Surgically resected ovarian cancer tissues and non-malignant control ovarian tissues were consecutively collected at the Affiliated Cancer Hospital of Xinjiang Medical University between September 2024 and August 2025. A total of 71 female subjects were enrolled in this study, and the age of patients with ovarian cancer ranged from 45–65 years old. The ovarian cancer cohort (n=51) was stratified by FIGO stage into stage I/II (n=21), stage III (n=21) and stage IV (n=9). Non-malignant controls (n=20) were histologically non-malignant ovarian tissues from patients with benign gynecological diseases. The final sample size was based on the availability of eligible surgical specimens. The small stage IV subgroup reflects clinical practice, in which primary surgical resection is uncommon for stage IV ovarian cancer. A post hoc power analysis for the primary endpoint (the difference in VISTA protein expression between advanced ovarian cancer and non-malignant controls) indicated that the sample size provided >80% statistical power at α=0.05. Baseline clinical characteristics are provided in Table SI. The present study was approved by Affiliated Tumor Hospital of Xinjiang Medical University ethics committee (approval No. K-2025116), and all participants provided written informed consent. Eligible participants had pathologically confirmed ovarian cancer without prior antitumor therapy; controls had no history of malignancy or autoimmune disease. Participants with other malignancies, severe infection, or incomplete clinical data were excluded.

After rinsing with physiological saline, the surgically resected tissues were divided into three aliquots for sectioning and immunofluorescence staining, reverse transcription-quantitative (RT-q) PCR and western blot analysis. All samples were processed within 30 min of collection and either frozen at −80°C or fixed in 4% paraformaldehyde at room temperature for 24 h. Patient information, including age, disease stage, and histologic type, was recorded. All samples were coded and managed under blinded conditions while maintaining sample traceability.

Peripheral blood mononuclear cells (PBMCs) were isolated from peripheral blood samples obtained from three independent healthy adult donors (two men and one woman; aged 25–35 years) with no history of malignancy, autoimmune disease, or acute or chronic infection. Written informed consent was obtained from all donors, and the protocol was approved by the Ethics Committee of the Affiliated Cancer Hospital of Xinjiang Medical University. All datasets reported as n=3 comprised three biological replicates, each derived from a different donor.

Human peripheral blood lymphocyte separation medium (Beijing Solarbio Science & Technology Co., Ltd.; cat. no. P8610) was used for PBMC isolation. Fresh anticoagulated whole blood was diluted 1:1 with RPMI 1640 medium (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 11875101). Diluted blood (6 ml) was gently layered over 3 ml of separation medium in a 15-ml centrifuge tube (ExCell Bio; cat. no. CS015-0001). After centrifugation at 400 × g (Shanghai Feige; DK-80) for 20 min at room temperature, cells from the intermediate buffy-coat layer were collected, washed twice with 5 ml PBS (300 × g, 5 min, room temperature), and resuspended in RPMI 1640 medium for subsequent use.

