Abstract

   Systemic lupus erythematosus (SLE) is an autoimmune disorder
   characterized by aberrant T cell activity and excessive autoantibody
   production. Follicular helper T cells (Tfh) play a pivotal role in
   promoting B cell-mediated autoantibody generation, contributing to SLE
   progression. Although mesenchymal stem cell-derived exosomes (MSC-Exos)
   exhibit immunomodulatory properties, their effects on Tfh in SLE and
   the underlying mechanisms remain unclear. To address this, we first
   analyzed sorted Tfh from an imiquimod-induced lupus murine model
   (IMQ-SLE) and found that MSC-Exos effectively suppressed Tfh function.
   Consistently, Tfh polarization assays demonstrated that MSC-Exos
   modulate Tfh differentiation in vitro. Subsequently, we evaluated the
   therapeutic potential of intravenous MSC-Exos administration and
   confirmed that MSC-Exos markedly inhibited Tfh expansion and function
   in vivo. Further RNA sequencing followed by validation experiments
   identified that MSC-Exos restore calcium homeostasis in Tfh.
   Mechanically, MSC-Exos down-regulate stromal interaction molecule 1
   (Stim1) and Orai1 expression, inhibiting nuclear factor of activated T
   cells (NFAT) and nuclear factor κB (NF-κB) activation. In parallel,
   MSC-Exos mitigate calcium overload-induced mitochondrial damage by
   suppressing mitochondrial calcium uniporter (MCU) expression. Finally,
   we observed that MSC-Exos also promote the differentiation of
   follicular regulatory T cells (Tfr) both in vivo and in vitro. These
   findings suggest that MSC-Exos ameliorate SLE by correcting cellular
   calcium dysregulation and mitochondrial damage in Tfh while
   simultaneously restoring the Tfh/Tfr imbalance, highlighting their
   potential as a therapeutic strategy for SLE.

Introduction

   Systemic lupus erythematosus (SLE) is a chronic autoimmune disease
   characterized by the involvement of multiple organs and the excessive
   formation of autoantibodies, which is one of its key pathogenic
   mechanisms. Follicular helper T cells (Tfh), a critical subset of CD4^+
   T cells in lupus, are characterized by high expression of CXCR5, PD-1,
   interleukin-21 (IL-21), and the specific transcription factor BCL6
   [[44]1]. They provide essential signals for the survival, affinity
   maturation, differentiation, and proliferation of B cells within
   germinal centers (GCs), facilitating the secretion of high-affinity
   antibodies [[45]2]. Increased and overactivated Tfh in SLE are
   positively correlated with disease activity, and our previous study
   indicated that inhibiting Tfh expansion can significantly alleviate
   lupus symptoms [[46]3]. While the role of Tfh in SLE is well
   established, the upstream regulators of their hyperactivity and the
   targeted intervention remain unexplored.

   Stem cell therapy shows promise in treating autoimmune diseases owing
   to its immunoregulatory capacity. Mesenchymal stem cells (MSCs)
   effectively mitigate lupus-related inflammation by suppressing T helper
   1 (Th1)/Th17 cell activity and promoting regulatory T cell (Treg)
   differentiation [[47]4,[48]5]. Our prior research demonstrated that
   bone marrow-derived MSCs suppress Tfh differentiation and alleviate
   lupus nephritis [[49]3]. However, concerns regarding MSC therapy
   include potential tumorigenic risks, immune rejection, and unstable
   production quality control [[50]6]. Compared to MSCs, extracellular
   exosomes derived from MSCs (MSC-Exos) offer a cell-free therapeutic
   approach with enhanced biological safety and serve as primary mediators
   of MSC immunoregulatory functions. Accumulating evidence highlights the
   systemic anti-inflammatory effects of MSC-Exos across diverse
   autoimmune conditions. In neurological, gastrointestinal, and
   respiratory disorders, MSC-Exos effectively drive macrophage
   polarization toward anti-inflammatory phenotypes, thereby alleviating
   various autoinflammatory conditions [[51]7–[52]9]. In T cell studies,
   MSC-Exos restore immune equilibrium by reducing Th1/Th17 populations
   and expanding Tregs, ameliorating tissue inflammation in inflammatory
   bowel disease (IBD) and highlighting their potential to regulate T
   cell-mediated immune responses, although specific mechanisms were not
   elucidated [[53]10]. Another study reveals that MSC-Exos suppress the
   activated T cell proliferation through the induction of cell cycle
   arrest, without specifying the T cell subset involved. Beyond immune
   cells, emerging mechanistic studies further demonstrate that MSC-Exos
   exert the anti-inflammatory effect by mitigating oxidative stress,
   stabilizing mitochondria, and suppressing mitochondrial DNA (mtDNA)
   release, thereby alleviating intracellular inflammation
   [[54]11,[55]12]. Regarding lupus studies, MSC-Exos enhance
   efferocytosis and the anti-inflammatory M2 polarization of macrophages
   while inducing Treg cells, thus alleviating lupus progression [[56]13].
   Despite these advances, critical gaps persist: While current studies
   predominantly focus on the MSC-Exos-mediated regulation of macrophage
   polarization, Th17, and Treg cells in lupus, their ability to modulate
   Tfh—the key drivers of pathogenic autoantibody production—remains
   entirely unexplored, leaving a key therapeutic opportunity unaddressed.

   To bridge this gap, our study systematically investigates MSC-Exos’
   regulatory effects on Tfh and their underlying mechanisms.
   RNA-sequencing (RNA-seq) and bioinformatics analyses reveal that Tfh in
   SLE exhibit aberrant calcium signaling, which can be modulated by
   MSC-Exos. By simultaneously addressing cytosolic and mitochondrial
   calcium dysregulation, MSC-Exos inhibit Tfh hyperactivation in SLE.
   This dual mechanism not only highlights novel therapeutic targets for
   Tfh in lupus but also underscores the potential of MSC-Exos in lupus
   treatment.

   The immune dysregulation observed in SLE is not attributed to a single
   T cell subset but results from the combined imbalance among various T
   cell subsets. Similar to Th17/Treg imbalance, the Tfh/Tfr
   imbalance—resulting in excessive immune activation—has emerged as a
   pivotal contributor to SLE [[57]14]. Tfr, a specialized subset of Treg
   cells sharing phenotypic similarities with Tfh, exhibit high expression
   of Foxp3 and IL-10. Through CTLA-4, IL-10, and transforming growth
   factor-β (TGF-β) secretion, Tfr antagonize Tfh-mediated B cell help,
   thereby restraining autoantibody production [[58]14]. Conversely,
   Tfh-derived IL-21 suppresses Foxp3 expression in Tfr [[59]15], creating
   a self-reinforcing cycle where Tfh hyperproliferation further
   destabilizes Tfr differentiation. Clinical evidence confirms that
   elevated Tfh/Tfr ratios in SLE patients correlate with disease severity
   [[60]14], and strategy addressing this imbalance such as low-dose IL-2
   shows promise in disease remission [[61]16]. However, current
   interventions often lack cell type specificity, underscoring the need
   for targeted therapies capable of simultaneously correcting Tfh
   hyperactivity while boosting Tfr function. Addressing this unmet need,
   our study demonstrates that MSC-Exos exert a bidirectional regulatory
   effect on the Tfh/Tfr balance, exhibiting MSC-Exos’ unique ability to
   restore immune homeostasis, opening new doors for SLE therapy.

Materials and Methods

MSC cultivation

   Adipose-derived MSC (AD-MSC) cell lines, sourced from adult
   liposuction, were obtained from Oricell (HUXMD-01001, China) and
   cultured in Dulbecco’s modified Eagle’s medium/F-12 medium (Gibco,
   8120254, USA) supplemented with 15% calf serum devoid of exosomes
   (Umibio, UR50202, China), along with 100 U/ml penicillin and
   streptomycin (Gibco, 15140122, USA). The cells were seeded at a density
   of 1 × 10^6 cells in 10-cm tissue culture dishes (Corning, USA) and
   maintained at 37 °C in a 5% CO[2] humidified atmosphere.

