Fetal bovine serum (FBS), penicillin-streptomycin (P/S), trypsin-ethylenediaminetetraacetic acid (EDTA) were purchased from Gibco (Grand Island, NY, USA). CD4 and CD8 MicroBeads and the AutoMACS system were purchased from Miltenyi Biotec. (Shanghai, China). Biolegend (San Diego, CA, USA) -provided antibodies included: APC-conjugated CD29 (303008), CD19 (115512), and F4/80 (123115); PE-conjugated CD11c (117307), Sca1 (108107), CD206 (141705), and inducible nitric oxide synthase (iNOS) (696805); with PE-Cy7-CD44 (103029) and FITC-CD11b (101205). Hieff qPCR SYBR Green Master Mix was acquired from Yeasen (Shanghai, China). FOXP3/Transcription Factor Staining Buffer set was purchased from eBioscience (San Diego, CA, USA). All primers were obtained from Sangon Biotech (Shanghai, China). TD139, Galectin-3 were purchased from Selleck&Bimake (Selleck&Bimake, Shanghai, China).

Male BALB/c (H-2d) and C57BL/6 (H-2b) mice (10–12 weeks old) were obtained from Shanghai Laboratory Animal Center (SLAC, Shanghai, China) and maintained in specific pathogen-free (SPF) conditions with autoclaved water and standard chow. All animal tests were conducted in compliance with the guidelines established by the Ethics Committee of Xinhua Hospital Affiliated with the Shanghai Jiao Tong University School of Medicine (Ethics Approval No. XHEC-F-2022-076). All mice in this study were euthanized by cervical dislocation, in strict accordance with the guidelines outlined by the Ethics Committee of Xinhua Hospital Affiliated with the Shanghai Jiao Tong University School of Medicine (Shanghai, China).

MSCs isolated from bone marrow of C57BL/6 mice were cultured in Dulbecco modified Eagle medium-low glucose (DMEM-L; Hyclone, Shanghai, China) containing 10% heat-inactivated fetal bovine serum (FBS; Gibco, USA) and 1% penicillin-streptomycin (10,000 U/mL penicillin, 10,000 µg/mL streptomycin) (1% P/S; Gibco, USA) at 37 °C in a humidified atmosphere of 5% CO₂. MSCs at passage 2 (P2- MSCs) were passaged by 0.25% trypsin-ethylenediaminetetraacetic acid (EDTA; Gibco, Burlington, Canada) at a 1:3 ratio when they were 80%–90% confluent, up to P12. Here, the MSCs were classified as early passages of MSCs (passage 4 to 6-MSCs) and late passages (passage 7 to 12-MSCs). MSCs exhibited typical fibroblast-like morphology, characterized by long, polygonal, plastic-adherent cells, and display the typical immunophenotypes, including Sca1 (+), CD29 (+), CD44 (+), CD19 (–), and CD11b (–) (Supplementary Fig. 1A,B).

The Raw264.7 cells, a macrophage cell line, were purchased from the American Type Culture Collection (Manassas, VA, USA). Cells were cultured in Dulbecco modified Eagle medium glucose (DMEM; Hyclone, China) containing 10% FBS and 1% P/S at 37 °C in a humidified atmosphere of 5% CO₂.

All cell lines (including Raw264.7) were validated by STR profiling and tested negative for mycoplasma contamination.

Firstly, total RNA was extracted from cells using EZ-press RNA Purification Kit (EZBioscience, CA, USA) according to manufacturer’s protocol. Total RNA meeting purity criteria (A260/A280 ratio 1.8–2.0) was converted to complementary deoxyribonucleic acid (cDNA) with PrimerScript™ RT Master Mix (Takara, Shiga, Japan) per manufacturer’s instructions. qRT-PCR was performed using relevant gene-specific primers and Hieff qPCR SYBR Green Master Mix (Yeasen, China) on Step One Plus Real-time PCR System (Life Technologies, CA, USA). The messenger ribonucleic acid (mRNA) expression levels of target genes were normalized to $\beta$-actin gene and calculated according to the 2^{-Δ⁢Δ⁢CT} method. All primers were obtained from Sangon Biotech. Primer sequences are shown in Table 1.