Naive CD4+ T cells were isolated using the Naive CD4+ T Cell Isolation Kit II (Miltenyi Biotec GmbH; cat. no. 130-094-131). The cells were resuspended in 10 µl of PBMCs and 40 µl of isolation buffer. The biotin antibody mixture (10 µl) and magnetic beads (20 µl) were added sequentially, and the suspension was incubated for 5 min at room temperature. LS Columns (Miltenyi Biotec GmbH; cat. no. 130-042-401) were placed in an OctoMACS^™ Separator (Miltenyi Biotec GmbH; cat. no. 130-042-108), rinsed with 3 ml of separation buffer, and the cell suspension was then added. Unlabeled cells were collected and rinsed with 3 ml of separation buffer. The retained cells were treated with 1 ml of Bead Release Reagent and subsequently washed with PBS for later use. Naive CD4+ T cells were differentiated into FOXP3+CXCR5+ T cells by culture in RPMI 1640 medium supplemented with fetal bovine serum (Excell Bio; cat. no. FND500), penicillin-streptomycin (Gibco; Thermo Fisher Scientific, Inc.; cat. no. 15070-063), CD3 monoclonal antibody [MultiSciences (Lianke Biotech); cat. no. F1100300], CD28 antibody (Santa Cruz Biotechnology, Inc.; cat. no. SC-53876; 0.5 µg/ml), IL-2 (Proteintech Group, Inc.; HZ-1015, 5 ng/ml), TGF-β1 (Proteintech Group, Inc.; HZ-1011, 5 ng/ml), and recombinant IL-23 (Proteintech Group, Inc.; cat. no. HZ-1254, 25 ng/ml), for 5 days. For flow cytometric (BD FACSCanto II; BD Biosciences) sorting of Tfr-like cells, cells were suspended in serum-free medium at 1×107 cells/ml and first stained with Zombie NIR fixable viability dye (BioLegend, Inc.; cat. no. 423106) for dead cell exclusion, followed by surface staining with CD25 (BD Biosciences; cat. no. 557753), CD127 (BD Biosciences; cat. no. 557938), CXCR5 (Thermo Fisher Scientific, Inc.; cat. no. 17-9185-42), and VISTA (Thermo Fisher Scientific, Inc.; cat. no. 25-1088-42) antibodies. The cells were incubated at 4°C in the dark for 30 min. Compensation controls were prepared using UltraComp eBeads (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 01-2222-42) stained with each individual fluorochrome, and compensation matrices were calculated using BD FACSuite software v1.6 (BD Biosciences) prior to each experiment. Fluorescence-minus-one controls were set up for each fluorochrome to define accurate gating boundaries, particularly for the VISTA channel with relatively dim expression. The sequential gating strategy was applied as follows: Lymphocytes were first identified by forward scatter area (FSC-A) vs. side scatter area (SSC-A); single cells were then discriminated using FSC-height (FSC-H) vs. FSC-A; dead cells were excluded based on viability dye negativity; CD25+CD127-CXCR5+ Tfr-like cells were further gated and subdivided into VISTA+ and VISTA- subsets for sorting. CD25+CD127-CXCR5+VISTA+ and CD25+CD127-CXCR5+VISTA- cells were sorted using a flow cytometer (BD FACSCanto II; BD Biosciences). Post-sort purity was verified by re-analyzing an aliquot of the sorted cell suspension and consistently exceeded 95% across all experiments. The gating strategy and post-sort purity results are shown in Fig. S1.

Lentivirus particles were packaged via a third-generation lentiviral vector system and 293T packaging cells (ATCC; cat. no. CRL-3216) were used for viral production. The lentiviral vectors used for VISTA overexpression and silencing were purchased from Shanghai GeneChem Co., Ltd. Then three groups were set up: Group M (conventional medium), Group A (supplemented with HiTransG A), and Group P (supplemented with HiTransG P), to compare the effects of different infection conditions. For lentiviral transduction, a cell suspension at 5×104 cells/ml was seeded into a 96-well plate at 100 µl per well (including three control wells), and cultured at 37°C for 24 h until the reported confluence reached 30%. The negative control virus was diluted to concentrations of 2×107, 1×107 and 1×106 TU/ml, and added to the corresponding systems at multiplicities of infection (MOIs) of 1, 10, and 20, respectively (Group M: 90 µl medium + 10 µl virus; Groups A/P: 86 µl medium + 10 µl virus + 4 µl infection solution). After 12 h of culture, the medium was replaced with normal medium. At 72 h, cells were examined microscopically and the condition and MOI yielding ~80% infection efficiency with acceptable cell morphology were selected for subsequent experiments. For establishment of stably transduced Tfr-like cell lines, puromycin at a final concentration of 2 µg/ml was used for continuous screening over 7 days; the stable cell pool was maintained in complete culture medium supplemented with 0.5 µg/ml puromycin during routine passaging. The targeting sequences of short hairpin (sh)RNA used in the present study were: sh-VISTA-1: CCCTGACTCTCCAAACTTTGA; sh-VISTA-2: CCGTATTCCCTGTATGTCTGT; sh-NC: TTCTCCGAACGTGTCACGT. Only the gene-targeting fragment sequences are presented.

