Ion Transport Basis of Diarrhea, Paneth Cell Metaplasia, and Upregulation of Mechanosensory Pathway in Anti-CD40 Colitis Mice

Abstract

Background

Anti-Cluster of differentiation (CD)-40-induced colitis, driven by innate inflammatory responses in the intestine, is a potent animal model exhibiting IBD pathophysiology including diarrhea. However, the ion transport basis of diarrhea and some key mucosal pathways (Paneth cells, stem cell niche, and mechanosensory) in this model have not been investigated.

Methods

Mucosal scrapings and intestinal tissue from control and CD40 antibody (150 µg) treated Rag2^−/− mice were examined for gut inflammation, Paneth cell numbers, expression of key transporters, tight/adherens junction proteins, stem cell niche, and mechanosensory pathway via hematoxylin and eosin staining, quantitative polymerase chain reaction, and western blotting.

Results

Compared with control, anti-CD40 antibody treatment resulted in a significant loss of body weight (P < .05) and diarrhea at day 3 postinjection. Distal colonic tissues of anti-CD40 mice exhibited increased inflammatory infiltrates, higher claudin-2 expression, and appearance of Paneth cell–like structures indicative of Paneth cell metaplasia. Significantly reduced expression (P < .005) of downregulated in adenoma (key Cl^- transporter), P-glycoprotein/multidrug resistantance-1 (MDR1, xenobiotic transporter), and adherens junction protein E-cadherin (~2-fold P < .05) was also observed in the colon of anti-CD40 colitis mice. Interestingly, there were also marked alterations in the stem cell markers and upregulation of the mechanosensory YAP-TAZ pathway, suggesting the activation of alternate regeneration pathway post-tissue injury in this model.

Conclusion

Our data demonstrate that the anti-CD40 colitis model shows key features of IBD observed in the human disease, hence making it a suitable model to investigate the pathophysiology of ulcerative colitis (UC).

Graphical Abstract

Graphical Abstract.

Made in BioRender.com

Key Messages.

What is already known?

The innate immune cell–mediated anti-CD40 colitis model exhibits features similar to other models of gut inflammation including diarrheal phenotype.

What is new here?

Mechanisms underlying diarrhea in this model involve downregulation of Cl^- transporter DRA. Also, reduced expression of xenobiotic transporter MDR-1 and adherens junction protein E-cadherin was observed. Additionally, key features of Paneth cell metaplasia and induction of claudin-2 and fetal-like signature via mechanosensory YAP-TAZ pathway is similar to what is seen in many cases of the human disease.

How can this study help patient care?

This research supports the suitability of this model to investigate the human disease and in future studies may lead to identification of new targets for intervention.

Introduction

The immune-based anti-CD40 colitis mouse model has gained attention as an important tool to study the involvement of innate immune cells in the pathogenesis of inflammatory bowel diseases (IBD). The advantage of utilizing this model is that it is a unique model driven by interleukin (IL) 23–producing gut resident macrophages and IL 22–producing group 3 innate lymphoid cells and is established within 4-5 days postinjection of the anti-CD40 antibody.¹ Therefore, this may be a great model to study the role of the innate immune system in colonic inflammation.

Although several models of immune-based colitis are available, administration of anti-CD40 antibody induces colitis in T and B cell-deficient mice (Rag2 KO) mice, allowing investigations into the CD40 activation pathway.² CD40 is a member of the tumor necrosis factor receptor (TNFR) superfamily and is expressed on the surface of various cell types including macrophages, dendritic cells (DCs), B cells, microglia, endothelial cells, keratinocytes, and epithelial cells.^3,4

The CD40/CD40 ligand (CD40L) system plays a critical role as a key regulator and amplifier of immune reactivity and contributes to the development and progression of many autoimmune diseases such as IBD.⁵ Additionally, CD40 is highly expressed by the colonic lamina propria antigen-presenting cells, including macrophages and DCs; therefore, activation of this pathway causes systemic and intestinal inflammation which propagates by secreted cytokines in the colon with minimal effects in the small bowel.

The unique involvement of innate immune cells in this model of colitis makes it a suitable platform to compare mucosal changes in this model to those which is mainly mediated or propagated via adaptive immune cells. Previous studies have demonstrated how colitis models present diarrheal phenotype and alterations in various membrane proteins, as well as mucosal changes similar to human IBD.⁶ While colitis-associated diarrhea is considered to be via downregulation of intestinal NaCl absorption, the ion transport basis of diarrhea associated with anti-CD40 colitis and mucosal changes related to this model have not been investigated. Also, previous studies have shown that changes in stem cell responses are associated with distinct patterns of cytokine regulation across different mouse models of colitis.⁷ Therefore, the purpose of the current work was to examine the effects of innate immune cell–based colitis on diarrheal phenotype, gut inflammation, and expression of key ion/drug transporters and tight/adherens junction proteins, as well as the mucosal stem cell compartment.

