Peripheral blood from RA patients and age and gender-matched healthy controls (HC) were obtained from the Department of Rheumatology and Immunology of the Second Affiliated Hospital of Dalian Medical University in China. Detailed clinical characteristics and laboratory features of RA patients and HC are shown in Table 1 . All RA patients in this study fulfilled the American College of Rheumatology (ACR) 1987 revised criteria. Ethics approval was given by the ethics committee of the Second Hospital of Dalian Medical University (2023-253).

Male DBA/1J mice aged 6–8 weeks were purchased from Shanghai SLAC Laboratory Animals Company and housed under pathogen-free conditions at the Laboratory Animal Center of Dalian Medical University. For the primary immunization, mice were subcutaneously injected in the tail with the emulsion formed by type II bovine collagen (Chondrex, USA) and Complete Freund adjuvant (CFA, Chondrex). Administer a booster injection consisting of type II bovine collagen and Incomplete Freund adjuvant (IFA, Chondrex) emulsion on day 21. The injection consisted of 100 µL of an emulsion containing 100 µg of collagen and 2 mg/mL of CFA or IFA.

The study covered the healthy control mice group (n=5), the CIA mice group (CIA mice were treated with vehicle, n=5), and the CIA+2-DG mice group (CIA mice were treated with 2-DG three times a week for 2 weeks on day 28 after immunization, 500 mg/kg, n=5), and the CIA+MitoQ mice group (CIA mice were treated with MitoQ three times a week for 2 weeks on day 28 after immunization, 2 mg/kg, n=5). The mice’s paw thickness and disease activity score were assessed. Disease activity scores were derived from the evaluation of clinical arthritis in all four limbs as reported by the scoring system for the evaluation of arthritis severity. The scoring system was defined as 0, no evidence of erythema and swelling; 1, erythema and mild swelling confined to the tarsals or ankle joint; 2, erythema and mild swelling extending from the ankle to the tarsals; 3, erythema and moderate swelling extending from the ankle to the metatarsal joints; and 4, erythema and severe swelling encompassing the ankle, foot and digits or ankylosis of the limb. All experiments complied with the guidelines established by the Institutional Animal Care and Use Committee (IACUC). The study was approved by the Ethics Committee of the Dalian Medical University (Approval No: AEE24013).

The mice’s ankle joints and paws were fixed overnight in 4% paraformaldehyde, decalcified in 14% EDTA, and embedded in paraffin. Sections of 5-μm thickness were generated from the paraffin tissue blocks. And the sections were stained with hematoxylin-eosin (HE, Solarbio, China), and Safranin O-fast green (SO/FG, Solarbio, China). The expression of hexokinase 2 (HK2, Cell Signaling Technology), lactate dehydrogenase (LDH, Cell Signaling Technology), IL-1β (Affinity), and IL-6 (Affinity) in synovium were analyzed by using immunohistochemistry. The sections were incubated with CD4-FITC (Biolegend) and PD-1-PE (Biolegend) antibodies in a 37°C wet box for 1 h, and the images were scanned under fluorescence microscopy.

For cell surface staining, cells were stained in PBS with Biolegend or eBioscience antibodies (Fixable viability dye (FVD), anti-CD4, anti-CXCR5, anti-PD-1, anti-ICOS, anti-CD40L, anti-CD27, anti-CD28, anti-CCR7, anti-CD107a and anti-Glut1 antibodies) for 20 min.

For intracellular staining of Perforin and Granzyme B, peripheral blood mononuclear cells (PBMCs) were fixed and permeabilized by Cytofix/Cytoperm Intracellular Staining Kit (BD Biosciences) for 60 min after cell surface staining, and labeled with anti-Perforin and anti-Granzyme B antibodies for 30 min.

For intracellular cytokines staining, cells were incubated with 50 ng/mL phorbol myristate acetate (PMA), 1 μg/mL Ionomycin, and 10 μg/mL Brefeldin-A (BFA) for 4 hours before staining with surface antibodies. Then cells were fixed and permeabilized to stain with cytokines antibodies (anti-IL-21, anti-CXCL13, anti-IL-4, anti-IFN-γ, and anti-TNF-α antibodies) for 30 min.

For transcription factor staining, cells were fixed and permeabilized using the Transcription Factor Fixation/Permeabilization Buffer Set (eBioscience) for 45 min. Then cells were stained with anti-BATF, anti-Bcl-6, anti-T-bet, and anti-Eomes.

All stained cells were analyzed on the Flow Cytometer (NovoCyte 2060R) and data were analyzed with NovoExpress software.

Levels of SA-β-Gal activity were assessed by Senescence Assay Kit (Beta Galactosidase, Fluorescence, Abcam). Collect and resuspend cells in 500 µL of fresh media containing 1.5 µL of Senescence Dye per tube. Cells were incubated for 1 h at 37°C, 5% CO2, then washed twice with 500 µL wash buffer, and stained with cell surface antibodies, and analyzed immediately using flow cytometry.

