Human umbilical cord mesenchymal stem cells alleviate colon epithelial apoptosis in ulcerative colitis by activating inositol-requiring enzyme 1/X-box binding protein 1 signaling

Abstract

Background

Ulcerative colitis (UC), a chronic inflammatory gastrointestinal disease, is characterized by disrupted intestinal barrier integrity and unresolved endoplasmic reticulum (ER) stress, which drives epithelial apoptosis and disease progression. While mesenchymal stem cells (MSCs), particularly human umbilical cord–derived MSCs (hUC-MSCs), have shown therapeutic potential in UC, their mechanisms in modulating ER stress remain unclear. This study aimed to investigate the role of hUC-MSCs in alleviating ER stress–induced epithelial damage and elucidate the underlying molecular pathways in a murine colitis model and in vitro systems.

Results

Intraperitoneal administration of hUC-MSCs significantly attenuated dextran sulfate sodium (DSS)-induced colitis in mice. Histological analysis revealed restored crypt architecture and reduced epithelial apoptosis. Transcriptomic profiling demonstrated that hUC-MSCs reduced differentially expressed genes in inflammatory bowel disease–related and ER stress response pathways in colon tissues. Mechanistically, hUC-MSCs activated the IRE1/XBP1 axis, increasing Xbp1 splicing and suppressing pro-apoptotic Bcl2l11 expression. In vitro, hUC-MSC-conditioned medium protected colon epithelial cells from TNF-α-induced apoptosis via IRE1/XBP1 activation, an effect abolished by the IRE1 inhibitor 4μ8C.

Conclusions

Our findings demonstrate that hUC-MSCs alleviate UC by mitigating ER stress through IRE1-mediated Xbp1 splicing, thereby reducing epithelial apoptosis and promoting mucosal repair. This study provides a mechanistic foundation for MSC-based therapies targeting ER stress in inflammatory bowel diseases.

Graphical abstract

Graphical Abstract.

Significance Statement.

Ulcerative colitis, a debilitating inflammatory bowel disease, remains challenging to treat because of persistent epithelial damage and unresolved inflammation. This study reveals that the treatment with human umbilical cord–derived mesenchymal stem cells (hUC-MSCs) enhances IRE1 phosphorylation and Xbp1 mRNA splicing. It alleviates the apoptosis of colonic epithelial cells caused by endoplasmic reticulum stress and promotes tissue repair. By demonstrating efficacy in preclinical models, this approach links mechanistic insights with translational potential, paving the way for MSC-based therapies to meet the unmet clinical needs in ulcerative colitis and other inflammatory diseases.

Introduction

Ulcerative colitis (UC) is a chronic inflammatory bowel disease characterized by persistent inflammation and damage to the colonic mucosa, significantly impacting patients’ quality of life.¹ One of the hallmarks of UC pathogenesis is the disruption and damage to the intestinal epithelium, which serves as a critical barrier to maintain gut homeostasis and protect against pathogens and inflammatory insults.^2,3 When this epithelial barrier is compromised, it leads to increased intestinal permeability, further driving inflammation and disease progression.^4,5

Current research indicates a close relationship between endoplasmic reticulum (ER) stress and the pathogenesis of various intestinal disorders, including UC.^6,7 ER stress, which arises from the accumulation of misfolded proteins in the ER, triggers the unfolded protein response (UPR) to restore homeostasis.⁸ However, prolonged or unresolved ER stress can lead to cellular apoptosis via CHOP-dependent or independent induction of Bcl2-like11 (Bcl2L11), NADPH oxidase activator (NOXA), and p53 unregulated apoptosis modulator (PUMA),^9,10 contributing to the exacerbation of inflammatory responses and tissue damage in UC.⁷ In particular, the inositol-requiring enzyme 1 (IRE1)-X-box binding protein 1 (XBP1) axis of the UPR plays a pivotal role in mitigating ER stress by promoting XBP1 mRNA splicing.^11,12 This spliced form of XBP1 enhances the production of ER chaperones and other proteins that help alleviate ER stress, thus protecting epithelial cells from stress-induced apoptosis.¹³ Genetic variants in XBP1, which lead to spontaneous enteritis, were identified to have link to increased risk of both Crohn’s disease and UC.¹³

Recent studies have highlighted the promise of mesenchymal stem cells (MSCs), particularly those derived from umbilical cord tissues, in modulating inflammation and promoting tissue repair.^14-16 In preclinical and clinical studies, MSCs and their derived exosomes have demonstrated beneficial effects in treating UC, primarily by reducing inflammation and enhancing mucosal healing.^17,18 However, the specific impact of MSC therapy on ER stress in UC has not yet been fully elucidated.