Anti-CD40 colitis model mediated via innate immune cell exhibits key features similar to other models of gut inflammation including diarrheal phenotype. In the current study, we demonstrate the downregulation of DRA (downregulated in adenoma, SLC26A3, key Cl^- transporter) and P-glycoprotein/MDR1 (xenobiotic transporter), which are features observed in human colitis. We also demonstrate a marked increase in the expression of claudin-2 and a decrease in the expression of E-cadherin (adherens junction protein), as well as Paneth cell metaplasia. Additionally, colitis driven by anti-CD40 antibody resulted in hyperproliferative crypts in the colon, with the induction of the fetal-like signature via upregulation of the mechanosensory YAP-TAZ pathway and differentially altered stem cell markers, making it a suitable model to investigate the involvement of innate immune cells in the pathogenesis of UC.

Methods

Mice

We purchased C57BL/6 (WT) and Rag2^−/− mice on the C57BL/6 background from Jackson Laboratories. All the mice were maintained, and the animal experiments were conducted at the Vanderbilt University Medical Center (VUMC) animal facilities, which were approved by Institutional Animal Care and Use Committee guidelines at VUMC.

Development of Anti-CD40 Colitis

Seven-week-old Rag2^−/− mice received either 150 µg of antimouse CD40 antibody (CD40 colitis) clone FGK4.5 (Bio X Cell) or phosphate-buffered saline (PBS) (control) intraperitoneally as previously described.⁸ Mice were weighed prior to injection and every day thereafter. Mice were monitored daily for signs of disease such as rectal bleeding, diarrhea, and scruffiness. Mice were euthanized 3 days postinjection, and intestinal tissues were utilized for analysis.

RNA Extraction and RT PCR

RNA was extracted from mucosal scrapings of mice ileum, and proximal and distal colons with the use of the RNeasy Mini Kit from Qiagen (Valencia, CA) per the manufacturer’s protocol. The extracted RNA was reverse-transcribed and amplified using SYBR green reagent with gene-specific primers listed in Table 1.

Table 1.

Realtime PCR mouse primers used in the study.

GAPDH	Forward: 5’-AGGTCGGTGTGAACGGATTTG-3’ Reverse: 5’-TGTAGACCATGTAGTTGAGGTCA-3’
DRA	Forward: 5’-TGGTGGGAGTTGTCGTTACA-3’ Reverse: 5’-CCCAGGAGCAACTGAATGAT-3’
NHE-3	Forward: 5’-TGAAAAGCAGGACAAGGAAATCT-3’ Reverse: 5’-TTGGCCGCCTTCTTATTCTGG-3’
MDR-1	Forward: 5’-CTGTTGGCGTATTTGGGATGT-3’ Reverse: 5’-CAGCATCAAGAGGGGAAGTAATG-3’
MPO	Forward: 5’-AGTTGTGCTGAGCTGTATGGA-3’ Reverse: 5’-CGGCTGCTTGAAGTAAAACAGG-3’
IL-23	Forward: 5’-ATGCTGGATTGCAGAGCAGTA-3’ Reverse: 5’-ACGGGGCACATTATTTTTAGTCT-3’
IL-22	Forward: 5’-ATGAGTTTTTCCCTTATGGGGAC-3’ Reverse: 5’-GCTGGAAGTTGGACACCTCAA-3’
LGR5	Forward: 5’-CAGCCTCAAAGTGCTTATGCT-3’ Reverse: 5’-GTGGCACGTAACTGATGTGG-3’
BMI1	Forward: 5’-ATCCCCACTTAATGTGTGTCCT-3’ Reverse: 5’-CTTGCTGGTCTCCAAGTAACG-3’
OLFM4	Forward: 5’-GCCACTTTCCAATTTCAC-3’ Reverse: 5’-CTTGCTGGTCTCCAAGTAACG-3’
LRIG1	Forward: 5’-TTGAGGACTTGACGAATCTGC-3’ Reverse: 5’-CTTGTTGTGCTGCAAAAAGAGAG-3’
HOPX	Forward: 5ʹ-TCTCCATCCTTAGTCAGACGC-3ʹ Reverse: 5ʹ-GGGTGCTTGTTGACCTTGTT-3ʹ