Identification and uptake of MSC-Exos

   AD-MSCs were cultured for 48 to 72 h until cell fusion reached
   approximately 80% to 90%, after which the supernatants were collected
   for exosome extraction. Only MSCs within 5 passages were used for
   exosome extraction. Exosomes were isolated through differential
   centrifugation to remove cells and debris, followed by
   ultracentrifugation at 100,000g for 60 min at 4 °C using a Beckman
   Coulter superhigh-speed centrifuge (Optima XPN-100, USA). The exosomal
   pellets were washed in phosphate-buffered saline (PBS), centrifuged
   again at 100,000g for 60 min at 4 °C, and then resuspended in PBS
   before storage at −80 °C. The protein concentration of MSC-Exos was
   determined using a bicinchoninic acid (BCA) protein assay kit (Thermo,
   [62]A55861, USA), while the size distribution and particle
   concentration were assessed using a Nanosight NS300 instrument (Malvern
   Instruments, Worcestershire, UK) (Fig. [63]S1). Transmission electron
   microscopy (TEM) (Tecnai G2 Spirit, FEI, USA) was utilized to examine
   the morphology of AD-MSC exosomes (MSC-Exos) (Fig. [64]S2), and Western
   blotting (WB) confirmed the expression of surface markers such as CD63
   and Alix, and the negative marker Histone 3 was also verified (Fig.
   [65]S3).

SLE patients

   For this study, 28 patients diagnosed with SLE meeting the criteria set
   forth by the Revised Criteria of the American College of Rheumatology
   and the 2019 European League Against Rheumatism/American College of
   Rheumatology (EULAR/ACR) classification criteria for SLE were enrolled.
   Ethical approval was obtained from the Ethics Committee of Zhongshan
   Hospital, Fudan University, Shanghai, China (approval no. B2023-0862R),
   and written informed consent was obtained from all participants. The
   study adhered to the principles outlined in the Helsinki Declaration,
   and all participants provided both written and verbal informed consent
   prior to participation.

Mice

   Specific pathogen-free (SPF) Balb/C mice (weight, 18 to 22 g; age, 6
   weeks) were obtained from the Animal Experiment Center of Fudan
   University (Shanghai, China) and housed in an SPF colony at the Animal
   Center of Shanghai Medical School, Fudan University. All experimental
   procedures conducted on the mice followed the National Institutes of
   Health Guidelines on Laboratory Research and were approved by the
   Animal Care Committee of Huashan Hospital (approval 2023-HSYY-174JZS).

IMQ-SLE model construction

   Mice were randomly divided into 3 groups: imiquimod (IMQ)-SLE with PBS
   injection, IMQ-SLE with MSC-Exos injection, and WT (wild-type) groups.
   Six-week-old WT mice were subjected to IMQ treatment according to
   established protocols to induce a lupus-like phenotype [[66]17,[67]18].
   Specifically, 5% IMQ cream was applied topically to the inner ear 3
   times a week for 10 weeks. Mice were sacrificed 14 d after injection.
   Spleens were isolated and mechanically dissociated through a 70-μm
   sterile filter to obtain single-cell suspensions.

Isolation of Tfh in IMQ-SLE

   Tfh were sorted using EasySep Naive CD4^+ T cell Isolation Kit
   (Stemcell, 19765, CA), allophycocyanin (APC)-conjugated anti-CXCR5
   (Invitrogen, 17-7185-82, USA), and EasySep APC Positive Selection Kit
   II (Stemcell, 17667, CA). The purity of sorted CD4^+ T cells was
   greater than 98% (Fig. [68]S4), and the purity of sorted CXCR5^+ T
   cells was greater than 60%. Sorted naive CD4^+ T cells were suspended
   in RPMI 1640 medium (GIBCO, 11875093, USA) supplemented with 100 U/ml
   penicillin, 100 mg/ml streptomycin, 0.05 mM β-mercaptoethanol, and 15%
   fetal bovine serum (Hyclone, SH30070, CA).

Induction of T follicular helper cell differentiation in vitro

   Naive CD4^+ T cells were isolated from the spleens of Balb/C mice using
   the EasySep Naive CD4^+ T Cell Isolation Kit (Stemcell, 19765, CA),
   achieving a purity exceeding 98% (Fig. [69]S5). These sorted cells were
   suspended in RPMI 1640 medium (GIBCO, 11875093, USA) supplemented with
   100 U/ml penicillin, 100 mg/ml streptomycin, 0.05 mM β-mercaptoethanol,
   and 15% fetal bovine serum (Hyclone, SH30070, CA). For Tfh
   polarization, the medium was additionally supplemented with 20 ng/ml
   IL-6, 10 μg/ml anti-interferon-γ (IFN-γ), 10 μg/ml anti-IL-4, 10 μg/ml
   anti-IL-2, and 10 μg/ml anti-TGF-β (PeproTech-BioGems, USA). Cell
   suspensions containing 2 × 10^6 cells were then seeded onto a 96-well
   plate precoated with 8 mg/ml anti-CD3 (PeproTech-BioGems, 05112-25-500,
   USA) and 8 mg/ml anti-CD28 (PeproTech-BioGems, 10312-25-500, USA) (Fig.
   [70]S6). Cells were cultured in a 5% CO[2] atmosphere for 72 h before
   analysis.

Flow cytometry analysis

   Flow cytometry was used to evaluate the expression of surface markers
   and intracellular molecules using BD FACSCanto II and FACSAria III flow
   cytometers (BD Biosciences, NJ, USA). For intracellular cytokine
   staining, cells were stimulated for 6 h with Cell Stimulation Cocktail
   Plus Protein Transport Inhibitors (ESscience Biotech, ECS005, ECS006,
   China). Intracellular staining of BCL6, IL-21, IL-10, and Foxp3 was
   performed using a Foxp3 staining kit (eBioscience, 2258710, USA). The
   following antibodies were used: Zombie Green Fixable Viability Kit
   (BioLegend, 423111, USA), APC-conjugated anti-CXCR5 (BioLegend, 145506,
   USA), phycoerythrin (PE)-conjugated anti-PD-1 (BioLegend, 135253, USA),
   PE/cyanine7-conjugated anti-PD-1 (BioLegend, 135215, USA),
   PE-conjugated Bcl-6 (BioLegend, 358504, USA), and
   PE/cyanine7-conjugated Foxp3 (Invitrogen, 2833598, USA). Data analysis
   was carried out using FlowJo software.

Real-time quantitative polymerase chain reaction analysis

   Total RNA was extracted from the cells using the RNA-Quick Purification
   Kit (ESscience Biotech, RN001, China) following the manufacturer’s
   protocol. Total RNA (1 μg) was reverse-transcribed into cDNA in a 10-μl
   reaction volume using the PrimeScript RT Reagent Kit (Takara, RR037A,
   Japan). The relative expression levels of BCL6, IL-21, CXCR5, IL-10,
   Foxp3, CaN, NFATc1, NFATc2, Orai1, Stim1, and Stim2 were assessed using
   real-time quantitative polymerase chain reaction (RT-qPCR) at the RNA
   level. For PCR amplification, 1.5 μl of cDNA was used as the template.
   The QuantStudio 6 RT-qPCR instrument (Thermo, USA) facilitated the
   reaction with a cycling program of 95 °C for 60 s, followed by 40
   cycles of 95 °C for 15 s and 60 °C for 34 s. Each sample was run in
   triplicate and normalized to the level of β-actin. The relative mRNA
   levels were determined using the 2−ΔΔCt method and represented as fold
   change compared with the control group. The primer sequences used for
   mRNA amplification are provided in Table [71]S1.

