Cutting Edge: Low-dose Recombinant IL-2 Treatment Prevents Autoantibody Responses in Systemic Lupus Erythematosus via Regulatory T Cell–independent Depletion of T Follicular Helper Cells

Loss of B cell tolerance and production of antinuclear Abs (ANAs) by Ab-secreting cells (ASCs) plays a major role in disease progression in several autoimmune diseases, including systemic lupus erythematosus (SLE) (1). ASCs develop via two main pathways: the germinal center (GC)-dependent route and the extrafollicular (EF) pathway. Although the exact developmental path leading to the onset of pathogenic autoreactive ASCs remains elusive, recent studies indicate that EF ASCs are critical drivers of SLE progression (2–4). Corresponding with this view, human SLE flares are characterized by EF ASC expansion, and their precursors are prevalent in active SLE and significantly increased in patients with active nephritis (5, 6). Studies in lupus mouse models also show a significant contribution of the EF pathway to the circulating ASC pool, further supporting the role of EF ASCs in SLE progression (7–14).

An increased frequency of CD4⁺ T follicular helper (Tfh) cells correlates with disease severity in both murine and human SLE, suggesting a potential role for Tfh cells in disease progression (1, 10, 15–18). Further supporting a role for Tfh cell help in promoting SLE pathogenesis, the absence of key factors involved in Tfh cell differentiation and maintenance, such as IL-6 (19) or ICOS (10), prevents lupus nephritis in lupus-prone mice. However, because these factors are involved in other immunological pathways, it is uncertain whether the improvements are due to the lack of Tfh cells or additional effects.

Tfh cells support B cell survival and differentiation (20), providing a plausible mechanism for how Tfh cells could contribute to promote disease activity. In agreement with this view, although some early studies suggest that cytokine stimulation or TLR signaling alone can support the development of autoreactive EF ASCs independently of T cells (21–23), most studies suggest a T cell–dependent origin for these cells (4, 24–27). Correspondingly, autoreactive EF ASCs in SLE patients and SLE-prone mice can arise from non-autoreactive B cell clones through somatic hypermutation, a process that requires T cell assistance (24–26). Furthermore, T cell–derived CD40L and IL-21 significantly enhance autoreactive EF ASC responses and autoantibody production (6, 27–30).

Importantly, however, IL-21 production is not unique to Tfh cells (31–33), and other non-Tfh cell subsets have been shown to efficiently help EF ASC responses even in the absence of conventional Tfh cells (30, 33). Thus, the requirement for true Tfh cell help in developing autoreactive EF ASC responses remains unclear. Hence direct proof is needed to confirm a cause-and-effect relationship between excessive conventional Tfh cell responses and the expansion of autoreactive ASCs in SLE.

Low-dose rIL-2 prevents immunopathology in SLE patients and lupus-prone mice, emerging as a promising approach for SLE treatment (34–37). It is generally believed that rIL-2 prevents SLE pathology by expanding Foxp3⁺ regulatory T cells (Tregs), prompting extensive efforts to selectively target rIL-2 to Tregs (38–40). However, studies in the context of protein immunization and viral infection demonstrate that excessive IL-2 signaling also inhibits Tfh cell development by preventing the upregulation of Bcl6, the master regulator of Tfh cell differentiation (41–45). Considering the putative involvement of Tfh cells in promoting autoantibody responses, we have previously hypothesized that suppression of Tfh cells by low-dose rIL-2 may be a major mechanism underlying its immunosuppressive effects (46, 47). In line with this, recent studies show reduced Tfh cell frequencies in low-dose rIL-2–treated SLE patients (48–50). However, whether rIL-2 mediates additional immunosuppression effects in SLE independently of Tregs remains to be formally demonstrated.

In this article, we show that treatment with low-dose rIL-2 efficiently depleted Tfh cells and prevented autoreactive ASC responses in lpr/lpr mice, which develop a SLE-like syndrome characterized by the aberrant expansion of EF ASCs (10, 11, 13, 14, 26, 51). This effect was confirmed in the B6.FcgR2b^−/−Yaa model. Importantly, low-dose rIL-2 treatment prevented Tfh cell responses independently of Tregs, because Tfh cells were efficiently eliminated in Treg-depleted B6.lpr/lpr/Foxp3^DTR/GFP mice after rIL-2 administration. To confirm that the lack of Tfh cells alone was sufficient to prevent Ab-mediated pathology, we generated Tfh cell conditional deficient B6.Bcl6^fl/fl/Cd4^cre/+/lpr/lpr mice. Despite normal frequencies of Treg cells in these animals, the absence of Tfh cells prevented autoreactive ASC responses, ANA production, and Ab-mediated pathology. Our findings underscore the critical role of canonical Tfh cells in promoting autoreactive ASC responses in lupus-prone mice. Moreover, our data reveal, to our knowledge, a novel immunosuppressive function of IL-2 that is independent of Tregs, providing an alternative mechanism to explain the clinical benefits of low rIL-2–based immunotherapies.