Here, we hypothesize that human umbilical cord–derived MSCs (hUC-MSCs) alleviate UC by targeting the IRE1/XBP1 axis to resolve ER stress and inhibit epithelial apoptosis. Using a DSS-induced murine colitis model and TNF-α-stimulated colon epithelial cells, we investigate whether hUC-MSCs restore ER homeostasis through XBP1 splicing, thereby bridging the gap between MSC therapy and ER stress modulation in UC. This study not only elucidates a novel mechanistic pathway but also highlights the translational potential of hUC-MSCs for clinical UC management.

Methods

Culture of hUC-MSCs

hUC-MSCs were purchased from Cyagen Bio sciences, Inc. MSCs were cultured in Dulbecco’s Modified Eagle Medium (DMEM)/F12 (HyClone, UT, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, NY, USA) and 0.5% penicillin–streptomycin (P/S) at 37 °C in a 5% CO₂ atmosphere. To prepare hUC-MSC conditioned medium (hUC-MSC-CM), the culture medium was replaced with a serum-free medium when the cell fusion rate reached about 80%, and supernatant was filtered using filters 0.22 μm and stored in −70 °C.

Induction of experimental colitis model and administration of hUC-MSCs

Female-specific pathogen-free (SPF) C57BL/6 mice (6–8 weeks old) were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd (Beijing, China) and housed in the Animal Facility at Shanxi Medical University. The experimental mice were randomly divided into three groups: DSS+PBS, DSS+MSCs, and control groups (n = 6 per group for survival analysis and histological examination, n = 3 per group for RNA sequencing). Starting from day 0, mice in the DSS + PBS and DSS + MSCs groups were administered with 4% DSS (molecular weight 36–50 kDa; MP Biomedicals, CA, USA) in drinking water for 7 days, followed by regular water. Mice in the control group received regular water throughout the experiment. For therapeutic studies, mice in DSS + MSCs group were injected with uc-MSCs (1.0 × 10⁶ cells per mouse, 100 μL) intraperitoneally on days 3, 5, and 7, while mice in DSS + PBS and the control groups received 100 μL of PBS via same route and on the same schedule. For survival analysis, body weight, disease activity index (DAI), and survival rate were monitored daily over a period of 14 days. For RNA-sequencing and histological examination, mice were euthanized on day 8. After euthanasia, colons were excised, rinsed with PBS, and measured for length. Portions of colon were used for myeloperoxidase (MPO) activity assay and RNA isolation. Tissue for histological examination was stored in 10% formalin.

MPO assay

Excised colon parts were homogenized and the MPO activity was measured in the supernatant using an MPO assay kit (NJJCBIO, Nanjing, China) according to the manufacturer’s instructions.

Colon histological inflammatory scores

Formalin-fixed colon tissues were embedded in paraffin routinely. Paraffin sections, 5 μm thick, were stained with hematoxylin and eosin (H&E). Inflammatory scores were determined using a score system¹⁹ as follows: Epithelium (E) score: 0, normal morphology; 1, loss of goblet cells; 2, loss of goblet cells in large areas; 3, loss of crypts; 4, loss of crypts in large areas. Infiltration (I) score: 0, no infiltration; 1, infiltrate around crypt basis; 2, infiltrate reaching to lamina muscularis mucosa layer; 3, extensive infiltration reaching the muscularis mucosa with abundant edema; and 4, infiltration of the submucosa layer. The histological score was calculated as the sum of these two parameters (total score = E + I).

Immunohistochemical staining for cleaved caspase-3 and Ki67

Formalin fixed colon tissues were embedded in paraffin routinely. Paraffin sections were incubated with antibodies to cleaved caspase-3 (1:500; Servicebio, Wuhan, China) or Ki-67 (1:200; Servicebio, Wuhan, China) overnight at 4 °C. Goat-anti-rabbit IgG was used as a secondary antibody and diaminobenzidine (DAB) staining was performed. The number of cleaved caspase-3 or Ki-67 positive intestinal epithelial cells was determined in 20 random visual fields (200×). The percentage of stained/total hepatocyte nuclei was designated as the Ki-67 labelling index.

In vivo tracking of hUC-MSCs

hUC-MSCs were labelled with a Dil fluorescent probe (Beyotime, Shanghai, China) according to the manufacturer’s instructions, and the labelling efficiency was assessed by flow cytometry. Mice were fed with 4% DSS, and labeled uc-MSCs were injected intraperitoneally on day 3. To examine whether Dil-labeled hUC-MSCs migrated to colon epithelial region, cryosections of colons tissue were stained with DAPI to mark nuclear. The colonic tissue sections were imaged using a Nikon fluorescence microscope.

Quantitative real-time PCR (qRT-PCR)

Excised colon parts were homogenized to extract total RNA using the TRIzol (TransGen Biotech, Beijing, China). To acquire cDNA samples, reverse transcription was conducted using PrimeScript reverse transcriptase (Sparkjade, Jinan, China). The qRT-PCR was carried out on an ABI Prism 7300 using the TB Green Premix Ex Taq Kit (Sparkjade, Jinan, China). The 2^△△Ct method was used to compare relative amounts of target mRNA to GAPDH. All primer sequences were listed in Table 1.

