Inhibition of JAK-STAT pathway corrects salivary gland inflammation and interferon driven immune activation in Sjögren’s Disease

Objectives: Inflammatory cytokines that signal through the JAK- STAT pathway, especially interferons (IFNs), are implicated in Sjögren’s Disease (SjD). Although inhibition of JAKs is effective in other autoimmune diseases, a systematic investigation of IFN-JAK-STAT signaling and effect of JAK inhibitor (JAKi) therapy in SjD-affected human tissues has not been reported. Methods: Human minor salivary glands (MSGs) and peripheral blood mononuclear cells (PBMCs) were investigated using bulk or single cell (sc) RNA sequencing (RNAseq), immunofluorescence microscopy (IF), and flow cytometry. Ex vivo culture assays on PBMCs and primary salivary gland epithelial cell (pSGEC) lines were performed to model changes in target tissues before and after JAKi. Results: RNAseq and IF showed activated JAK-STAT pathway in SjD MSGs. Elevated IFN-stimulated gene (ISGs) expression associated with clinical variables (e.g., focus scores, anti-SSA positivity). scRNAseq of MSGs exhibited cell-type specific upregulation of JAK-STAT and ISGs; PBMCs showed similar trends, including markedly upregulated ISGs in monocytes. Ex vivo studies showed elevated basal pSTAT levels in SjD MSGs and PBMCs that were corrected with JAKi. SjD-derived pSGECs exhibited higher basal ISG expressions and exaggerated responses to IFNβ, which were normalized by JAKi without cytotoxicity. Conclusions: SjD patients’ tissues exhibit increased expression of ISGs and activation of the JAK-STAT pathway in a cell type-dependent manner. JAKi normalizes this aberrant signaling at the tissue level and in PBMCs, suggesting a putative viable therapy for SjD, targeting both glandular and extraglandular symptoms. Predicated on these data, a Phase Ib/IIa randomized controlled trial to treat SjD with tofacitinib was initiated.


Single Cell RNA Sequencing of the MSG and PBMCs
Tissue dissociation.Seven SjD and five non-SjD subjects provided MSG biopsies and samples were processed for scRNAseq as previously described (online supplemental table 1). 4 5 After excision, MSGs (2-3/per patient) were placed immediately in ice-cold RPMI.MSGs were placed in a sterile 100 mm tissue culture dish and delicately dissected into uniform ~1 mm lobules.Lobules were dissociated using the Miltenyi Multi-tissue Dissociation Kit A using the Multi_A01 in C-type tubes at 37 °C in an OctoMACS tissue disruptor using heated sleeves.Single-cell suspensions were filtered through 70-and 30-μm filters and rinsed with 1× Hanks' buffered salt supplemented with 1% ultrapure, DNase/RNase-free, bovine serum albumin solution.Cells were centrifuged at 300g for 10 min at 4 °C and washed twice with 1× Hanks' buffered salt solution.Cell counting and viability were determined using a Trypan blue exclusion assay.Suspensions with greater than 75% viability were used for subsequent sequencing.In all instances, adequate numbers of glands were submitted for histopathological assessment and standardized focus scoring as described above.
Single-cell capture, library preparation and sequencing.Single-cell suspensions targeting approximately 10,000 cells were prepared as described above and loaded onto a 10x Genomics Chromium Next GEM Chip B (10x Genomics) following the manufacturer's recommendations.
After cell capture, single-cell library preparation was performed following the instructions for the 10x Chromium Next GEM Single Cell 3' kit v3 (10x Genomics).The libraries were pooled and sequenced on four lanes of a NextSeq500 sequencer (Illumina), adopting the read configuration indicated by the manufacturer.
scRNAseq data processing, quality control, and analysis.
Read processing was performed using the 10x Genomics workflow (10x Genomics).Briefly, the Cell Ranger v3.0.1 Single-Cell Software Suite was used for demultiplexing, barcode assignment and UMI quantification (http://software.10xgenomics.com/single-cell/overview/welcome).Sequencing reads were aligned to the hg38 reference genome (Genome Reference Consortium Human Build 38) using a pre-built annotation package obtained from the 10x Genomics website (https:// www.10xgenomics.com/).Samples were demultiplexed using the 'cell ranger mkfastq' function, and gene count matrices were generated using the 'cellranger coun' function.