Protein Extraction and Western Blotting

Small amounts of distal and proximal colonic mucosal scrapings were harvested in RIPA Lysis buffer (Cell signaling, Danvers, MA) with protease inhibitor (Roche, Basel, Switzerland) and phosphatase inhibitor (Sigma-Aldrich, St. Louis, MO). These samples were then homogenized using a homogenizing gun and sonicated until a homogenous lysate was prepared. Samples were then pelleted down by centrifugation at 7500 g for 7 minutes, and the supernatants were used for determining protein concentration with Bradford assay (Bio-Rad, Hercules, CA). Twenty-five μg of protein from each sample were loaded onto a 7.5% precast SDS PAGE gel (Bio-Rad), then transferred to a polyvinylidene difluoride (PVDF) membrane (Bio-Rad) and probed with the respective primary antibodies based on the protein of interest at 4°C overnight after blocking in 5% PBS milk for 1 hour at room temperature. The primary antibodies include DRA rabbit polyclonal antibody, 1:500 in 1% PBS milk. The DRA antibody was raised against the C-terminal amino acid (745-764) sequence: INTNGGLRNRVYEPVETKF of SLC26A3 (accession number: BC025671) at Research Resource Centre (RRC) of Univ. of Illinois at Chicago; NHE3 rabbit antibody, 1:3000 in 1% PBS milk (gift from Dr. Chris Yun, Emory University), E-cadherin (Cell Signaling), LYZ-1 (Santa Cruz), Caludin-2 (Invitrogen), Annexin A-1 (Invitrogen), TEAD1 (Cell Signaling), TAZ (Cell Signaling), FAK (Invitrogen), COX-2 (Invitrogen), P-glycoprotein/MDR1 (Santa Cruz), Sca-1 (Invitrogen), YAP (Cell Signaling), FOXM1 (Cell Signaling), and GAPDH rabbit antibody (Sigma), 1:8000 in 1% PBS milk. The membranes were then washed and probed with secondary horseradish peroxidase (HRP)-conjugated antibodies at a ratio of 1:5000 (Promega, Madison, WI) for 1 hour at room temperature. After washing the membranes, they were processed as per the manufacturer’s protocol (Bio-Rad) for visualizing protein bands via enhanced chemiluminescence.

Immunofluorescence Staining

Frozen sections of 5 μm from mouse distal colon were fixed with 4% paraformaldehyde for 20 minutes. These sections were then permeabilized with 0.3% Nonidet-P40 (NP 40) (Fisher scientific) for 5 minutes and blocked with 5% normal goat serum (NGS) in PBS for 2 hours at room temperature. These were then incubated with primary antibodies for DRA (rabbit polyclonal antibody raised against the same sequence as previously described at Pocono Rabbit Farm, Canadensis, PA), mouse villin antibody (Invitrogen, Carlsbad, CA), Occludin (Invitrogen), Zona occludens-1 (Invitrogen), E-cadherin (Cell Signaling, Danvers, MA), YAP (Cell Signaling), LYZ-1 (Biogenex, Fremont, CA), and Ki67 (Invitrogen) at 1:100 ratio in 1% NGS in PBS. After 3 washes of PBS, the sections were incubated with fluorescently tagged secondary antibodies (Invitrogen). The slides were once again washed in PBS and mounted with DAPI (Invitrogen) using cover slips. The slides were sealed with clear nail polish and stored at −20°C until imaged. The images were captured using the BX-100 fluorescent microscope.

Hematoxylin and Eosin Staining

Formalin-fixed paraffin-embedded distal colonic tissues were sectioned at 5-μm thickness using a microtome and stained with hematoxylin and eosin according to the manufacturer’s protocol (Sytek Laboratories, Upsala, MN). These sections were imaged at 20x magnification using Olympus BX light microscope.

Periodic Acid Schiff Staining and Quantitation

Paraffin embedded 5 µm thickness colonic tissues were stained with periodic acid schiff (PAS) staining kit (Sigma) as per manufacturers protocol. Quantitation of PAS-positive crypts were conducted in a blinded fashion in at least 5 mice per group as described previously.

Statistical Analysis

Statistical significance between the experimental groups was determined by application of an unpaired 2-tailed Student t test using Prism version 9; P < .05 was considered to be statistically significant.

Results

Anti-CD40 Mediated-colitis Developed After 3 Days Postinjection and Demonstrated Significant Loss of Body Weight and Diarrheal Phenotype

Administration of anti-CD40 antibody in Rag2^−/− mice gave rise to a drastic change in their body weight starting from day 2 postinjection (Figure 1A). The mice continuously lost weight compared with PBS-injected control animals (>10% from initial weight) until the end point of the study. Whole excised colons of anti-CD40 colitis mice displayed marked diarrheal phenotype as depicted in Figure 1B. The prominent solid fecal pellets observed in the control mice were absent in the anti-CD40 colitis mice.

Figure 1.