WB analysis

   Cell lysates were prepared by treating cells with cell lysis buffer
   (Beyotime, P0013C, China) supplemented with protease and phosphatase
   inhibitors (Beyotime, P1045, China). The supernatant, collected after
   centrifugation at 12,000g for 20 min at 4 °C, was used to determine
   protein concentration using a BCA protein assay kit (Thermo,
   [72]A55861, USA). Proteins were then separated by sodium dodecyl
   sulfate–polyacrylamide gel electrophoresis and transferred onto
   polyvinylidene fluoride membranes. These membranes were blocked with 5%
   nonfat milk at 25 °C for 1 h. Primary antibodies targeting CaN
   (Proteintech, 13422-1-AP, China), NFATc2 (Proteintech, 22023-1-AP,
   China), P65 (Affinity, AF5006, USA), p-P65 (Affinity, AF2006, USA), IκB
   (Abmart, [73]T55026, China), p-IκB (Abmart, TA2002, China), Stim1
   (Proteintech, 11565-1-AP, China), Orai1 (Proteintech, 28411-1-AP,
   China), and MCU (Abways, BY0101, China) were incubated overnight at 4
   °C. Chemiluminescence detection was performed using enhanced
   chemiluminescence reagents, and images were captured using film
   cassette exposure (TANON Science & Technology, China).

ELISA analysis

   Enzyme-linked immunosorbent assay (ELISA) assays were conducted using
   the following kits: the Mouse anti-double-stranded DNA (dsDNA)
   immunoglobulin G (IgG)-specific ELISA Kit (Alpha Diagnostic
   International, 5120, USA), the Mouse Anti-Nuclear Antigens (ANA/ENA)
   Ig’s (total IgA, IgG, and IgM) ELISA Kit (Alpha Diagnostic
   International, 5210, USA), and the Mouse Complement Component 3 (C3)
   ELISA Kit (Alpha Diagnostic International, 6270, USA). All procedures
   were performed according to the manufacturer’s instructions.

Measurement of the Ca^2+ concentration

   The following probes were used: Fluo-4 AM (YEASEN, 40704ES50, China)
   and Rhod-2 AM (YEASEN, 40776ES50, China). Intracellular Ca^2+
   concentration was measured using a Ca^2+ probe Fluo-4 AM according to
   the manufacturer’s protocol. Mitochondrial Ca^2+ were monitored in live
   cells using Rhod-2 AM according to the manufacturer’s protocol. Tfh
   were loaded with 4 μM Fluo-4 AM or Rhod-2 AM in serum-free medium for
   30 min at 37 °C, followed by washes with Hanks’ balanced salt solution
   (HBSS) buffer to remove excess dye. After washing, the cells were
   incubated with HBSS for 30 min before testing. Flow cytometry was used
   to evaluate the Ca^2+ concentration using BD FACSCanto II (BD
   Biosciences, NJ, USA), with Fluo-4 exhibiting increased green
   fluorescence intensity in positive cells, and Rhod-2 exhibiting
   increased red fluorescence intensity in positive cells. Data analysis
   was carried out using FlowJo software.

Measurement of the ROS, MMP, and mPTP

   Intracellular reactive oxygen species (ROS) levels were measured by
   DCFH-DA (Beyotime, S0033S, China). Mitochondrial membrane potential
   (MMP, ΔΨm) was measured by JC-1 assay kit (Beyotime, C2006, China).
   Mitochondrial permeability transition pore (mPTP) opening levels were
   measured by calcein (Abbkine, BMD0064, China). Tfh were stained with
   DCFH-DA, JC-1, or calcein according to their manufacturer’s manual and
   analyzed by BD FACSCanto II (BD Biosciences, NJ, USA), respectively.

RNA sequencing

   Naïve CD4^+ T cells were collected from 3 WT, then cultured for 72 h
   under Tfh polarizing conditions, and intervened with MSC-Exos (n = 3)
   or PBS (n = 3). Three sets of paired cell samples were collected for
   RNA-seq. The data that support the findings of this study have been
   deposited into CNGB Sequence Archive (CNSA) [[74]19] of China National
   GeneBank DataBase (CNGBdb) with accession number CNP0006973. All
   relevant data are available upon reasonable request.

TEM analysis

   Cells were fixed with TEM fixer (Solarbio, P1126, China) at 4 °C for 12
   h, followed by pre-embedding in 1% agarose and immobilization in 1%
   osmium tetroxide. After ethanol dehydration at room temperature,
   samples were embedded in Poly/Bed 812 resin and polymerized at 65 °C.
   Ultrathin sections were stained with 2% uranyl acetate and 2.6% lead
   citrate, and analyzed using a transmission electron microscope
   (Hitachi, Japan).

Statistical analysis

   Statistical data were expressed as mean ± standard error of the mean. A
   two-tailed unpaired Student’s t test was used to evaluate differences
   between two groups. Multiple comparisons were conducted using one-way
   analysis of variance (ANOVA) with Tukey’s multiple comparisons test.
   Statistical significance was determined at a P value of <0.05. All
   statistical analyses were performed using GraphPad Prism version 8.0
   (GraphPad Software, USA).

Results

Tfh expansion accompanied by Tfh/Tfr imbalance in lupus

   The lupus mouse model induced by the Toll-like receptor 7 (TLR7)
   agonist IMQ effectively mimics various clinical manifestations of lupus
   [[75]17,[76]18]. In our study, we administered IMQ to 6-week-old female
   Balb/C mice over a 10-week period (Fig. [77]1A). Post-induction, these
   IMQ-treated mice exhibited significant splenomegaly (Fig. [78]1B). Tfh
   were identified by surface markers, including high expression of CXCR5
   and PD-1, along with significant levels of intracellular BCL6 and
   IL-21. Flow cytometry analysis revealed a significant increase in the
   proportion of CD4^+CXCR5^+PD-1^+ cells in the spleens of IMQ-SLE mice
   compared to WT controls (Fig. [79]1C and D). Tfr shared similar surface
   markers with Tfh but also expressed Foxp3; thus, we also investigated
   the frequency of Tfr simultaneously. Conversely, the proportion of
   CD4^+CXCR5^+PD-1^+Foxp3^+ Tfr was significantly reduced, leading to an
   inverse relationship between Tfh and Tfr (Fig. [80]1C, E, and I).

Fig. 1.

   Fig. 1.
   [81]Open in a new tab

   The imbalance of Tfh/Tfr in IMQ-SLE mice and lupus patients. (A)
   Establishment process of the IMQ-SLE model. (B) Comparison of spleen
   sizes between WT and IMQ-SLE. (C) Murine CXCR5^+PD-1^+ cells were gated
   on CD4^+ cells, and Foxp3^− Tfh and Foxp3^+ Tfr cells were gated on
   CXCR5^+PD-1^+ cells in WT and IMQ-SLE mice. (D) Splenic
   CD4^+CXCR5^+PD-1^+ frequencies in IMQ-SLE mice (n = 8) versus WT (n =
   7). (E) Proportions of Foxp3^− Tfh and Foxp3^+ Tfr gated on
   CXCR5^+PD-1^+ cells in WT (n = 7) versus IMQ-SLE (n = 8) mice. (F)
   Human CXCR5^+PD-1^+ cells were gated on CD4^+ cells, and Tfh
   (CD25^−CD127^High) and Tfr (CD25^+CD127^Low) cells were gated on
   CXCR5^+PD-1^+ cells in healthy controls (HC) and SLE patients. (G)
   Peripheral CXCR5^+PD-1^+ cell frequencies in HC (n = 24) versus SLE
   patients (n = 23). (H) Proportions of CD25^−CD127^High Tfh and
   CD25^+CD127^Low Tfr in HC (n = 24) versus SLE patients (n = 23). (I)
   Tfh accumulation and Tfr depletion in SLE mice and patients. (J)
   Frequencies of Tfh and Tfr and their correlation in patients with
   different SLEDAI scores. Bars indicate the means ± SEM. *P < 0.05, **P
   < 0.01, ***P < 0.001, ****P < 0.0001 determined by 2-tailed unpaired
   Student’s t test.

