Ifidancitinib is a selective next-generation JAK1/3 inhibitor predicted to disrupt γc cytokine and IFN-γ signaling. We characterized the effects of selective inhibition of JAK/STAT signaling in mouse T cells by treatment with Ifidancitinib. We found that treatment of mouse T cells with Ifidancitinib over a range of doses strongly inhibited IL-2-stimulated STAT5 phosphorylation ( Figure 1A ). Ifidancitinib also exhibited potent inhibitory effects on JAK1/2-mediated IFN-γ signaling, as shown by its suppression of STAT1 phosphorylation upon stimulation with IFN-γ ( Figure 1B ). These findings demonstrate that Ifidancitinib is a potent JAK1/3 inhibitor.

To directly test the effect of Ifidancitinib-mediated JAK1/3 inhibition on the development of NKG2D⁺CD8⁺ T cells in vitro, we treated IL-15 stimulated T cells with Ifidancitinib at the start of culture. The presence of NKG2D⁺CD8⁺ T cells was quantified after 4 days, and as expected, IL-15 robustly induced the generation of NKG2D⁺CD8⁺ T cells ( Figure 1C ). Compared to vehicle control, however, Ifidancitinib treatment significantly reduced the population of NKG2D⁺CD8⁺ T cells in a dose-dependent manner ( Figure 1D ). Accordingly, the proportions of IFN-γ producing CD8⁺ T cells were also markedly reduced by Ifidancitinib treatment ( Figures 1E, F ). These results indicated that Ifidancitinib-mediated JAK inhibition suppressed the generation and differentiation of IFN-γ producing NKG2D⁺CD8⁺ T cells in vitro.

C3H/HeJ mice with AA exhibit a striking cutaneous ly0mphadenopathy with increased total cellularity, which may be mediated by a milieu of γc cytokines and other helper cells (7). Therefore, we examined the effect of Ifidancitinib on in vitro T cell proliferation driven by IL-2, a known γc cytokine important for T cell proliferation (18). CellTrace Violet-labeled T cells were co-stimulated with anti-CD3 Ab and IL-2 and subsequently treated with either Ifidancitinib or vehicle control. As determined by dye dilution and subsequent flow cytometry analysis, IL-2 strongly promoted T cell proliferation in the vehicle control as expected ( Figures 1G, H ). In contrast, Ifidancitinib strongly suppressed T cell proliferation in a dose-dependent manner ( Figures 1G, H ). Next, to test whether Ifidancitinib could block γc cytokine-driven T cell proliferation in vivo, we adoptively transferred CellTrace Violet-labeled B6 CD45.1 T cell blasts into B6 CD45.2 recipient mice, which were then treated with IL-2 for two days in order to promote T cell proliferation (18). Recipient mice were also treated with varying doses of Ifidancitinib over the same time period and the proliferation of transferred T cells in their lymphoid organs was determined after the two days of treatment. As determined by dye dilution, Ifidancitinib inhibited IL-2-driven T cell proliferation in a dose-dependent manner ( Figures 1I, J ). Taken together, these results indicated that Ifidancitinib robustly inhibited IL-2-driven mouse T cell proliferation both in vitro and in vivo.

We next evaluated the effects of Ifidancitinib on human T cells. Purified CD8⁺ T cells were stimulated with anti-CD3 Ab in the presence of IL-15 to induce NKG2D expression (7). Compared to the control, Ifidancitinib not only robustly inhibited NKG2D expression ( Figures S1A, B ), but also significantly decreased the proportions of IFN-γ producing CD8⁺ T cells ( Figures S1C, D ). Furthermore, Ifidancitinib reduced the proliferative response of T cells in response to anti-CD3 and IL-2 stimulation in a dose-dependent manner ( Figures S1E, F ). These results demonstrate that pharmacological inhibition of JAK1/3 signaling via Ifidancitinib impaired human T cell function in vitro.

