{"paper_id":"1740c79e-d0ca-45cb-8e83-60a9bbb8b74c","body_text":"1\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-001468\nMesenchymal stem cells improve \novarian function by suppressing fibrosis \nthrough CTGF/FAK signalling in \nsystemic lupus erythematosus\nHaiwei Zhang    ,1 Hui Yang    ,2 Yingyi Wu,3 Yirui Shi,1 Min Xu,2 Yueyang Zhang,3 \nHongwei Chen    ,2 Lingyun Sun    1\nTo cite: Zhang H, Yang H, \nWu Y, et al. Mesenchymal \nstem cells improve ovarian \nfunction by suppressing fibrosis \nthrough CTGF/FAK signalling in \nsystemic lupus erythematosus. \nLupus Science & Medicine \n2025;12:e001468. doi:10.1136/\nlupus-2024-001468\n ► Additional supplemental \nmaterial is published online only. \nTo view, please visit the journal \nonline (https:// doi. org/ 10. 1136/ \nlupus- 2024- 001468).\nHZ and HY contributed equally.\nReceived 29 November 2024\nAccepted 10 July 2025\nFor numbered affiliations see \nend of article.\nCorrespondence to\nDr Hongwei Chen;  chenhw@ \nnju. edu. cn and Dr Lingyun Sun;  \nlingyunsun@ nju. edu. cn\nReproductive health and APS\n© Author(s) (or their \nemployer(s)) 2025. Re- use \npermitted under CC BY- NC. No \ncommercial re- use. See rights \nand permissions. Published by \nBMJ Group.\nAbstrAct\nObjective SLE is a multisystem autoimmune disease \ncharacterised by chronic inflammation and progressive \norgan damage, including ovarian dysfunction. This \nstudy investigated the therapeutic efficacy of umbilical \ncord- derived mesenchymal stem cells (UC- MSCs) in \nameliorating ovarian impairment and restoring ovarian \nfunction through the inhibition of fibrosis in a lupus mouse \nmodel.\nMethods Serum levels of sex hormones were quantified \nvia ELISA. Ovarian tissue samples were histologically \nevaluated for follicle count and fibrosis via H&E and \nMasson’s trichrome staining. Quantitative reverse- \ntranscriptase- PCR, western blot, immunofluorescence \nand immunohistochemistry were employed to evaluate \ninflammatory cytokines, fibrotic factors, hormone receptors \nand signalling proteins. Primary granulosa cells (GCs) \nisolated from lupus mice (MRL/lpr) were cocultured with \nMSCs and the expression of fibrotic factors was analysed \nby western blot. Additionally, a human GC line (KGN) was \nused to further explore the relationships among connective \ntissue growth factor (CTGF), focal adhesion kinase (FAK)/\nFAK- Tyr576/577 phosphorylation and fibrosis. This was \nachieved through stimulation with recombinant CTGF, the \nCTGF antagonist FG- 3019 or the FAK inhibitor SU6656.\nResults UC- MSC transplantation significantly \ndownregulated the expression of proinflammatory \ncytokines (Tnf-α, Il- 1β) and fibrotic markers (Ctgf, α-Sma) \nwhile upregulating the expression of key hormone \nreceptors (Amh, Esr1, Esr2). Additionally, a reduction in \nCD3\n+/CD4+ T- cell infiltration, C3 complement deposition \nand IgG levels was observed, accompanied by an increase \nin regulatory T cells. Further analysis revealed that fibrotic \nmarkers and FAK- Tyr576/577 phosphorylation were \nmarkedly suppressed in primary ovarian GCs following \nMSC transplantation. In vitro experiments demonstrated \nthat recombinant CTGF promoted fibrogenesis in the \nhuman GC line KGN. Conversely, MSC treatment inhibited \nphosphorylated FAK- Tyr576/577 and downregulated the \nexpression of Collagen 1 and α-SMA, suggesting that \nUC- MSCs alleviate ovarian fibrosis by suppressing FAK- \nTyr576/577 phosphorylation.\nConclusion This study demonstrated that UC- MSC \ntreatment ameliorated ovarian dysfunction and attenuated \novarian fibrosis in lupus mice by modulating the CTGF/FAK- \nTyr576/577 phosphorylation pathway.\nIntROduCtIOn\nProlonged chronic inflammation in SLE inev-\nitably results in multi- organ damage. Notably, \na growing number of female patients with SLE \nexhibit reproductive system impairments, \nincluding irregular menstruation, amenor -\nrhoea, premature ovarian failure (POF) and \nother reproductive system injuries.\n1 Indeed, \nchronic inflammatory stimulation by TNF-α \nand interleukin (IL)- 6 seems to play a role in \nthe pathogenesis of polycystic ovary syndrome \n(PCOS).\n2 In patients with SLE, elevated \nlevels of proinflammatory cytokines (IL- 1β \nand IL- 18) in serum, which are involved in \nfibrosis,3 4  presumably contribute to ovarian \ndamage in turn. In addition, ovarian hard-\nening and insufficiency are also associated \nwith elevated concentrations of inflamma-\ntory cytokines induced by the autoantibody- \ncomplement complex in patients with SLE.\n5–9 \nMoreover, increased infiltration and prolifera-\ntion of peripheral neutrophils and Th17 cells \nWHAt IS ALREAd Y KnOWn On tHIS tOPIC\n ⇒ Fibrosis and chronic inflammation can cause ovarian \ndysfunction in SLE. Mesenchymal stem cells (MSCs) \nhave shown therapeutic efficacy in alleviating lupus \nnephritis. However, their effects on ovarian function \nin patients with SLE remain poorly understood.\nWHAt tHIS Stud Y AddS\n ⇒ Umbilical cord- MSC treatment was found to improve \novarian function and reduce ovarian fibrosis in lupus \nmice by modulating the connective tissue growth \nfactor/focal adhesion kinase- Tyr576/577 pathway.\nHOW tHIS Stud Y MIGHt AFFECt RESEARCH, \nPRACtICE OR POLICY\n ⇒ This study suggests that MSCs might be potential \ntherapeutic agents for the treatment of ovarian dys-\nfunction in patients with lupus in the future, espe-\ncially for the treatment of ovarian inflammation and \nfibrosis.\nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-0014682\nLupus Science & Medicine\nalso seem to participate in the process of impairing the \novaries.10 Although insufficient blood supply, deposition \nof immune complexes, autoreactive immune cell infil-\ntration and dysregulated gonadal hormones (eg, oestra-\ndiol and Müllerian hormone) are primarily observed in \nthe ovaries of SLE, how the onset of lupus disease causes \novarian dysfunction remains mostly unknown.\n11–13\nFibrosis, a marker of tissue ageing and dysfunction, \nplays an important role in the pathogenesis of organ \ninsufficiency, including ovarian dysfunction. Many proin-\nflammatory factors can induce fibrosis in multiple organs, \nresulting in their dysfunction. TNF-α and IL- 17 have toxic \neffects on liver fibrosis.\n14 Blockade of IL- 6 signalling \nreportedly reduces renal fibrosis. 15 In osteoarthritis, the \nupregulation of proinflammatory cytokines (IL- 1β and \nIL- 18) and fibrosis markers (TGF-β, PLOD2, COL1A1 and \nTIMP1) affects the recovery of arthritis.16 Notably, fibrosis \nalso affects ovarian function. Lliberos et al reported that \nthe failure of follicles resulted from ovarian ageing due \nto chronic inflammation and fibrosis with the activation \nof proinflammatory cytokines (TNF-α, IL- 1α/β and IL- 6), \ninflammasome genes (ASC and NLRP3) and Collagen I \nand III.\n17 Collagen and hyaluronic acid have been shown \nto regulate ovarian stiffness, which can affect follicular \ndevelopment and oocyte quality.\n18 19 In addition, NLRP3 \ninflammasome activation via Toll- like receptor 4 in PCOS \novaries affects follicular dysfunction by increasing the \nexpression of fibrotic factors, such as TGF-β, connective \ntissue growth factor (CTGF), α-SMA, β-catenin, type I \ncollagen and β-collagen, in ovarian cells.