   Consistent with murine findings, SLE patients exhibited elevated
   peripheral CXCR5^+PD-1^+ cell frequencies versus healthy controls (Fig.
   [82]1F and G). Subpopulation analysis defined Tfh (CD25^−CD127^High)
   and Tfr (CD25^+CD127^Low), demonstrating significant Tfh expansion and
   Tfr depletion in SLE patients (Fig. [83]1F, H, and I). Clinically, Tfh
   abundance exhibited a significant positive correlation with Systemic
   Lupus Erythematosus Disease Activity Index (SLEDAI) scores, while Tfr
   levels inversely correlated with disease activity (Fig. [84]1J). These
   results demonstrate an imbalance of Tfh/Tfr in SLE, indicating that the
   restoration of this imbalance could serve as a potential therapeutic
   target for SLE.

MSC-Exos restrain the differentiation and function of Tfh both in vitro and
in vivo

   In this study, we utilized exosomes derived from AD-MSCs, referred to
   as MSC-Exos. To assess the potential of MSC-Exos in inhibiting Tfh, we
   first conducted in vitro experiments. Using magnetic-activated cell
   sorting (MACS), we isolate CD4^+ T cells from spleens and further
   enriched for Tfh by selecting those with high CXCR5 expression,
   achieving a purity exceeding 60%, indicative of functionally mature Tfh
   in IMQ-SLE (Fig. [85]2A). Subsequently, we cocultured these sorted Tfh
   with MSC-Exos at different concentrations for 24 h. The concentration
   gradient of MSC-Exos was as follows: low concentration: 0.8 × 10^8
   particles/ml; medium concentration: 1.6 × 10^8 particles/ml; high
   concentration: 3.2 × 10^8 particles/ml. We then examined IL-21
   expression in sorted Tfh after coculture with MSC-Exos or PBS. Results
   showed that the medium concentration of MSC-Exos significantly
   suppressed IL-21 expression, while low and high concentrations showed
   no significant difference (Fig. [86]2B and C). Consequently, the
   moderate concentration of MSC-Exos was chosen for subsequent
   experiments (1.6 × 10^8 particles/ml). We also assessed BCL6 expression
   in Tfh cocultured with MSC-Exos, which showed a significant reduction
   (Fig. [87]2D). CXCR5 also decreased notably, along with a decrease in
   CXCR5^+IL-21^+ double-positive cells (Fig. [88]2E and F). To determine
   if MSC-Exos have similar effects on Tfh from lupus patients, peripheral
   blood mononuclear cells (PBMCs) were isolated and cocultured with
   MSC-Exos. Flow cytometry indicated that MSC-Exos could also inhibit Tfh
   derived from lupus patients in vitro (Fig. [89]2G). These results
   indicated that MSC-Exos significantly suppress the function of Tfh in
   IMQ-SLE mice and lupus patients, as evidenced by reduced secretion of
   IL-21 by Tfh and decreased expression levels of BCL6 and CXCR5.

Fig. 2.

   Fig. 2.
   [90]Open in a new tab

   MSC-Exos inhibited the function of mature Tfh cells sorted from IMQ-SLE
   in vitro. (A) Isolation of mature Tfh in IMQ-SLE (n = 14). (B and C)
   Mature Tfh from IMQ-SLE (n = 14) were cocultured with varying
   concentrations of MSC-Exos for 24 h, and IL-21 secretion levels were
   detected using flow cytometry. (D) MSC-Exos were cocultured with mature
   Tfh from IMQ-SLE (n = 5) for 24 h, and BCL6 expression levels were
   measured using flow cytometry. (E) MSC-Exos were cocultured with mature
   Tfh from IMQ-SLE (n = 7) for 24 h, and CXCR5 expression levels were
   examined using flow cytometry. (F) MSC-Exos were cocultured with mature
   Tfh from IMQ-SLE (n = 7) for 24 h, and the proportion of IL-21^+CXCR5^+
   cells was assessed using flow cytometry. (G) MSC-Exos were cocultured
   with PBMCs of SLE patients (n = 8) for 24 h, and the proportion of
   CD4^+PD-1^+CXCR5^+ cells was evaluated using flow cytometry. Bars
   indicate the means ± SEM. ns, not significant; *P < 0.05; **P < 0.01;
   ***P < 0.001; ****P < 0.0001 determined by t test or one-way ANOVA with
   Tukey’s multiple comparisons test or 2-tailed unpaired Student’s t
   test.

   To investigate the impact of MSC-Exos on differentiation of Tfh, we
   induced Tfh under Tfh polarization condition in vitro. First, naïve
   CD4^+ T cells were isolated followed by in vitro Tfh polarization for
   72 h (Fig. [91]S5 and [92]S6). The condition for Tfh polarization
   included 50 ng/ml IL-6, 10 μg/ml anti-IFN-γ, 10 μg/ml anti-IL-4, 10
   μg/ml anti-TGF-β, and 5 μg/ml anti-IL-2 [[93]20,[94]21]. Concurrently,
   naïve CD4^+ T cells under nonpolarization conditions (Th0) were used as
   a negative control. After 72 h, Tfh expressing high levels of PD-1 and
   CXCR5 were stably induced in vitro (Fig. [95]3A). Based on previous
   results, we cocultured the Tfh with a medium concentration (1.6 × 10^8
   particles/ml) of MSC-Exos under polarization conditions to achieve
   optimal inhibitory effects. The results showed that MSC-Exos
   significantly reduced the proportion of CXCR5^+PD-1^+ Tfh and IL-21
   expression in vitro (Fig. [96]3A to D). Furthermore, RT-qPCR analysis
   revealed that MSC-Exos significantly suppressed the elevated mRNA
   levels of IL-21 and BCL6 in the Tfh polarization system, while CXCR5
   expression showed no statistically significant difference (Fig.
   [97]3E). These findings collectively indicate that MSC-Exos effectively
   suppress Tfh differentiation in vitro.

Fig. 3.

   Fig. 3.
   [98]Open in a new tab

   MSC-Exos suppressed Tfh differentiation in vitro. (A and B) Naïve CD4^+
   T cells were extracted from WT Balb/C mice and cultured under Tfh
   polarization conditions for 72 h with CD3/28 stimulation in vitro, with
   coculture of MSC-Exos or PBS as control. Th0 represents naïve CD4 T
   cells cultured under CD3/28 stimulation for 72 h, without Tfh
   polarization conditions. The proportion of CXCR5^+PD-1^+ Tfh cells was
   determined using flow cytometry. (C and D) IL-21 expression was
   analyzed in the Tfh-induced differentiation system after coculture with
   MSC-Exos or PBS by flow cytometry. (E) IL-21, BCL6, and CXCR5 mRNA
   expression levels were analyzed in the Tfh polarization condition
   system after coculture with MSC-Exos or PBS by RT-qPCR. Bars indicate
   the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001
   determined by one-way ANOVA with Tukey’s multiple comparisons test.

   To evaluate the in vivo impact of MSC-Exos on Tfh in the IMQ-SLE model,
   we intravenously administered 100 μg of MSC-Exos (2.25 × 10^9
   particles) to IMQ-SLE mice, with PBS-treated mice serving as controls.
   After 14 d, all mice were sacrificed for further analysis (Fig.
   [99]4A). The MSC-Exos-treated group exhibited a significantly lower
   spleen index compared to the PBS control group, suggesting a reduction
   in splenomegaly (Fig. [100]4B and C). Histological analysis of kidney
   sections showed that MSC-Exos reduced glomerular capillary
   intracellular proliferation, indicating potential reno-protective
   effects. (Fig. [101]4D). Serological analysis via ELISA demonstrated
   that MSC-Exos significantly decreased serum levels of anti-dsDNA and
   anti-nuclear autoantibodies (ANA) while concurrently increasing serum
   complement C3 levels compared to the PBS control group, suggesting an
   improvement in autoimmune pathology (Fig. [102]4E to G).

Fig. 4.