C57BL/6 (wild-type [WT]), B6.MRL-Fas^lpr/J (LPR), B6.129S2-IL6^tm1Kopf/J (Il-6^−/−), B6.129(Cg)-Foxp3^{tm3(Hbegf/GFP)Ayr/J} (B6.Foxp3^DTR/GFP), B6.SB-Yaa/J, B6;129S-Fcgr2b^tm1Ttk/J, and Tg(Cd4-cre)1Cwi/BfluJ (Cd4-cre) were originally purchased from The Jackson Laboratories. LPR mice were crossed to Il-6^−/− to generate LPR.Il-6^−/− mice and to B6.Foxp3^DTR/GFP to obtain B6.LPR.Fox-3^DTR/GFP mice (LPR.Foxp3^DTR). B6.SB-Yaa/J mice were crossed to B6;129S Fcgr2b^tm1Ttk/J mice to obtain FcgR2b^−/−Yaa (Yaa) mice. B6.Bcl6^fl/fl mice (52) were originally obtained from Dr. C. Xiao (Scripps Research Institute) and crossed to CD4-cre mice to generate Tfh-deficient (B6.Bcl6^fl/fl/Cd4^cre/+) mice. B6.Bcl6^fl/fl/Cd4^cre/+ mice were then crossed to LPR mice to generate B6.LPR/Bcl6^fl/fl/Cd4^cre/+ mice (LPR.Tfh^−/− mice). All mice were bred in the University of Alabama at Birmingham animal facility. All experiments with LPR mice and related strains were performed with female mice. Experiments with Yaa mice were performed with male mice. In the indicated experiments, mice were i.p. treated with human rIL-2 (National Cancer Institute) at the indicated time points and doses. In the indicated experiments, experimental animals received an i.p. injection of 50 µg/kg diphtheria toxin (DT; Sigma) at the indicated time points. All experimental procedures involving animals were approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee and were performed according to guidelines outlined by the National Research Council.

Lymph nodes (LNs) and spleens were harvested and mechanically disaggregated to obtain cell suspensions. Spleen RBCs were lysed with RBC lysis buffer (150 mM NH₄Cl, 10 mM KHCO₃, and 0.1 mM EDTA) for 5 min. All cell suspensions were filtered through a 70-μm nylon strainer. Cells were resuspended and stained with fluorochrome-conjugated Abs diluted in 2% calf serum-PBS. Fluorochrome-labeled Bcl6 PE (clone K112-91), CD11c (clone HL3), CD138 (clone 281-2), CD19 (clone 1D3), CD25 (clone PC61), CD4 (clone RM4-5), CD44 (clone IM7), CD8 (clone 53-6.7), CXCR5 (clone 2G8), TCRβ (clone H57 597), and CD3 (clone 17A2) were obtained from BD Biosciences. Fluorochrome-labeled CD38 (clone 90) and IRF4 (clone IRF4.3E4) were purchased from BioLegend. Fluorochrome-labeled FOXP3 PE-Cy7 (clone FJK-16 s), GL7 (clone GL-7), and PD1 (clone J43) were obtained from eBioscience. The fluorochrome-labeled polyclonal anti-mouse Ig was from Southern Research. Biotin-conjugated primary Abs were detected with fluorochrome-labeled streptavidin from BD Biosciences. Dead cell exclusion was performed using 7-AAD (Calbiochem). Intracellular staining for transcription factors was performed with the mouse Treg staining kit (eBioscience), following the manufacturer’s instructions. Flow cytometry was performed using an Attune NxT Flow Cytometer (Thermo Fischer Scientific). Data were analyzed using FlowJo v.10.8.0 software (Treestar).