\n20 Furthermore, \nfocal adhesion kinase (FAK) can link mechanical force \nto certain types of fibrosis via inflammatory signalling.\n21 \nHence, ovarian fibrosis seems to directly affect ovarian \nfunction. Blocking the excessive activation of fibrosis is \nassumed to be an effective strategy to alleviate ovarian \npathological remodelling.\nRecently, human umbilical cord- derived mesenchymal \nstem cells (UC- MSCs) have shown broad application \nprospects in the treatment of autoimmune diseases and \nregenerative medicine.\n22 Transplantation of UC- MSCs has \nachieved beneficial results in SLE, rheumatoid arthritis, \nSjögren’s syndrome, etc.\n23–25 MSC therapy has also been \nshown to be one of the most effective treatments for \nrestoring fertility and pregnancy. Notably, UC- MSCs are \nthe most appropriate MSCs for treating immune- related \novarian insufficiency because of their painless extraction \nprocedure.\n26 27  MSCs can improve ovarian function by \nameliorating ovarian inflammation and granulosa cell \n(GC) apoptosis and restoring functional hormone levels \nin POF and PCOS mice.\n28 29  However, there are few \nreports on the treatment of ovarian fibrosis with MSCs. \nCui et al reported that human UC- MSC transplantation \n(MSCT) significantly inhibited the expression of α-SMA \nin primary ovarian insufficiency and the production of \ncollagen I and collagen III but did not provide strong \nevidence of functional improvement.\n30 Therefore, the \nunderlying mechanism by which MSCs improve ovarian \nfunction in SLE is still unclear.\nIn this study, we used a Fas mutant lupus mouse model \n(LPR) to investigate the effect of MSCs on ovarian fibrosis \nin SLE. We analysed the changes in ovarian fibrosis, the \nimmune microenvironment and ovarian function after \nthe onset of SLE and after MSCT in LPR mice. We further \nexplored the important role of the CTGF/FAK signalling \npathway in suppressing fibrosis and restoring ovarian \nfunction in SLE.\nMAtERIALS And MEtHOdS\nAnimals\nFemale MRL/lpr mice (LPR) at the ages of 7 weeks \n(n=5) and 22 weeks (n=5) and female MRL/mpj mice \n(MPJs) at the ages of 7 weeks (n=5) and 22 weeks (n=5) \nwere purchased from Sibeifu (Beijing, China) for in \nvivo experiments. The mice were housed in cages \nunder conditioned air and had free access to food \nand water under specific pathogen- free conditions. \nAll animal experiments followed the institutional \nguidelines of the Affiliated Drum Tower Hospital of \nMedical School of Nanjing University.\nCulture and transplantation of MSCs\nHuman UC- MSCs were acquired from the Stem Cell \nCenter of Jiangsu Province. The details of isola-\ntion, purification and identification were described \npreviously.\n30 Dulbecco’s Modified Eagle’s Medium/\nnutrient mixture F- 12 (DMEM/F12, Gibco) supple-\nmented with 10% fetal bovine serum (FBS, Gibco) was \nused to culture the MSCs. For animal experiments, \nMSCs at passage six were harvested after removal \nof the cell culture media, suspended in phosphate-  \nbuffered saline (PBS) and injected into LPR mice \n(5×10\n5 cells/mouse) through the tail vein at 20 weeks \nof age. The mice were then sacrificed, and the tissues \nwere collected at the age of 22 weeks for further \nexperiments. To track the engraftment of UC- MSCs \nafter transplantation in 20- week- old LPR mice, the \ncells were labelled with a PKH26 red fluorescent cell \nlinker (Sigma, USA) according to the manufacturer’s \nprotocol. One hour after the transplantation of the \nPKH26 labelled UC- MSCs, the LPR mice were sacri-\nficed and their ovaries were collected for further \nimmunofluorescence examination.\nHistological examination\nMurine ovaries were fixed, embedded and sectioned \nat 5 µm. Sections of the largest area of the whole ovary \nwere dewaxed with xylene, dehydrated in a graded \nalcohol series and dyed with H&E and Masson. Pictures \nwere taken via a photomicroscope (Olympus, Tokyo, \nJapan) at 100× magnification. Histological scores \nwere measured with ImageJ according to the protocol \ndescribed previously.\n31 Independent sample Student’s \nt tests were used to compare the measurement data \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-001468 3\nReproductive health and APS\nbetween the two groups. p<0.05 was considered statis-\ntically significant.\nImmunohistochemistry\nThe ovary sections were treated with 3% hydrogen \nperoxide to block endogenous peroxidases and then \nincubated overnight with rabbit anti- mouse primary \nantibodies, including α -SMA (1:200, Proteintech, \n55135–1- AP), COLLAGEN1 (1:500, Proteintech, \n14695–1- AP), FSHR (1:200, Proteintech, 22665–1-  \nAP), CTGF (1:500, Proteintech, 25474–1- AP), F4/80 \n(1:1000, Servicebio, GB113373- 100, Wuhan China) \nand Ly6G (1:1000, Servicebio, GB11229- 100), at 4°C. \nThe secondary antibodies used were horseradish \nperoxidase (HRP)- conjugated goat anti- rabbit/mouse \nIgG (1:200, Servicebio, GB23303/GB23301). The \n3,3- N- diaminobenzidine tetrahydrochloride (DAB) \nchromogen (Servicebio, G1211) was further applied \nfor colour reactions, and the development time was \ncontrolled under a microscope, followed by haema-\ntoxylin staining and sealing of the sections.\nImmunofluorescence\nFrozen sections of ovaries were 10 microns thick and \nfixed at room temperature with 4% PFA (paraformal-\ndehyde) for half an hour. After blocking endogenous \nperoxidase with 1% hydrogen peroxide, the sections \nwere blocked with 3% bovine serum albumin (BSA) \nfor 1 hour and incubated with rabbit and mouse \nprimary antibodies at 4°C overnight. The primary anti-\nbodies used were C3 (1:100, Abcam 200999), FSHR \n(1:300, Proteintech 14 695–1- AP), α -SMA (1:300, \nProteintech 55 135–1- AP), Collagen Type 1 (1:100, \nProteintech 14 695–1- AP) and CTGF (1:100, Protein-\ntech 25 474–1- AP). After rinsing the slides with PBS \n(pH 7.4), goat anti- mouse IgG conjugated to Alexa \nFluor 488 (1:1000, CST 4408S) or PE (1:1000, CST \n8887S), and goat anti- rabbit IgG conjugated to Alexa \nFluor 488 (1:1000, CST 4412S) or Alexa Fluor 555 \n(1:1000, CST 4413S) were added to the slides, which \nwere incubated for 1 hour at room temperature. The \nslides were washed with PBS three times, nucleated \nwith Hoechst 33 258 (1:200, Sigma 94403) at room \ntemperature for 10 min, washed again with PBS and \nsealed with a fluorescence quencher. Slice imaging \nwas performed using a laser confocal scanning micro-\nscope (Olympus FV3000, Tokyo, Japan).\nELISA and flow cytometry\nSerum from each LPR mouse was collected, centri-\nfuged and stored at −80°C for analysis. The levels of \nanti- Müllerian hormone (AMH) and oestradiol (E2) \nwere measured using an ELISA kit according to the \nmanufacturer’s instructions (Multi Sciences, China).\nCells were isolated from 22- week- old LPR murine \novaries according to a previously described isolation \nmethod.\n32 33  The cells were stained with the following \nanti- mouse antibodies: anti- CD25- APC (BioLegend, \n101909), anti- CD4- FITC (BioLegend, 100510) and \nanti- FOX3- PE (BioLegend, 126403). The stained cells \nwere analysed by flow cytometry (BD LSRFortessa), \nand the data were analysed by FlowJo software.