Levels of mtROS were assessed by MitoSOX Red probe (Invitrogen). After staining with surface antibodies, cells were stained with 5 μM MitoSOX Red probe in Hanks’ balanced salt solution buffer for 30 min at 37°C. Then wash the cells gently three times for flow cytometry.

Mitochondrial membrane potential (MMP) was determined by Mitochondrial Membrane Potential Kit (JC-10 assay; Solarbio) according to the instruction of the manufacturer. Briefly, after staining with surface antibodies, cells were stained with JC-10 staining solution at 37°C for 20 min. Then wash the cells gently three times with warm Hank’s for flow cytometry.

For glucose uptake assay, PBMCs from RA patients were resuspended and cultured in glucose free RPMI 1640 medium (Gibco) for 30 min and then cultured with 50 μM D-glucose analog 2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino]-2-deoxy-D- glucose (2-NBDG) (Sigma-Aldrich) for 30 min at 37°C. For HK2 staining, the cells were fixed in 4% paraformaldehyde and permeabilized in 0.5% Triton X-100 before anti-HK2 antibody staining. For LDH staining, the cells were fixed in 4% paraformaldehyde and permeabilized in 90% methanol before LDH antibody staining. All stained cells were analyzed by using flow cytometry.

CD4⁺ T cells (purity range 95%-99%) were purified from PBMCs of RA patients using the Human CD4⁺ T Cell Isolation Kit (BioLegend), and cultured with plate-coated anti-CD3 antibody (5 μg/mL, eBioscience) and anti-CD28 antibody (2 μg/mL, eBioscience) in RMPI-1640 medium containing 10% fetal bovine serum and 1% penicillin/streptomycin for 72 h. In several experiments, CXCL13 (100 μg/mL, Peprotech), CCL2 (100 μg/mL, Peprotech), 2-DG (1 mM, Solarbio) or MitoQ (200 nM, Biovision) was administrated to the cell culture.

Data is presented as means ± standard errors of the mean (SD). GraphPad Prism 9 was used to conduct all statistical analyses. Statistical differences were analyzed by paired t-test, Wilcoxon rank sum test, unpaired t-test, Welch’s t-test, and Mann-Whitney test. Paired t-tests, unpaired t-tests, or Welch t-tests were used to compare parameter data. The Mann-Whitney test or Wilcoxon matched-pairs signed rank test was performed for non-parametric data. The P-values < 0.05 were considered significant. Asterisks mark the significant differences between different groups (*P<0.05; **P<0.01 and ***P<0.001).

To delve into the precise pathogenic characteristics of cTfh and Tph cells, we initially determined the percentages of these two T cell subsets in the peripheral blood of HC and RA patients using flow cytometry. CXCR5⁺PD-1⁺ cTfh and CXCR5^-PD-1⁺ Tph cells were gated from CD4⁺ T cells ( Figures 1A, B ). Results show that RA patients have a higher number of cTfh and Tph cells than HC ( Figure 1C ). And the frequency of Tph cells is higher than that of cTfh cells in RA patients ( Figure 1D ). Further, the percentages of cTfh cells are positively correlated with the levels of serum anti-CCP antibody, and the percentages of Tph cells are positively correlated with erythrocyte sedimentation rate (ESR) in RA patients. No significant correlation is found between the percentages of cTfh or Tph cells and the levels of RF and C-reactive protein (CRP) in RA patients ( Figure 1E ). What’s more, immunofluorescence results show that CD4⁺PD-1⁺ T cells are increased in ankle joint of CIA mice ( Figure 1F ). Compared to control mice, CIA mice show higher levels of Tfh and Tph cells in peripheral blood and spleen, and the frequency of Tph cells is higher than that of cTfh cells ( Figure 1G ). These results indicate that the accumulated cTfh and Tph cells may play distinct roles in the pathogenesis of RA.

Studies have demonstrated that the transcription factor BATF facilitates Tfh cell differentiation by promoting Bcl6 expression in CD4⁺ T cells (20). ICOS, CD40L, CXCL13, IL-4 and IL-21 of Tfh cells are important for B cell activation and differentiation (8, 21–24). Our results show that compared to Tph cells, cTfh cells express higher levels of Bcl6 and BATF ( Figure 2A ), ICOS, CD40L, CXCL13, IL-4 and IL-21 ( Figure 2B ) in RA patients and HC, which indicate that cTfh cells have a more potent ability for B-cell recruitment and assistance than Tph cells.