For anti-dsDNA IgG determination, 96-well plates (Corning Clear Polystyrene 96-Well Microplates) were treated with poly-l-lysine (Sigma Aldrich) for 2 h and coated overnight with 5 μg/ml calf thymus DNA (Sigma Aldrich). For anti-histone IgG quantification, ELISA plates were coated overnight with 16.6 μg/ml Histone from calf thymus (Sigma Aldrich). Coated plates were then blocked for 1 h with 1% BSA-PBS. Serum from mice was collected and serially diluted in PBS with 10 mg/ml BSA and 0.1% Tween 20 before incubation on coated plates. After washing, the bound Ab was detected with HRP-conjugated goat anti-Mouse IgG (Southern Biotech). 2,2′-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) was used as the colorimetric substrate for HRP, and the resulting product was quantified by spectrophotometry at 405 nm (OD).

For ELISPOT assays, multiscreen 96-well plates (MAHAS4510; Millipore) were treated with poly-l-lysine (Sigma Aldrich) for 2 h and coated overnight with calf thymus DNA (Sigma Aldrich). Plates were washed with PBS and blocked with complete medium (RPMI supplemented with 10% FBS, 0.5% penicillin [100×], 0.5% streptomycin [100×], 1% glutamine [200 mM], 1% sodium pyruvate (100 mM), 1% HEPES (pH 7.4, 1 M), 0.15% sodium bicarbonate, 1.2% amino acids (50×), 1.2% nonessential amino acids (100×), 1.2% vitamins (100×), 0.7% glucose, and 0.1% 2-ME (1000×, 55 mM). Single-cell suspensions from LNs were prepared as described earlier, washed, diluted in complete medium, and cultured on coated plates. After 5 h, the wells were washed with PBS containing 0.5% BSA and 0.05% Tween 20, and IgG was detected using alkaline phosphatase–conjugated goat anti-mouse IgG (Jackson Immunoresearch). Plates were washed with 0.2% Tween 20 in PBS and developed with BCIP/NBT (Moss Substrates). Spots were recorded using a CTL Immunospot S6 Macroplate Imager Reader (New Life Scientific) and counted manually.

Kidneys were frozen in OCT (Tissue-Tek; Sakura). Frozen kidney sections (5–7 μm) were obtained using a Leica CM1850 cryostat (Leica), fixed with acetone, blocked with a 2% albumin-PBS, and probed with Alexa Fluor 488–labeled anti-IgG H + L Ab (Invitrogen). Slides were mounted using Fluoroshield with DAPI histology mounting medium (Sigma Aldrich). Images were acquired with a Nikon ECLIPSE Ti microscope. Corrected total cell fluorescence (CTCF) was calculated using the FIJI software (version 1.53) as previously described (53): CTCF = integrated density − (area of selected cell × mean fluorescence of background).

High-powered (×40 objective magnification) images of H&E-stained kidney specimens, focusing on renal glomeruli and the surrounding interstitium, were evaluated for pathological lesions adapted from an established scoring scheme (54). These evaluations were conducted based on blinded, semiquantitative assessments by a board-certified veterinary anatomic pathologist (J.B.F.). The spectrum of lesions assessed included glomerular cellularity, necrosis, crescent formation, and glomerular sclerosis for glomeruli, as well as tubular degeneration, interstitial fibrosis, and inflammation for the renal interstitium. For the renal interstitium/tubules, the scores were derived by averaging the values for each parameter across 8–14 high-power fields (×40 magnification). These average scores were then summed to obtain the total renal interstitial score. The overall renal pathology score for each individual was determined by summing the average scores of both glomerular and renal interstitial assessments.

GraphPad Prism software (Version 10) was used for data analysis. The statistical significance of differences in mean values was determined using a two-tailed Student t test or one-way ANOVA with post hoc Tukey’s multiple comparison test. The p values <0.05 were considered statistically significant: *p < 0.05, **p < 0.01 and ***p < 0.001.

Data in this study will be made available upon request from the corresponding author.

Somatically hypermutated EF ASCs characteristically expand in mice carrying the homozygous lymphoproliferation spontaneous mutation Fas^lpr (lpr/lpr mice) (10, 11, 13, 14, 26, 51). To determine whether CD4⁺ T cells with a Tfh cell-like phenotype accrued in C57BL/6.lpr/lpr mice (herein referred to as LPR mice), we enumerated PD-1^hiCXCR5^hiBcl6^hi CD4⁺ T cells (Supplemental Fig. 1A) in the LNs of 6-mo-old LPR and C57BL/6 WT mice (Fig. 1A, 1B). The frequency (Fig. 1A) and number (Fig. 1B) of Tfh cells were significantly increased in LPR compared with WT mice.