The induced Tfr-like cells were sorted by flow cytometry to obtain VISTA+ Tfr-like cells and VISTA- Tfr-like cells. For VISTA- Tfr-like cells, three groups were set up: blank group, OE negative control group (transduced with negative control virus), and VISTA overexpression group (transduced with LV–VISTA overexpression lentivirus). For VISTA+ Tfr-like cells, three groups were set up: blank group, shRNA negative control group (transduced with negative control virus), and VISTA-silencing group (transduced with VISTA-RNAi). Prior to lentiviral transduction, cells in each group were stained with the CellTrace^™ 5,6-carboxyfluorescein diacetate succinimidyl ester (CFSE) cell proliferation kit (Thermo Fisher Scientific, Inc.; cat. no. C34570) at 0.5 µM with a cell density of 1×106 cells/ml. and the staining reaction was incubated at 37°C for 20 min in the dark. After lentiviral transduction, cells in each group were cultured individually. At the end of culture, cells in each group were collected, and cell viability was evaluated by trypan blue exclusion to determine whether cell death confounded the proliferation findings. Total cell apoptosis rate was calculated as the sum percentage of early apoptotic cells and late apoptotic cells. No statistically significant difference in cell viability was observed among the groups. Subsequently, cells were washed once with PBS, resuspended in 500 µl PBS, and CFSE fluorescence intensity was measured by flow cytometry using a BD FACSCanto II flow cytometer (BD Biosciences). Three independent experiments were performed, and all replicates were included in the statistical analysis. Flow cytometric raw data were acquired and analyzed using BD FACSDiva software v8.0 (BD Biosciences).

Induced VISTA+ Tfr-like cells and VISTA- Tfr-like cells were sorted by flow cytometry and assigned to blank control, negative control, and target intervention groups as aforementioned. After lentiviral transduction, cells in each group were cultured separately. At the end of culture, the culture supernatant from each group was collected, and cytokine concentrations were measured by ELISA using a microplate reader (Bio-Rad Laboratories, Inc.; xMark^™). For each monoculture group, IL-10, TGF-β1, IL-21 and granzyme B were detected (detection kits from Multi Sciences; cat. nos. EK110, EK981, EK121 and EK158), and the assays were performed according to the manufacturers' instructions. Three independent experiments were performed, and all replicates were included in the statistical analysis.