Given the striking effect of Ifidancitinib on the suppression of γc cytokine-driven T cell proliferation and function, we next tested the effect of Ifidancitinib on disease development in the C3H/HeJ skin graft model of AA. Upon engraftment of AA affected skin from C3H/HeJ AA donor mice, young recipient C3H/HeJ mice were fed either normal chow or Ifidancitinib-incorporated chow diet (0.5 g/kg) (19). In the control group, all mice developed AA by 8 weeks after engraftment. In contrast, none of Ifidancitinib treated-mice developed AA by 12 weeks after engraftment ( Figures 2A, B ). Concordant with the prevention of disease, immunofluorescence staining of skin revealed scant CD8⁺ infiltrates in Ifidancitinib-treated animals, compared to striking infiltration of CD8⁺ T cells in the skin of control mice ( Figure 2C ). Ifidancitinib-treated mice also showed significantly reduced staining for MHC class I and II on HF epithelium compared to controls ( Figure 2C ). Flow cytometry of single cell suspensions of skin samples confirmed reduced expression of MHC class I and II on HF (CD45-EpCAM+CD200+) in Ifidancitinib-treated mice compared to controls ( Figure 2D ) (20). Furthermore, flow cytometry of single cell suspensions of skin samples further showed that CD45⁺ inflammatory infiltrates, including CD4⁺ T cells and CD8⁺ T cells, were significantly diminished in skin from Ifidancitinib-treated mice compared to controls ( Figures 2E, F ). In addition, Ifidancitinib significantly reduced the frequencies and total cell numbers of both CD8+NKG2D+ T cells and CD8+CD44+CD62L- effector memory T cells in skin-draining lymph nodes (SDLN) ( Figures 2G–I ). Taken together, these results indicate that JAK1/3 inhibition by Ifidancitinib prevented disease onset in the C3H/HeJ skin grafted mice.

To address whether Ifidancitinib effectively reverses disease in AA-affected mice, C3H/HeJ mice that developed AA 8-10 weeks after engraftment were treated with Ifidancitinib-incorporated chow diet (0.5 g/kg) (19). Following 12 weeks of treatment, robust hair regrowth was observed in Ifidancitinib-treated mice, compared to sustained and progressive hair loss in control mice ( Figures 3A, B ). Histological analyses on skin biopsies taken after treatment showed that Ifidancitinib-treated mice exhibited substantially reduced AA-associated inflammatory infiltrates and inflammatory markers (CD8, MHC class I and II) ( Figure 3C ). Additional analyses by flow cytometry of single cell suspension of skin samples confirmed that Ifidancitinib significantly reduced inflammatory infiltrates in the skin ( Figures 3D, E ). Role of tissue-resident memory T cells (T_RM) in pathogenesis of AA was implicated in patients with AA but remains poorly understood (21, 22). The percentage and total number of CD69⁺CD103⁺CD8⁺ T_RM were markedly reduced in the skin of Ifidancitinib-treated mice compared to skin of control mice. Flow cytometry showed that the proportions of IFN-γ producing CD8⁺ T cells in the skin were also markedly reduced in Ifidancitinib-treated mice ( Figures 3D, E ).

To gain further insight into the molecular responses to Ifidancitinib, we performed bulk RNA sequencing (RNA-seq) on the skin of treated mice. We found that Ifidancitinib induced dramatic changes in gene expression in treated mice at 10 weeks and shifted gene expression patterns towards the pattern of non-alopecic C3H/HeJ mice ( Figure S2A ). We showed that the sets of genes selected to comprise our IFN and CTL signatures-Cd8a, Gzmb, Icos and Prf1 for the CTL signature and Cxcl9, Cxcl10, Cxcl11, Mx1, and Stat11 for the IFN signature were significantly decreased in mice responsive to treatment with Ifidancitinib ( Figure S2B ) (7). In summary, these results indicate that the JAK1/3 inhibitor Ifidancitinib reduced the presence of inflammatory infiltrates and attenuated their function in skin, leading to effective reversal of established disease.

We next investigated mechanisms underlying the reduction of infiltrating T cells in the skin during Ifidancitinib treatment. We examined whether treatment affected the number, phenotype, or function of T cells or T cell subsets in peripheral lymphoid organs. We found that AA mice treated with Ifidancitinib had a significantly reduced absolute number of total CD3 T cells, CD4 and CD8 T cells in the SDLNs compared with controls ( Figure S3 ). CD8⁺NKG2D⁺ T cells and CD8⁺CD44⁺CD62L^- effector memory T cells in SDLNs, which are associated with AA pathogenesis, were diminished in terms of both absolute number and percentage of total T cells ( Figure S3 ). These results indicate that Ifidancitinib treatment resulted in a profound reversal of T cell activation and proliferation in peripheral lymphoid tissues.