   Fig. 4.
   [103]Open in a new tab

   IMQ-SLE mice received an intravenous administration of 100 μg of
   MSC-Exos (2.25 × 10^9 particles) and were sacrificed after 14 d. (A)
   Serum was collected to measure anti-dsDNA antibody, anti-ANA antibody,
   and C3 levels. Kidneys were harvested for periodic acid–Schiff (PAS)
   staining. Spleens were extracted to evaluate the inhibitory effect of
   MSC-Exos on Tfh in vivo. (B and C) Representative spleens from each
   group. MSC-Exos treatment inhibited spleen enlargement and reduced the
   spleen index (n = 6). (D) PAS staining of kidney samples. Scale bar, 50
   μm. (E to G) MSC-Exos treatment reduced serum anti-dsDNA and anti-ANA
   antibody levels and increased serum C3 levels (n = 3). (H to J)
   Proportions and mean fluorescence intensity (MFI) of IL-21% cells in WT
   (n = 6), IMQ-SLE + PBS (n = 3), and IMQ-SLE + Exo (n = 6) mice. (K and
   L) Proportions of BCL6^+CXCR5^+ % cells in WT (n = 6), IMQ-SLE + PBS (n
   = 6), and IMQ-SLE + Exo (n = 6) mice. Bars indicate the means ± SEM. *P
   < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 determined by one-way
   ANOVA with Tukey’s multiple comparisons test.

   To further assess the impact of MSC-Exos on Tfh in vivo, CD4^+ T cells
   were extracted from the spleens and Tfh populations were quantified
   using flow cytometry. Results indicated a notable reduction in IL-21
   expression levels and a significant decrease in the proportion of
   CXCR5^+BCL6^+ cells among CD4^+ T cells in the MSC-Exos-treated group,
   compared to the PBS-treated controls (Fig. [104]4H, I, K, and L).

   Collectively, these findings indicate that MSC-Exos have significant
   therapeutic effects on IMQ SLE mice, and further suggest that MSC-Exos
   exert therapeutic effects on lupus mice by inhibiting Tfh.

MSC-Exos regulate calcium homeostasis in Tfh via Stim1–Orai1 signaling

   To further elucidate the specific mechanisms by which MSC-Exos regulate
   Tfh, we performed RNA-seq on MSC-Exos-treated Tfh cells, with
   PBS-treated cells serving as control. Differentially expressed gene
   (DEG) analysis followed by Kyoto Encyclopedia of Genes and Genomes
   (KEGG) pathway enrichment analysis revealed significant enrichment of
   the calcium signaling pathway, alongside notable enrichment of pathways
   associated with SLE (Fig. [105]5A and B). Additionally, Gene Ontology
   (GO) enrichment analysis of dataset [106]GSE157648 from the Gene
   Expression Omnibus (GEO) database comparing Tfh to naïve CD4^+ T cells
   demonstrated significant enrichment of calcium-related pathways (Fig.
   [107]5C). These findings suggest that the calcium signaling pathway is
   markedly activated in Tfh, and MSC-Exos may exert their regulatory
   effects on Tfh by modulating calcium homeostasis.

Fig. 5.

   Fig. 5.
   [108]Open in a new tab

   MSC-Exos regulate calcium homeostasis in Tfh. (A) Volcano diagram of
   DEGs of Tfh between control group and Exo group (n = 3). (B) KEGG
   enrichment analysis of Tfh between control group (n = 3) and Exo group
   (n = 3). These pathways were significant differences in the diagram.
   (C) GO enrichment analysis of DEGs between Tfh and naïve CD4^+ T cells
   in C57 based on the [109]GSE157648 dataset. (D) Fluo-4 AM-loaded Th0
   and Tfh treated with/without MSC-Exos in HBSS medium for 30 min, and
   intracellular Ca^2+ levels were detected by flow cytometry using Fluo-4
   AM as the probe (n = 5). (E) mRNA expression levels of Stim1, Stim2,
   and Orai1 in Tfh with PBS or MSC-Exos coculture were detected by
   RT-qPCR. (F) WB analysis of Stim1 and Orai1 protein expression in Th0
   and Tfh after treatment with MSC-Exos or PBS. (G) WB analysis of Stim1
   and Orai1 protein expression in Tfh treated with/without MSC-Exos, and
   ionomycin (5 μM) was used to elevate intracellular Ca^2+ levels. (H)
   RT-qPCR analysis of NFATc1, NFATc2, and CaN mRNA expression levels in
   Tfh after treatment with MSC-Exos or PBS. (I) WB analysis of CaN and
   NFATc2 protein expression in Tfh treated with/without MSC-Exos. (J) WB
   analysis of IκB, p-IκB, P65, and p-P65 protein expression in Tfh
   treated with/without MSC-Exos. Bars indicate the means ± SEM. *P <
   0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 determined by one-way
   ANOVA with Tukey’s multiple comparisons test.

   The calcium-related pathway plays a crucial role in CD4^+ T
   cell-mediated autoimmune responses [[110]22]. To investigate calcium
   dynamics in Tfh, we utilized Fluo-4 staining to measure intracellular
   Ca^2+ concentrations. The results revealed a significant elevation in
   Ca^2+ levels in Tfh compared to Th0. Notably, treatment with MSC-Exos
   markedly reduced intracellular calcium levels (Fig. [111]5D),
   suggesting a regulatory role of MSC-Exos in calcium homeostasis.

   The Stim1–Orai1 signaling is a critical regulator of intracellular
   calcium homeostasis. Upon T cell receptor (TCR) stimulation,
   endoplasmic reticulum (ER) calcium stores are depleted, leading to the
   activation and translocation of Stim1 to ER–plasma membrane junctions
   [[112]23]. Stim then binds to and opens Orai1 channels, facilitating
   extracellular calcium influx. To validate the regulatory effect of
   MSC-Exos on the Ca^2+ upstream signaling, we first assessed the mRNA
   levels of the key initiators: Stim1, Stim2, and Orai1. RT-qPCR results
   demonstrated that Stim1 mRNA was significantly up-regulated in Tfh,
   while MSC-Exos treatment markedly reduced Stim1, Stim2, and Orai1 (Fig.
   [113]5E). These findings were further confirmed by WB analysis (Fig.
   [114]5F). Ionomycin, a commonly used stimulator of intracellular
   calcium signaling, was subsequently employed to further validate the
   effects of MSC-Exos as positive control. WB results revealed that the
   down-regulation of Stim1 and Orai1 by MSC-Exos was attenuated in the
   presence of ionomycin (Fig. [115]5G). Collectively, these results
   indicate that MSC-Exos can modulate the activation of the Ca^2+
   upstream signaling in Tfh through Stim1 and Orai1.

   Elevated cytoplasmic calcium levels activate calcineurin (CaN), a
   serine/threonine phosphatase, which dephosphorylates and activates
   downstream targets—NFAT [[116]22]. Dephosphorylated NFAT translocates
   to the nucleus, where it initiates transcription of genes essential for
   T cell proliferation, differentiation, and pro-inflammatory cytokine
   production [[117]23]. We first confirmed the up-regulation of CaN and
   NFATc2 in lupus Tfh in public dataset [118]GSE157648 (Fig. [119]S7). To
   investigate the effects of MSC-Exos on the Ca^2+ downstream signaling
   in Tfh, we quantified CaN and NFAT levels. Tfh exhibited significantly
   higher CaN and NFATc2 mRNA levels compared to Th0. MSC-Exos treatment
   reduced CaN expression by >50% and normalized NFATc2 to Th0 levels,
   whereas NFATc1 exhibited minimal variation (Fig. [120]5H). WB confirmed
   parallel reductions in CaN and NFATc2 protein levels induced by
   MSC-Exos (Fig. [121]5I).