Table 1.

List of primer sequences.

Gene	Forward 5ʹ-3ʹ	Reverse 5ʹ-3ʹ
Gapdh	AGGTCGGTGTGAACGGATTTG	TGTAGACCATGTAGTTGAGGTCA
Tnf	TCTTCTCATTCCTGCTTGTGG	CACCCCGAAGTTCAGTAGACA
Nupr1	CCCTTCCCAGCAACCTCTAAA	TCTTGGTCCGACCTTTCCGA
Casp12	GAGCCACCCCTTCTGAGTGTT	AAGCAAGGATGTGGGGAACCA
Ifng	ATCTGGAGGAACTGGCAAAA	TTCAAGACTTCAAAGAGTCTGAGG
Agr2	GGAGCCAAAAAGGACCCAAAG	CTGTTGCTTGTCTTGGATCTGT
Xbp1-U	AAGAACACGCTTGGGAATGG	ACTCCCCTTGGCCTCCAC
Xbp1-S	GAGTCCGCAGCAGGTG	GTGTCAGAGTCCATGGGA
Bcl2l11	GACAGAACCGCAAGACAGGA	CTGCAGCCTTCCTTGGTGTA

Western blotting analysis

For total protein extraction of colon tissues, colon tissues were washed thoroughly in cold PBS and contents were flushed away. Then, the colon tissues were homogenized with RIPA lysis buffer (containing PMSF and phosphatase inhibitors, Solarbio, Beijing, China) and incubated for 30 min. For total protein extraction of cultured cells, extraction of cytoplasmic protein, and nuclear protein from colon tissue and cell samples, cells were lysed by RIPA buffer (containing PMSF and phosphatase inhibitors) for 30 min. Then, the lysate was centrifuged at 13 500 g at 4 °C for 15 min, and the supernatant was collected. The BCA quantitative kit (KTD3001, Abbkine, Wuhan, USA) was used to determine protein concentration. An equal amount of protein (30 μg) was isolated on SDS-PAGE gel (Epizyme, Shanghai, China). Subsequently, the protein was transferred to the PVDF membrane (Merk, Darmstadt, USA) and then blocked using 5% skim milk for 2 h under room temperature before incubating with the primary antibody. Then, the second antibody conjugated with IgG-HRP was incubated for 1 h and then developed by SuperKine™ West Femto Maximum Sensitivity Substrate (Abbkine, Wuhan, China) using ChemiDOC™ XBS imaging systems (Bio-Rad, CA, USA). Anti-β-Actine and Xbp1 antibodies were from Abcam (Cambridge, UK), anti-IRE1 antibody was from Abmart (Shanghai, China), and anti-BCI2L11 antibody was from Abways (Beijing, China).

Culture of colon epithelial cells

Immortalized mouse colon epithelial cells (MCECs）(IMMOCELL, Guangzhou, China) were cultured in DMEM (Hyclone, UT, USA) supplemented with 10% FBS and 0.5% P/S for routine cultures and in DMEM with 0.5% serum for cell function experiments. Cells were maintained in a humidified atmosphere of 5% CO₂ at 37 °C. The culture medium was refreshed every 2–3 days, and cells were passaged upon reaching approximately 80% confluence.

In vitro cell function assay

TNF-α (Yisheng, Shanghai, China) was applied to simulate inflammation-induced intestinal epithelial damage, as observed in UC. 4μ8C (Selleck, TX, USA) was used as selective inhibitor of IRE1α-induced XBP1 mRNA cleavage. Apoptosis was evaluated by Annexin V and PI staining (Apexbio, TX, USA), and cells were analyzed using a flow cytometer (Aria III, BD Biosciences).

Transcriptome sequencing and data analysis

Colon segments were transported on dry ice to Biomarker Technology Co. (Beijing, China), where total RNA was isolated and RNA quality control was conducted. After passing the quality inspection, cDNA library preparation and transcriptome sequencing were conducted. All nine libraries were sequenced using the Illumina NovaSeq 5000 platform (San Diego, CA, USA). The raw sequencing data were analyzed with FastQC using Cutadapt to remove joints and Trimmomatic²⁰ to remove low-quality bases and reads at both ends. The clean reads were mapped to the Mus musculus reference genome assembly GRCm38 using Hisat2.²¹ The number of fragments that overlap each Entrez gene was summarized using featureCounts and differential expression analysis between the DSS + PBS group and DSS + MSC group, and the control group was performed using the limma software package.²² A P-value cutoff < 0.05 with an absolute |log2FC| >1 was used to determine differential expression genes (DEGs). Volcano and PCA plot were generated with the R package ggplot2. Heatmaps were generated with the R package “pheatmap.”

xCell analysis

Estimation of macrophages, monocytes, neutrophils DC, and epithelial cells fractions from colon tissues of UC patients was determined through gene expression analysis using xCell²³ (https://xcell.ucsf.edu/). Statistical analysis was performed with GraphPad Prism 8.0 software (GraphPad Software). Results were considered significant when P value was <0.05.