In the setting of chronic viral infection, which results in the sustained disruption of the γc cytokine network, effector T cells might undergo exhaustion (23). Exhausted T cells display high levels of PD-1 and reduced production of effector cytokines, such as IFN-γ (23, 24). Previous studies showed that Jak3 knockout T cells have impaired γc cytokine signaling and show increased expression of PD-1 and LAG-3 (25). Due to its properties as a potent JAK1/3 inhibitor, we postulated that in addition to inhibiting the generation and proliferation of alopecic T cells, Ifidancitinib might also may also selectively induce effector T cell exhaustion in treated mice. Indeed, we observed an increase in the percentage of CD44⁺ effector T cells coexpressing high levels of PD-1 and TIM-3 in the SDLNs and spleen in Ifidancitinib-treated mice compared to control mice ( Figures 4A, B ). These PD-1-expressing T cells also expressed high levels of Eomes that is known to be highly expressed in exhausted T cells co-expressing inhibitory receptors ( Figures 4C, D ) (26). We next examined cytokine secretion by PD-1⁺TIM-3⁺CD8⁺ T cells in response to stimulation. PD-1⁺TIM-3⁺CD8⁺ T cells from Ifidancitinib-treated mice displayed decreased production of IFN-γ compared to the T cells from control mice ( Figures 4E, F ). Taken together, these findings are consistent with the functional exhaustion of effector CD8⁺ T cells associated with coexpression of PD-1 and TIM-3 and decreased IFN-γ production. Thus, the effect of Ifidancitinib on AA prevention and reversal may be due to the induction of T cell exhaustion, in addition to inhibitory effects on T cell proliferation, differentiation and survival.

To gain insight into the underlying mechanisms of the upregulated co-inhibitory receptor expression on T cells after JAK1/3 inhibition, we stimulated T cells in the presence of Ifidancitinib or vehicle control in vitro. T cells increased expression of PD-1 after stimulation. Ifidancitinib treatment increased the expression levels of PD-1 on both CD4 T and CD8 T cells compared to vehicle control ( Figures 5A, B and Figure S4A ). Similar to Ifidancitinib, STAT5-IN-1 (STAT5 inhibitor) treatment also increased PD-1 expression on both CD4 T and CD8 T cells ( Figures 5C, D and Figure S4B ). Similarly, JAK1 or JAK3 inhibition by JAK1-selective inhibitor Itacitinib, JAK3-selective inhibitor Ritlecitinib, JAK1/2-selective inhibitor Ruxolitinib, or pan-JAK inhibitor Tofacitinib but not JAK2-selective inhibitor Fedratinib (27), increased PD-1 expression on both CD4 T and CD8 T cells ( Figures 5E, F and Figure S4C ), indicating that JAK1/3-STAT5 pathway might regulate PD-1 expression on activated T cells (28).

We next evaluated the potential contribution of endogenous γ chain cytokine produced by activated T cells to regulate PD-1 expression on activated T cells. Since activated T cells are the major source of IL-2, we simulated T cells with anti-CD3 and anti-CD28 in the presence of anti-IL-2 neutralizing mAbs. We observed that IL-2 neutralization increased PD-1 expression compare to Abs isotype control ( Figures 5G, H and Figure S4D ), indicating that endogenous IL-2 was critical for PD-1 downregulation in vitro (28).

Transcription factors thymocyte selection-associated HMG box (TOX) and EOMES have been shown to promote T cell exhaustion in cancer and chronic viral infections (26, 28–30). We measured the expression levels of TOX and EOMES in T cells after individual JAK-selective inhibitor treatment. We observed that JAK1 or JAK3 inhibition by selective inhibitors robustly increased the frequency of Eomes⁺TOX⁺ in both CD4 T and CD8 T cells T cells compared to vehicle and JAK2 inhibitor treatment in vitro ( Figure 5I and Figure S4E ). Similar to JAK1 or JAK3 inhibition by selective inhibitors, IL-2 neutralization promoted the generation of Eomes⁺TOX⁺ T cells in vitro ( Figure 5J ).