\nCollection, culture and treatment of ovarian granulosa cells \nfrom ovaries\nOvarian GCs were isolated from 22- week- old LPR females. \nThe GCs were isolated by piercing the follicles with a \nsterile syringe under a stereoscope (Olympus, Tokyo, \nJapan).\n34 35 The remaining ovarian tissue was cut into 1 \nmm3 pieces with scissors and washed three times with \nPBS. Mixed digestive juices (0.4% collagenase intrave-\nnous, 0.1% deoxyribonuclease I, 0.2% dispase II and 0.2% \nhyaluronidase) dissolved in DMEM/F12 medium were \nused to digest the tissues at 37°C for 30 min. After diges-\ntion, the strain was strained through a 100 µm strainer \nand centrifuged at 1000 rpm for 5 min. The precipitates \nwere suspended in DMEM/F12 medium supplemented \nwith 10% FBS (ExCell Bio, FSP500), 1% 100 U/mL strep-\ntomycin sulfate and 100 U/mL penicillin G and cultured \nin a humid incubator containing 5% CO2 at 37°C.\ntranswell assays\nPrimary GCs were cocultured with MSCs in a transwell \nchamber (0.4 µm pore size, Corning, New York, USA), \nand the control group was cultured without MSCs. The \nupper chamber, which contained 300 µl of serum- free \nDMEM/F12 medium, was filled with a density of 1×10\n4 \nMSCs. The lower compartment contained GCs filled with \n1 mL of DMEM/F12 containing 10% FBS. After incuba-\ntion at 37°C for 24 hours, 200 µl of radioimmunoprecipi-\ntation assay (RIPA) (NCM Biotech, WB3100) lysis solution \ncontaining 1% phosphatase inhibitor and 1% protease \ninhibitor (Epizyme BioTech, GRF101) was added to one \npart, and 1 mL of TRIzol (Vazyme, R401- 01) was added to \nthe other part and stored at −80°C.\nQuantitative reverse-transcriptase-PCR\nTotal RNA was extracted from cells or whole ovarian \nlysate tissues with TRIzol Reagent (Vazyme, R401-  \n01) and reverse transcribed according to the manu-\nfacturer’s instructions. Complementary DNA was \nsynthesised using the PrimeScript RT Kit (Takara \nBiotechnology, Tokyo, Japan). A QuantStudio 6 Flex \n(Foster City Applied Biosystems, USA) was used for \nPCR amplification. All reactions were performed \nusing a FastStart DNA Master SYBR Green I light \ncycler (Takara Biotechnology, Tokyo, Japan) \naccording to the manufacturer’s instructions. All \nprimer sequences were synthesised and verified via \nBasic Local Alignment Search Tool (BLAST) searches \nvia the National Center for Biotechnology Informa-\ntion (NCBI) software tool Primer- BLAST (https://\nwww.ncbi.nlm.nih.gov/tools/primer-blast/). The \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-0014684\nLupus Science & Medicine\nFigure 1 Deterioration of the ovarian microenvironment with impaired ovarian function in LPR mice after disease onset. \n(A) H&E staining was performed in 7- week- old MRL/lpr (LPR) and MRL/mpj (MPJ) mice compared with 22- week- old LPR and \nMPJ mice. Scale bar: 50 µm. (B) Masson staining of the ovaries of 7- week- old MRL/lpr and MRL/mpj mice and 22- week- old \nMRL/lpr and MRL/mpj mice. (C) The number of mature nucleated follicles in the two groups of mice aged 7 weeks and 22 \nweeks. (D) Collagen volume fraction (CVF%) statistics of 7- week- old and 22- week- old mice in the two groups. (E) Comparison \nof Tnf-α, Il- 1β, Il- 6 and Il- 18 mRNA expression levels in the ovaries of 22- week- old LPR and MPJ mice. (F) Comparison of the \nmRNA expression levels of the ovarian fibrosis factors Ctgf, Collagen I, α-Sma and Tgf-β1 between LPR and MPJ mice at 22 \nweeks. (G) Comparison of the mRNA expression levels of the ovarian hormone function receptors Amh, Esr1, Esr2 and Fshr \nin LPR and MPJ mice at 22 weeks. Data are presented as the mean±SEM; two- way ANOVA (C–D, n=5) and Student’s t- test \n(E–F , n=5); *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. AMH, anti- Müllerian hormone; ANOVA, analysis of variance; CTGF , \nconnective tissue growth factor; ESR, oestrogen receptor; FSHR, follicle stimulating hormone receptor; IL, interleukin; mRNA, \nmessenger RNA; SMA, smooth muscle actin; TGF , transforming growth factor; TNF , tumour necrosis factor.\nprimers were synthesised by GenScript (Nanjing, \nChina) and are listed in online supplemental table \nS1.\nWestern blotting\nCells or tissues were lysed in RIPA buffer containing \na protease and phosphatase inhibitor cocktail. The \nwhole process of ice cracking takes half an hour. \nThe protein samples were resolved by 12% sodium \ndodecyl sulfate–polyacrylamide gel electrophoresis \n(SDS- PAGE) and blotted onto polyvinylidene difluo-\nride membranes. After being blocked with 5% non- fat \nmilk in Tris- buffered saline- Tween (0.2%) for 1 hour, \nthe blots were incubated overnight with appropriate \nprimary antibodies and then with HRP- conjugated \nsecondary antibodies. The signal was visualised with \nan enhanced chemiluminescence system.\nStatistical analysis\nQuantitative results are expressed as the mean±SEM. \nAll the data were analysed using GraphPad Prism \nV.8.0. Statistical differences were analysed by using \nStudent’s t- test between two groups and using two- way \nanalysis of variance among multiple groups. p<0.05 \nwas considered statistically significant.\nRESuLtS\ndeterioration of the ovarian microenvironment with impaired \novarian function in LPR mice after disease onset\nTo determine whether lupus activity affects ovarian func-\ntion, we first performed H&E and Masson staining of the \novaries of LPR mice and their control MPJ mice. As antic-\nipated, no significant differences were observed between \nthe two groups at 7 weeks of age in terms of the number \nof nucleated follicles or the area of collagen fibres, an indi-\ncator of fibrosis (figure 1A,B). However, by 22 weeks of age, \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-001468 5\nReproductive health and APS\nFigure 2 UC- MSC transplantation reduces inflammation and reshapes the immune microenvironment in the ovaries of \nLPR mice. (A) Comparison of Tnf-α, Il- 1β, Il- 6 and Il- 18 mRNA expression levels in the ovaries of the MSCT and PBS groups. \n(B) Immunofluorescence showed that the levels of IgG and C3 immune complexes were decreased in the MSCT group. Scale \nbar: 50 µm. (C) The expression of CD3 and CD4 in mouse ovaries decreased after UC- MSC transplantation. Scale bar: 50 µm. \n(D) The proportion of Tregs in the ovaries of the two groups of mice was analysed by flow cytometry. Treg cells increased in the \novaries of the mice in the MSCT group. The data are presented as the means±SEMs; Student’s t- test (A, n=5; D, n=4); *p<0.05, \n**p<0.01. IL, interleukin; mRNA, messenger RNA; MSCT, mesenchymal stem cell transplantation; PBS, phosphate- buffered \nsaline; TNF , tumour necrosis factor; Tregs, regulatory T cells; UC- MSC, umbilical cord- derived mesenchymal stem cell.