Protective effect of hUC-MSCs regulating peritoneal mesothelial cell EMT through cGAS-STING signaling pathway in high glucose-induced peritoneal fibrosis

Abstract

BACKGROUND

Peritoneal dialysis represents one of the primary alternative therapies for end-stage renal disease. Peritoneal fibrosis (PF) may, however, develop as a consequence of the peritoneum being exposed to non-physiological peritoneal dialysis fluid over extended durations. PF is the main factor that forces end-stage renal disease patients to discontinue long-term peritoneal dialysis. Therefore, how to delay and block the progression of PF has surfaced as an issue demanding timely scholarly attention. Human umbilical cord-derived mesenchymal stem cells (hUC-MSCs) have a reversing effect on the fibrosis levels of various tissues. We hypothesized that hUC-MSCs may afford protection in PF.

AIM

To investigate the modulatory potential of hUC-MSCs across both in vivo and in vitro experimental models of PF.

METHODS

Both the in vivo and in vitro PF models displayed differential expression of cyclic GMP-AMP synthase (cGAS) and stimulator of interferon genes (STING). Next, hUC-MSC-mediated effects on epithelial-mesenchymal transition (EMT) driven by high glucose (HG) in HMrSV5 cells were determined. We also performed a rescue experiment with the STING agonist DMXAA to validate the engagement of the cGAS-STING signaling cascade in this event. The anti-fibrotic potential of hUC-MSCs was lastly investigated in a mouse model of PF.

RESULTS

Heightened levels of cGAS and STING proteins were discernible in both HG-exposed HMrSV5 cells and the mouse PF model. The presence of hUC-MSCs in co-culture exerted a potent suppressive influence on HG-evoked migratory, invasive, and EMT-related molecular changes in HMrSV5 cells. DMXAA-mediated intervention nullified the protective capacity of hUC-MSCs against HG-occasioned EMT in HMrSV5 cells. In addition, hUC-MSC intervention in PF mice dually mitigates peritoneal fibrotic alterations and inflammatory infiltration while lowering the protein expression of cGAS-STING pathway components.

CONCLUSION

The cGAS-STING pathway is activated in PF models both in vivo and in vitro, hUC-MSCs inhibit this pathway to suppress EMT in HMrSV5 cells and retard PF progression in mice.

Key Words: Human umbilical cord mesenchymal stem cells; Peritoneal fibrosis; Cyclic GMP-AMP synthase-stimulator of interferon genes signaling pathway; Epithelial-mesenchymal transition; Inflammation

Core Tip: The cyclic GMP-AMP synthase (cGAS)-stimulator of interferon genes (STING) pathway is activated in peritoneal fibrosis (PF). This study shows that human umbilical cord-derived mesenchymal stem cells (hUC-MSCs) alleviate high glucose-induced epithelial-mesenchymal transition and murine PF. Mechanistically, hUC-MSCs exert protective effects by inhibiting the cGAS-STING axis. These findings identify hUC-MSCs as a promising therapeutic strategy to retard PF progression by targeting the cGAS-STING pathway in patients undergoing long-term peritoneal dialysis.

INTRODUCTION

As a frontline renal replacement therapy, peritoneal dialysis is extensively utilized in treating end-stage renal disease (ESRD)[1,2]. Its global reach is considerable, especially across developing countries, driven by the simplicity, safety, efficacy, and home-based convenience it affords[3]. Previous studies have demonstrated that peritoneal dialysis provides safety and efficacy comparable to hemodialysis, better preserves residual renal function, and imposes a lower cardiovascular burden[4-6]. Clinical data nonetheless corroborate that the peritoneum undergoes structural and functional damage upon prolonged exposure to non-physiological dialysis solutions, including desquamation of the peritoneal mesothelial cell (PMC) layer, phenotypic transformation, enhanced neovascularization, extracellular matrix deposition, and inflammatory cell infiltration. These pathological changes ultimately culminate in peritoneal fibrosis (PF)[7-9]. Therefore, elucidating strategies to delay or inhibit the progression of PF is of considerable scientific and clinical significance for improving dialysis efficiency and prolonging patient survival.