Anti-CD40 colitis mice lost significant body weight and displayed diarrheal phenotype. A, Body weight change in control vs anti-CD40 colitis mice. 7-week-old Rag2^−/− mice received either 150 μg anti-CD40 antibody (anti-CD40 colitis) or PBS (control) via intraperitoneal injections. Mice were euthanized 3 days postinjection. Body weight was measured daily. B, Representative images of whole excised colon of control and anti-CD40 colitis mice. Arrowheads indicate solid fecal pellets. Data represented as mean ± standard error of mean n = 5-8, ***P < .0005, ****P < .0001 vs control.

Inflammation Due to Anti-CD40 Colitis Was Most Prominent in the Colon

The inflammation that developed after anti-CD40 administration was mainly apparent in the colon of these mice. The mRNA expression of the key driving cytokines including interleukin (IL)-23 and interleukin (IL)-22 was significantly higher in the distal colon (Figure 2A, B). These were much less prominent in the proximal colon and were absent in the ileal mucosa (data not shown). In addition, neutrophil-mediated myeloperoxidase (MPO) expression was also significantly higher in the distal colon of anti-CD40 colitis mice (Figure 2C). This was also evident in the distal colonic histological micrographs (Figure 2D), where inflammatory infiltrate and luminal exudates were only present in anti-CD40 colitis mice. Another important enzyme in colonic inflammation, cyclooxygenase-2 (COX-2), was also markedly elevated in anti-CD40 colitis colons compared with controls (Figure 2E). In addition, as shown before, anti-CD40 colitis mice also had significantly lower goblet cell numbers as determined by PAS-positive cells per crypts (Figure 2F, G). Apart from these inflammatory mediators, IL-33 and IL-6 were also elevated in the colons of anti-CD40 colitis mice (data not shown). Since the presence of diarrhea was apparent in anti-CD40 colitis, we next examined the expression of important ion transporters.

Figure 2.

Inflammation was prominent in the distal colon of mice with anti-CD40 colitis. Representative bar graphs showing relative mRNA expression of (A) IL-23; (B) IL-22; (C) Myeloperoxidase (MPO) in distal colonic mucosa normalized to internal control Glyceraldehyde 3-phosphate dehydrogenase (GAPDH). D, Representative micrographs of distal colonic tissue sections of 5-μm thick stained with hematoxylin and eosin stain. Scale bar- 50 μm. E, Representative western blot and bar graph showing densitometric analysis of COX2 normalized to GAPDH in distal colonic mucosa of control and anti-CD40 colitis mice. F, Representative micrographs of Periodic acid and Schiff-stained distal colonic sections of control and anti-CD40 colitis mice. G, Violin plot showing quantitated PAS positive goblet cells per crypt in colonic tissues of control VS CD40 colitis mice. Data represented as mean ± standard error of mean n = 5-8, *P < .05, **P < .005, ****P < .0001 vs control.

Colonic Tissues of Anti-CD40 Colitis Mice Had Significantly Lower Expression of Important Ion and Xenobiotic Transporters

Upon examining the expression of important sodium chloride transporters it was evident that both the mRNA and protein expression of DRA (Slc26a3 or downregulated in adenoma), the major chloride transporter in the distal colon, were drastically reduced in distal colon of mice with anti-CD40 colitis (Figure 3A, B). These results were further corroborated with immunofluorescence staining of DRA (green), where apical expression compared with villin (red) was severely affected in CD40 colitis, as depicted in Figure 3C. In addition, Slc26a6, another chloride transporter expressed minimally in the large intestine, or putative anionic transporter-1 (PAT1), was also significantly reduced in the colons (data not shown). Next, the sodium counterpart transporter Slc9a3 or sodium hydrogen exchanger-3 (NHE3) mRNA and protein expression were determined. Interestingly although the mRNA expression was reduced in colitis mice (Figure 3D), the expression of protein of NHE3 remained unchanged (Figure 3E). This is similar to what has been observed in other models of colitis; and since loss of DRA in itself can cause diarrhea, this could be the main cause for the observed diarrheal phenotype in the colon. Another important xenobiotic transporter multidrug-resistance protein-1 (MDR-1) was next investigated. Multidrug-resistance protein-1 is highly expressed in the colon, and its loss is implicated in the pathogenesis of IBD. We observed a marked reduction in both the mRNA and protein expression of this membrane transporter, as shown in Figure 3F, G. These results attest to the manifestation of diarrhea in this model and show vast similarities to other models of colitis where all immune cells are present.

Figure 3.