For the extraction of total protein from ovarian cancer tissues, RIPA lysis buffer (Wuhan Boster Biological Technology, Ltd.; cat. no. AR0105) containing a protease inhibitor (Wuhan Boster Biological Technology, Ltd.; cat. no. AR1178) was used. The tissues were ground in liquid nitrogen and mixed with prechilled lysis buffer. The mixture was incubated at 4°C for 60 min, followed by centrifugation at 13,500 × g, 4°C (Shanghai Lishen Scientific Equipment Co., Ltd.; Neofuge) for 15 min to collect the supernatant. Protein concentration was determined using the Easy II Protein Quantitative Kit (TransGen Biotech Co., Ltd.; cat. no. DQ111-01), and total protein quantification was performed via the bicinchoninic acid (BCA) colorimetric assay. The absorbance was measured at 562 nm with a microplate reader (Bio-Rad Laboratories, Inc.; xMark), and the protein concentration was calculated accordingly. Protein samples were mixed with SDS-PAGE loading buffer (containing β-mercaptoethanol; Amresco, LLC; cat. no. 0482), denatured at 100°C for 5 min, and centrifuged at 13,500 × g, 4°C for 5 min. Then, 50 µg of protein per well and a prestained protein marker (Beijing Solarbio Science & Technology Co., Ltd.; cat. no. PR1910) were loaded onto 10% separating SDS-PAGE gel with 4% stacking gel. Following electrophoresis, proteins were transferred to a 0.45 µm PVDF membrane (MilliporeSigma; cat. no. IPVH00010) using a Mini-PROTEAN Tetra System transfer apparatus (Bio-Rad Laboratories, Inc.). The membrane was blocked with 5% non-fat milk blocking solution at room temperature for 1 h. All TBST washing buffer contained 0.05% (v/v) Tween-20. After rinsing with TBST, the membrane was incubated with primary antibodies separately: anti-β-actin (SinoBiological, cat. no. 100166-MM10, 1:1,000 dilution) and anti-VISTA (Abcam; cat. no. ab300042, 1:1,000 dilution) at 4°C overnight. After rinsing with TBST, the corresponding secondary antibodies were added: Goat anti-mouse IgG H&L (HRP) (Abcam; cat. no. ab205719, 1:10,000 dilution) and goat anti-rabbit IgG H&L (HRP) (Abcam; cat. no. ab205718, 1:5,000 dilution), followed by incubation at room temperature for 1 h. After rinsing with TBST, the membrane was developed with SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Fisher Scientific, Inc.; cat. no. 34580), and imaged using a ChemiScope 3000 chemiluminescence imaging system (Clinx Science Instruments Co., Ltd.). For quantitative analysis, Band intensities were measured by manually drawing rectangular regions of interest around each target band in ImageJ, band gray value quantification was completed using ImageJ software v1.54f (National Institutes of Health); total protein levels were normalized to β-actin as the loading control; relative phosphorylation levels of AKT and ERK1/2 were calculated as the ratio of phosphorylated (p-) protein to the corresponding total protein (p-AKT/total AKT, p-ERK1/2/total ERK1/2). Three independent experiments were performed, and all replicates were included in the statistical analysis.

Total RNA extraction from human ovarian tissues was performed using TRIzol^® Reagent (Thermo Fisher Scientific, Inc.; cat. no. 15596026), and all procedures for RNA isolation, cDNA reverse transcription and quantitative real-time PCR strictly followed the standard protocols provided by each reagent manufacturer. TRIzol^® Reagent was used as the dedicated lysis/extraction buffer for total RNA isolation, which contains phenol, guanidinium isothiocyanate and other proprietary components for simultaneous RNA, DNA and protein separation. Tissue (50 mg) was ground in liquid nitrogen, then 1 ml of TRIzol was added. The mixture was incubated at room temperature for more than 15 min, followed by addition of 200 µl of chloroform, thorough mixing, and incubation for 5 min. Centrifugation was conducted at 13,500 × g, 4°C (Thermo Fisher Scientific, Inc.; Heraeus Multifuge X1R) for 15 min. An equal volume of isopropanol was added to the supernatant, and the mixture was placed at −20°C for 60 min, followed by centrifugation at 13,500 × g, 4°C for 15 min. The precipitate was washed twice with 75% ethanol (centrifugation at 13,500 × g, 4°C for 15 min each time) then dissolved in RNase-free water. RNA concentration and purity were measured using a micro-spectrophotometer (Hangzhou Allsheng Instrument Co., Ltd.; Nano-100), with an A260/A280 ratio of 1.8–2.1 considered acceptable. RNA integrity was assessed by agarose gel electrophoresis (agarose: Sangon Biotech Co., Ltd.; cat. no. A610013-0250; electrophoresis power supply: Beijing Baygene Biotech Co., Ltd.; BG-Power600i; electrophoresis tank: Beijing Liuyi Biotechnology; cat. no. DYCZ-21) combined with a gel imaging system (Tanon Science and Technology Co., Ltd.; 2500), with clear 28S and 18S bands considered acceptable. First-strand cDNA synthesis was performed using 5X All-In-One RT MasterMix (Applied Biological Materials; cat. no. G492). The reaction system contained 1,000 ng of total RNA, 4 µl of 5X RT MasterMix, and RNase-free water to make up the total volume to 20 µl. The reaction was carried out in a PCR instrument (Bio-Rad Laboratories, Inc.; MyCycler Thermal Cycler) under the following conditions: 37°C for 15 min, 60°C for 10 min, and 95°C for 3 min. The synthesized cDNA was stored at −80°C and diluted 1:1 before use. For qPCR, EvaGreen Express 2XqPCR MasterMix-Low Rox (Applied Biological Materials; cat. no. G891) was used. The primers were VSIR-F (TGTACCCATCCTCCTCCCAG), VSIR-R (GAATGTTGCTGTCCATCCGC), GAPDH-F (TGTTGCCATCAATGACCCCTT), and GAPDH-R (CTCCACGACGTACTCAGCG). The qPCR reaction system contained 10 µl of 2XMasterMix, 0.5 µl of each forward and reverse primer, 2 µl of cDNA, and RNase-free water to adjust the total volume to 20 µl. The reaction was run on a Real-Time PCR instrument (Applied Biosystems; Thermo Fisher Scientific, Inc.; 7500 Fast) with the following program: Pre-denaturation at 95°C for 3 min, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing/extension at 60°C for 1 min. The relative expression level of the target gene was calculated using the 2^−ΔΔCq method (31). Results were obtained from three independent experiments, with three technical replicates for each gene.