Mesenchymal stem cells (MSCs) are undifferentiated cells known to retain self-renewing ability, multilineage regenerative potential, and migratory behavior[10,11]. Under specific induction conditions, MSCs have demonstrated therapeutic potential in repairing tissue injury and fibrosis in organs such as the kidney[12], heart[13], lung[14], and liver[15]. First identified in bone marrow, MSCs were primarily derived from this source[16]. However, clinical acquisition of bone marrow presents challenges owing to limited donor availability and the invasive nature of the procedure. Moreover, with advancing age, both the proliferative capacity and multilineage differentiation potential of bone marrow-derived MSCs (BMSCs) decline[17]. The potential risk of viral contamination further limits the broader clinical application of BMSCs[18]. In contrast, the umbilical cord, typically discarded after childbirth, yields a prolific and ethically approved MSC source. Compared with BMSCs, human umbilical cord MSCs (hUC-MSCs) exhibit superior proliferative and differentiation capabilities. Moreover, owing to their low immunogenicity and immunomodulatory nature, hUC-MSCs carry a diminished risk of eliciting immune rejection in both allogenic and even xenogenic transplantation settings[19-21]. Nevertheless, research addressing the therapeutic actions of hUC-MSCs in PF is still in its early stages, and molecular mechanisms underlying their anti-fibrotic effects require further clarification.

In light of the above, this study pursues a detailed investigation of the mechanisms mediating hUC-MSC-driven amelioration of PF. Dual PF models, encompassing in vivo and in vitro approaches, were engineered for the assessment of hUC-MSC therapeutic effects. We further explored the extent to which hUC-MSCs influence high glucose (HG)-initiated epithelial-mesenchymal transition (EMT) in HMrSV5 human PMCs.

MATERIALS AND METHODS

Reagents

The human PMC line HMrSV5 was commercially acquired from BeNa Culture Collection (Beijing, China). The hUC-MSCs were obtained from iCell Bioscience Inc. (Shanghai, China). Antibodies against cyclic GMP-AMP synthase (cGAS), stimulator of interferon genes (STING), interferon regulatory factor 3 (IRF3), TANK-binding kinase-1 (TBK1), epithelial cadherin (E-cadherin), vimentin, alpha-smooth muscle actin (α-SMA), transforming growth factor-beta1 (TGF-β1), fibronectin, and β-actin were obtained from Affinity Biosciences (Jiangsu Province, China). PE anti-human CD29 antibody, PE anti-human CD34 antibody, and PerCP anti-human CD45 antibody were purchased from Biolegend (San Diego, CA, United States). CD44 monoclonal antibody PE-Cyanine7 was purchased from Invitrogen (Carlsbad, CA, United States). The BeyoClick™ EdU Cell Proliferation Kit with AF488 was commercially obtained from Beyotime Biotechnology (Shanghai, China). MedChemExpress (NJ, United States) served as the vendor for DMXAA and DIR. Mouse tumour necrosis factor (TNF)-α and interleukin (IL)-6 enzyme-linked immunosorbent assay (ELISA) Kits were purchased from Shanghai Enzyme-linked Biotechnology Co., Ltd. (Shanghai, China).

Cell culture and treatment

We culture HMrSV5 cells, along with hUC-MSCs, in DMEM. This medium received supplementation comprising 15% fetal bovine serum and 1% penicillin/streptomycin. Cell maintenance was implemented at 37 °C in a 5% CO₂-enriched environment. HMrSV5 cells received 42.5 g/L HG treatment for 24 hours, 48 hours, and 72 hours to simulate the in vitro PF model. Subsequently, hUC-MSCs were co-cultured in a non-contact manner with HMrSV5 cells that had been pre-treated with HG. The co-culture system was arranged with hUC-MSCs in the upper chamber and HG-exposed HMrSV5 cells in the lower chamber. The co-culture ratio of hUC-MSCs to HMrSV5 cells was maintained at 2:1. A 48-hour co-culture period preceded cell collection for the ensuing experiments.

DMXAA, serving as a STING agonist, was used in a rescue experiment to further confirm STING pathway implication. HMrSV5 cells were treated with 50 μg/mL DMXAA and subsequently co-cultured with hUC-MSCs under the same conditions described above. After 48 hours of treatment, HMrSV5 cells were harvested for further analyses.

Identification of hUC-MSCs

Centrifugation was implemented on the collected hUC-MSCs to discard the supernatant. The collected pellet received one phosphate-buffered saline (PBS) wash (1 mL) prior to 5-minute centrifugation at 1500 rpm. Supernatant disposal was followed by resuspension of the cells in a fitting volume of PBS. The suspension was apportioned uniformly into five tubes. One tube served as the negative control, while 5 μL of CD29 PE, CD34 PE, CD44 PE/Cy7, and CD45 PerCP antibodies were added to the remaining four tubes, respectively. Samples were softly mixed and then subjected to a 20-minute light-protected incubation at room temperature. Subsequently, the tubes each received 1 mL PBS, after which centrifugation was performed at 1500 Two additional washes were performed following supernatant disposal. The concluding procedure involved resuspending the cells in 500 μL PBS, mixing well, and performing flow cytometric evaluation.