\nthe LPR mice exhibited a marked reduction in the number \nof nucleated follicles, accompanied by significant ovarian \nfibrosis (figure 1A,B). Each group of mice presented fewer \nfollicle numbers and a greater collagen volume fraction \n(CVF) from the ages of 7 to 22 weeks (figure 1C,D), which \nwas probably due to age. Importantly, following the active \nlupus activity confirmed by dramatically elevated serum \nanti-double- stranded DNA (dsDNA) antibody at 22 weeks \nof age (online supplemental figure S1A), the number of \nfollicles was significantly reduced by approximately fivefold \n(figure 1C) in the LPR mice compared with the MPJ mice, \nwhereas the CVF increased by more than threefold in the \nLPR mice (figure 1D), suggesting that active lupus activity \nindeed impairs ovarian development.\nTo understand how an ovary is injured under lupus condi-\ntions, we examined the expression of inflammatory factors in \nthe ovarian microenvironment, as previous studies reported \nthat proinflammatory factors can promote fibrosis and \naffect ovarian function.\n3 4 Indeed, our quantitative reverse- \ntranscriptase- PCR (RT‒qPCR) results revealed a marked \nupregulation of proinflammatory factors (Tnf-α, Il- 1β, Il-6 \nand Il- 18) in the LPR mice compared with the MPJ mice at \nthe age of 22 weeks (figure 1E). In contrast, there was no \nsignificant difference in terms of the expression of most \ninflammatory factors between the LPR and MPJ mice before \ndisease onset at 7 weeks (online supplemental figure S1B), \nat which time only the expression level of IL- 1β, a pioneer \nfactor driving the pathogenesis of lupus, was significantly \ngreater in the LPR mice than in the MPJ mice. Consistently, \nin the ovaries of mice with active lupus at 22 weeks of age, \nthe expression of fibrotic markers (Ctgf, collagen I and a- Sma) \nwas significantly upregulated, and Tgfβ1 also tended to be \nupregulated, although the difference was not significant \n(figure 1F). However, these fibrotic markers were not upreg-\nulated at 7 weeks of age (online supplemental figure S1C).\nAdditionally, to confirm the impairment of ovarian func-\ntion associated with the upregulation of proinflammatory \nfactors and fibrosis in LPR mice, we next detected the expres-\nsion of ovarian hormone (Amh) and functional receptors \n(Esr1, Esr2 and Fshr). The results revealed significant down-\nregulation of these genes in the ovaries of 22- week- old LPR \nmice (figure 1G), whereas no such changes were observed \nin 7- week- old LPR mice (online supplemental figure S1D). \nThus, lupus disease activity seems to deteriorate the ovarian \nmicroenvironment through excessive inflammatory factors \nand fibrosis, presumably breaking down the function of \novarian hormone secretion and eventually leading to ovarian \ninsufficiency in lupus mice.\nuC-MSC transplantation alleviates ovarian inflammation and \nreshapes the immune microenvironment in LPR mice\nUC- MSCs have shown potential immune modulation \ncapabilities in many diseases. To investigate whether they \ncan also have immunosuppressive effects that reset the \nimmune niche of the ovaries in LPR mice, we first exam-\nined the expression of proinflammatory factors that were \npreviously significantly upregulated in the ovaries of LPR \nmice. As expected, the RT‒qPCR results demonstrated \nsignificant downregulation of Tnf-α, Il- 1β, Il- 6 and Il- 18 \nin the ovaries of the LPR mice after UC- MSC treatment \n(figure 2A and online supplemental figure S2). Next, \nimmunofluorescence staining verified that UC- MSCT also \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-0014686\nLupus Science & Medicine\nFigure 3 UC- MSC transplantation improves ovarian \nfunction and reduces ovarian fibrosis. (A) The area of \nMasson- stained ovaries in the MSCT group was significantly \nsmaller than that in the PBS group. Scale bar: 50 µm. \n(B) Collagen volume fraction (CVF%) statistics in the PBS \nand MSCT groups. (C) Ctgf, Collagen I, a- Sma and Tgf-β1 \nmRNA expression levels in the PBS and MSCT groups. \n(D) H&E staining results of ovaries in the PBS group and \nMSCT group. Scale bar: 50 µm. (E) The number of ovarian \nfollicles in mice before and after MSC transplantation. \n(F) Comparison of the mRNA expression levels of the ovarian \nfunctional receptors Amh, Esr1 and Esr2 between the PBS \ngroup and the MSCT group. (G) Comparison of the ovarian \nmorphology of 22- week- old mice in the PBS, MSCT and \nMPJ groups. (H) Serum AMH concentrations of the mice in \nthe PBS group and MSCT group were detected via ELISA. \n(I) Serum E2 concentrations of the mice in the PBS group \nand MSCT group were detected via ELISA. The data are \npresented as the means±SEMs; Student’s t- test (B–C, E–F , \nH–I, n=5); *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. AMH, \nanti- Müllerian hormone; CTGF , connective tissue growth \nfactor; E2, oestradiol; ELISA, enzyme- linked immunosorbent \nassay; ESR, oestrogen receptor; mRNA, messenger RNA; \nMSCT, mesenchymal stem cell transplantation; PBS, \nphosphate- buffered saline; SMA, smooth muscle actin; TGF , \ntransforming growth factor; UC- MSC, umbilical cord- derived \nmesenchymal stem cell.\nreduced the deposition of IgG antibodies and C3 comple-\nment in the ovaries of the MSCT group (figure 2B). \nSince T cells are implicated in ovarian dysfunction,\n10 \nwe further assessed changes in T- cell infiltration in the \novaries through immunohistochemistry. The results \nrevealed a noticeable decrease in the numbers of both \nCD3\n+ T cells and CD4+ T cells in the LPR mice following \nMSCT (figure 2C), whereas there was no difference in the \nnumber of CD8\n+ T cells between the LPR mice with or \nwithout MSCT (online supplemental figure S1E). Inter -\nestingly, the proportion of regulatory T cells, a well- known \ninhibitory subtype of T cells, significantly increased from \nan average of 8.1–19.2% in the ovaries of LPR mice in \nthe MSCT group (figure 2D). With respect to other \nimmune cells, F4/80\n+ macrophages but not Ly6 + neutro-\nphils in the ovaries also seemed to be affected following \nMSC treatment (online supplemental figure S3). Taken \ntogether, these results indicate that MSCT reversed the \novarian immune microenvironment and improved the \ninflammatory state in LPR mice.\nuC-MSC transplantation improves ovarian function and \nreduces ovarian fibrosis\nGiven that the immune microenvironment is reset and \nthat proinflammatory conditions are reduced in the \novaries of LPR mice following UC- MSC treatment, we \nspeculated that ovarian fibrosis would be correspond-\ningly suppressed, leading to the restoration of ovarian \nfunction. To test this hypothesis, Masson staining was \nconducted to assess the extent and distribution of ovarian \nfibrosis. The results revealed that widespread fibrosis \nwas predominantly localised around the follicles in the \novaries of the PBS group, whereas this fibrotic pattern was \nmarkedly diminished in the MSCT group (figure 3A,B). \nFurther RT- qPCR results confirmed the significant down-\nregulation of fibrosis markers (Ctgf, Collagen I, a- Sma and \nTgf-β1) after MSC treatment (figure 3C). In terms of \novarian function, H&E staining revealed that the number \nof nucleated follicles in the ovaries of the MSCT group \nmore than doubled compared with that in the PBS \ngroup (figure 3D,E). Similarly, RT- qPCR analysis showed \nthat the expression levels of ovarian hormone function \nmarkers (Amh, Esr1, Esr2 and Fshr) were greater in the \nMSCT group than in the PBS group (figure 3F). MSCT \nsignificantly improved the number of ovarian cysts in the \novarian tissue (figure 3G). More importantly, the concen-\ntrations of serum AMH and E2 in the MSCT group were \nsignificantly higher than those in the PBS group, as \ndetected by ELISA, indicating that ovarian function was \nsuccessfully improved as ovarian fibrosis was reduced \n(figure 3H,I).