To evaluate the potential contribution of γ chain cytokine in regulating PD-1 expression and T cell exhaustion in vivo, we treated C3H/HeJ AA mice with a combination of IL-2 neutralizing mAbs, IL-9 neutralizing mAbs and IL-15 neutralizing mAbs (anti-IL-2/9/15). We observed increased frequency of PD-1⁺TOX⁺CD44⁺ CD8 T cells with anti-IL-2/9/15 treated mice compared to isotype control treated mice ( Figures 5K–M ). Further, JAK3-selective inhibitor Ritlecitinib increased frequency of PD-1⁺TOX⁺CD44⁺ CD8 T cells ( Figures 5N–P ). These results indicated that γc cytokine regulates T cell survival and exhaustion, and pharmacological inhibition of JAK1 or 3 signaling or neutralizing γc cytokine promote effector T cell exhaustion (23).

To determine the effects of JAKi on induction of exhaustion phenotype in CD8+ T cells, we utilized scRNAseq to characterize T cell composition in AA skin after treatment with the pan-JAK inhibitor, Tofactinib (Tofa) ( Figure S5 ). We analyzed single cell suspensions of pre-sorted CD45+ cells isolated from whole skin from C3H/HeJ untreated mice or mice treated with Tofa. 19 clusters representing diverse immune cell types have been identified by scRNAseq ( Figure 6A , Supplementary Table 1 ) (31). We identified a significant reduction in the proportion of TRM cells and Eff_CD8T cells (cluster 0 and cluster 1, respectively) in the skin of Tofa-treated mice, and a significant expansion in the proportion of ex_CD8T cells characterized by the expression of Tox (cluster 9) ( Figures 6A–C ). We used RNA velocity trajectory analysis to predict the origin trajectory of the ex_CD8T cell cluster ( Figures 6D, E ) (32). We identified that similar to TRM cells, ex_CD8T cells in the skin originate from the Eff_CD8T cell cluster ( Figures 6D, E ).

C3H/HeJ, C57BL/6 (B6), CD45.1 (B6.SJL-Ptprca Pepcb/BoyJ) were purchased from The Jackson Laboratory and were maintained under specific pathogen-free conditions at the animal facility at Columbia University Medical Center (CUMC). All protocols were approved by the Institutional Animal Care and Use Committee of CUMC.

Surgical methods for skin grafts to induce AA in C3H/HeJ mice were described previously (7). For the skin graft model, around 10 mm diameter skin from C3H/HeJ AA mouse was grafted to 10-week-old C3H/HeJ recipient and covered with bandage for 10 days. Hair status was examined twice weekly.

Antibodies used in the experiments with mouse cells were obtained from Biolegend unless otherwise stated: anti-CD3 (17A2); anti-CD4 (GK1.5); anti-CD8 (53-6.7); anti-B220 (RA3-6B2); anti-CD25 (PC61.5); anti-CD44 (IM7); anti-CD45 (30-F11); anti-CD62L (MEL-14); anti-CD45.1(A20); anti-CD45.2 (104); anti-NKG2D (CX5); anti-PD-1 (29F.1A12); anti-TIM-3 (RMT3-23); anti-CD200 (OX-90); anti-EpCAM (G8.8); anti-MHC I (36-7-5); anti-MHC II (M5/114.15.2); anti-LAG-3 (C9B7W; Thermo Fisher Scientific); anti-pan NK cells (DX5; Thermo fisher); anti-Foxp3 (FJK-16s; Thermofisher); IL-17 (TC11-18H10.1); anti-TNF-α (TN3-19.12); anti-IFN-γ (XMG1.2); anti-Eomes (Dan11mag; Thermo fisher); anti-TOX (REA473; Miltenyi Biotec), anti-pStat1 (Tyr701) (58D6; Cell Signaling); anti-pStat5 (Tyr694) (C71E5; Cell Signaling); anti-Actin (SCBT). The following mAbs raised against human antigens were all purchased from Biolegend: CD3 (OKT3); CD4 (OKT4); CD8 (SK1); NKG2D (1D11); IL-17 (BL168); IFN-γ (B27). Viable cell populations were gated based on forward and side scatters and by fixable blue (Thermo Fisher Scientific) staining. Samples were collected on an LSR II and analyzed with FlowJo software (Tree Star).