5-ethynyl-2’-deoxyuridine cell proliferation assay

Strict adherence to the BeyoClick™ EdU Cell Proliferation Kit protocol guided the 5-ethynyl-2’-deoxyuridine staining of HMrSV5 cells. Cell proliferation activity in different HMrSV5 groups was subsequently analyzed via flow cytometry.

Western blotting

We adopted the western blotting (WB) protocol delineated by Dong et al[22]. HMrSV5 cells and mouse peritoneal tissue specimens were gathered for total protein extraction. Protein quantification was implemented by use of the BCA assay. Proper denaturation of protein samples preceded their separation via sodium-dodecyl sulfate gel electrophoresis (1.5 hours) and subsequent transfer to polyvinylidene fluoride membranes at a 300 mA constant current for 1.5 hours. An incubation step at 4 °C with primary antibodies was then applied to the membrane. Room-temperature incubation with matching secondary antibodies for 2 hours was performed on the subsequent day. Post-wash treatment involved chemiluminescent substrate application and signal detection via an ultra-sensitive chemiluminescence imaging system.

Wound healing assay

Each well was scored with a linear scratch using a 200 μL pipette tip. Three PBS rinses following medium removal served to clear any detached cells. Serum-free medium replaced the original culture medium, and 0 hour scratch photographs were immediately obtained. Continued culture for 24 hours in the incubator was followed by re-capture of images from the same fields (24 hours). Wound healing rate calculation for HMrSV5 cells was based on scratch area measurements at 0 and 24 hours.

Cell invasion assay

Post-incubation medium removal and gentle PBS washing for 5 minutes were sequentially performed. Cells were treated with 0.1% crystal violet for fixation and staining (1 hour, room temperature). Post-staining removal of cells from the upper membrane surface was performed with a cotton swab. The inverted inserts were mounted on glass slides and imaged under a microscope. Image capture preceded removal of the staining solution and addition of 1 mL of 33% acetic acid to dissolve crystal violet in invaded cells. The solution was mixed thoroughly and incubated briefly. We aliquoted 200 μL per well into a 96-well plate for measurement. The multifunctional microplate reader was used to determine absorbance at 562 nm.

Quantitative real-time polymerase chain reaction

RNA extraction from HMrSV5 cells was achieved with TRIzol reagent, after which reverse transcription yielded cDNA. Quantitative real-time polymerase chain reaction (qRT-PCR) was implemented with SYBR Green I dye-based detection and β-actin as the normalization reference. For relative quantification of target gene mRNA expression, the 2^-∆∆Ct method was utilized. Thermal cycling parameters were established at 95 °C for 10 minutes initially, with 40 repeat cycles of denaturation at 95 °C for 10 seconds and annealing at 58 °C for 30 seconds. Primer sequences are tabulated in Table 1.

Table 1 Primer sequence.

Gene symbol	Forward primer	Reverse primer
cGAS	AAGGCCTGCGCATTCAAAACT	AAGGATAGCCGCCATGTTTCTTCTT
STING	CCGGGAGGCAGAAGATGC	CAGGCACTCAGCAGAACCAA
IRF3	CACTGAAGCGGCTGTTGGT	AGGAGATGGTCTGCTGGAAG
TBK1	AGACAAAGCAGAACGTAGATTAGC	GCGTCATAGCTTTTGTGGCA
E-cadherin	AATCTGAAAGCGGCTGATACTG	CCATTCGTTCAAGTAGTCATAGTCC
Vimentin	GGATTCACTCCCTCTGGTTG	TGATGCTGAGAAGTTTCGTTG
α-SMA	GCGATCTCACCGACTACCTG	GCCGACTCCATACCGATGAA
TGF-β1	CCGACTACTACGCCAAGGA	AACCACTGCCGCACAACTC
β-actin	TGGCACCCAGCACAATGAA	CTAAGTCATAGTCCGCCTAGAAGCA

cGAS: Cyclic GMP-AMP synthase; STING: Stimulator of interferon genes; IRF3: Interferon regulatory factor 3; TBK1: TANK-binding kinase-1; E-cadherin: Epithelial cadherin; α-SMA: Alpha-smooth muscle actin; TGF-β1: Transforming growth factor-beta1.

Immunofluorescence

Following the respective treatments, HMrSV5 cells were processed through fixation (4% paraformaldehyde, 30 minutes), permeabilization (0.5% Triton X-100, 20 minutes), and blocking (5% BSA, 37 °C, 30 minutes). Overnight probing of the samples with anti-fibronectin primary antibody was performed at 4 °C. Cy3-conjugated Goat Anti-Rabbit IgG (H + L) secondary antibody was administered after washing. The slides were DAPI-counterstained, mounted, and analyzed via fluorescence microscopy.