   Nuclear factor κB (NF-κB) is also located on the downstream of the
   calcium signaling pathway, sharing similar stimuli with NFAT by Ca^2+
   [[122]24]. In resting T cells, the NF-κB subunit P65 (RelA) binds to
   inhibitory IκB proteins. TCR stimulation triggers IκB kinase
   (IKK)-mediated IκB phosphorylation, leading to its proteasomal
   degradation and subsequent P65 nuclear translocation to drive
   inflammatory gene expression [[123]25]. This classical process in Tfh
   is responsible for up-regulating CXCR5, IL-21, and ICOS expression,
   contributing to Tfh differentiation and function [[124]26]. Given the
   interplay between NFAT and NF-κB pathways in immune regulation
   [[125]27], we further investigated NF-κB signaling. By analyzing the
   [126]GSE31702 from GEO database, we confirmed the significant
   activation of the NF-κB pathway in Tfh from SLE compared to those from
   healthy controls (Fig. [127]S7). To determine whether MSC-Exos also
   regulate on this downstream calcium signaling, we examined the
   associated protein expression. Tfh showed a significant increase in IκB
   protein levels compared to Th0. In contrast, coculturing with MSC-Exos
   led to a notable decrease in IκB phosphorylation levels (Fig. [128]5J).
   Similarly, the phosphorylation of P65 in Tfh significantly decreased
   after coculture with MSC-Exos, while total P65 protein levels remained
   unchanged (Fig. [129]5J). These results suggest that MSC-Exos stabilize
   IκB, thereby inhibiting the dissociation of P65 from IκB, reducing the
   activation level of P65 phosphorylation and preventing its nuclear
   translocation. Collectively, MSC-Exos exert an inhibitory effect on Tfh
   differentiation and function by regulating the Ca^2+-mediated NFAT and
   NF-κB pathway.

MSC-Exos alleviate mitochondrial damage by inhibiting MCU-induced calcium
overload

   Intracellular calcium homeostasis is crucial for the mitochondrial
   quality control. For mitochondria, calcium is a double-edged sword: Low
   levels are essential for optimal adenosine triphosphate (ATP)
   production, while calcium overload impairs its function [[130]28]. We
   have demonstrated that Tfh exhibited disrupted calcium homeostasis,
   which was restored by MSC-Exos via inhibition of Stim1/Orai1-mediated
   calcium signaling. To further explore the consequences of calcium
   dysregulation, we employed electron microscopy, which revealed
   significant mitochondrial morphological abnormalities and increased
   autophagy in Tfh. Notably, these pathological changes were markedly
   alleviated by MSC-Exos treatment (Fig. [131]6A). Given the close
   association between mitochondria and calcium homeostasis, we further
   investigated the impact of MSC-Exos on mitochondrial calcium
   regulation.

Fig. 6.

   Fig. 6.
   [132]Open in a new tab

   MSC-Exos alleviate mitochondrial damage in Tfh by inhibiting
   MCU-induced calcium overload. (A) Representative TEM images of
   mitochondria in Th0 and Tfh, with coculture of MSC-Exos or PBS as
   control. Scale bars, 200 nm. (B) Rhod-2 AM-loaded Th0 and Tfh treated
   with/without MSC-Exos in HBSS medium for 30 min, and mito-Ca^2+ influx
   rates were detected by flow cytometry using Rhod-2 AM as the probe. Tfh
   was stimulated with anti-CD3-biotin (5 μg/ml). Ionomycin (5 μM) was
   used to increase mito-Ca^2+ levels. (C) Measurements of mito-Ca^2+
   levels in Tfh after 60-s stimulation of anti-CD3-biotin (5 μg/ml) using
   Rhod-2 AM as the probe. Ionomycin (5 μM) was used to increase
   mito-Ca^2+ levels (n = 10). (D) Representative flow cytometric analysis
   of ROS production in Tfh using DCFH-DA as the probe. Ionomycin (5 μM)
   was used to increase mito-Ca^2+ levels (n = 10). (E) Representative
   flow cytometric analysis of mPTP opening levels in Tfh using calcein as
   the probe. The higher degree of mPTP opening, the lower calcein
   fluorescence intensity (n = 10). (F) JC-1 assay evaluation of the MMP
   level of Th0 and Tfh treated with/without MSC-Exos. Ionomycin (5 μM)
   was used to increase mito-Ca^2+ levels (n = 10). (G and H) WB analysis
   of MCU protein expression in Tfh treated with/without MSC-Exos, and
   ionomycin (5 μM) was used to elevate intracellular Ca^2+ levels. Bars
   indicate the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P <
   0.0001 determined by one-way ANOVA with Tukey’s multiple comparisons
   test.

   First, we monitored the mitochondrial calcium influx rate using Rhod-2
   AM, which is widely used to test the mitochondrial Ca^2+ [[133]29].
   Results revealed that MSC-Exos treatment significantly suppressed the
   mitochondrial Ca^2+ influx rate in Tfh (Fig. [134]6B). Additionally,
   MSC-Exos effectively inhibited the ionomycin-induced increase in
   mitochondrial Ca^2+ influx rate, even in the ionomycin-treated group
   (Fig. [135]6B). Subsequent measurements showed that mitochondrial Ca^2+
   levels in Tfh were significantly elevated compared with Th0, and
   ionomycin further accentuated this difference. Notably, MSC-Exos
   effectively prevented mitochondrial calcium overload in Tfh in the
   presence and absence of ionomycin (Fig. [136]6C).

   Mitochondrial calcium overload leads to a range of mitochondrial damage
   [[137]28]. Given the pivotal role of mitochondria as cellular
   powerhouses, their dysfunction directly contributes to oxidative
   stress. To investigate this, we assessed ROS levels in Tfh. Flow
   cytometry analysis demonstrated that, as expected, Tfh exhibited
   significantly higher ROS levels than Th0. Notably, treatment with
   MSC-Exos markedly reduced ROS levels, even after ionomycin
   treatment—which further increases mitochondrial Ca^2+—MSC-Exos
   continued to alleviate oxidative stress effectively (Fig. [138]6D).
   Furthermore, we evaluated the quality of mitochondria by examining the
   mitochondrial permeability transition pore (mPTP) and mitochondrial
   membrane potential (MMP, ΔΨm). Under normal conditions, mPTP remains
   closed, but stimuli such as oxidative stress or calcium overload induce
   its opening. The opening of the mPTP disrupts the mitochondrial
   membrane potential (MMP, ΔΨm) and consequently leads to mitochondrial
   dysregulation and impaired ATP production. We further evaluated the
   levels of mPTP opening and MMP using flow cytometry. Calcein was
   employed to assess the extent of mPTP opening, with a higher degree of
   mPTP opening corresponding to a lower calcein fluorescence intensity.
   JC-1 was employed to evaluate MMP. The results demonstrated that Tfh
   exhibited a higher degree of mPTP opening and a more pronounced
   reduction in MMP compared to Th0, and MSC-Exos effectively ameliorated
   these mitochondrial abnormalities (Fig. [139]6E and F). Moreover, in
   the presence of ionomycin to exacerbate mitochondrial calcium overload,
   Tfh suffered more severe mitochondrial damage—characterized by an even
   greater mPTP opening and further MMP declines—yet MSC-Exos maintained
   their protective effect against mitochondrial injury (Fig. [140]6E and
   F).

   MCU, a selective calcium channel on the inner mitochondrial membrane,
   is a key contributor to mitochondrial calcium overload when its
   expression is dysregulated, and it has recently been linked to CD4^+ T
   cell activation [[141]30]. In line with our findings on mitochondrial
   damage in Tfh, we next examined MCU protein expression to elucidate how
   MSC-Exos modulate mitochondrial calcium overload. WB analysis revealed
   that MCU expression was significantly up-regulated in Tfh compared with
   Th0, and treatment with MSC-Exos markedly normalized this aberrant
   expression. Furthermore, when ionomycin was applied to exacerbate
   calcium overload, MCU overexpression in Tfh was further amplified;
   however, MSC-Exos continued to significantly reduce MCU levels.

   Collectively, these findings demonstrate that mitochondrial calcium
   overload in Tfh induces mitochondrial damage, as evidenced by ROS
   burst, abnormal mPTP opening, and reduced MMP, and MSC-Exos effectively
   mitigate this overload through modulation of MCU, thereby suppressing
   Tfh differentiation and function.