Recombinant murine and human IL-2, IL-15 and IFN-γ were from Peprotech. Ifidancitinib chow diet and Ifidancitinib were provided by Rigel Pharmaceuticals and Aclaris Therapeutics. Itacitinib (catalog HY-16997, MedChemExpress), Fedratinib (catalog 202893, Medkoo), Ritlecitinib (catalog PZ0316, MilliporeSigma), Ruxolitinib (catalog S1378, Selleck), Tofacitinib (catalog 200811, Medkoo)

Newly grafted C3H/HeJ mice or long standing C3H/HeJ AA mice were fed with Ifidancitinib incorporated chow diet (0.5 g/kg) for indicated time. Mice were scored weekly for signs of hair regrowth and loss. Mice were euthanized and organs were collected for analysis after treatment.

Ritlecitinib was delivered through an ALZET osmotic pump (MODEL 1002, DURECT). For topical treatment, C3H/HeJ AA mice were topically treated with 2% (w/w) Tofacitinib in Aquaphor (Aquaphor) twice daily. For antibody treatment, newly onset C3H/HeJ mice with AA were administered through IP injection with 0.1 mg combination of IL-2 neutralizing mAbs (JES6-1A12; Bioxcell), IL-9 neutralizing mAbs (9C1; Bioxcell) IL-15 neutralizing mAbs (AIO.3; Bioxcell) or Rat IgG2a (Bioxcell) twice weekly for 8 weeks.

Skin was finely minced and digested for 45 min at 37°C with 2 mg/ml collagenase type IV (Worthington) and 0.05 mg/ml DNase I (Sigma-Aldrich) in RPMI 1640. Digested skins were filtered through a 70 μm cell strainer. SDLNs or spleens were homogenized and filtered through a 70 μm cell strainer. Splenocytes were depleted of erythrocytes by RBC Lysis Buffer (Thermo Fisher Scientific). Cells were either re-suspended in FACS buffer (PBS with 2% FCS) for flow cytometric staining, stimulated with PMA and ionomycin for cytokine production or single-cell library construction.

For NKG2D⁺CD8⁺ T cell differentiation, purified C3H/HeJ naïve CD8⁺ T cells were stimulated with IL-15 (20 ng/ml) in the presence of Ifidancitinib or vehicle control (DMSO) for 4 d. For in vitro proliferation, CellTrace Violet-labeled T cells were stimulated with plate-bound anti-CD3 Ab (0.5 μg/ml) and IL-2 (5 ng/ml), and cell proliferation was assessed by dye dilution. For PD-1 induction, purified C3H/HeJ naïve CD4⁺ T or CD8⁺ T cells were stimulated with 500 ng/ml anti-CD3 in the presence of indicated regents in vitro for 4 d described in Figure 5 .

Purified T cells from CD45.1-congenic B6 mice were stimulated with plate-bound anti-CD3 Ab (1 μg/ml) and IL-2 (5 ng/ml) in vitro for 3d. CellTrace Violet-labeled T cells were adoptively transferred i.v. to B6 (CD45.2) mice. After 12 h, mice were treated with various dosages of Ifidancitinib and 20 μg rhIL-2 as described previously. On day 3, mice were sacrificed and the proliferation of donor T cells from their spleens and SDLNs were analyzed by flow cytometry.

SDLN or skin single cell suspensions were cultured with 1× Cell Stimulation Cocktail (Thermo Fisher Scientific). After 1 h, 1× Protein Transport Inhibitors) (500X) (Thermo Fisher Scientific) was added, followed by additional 4 h incubation at 37° C. The cells were then fixed, and permeabilized using FoxP3 fixation/permeabilization kit (Thermo Fisher Scientific) and stained intracellularly with anti-IFN-γ, IL-4, IL-17 or TNF-α for 60 min at 4° C.