Fluorescein conjugated monoclonal antibodies (mAbs) used for experiments include phycoerythrin (PE)-conjugated mAbs to mouse, Fluorescein isothiocyanate (FITC)-conjugated mAbs to mouse, allophycocyanin (APC)-conjugated mAbs to mouse and R-Phycoerythrin-Cyanine 7 (PE-Cy7)-conjugated mAbs to mouse. Antibody information are provided in the Reagents and Animals section. In brief, for cell surface antigen staining, cells were collected, washed with phosphate buffered solution (PBS) and incubated with relevant fluorescein-conjugated mAb in the dark for 30 minutes at 4 °C. For cell nuclear/intracellular antigen staining, after stained with cell surface antigen, cells were fixed and permeabilized with eBioscience™ FOXP3/Transcription Factor Staining Buffer set (Invitrogen, CA, USA) according to the manufacture’s protocol followed by intracellular staining with relevant fluorescein-conjugated mAb in the dark for 30 minutes at room temperature. Finally, cells were washed with PBS and performed flow cytometric analysis. Flow cytometry (FCM) was performed by using BD FACS CantoII (Becton-Dickinson Bioscience, Franklin Lakes, NJ, USA) and analysis was performed by using FlowJo software (Version 10.8.1; BD Biosciences, Ashland, OR, USA).

Mice underwent bone marrow (BM) transplantation (BMT) according to the protocol described previously [34]. Briefly, recipient mice (BALB/c, male, 10–12 weeks) were randomly divided into three groups as follows: the T cell depleted-bone marrow cells (TCD-BMC) + T cells + vehicle group, the TCD-BMC + T cells + P6-MSCs group and the TCD-BMC + T cells + P12-MSCs group. At day 0, after BALB/c recipients received 7.0 Gy (¹³⁷Cesium source at 100 cGy/min) total body irradiation (TBI), TCD-BMC (5 $\times$ 10⁶ cells/mouse) together with splenic CD3⁺ T cells (2 $\times$ 10⁶ cells/mouse) from allogeneic donor mice (C57BL/6) were collected and co-injected into BALB/c recipients via tail vain. The recipients were transfused with vehicle, P6-MSCs (3 $\times$ 10⁵ cells/mouse) or P12-MSCs (3 $\times$ 10⁵ cells/mouse) intravenously at day 1, 3, 5 according to groups they were assigned to. The severity of GVHD was evaluated every 2 days post-transplantation using a standardized clinical scoring system, based on five parameters: weight loss, posture, activity, fur texture, and skin integrity. Scoring criteria conformed to established protocols as previously described [35]. The mice were euthanized via cervical dislocation on day 21 post-TBI without prior anesthesia, following which peritoneal cavity-derived macrophages were collected for subsequent analysis. Target organs (liver, lung, small intestine) were collected from per experimental group, fixed in 10% formalin, paraffin-embedded, hematoxylin and eosin (H&E)-stained, and blindly evaluated for GVHD pathology.

Spleens harvested from C57BL/6 mice were grinded through 40-µm nylon cell strainers and obtained splenic cell suspension. CD4⁺/CD8⁺ T cell isolation or depletion was performed using CD4 and CD8a MicroBeads and the AutoMACS system (Miltenyi Biotec, Germany) according to the manufacturer’s protocol. Briefly, 10 µL of CD4 MicroBeads and 10 µL of CD8a MicroBeads were added per 1 $\times$ 10⁷ cells, and the mixture was incubated for 15 minutes at 4 °C in the dark. The cell suspension was then passed through a pre-rinsed LS column mounted on the AutoMACS Pro Separator to retain T cells (CD4⁺ and CD8⁺) via magnetic separation.

Bone marrow cells (BMC) were isolated from the femurs and tibias of C57BL/6 mice following humane euthanasia. The marrow cavity was flushed with ice-cold DMEM-complete medium using a 25-gauge needle to obtain a single-cell suspension, which was sequentially filtered through a 40-µm nylon cell strainer to remove debris. Red blood cells in the suspension were lysed with ACK lysis buffer, and the resulting nucleated cell pellet was resuspended in MACS Buffer at a concentration of 1 $\times$ 10⁸ cells/mL. T cell depletion (TCD) was performed using CD4 and CD8a MicroBeads according to the manufacturer’s protocol. The efficiency of T cell depletion was validated by flow cytometry using anti-mouse CD3-PE antibody. Only TCD-BMC preparations with residual CD3⁺ T cells $\le$1% (consistent with purity validation in Supplementary Fig. 2) were used for transplantation experiments.