Screening for response to ER stress-related DEGs

The DEGs induced by certain treatment were intersected with “Response to endoplasmic reticulum stress’” (RERS) gene set from Molecular Signatures Database (MSigDB) to obtain RERS related DEGs, and the result was visualized using the R package “VennDiagram.”

Functional enrichment analysis and protein–protein interaction analysis of RERS-related DEGs

The gene ontology (GO) enrichment analyses were conducted for gene intersections using the R package “clusterProfiler.”²⁴ STRING online database (http://string-db.org) was used for analyzing the protein–protein interaction (PPI) of RERS-related DEGs.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 8.0 (GraphPad Software). Data were expressed as mean ± standard deviation (SD) format with at least three replicates in each group. Analysis of variance for repeated measurements (RM-ANOVA) was used to compare body weights and DAI. One-way ANOVA analysis was used when comparing three or more groups. A P-value < 0.05 was considered statistically significant.

Results

Intraperitoneal hUC-MSCs treatment alleviated the DSS-induced colitis in mice model

To investigate the therapeutic efficacy of intraperitoneal injection of hUC-MSCs against ulcerative colitis, we established a preclinical model by administering a solution of 4% DSS in drinking water to mice for 7 days (Figure 1A). At the end of this period, mice are taken back to normal water. DSS-induced colitis mice were intraperitoneally injected with 1.0 × 10⁶ hUC-MSCs on day 3, day 5, and day 7 (Figure 1A), and the disease activity, body weight, and survival rate of the mice were monitored.

Figure 1.

Intraperitoneal injection of MSCs alleviated DSS-induced colitis in mice. (A) Schematic overview of experimental design. (B, C, and D) Therapeutic efficacy evaluation of acute DSS model for 14 days: (B) disease activity index, (C) body weight change, and (D) probability of survival. (E and F) Colon appearance (E) and length (F) on day 8. (G) MPO concentrations in colons on day 8. (H) Relative TNF-α gene expression levels in colons on day 8. Data are expressed as mean ± SD. **P < .01, ***P < .001, and ****P < .0001.

Upon DSS administration, regardless of hUC-MSC injection, DAI values began to increase, peaking on day 7 or 8, followed by a gradual decline. However, after the recovery of normal water, the DAI decreased more rapidly in the presence of hUC-MSCs (Figure 1B). The body weight of mice sharply declined from day 4 to day 9 post-DSS administration, but recovery was faster in the hUC-MSC-treated group after day 9 (Figure 1C). Consistent with these findings, the survival rate was significantly improved in the presence of hUC-MSCs (Figure 1D).

On day 8 after DSS treatment, colon length was measured before histological examination. Shortening of the colon, which is the characteristic pathological feature of colitis, was substantially inhibited by hUC-MSCs treatment (Figure 1E and F). MPO level, a commonly used biomarker of colitis, is a neutrophil granule constituent, where its level is proportional to the number of migrated neutrophils. In the MSCs treated group, the MPO activity in colon tissues decreased by 45% compared to the PBS–treated mice, implying that hUC-MSCs treatment suppressed the infiltration of neutrophils (Figure 1G). The expression of TNF-α, the major cytokine involved in ulcerative colitis, was markedly reduced in the MSCs treated mice (Figure 1H).

Collectively, MSCs showed preventive and therapeutic effects in the acute colitis models.

hUC-MSC treatment attenuated DSS-induced colon epithelial injury

In the DSS-induced colitis model, administration of DSS disrupts the intestinal epithelial barrier. Histological analysis revealed loss or shortening of crypts and marked infiltration of inflammatory cells in both the lamina propria and muscularis mucosa (Figure 2A), accompanied by significantly increased histopathological scores (Figure 2B). Previous studies have suggested that DSS exerts direct cytotoxic effects on intestinal epithelial cells in the basal crypts, leading to increased epithelial apoptosis. Consistent with this, immunohistochemical analysis showed that DSS treatment increased the number of active caspase (Casp)-3-positive cells in colonic tissues (Figure 2C and D) and reduced the number of Ki67-positive cells (Figure 2E and F), indicating enhanced epithelial cell apoptosis and impaired proliferation, respectively.

Figure 2.

hUC-MSC treatment attenuated DSS-induced colon epithelial injury. (A) Representative H/E-staining images of colon tissues were shown. (B) Histopathological scores of colons. (C and D) Representative images of immunohistochemistry with cleaved Caspase3 (C) and quantitation of the cleaved Caspase3-positive area (D). (E and F) Representative images of immunohistochemistry with Ki67 (E) and quantitation of the Ki67-positive area (F). (G) Frozen sections of colon from mice with colitis 24 h after Dil-labeled UCMSCs injection. Data are expressed as mean ± SD (n = 6 mice per group). ****P < .0001.