Human ovarian tissue sections were incubated at 65°C in a constant-temperature incubator (Shanghai Jinghong; Model DNP-9272) for 1.5–2 h. All ovarian tissue specimens were fixed in 4% paraformaldehyde at room temperature for 24 h before paraffin embedding and sectioning. Subsequently, dewaxing was performed with xylene twice, followed by hydration through absolute ethanol I and II, 95, 90, 80 and 70% ethanol and distilled water. The sections were placed in 0.01 M citrate buffer (pH 6.0) and heated in a pressure cooker; heating was discontinued after steam release. The sections were soaked for 10 min, then the lid was opened to cool to room temperature. PBS was used as the uniform washing reagent for all subsequent washing steps throughout the immunofluorescence staining procedure. After washing with PBS, the tissue boundaries were marked with a hydrophobic barrier pen (Fuzhou Maxim; cat. no. PEN-0002). Then, 3% H₂O₂ was added dropwise to cover the tissue and the sections were incubated at room temperature in the dark for 25 min to inactivate endogenous peroxidase. After washing with PBS, excess buffer was removed and 3% goat serum (Wuhan Servicebio Technology Co., Ltd.; cat. no. G1209) was added dropwise for blocking. Diluted primary antibodies were applied sequentially: anti-VISTA antibody (Abcam; cat. no. ab300042; 1:2,000 dilution), anti-CD4 antibody (Abcam; cat. no. ab133616; 1:2,000 dilution), CXCR5 antibody (Abcam; cat. no. ab254415; 1:2,000 dilution), and FOXP3 antibody (Santa Cruz Biotechnology, Inc.; cat. no. sc-53876; 1:1,000 dilution). The sections were incubated overnight at 4°C in a humidified chamber, then washed with PBS. HRP-conjugated goat anti-rabbit or anti-mouse secondary antibodies were applied as appropriate, followed by incubation at room temperature for 50 min. The sections were then washed with PBS. Corresponding tyramide conjugates (100 µl each) were added dropwise: iF440-Tyramide (Wuhan Servicebio Technology Co., Ltd. cat. no. G1250), iF546-Tyramide (Wuhan Servicebio Technology Co., Ltd.; cat. no. G1251), iF488-Tyramide (Wuhan Servicebio Technology Co., Ltd.; cat. no. G1231), and iF647-Tyramide (Wuhan Servicebio Technology Co., Ltd.; cat. no. G1232) (all at 1:500 dilution). The sections were incubated at room temperature in the dark for 10 min, then washed with PBS. DAPI staining solution (Wuhan Boster Biological Technology, Ltd.; cat. no. AR1176) was added dropwise, and the sections were incubated at room temperature in the dark for 10 min, followed by washing with PBS. Antifade mounting medium (Wuhan Boster Biological Technology, Ltd.; cat. no. AR0036) was added dropwise for mounting. Images were acquired using a laser confocal microscope (Nikon Corporation; Model AIRHD25).