In addition to B cell-helping function, Tph cells were reported to exhibit cytotoxic activity (16, 17). Transcription factors Blimp1 and T-bet are important for cytotoxic effector T cell differentiation and function (25). Our results show that compared to cTfh cells, Tph cells express higher levels of Blimp1 and T-bet ( Figure 2C ), IFN-γ, TNF-α, perforin, granzyme B and CD107a ( Figure 2D ) in RA patients and HC. According to reports, senescent T cells are characterized by down-regulation of CCR7, CD28 and CD27, and up-regulation of NK receptors and innate-like killing effects (26, 27). Thus, we detected the cellular senescence biomarkers and cytotoxicity-associated molecules in cTfh and Tph cells of RA patients. Results show that compared to cTfh, Tph cells show lower levels of CD28, CD27 and CCR7 ( Figure 2E ), and higher levels of senescence associated SA-β-gal activity ( Figure 2F ) in RA patients and HC. These results indicate that Tph cells have more pronounced cellular senescence characteristics and stronger cytotoxic activity than cTfh cells.

Our previous study found that Iguratimod restrained RA-cTfh cell functions by inhibiting HIF1α-HK2 axis mediated glucose metabolism (6). To explore whether the distinct functions of cTfh and Tph cells are related to cellular glucose metabolism, we detected the glucose uptake ability and the expression of key glycolytic molecules in cTfh and Tph cells of RA patients and HC. Despite these two T cell subsets showing similar glucose uptake ability ( Figure 3A ) in RA patients, cTfh cells express higher levels of glucose transporter 1 (GLUT1), HK2 and LDH than Tph cells ( Figures 3B-D ) in RA patients and HC. These results suggest that cTfh cells exhibit a higher level of glycolysis than Tph cells. Increased ROS production contributes to T cell senescence, and oxidation of NADPH produced by oxidative phosphorylation is one of the main sources of ROS production (28, 29). Given the lower levels of glycolytic molecules observed in Tph cells than cTfh cells, we hypothesized that Tph cell functions are associated with mitochondrial metabolism. In this study, JC-10 and MitoSOX were used to detect mitochondrial function. Results show that Tph cells exhibit higher levels of mitochondrial membrane potential and mtROS ( Figures 3E-F ) in RA patients, which suggest that Tph cells show mitochondrial dysfunction, resulting in elevated levels of mtROS.

According to reports, despite a lack of CXCR5 expression, Tph cells are characterized by expressing CCR2 (8). Here, our results confirm that CXCR5^-PD-1⁺ Tph cells expressed higher level of CCR2 than CXCR5⁺PD-1⁺ cTfh cells from RA patients ( Figure 4A ). Next, to verify the distinct metabolism patterns in cTfh and Tph cells of RA patients, we measured the levels of GLUT1, HK2, LDH and mtROS in RA CD4⁺ T cells stimulated with CXCL13 (ligand for CXCR5) or CCL2 (ligand for CCR2). Results show that CXCL13-CXCR5 signaling up-regulate the expression of GLUT1 and HK2 in cTfh cells, while CCL2-CCR2 signaling promote the production of mtROS in Tph cells ( Figures 4B-E ). These results suggest that the enhanced glycolysis in RA-cTfh cells is related to CXCL13-CXCR5 signaling, while the increased mtROS in RA-Tph cells is related to CCL2-CCR2 signaling.

To further investigate the roles of glycolysis and mtROS in the differential functions of cTfh and Tph cells, we inhibited glycolysis using the glucose analog 2-DG and mitigated mtROS with the mitochondria-targeted antioxidant MitoQ in CD4⁺ T cells from RA patients, we found that 2-DG could reduce the proportion of cTfh, and MitoQ could decrease the proportion of both cTfh and Tph cells ( Figure 5A ). Results show that glycolysis inhibition down-regulate the expression of ICOS, CD40L, IL-4 and IL-21 in both cTfh and Tph cells ( Figures 5B, C ). mtROS scavenging significantly down-regulate the expression of IFN-γ, TNF-α, perforin and granzyme B in both Tph and cTfh cells ( Figures 5D, E ). These results suggest that in RA patients, the enhanced B-cell recruitment and helper functions of cTfh cells may rely on increased glycolysis, while the heightened cytotoxic activity of Tph cells may be associated with elevated mtROS levels.

To confirm the roles of glycolysis and mtROS in the differential functions of cTfh and Tph cells in vivo, we administered 2-DG or MitoQ to CIA mice. Results show that both 2-DG and MitoQ reduce paw swelling and arthritis score ( Figure 6A ), and the expression of HK2 and LDH in ankle joints ( Figure 6B ). Additionally, 2-DG and MitoQ improve the pervasive infiltration of inflammatory cells, cartilage and bone tissue erosion, and synovial hyperplasia ( Figure 6C , upper two panels). 2-DG and MitoQ also reduce the expression of inflammatory cytokines IL-1β and IL-6 ( Figure 6C , lower two panels) and the presence of CD4⁺PD-1⁺ T cells in ankle joints ( Figure 6D ). Furthermore, 2-DG and MitoQ down-regulate the percentages of cTfh and Tph cells in the peripheral blood of CIA mice, although not in the spleen ( Figure 6E ). These results suggest that inhibition of glycolysis or mtROS production alleviate disease severity in CIA mice via restraining the development and recruitment of cTfh and Tph cells.