Compared to the DSS+PBS group, colonic tissues in the DSS+MSC group exhibited more intact crypts, reduced infiltration of inflammatory cells (Figure 2A), and significantly lower histopathological scores (Figure 2B). Immunohistochemical analysis further demonstrated that hUC-MSC treatment markedly reduced epithelial apoptosis (Figure 2C and D). Moreover, hUC-MSC treatment significantly restored Ki67 expression levels, indicating a reversal of colon epithelial cells loss (Figure 2E and F).

To determine the specific localization of transplanted MSCs, Dil-labeled hUC-MSCs were intraperitoneally injected into DSS-induced colitis mice. The labeled hUC-MSCs were detected in the inflamed colonic tissues 24 h post-injection. Interestingly, hUC-MSCs were observed in both the epithelium and lamina propria of the inflamed colon, with a higher accumulation in the epithelium (Figure 2G).

These findings indicate that hUC-MSC treatment reduces epithelial injury in DSS-induced colitis by alleviating apoptosis, promoting colon epithelial cells regeneration, and preferentially localizing to and accumulating in the colonic epithelium.

hUC-MSCs treatment reduced colon tissue gene expression profile change in DSS-induced colitis

We next investigated the genetic regulatory network regulated by hUC-MSC in colitis. Total RNA was extracted from colons of three NC group mice, three DSS+PBS group mice, and three DSS+MSC group mice. RNA sequencing data demonstrated that 1491 upregulated and 940 downregulated differentially expressed genes (DEGs, P-adjust <.05, absolute log 2 [fold change] > 1) induced by DSS treatment, respectively (Figure 3A and D). However, much less genes were regulated (558 genes were upregulated and 485 genes were downregulated) with hUC-MSC treatment with the same criteria compared to untreated (Figure 3B and D). Further analyses found that 526 and 872 genes were upregulated and downregulated in DSS-MSC group when compared to DSS-PBS group mice, respectively (Figure 3C and D). Principal component analysis also showed that the mRNA expression profile in the colon of DSS-MSC group mice was separated from DSS-PBS group mice but closed to NC group mice (Figure 3E). Hierarchical clustering analysis also indicated distinguishable transcription patterns between the DSS+PBS and DSS+MSC groups (Figure 3F).

Figure 3.

RNA-sequencing analysis of colon tissues in DSS-induced colitis and the effects of hUC-MSC treatment. Transcriptomic analysis was performed to investigate differential gene expression at the transcriptional level in colon tissues from NC, DSS+PBS, and DSS+MSC groups. (A, B, C) Volcano plots showing the distribution of differentially expressed genes (DEGs). (P-adjust < .05, |log₂ fold change| > 1). (A) DEGs in the DSS+PBS group compared to the NC group. (B) DEGs in the DSS+MSC group compared to the NC group. (C) DEGs in the DSS+MSC group compared to the DSS+PBS group. (D) Bar chart summarizing the number of upregulated and downregulated DEGs across comparisons. (E) Principal component analysis (PCA) showing the clustering of mRNA expression profiles in colon tissues. (F) Heatmap showing hierarchical clustering of DEGs across samples, highlighting distinct transcriptional patterns between the DSS+PBS and DSS+MSC groups. (G) Venn diagram comparing DEGs with the “Inflammatory Bowel Disease” gene set from the GeneCards database. (H) The enrichment of innate immune cells and epithelial cells in the colon using the xCell algorithm. Data are expressed as mean ± SD (n = 3 mice per group). *P < .05, **P < .01, and ***P < .001.

Various chemokines and their receptors (Cxcl11, Cxcl13, and Ccr1), cytokines and their receptors (Csf3, IL1r2, Il6ra, and Il7r), as well as inflammation-related genes (Saa3, Tlr5, S100a8, and S100a9) were upregulated upon DSS induction (Figure 3A) but downregulated with MSC treatment (Figure 3C). Of note, Nlrp10 (a multifunctional negative regulator of inflammation and apoptosis) and Cldn4 (an intestinal barrier-related gene) were detected to be downregulated upon DSS induction (Figure 3A), but were upregulated in hUC-MSC treated mice as compared to the DSS + PBS group (Figure 3C). We further compared the DEGs induced by DSS with the “Inflammatory bowel disease” gene set in the GeneCards database, and 832 genes overlapped (Figure 3G). However, only 343 genes overlapped the “Inflammatory bowel disease” gene set among the DEGs between DSS + MSC group and NC group (Figure 3G). These RNA-seq data uncover that hUC-MSC treatment regulates the pro-inflammatory transcriptional profile in DSS-induced colitis.