Animals

Male C57BL/6J mice (n = 33, specific pathogen-free-grade, 6-8 weeks old) were purchased from Jiangsu Jicui Yaoke Biotechnology Co., Ltd. [license No. SCXK (Su) 2023-00099]. Compliance with the Medical Ethics Committee-approved ethical guidelines (No. 20240819) was maintained throughout all animal experiments. After one week of adaptive feeding, a 28-day regimen of daily intraperitoneal injections (3 mL, 4.25% peritoneal dialysis fluid) was employed to establish the PF model. A randomization procedure assigned the mice to three groups (control, PF, hUC-MSCs), each comprising 10 animals. The hUC-MSCs group received intravenous hUC-MSC injections (1 × 10⁶ cells per mouse) via the tail vein once weekly during the modeling period for four consecutive weeks. To track the distribution of transplanted stem cells in vivo, hUC-MSCs were labeled with the near-infrared dye DIR prior to injection. The labeling step involved resuspending hUC-MSCs (1 × 10⁶ cells). Mice (n = 3) were administered DIR-labeled hUC-MSCs by means of tail vein injection. hUC-MSC distribution in mice was tracked by in vivo imaging at the 24-hour post-injection time point. Afterwards, 2.5% isoflurane-induced anesthesia was administered, followed by euthanasia. Peritoneal tissues were collected for subsequent experiments.

Hematoxylin and eosin staining

Harvested peritoneal tissues were preserved in 10% paraformaldehyde. Standard histological processing was performed: Graded ethanol dehydration, xylene clearing, dewaxing, paraffin embedding, thin section cutting, and hematoxylin and eosin (HE) staining. Histopathological changes in the peritoneal tissues of different mouse groups were examined under a light microscope.

Masson trichrome staining

Mouse peritoneal tissues were processed for paraffin embedding and sectioning. We then baked, deparaffinized, and sequentially rehydrated the sections through graded ethanol. For Masson trichrome staining, a three-color staining solution was applied. Sequential staining with potassium dichromate (3 minutes) and hematoxylin (3 minutes) was implemented. Acidic ethanol differentiation solution was used for differentiation. A 10-minute lichun red/fuchsin stain and a 10-minute phosphomolybdic acid treatment were applied sequentially. Toluidine blue staining was subsequently applied for 5 minutes. Post-staining processing included graded alcohol dehydration, xylene clearing, mounting, and observation under a light microscope.

ELISA

We utilized ELISA kits, per the manufacturer’s protocol, to quantify TNF-α and IL-6 in the peritoneal tissues of different mice.

Immunohistochemistry

Deparaffinization in xylene and graded ethanol rehydration followed a 2-hour baking step at 60 °C for the peritoneal tissue sections (to ensure optimal tissue adherence)[23]. PBS was applied to wash the sections three times. Two microwave-based antigen retrieval steps in citrate buffer (pH = 6.0) were followed by triple PBS washing. A 20-minute incubation with hydrogen peroxide working solution under dark conditions served to quench endogenous peroxidase activity before three PBS washes. Sections underwent overnight primary antibody incubation at 4 °C under humidified conditions. Rewarming at 37 °C for 30 minutes and three PBS washes were performed the next day. Incubation with diluted secondary antibody (37 °C, 1 hour, humidified chamber) followed by three PBS washes was then implemented. Hematoxylin counterstaining, graded alcohol dehydration, xylene clearing, neutral balsam mounting, and light microscopic analysis were performed in sequence.

Statistical analysis

Data visualization and statistical analysis were accomplished by use of GraphPad Prism 9.0 software. All experiments were conducted in triplicate, with data presented as mean ± SD. One-way ANOVA was employed to compare multiple groups of quantitative measurements, with pairwise comparisons performed via the Student-Newman-Keuls method. Results with P < 0.05 were regarded as statistically significant.

RESULTS

Activation of the cGAS-STING signaling pathway in vitro and in vivo models of PF

HMrSV5 cell viability following 4.25% HG treatment was assessed via a 5-ethynyl-2’-deoxyuridine assay (Figure 1A). The results showed that cell viability was significantly compromised by 4.25% HG treatment for 48 hours and 72 hours (P < 0.05) (Figure 1B). HMrSV5 cells treated with HG were subsequently subjected to WB analysis for detection of cGAS and STING protein expression (Figure 1C). Quantitative analysis is presented in Figure 1D and E. cGAS and STING protein levels progressively and significantly increased across the 24 hours, 48 hours, and 72 hours HG groups compared with controls (P < 0.05). According to these findings, the 48-hour HG treatment was designated for in vitro PF model generation using HMrSV5 cells. Furthermore, WB was employed to examine cGAS and STING protein expression in mouse peritoneal tissues at 7 days, 14 days, and 28 days following intraperitoneal administration of 4.25% peritoneal dialysis fluid (Figure 1F). Quantitative results are shown in Figure 1G and H. Compared with the control group, the 7-day, 14-day, and 28-day PF groups demonstrated significantly increased expression levels of cGAS and STING proteins in peritoneal tissues (P < 0.05). Moreover, protein expression progressively increased with prolonged treatment duration. Therefore, a 28-day treatment regimen with 4.25% peritoneal dialysis fluid was selected to establish the in vivo PF model.

