primers used for Muc4 expression in conventional PCR are listed in Supplementary Table-1 (B).

Real-time PCR was performed using LightCycler 480 SYBR Green I Master Mix (Roche) with LightCycler 480 II (Roche) Real-Time PCR system. The relative amount of expression was calculated using 2−ΔΔCt method. The average of three independent analyses for each gene and sample was calculated and was normalized to the endogenous reference control gene Gapdh. The primers used for expression analysis of mucin genes (MUC2, Muc1, Muc2, Muc3, Muc4 and Muc13) and inflammatory cytokines (IL-1β, TNF-α, Lysozyme M, SLPI and IL-10) are listed in Supplementary Table-1 (C).

Statistical methods

Statistical analyses were carried out using two-tailed Student’s t-test or Wilcoxon rank sum test and P ≤ 0.05 was considered to be statistically significant. The Kaplan–Meier method was used to estimate survival distribution and log-rank test was used to compare survival distributions between the groups. SAS software (SAS Institute Inc., Cary, NC, USA) was used for data analysis and R software (R Foundation for Statistical Computing, Vienna, Austria) was used for making box plots. For Kaplan–Meier survival analyses, 15 age-matched animals per group (WT and Muc4^−/−) were randomly distributed and the final analysis was carried out in a blinded manner. For DSS-induced colitis experiment, age-matched 18 WT (males and females combined) and 18 Muc4^−/− (males and females combined) mice were randomly distributed and all the analyses associated with these animals were performed in a blinded manner. Those animals that died during the course of the experiment were excluded from the analyses. For AOM/DSS-induced CRC experiment, 10 age-matched animals per group (WT and Muc4^−/−) were randomly distributed and the final analysis was carried out in a blinded manner. Animals that died during the course of the experiment were excluded from the analyses. All the animal experiments were performed once and mean values were presented from each group of animals.

ABBREVIATIONS

AOM

azoxymethane

CD

Crohn’s disease

CRC

colorectal cancer

DSS

dextran sodium sulfate

IBD

inflammatory bowel disease

IL

interleukin

TNF

tumor necrosis factor

UC

ulcerative colitis

WT

wild type