Expression of Important Intestinal Sodium Chloride and Xenobiotic Transporters Were Downregulated in anti-CD40 Colitis Colon. A, Representative bar graph displaying relative mRNA expression of DRA mRNA expression normalized to internal control GAPDH. B, Representative western blot and bar diagram showing densitometric analysis of DRA protein expression normalized to GAPDH. C, Immunofluorescence images showing DRA (green) staining along with the apical marker Villin (Red) and nuclear stain DAPI (blue). Scale bar- 10 μm. D, Bar graph depicting relative mRNA expression of NHE3 in distal colon of control vs anti-CD40 colitis mice. E, Representative western bot and densitometric analysis of NHE3 expression normalized to GAPDH in control vs anti-CD40 Colitis mice. F, Bar graph depicting relative mRNA expression of MDR1 in distal colon of control vs anti-CD40 colitis mice. G, Representative western bot and densitometric analysis of MDR1 expression normalized to GAPDH in control vs anti-CD40 Colitis mice. Data represented as mean ± standard error of mean n = 5, *-P < .05, **-P < .005, ***-P < .0005, ****-P < .0001 vs control.

Important Tight and Adherens Junction Proteins Are Dysregulated in Anti-CD40 Colitis Colons

We and others have shown how inflammation causes alterations in expression of tight (TJ) and adherens junction (AJ) proteins. Since there was marked inflammation and loss of other membrane transporters, it was of interest to determine the expression of TJ/AJ proteins in the colons of anti-CD40 colitis mice. Surprisingly, the expression of important TJ proteins Zona occludens-1 (ZO-1) and Occludin was unaffected by anti-CD40 colitis, as clearly demonstrated in Figures 4A and 4B. Both these proteins were intact, and the expression levels at the apical membrane were similar to controls, as the punctate appearance on cross-section view of colonic mucosa was observed after colitis development. However, the leaky pore forming claudin-2 levels were markedly elevated in the colons of colitis mice (Figure 4C). Interestingly, adherens junction protein E-cadherin was reduced in colitis colons, which is presented as both total mucosal protein level and its expression after immunofluorescence staining (Figure 4D, E).

Figure 4.

Important tight and adherens junction proteins were dysregulated in anti-cd40 colitis colon. A, Representative images of distal colonic tissues stained for (A) Zona occludens-1 (ZO-1/red) (B) Occludin (red) and Villin (green) and DAPI (blue). C, Representative western blot and densitometric analysis of Claudin-2 protein expression normalized to GAPDH. D, Representative western blot and densitometric analysis of E-cadherin protein expression normalized to GAPDH. E, Immunofluorescence images showing E-cadherin (red) and DAPI (blue) staining in control and anti-CD40 colitis mice distal colonic tissues. Scale bar-10 μm. Data represented as mean ± standard error of mean n = 5-10, *P < .05 vs control.

Anti-CD40 Colitis Mice Displayed Paneth Cell–like Structures in the Colon

Paneth cell metaplasia is observed in a population of IBD patients, and the mechanism behind this phenomenon still remains elusive. Interestingly, in anti-CD40 colitis mice, we observed a robust induction of lysozyme-1 (LYZ1) protein expression (Figure 5A). These observations were further validated by utilizing hematoxylin and eosin staining, where Paneth cell–like structures (identified by presence of granules) are shown in Figure 5B. Further, we confirmed the presence of LYZ-1 via immunostaining (Figure 5C). This is a striking feature observed in this innate immune cell–based model, which requires further investigation.

Figure 5.

Anti-CD40 colitis mice displayed paneth cell like structures in the colon. A, Representative histological micrograph of control and anti-CD40 colitis colon stained in H and E stain. Arrow heads point at Paneth cell-like structures. Scale bar- 50μm. B, Representative western blot and densitometric analysis of Lysozyme-1 (LYZ1) protein expression normalized to GAPDH. C, Fluorescent micrograph of distal colonic tissues stained with LYZ1 antibody (green) and DAPI (blue). Scale bar-,10 μm. Data represented as mean ± standard error of mean n = 3-5, **P < .005 vs control.

Colonic Stem Cell Markers Were Differentially Modulated in Anti-CD40 Colitis

Since colitis can affect the stem cell compartment and is observed in other models of colitis, we next aimed at determining mRNA levels of various colonic stem cell markers in the colonic tissues. We observed no alterations in the crypt base columnar cell (CBC) adult intestinal stem cell markers Lgr5 (Leucine-rich repeat-containing G-protein coupled receptor 5) and Olfm (Olfactomedin 4). As shown in Figure 6A and B, CBC cell population remained unaltered in anti-CD40 colitis and was similarly expressed in control vs colitis mice colons. Next, the +4 stem cell markers, which are known to be irradiation-resistant and are more quiescent, were determined. Interestingly, the markers Lrig1 (Leucine-rich repeats and immunoglobulin-like domains protein 1) and HOPX (Homeodomain-only protein) were both significantly reduced in anti-CD40 colitis colons (Figure 6C, D). However, the marker Bmi1 (Polycomb complex protein BMI-1) was significantly increased in colitis mice (Figure 6E). These alterations prompted us to determine if colitis-associated regeneration pathway with fetal-like gene signature was active in this model of colitis.