Figure 1 Activation of the cyclic GMP-AMP synthase-stimulator of interferon genes signaling pathway in in vitro and in vivo models of peritoneal fibrosis. A: 5-ethynyl-2’-deoxyuridine assay was performed to assess the proliferation of HMrSV5 cells under different treatments; B: Quantitative analysis of 5-ethynyl-2’-deoxyuridine assay results; C: Western blot analysis of cyclic GMP-AMP synthase (cGAS) and stimulator of interferon genes (STING) protein expression in HMrSV5 cells; D and E: Quantitative analysis of protein expression in HMrSV5 cells: CGAS (D); STING (E); F: Western blot analysis of cGAS and STING protein expression in peritoneal tissues from different mouse groups; G and H: Quantitative analysis of protein expression in mouse peritoneal tissues: CGAS (G); STING (H). Data are presented as mean ± SD. ^aP < 0.05 vs control. cGAS: Cyclic GMP-AMP synthase; STING: Stimulator of interferon genes; EdU: 5-ethynyl-2’-deoxyuridine.

Effects of hUC-MSCs on migration and invasion of HG-treated HMrSV5 cells

The phenotypic surface markers CD29, CD44, CD34, and CD45 of hUC-MSCs were analyzed through flow cytometry (Figure 2A). The results confirmed that hUC-MSCs were positive for CD29 and CD44 and negative for CD34 and CD45, a pattern that conforms to the phenotypic profile of MSCs. The role of hUC-MSCs in regulating the migratory potential of HG-treated HMrSV5 cells was evaluated by a wound healing assay (Figure 2B). Figure 2C presents the quantitative findings. The migratory rate of cells was significantly higher in the HG group than in the control group (P < 0.05). Compared with the HG group, the hUC-MSCs group showed significantly reduced migratory capacity (P < 0.05). A further evaluation of hUC-MSC-mediated effects on invasion in HG-treated HMrSV5 cells was performed using a Transwell assay (Figure 2D). As shown in Figure 2E, the HG group showed a significant rise in invasive behavior compared with the control group (P < 0.05). In contrast, co-culture with hUC-MSCs markedly attenuated the invasive ability of HMrSV5 cells compared with the HG group (P < 0.05).

Figure 2 Effects of human umbilical cord-derived mesenchymal stem cells on migration and invasion of high glucose-treated HMrSV5 cells. A: Identification of human umbilical cord-derived mesenchymal stem cell surface markers via flow symmetry (negative markers: CD34, CD45; positive markers: CD29, CD44); B and C: Migration ability of HMrSV5 cells assessed via wound healing assay and corresponding quantitative analysis; D and E: Invasion ability of different HMrSV5 cells assessed via Transwell assay and corresponding quantitative analysis. Data are presented as mean ± SD. ^aP < 0.05 vs control; ^bP < 0.05 vs high glucose. hUC-MSCs: Human umbilical cord-derived mesenchymal stem cells; HG: High glucose; CD34: Cluster of differentiation 34; CD45: Leukocyte common antigen; CD29: Integrin beta 1; CD44: CD44 antigen.

Effects of hUC-MSCs on EMT-related gene and protein expression in HG-treated HMrSV5 cells

qRT-PCR analysis demonstrated that HG stimulation significantly upregulated the transcription of mesenchymal markers (vimentin, α-SMA, and TGF-β1) while downregulating the epithelial marker E-cadherin (Figure 3A-D). Notably, hUC-MSCs treatment markedly reversed these alterations, restoring E-cadherin expression and suppressing mesenchymal marker expression (P < 0.05). WB analysis corroborated these transcriptional findings at the protein level (Figure 3E), with quantitative results shown in Figure 3F-I. These results indicated that hUC-MSCs effectively suppress HG-induced EMT in HMrSV5 cells.