MSC-Exos restore the imbalance of Tfh/Tfr in SLE

   During the validation of the effects of MSC-Exos on Tfh
   differentiation, we observed that although MSC-Exos significantly
   suppressed IL-21 and BCL6 expression and markedly reduced CXCR5^+PD-1^+
   Tfh population, their inhibitory effect on CXCR5 mRNA expression was
   not apparent (Fig. [142]3E). Given that Tfr and Tfh share the surface
   marker CXCR5 and that the restoration of the Tfr/Tfh imbalance is
   closely associated with the amelioration of SLE [[143]31], we
   hypothesized that MSC-Exos may concurrently inhibit Tfh while promoting
   Tfr expansion and function.

   To test this, we firstly isolated CD4^+CXCR5^+ cells—which comprise
   both mature Tfh and Tfr populations—from the spleens of IMQ-SLE mice
   and cocultured them with MSC-Exos at the same 3 concentrations used in
   previous experiments. Flow cytometry analysis revealed that
   IMQ-SLE-derived CXCR5^+ cells exhibited significantly reduced levels of
   Foxp3, a key marker of Tfr, compared with controls (Fig. [144]7A and
   C). Both moderate and high concentrations of MSC-Exos significantly
   increased the proportions of Foxp3^+ and IL-10^+ cells, indicating
   enhanced Tfr functionality (Fig. [145]7B and C). Notably, the effect on
   IL-10 expression was most pronounced at the moderate MSC-Exos
   concentration (Fig. [146]7B and C), mirroring our earlier observations
   regarding IL-21 suppression (Fig. [147]2B and C). Specifically,
   MSC-Exos at moderate concentrations significantly inhibited Tfh
   functionality while concurrently most effectively promoting Tfr
   function. Additionally, we also evaluated the influence of MSC-Exos on
   the differentiation of Tfr in vitro. RT-qPCR and flow cytometry results
   confirmed that MSC-Exos significantly increased Foxp3 and IL-10 levels
   in CXCR5^+ cells under Tfh polarization condition (Fig. [148]7D to H).

Fig. 7.

   Fig. 7.
   [149]Open in a new tab

   MSC-Exos suppressed Tfh while promoting the differentiation and
   function of Tfr. (A to C) Isolated CXCR5^+ cells from IMQ-SLE (n = 4)
   were cocultured with gradient concentrations of MSC-Exos in vitro for
   24 h, and Foxp3 and IL-10 expression levels were assessed by flow
   cytometry. (D) The mRNA expression levels of Foxp3 and IL-10 in the Tfh
   polarization conditions were measured using RT-qPCR. (E to H) The
   expression levels of Foxp3 and IL-10 were detected in CXCR5^+ cells
   within Tfh polarization conditions by flow cytometry. (I to L)
   Proportions of CD4^+CXCR5^+Foxp3^+ cells and CD4^+IL-10^+ cells in WT
   (n = 5), IMQ-SLE + PBS (n = 5), and IMQ-SLE + Exo (n = 5) mice. Bars
   indicate the means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P <
   0.0001 determined by one-way ANOVA with Tukey’s multiple comparisons
   test.

   Extending our findings in vivo, we further investigated the effects of
   MSC-Exos on Tfr in IMQ-SLE mice. The results revealed a significant
   decrease in the frequency of CD4^+CXCR5^+Foxp3^+ Tfr in IMQ-SLE
   compared to WT mice, which was markedly restored following intravenous
   injection of 100 μg of MSC-Exos (Fig. [150]7I and J). Moreover,
   MSC-Exos treatment significantly elevated IL-10 expression levels in
   IMQ-SLE mice (Fig. [151]7K and L).

   Taken together, these results demonstrate that MSC-Exos exert a
   bidirectional regulatory effect on the Tfh/Tfr balance—suppressing Tfh
   differentiation and function while simultaneously enhancing Tfr
   expansion and functionality. This dual action may underlie the
   therapeutic potential of MSC-Exos in ameliorating SLE.

Discussion

   SLE is a chronic autoimmune disease driven by autoantibody production,
   leading to multi-organ damage and increased mortality [[152]32]. Tfh
   are crucial in assisting B cells to produce high-affinity antibodies
   and are highly activated and proliferative in SLE [[153]2]. Given the
   importance of Tfh in lupus pathogenesis and the therapeutic potential
   of MSC-Exos in lupus [[154]13,[155]33], we investigated the effect and
   mechanism of MSC-Exos on Tfh in SLE. Our findings demonstrate that
   MSC-Exos effectively inhibit Tfh function and differentiation by
   restoring calcium homeostasis through Stim1–Orai1 pathway and calcium
   overload-induced mitochondrial damage. Moreover, we demonstrated for
   the first time the bidirectional regulatory effect of MSC-Exos on
   Tfh/Tfr balance in SLE (Fig. [156]8). Interestingly, we observed a
   nonlinear dose–response, with medium concentrations of MSC-Exos showing
   greater efficacy than higher doses. This may reflect cytokine overload,
   uptake saturation, or adaptive resistance within the inflamed lupus
   microenvironment.

Fig. 8.

   Fig. 8.
   [157]Open in a new tab

   MSC-Exos regulate Tfh differentiation and function via a multi-target
   network, restoring the Tfh/Tfr balance and alleviating SLE.
   Mechanistically, MSC-Exos inhibit Tfh activation by correcting
   intracellular calcium dysregulation and preventing mitochondrial
   calcium overload. In the cytoplasm, MSC-Exos restrict calcium influx
   through inhibition of the Stim1/Orai1 expression, thereby reducing NFAT
   and NF-κB activation. In the mitochondria, MSC-Exos suppress aberrant
   MCU expression, mitigating calcium overload-induced mitochondrial
   damage, and restoring mitochondrial homeostasis in Tfh. Figure was
   created with BioGDP [[158]50].

   The hyperactivation of the TLR7 signaling pathway is a key factor in
   the pathogenesis of lupus [[159]34]. In this study, we established a
   lupus model in mice using IMQ, a TLR7 agonist, which is widely used in
   lupus research [[160]17]. Female mice were used to reflect the
   female-biased prevalence of SLE and to remain consistent with commonly
   used lupus models [[161]33]. We confirmed that IMQ-SLE mice and SLE
   patients had higher frequency of Tfh, along with a lower number of Tfr,
   indicating an imbalance in the Tfh/Tfr, consistent with other lupus
   models and patients [[162]14,[163]35]. Several studies have
   demonstrated the ability of MSCs to inhibit the activation of T cells
   in lupus [[164]3]. MSC-Exos, which carry MSC immunoregulatory
   capabilities, have also shown great therapeutic value in autoimmune
   diseases [[165]13]. However, whether MSC-Exos can effectively target
   Tfh and their potential underlying mechanisms remain unclear. In this
   study, we revealed that MSC-Exos suppress the excessive frequency of
   Tfh in SLE both in vivo and in vitro. Meanwhile, MSC-Exos promote the
   population and function of Tfr, thereby regulating the Tfh/Tfr
   imbalance in SLE. Mechanically, MSC-Exos effectively regulate the
   calcium homeostasis in Tfh in both cytoplasm and mitochondria. For
   cytoplasm, MSC-Exos can target the Stim1–Orai1–NFAT/NF-κB pathway to
   inhibit Tfh differentiation and function. For mitochondria, MSC-Exos
   alleviate mitochondrial damage in Tfh by inhibiting MCU-induced calcium
   overload.

   Dysregulation of calcium homeostasis leads to excessive T cell
   activation and proliferation, thereby exacerbating autoimmune
   responses. Stim1–Orai1 signaling is the key regulator of cellular
   calcium homeostasis. Stim1 and Orai1 control the initiation of calcium
   influx upstream, where Stim1 functions as the ER calcium sensor and
   Orai1 serves as the plasma membrane calcium channel [[166]24]. Studies
   have demonstrated significantly elevated expression of Stim2 and Orai1
   in lupus nephritis, and intervention with Orai1-specific small
   interfering RNA (siRNA) has been shown to markedly alleviate renal
   damage [[167]36]. Vaeth et al. [[168]37] revealed that knockdown of
   either Stim1 or Stim2 significantly impaired Tfh function by
   suppressing NFAT-mediated BCL6 transcription. In this study, MSC-Exos
   down-regulated Stim1/Orai1 expression, thereby modulating calcium
   influx upstream and ameliorating calcium homeostasis imbalance in Tfh.