Figure 6.

Distal colonic crypt stem cell populations were altered due to anti-CD40 colitis. Representative bar graphs showing relative mRNA expression of stem cell markers (A) LGR5; (B) OLFM4; (C) LRIG1; (D) HOPX (E) BMI1 normalized to housekeeping control GAPDH in distal colonic tissues of control vs anti-CD40 colitis mice. Data represented as mean ± standard error of mean n = 4-8, *-=P < .05, ***P < .0005 vs control.

Fetal-like Signature Mediated via YAP/TAZ Pathway Was Activated in Anti-CD40 Colitis Colons

The mechanosensory YAP/TAZ pathway has been implicated in colitis-associated regeneration. Although we did not observe a significant reduction in CBC stem cell markers, there were significant alterations in quiescent stem cell markers. Therefore, we focused our attention to the mechanosensory YAP/TAZ pathway, which is driven by stem cell marker Sca-1. In fact, anti-CD40 colitis mice had a robust elevation in the protein expression of Sca-1 (Figure 7A), along with YAP (Figure 7B). These proteins were colocalized at the crypts of the colon, and activated YAP was prominently present in the nucleus (Figure 7C), demonstrating activation of this pathway.

Figure 7.

Colonic crypts of Anti-CD40 Colitis Mice Expressed Increased YAP and Fetal Stem Cell Marker Sca-1. A, Representative western blot and densitometric analysis of Sca-1 (Stem cells antigen-1) protein expression normalized to GAPDH. B, Representative western blot and densitometric analysis of YAP (Yes associated protein) protein expression normalized to GAPDH. C, Immunofluorescence images showing Sca-1 (red), YAP (green) and DAPI (blue) staining in control and anti-CD40 colitis mice distal colonic tissues. Scale bar- 50 μm. Data presented as mean ± standard error of mean n = 8-10, ****P < .0001 vs control.

Colonic crypts of anti-CD40 colitis mice were hyperproliferative and had significantly more staining for Ki67 in crypts compared with control mice (Figure 8A). To further explore the activation of fetal signature-mediated crypt proliferation via YAP/TAZ pathway in the colons of these mice, we determined protein expression of various other players shown in Figure 8B. In this regard, we observed a robust induction in YAP binding partner TAZ (Figure 8C) and TEAD1 (Figure 8D) protein expression. Since the expression of YAP/TAZ pathway was observed in the colons, we next investigated if it was via mechanosensory cues by activation of β-integrin, which in turn increases FAK protein expression (Figure 8E). Since FAK expression was also markedly elevated, we determined the protein levels of another important protein, FOXM1, known to increase activation of YAP pathway in cancers. FOXM1 has been identified as a protein that stabilizes YAP and prevents its degradation; FOXM1 was almost 2-fold increased in anti-CD40 colitis mice.

Figure 8.

Mechanosensory YAP/TAZ pathway and downstream proteins were upregulated in the hyperproliferative colonic crypts of anti-CD40 colitis. A, Representative immunofluorescence image of control and anti-CD40 colitis mice stained with Ki67 (green) and DAPI showing hyperproliferative crypts. Scale bar- 20 μm. B, Schematic diagram showing important proteins involved in activation of the mechanosensory YAP/TAZ fetal signature pathway. Abbreviations: YAP, Yes associated protein; TAZ, transcriptional co-activator with PDZ-binding motif; FAK, focal adhesion kinase; TEAD-TEA/ATTS domain; FOXM1, Forkhead box protein M; Sca-1-Stem cells antigen-1). C, Representative western blot and densitometric analysis of TAZ protein expression normalized to GAPDH. D, Representative western blot and densitometric analysis of TEAD protein expression normalized to GAPDH. E, Representative western blot and densitometric analysis of FAK protein expression normalized to GAPDH. F, Representative western blot and densitometric analysis of FOXM1 protein expression normalized to GAPDH. G, Representative western blot and. H, Densitometric analysis of Annexin-A1 protein expression normalized to GAPDH. Data presented as mean ± standard error of mean n = 4-10, **P < .005, ***P < .0005, ****P < .0001 vs control.

Finally, another fetal marker, Annexin-A1, was also markedly elevated in anti-CD40 colitis, along with Sca-1 which has also been observed in human colitis tissues. Overall, these results demonstrate that anti-CD40 colitis affects the stem cell compartment, and the regeneration of the damaged tissue is being mediated by fetal-like gene activation.