Figure 3 Effects of human umbilical cord-derived mesenchymal stem cells on epithelial-mesenchymal transition-related gene and protein expression in high glucose-treated HMrSV5 cells. A: Epithelial cadherin mRNA; B: Vimentin mRNA; C: Alpha-smooth muscle actin mRNA; D: Transforming growth factor-beta1 mRNA levels detected via quantitative real-time polymerase chain reaction; E: Representative western blot bands; F-I: Quantitative analysis of protein expression: Epithelial cadherin (F); vimentin (G); alpha-smooth muscle actin (H); transforming growth factor-beta1 (I). Data are presented as mean ± SD. ^aP < 0.05 vs control; ^bP < 0.05 vs high glucose. hUC-MSCs: Human umbilical cord-derived mesenchymal stem cells; HG: High glucose; E-cadherin: Epithelial cadherin; α-SMA: Alpha-smooth muscle actin; TGF-β1: Transforming growth factor-beta1.

Effects of hUC-MSCs on cGAS-STING signaling activation in HMrSV5 cells

To elucidate the underlying mechanism, the involvement of the cGAS-STING signaling pathway was investigated through qRT-PCR and WB analysis. As shown in Figure 4A and B, compared with the control group, the HG group exhibited significantly increased mRNA expression levels of cGAS and STING (P < 0.05). Compared with the HG group, the hUC-MSCs group showed significantly decreased mRNA expression levels of cGAS and STING (P < 0.05). Subsequently, WB analysis was employed to detect the effect of hUC-MSCs on the cGAS and STING protein expressions of HMrSV5 cells in different groups (Figure 4C). Compared with the control group, the HG group demonstrated significantly increased protein expression levels of cGAS and STING (P < 0.05). Compared with the HG group, the hUC-MSCs group exhibited significantly decreased protein expression levels of cGAS and STING (P < 0.05) (Figure 4D and E). These results indicated that hUC-MSCs inhibited activation of the cGAS-STING signaling pathway in HMrSV5 cells.

Figure 4 Effects of human umbilical cord-derived mesenchymal stem cells on activation of the cyclic GMP-AMP synthase-stimulator of interferon genes signaling pathway in various HMrSV5 cells. A: Cyclic GMP-AMP synthase mRNA; B: Stimulator of interferon genes mRNA levels detected via quantitative real-time polymerase chain reaction; C: Representative western blot bands; D and E: Quantitative analysis of protein expression: Cyclic GMP-AMP synthase (D); stimulator of interferon genes (E). Data are presented as mean ± SD. ^aP < 0.05 vs control; ^bP < 0.05 vs high glucose. hUC-MSCs: Human umbilical cord-derived mesenchymal stem cells; cGAS: Cyclic GMP-AMP synthase; STING: Stimulator of interferon genes; HG: High glucose.

HUC-MSCs suppress HG-induced EMT in HMrSV5 cells by inhibiting the cGAS-STING signaling pathway

To further establish a causal relationship, rescue experiments were performed using DMXAA, a pharmacological agonist of STING. Immunofluorescence staining was conducted to evaluate fibronectin expression in different treatment groups (Figure 5A). Quantitative analysis is shown in Figure 5B. Compared with the HG group, the DMXAA group exhibited significantly increased fibronectin protein expression (P < 0.05). Compared with the DMXAA group, the hUC-MSCs + DMXAA group showed significantly decreased fibronectin protein expression (P < 0.05).

Figure 5 Human umbilical cord-derived mesenchymal stem cells suppress high-glucose-induced epithelial-mesenchymal transition in HMrSV5 cells by inhibiting the cyclic GMP-AMP synthase-stimulator of interferon genes signaling pathway. A: Representative immunofluorescence images of fibronectin expression in different HMrSV5 groups; B: Quantitative analysis of fibronectin fluorescence intensity; C: Representative images of the wound healing assay; D: Quantitative analysis of cell migration; E-H: TANK-binding kinase 1 mRNA (E); interferon regulatory factor 3 mRNA (F); alpha-smooth muscle actin mRNA (G); transforming growth factor-beta1 mRNA (H) levels detected via quantitative real-time polymerase chain reaction; I: Representative western blot bands; J-M: Quantitative analysis of protein expression: TANK-binding kinase 1 (J); interferon regulatory factor 3 (K); alpha-smooth muscle actin (L); transforming growth factor-beta1 (M). Data are presented as mean ± SD. ^aP < 0.05 vs high glucose; ^bP < 0.05 vs DMXAA. hUC-MSCs: Human umbilical cord-derived mesenchymal stem cells; HG: High glucose; EMT: Epithelial-mesenchymal transition; cGAS: Cyclic GMP-AMP synthase; STING: Stimulator of interferon genes; TBK1: TANK-binding kinase 1; IRF3: Interferon regulatory factor 3; α-SMA: Alpha-smooth muscle actin; TGF-β1: Transforming growth factor-beta1.