Statistical analysis was performed using SPSS 19.0 software (IBM Corp.). Unless otherwise specified, in vitro data were derived from three independent experiments. The Shapiro-Wilk test was used to assess data normality, and Levene's test was used to assess homogeneity of variance across groups. Normally distributed continuous data were summarized as the mean ± standard deviation. One-way analysis of variance (ANOVA) was used for overall comparisons among multiple groups, followed by pairwise comparisons using Tukey's multiple comparisons test (for homogeneous variance) or Dunnett's T3 test (for heterogeneous variance). Non-normally distributed continuous data were summarized as the median (interquartile range) [M (P₂₅-P₇₅)]. The Kruskal-Wallis H test was used for overall comparisons among multiple groups, and pairwise post-hoc comparisons were performed using Dunn's test with false discovery rate correction for multiple testing, which is consistent with the analytical workflow implemented in GraphPad Prism software (Dotmatics). Spearman rank correlation was used to assess the association between VISTA expression and Tfr-like cell infiltration density. Exact P-values are reported for all statistical comparisons. P<0.05 was considered to indicate a statistically significant difference (Table SII, Table SIII, Table SIV, Table SV, Table SVI, Table SVII, Table SVIII, Table SIX). Graphs were generated using GraphPad Prism 9.0 software (Dotmatics).

Ovarian tissue sections from non-malignant controls and patients with different pathological stages (Stage I/II, III, and IV) of ovarian cancer were collected, and CD4, FOXP3, CXCR5, and VISTA expression was assessed by immunofluorescence staining (Fig. 1A). CD4 protein (red fluorescence, primarily localized to the cell membrane) showed the highest mean fluorescence intensity in the non-malignant controls group, which gradually decreased with advancing ovarian cancer stage (Stage I/II → III → IV). By contrast, the expression of FOXP3 (yellow fluorescence, nuclear localization), CXCR5 (green fluorescence, membrane localization) and VISTA (cyan fluorescence, membrane localization) showed the opposite pattern, with low expression in non-malignant controls and progressive increases across disease stages. A small population of VISTA-positive Tfr-like cells coexpressing all four proteins was also observed in ovarian tissues, and their apparent abundance and fluorescence intensity increased with advancing tumor stage. Compared with the non-malignant controls group, CD4 expression was markedly decreased in all stages of ovarian cancer, while the other markers were markedly increased (Fig. 1B-F). Differences between adjacent stages were also statistically significant, suggesting that changes in these four markers and VISTA+ Tfr-like cells may be associated with ovarian cancer progression.

qPCR and western blotting were used to examine VISTA expression across the clinical groups. VISTA mRNA expression was low in non-malignant control tissues (Fig. 2A). Across ovarian cancer stages, VISTA mRNA expression showed an overall upward trend from stage I/II to stages III and IV. However, one-way ANOVA indicated that the between-group differences were not statistically significant. Nevertheless, some individual stage III/IV samples showed mRNA levels approximately threefold higher than those in non-malignant controls, indicating substantial interindividual variability. By contrast, changes in VISTA protein levels were more pronounced. VISTA protein expression differed markedly among the stage I/II, stage III, and stage IV groups and associated positively with FIGO stage. These findings indicate that upregulation of VISTA protein expression may be associated with ovarian cancer progression (Fig. 2B and C).