Purified human CD8⁺ T cells were stimulated in vitro with IL-15 (20 ng/ml) in the presence of Ifidancitinib or vehicle control (DMSO) for 4 d. The frequency of NKG2D⁺CD8⁺ T cell was assessed by flow cytometric analysis. For cell proliferation, CellTrace Violet-labeled T cells were stimulated with plate-bound anti-CD3 Ab (1 μg/ml) and IL-2 (10 ng/ml), and cell proliferation was assessed by dye dilution.

Immunofluorescence staining of frozen skin sections was performed as previously described., The individual primary Ab was used at optimal concentrations for detection (final concentration 10ug/ml), followed fluorochrome-conjugated goat anti-rat secondary Abs (Thermo Fisher Scientific). The immunofluorescence staining of mouse MHC class I was performed using biotin labeled mouse anti-H-2Kk, followed by fluorochrome-conjugated streptavidin (Thermo Fisher Scientific). Murine T cell blasts were pretreated in the absence or presence of Ifidancitinib (0.1, 0.3 and 1 μM) for 1 h and then stimulated with IL-2 (20 ug/ml) or IFN-γ (40 ug/ml) for 15 min. Total cell lysate (40 ug) was separated by electrophoresis and probed with Abs as described previously.

Total cellular RNA was extracted using RNeasy Kit (Qiagen) from skin homogenates. RNA quality and quantity were determined using an Agilent BioAnalyzer (Agilent Technologies). RNA isolation, library construction and sequencing were performed at GENEWIZ. Gene-expression analysis further was performed with the Seurat R package (27).

To prepare mouse skin single-cell suspensions of immune cells, live CD45+ cells were FACS sorted. The samples were submitted to the JP Sulzberger Columbia Genome Center’s Single Cell Analysis Core. There, the Chromium Next GEM Single Cell 5’ Reagent Kit V2 was used to prepare cell libraries according to the manufacturer’s instructions, with a target of 5000 cells and 350 million reads. Libraries were sequenced on an Illumina NextSeq 500/550.

After sequencing, FASTQ files were processed using the CellRanger v6.1.2 and aligned to the mm10-2020-A reference transcriptome. The output CellRanger files of the individual samples were merged into a single Seurat object and analyzed using Seurat package v4.0.6 in R v4.1.0 (31). For quality control, we removed cells with lower than 500 molecules detected within a cell and cells with more than 5% of UMIs were derived from mitochondrial genes, and included cells which had more than 200 genes but less than 3000 genes. Post quality control, 7127 cells for untreated skin and 4230 cells for Tofa-treated skin remained for downstream bioinformatics analyses. We then used Seurat package to integrate the individual samples, perform a global-scaling normalization and linear transformation using ‘NormalizeData’, and ‘ScaleData’ functions, respectively. We next used the ‘Elbowplot’ function, to determine the dimensionality of the dataset, and subsequently used 30 PCAs to cluster the cells using the Louvain algorithm. We then performed uniform manifold approximation and projection (UMAP) dimensionality reduction to explore and visualize the dataset. We used ‘FindAllMarkers’ function in Seurat to find differentially expressed markers for each cluster and used canonical markers to match the unbiased clustering to known cell types as outlined in Supplementary Table 1 . We used ‘VlnPlot’ and ‘FeaturePlot’ functions in Seurat to show expression probability distributions across clusters and visualize feature expression on a PCA plot.

To perform RNA velocity analysis in our scRNASeq dataset, we used the Python v.3.11 based package scVelo, which uses normalized Seurat object in conjunction with RNA Velocity analysis (32). Briefly, we used Samtools v1.10 and Velocyto to generate Loom files for the individual single cell samples we analyzed using Seurat. Then, we integrated the individual loom files and the Seurat meta-data using ‘anndata’ function. Then, we merged the individual samples into one anndata object and added UMAP coordinates from the merged Seurat object to match the order of the Cell IDs in the anndata object. At last, we used scVelo to generate RNA Velocity plot based on Seurat UMAP coordinates and the ‘subset’ function to generate RNA Velocity plot for the CD8 T cell subsets (cluster 0, 1, and 9).

Statistical analyses were performed using the GraphPad Prism 7.0 software. Data were expressed as mean plus or minus SEM for each group and results were compared with 2-tailed t tests, log-rank test, or one-way ANOVA. n.s.> 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.