\nuC-MSC transplantation suppresses ovarian fibrosis by \ninhibiting FAK/FAKp-tyr576/577 phosphorylation in granulosa \ncells\nTo investigate the mechanism by which UC- MSCT reduces \novarian fibrosis, we performed immunohistochemistry to \nassess the expression of CTGF, COLLAGEN I and α-SMA. \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-001468 7\nReproductive health and APS\nFigure 4 UC- MSC transplantation ameliorates ovarian fibrosis by inhibiting FAK/FAKp- Tyr576/577 phosphorylation in \ngranulosa cells (A) Immunohistochemical expression levels of the fibrotic protein CTGF in the ovaries of the PBS group and \nMSCT group. Scale bar: 50 µm. (B) Immunohistochemical expression levels of the fibrotic protein COLLAGEN I in the ovaries \nof the PBS group and MSCT group. Scale bar: 50 µm. (C) Expression levels of the fibrotic protein α-SMA in the ovaries of \nthe PBS group and MSCT group as determined by immunohistochemistry. Scale bar: 50 µm. (D) FSHR identification by \nimmunofluorescence in the renal podocyte line MPC5 (negative control) and primary granulosa cells from 22- week- old MRL/\nlpr mice. Scale bar: 25 µm. (E)  Ctgf, Collagen I and a- Sma mRNA expression levels in primary granulosa cells from the Ctrl and \nMSCT groups. (F) The expression levels of CTGF , COLLAGEN I, α-SMA and TGF-β1 in primary mouse granulosa cells were \ndetected by western blot after coculture with MSCs. (G) The expression levels of FAK and p- FAK were detected by western blot \nafter coculture with MSCs in primary mouse granulosa cells. (H) The expression levels of CTGF , COLLAGEN I and α-SMA in \nmouse ovaries from the PBS group and MSCT group were detected by western blotting. (I) The expression levels of FAK and p- \nFAK in mouse ovaries from the PBS group and MSCT group were detected by western blotting. The data are presented as the \nmeans±SEMs, Student’s t- test (E, n=4); *p<0.05, ****p<0.0001. AMH, anti- Müllerian hormone; CTGF , connective tissue growth \nfactor; FAK, focal adhesion kinase; FSHR, follicle stimulating hormone receptor; GC, granulosa cell; mRNA, messenger RNA; \nMSCT, mesenchymal stem cell transplantation; PBS, phosphate- buffered saline; SMA, smooth muscle actin; TGF , transforming \ngrowth factor; UC- MSC, umbilical cord- derived mesenchymal stem cell.\nThe results revealed that all three fibrosis markers were \nsignificantly less abundant in the MSCT group than in the \nPBS group (figure 4A–C). Interestingly, both α-SMA and \nCOLLAGEN I were localised primarily within GCs inside \nthe follicles or around the follicular structures. Next, we \nisolated primary FSHR\n+ GCs from LPR mice to further \ninvestigate the role of UC- MSC treatment in suppressing \nfibrosis within GCs (figure 4D). The RT- qPCR results \ndemonstrated that the expression levels of Ctgf, Collagen \nI and α-SMA decreased in primary GCs following MSC \ntreatment (figure 4E), a finding that was further corrob-\norated by western blotting (figure 4F). Unexpectedly, \nthe levels of TGF-β1, a well- known factor that promotes \nfibrosis, remained unchanged in the ovaries after MSC \ntreatment. Because FAK can mediate extracellular matrix \nsignals to regulate fibrosis,\n36 we then examined FAK \nexpression at the protein level. The results revealed that \nthe phosphorylation of FAK- Tyr576/577 but not FAK was \ninhibited in GCs after MSC treatment (figure 4G). Addi-\ntionally, compared with the PBS group, the MSCT group \npresented reduced expression of CTGF, COLLAGEN I, \nα-SMA and phosphorylated FAK- Tyr576/577 in whole \novarian tissues (figure 4H,I). These findings suggest \nthat MSCs may downregulate the expression of fibrosis- \nrelated factors by inhibiting the phosphorylation of FAK- \nTyr576/577 and α-SMA, potentially through the action of \nCTGF.\nMSCs inhibit fibrosis in granulosa cells through the CtGF-\nmediated FAK/FAKp-tyr576/577 signalling pathway\nSince MSCT was found to reduce both α-SMA and CTGF \nexpression in ovarian tissue, particularly in FSHR\n+ GCs \nof LPR mice (figure 5A,B), we further investigated the \neffects of CTGF on the expression of α-SMA, COLLAGEN \nI and FAK/FAKp- Tyr576/577As shown in figure 5C, \nrecombinant CTGF alone stimulated human GCs (KGNs) \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-0014688\nLupus Science & Medicine\nFigure 5 MSCs inhibit fibrosis via the FAK/FAKp- Tyr576/577 pathway via CTGF in granulosa cells (A) FSHR and α-SMA \nexpression in the ovaries of mice in the PBS group and MSCT group. Scale bar: 50 µm. (B) CTGF and α-SMA expression in the \novaries of mice in the PBS group and MSCT group. Scale bar: 50 µm. (C) Representative western blot results of COLLAGEN I, \nα-SMA and FAK/FAKp- Tyr576/577 pathway proteins in KGN cells treated with CTGF and FG- 3019. (D) Representative western \nblot results of COLLAGEN I, α-SMA and FAK/FAKp- Tyr576/577 pathway proteins in KGN cells treated with CTGF and SU6656. \n(E) Representative western blot results of COLLAGEN I, α-SMA and FAK/FAKp- Tyr576/577 pathway proteins in KGN cells \n(pretreated with CTGF for 24 hours) cotreated with SU6656 and MSCs. (F) Representative western blot results of COLLAGEN \nI, α-SMA and FAK/FAKp- Tyr576/577 pathway proteins in the coculture of KGN cells with CTGF , SU6656 and MSCs. CTGF , \nconnective tissue growth factor; FAK, focal adhesion kinase; FSHR, follicle stimulating hormone receptor; KGN, human ovarian \ngranulosa cells; MSC, mesenchymal stem cell; MSCT, mesenchymal stem cell transplantation; PBS, phosphate- buffered saline; \nSMA, smooth muscle actin.\nto produce fibrosis- related proteins, including α-SMA and \nCOLLAGEN I, which was accompanied by the upregula-\ntion of phosphorylated FAKp- Tyr576/577. This finding \nimplies that CTGF plays a role in regulating fibrosis in \nKGNs. The introduction of a CTGF antagonist (FG- 3019) \neffectively suppressed fibrosis by downregulating α-SMA, \nCOLLAGEN I and FAK/FAKp- Tyr576/577 in KGN cells. \nNotably, this suppression was reversed when KGN cells \nwere treated simultaneously with both CTGF and FG- 3019. \nThese results indicated that CTGF promoted the expres-\nsion of fibrosis markers (α-SMA and COLLAGEN I) in \nKGNs, possibly via FAKp- Tyr576/577.\nTo assess whether FAK signalling mediates the effect \nof CTGF in promoting fibrosis in KGNs, we treated the \ncells with SU6656, a FAK phosphorylation inhibitor. \nThe western blotting results revealed that α-SMA and \nCOLLAGEN I were downregulated in the KGNs following \nSU6656 treatment, which was accompanied by a decrease \nin the protein level of FAKp- Tyr576/577 (figure 5D). These \nfindings indicate that the FAK/FAKp- Tyr576/577 pathway \nindeed mediates CTGF- induced fibrosis. We subsequently \nperformed transwell assays to investigate whether MSCs \ninhibit GC fibrosis via the FAKp- Tyr576/577 pathway. \nKGN cells were pretreated with CTGF for 24 hours to \ninduce the expression of fibrosis- related genes, followed \nby indirect co- culture with MSCs. MSC treatment resulted \nin the downregulation of α-SMA and COLLAGEN I, along \nwith the inhibition of FAKp- Tyr576/577, although FAK \nlevels remained unchanged, as demonstrated by western \nblotting (figure 5E). In contrast, the addition of SU6656, \neither alone or in conjunction with MSCs, further reduced \nthe protein levels of FAKp- Tyr576/577 and α-SMA. These \nfindings suggest that MSCs exert their antifibrotic effects \nthrough the FAKp- Tyr576/577 pathway, leading to the \ninhibition of α-SMA and COLLAGEN I expression.