Naive CD4+ T cells were isolated using the Naive CD4+ T Cell Isolation Kit II according to the manufacturer's instructions. The isolated naive CD4+ T cells were stimulated with CD3 monoclonal antibody, CD28 monoclonal antibody, recombinant IL-2, recombinant TGF-β1, and recombinant IL-23 for 5 days to induce differentiation (Fig. 3A). After induction, phenotypic characterization and sorting were performed using flow cytometry. CD25+CD127- cells were selected, and the Tfr-like subset was further identified by CXCR5 expression. VISTA+ Tfr-like cells (CD25+CD127-CXCR5+VISTA+) and VISTA- Tfr-like cells (CD25+CD127-CXCR5+VISTA-) were sorted (Fig. 3B). To investigate the functional role of VISTA, VISTA expression was manipulated by lentiviral transduction: VISTA+ Tfr-like cells were transduced with VISTA-shRNA lentivirus (knockdown group), and VISTA- Tfr-like cells were transduced with LV–VISTA overexpression lentivirus (overexpression group). Blank control and negative control virus groups were also included. qPCR results showed that VISTA mRNA levels were decreased in the knockdown group and increased in the overexpression group compared with the control groups (Fig. 3C and F). Western blot analysis confirmed corresponding changes at the protein level, supporting effective modulation of VISTA expression (Fig. 3D, E, G and H).

CFSE fluorescence decreases by approximately half with each cell division, allowing flow-cytometric assessment of proliferation under 488-nm excitation. Proliferation assays showed that the CFSE fluorescence decay ratio was markedly higher in VISTA+ Tfr-like cells than in VISTA- Tfr-like cells, indicating greater proliferative activity in the VISTA+ subset (Fig. 4A and B). Following VISTA overexpression in VISTA- Tfr-like cells, the CFSE fluorescence decay ratio was markedly higher than in the blank control and OE negative control groups (Fig. 4C). Conversely, silencing VISTA in VISTA+ Tfr-like cells resulted in a markedly lower ratio than in the blank control and shRNA negative control groups (Fig. 4D). These findings indicate that VISTA promotes Tfr-like cell proliferation, whereas VISTA silencing reduces proliferation. ELISA results revealed that the levels of IL-10 and TGF-β1 in the supernatant of VISTA+ Tfr-like cells were markedly higher than those in VISTA- Tfr-like cells, whereas the levels of IL-21 and granzyme B were markedly lower (Fig. S2A-D). After VISTA overexpression in VISTA- Tfr-like cells, IL-10 and TGF-β1 levels markedly increased, while IL-21 and granzyme B levels markedly decreased (Fig. 4E). By contrast, silencing VISTA in VISTA+ Tfr-like cells led to markedly reduced IL-10 and TGF-β1 levels and markedly elevated IL-21 and granzyme B levels (Fig. 4F). These findings indicate that VISTA promotes anti-inflammatory cytokine secretion and suppresses immune effector molecules, producing a profile consistent with enhanced immunosuppressive function in Tfr-like cells.

Following VISTA overexpression in VISTA- Tfr-like cells, the levels of p-AKT and p-ERK1/2 were markedly lower than in the blank control and negative control groups. Conversely, silencing VISTA in VISTA+ Tfr-like cells using VISTA shRNA resulted in markedly higher p-AKT and p-ERK1/2 levels than in the blank and negative control groups (Figs. 5A-I, S2E-H). By contrast, total AKT and total ERK1/2 protein levels showed no significant differences across groups, suggesting that VISTA alters pathway phosphorylation without affecting total AKT or ERK1/2 abundance.

Previous studies suggest that AKT phosphorylation can inhibit Treg function³¹ and may correlate negatively with the abundance of CXCR5+FOXP3+ cells, a characteristic Tfr-like phenotype, in a model of acute colitis (32). In addition, ERK inhibition can promote FOXP3 expression (33) and FOXP3 is a key transcriptional regulator of Treg and Tfr-like cell function. Together, these observations are consistent with an inverse relationship between AKT/ERK1/2 phosphorylation and Tfr-like cell proliferation and immunosuppressive activity. These data revealed that altered VISTA expression coincided with changes in AKT and ERK1/2 phosphorylation alongside shifts in Tfr-like cell proliferation and immunosuppressive activity. These observations identified correlative changes between VISTA levels, AKT/ERK1/2 phosphorylation and Tfr-like cell functional phenotypes.