Macrophages were harvested from the peritoneal lavage of recipient mice at day 21 after BMT. After intraperitoneal injection of pre-cooled PBS, gentle abdominal massage was performed prior to peritoneal lavage fluid collection. Finally, the harvested cells were stained for flow cytometric analysis using anti-mouse CD11b-FITC, anti-mouse F4/80-APC and anti-mouse CD206-PE and anti-mouse iNOS-PE antibodies. CD11b⁺F4/80⁺ double-positive cells were identified as peritoneal macrophages (Supplementary Fig. 3).

To investigate the effects of MSCs on macrophages at different passages, P6-MSCs and P12-MSCs were seeded into the lower chamber of transwell (Corning, USA, catalog number 3412, pore diameter 0.4 µm, polycarbonate membrane) at a density of 2 $\times$ 10⁵ per well. When they reached approximately 70% confluence, Raw 264.7 cells were seeded into the upper chamber of transwell at a density of 1 $\times$ 10⁷ per well and co-cultured with P6-MSCs or P12-MSCs for 48 hours. Besides, the co-culture system was treated with vehicle, Gal-3 (500 ng/mL), TD139 (100 µM), LPS (100 ng/mL) + vehicle, LPS (100 ng/mL) + galectin-3 (500 ng/mL), or LPS (100 ng/mL) + TD139 (100 µM). Finally, Raw264.7 cells were collected and stained with PE-conjugated anti-mouse CD206 or iNOS antibodies for flow cytometry.

Raw264.7 cells were seeded into 24-well plates at a density of 3 $\times$ 10⁶ per well and treated with vehicle, galectin-3 (100 ng/mL), galectin-3 (500 ng/mL) for 48 hours. Galectin-3 was dissolved in ddH₂O. Raw264.7 cells were collected and stained with anti-mouse CD206-PE or anti-mouse iNOS-PE antibodies for flow cytometry. Besides, expression of macrophage polarization-related mediator (iNOS, interleukin-1beta (IL-1$\beta$), interleukin-6 (IL-6), Tumor necrosis factor-$\alpha$ (TNF-$\alpha$), CC-chemokine ligand 2 (CCL2), CC-chemokine ligand 3 (CCL3), CC-chemokine ligand 4 (CCL4), cyclooxygenase 2 (COX2), interleukin-10 (IL-10), YM1, Transforming growth factor-beta1 (TGF-$\beta$1), Arginase (Arg) and CD206) in Raw264.7 cells were determined by qPCR across experimental groups.

MSCs at different passages were seeded in 96-well cell culture plates at 3 $\times$ 10³ per well. The cells were treated with vehicle or TD139 at 100 µM concentration for 72 h. The absorbance at 450 nm was measured every 24 h with a microplate reader (Thermo Fisher) according to the instructions of the CCK-8 kit (cell counting kit-8, Beyotime, China). The total numbers of migrated cells were counted to evaluate cell proliferation.

Raw 264.7 macrophages were seeded in 24-well cell culture plates at 1.5 $\times$ 10⁶ per well. TD139 was dissolved in DMSO. The cells were treated with vehicle or TD139 (100 µM) for 48 h. Then, the Raw 264.7 macrophages were harvested and stained for flow cytometric analysis using anti-mouse CD206-PE or anti-mouse iNOS-PE antibodies.

All experiments were independently repeated a minimum of two to three times. The figure legends provide detailed information on reproducibility and sample sizes. Data are presented as mean $\pm$ standard deviation. Statistical analyses were performed using one-way or two-way analysis of variance (ANOVA) or Student’s t-test with GraphPad Prism software. Non-parametric murine survival data were analyzed using the log-rank test. All p values were two-sided, with p 0.05 considered statistically significant.