\nTo further elucidate the effect of MSCs on exogenous \nCTGF- induced fibrosis in KGNs, we cocultured KGNs \n(without prior CTGF pretreatment) with MSCs. Western \nblotting revealed that MSCs inhibited the phosphory-\nlation of FAKp- Tyr576/577 as well as the expression of \nα-SMA and COLLAGEN I (figure 5F). Collectively, these \nfindings suggest that MSCs inhibit fibrosis in KGNs \nthrough the CTGF/pFAK- Tyr576/577 signalling pathway.\ndISCuSSIOn\nIn this study, we identified a spectrum of pathological \nalterations in the ovaries of lupus- prone mice following \ndisease onset, including disruption of the immune \nmicroenvironment, fibrosis, gonadal hormone imbal-\nance and follicular depletion. Notably, these patholog-\nical changes were ameliorated, and ovarian function was \nrestored following treatment with UC- MSCs. Mechanisti-\ncally, UC- MSCs modulated the CTGF/FAKp- Tyr576/577 \nsignalling pathway, thereby attenuating ovarian fibrosis.\nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-001468 9\nReproductive health and APS\nOvarian insufficiency remains a significant clinical chal-\nlenge for women with SLE, particularly those seeking to \nconceive. However, the precise mechanisms underlying \nthis condition remain poorly understood. Although \novarian dysfunction in SLE has been attributed to disease \nactivity, as evidenced by the progressive decline in serum \nE2 and AMH levels with active disease, 37 38  the specific \npathways leading to ovarian injury and functional impair-\nment are not well defined. In this study, we demonstrated \nthat ovarian insufficiency in lupus mice was closely \nrelated to fibrosis and an imbalanced immune microen-\nvironment. The presence of excessive proinflammatory \ncytokines in the ovaries of LPR mice presumably leads \nto or exacerbates fibrosis, as other studies have shown \nthat fibrosis is associated with an inflammatory environ-\nment and immune cell infiltration.\n39 40  The elevated \nlevels of proinflammatory cytokines (TNF-α, IL-1β, IL- 6 \nand IL- 18) in the ovaries of LPR mice likely originated \nfrom peripheral blood, infiltrating CD3\n+/CD4+ T cells \nand/or resident ovarian cells (eg, macrophages, GCs and \nconnective cells) affected by the deposition of IgG and \nC3 complement. This inflammatory milieu is thought \nto drive or exacerbate fibrosis, which is consistent with \nprior studies linking fibrosis to inflammatory environ-\nments and immune cell infiltration. Consequently, the \ndisrupted immune niche in the ovary was associated with \nsevere fibrosis, as evidenced by the upregulation of profi-\nbrotic markers (CTGF, α-SMA and Collagen 1). Impor -\ntantly, UC- MSCT effectively suppressed these pathological \nchanges, restoring ovarian function.\nUC- MSCT has been shown to improve autoimmune \ndiseases such as lupus nephritis, Sjogren’s syndrome and \nscleroderma.\n22 25 41 42  However, its therapeutic poten-\ntial for treating ovarian dysfunction in these diseases \nremains underexplored, despite emerging evidence \nthat exosomes derived from adipose- derived stem cells \ncan ameliorate PCOS.\n43 In this study, we demonstrated \nthat UC- MSCT alleviated immune dysregulation, inhib-\nited ovarian fibrosis and restored ovarian function in \nlupus mice. These effects are mediated, at least in part, \nthrough GCs, which play a critical role in folliculogenesis \nand hormone production. GCs differentiate into pari-\netal cells of the basement membrane and cumulus cells \nand contribute to the formation of the corpus luteum.\n44 \nNotably, AMH and E2, key biomarkers of the ovarian \nreserve, are secreted primarily by GCs. In 22- week- old \nLPR mice, ovarian reserve function was significantly \ncompromised, as reflected by reduced serum levels of \nAMH and E2 and the presence of fibrosis in follicular and \nperifollicular regions, where GCs are normally localised.\n35 \nFollowing UC- MSCT, markers of ovarian reserve (AMH \nand E2) and oestrogen receptors (ESR1 and ESR2) were \nupregulated, accompanied by an increase in healthy folli-\ncles and a reduction in atretic follicles within 1–2 weeks. \nThese findings suggest that UC- MSC treatment improves \nGC function, thereby mitigating fibrosis and restoring \novarian function.\nMost strikingly, we identified the CTGF/FAKp- \nTyr576- 577 signalling pathway as a key mechanism \nunderlying the antifibrotic effects of UC- MSCs. Although \nTGF-β1/Smad3 signalling contributes to tissue fibrosis, \nincluding ovarian fibrosis,\n45 CTGF, rather than TGF-β1, \nplays a more important role in inducing ovarian fibrosis \nunder autoimmune conditions in LPR mice since TGF-β1 \nlevels did not differ significantly between MRL/lpr and MPJ \nmice at 22 weeks of age. Further experiments confirmed \nthat the upregulation of fibrosis markers (COLLAGEN1 \nand α-SMA) in ovarian tissues, including GCs, was associ-\nated with the elevated CTGF expression both in vivo and \nin vitro. Moreover, inhibition of FAKp- Tyr576/577 phos-\nphorylation in cultured GCs blocked collagen produc-\ntion, demonstrating that CTGF- induced fibrosis in GCs \nis mediated by FAKp- Tyr576/577 signalling, a pathway \nalso implicated in fibrosis in other organs.\n46 47  Notably, \nMSC treatment can suppress FAKp- Tyr567/577 signalling \nin GCs, leading to the downregulation of fibrosis- related \nmarkers and the restoration of the hormone- secreting \nfunction of GCs, which is essential for supporting oocyte \ndevelopment and maintaining the ovarian reserve.\nHowever, much work is needed to overcome the \nlimitations of this study. In the future, further explo-\nration of the molecular mechanism by which the \nCTGF/FAK- Tyr576/577 pathway regulates the inter -\nactions among inflammation, fibrosis and ovarian \nfunction in lupus is vital. This information is the key \nto understanding what proinflammatory cytokines \nauthentically dictate fibrotic transformation within \nGCs and why lupus ovarian function can be restored \nfollowing UC- MSCT. In addition, although a local \novarian niche with reduced inflammatory and fibrotic \nconditions was found to help restore ovarian func-\ntion after MSC treatment in this study, other organs, \nsuch as the kidney,\n22 48  brain, 49 joints 50 and skin, 42 \nmay benefit from cell therapy simultaneously, as \nreported previously, thereby synergically protecting \nlupus ovarian function and fertility capacity through \nameliorating the endocrine system, particularly via \nthe hypothalamic‒ pituitary‒ gonadal axis. Another \nlimitation is that all the findings were only based on \nLPR mice, a genetic lupus model in which Fas muta-\ntion causes autoimmunity, which has high disease \npenetrance and a non- representative inflammatory \nprofile compared with human SLE, particularly with \nrespect to ovarian fibrosis. We must be cautious to \ndirectly translate the present achievements to human \npatients before their therapeutic effects on UC- MSCs \ncan be further validated to treat ovarian insufficiency \nin other lupus models with profiles that are more \ncompatible with those of human SLE and in clinical \ntrials with a large number of participants. Moreover, \nalthough UC- MSC treatment is considered a safe ther -\napeutic strategy without the risk of immune rejection, \ntransplanted allogenic cells are ultimately recognised \nand discriminated by the recipient’s immune system \nwithin a short period of time. Hence, identifying \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-00146810\nLupus Science & Medicine\nthe optimal subgroup of UC- MSCs with the best \ntherapeutic effects on ovarian function is essential \nbecause heterogeneous MSCs comprise different \nsubpopulations.