We next enumerated ASCs (CD138⁺IRF4^hiIg⁺; Supplemental Fig. 1B) in WT and LPR mice (Fig. 1C, 1D). As expected, correlating with the presence of serum anti-dsDNA Abs (Supplemental Fig. 1C) and glomerular IgG deposits in the kidneys (Supplemental Fig. 1D, 1E), the frequency (Fig. 1C) and number (Fig. 1D) of ASCs were significantly increased in the LPR mice compared with their WT counterparts. Similarly, results were obtained in the spleen (Fig. 1E, 1F). Despite the abnormal accumulation of ASCs, the frequency of spontaneous GC B cells was similar in both groups (Supplemental Fig. 1F, 1G). This observation is consistent with previous studies suggesting a predominant EF origin of the ASC pool in the LPR model (10, 11, 13, 14, 26, 51). Supporting this view, LPR ASCs expressed high levels of T-bet, CD11c, and CXCR3 (Supplemental Fig. 1H, top panel), which are highly expressed by EF ASCs (6, 55, 56). In contrast, GC B cells were T-bet^loCD11c^loCXCR3^lo relative to ASCs (Supplemental Fig. 1H, bottom panel).

Next, we examined whether treatment with low-dose rIL-2 could prevent Tfh cell responses in LPR mice. Three-month-old LPR mice were already ANA positive compared with controls (Supplemental Fig. 1I). Thus, we administered 30,000 U of rIL-2 or vehicle to LPR mice between 3 and 4 mo of age twice a week for 10 wk (Supplemental Fig. 1J). This dose has been shown to prevent Tfh cell development in vivo (41, 43, 45). As predicted, the treatment significantly reduced the frequency (Fig. 1G) and number (Fig. 1H) of Tfh cells in rIL-2–treated LPR mice compared with their vehicle-treated counterparts. No differences were detected in CD4⁺Foxp3⁺PD-1^hiCXCR5^hiBcl6^hi T follicular regulatory cells (Supplemental Fig. 1K–M). Correlating with the diminished Tfh cell response, we observed a reduction in the frequency (Fig. 1I) and number (Fig. 1J) of ASCs in the IL-2–treated mice relative to controls. In addition, we enumerated anti-dsDNA–specific ASCs by ELISPOT. The number of anti-dsDNA, IgG-secreting ASCs was significantly reduced in the rIL-2–treated group compared with their control counterparts (Fig. 1K). In line with the reduction in the effector B cell response, the titers of anti-dsDNA IgG Abs (Fig. 1L), the glomerular deposits of IgG in the kidney (Fig. 1M, 1N), and the glomerulonephritis histopathology score (Fig. 1O) were decreased in mice receiving rIL-2. These data indicate that low-dose rIL-2 treatment effectively inhibits Tfh cell and autoreactive ASC responses in LPR mice.

To confirm that the effect of the low-dose rIL-2 treatment in preventing Tfh and effector B cell responses was not unique to the LPR model, we treated 2-mo-old Yaa mice with ongoing disease (Supplemental Fig. 2A) with either rIL-2 or PBS. The Yaa mice carry the Yaa mutation, a duplication and translocation of a segment of the X chromosome containing several genes, including TLR7 (Tlr7), to the Y chromosome. When crossed with FcgR2b^−/− mice, the resulting Yaa mice spontaneously develop effector B cell responses and an aggressive SLE-like disease (57–59). For simplicity, we will refer to these mice as Yaa mice. Similar to the LPR model, studies suggest that the EF pathway strongly contributes to the autoantibody production in the Yaa mice (6, 60–62).

The frequency (Fig. 2A) and number (Fig. 2B) of Tfh cells were significantly reduced in Yaa mice after low-dose rIL-2 treatment. Correlating with this observation, the ASC response was significantly diminished in the rIL-2–treated Yaa mice relative to the control-treated counterparts (Fig. 2C, 2D). In addition, the titers of anti-dsDNA IgG Abs (Fig. 2E) and glomerular deposits of IgG in the kidneys (Fig. 2F) decreased in the rIL-2–treated mice compared with vehicle-treated controls. Collectively, these findings indicate that low-dose rIL-2 treatment prevents the accumulation of Tfh cells in lupus-prone mice, an effect that correlates with diminished autoreactive ASC response after rIL-2 treatment.