Discussion

In the current study we demonstrated that anti-CD40 colitis resulted in (1) significant loss of body weight and diarrheal phenotype in the colon; (2) increase in mRNA expression of pro-inflammatory cytokines; (3) Paneth cell metaplasia; (4) reduction in key Cl transporter SLC26A3 and important xenobiotic transporter MDR1 mRNA and protein expression; (5) increased claudin-2 tight junction protein and decreased expression of adherens junction protein E-cadherin; and (6) upregulation of the mechanosensory YAP-TAZ pathway.

Earlier reports have demonstrated that colitis-associated diarrhea is governed by downregulation of intestinal transporters involved in NaCl absorption. Electroneutral NaCl absorption is the predominant mechanism for Na⁺, Cl⁻, and fluid absorption in the ileum and colon.⁶ Dysregulation of this electrolyte absorption via coupled operation of NHE3 and DRA has been indicated as a major contributing factor to IBD-associated diarrhea.^9-11 Similar to other models of IBD, in the current study we also observed a marked downregulation of chloride transporters DRA and PAT1 in the colonic mucosa of anti-CD40 treated mice. In addition, a decrease in NHE3 (the major transporter for Na⁺ absorption in the intestine)-mediated Na⁺ absorption has also been linked to IBD associated diarrhea.^6,12,13 In the current study, NHE3 mRNA expression was significantly decreased in the colon of anti-CD40 colitis mice compared with controls, with no change in the protein expression. Studies utilizing UC patient biopsies and murine inflammatory models have shown that loss of NHE3 activity (as noted in IBD) does not necessarily correlate with decrease in NHE3 gene expression and/or altered protein expression or apical localization.⁶ In the current study, while NHE3 activity was not assessed, unchanged protein expression with significantly downregulated NHE3 mRNA expression also indicates that decreased electroneutral NaCl absorption appears to be the mechanism underlying anti-CD40 colitis-associated diarrhea.

In IBD models and patients, loss of DRA protein has been shown to be associated with intestinal inflammation.⁶ The DRA gene has been established as the key mediator of chloride/bicarbonate exchange in the electroneutral NaCl absorption process. Mutations in the DRA gene result in congenital chloride diarrhea, and its deficiency has been shown to be the most critical event in IBD and associated diarrhea.⁶ Our current study suggests that similar to IBD patients and other animal models of IBD, a marked reduction in DRA expression underlies anti-CD40 colitis-induced diarrhea.

The DRA gene has recently emerged to play a critical role in maintenance of gut epithelial homeostasis.^14,15 Additionally, DRA KO mice phenotypically display high IBD risk, loss of mucus barrier, and altered tight junctions/adherens junctions, dysregulated lamina propria lymphocytes, and gut dysbiosis.^14-17 Collectively, these earlier reports on DRA and NHE3 (as discussed previously) indicate that in addition to their key roles in electrolyte absorption and hence in diarrheal phenotype,¹⁸ altered DRA and NHE3 may also contribute to the inflammation associated with the anti-CD40 colitis-associated epithelial barrier loss (increase in claudin-2 and decrease in E-cadherin).

Additionally, we observed marked loss in mRNA and protein expression of MDR1 in distal colon of anti-CD40 colitis mice compared with control mice. Multidrug resistantance-1 is highly expressed on epithelial surfaces and mediates efflux of drugs/xenobiotics.¹⁹ Multidrug resistantance-1 also regulates neutrophil transmigration across intestinal epithelium, conferring a critical link between intestinal barrier and innate immune system.¹⁹ Loss of MDR1 is associated with increased susceptibility to colitis in both human cohorts and mouse models.^20-22 In the current study, reduced MDR1 levels in anti-CD40 colitis mice could be important contributors to the development of colonic inflammation. Moreover, reports have shown that MDR1 C1236T polymorphisms increase susceptibility to UC.²³

In IBD, besides changes in overall epithelial membrane proteins, various cell types comprising the intestinal epithelium undergo changes with respect to their abundance, activity, and location.²⁴ We show aberrant appearance of Paneth-like cells and hyperproliferation of colonic crypts in anti-CD40 colitis mice. Paneth cells (PCs), localized to the crypt base in the small intestine, play an important role in maintaining gut homeostasis via secretion of an array of antimicrobial peptides and regulating innate mucosal immunity. Impairment in PCs have been shown to determine the inflammatory tone in the intestine and contribute to IBD.^25,26 Although, PCs are absent in rodent colonic epithelium, exacerbation of colitis and dysbiosis, in response to ectopic expression of lysozyme in colon, has also been observed.²⁶ Interestingly, aberrant appearance of PCs in the colon, referred to as Paneth cell metaplasia, has also been observed in gastrointestinal disorders such as colorectal cancer (CRC) and inflammatory bowel disease.^27,28 In our current study, we observed a marked increase in lysozyme-1 (LYZ1) protein expression in the colon of anti-CD40-treated mice. A possible explanation for the appearance of Paneth-like cells in these mice could be an adaptive response to pronounced inflammation in anti-CD40 mouse model of colitis, reflective of Paneth cell plasticity and their ability to acquire stem cell-like features, thereby aiding in recovery of damaged epithelium. Detailed mechanistic studies are warranted to delineate the signaling pathways that could invoke cellular metaplasia in anti-CD40 treated mice.