We estimated the enrichment of innate immune cells and epithelial cells in the colon using the xCell algorithm applied to the RNA-seq data. The xCell enrichment scores, which estimate the proportion of cell types for each sample, indicated that the levels of macrophages, monocytes, neutrophils, and dendritic cells (DC) were significantly higher in the colons of DSS + PBS group mice compared to NC group. hUC-MSC treatment reduced the enrichment of these innate immune cells in the colon to varying degrees; however, the differences were not statistically significant (Figure 3H). In contrast, intestinal epithelial cells (IEC) were almost completely depleted by DSS induction. Remarkably, hUC-MSC treatment significantly restored IEC enrichment (Figure 3H).

hUC-MSC treatment suppressed pathological ER stress induced by DSS

The induction of ER stress is a hallmark of colitis and contribute to the progression of inflammation by triggering IEC death. To investigate the involvement of ER stress in hUC-MSC treatment for UC, we analyzed the overlap between the “Response to endoplasmic reticulum stress (RERS)” gene set and DEGs induced by DSS + PBS or DSS + MSC. Specifically, compared to healthy controls, DSS induction significantly altered the expression of 29 RERS genes, 21 of which were restored to near-normal levels following MSC treatment (Figure 4A and Figure S1, see online supplementary material for a color version of this figure). Functional annotations of these DSS induced RERS-related DEGs via Gene Ontology (GO) analyses revealed enrichment in biological processes such as response to ER stress, response to topologically incorrect protein, response to unfolded protein, autophagy, intrinsic apoptotic signaling pathway in response to ER stress, regulation of inflammatory response, positive regulation of apoptotic signaling pathway, and IRE1-mediated unfolded protein response (Figure 4B). The PPI network of the DSS-induced RERS-related DEGs was shown in Figure 4C.

Figure 4.

hUC-MSC treatment suppressed pathological ER stress induced by DSS. (A) Venn diagram comparing DEGs with the RERS gene set from the GeneCards database. (B) GO terms enrichment of RERS-related DEGs. (C) The PPI network of the DSS-induced RERS-related DEGs. (D to H) Relative Caspase12 (D), Nupr1 (E), IFN-γ (F), Agr2 (G), and Bcl2L1 (H) gene expression levels in colons on day 8 of group NC, DSS + PBS, and DSS + MSC. (I and J) Western blot analysis of BCL2L1 protein expression in colons on day 8 across the three groups. Representative western blotting results (I) and relative grayscale (J). (K and L) Relative mRNA abundance of Xbp1-U (K) and Xbp1-S (L) in colons on day 8 of group NC, DSS + PBS, and DSS + MSC. (M to O) Western blot analysis of p-IRE1 and XBP1-S protein expression in colons on day 8 across the three groups. Representative western blotting results (M) and relative grayscale of p-IRE1 (N) and XBP1-S (O). Data are presented as mean ± SD (n = 6). *P < .05, **P < .01, ***P < .001, and ****P < .0001.

The expression levels of the key RERS genes were further validated in colon tissues from DSS + PBS and DSS + MSC group mice (Figure 4D–H). Casp12, a caspase family member specifically activated by ER stress, was upregulated in the DSS + PBS group but returned to baseline levels with MSC treatment (Figure 4D). Similarly, Nupr1, an ER stress-inducible transcription factor,^25,26 was significantly upregulated in the DSS + PBS group but was reversed in the DSS + MSC group. The expression of interferon-γ (IFN-γ), a key cytokine involved in IBD, was markedly reduced in the hUC-MSC treated mice compared to only DSS-induced mice (Figure 4F). In both human and mice, anterior gradient 2 (Agr2) deficiency is associated with mucus barrier defect,²⁷ impaired ability to mitigate ER stress, and infantile-onset IBD.²⁸ Consistent with these findings, Agr2 expression was downregulated in DSS-induced colitis model, whereas MSC treatment restored its expression (Figure 4G). Bcl2l1, a key mediator of ER stress–induced apoptosis,¹⁰ was upregulated at both mRNA and protein levels in the DSS + PBS group but was restored to baseline levels following MSC treatment (Figure 4H–J).

During ER stress, IRE1 activation induces splicing of Xbp1 mRNA, generating spliced Xbp1 (Xbp1-S), which plays a critical role in damage repair. Compared with DSS-induced colitis model group, hUC-MSC treatment increased IRE1 Ser-724 phosphorylation levels (Figure 4M and N), enhanced Xbp1-S mRNA abundance (Figure 4L), and Xbp1-S protein expression (Figure 4M and N), while reducing unspliced Xbp1(Xbp1-U)mRNA levels (Figure 4L). These findings suggest that MSC transplantation mitigates ER stress-induced intestinal injury by promoting Xbp1 splicing.

hUC-MSC inhibited TNF-α-induced ER stress and intestinal epithelial cell apoptosis via paracrine signaling

The predominant colitogenic cytokine, TNF-α, has been reported to serve as a physiological ER stress inducer in IECs.²⁹ We analyzed the overlap between the “RERS” gene set and DEGs in IECs from TNF-α treated mice (GSE268933). Specifically, TNF-α induction significantly altered the expression of 20 RERS genes compared to healthy controls, including Bcl2l11, Nupr1, and Cebpb, which were also upregulated in DSS-induced colitis in mice (Figure 5A–C).