Previous studies have shown that low-dose rIL-2 treatment promotes Foxp3⁺ Treg cell expansion in vivo (39, 47). Accordingly, the frequency of Foxp3⁺ Treg cells was increased in rIL-2–treated LPR and Yaa mice compared with controls (Fig. 3A, 3B). Because Treg cells are known to suppress T cell responses, we next decided to test whether the reduction in the frequency of Tfh cells observed in these mice following rIL-2 treatment mice was Treg cell dependent. To do this, we crossed LPR mice with B6.Foxp3^DTR/GFP mice to generate B6.LPR.Foxp3^DTR/GFP mice (LPR.Foxp3^DTR). We then treated 7-mo-old LPR.Foxp3DTR and control LPR mice with 30,000 U of rIL-2 or vehicle for 7 d and concurrently administered DT to deplete Treg cells in the LPR.Foxp3^DTR mice. As a control, a group of LPR.Foxp3^DTR mice did not receive DT (Fig. 3C). As expected, DT administration effectively depleted Treg cells in rIL-2–treated LPR.Foxp3^DTR mice, but not in the rIL-2–treated control LPR mice (Fig. 3D, 3E). Importantly, despite the absence of Treg cells, the frequency and number of Tfh cells was similarly reduced in the Treg-depleted LPR.Foxp3^DTR mice and the Treg cell–sufficient LPR.Foxp3^DTR and LPR mice following low-dose rIL-2 administration (Fig. 3F, 3G). These data indicate that Tfh cell depletion following low-dose rIL-2 administration does not depend on the presence of Treg cells. This observation aligns with previous studies showing that intrinsic IL-2/STAT5 signaling in CD4⁺ T cells directly represses Tfh cell responses by inhibiting the upregulation of Bcl6, the master regulator of Tfh cell differentiation (41–45). The autoreactive B cell response was not assessed in these mice because Treg cell depletion in adult mice leads to catastrophic T cell overactivation (63). Consequently, the mice become critically ill within the first 2 wk after the initial DT administration and succumb to terminal disease (63).

Although IL-2 inhibits Tfh cell differentiation, evidence from immunization and infection models indicates that IL-6/STAT3 signaling is crucial for optimal Tfh cell responses to exogenous Ags (45, 64). Whether IL-6 is also required for Tfh cell responses to self-antigens is unknown. Notably, previous studies have shown that IL-6 deficiency prevents lupus nephritis in lupus-prone MRL-LPR mice (19). However, these studies did not assess whether the reduced pathology observed in IL-6–deficient MRL-LPR mice was associated with changes in Tfh cells, Treg cells, or alterations in ASC responses.

To assess whether IL-6 deficiency impacted these responses in lupus-prone mice, we generated LPR.Il6^−/− mice. We observed a significant decrease in the frequency (Fig. 4A) and number (Fig. 4B) of Tfh cells in LPR.Il6^−/− mice compared with control LPR mice. Correlating with the relative lack of Tfh cells, ASCs failed to accumulate in IL-6–deficient mice compared with their control counterparts (Fig. 4C, 4D). Consistent with the decline in the ASC response, LPR.Il6^−/− mice exhibited decreased titers of anti-dsDNA IgG Abs (Fig. 4E) and reduced glomerular deposits of IgG in the kidneys (Fig. 4F) relative to controls. Thus, similar to what is observed in low-dose rIL-2–treated mice, the absence of Tfh cells in LPR.Il6^−/− mice was associated with reduced effector B cell responses in these mice.

Importantly, we found that both the frequency (Fig. 4G) and the number (Fig. 4H) of Treg cells were significantly reduced in LPR.Il6^−/− mice compared with control LPR mice, suggesting that the diminished effector B cell response in these mice was unlikely due to enhanced Treg cell activity. Collectively, our data indicate that depletion of Tfh cells leads to the prevention of autoreactive ASC responses, regardless of whether Treg cell responses are enhanced, as observed in the rIL-2–treated LPR mice, or diminished, as seen in LPR.Il6^−/− mice.