To determine whether passage number of MSCs would affects the therapeutic efficacy of MSCs in GVHD, recipient mice were received MSCs at different passages after BMT (Fig. 1A). Although there were no significant differences in survival of mice among three groups (Fig. 1B), the clinical GVHD scores were significantly higher in recipients treated with P12-MSCs compared to those treated with P6-MSCs (Fig. 1C, p 0.0001). To confirm these results, histopathological examination was performed at day 21 post-BMT. As shown in Fig. 1D, the histopathological images showed more severe tissue damage in GVHD-target organs including small intestine, liver, lungs, in P12-MSCs-treated recipients compared to those treated with P6-MSCs. Taken together, these results suggest that MSCs at late passage (P12) are less effective in alleviating the severity of acute GVHD (aGVHD) compared to MSCs at early passage (P6).

Fig. 1.

Effects of MSCs at different passages on the GVHD mice. (A) Experimental scheme for GVHD model establishment and treatment. (B) Kaplan-Meier survival curve of GVHD mice treated with PBS, P6-MSCs, P12-MSCs. (C) Clinical GVHD score of GVHD mice treated with PBS, P6-MSCs, P12-MSCs. (D) Representative histopathology images of liver, lung and small intestine from GVHD mice treated with PBS, P6-MSC and P12-MSC on day 21 after transplantation. Scale bar, 100 µm. Data were pooled from 3 independent experiments. N = 3–5 in each group. ****p 0.0001. GVHD, graft-versus-host-disease; PBS, phosphate buffered solution.

Next, we sought to explore the potential mechanisms underlying the variation in MSCs at different passages in the context of GVHD. Given the powerful immunoregulatory effects of MSCs on macrophages, we next assessed the polarization of macrophages isolated from peritoneal lavage liquid of GVHD mice model treated with MSCs at different passages. Compared to the P6-MSCs-treated mice, the proportion of F4/80⁺ CD11b⁺ CD206⁺ M2-like macrophages was lower in the P12-MSCs-treated mice (Fig. 2A, p 0.0001). Meanwhile, we utilized Raw246.7 cell line to investigate the effects of MSCs at different passages on macrophages in vitro. After co-culturing with MSCs at different passages, the proportion of CD206⁺ M2-like macrophages was also lower in P12-MSCs-treated group than that in P6-MSCs-treated group (Fig. 2B, p 0.05). Moreover, the proportion of iNOS⁺ M1-like macrophages in Raw264.7 cells was higher in P12-MSCs-treated group than in P6-MSCs-treated group (data not shown). Collectively, our results indicated that MSCs at later passages exhibit a significantly reduced ability to induce M2 macrophage polarization.

Fig. 2.

The effect of MSCs on macrophage polarization both in vivo and in vitro. (A) Flow cytometric analysis of CD206⁺ M2-like macrophage proportion within F4/80⁺CD11b⁺ peritoneal macrophages from GVHD mice treated with PBS, P6-MSCs or P12-MSCs on day 21 post-transplantation. (B) Flow cytometric analysis of CD206⁺ M2-like macrophages percentages in Raw264.7 cells co-cultured with vehicle control (culture medium alone), P6-MSCs, or P12-MSCs without LPS stimulation for 48 hours. Data were pooled from 3 independent experiments. *p 0.05, ****p 0.0001. LPS, lipopolysaccharide.

To simulate the in vivo inflammatory environment, we supplemented the MSCs-Raw264.7 macrophage co-culture system with LPS. Although the proportion of CD206⁺ M2-like macrophages and iNOS⁺ M1-like macrophages did not differ significantly between P6-MSCs-treated and P12-MSCs-treated groups, mRNA expression of M2-associated anti-inflammatory mediators, such as IL-10 was reduced, while the mRNA expression of M1-associated pro-inflammatory mediators, including L-6, CCL4, and COX2, was elevated in the P12-MSCs-treated group compared to the P6-MSCs-treated group (Fig. 3A,B). Our results indicated that the therapeutic efficacy of MSCs at different passages for acute GVHD could be related to macrophage polarization and the effect of MSCs at late passage on M2-like macrophages polarization was notably weakened.

Fig. 3.