Figure 5.

TNF-α-induced ER stress in intestinal epithelial cell (A) Venn diagram comparing TNF-α-induced DEGs with the RERS gene set from the GeneCards database. (B) Heatmap showing the TNF-α-induced RERS-related DEGs. (C) The PPI network of the TNF-α-induced RERS-related DEGs.

Recent studies have increasingly highlighted that the therapeutic efficacy of MSC transplantation is primarily mediated through paracrine signaling.³⁰ To further explore this mechanism, we investigated the protective effects of MSC-conditioned medium (MSC-CM) on TNF-α-induced intestinal epithelial cell damage in vitro. Apoptosis analysis revealed that MSC-CM effectively suppressed TNF-α-induced cell apoptosis (Figure 6A and B). qPCR and Western blotting analyses showed MSC-CM inhibited TNF-α-induced BCL2L11 expression (Figure E, F, and I). Mechanically, both IRE1 phosphorylation (Figure 6F and G) and Xbp1 mRNA splicing (Figure 6C, D, F, and H) were promoted by MSC-CM treatment. However, the protective effects of MSC-CM were abrogated upon addition of the Xbp1 splicing-specific inhibitor 4μ8C (Figure 6A–I).

Figure 6.

MSC-CM protected against TNF-α-induced intestinal epithelial cell apoptosis by suppressing pathological ER stress apoptosis of MCECs was induced by treating cells with 100 ng/mL TNF-α in the culture medium. (A and B) Flow cytometry analysis of Annexin V and PI staining revealed that MSC-CM treatment significantly reduced the percentage of Annexin V+ apoptotic cells, while the protective effect of MSC-CM was reversed by the Xbp1 splicing inhibitor 4μ8C. (A) Representative flow cytometry plots; (B) quantification of apoptotic cells. (C to E) Relative mRNA levels of Xbp1-U (C), Xbp1-S (D), and Bcl2l11 (E) in MCECs treated with TNF-α alone or in combination with MSC-CM and 4μ8C. (F to G) Western blot analysis of p-IRE1, XBP1-S, and BCL2L11 protein levels in MCECs treated with TNF-α alone or with MSC-CM and 4μ8C. (F) Representative western blotting results; quantification of relative grayscale of p-IRE1 (G) and XBP1-S (H) and BCL2L11 (I). Data are presented as mean ± SD (n = 6). *P < .05, **P < .01, ***P < .001, and ****P < .0001.

These findings suggest that MSC-CM exerts its protective effects on TNF-α-induced intestinal epithelial cell apoptosis by promoting Xbp1 mRNA splicing, thereby alleviating ER stress and reducing cell death.

Discussion

MSCs derived from various tissue origins³¹ or administered through different transplantation routes³² have exhibited distinct therapeutic potentials in different disease models. This study revealed the therapeutic efficacy of hUC-MSC, delivered via intraperitoneal injection, in mitigating epithelial injury and inflammation in a DSS-induced colitis model. hUC-MSCs’ treatment significantly alleviated the disruption of the intestinal epithelial barrier, restored crypt structure, and reduced inflammatory cell infiltration, as evidenced by histological and immunohistochemical analyses.

Previous studies have shown the potentials of MSCs and MSC-derived products in the treatment of UC. The therapeutic mechanisms include immunomodulation, promotion of intestinal epithelial cell regeneration, and normalization of the intestinal microbiome.¹⁸ In this study, hUC-MSCs were demonstrated to modulate ER stress pathways by promoting Xbp1 mRNA splicing, thereby reducing the expression of key pro-apoptotic genes such as Bcl2l11(Bim).³³ These effects were observed both in DSS-induced UC mouse model and in vitro TNF-α-induced ER stress and apoptosis in cultured intestinal epithelial cells. Collectively, these findings provide novel insights into the mechanisms by which hUC-MSCs and their secreted factors alleviate ER stress–induced epithelial injury and inflammation, offering a promising therapeutic strategy for UC.

The XBP1 pathway is critical for maintaining gut barrier function and immune balance. The pathogenesis of UC is thought to result from a complex interplay of genetic, environmental, microbial, and immune factors.³⁴ Genetic variants in XBP1 increase the risk of UC,¹³ while certain microbial or dietary compounds—such as trierixin from Streptomyces—can suppress its normal activity in epithelial cells.³⁵ Animal studies show that loss of Xbp1 in intestinal epithelial cells triggers spontaneous inflammation and worsens colitis.¹³ Additionally, the IRE1/XBP1 pathway contributes to intestinal homeostasis by augmenting cytokine production in group 3 innate lymphoid cells (ILC3s), a key cell type for mucosal defense; disrupting this pathway in ILC3s reduces the expression of protective cytokines like Il22, leading to increased vulnerability to infections and colitis.³⁶

Earlier network analyses found that XBP1 levels are increased in UC and correlate with more severe disease and higher macrophage infiltration, but with fewer CD8⁺ T cells.³⁷ However, these studies did not account for differences between the spliced (active) and unspliced forms Xbp1. Because these forms have distinct roles in epithelial versus immune cells, overlooking this distinction may explain the inconsistencies among previous studies.