The migratory ability of HMrSV5 cells was assessed via a wound healing assay (Figure 5C), with quantitative results shown in Figure 5D. Compared with the HG group, the DMXAA group demonstrated significantly enhanced HMrSV5 cell migration (P < 0.05). However, compared with the DMXAA group, the hUC-MSCs + DMXAA group exhibited significantly reduced migratory capacity (P < 0.05). Subsequently, the TBK1, IRF3, α-SMA, and TGF-β1 mRNA expressions in different HMrSV5 cells were measured via qRT-PCR (Figure 5E-H). Compared with the HG group, the DMXAA group showed significantly increased RNA expression levels of TBK1, IRF3, α-SMA, and TGF-β1 (P < 0.05). However, compared with the DMXAA group, the hUC-MSCs + DMXAA group exhibited significantly decreased mRNA expression levels of TBK1, IRF3, α-SMA, and TGF-β1 (P < 0.05). WB analysis was subsequently performed to detect TBK1, IRF3, α-SMA, and TGF-β1 protein expression (Figure 5I). Compared with the HG group, the DMXAA group demonstrated significantly elevated protein expression levels of TBK1, IRF3, α-SMA, and TGF-β1 (P < 0.05). However, compared with the DMXAA group, the hUC-MSCs + DMXAA group showed significantly reduced protein expression levels of TBK1, IRF3, α-SMA, and TGF-β1 (P < 0.05) (Figure 5J-M).

Effects of hUC-MSCs on PF and inflammatory factors in PF mice

To further evaluate the therapeutic effects of hUC-MSCs in vivo, we investigated the impact on PF in PF mice. In vivo imaging was first performed to determine the distribution of DIR-labeled hUC-MSCs (Figure 6A and B). The results showed that DIR-labeled hUC-MSCs were mainly distributed in the peritoneal, liver, and lung tissues of mice. Subsequent fluorescence imaging of excised peritoneal tissues confirmed strong red fluorescence signals in mice receiving hUC-MSCs (Figure 6C), indicating successful homing of transplanted cells. Histopathological changes were assessed via HE staining and Masson trichrome staining (Figure 6D and E). The peritoneal structure in the control group was intact and exhibited no obvious pathological abnormalities. In contrast, the PF group displayed marked thickening of the fibrous layer, extensive inflammatory cell infiltration, and a pronounced increase in PF. However, the hUC-MSCs group showed substantially reduced inflammatory cell infiltration, thinner fibrous layers, and substantially attenuated fibrosis. Moreover, TNF-α and IL-6 levels in the peritoneal tissues of different mice were determined via ELISA (Figure 6F and G). Compared with the control group, the PF group exhibited significantly increased levels of TNF-α and IL-6 in peritoneal tissues (P < 0.05). However, the hUC-MSCs group had significantly reduced TNF-α and IL-6 concentrations relative to the PF group (P < 0.05).

Figure 6 Effects of human umbilical cord-derived mesenchymal stem cells on peritoneal fibrosis and inflammatory cytokines in peritoneal fibrosis mice. A-C: Tracking the distribution of human umbilical cord-derived mesenchymal stem cells (hUC-MSCs) injected in mice. In vivo imaging of whole-body fluorescence (A). Fluorescent imaging of isolated organs (B). Fluorescence imaging of peritoneal tissues following hUC-MSC injection (C); D: Representative hematoxylin and eosin staining images; E: Representative Masson staining images; F and G: Tumor necrosis factor-alpha (F) and interleukin-6 (G) levels in peritoneal tissues were detected via enzyme-linked immunosorbent assay. Data are presented as mean ± SD. ^aP < 0.05 vs control; ^bP < 0.05 vs peritoneal fibrosis. hUC-MSCs: Human umbilical cord-derived mesenchymal stem cells; PF: Peritoneal fibrosis; HE: Hematoxylin and eosin; TNF-α: Tumor necrosis factor-alpha; IL-6: Interleukin-6.

Effects of hUC-MSCs on fibrosis-related protein expression in mouse peritoneal tissue

Immunohistochemistry was performed to examine the expression of fibrosis-related proteins (α-SMA, vimentin, and E-cadherin) in peritoneal tissues (Figure 7A-F). Compared with the control group, the PF group demonstrated significantly increased expression of α-SMA and vimentin (P < 0.05) and significantly decreased expression of E-cadherin (P < 0.05). However, compared with the PF group, the hUC-MSCs group exhibited significantly reduced expression of α-SMA and vimentin (P < 0.05), while E-cadherin expression was significantly increased (P < 0.05). WB analysis was subsequently performed to validate these findings (Figure 7G). The quantitative results (Figure 7H-J) were consistent with the immunohistochemical data.