\n48\nIn summary, our findings highlight the critical role \nof the CTGF/FAKp- Tyr567/577 signalling pathway \nin the recovery of ovarian function in lupus mice \nthrough the inhibiting of ovarian fibrosis following \nUC- MSCT. However, the precise mechanisms by which \nproinflammatory cytokines induce ovarian fibrosis \nremain to be fully elucidated. Additionally, further \nstudies are needed to clarify how GCs contribute to \nthe restoration of ovarian function after UC- MSCT.\nCOnCLuSIOn\nUC- MSCT can ameliorate ovarian dysfunction in \nlupus mice by suppressing fibrosis and restoring \nthe immune microenvironment through the CTGF/\nFAKp- Tyr567/577 signalling pathway. This study \nprovides a foundation for the development of novel \ntherapeutic strategies for treating ovarian insuffi-\nciency in SLE.\nAuthor affiliations\n1Department of Rheumatology and Immunology, Nanjing Drum Tower Hospital, \nClinical College of Xuzhou Medical University, Nanjing, China\n2Department of Rheumatology and Immunology, Nanjing Drum Tower Hospital, \nClinical College of Nanjing University of Chinese Medicine, Nanjing, China\n3Department of Rheumatology and Immunology, China Pharmaceutical University \nNanjing Drum Tower Hospital, Nanjing, China\nAcknowledgements This work was supported by Xuzhou Medical University and \nthe Institute of Rheumatology and Immunology, Nanjing Drum Tower Hospital.\nContributors Conceptualised and supervised the study: LS and HC. Data \nacquisition: HZ, HY, YW, YS, MX, YZ. Drafting the manuscript: HZ and HC. Critically \nrevising the manuscript for important intellectual content: HC. Guarantor: HC and \nLS.\nFunding This work was supported by the National Key R&D Program of China \n(2020YFA0710804), Projects of the National Natural Science Foundation of \nChina (81930043 and 82271843) and the Jiangsu Provincial Key Research and \nDevelopment Program (BE2020621).\nCompeting interests None declared.\nPatient and public involvement Patients and/or the public were not involved in \nthe design, or conduct, or reporting, or dissemination plans of this research.\nPatient consent for publication Not applicable.\nEthics approval The study was conducted in accordance with the National \nResearch Council Guidance for Care and Use of Laboratory Animals and approved \nby the Committee of Experimental Animal Administration of the Affiliated Drum \nTower Hospital of Medical School of Nanjing University (Approval number: \n2022AE01012).\nProvenance and peer review Not commissioned; externally peer reviewed.\ndata availability statement Data are available upon reasonable request. All the \ndata generated and analysed during this study are included in this published article \nand supplementary information file.\nSupplemental material This content has been supplied by the author(s). It has \nnot been vetted by BMJ Publishing Group Limited (BMJ) and may not have been \npeer- reviewed. Any opinions or recommendations discussed are solely those \nof the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and \nresponsibility arising from any reliance placed on the content. Where the content \nincludes any translated material, BMJ does not warrant the accuracy and reliability \nof the translations (including but not limited to local regulations, clinical guidelines, \nterminology, drug names and drug dosages), and is not responsible for any error \nand/or omissions arising from translation and adaptation or otherwise.\nOpen access This is an open access article distributed in accordance with the \nCreative Commons Attribution Non Commercial (CC BY- NC 4.0) license, which \npermits others to distribute, remix, adapt, build upon this work non- commercially, \nand license their derivative works on different terms, provided the original work is \nproperly cited, appropriate credit is given, any changes made indicated, and the use \nis non- commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.\nORCId ids\nHaiwei Zhang http://orcid.org/0009-0003-7757-3974\nHui Yang http://orcid.org/0000-0001-8485-1029\nHongwei Chen http://orcid.org/0000-0002-9400-6053\nLingyun Sun http://orcid.org/0000-0002-8563-2036\nRefeRences\n 1 Oktem O, Yagmur H, Bengisu H, et al. Reproductive aspects of \nsystemic lupus erythematosus. J Reprod Immunol 2016;117:57–65. \n 2 Hong L, Zhang Y , Wang Q, et al. Effects of interleukin 6 and tumor \nnecrosis factor-α on the proliferation of porcine theca interna cells: \nPossible role of these cytokines in the pathogenesis of polycystic \novary syndrome. Taiwan J Obstet Gynecol 2016;55:183–7. \n 3 Dabravolski SA, Nikiforov NG, Eid AH, et al. Mitochondrial \nDysfunction and Chronic Inflammation in Polycystic Ovary \nSyndrome. Int J Mol Sci 2021;22:3923. \n 4 Yang Y , Xia J, Yang Z, et al. The abnormal level of HSP70 is related to \nTreg/Th17 imbalance in PCOS patients. J Ovarian Res 2021;14:155. \n 5 Mende R, Vincent FB, Kandane- Rathnayake R, et al. Analysis \nof Serum Interleukin (IL)- 1β and IL- 18 in Systemic Lupus \nErythematosus. Front Immunol 2018;9:1250. \n 6 Ihim SA, Abubakar SD, Zian Z, et al. Interleukin- 18 cytokine in \nimmunity, inflammation, and autoimmunity: Biological role in \ninduction, regulation, and treatment. Front Immunol 2022;13:919973. \n 7 Suárez A, López P , Mozo L, et al. Differential effect of IL10 and \nTNF{alpha} genotypes on determining susceptibility to discoid and \nsystemic lupus erythematosus. Ann Rheum Dis 2005;64:1605–10. \n 8 Ayers CD, Carlson KS. Spontaneous Pregnancy in the Setting \nof Primary Ovarian Insufficiency and Breastfeeding: Does \nImmunosuppression Play a Role? Am J Case Rep 2020;21:e926980. \n 9 Rudnicka E, Suchta K, Grymowicz M, et al. Chronic Low Grade \nInflammation in Pathogenesis of PCOS. Int J Mol Sci 2021;22:3789. \n 10 Zhai Y , Pang Y . Systemic and ovarian inflammation in women with \npolycystic ovary syndrome. J Reprod Immunol 2022;151:103628. \n 11 Deroux A, Dumestre- Perard C, Dunand- Faure C, et al. Female \nInfertility and Serum Auto- antibodies: a Systematic Review. Clin Rev \nAllergy Immunol 2017;53:78–86. \n 12 Mobeen H, Afzal N, Kashif M. Polycystic Ovary Syndrome May Be an \nAutoimmune Disorder. Scientifica (Cairo) 2016;2016:4071735. \n 13 Mackiewicz Z, Sukura A, Povilenaité D, et al. Increased but \nimbalanced expression of VEGF and its receptors has no positive \neffect on angiogenesis in systemic sclerosis skin. Clin Exp Rheumatol \n2002;20:641–6.