In this study, intraperitoneal injection of hUC-MSCs significantly increased Xbp1 mRNA splicing in intestinal epithelial cells, which may play a critical protective role against UC pathogenesis. This finding demonstrates the importance of Xbp1 splicing levels and cell type–specific functions in UC and reveals a novel mechanism for MSC therapy in treatment of colitis.

In active UC, the rate of apoptosis of epithelial cells is significantly elevated. DSS results in the death of epithelial cells, leading to the disruption of the epithelial monolayer lining. This disruption allows the entry of luminal bacteria and associated antigens into the mucosa and causes an altered mucosal immune response. Mucosal healing has become a key therapeutic goal for treatment of patients with UC suggested by its strong association with improved clinical outcomes and long-term disease remission.^2,3 A recent Immunity study further demonstrated that XBP1 safeguards the intestinal mucus barrier and suppresses necroptosis-induced colitis, underscoring its essential role in maintaining epithelial homeostasis and barrier function.³⁸ Together with our findings, these data highlight the multifaceted protective effects of XBP1 signaling in colitis. Compared to conventional anti-inflammatory therapies, targeting downstream inflammatory mediators, hUC-MSCs offer a more effective therapeutic strategy by addressing both epithelial damage and immune dysregulation, a promising treatment for UC.

Limitations

This study also has limitations. The DSS-induced colitis model, although widely used, does not fully recapitulate the chronic relapsing nature and multifactorial pathogenesis of human UC. Furthermore, our experiments were limited to preclinical models, and the therapeutic efficacy of hUC-MSCs in human UC patients requires careful evaluation in clinical settings. These caveats should be acknowledged when interpreting our findings. From a translational perspective, while MSC-based therapies hold promise, several challenges remain before clinical application. Critical considerations include ensuring the long-term safety of MSC administration, determining optimal dosing regimens, and assessing the durability of therapeutic benefits. Potential risks such as immunogenicity, tumorigenicity, and unintended differentiation must be addressed in future studies. Additionally, regulatory and manufacturing challenges, including standardization of MSC isolation and expansion protocols, will influence the feasibility of clinical translation. Addressing these issues will be essential for advancing hUC-MSC therapy from experimental models to clinical practice.

Conclusions

Our current study revealed that intraperitoneally injection of hUC-MSC alleviated DSS-induced colitis in mice. In vivo tracking experiment demonstrated that injected MSCs can homing to both the epithelium and lamina propria of the inflamed colon. RNA-sequencing and subsequent qPCR and western blotting analysis confirmed that hUC-MSCs confer protection in UC through the modulation of ER stress and promotion of Xbp1 splicing in colon epithelial cells. Our study has strongly suggested the clinical potential of hUC-MSCs in management of UC. It also suggested targeting ER stress in developing future therapeutic strategies for UC.

Author contributions

Taoran Zhao (Formal analysis [lead], Funding acquisition [lead], Investigation [lead], Software [lead], Visualization [lead], Writing—original draft [lead]), Wenyi Hou (Data curation [lead], Formal analysis [lead], Visualization [lead]), Mengwei Wang (Data curation [supporting], Formal analysis [supporting], Visualization [supporting], Writing—original draft [supporting]), Zhenyu Feng (Validation [equal], Writing—original draft [equal], Writing—review & editing [equal]), Xiaoni Feng (Data curation [supporting], Formal analysis [supporting], Writing—original draft [supporting]), Heng Wang (Data curation [supporting], Visualization [supporting]), Hong Zhao (Project administration [equal], Resources [equal]), Xiujuan Li (Resources [supporting]), Shulin Hou (Resources [equal], Writing—original draft [supporting]), Guoping Zheng (Conceptualization [equal], Methodology [equal], Project administration [equal], Writing—review & editing [equal]), Xiaozheng Zhang (Conceptualization [equal], Methodology [equal], Project administration [equal], Writing—review & editing [equal]), and Jun Xie (Conceptualization [lead], Methodology [equal], Project administration [equal])

Supplementary material

Supplementary material is available at Stem Cells Translational Medicine online.

Funding

This study was supported by the Natural Science Foundation of Shanxi Province (No. 20210302123315, No.202303021221128), Shanxi Key Laboratory Open Project (No. CPSY 202204), and PhD Start-up Foundation of Shanxi Medical University (No. XD1912).

Conflict of interest

None declared.

Data Availability

RNA-sequencing data will be made publicly available through Genome Sequence Archive (GSA) (CRA022678) upon publication. All other datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval

All animal experiments were approved by and conducted in accordance with the guidelines of the Ethics Committee of Animal Care and Use Committee of Shanxi Medical University (NO.DWYJ-2024-163).