Figure 7 Effect of human umbilical cord-derived mesenchymal stem cells on fibrosis-related protein expression in mouse peritoneal tissues. A-F: Alpha-smooth muscle actin (A and B); vimentin (C and D); and epithelial cadherin (E and F) expression detected by immunohistochemistry; G: Representative western blot bands; H-J: Quantitative analysis of protein expression: Alpha-smooth muscle actin (H); vimentin (I); epithelial cadherin (J). Data are presented as mean ± SD. ^aP < 0.05 vs control; ^bP < 0.05 vs peritoneal fibrosis. hUC-MSCs: Human umbilical cord-derived mesenchymal stem cells; α-SMA: Alpha-smooth muscle actin; E-cadherin: Epithelial cadherin; PF: Peritoneal fibrosis.

DISCUSSION

As a predominant pathological complication of extended peritoneal dialysis, PF has attracted increasing clinical attention[24,25]. Peritoneal dialysis is an indispensable therapeutic option for renal replacement, with solute and water exchange being achieved through the normal physiological activity of the peritoneum. However, with prolonged exposure to dialysis solutions, the peritoneal membrane may undergo progressive structural and functional disruption, among which PF represents the most prominent pathological manifestation[26-28]. The pathogenesis of PF entails elaborate cross-talk among inflammatory responses, mechanical stimuli, and multiple signaling pathways. Therefore, elucidating the molecular mechanisms underlying PF and identifying effective therapeutic strategies remain critical research priorities. Stem cell therapy has emerged as a focus of considerable research interest recently, owing to its distinctive regenerative and immunomodulatory potential[29]. Among various stem cell sources, hUC-MSCs possess distinct advantages, including abundant availability, non-invasive procurement, and low immunogenicity, making them promising candidates for clinical application[30]. Our data in the present study supported that hUC-MSC treatment alleviated PF in PF mice. Specifically, hUC-MSCs reduced the levels of pro-inflammatory cytokines (TNF-α and IL-6) and fibrosis-related proteins (α-SMA and vimentin) while restoring the expression of E-cadherin in peritoneal tissues. For example, Li et al[31] showed that hUC-MSCs attenuated PF in a rat PF model, supporting the reproducibility and reliability of our results.

The cGAS-STING signaling pathway is a crucial component of innate immunity[32]. As a cytoplasmic DNA sensor, cGAS mainly detects exogenous DNA derived from pathogens. When activated, cGAS catalyzes cGAMP synthesis, which subsequently binds to STING located in the endoplasmic reticulum. This interaction facilitates the translocation of STING to the Golgi apparatus, thereby activating the phosphorylation of TBK1 and IRF3 and inducing the expression of various inflammatory cytokines[33,34]. The cGAS-STING signaling pathway has been found to be aberrantly activated in a number of inflammatory and fibrotic disorders[35-37]. Although chronic inflammation is commonly viewed as a major cause of PF progression, how the cGAS-STING signaling pathway participates in PF remains to be clarified. In our study, compared with the control group, the PF group showed a significant rise in cGAS and STING protein expression in peritoneal tissues. A marked decline in cGAS and STING protein levels was detected in the hUC-MSCs group relative to the PF group.

The EMT of PMCs represents an early and potentially reversible stage in the development of PF[38-40]. Upon exposure to non-biocompatible peritoneal dialysis fluids, PMCs undergo significant structural alterations, including loss of intercellular tight junctions, impairment of cellular polarity, and damage to the basement membrane[41]. During this process, PMCs acquire myofibrotic cells, enabling their migration beneath the basement membrane and contributing to excessive extracellular matrix deposition[42]. This process contributes to increased matrix synthesis and ultimately results in PF[43]. Our findings demonstrated that the cGAS-STING signaling pathway was activated following HG-induced EMT in PMCs. Co-culture with hUC-MSCs markedly suppressed HG-induced EMT in PMCs. Furthermore, compared with the HG group, the hUC-MSCs + HG group exhibited significantly reduced activation of the cGAS-STING signaling pathway in PMCs. The cumulative findings suggest that hUC-MSCs may hamper HG-induced EMT in PMCs by inhibiting cGAS-STING pathway activity, thereby exerting a therapeutic effect against PF in PF mice. However, this study also has some limitations. Although in vitro rescue experiments using the STING agonist DMXAA confirm that hUC-MSCs alleviate HG-induced EMT by suppressing cGAS-STING signaling, in vivo rescue experiments remain insufficient. Future studies will include in vivo validation using STING gene knockout mice to further substantiate the mechanistic role of this pathway in PF.

CONCLUSION

In summary, the cGAS-STING signaling pathway is activated both in vivo and in vitro models of PF. Moreover, hUC-MSCs attenuate HG-provoked EMT in PMCs by blocking cGAS-STING pathway activation, thereby alleviating PF in PF mice.