\n 14 Wree A, McGeough MD, Inzaugarat ME, et al. NLRP3 inflammasome \ndriven liver injury and fibrosis: Roles of IL- 17 and TNF in mice. \nHepatology 2018;67:736–49. \n 15 Chen W, Yuan H, Cao W, et al. Blocking interleukin- 6 trans- signaling \nprotects against renal fibrosis by suppressing STAT3 activation. \nTheranostics 2019;9:3980–91. \n 16 Zhang L, Xing R, Huang Z, et al. Inhibition of Synovial Macrophage \nPyroptosis Alleviates Synovitis and Fibrosis in Knee Osteoarthritis. \nMediators Inflamm 2019;2019:2165918. \n 17 Lliberos C, Liew SH, Zareie P , et al. Evaluation of inflammation and \nfollicle depletion during ovarian ageing in mice. Sci Rep 2021;11:278. \n 18 Amargant F , Manuel SL, Tu Q, et al. Ovarian stiffness increases \nwith age in the mammalian ovary and depends on collagen and \nhyaluronan matrices. Aging Cell 2020;19:e13259. \n 19 Umehara T, Winstanley YE, Andreas E, et al. Female reproductive life \nspan is extended by targeted removal of fibrotic collagen from the \nmouse ovary. Sci Adv 2022;8:eabn4564. \n 20 Landry DA, Yakubovich E, Cook DP , et al. Metformin prevents age- \nassociated ovarian fibrosis by modulating the immune landscape in \nfemale mice. Sci Adv 2022;8:eabq1475. \n 21 Khallaf WAI, Sharata EE, Attya ME, et al. LCZ696 (sacubitril/\nvalsartan) mitigates cyclophosphamide- induced premature ovarian \nfailure in rats; the role of TLR4/NF-κB/NLRP3/Caspase- 1 signaling \npathway. Life Sci 2023;326:121789. \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.\n\n\nZhang H, et al. Lupus Science & Medicine 2025;12:e001468. doi:10.1136/lupus-2024-001468 11\nReproductive health and APS\n 22 Deng D, Zhang P , Guo Y , et al. A randomised double- blind, placebo- \ncontrolled trial of allogeneic umbilical cord- derived mesenchymal \nstem cell for lupus nephritis. Ann Rheum Dis 2017;76:1436–9. \n 23 Yuan X, Qin X, Wang D, et al. Mesenchymal stem cell therapy \ninduces FLT3L and CD1c+ dendritic cells in systemic lupus \nerythematosus patients. Nat Commun 2019;10:2498. \n 24 Tyndall A, Houssiau FA. Mesenchymal stem cells in the treatment of \nautoimmune diseases. Ann Rheum Dis 2010;69:1413–4. \n 25 Xu J, Wang D, Liu D, et al. Allogeneic mesenchymal stem cell \ntreatment alleviates experimental and clinical Sjögren syndrome. \nBlood 2012;120:3142–51. \n 26 Garg K, Zilate S. Umbilical Cord- Derived Mesenchymal Stem Cells \nfor the Treatment of Infertility Due to Premature Ovarian Failure. \nCureus 2022;14:e30529. \n 27 Li Z, Zhang M, Tian Y , et al. Mesenchymal Stem Cells in Premature \nOvarian Insufficiency: Mechanisms and Prospects. Front Cell Dev \nBiol 2021;9:718192. \n 28 Zhang H, Luo Q, Lu X, et al. Effects of hPMSCs on granulosa cell \napoptosis and AMH expression and their role in the restoration of \novary function in premature ovarian failure mice. Stem Cell Res Ther \n2018;9:20. \n 29 Xie Q, Xiong X, Xiao N, et al. Mesenchymal Stem Cells Alleviate \nDHEA- Induced Polycystic Ovary Syndrome (PCOS) by Inhibiting \nInflammation in Mice. Stem Cells Int 2019;2019:9782373. \n 30 Cui L, Bao H, Liu Z, et al. hUMSCs regulate the differentiation of \novarian stromal cells via TGF-β\n1/Smad3 signaling pathway to inhibit \novarian fibrosis to repair ovarian function in POI rats. Stem Cell Res \nTher 2020;11:386. \n 31 Wei Q, Kong N, Liu X, et al. Pirfenidone attenuates synovial fibrosis \nand postpones the progression of osteoarthritis by anti- fibrotic \nand anti- inflammatory properties in vivo and in vitro. J Transl Med \n2021;19:157. \n 32 Sun B, Ma Y , Wang F , et al. miR- 644- 5p carried by bone \nmesenchymal stem cell- derived exosomes targets regulation of \np53 to inhibit ovarian granulosa cell apoptosis. Stem Cell Res Ther \n2019;10:360. \n 33 Jin J, Ma Y , Tong X, et al. Metformin inhibits testosterone- induced \nendoplasmic reticulum stress in ovarian granulosa cells via \ninactivation of p38 MAPK. Hum Reprod 2020;35:1145–58. \n 34 Kreeger PK, Deck JW, Woodruff TK, et al. The in vitro regulation of \novarian follicle development using alginate- extracellular matrix gels. \nBiomaterials 2006;27:714–23. \n 35 An R, Wang X, Yang L, et al. Polystyrene microplastics cause \ngranulosa cells apoptosis and fibrosis in ovary through oxidative \nstress in rats. Toxicology 2021;449. \n 36 Li R, Frangogiannis NG. Integrins in cardiac fibrosis. J Mol Cell \nCardiol 2022;172:1–13. \n 37 Shabanova SS, Ananieva LP , Alekberova ZS, et al. Ovarian function \nand disease activity in patients with systemic lupus erythematosus. \nClin Exp Rheumatol 2008;26:436–41.\n 38 Lawrenz B, Henes J, Henes M, et al. Impact of systemic lupus \nerythematosus on ovarian reserve in premenopausal women: \nevaluation by using anti- Muellerian hormone. Lupus (Los Angel) \n2011;20:1193–7. \n 39 Li Y , Zheng Q, Sun D, et al. Dehydroepiandrosterone stimulates \ninflammation and impairs ovarian functions of polycystic ovary \nsyndrome. Journal Cellular Physiology 2019;234:7435–47. \n 40 Bukulmez O, Arici A. Leukocytes in ovarian function. Hum Reprod \nUpdate 2000;6:1–15. \n 41 Ma H, Liu C, Shi B, et al. Mesenchymal Stem Cells Control \nComplement C5 Activation by Factor H in Lupus Nephritis. \nEBioMedicine 2018;32:21–30. \n 42 Yun JH. Autologous Mesenchymal Stem Cells Foster \nRevascularization of Ischemic Limbs in Systemic Sclerosis. Ann \nIntern Med 2011;155:65. \n 43 Cao M, Zhao Y , Chen T, et al. Adipose mesenchymal stem cell- \nderived exosomal microRNAs ameliorate polycystic ovary syndrome \nby protecting against metabolic disturbances. Biomaterials 2022;288. \n 44 Stocco C, Telleria C, Gibori G. The molecular control of corpus \nluteum formation, function, and regression. Endocr Rev \n2007;28:117–49. \n 45 Shen H, Wang Y . Activation of TGF-β1/Smad3 signaling pathway \ninhibits the development of ovarian follicle in polycystic ovary \nsyndrome by promoting apoptosis of granulosa cells. J Cell Physiol \n2019;234:11976–85. \n 46 Ge C, Tan J, Dai X, et al. Hepatocyte phosphatase DUSP22 \nmitigates NASH- HCC progression by targeting FAK. Nat Commun \n2022;13:5945. \n 47 Zhang P , Liu J, Zong R. Triptolide protects against TGF-β1- induced \npulmonary fibrosis by regulating FAK/calpain signaling. Exp Ther \nMed 2019;18:4781–9. \n 48 Chen H, Wen X, Liu S, et al. Dissecting Heterogeneity Reveals a \nUnique BAMBI\nhigh MFGE8high Subpopulation of Human UC- MSCs. \nAdv Sci (Weinh) 2022;10:e2202510.\n 49 Han X, Wang D, Chen L, et al. MSC transplantation ameliorates \ndepression in lupus by suppressing Th1 cell- shaped synaptic \nstripping. JCI Insight 2025;10:e181885. \n 50 Liu S, Liu D, Chen C, et al. MSC Transplantation Improves \nOsteopenia via Epigenetic Regulation of Notch Signaling in Lupus. \nCell Metab 2015;22:606–18. \nLupus Science & Medicine: first published as 10.1136/lupus-2024-001468 on 5 August 2025. Downloaded from https://lupus.bmj.com on 26 May 2026 by guest.\nProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies.","source_license":"CC0","license_restricted":false}