We also observed colonic crypt hyperplasia in anti-CD40 colitis mice compared with control mice, indicating remodeling of the stem cell compartment. Recent evidence indicates 2 distinct pools of stem cells participating in regeneration and restoration of intestinal epithelium.²⁹ Efficient regeneration of intestinal epithelium in response to acute injury depends on Lgr5⁺ stem cells that serve as drivers of homeostatic constant regeneration of intestinal epithelium.²⁹ These Lgr5⁺ stem cells have been shown as dispensable for repair during colonic inflammation.³⁰ In mice, this stem cell population is lost chemically induced colitis and is dispensable in immune based 2,4,6-tninitrobenzenesulfonic acid (TNBS) model.⁷ In a chronic colitis model (anti-IL10R colitis model), Lgr5⁺ stem cells are required for disease restriction.⁷ In our anti-CD40 colitis model, expression of Lgr5 and Olfm4 markers of homeostatic stem cells did not change compared with control mice. This suggested that in this model, similar to the TNBS model, Lgr5⁺ stem cells are dispensable. Slow proliferating reserve intestinal stem cells, termed as +4 stem cells, have been shown to fulfill regenerative tasks under chronic injury or physiologic stress.²⁹ Analysis of the expression of markers of +4 stem cells in our model revealed significant decline in Lrig1 and Hopx but marked increase in Bmi1 in colon of anti-CD40 mice compared with control mice. Specific stem cell marker positive cells enriched for fetal-like markers have been shown to develop and participate in epithelial regeneration.³¹ Similar fetal-like reversion has been noted in DSS colitis model and helminth infection.^32,33

Fetal-like stem cells aid in repair and regeneration. Our data also showed a significant upregulation of fetal-like stem cell markers Sca1 and Annexin-A1 (characteristic of regenerative epithelium). Recently, a critical role of Sca1 and activation of YAP/TAZ in intestinal epithelial regeneration has been highlighted in murine colitis models.³⁴ Activation of YAP/TAZ is required for the recovery of colonic tissue injury in the DSS and TNBS colitis models. Activation of YAP via IL-6 (intestinal inflammation route)³⁵ or through the mechano-sensory pathway (extracellular matrix remodeling and FAK/Src signaling route)³³ in DSS colitis models was determined essential for epithelial regeneration. Evidently, inhibition of YAP impairs intestinal regeneration in the DSS colitis model.^36,37 These results were recapitulated in our anti-CD40 colitis mouse model where elevated levels of activated YAP were localized to the nucleus, leading to increased TEAD1 and hyperproliferation of the crypts (marked increase in Ki67 immunofluorescence). The YAP mRNA expression has also been shown to be increased in intestinal epithelial cells of CD patients and mice with colitis.³⁸ Its role in mucosal regeneration has been challenged with increased susceptibility to tumorigenesis^37,38 and intestinal fibrogenesis.³⁹ Thus, activation of FAK/Src signaling and YAP in anti-CD40 model requires further investigation with regard to fibrosis and tumorigenic potential.

In conclusion, we have shown that an innate immune cell–mediated anti-CD40 colitis model exhibits Paneth cell metaplasia and ion transport defects, specifically the decreased expression of chloride bicarbonate exchanger DRA underlying diarrheal phenotype. An induction in the fetal-like signature via upregulation of mechanosensory pathway and altered stem cell marker levels suggests that this is a unique model to study the role of innate immune cells in mucosal regeneration and IBD pathogenesis.

Funding

These studies were supported by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Biomedical Laboratory Research and Development: Merit Review Award, BX002011 (P.K.D.), VA BCCMA Award I01BX005862 (P.K.D.), BX002867 (S.S.), VA CDA2 Award BX004719 (A.K.) and VA Senior Research Career Scientist Award: 1IK6BX005242 (P.K.D.). The studies were also supported by NIH/NIDDK grants, R01 DK 54016 and 2R56DK092441 (P.K.D.), and 111671 (D.O-V.).

Conflicts of Interest

Authors have no conflict of interest statement to declare.