The effect of MSCs on macrophage polarization in the present of LPS. Raw264.7 cells were cocultured with vehicle control (culture medium alone), P6-MSCs or P12-MSCs in the present of 100 ng/mL LPS for 48 h, distinct from Fig. 2 (no LPS). (A) Flow cytometric analysis of CD206⁺ cell percentages in Raw264.7 macrophages. (B) qPCR analysis of macrophage polarization-related mediator expression (IL-10, IL-6, CCL4, and COX2) in Raw264.7 macrophages. Data were pooled from 3 independent experiments. Data are presented as the mean $\pm$ SD. ns (p $>$ 0.05), *p 0.05, **p 0.01, ****p 0.0001. SSC, Side Scatter.

Previous studies have demonstrated that the activation of Gal-3 enhances inflammation [23], whereas the downregulation of Gal-3 can suppress inflammatory responses [24]. Several studies have indicated that Gal-3 promotes a pro-inflammatory response by inhibiting M2 macrophage polarization and inducing apoptosis [21, 22]. MSCs constitutively express Gal-3 at both the mRNA and protein levels, thereby exerting immunomodulatory effect via galectins [36]. Moreover, Gal-3 is a new informative model to predict aGVHD development [37]. Based on this, we inferred that Gal-3 expression may be linked to the effects of MSCs on macrophages, to verify this point, we first analyzed the expression of Gal-3 in MSCs at different passages. As shown in Fig. 4, Gal-3 expression escalated starting at passage10 and markedly higher in MSCs at later passages compared to those at earlier passages. Then, we assessed the characteristics of Raw 264.7 macrophages in the presence of Gal-3. As shown in Fig. 5A, the proportion of iNOS⁺ M1-like macrophages notably increased in Gal-3-treated Raw 264.7 cells in a dose-dependent manner after 48 h. The M1-like macrophage phenotype is characterized by the expression of chemokines such as iNOS, CCL2, CCL3, CCL4 and the induction of pro-inflammatory mediators including IL-1$\beta$, IL-6, TNF-$\alpha$, COX2. Our data showed that, in addition to IL-1$\beta$, the mRNA expression of these M1-associated mediators was significantly upregulated in Raw 264.7 macrophages treated with high doses of Gal-3 (Fig. 5B, p 0.01). Furthermore, the proportion of CD206⁺ M2-like macrophages in Gal-3 treated group was significant decreased in a dose-dependent manner compared to the control groups (Fig. 5C, p 0.01). Meanwhile, anti-inflammatory mediators associated with the M2 macrophage phenotype, including IL-10, TGF-$\beta$1, Arg, and Ym1, were decreased in Gal-3-treated Raw264.7 macrophages compared to the control groups (Fig. 5D, p 0.0001). All these data suggested that the differential effects of MSCs at different passages on macrophage polarization may be related to Gal-3. As MSCs progress through passages, they express higher levels of Gal-3, which could concentration-dependently induce macrophage polarization toward the iNOS⁺ M1 phenotype while suppressing CD206⁺ M2 macrophage polarization.

Fig. 4.

The mRNA expression of Galectin-3 in MSCs at different passages by qPCR. Data were pooled from 3 independent experiments. Data are presented as the mean $\pm$ SD. ****p 0.0001.

Fig. 5.

The effect of Gal-3 on macrophage polarization. Raw264.7 cells were treated with vehicle, Gal-3 (100 ng/mL) or Gal-3 (500 ng/mL) for 48 h. (A) Flow cytometry analysis of iNOS⁺ M1 macrophage proportion in Raw264.7 macrophages. (B) qPCR analysis of iNOS⁺ M1 macrophage polarization-related mediator expression (iNOS, IL-1$\beta$, IL-6, TNF-$\alpha$, CCL2, CCL3, CCL4 and COX2) in Raw264.7 macrophages. (C) Flow cytometry analysis of CD206⁺ M2 macrophage proportion in Raw264.7 macrophages. (D) qPCR analysis of CD206⁺ M2 macrophage polarization-related mediator expression (IL-10, YM1, TGF-$\beta$1, Arg and CD206) in Raw264.7 macrophages. Data were pooled from 3 independent experiments. Data are presented as the mean $\pm$ SD. ns (p $>$ 0.05), **p 0.01, ****p 0.0001. Gal-3, Galectin-3.