Our findings suggest a cause-and-effect relationship between the depletion of Tfh cells and the prevention of autoreactive ASC responses. To directly confirm that the absence of Tfh cells is sufficient to prevent autoreactive ASC responses, we generated Tfh-deficient LPR mice. To do this, we crossed LPR mice with B6.Bcl6^fl/flCd4^cre/+ mice, resulting in the generation of B6.LPR/Bcl6^fl/fl/Cd4^cre/+ mice. For simplicity, these mice will be referred to as LPR-Tfh^−/− mice. As a control, we used B6.LPR/Bcl6^fl/fl control littermates, referred to as LPR-WT.

As predicted, because Bcl6 is required for the differentiation of Tfh cells (20), Tfh cells failed to accumulate in 8-mo-old LPR-Tfh^−/− mice compared with LPR-WT controls (Fig. 5A, 5B). In contrast, no differences were detected in the frequency and number of Treg cells (Fig. 5C, 5D). These results confirm that Tfh cells do not develop in LPR-Tfh^−/− mice, whereas Treg cells accumulate normally.

We next enumerated ASCs in LPR-Tfh^−/− and LPR-WT mice. In line with the absence of Tfh cells, both the frequency (Fig. 5E) and the number (Fig. 5F) of ASCs were significantly diminished in LPR-Tfh^−/− mice compared with LPR-WT mice. Correspondingly, the serum titers of anti-dsDNA and anti-Histone IgG Abs (Fig. 5G) and the glomerular deposits of IgG in the kidneys (Fig. 5H, 5I) were reduced to background levels in LPR-Tfh^−/− mice relative to their control counterparts. These data indicate that Tfh cell deficiency is sufficient to prevent the development of autoreactive B cell responses in lupus-prone mice.

In summary, we show in this study that blocking spontaneous Tfh cell responses—whether through low-dose rIL-2 treatment, IL-6 deficiency, or targeted Tfh cell depletion—effectively reduces autoreactive effector B cell responses in lupus-prone mice, independent of the magnitude of the Treg cell response. These data demonstrate a direct link between the accumulation of self-reactive Tfh cells and the expansion of autoreactive ASCs. Considering the dominant contribution of the EF pathway to autoantibody production in LPR mice, these findings evidence the therapeutic potential of depleting Tfh cells to prevent EF ASC responses in SLE.

Furthermore, our data reveal an immunosuppressive function of IL-2 that operates independently of Treg cells, shedding light on the potential mechanisms underlying the clinical benefits of low-dose rIL-2–based therapies. Thus, the effects of rIL-2–based immunotherapies on Tfh cells should also be considered when evaluating the effects of IL-2–based immunotherapies in preclinical studies and clinical trials. Based on our results, we propose a two-arm model in which the combination of depleting Tfh cells and enhancing Treg cell responses synergistically contributes to preventing SLE pathology after rIL-2 treatment. In this model, enhancing Treg cell activity counteracts aberrant effector T cell and innate cell responses through direct immunosuppression. In contrast, reducing Tfh cells curtails Ab-mediated pathology by preventing the continuous replenishment of autoreactive ASCs, thereby limiting pathogenic B cell responses. Together, these mechanisms synergistically mitigate SLE immunopathology in response to IL-2–based immunotherapy. Thus, targeting both Tfh and Treg cells, rather than Tregs alone, may offer a more effective strategy for addressing the complex immune dysregulation in SLE and other Ab-mediated disorders.

Recent studies have highlighted the role of T peripheral helper (Tph) cells in supporting the pathological production of autoantibodies within inflamed nonlymphoid tissues in various autoimmune diseases, including rheumatoid arthritis and SLE (33, 65, 66). Importantly, similar to Tfh cells, IL-2/STAT5 signaling inhibits Tph cell differentiation (67–69). Therefore, alongside depleting Tfh cells in secondary lymphoid organs, the simultaneous suppression of Tph cell activity in peripheral tissues could further limit autoantibody responses after low-dose rIL-2 administration. Future studies will be required to confirm the effect of rIL-2–based immunotherapies on Tph cells.

Collectively, our results highlight the crucial role of Tfh cells in promoting autoreactive ASC responses and provide proof of concept for therapeutic interventions aimed at targeting pathogenic Tfh cells by modulating the relative availability of IL-2 and IL-6 in vivo.

The authors have no financial conflicts of interest.

We thank Thomas S. Simpler, Kelsey Browning, and Rebecca Burnham for animal husbandry.