   In the downstream of calcium influx, MSC-Exos effectively blocked NFAT
   and NF-κB pathway activity in Tfh, which play crucial roles in
   regulating T cell activation and differentiation [[169]38]. The NFAT
   pathway is commonly activated in SLE [[170]23], with an observed
   up-regulation of CaN-dependent NFATc2 in CD4^+ T cells, leading to
   enhanced inflammatory responses. However, the specific CD4^+ T cell
   subset involved remains unclear [[171]39]. It is speculated that
   elevated expression of NFAT in Th17 cells in lupus may contribute to
   the up-regulation of CD40L expression, exacerbating lupus symptoms
   [[172]23]. Inhibiting the NFAT pathway has significant therapeutic
   value in SLE, with drugs like cyclosporine A and tacrolimus offering
   protection against lupus nephritis [[173]40]. Evidences have shown that
   NFATc2 facilitates the maintenance of Tfh [[174]37], and NFATc2
   directly influences IL-21 and CD40L expression [[175]41]. Additionally,
   tacrolimus notably suppresses Tfh in transplant recipients, mitigating
   rejection reactions [[176]42]. These findings suggest that inhibiting
   the NFAT pathway in Tfh may significantly suppress immune responses in
   SLE, which is consistent with our observations. Similarly, the NF-κB
   pathway, a central regulator of inflammatory responses, is critical for
   Tfh function. NF-κB deficiency results in impaired CXCR5 expression
   [[177]43], and NF-κB directly regulates IL-21 transcription,
   underscoring its importance in Tfh differentiation and function
   [[178]26]. Here, we showed that MSC-Exos markedly inhibit the NF-κB
   pathway in Tfh, consistent with a recent evidence showing the
   anti-inflammatory effects of bone marrow-derived MSCs derived exosomes
   (BMSC-Exos) through NF-κB down-regulation to reduce intervertebral disc
   degeneration [[179]44]. Furthermore, NFATc2 shares 142 common immune
   response gene targets with P65, allowing dynamic integration of the
   NFAT and NF-κB pathways in regulating T cell immune responses
   [[180]27], which aligns with our research findings that MSC-Exos
   modulate both NFAT and NF-κB pathways in Tfh.

   MCU-mediated mitochondrial calcium homeostasis is critical for various
   cellular processes, including energy metabolism and cell survival
   [[181]45]. Dysregulation of MCU expression leads to mitochondrial
   calcium overload by facilitating Ca^2+ influx from the cytosol to the
   mitochondrial matrix [[182]30]. This overload impairs mitochondrial
   function primarily by inhibiting ATP hydrolase activity [[183]28] and
   subsequently induces mitochondrial swelling and membrane rupture,
   triggering ROS burst and mPTP opening, ultimately resulting in
   downstream inflammatory cascades [[184]30]. T cell-specific MCU
   knockout in an experimental autoimmune encephalomyelitis (EAE) model
   alleviated neuroinflammation by suppressing mitochondrial calcium
   overload, suggesting that MCU is a promising therapeutic target for
   autoimmune diseases [[185]30]. Our findings indicate that calcium
   dysregulation not only initiates NFAT/NF-κB activation but also drives
   MCU-dependent mitochondrial calcium overload. Targeting the
   Stim/Orai1–MCU axis may therefore represent a promising strategy for
   SLE. In this study, MSC-Exos effectively counteract mitochondrial
   calcium overload via MCU in Tfh, even under ionomycin-induced
   exaggerated calcium overload conditions. Moreover, MSC-Exos
   significantly alleviate mitochondrial oxidative damage and restore
   mitochondrial homeostasis in Tfh, thereby exerting anti-inflammatory
   effects. These findings reveal that the regulation of Tfh by MSC-Exos
   in this study is not confined to a single target but involves
   simultaneous network regulation of Stim1–Orai1–NFAT/NF-κB axis and
   MCU-dependent mitochondrial calcium overload, resulting in potent
   inhibition of Tfh in SLE.

   Moreover, we found that MSC-Exos not only suppress Tfh but also promote
   the differentiation and function of Tfr. NFATc1 is expressed in resting
   T cells, while NFATc2 expression can be up-regulated by NFATc1 upon T
   cell stimulation, creating a positive feedback loop that amplifies NFAT
   pathway [[186]22]. Foxp3, a key transcription factor in Tregs, can bind
   to the P1 region of the NFATc2 promoter, inhibiting its interaction
   with NFATc1 and thereby suppressing NFATc2 production [[187]46]. Our
   data suggest that MSC-Exos primarily target NFATc2 rather than NFATc1,
   likely through selective up-regulation of Foxp3, which directly
   inhibits NFATc2 transcription independently of NFATc1. However, this
   finding appears to contrast with studies showing that NFATc2 deficiency
   impairs Tfr cell numbers and function, exacerbating autoimmune
   responses [[188]37]. This discrepancy contradicts our findings,
   suggesting that MSC-Exos may intervene in the Tfh/Tfr balance through
   multi-target mechanisms. For instance, the promotion of Tfr cells could
   be mediated by direct enhancement of Foxp3 expression or through
   alternative pathways, which warrants further investigation.
   Additionally, Foxp3 up-regulation may also contribute to the
   suppression of Tfh, as Foxp3 directly inhibits the activity of NF-κB
   and NFAT transcription factors in T helper cells [[189]47]. This
   inhibitory effect is well documented in Tregs and certain tumor cells,
   where the Foxp3–NF-κB axis plays a critical role in Treg
   differentiation and function [[190]48,[191]49]. Notably, Foxp3
   deficiency is associated with heightened NF-κB activity [[192]49].
   These findings suggest that, beyond modulating calcium homeostasis to
   inhibit NFAT and NF-κB pathways, MSC-Exos may further enhance their
   suppressive effects on Tfh by inducing Foxp3. Therefore, we hypothesis
   that MSC-Exos achieve dual regulation of Tfh and Tfr by promoting Foxp3
   expression, which simultaneously suppresses Tfh hyperactivation and
   enhances Tfr-mediated immune regulation.

   The limitations of this study should be noted. Firstly, we did not use
   lentivirus or gene knockout mice for reverse validation, mainly due to
   the scarcity and instability of Tfh and Tfr in vitro. Secondly, the
   application of MSC-Exos in mice lasted only 14 d, necessitating further
   investigation into the long-term effects of MSC-Exos. Additionally, the
   lack of a mature in vitro polarization condition system for Tfr hampers
   a comprehensive understanding of the specific mechanisms through which
   MSC-Exos regulate Tfr. Lastly, while our data establish MSC-Exos as an
   effective corrector of Tfh/Tfr imbalance in SLE, this mechanism may
   extend to other inflammatory conditions where Tfh/Tfr dysregulation
   serves as an indicator of autoimmune pathology severity.

Conclusion

   Our study elucidates that MSC-Exos significantly inhibit Tfh
   differentiation and function in SLE. Mechanistically, MSC-Exos inhibit
   Tfh activation by correcting intracellular calcium dysregulation and
   preventing mitochondrial calcium overload. In the cytoplasm, MSC-Exos
   restrict calcium influx through inhibition of the Stim1/Orai1
   signaling, thereby reducing NFAT and NF-κB activation. In the
   mitochondria, MSC-Exos prevent mitochondrial calcium overload by
   down-regulating MCU, reducing mitochondrial damage in Tfh.
   Additionally, we demonstrate the bidirectional regulatory effect of
   MSC-Exos on Tfh/Tfr balance in SLE. These results not only highlight
   the therapeutic potential of MSC-Exos in SLE but also underscore the
   critical role of calcium homeostasis in SLE management.

Ethical Approval

   Ethical approval was obtained from the Ethics Committee of Zhongshan
   Hospital, Fudan University, Shanghai, China (approval no. B2023-0862R).

Acknowledgments