Based on the above data, it is evident that elevated expression of Gal-3 in MSCs at later passages results in a diminished effect of MSCs on M2-like macrophage polarization. In this study, we sought to assess the potential applications of Gal-3 inhibitors. While prior studies have focused on Gal-3, our data demonstrated that Gal-1 is also expressed in MSC (Supplementary Fig. 4A). TD139, a selective small-molecule inhibitor of Gal-3, with no impact on Gal-1 (Supplementary Fig. 4B), and is currently undergoing Phase II clinical trials for idiopathic pulmonary fibrosis [28, 38]. Treatment with TD139 at concentrations ranging from 10–100 µM for 48 hours led to a significant dose-dependent reduction of Gal-3 expression in P12-MSCs (Fig. 6A). Based on this observation, a concentration of 100 µM TD139 was selected for subsequent experiments. Similarly, Gal-3 expression was also inhibited in P6-MSCs following high-dose TD139 treatment (Fig. 6B, p 0.0001). In addition, our data showed that TD139 treatment did not alter MSCs proliferation and viability (Supplementary Fig. 5).

Fig. 6.

The effect of TD139 on Gal-3 expression of MSCs and macrophage polarization induced by MSCs. MSCs at different passages were treated with DMSO, TD139 (10 µM), TD139 (50 µM) or TD139 (100 µM) for 48 h (A,B). MSCs at different passages cocultured with Raw264.7 macrophages were treated with vehicle, Gal-3 (500 ng/mL) or TD139 (100 µM) for 48 h (C,D). (A) qPCR analysis of Gal-3 expression in P12-MSCs. (B) qPCR analysis of Gal-3 expression in P6-MSCs. (C) Flow cytometry analysis of CD206⁺ M2 macrophage proportion in Raw264.7 macrophages cocultured with P6-MSCs. (D) Flow cytometry analysis of CD206⁺ M2 macrophage proportion in Raw264.7 macrophages cocultured with P12-MSCs. Data were pooled from 3 independent experiments. Data are presented as the mean $\pm$ SD. **p 0.01, ****p 0.0001.

To determine whether TD139 affect other profiles of MSCs, we also examined the immunophenotype of MSCs in presence of TD139. The results revealed that MSCs cultured with or without TD139 exhibited typical fibroblast-like morphology, characterized by long, polygonal, plastic-adherent cells, and displayed the expected immunophenotypes of MSCs, including Sca1 (+), CD29 (+), CD44 (–), CD19 (–), and CD11b (–) (data not shown). These findings indicated that TD139 effectively inhibits Gal-3 expression in MSCs without altering their fundamental biological properties, including morphology and immunophenotype.

Since TD139 inhibits Gal-3 expression in MSCs, we inferred that TD139 could improve the immunoregulatory function of MSCs. In this study, we investigated the impact of TD139 on macrophage polarization induced by MSCs. As shown in Fig. 6, after incubation with TD139 at a concentration of 100 µM, the proportion of CD206⁺ M2-like macrophages relative to Raw264.7 macrophages was significantly increased in the TD139-treated P6-MSCs group compared to the vehicle-treated P6-MSCs group (Fig. 6C, p 0.05). Besides, the proportion of CD206⁺ M2-like macrophages relative to Raw264.7 macrophages was also increased in the TD139-treated P12-MSCs group compared to the vehicle-treated P12-MSCs group (Fig. 6D, p 0.001). Moreover, although no significant differences were observed among groups containing P6-MSCs, the percentage of iNOS⁺ M1-like macrophages was decreased in the TD139-treated P12-MSCs group compared to the vehicle-treated P12-MSCs group (data not shown). Taken together, these results demonstrated that TD139 enhances the immunoregulatory effect of MSCs at early passages on macrophages by promoting CD206⁺ M2-like macrophages polarization. Furthermore, TD139 also saves the immunoregulatory function of MSCs at later passages via inducing CD206⁺ M2-like macrophages polarization and suppressing iNOS⁺ M1-like macrophages polarization. Additionally, both the proportion of CD206⁺ M2-like macrophages and iNOS⁺ M1-like macrophages relative to Raw264.7 macrophages in the TD139 treated P12-MSCs group was closely resembled the results observed in the TD139-treated P6-MSCs group. This indicated that TD139 treatment can restore the immunoregulatory capacity of MSCs at later passages to levels comparable to those at early passages, further confirming that TD139 can preserve the immunoregulatory function of MSCs at late passages.