Single-cell RNA sequencing data was analyzed in Python using Scanpy.
Cells containing less than 100 genes and genes expressing in less than 10 cells were filtered out.Raw data was normalized as count per ten thousand and then logarithmized.Individual sample libraries were combined and processed as a single library using the Batch balanced kNN procedure. 6Cell clustering was performed by the Leiden graph-clustering method 7 and displayed in Uniform Manifold Approximation and Projection (UMAP) format.Type I IFN score was calculated based on the average expression of 21 Type I IFN stimulated genes (ISGs) using the "scanpy.tl.score_genes" command. 8Differentially expressed genes (DEGs) were identified using the Wilcoxon model in the "scanpy.tl.rank_genes_groups" function after excluding ribosomal and mitochondrial genes.Enriched pathway analysis in DEGs was performed using DAVID Bioinformatics Resources (https://david.ncifcrf.gov).
Cells containing more than 200 and fewer than 2,500 unique features were retained.From this set, cells with greater than 15% of read counts attributed to mitochondrial DNA were filtered out.We adjusted this value from 5% to 15% to increase the yield from each sample and did not observe substantive changes in our results after adjustment.

Detection of JAK1 and JAK3 in FFPE biopsies of minor salivary glands.
Sections (5 µm) from FFPE MSG were baked for 1hrs at 65°C, deparaffinized and rehydrated following standard procedures.Antigen retrieval was performed using R-UNIVERSAL Epitope Recovery Buffer (1x) (Electron Microscopy Sciences) in microwave pressure cooker using 80% power for 15 min.Sections were blocked for 30 min RT using 10% FBS buffer in PBS plus saponin 0.01%, and incubated in diluted primary antibodies overnight at 4 °C.Primary antibodies were washed in PBS for 10 min, 3 times, and a mix of secondary antibodies were applied in blocking buffer (online supplemental table 2).After 45 minutes of incubation, secondaries antibodies were rinsed, and tissues were mounted using mounting solution.Whole slides were digitally scanned using the 40X objective using an Axioscan Z1 (Zeiss), with the following exposure times (DAPI, 20ms; JAK1, 200ms; JAK3, 150ms; and Cytokeratin-18, 50ms).
Whole slide images were uploaded into Visiopharm Image Analysis software V2022.11(Visiopharm A/S).Digitized tissue sections were identified, cells were detected and segmented using deep-learning-based nuclear segmentation, cellular phenotyping was assessed using the PhenoApp ® module using thresholds based on visual assessment of positivity.PhenoApp ® was trained on all the analyzed samples.After establishment of cellular phenotypes, the median fluorescence intensity per cell was normalized based on the background autofluorescence per slide and the max and min intensity in the set.Dimensionality reduction using t-Stochastic Nearest-neighbor Embedding (tSNE) was used to cluster segmented cells and visualize individual cellular phenotypes based on protein expression of JAK1, JAK3, and Cytokeratin-18.

Immunofluorescence in primary cell culture assays.
Primary salivary gland epithelial cells (pSGECs) were generated according to methods published by Jang et al., 9 from patients fulfilling ACR 2016 classification criteria or HV. 10 pSGECs were plated on chamber slides and stimulated as indicated.The cells were fixed for 2h at 4°C and permeabilized for 10 min at −20°C.Cells were stained with primary antibodies overnight and then stained by secondary antibody for 2h (online supplemental table 2).Images were acquired on Nicon A1 HD (Nicon) confocal microscope and processed with CellProfiler in ImageJ (Broad Institute). 11

Assessment of serum proteome
Proteomic profiles were measured in serum (50uL) using the SOMAscan Assay V1.3 (SomaLogic, Inc.) at the Trans-NIH Center for Human Immunology, Autoimmunity, and Inflammation (CHI), National Institutes of Health (Bethesda, MD, USA) as previously reported. 12Following the manufacturer's data processing guidelines, the raw data were normalized by control hybridization, then median signal normalization and finally, inter-plate calibration normalization.Once normalized, the data levels of protein were compared.The type I IFN protein (IFNP) scores were calculated based on Smith et al. 13 Prior to calculation of the IFNP, the four proteins characterized in the IFNP were log 2 transformed and scaled to the mean and standard deviation of the respective sample distribution.Data processing was performed in -R 3.1.1.

Flow cytometry
Minor Salivary Gland Flow Cytometry.Freshly biopsied minor salivary glands were dissociated and enumerated as described above.Cells were fixed using BD Cytofix Fixation Buffer (BD) containing 4.2% formaldehyde, washed with staining buffer, then permeabilized with BD phosflow Perm Buffer Ⅲ (BD).Multicolor flow cytometry was used to quantify the phosphorylation status of each pSTAT in gated cell subset populations (online supplemental table 2 and online supplemental method 1).Cells were acquired using a FACS Symphony (BD Biosciences) flow cytometer and analyzed with FlowJo ™ v10.8 (BD Life Sciences).
Peripheral blood mononuclear cells.Flow cytometry analysis was performed using cryopreserved peripheral blood mononuclear cells (PBMCs) isolated by BD Vacutainer® CPT™ Cell Preparation Tube (BD Biosciences) for basal or experimental analyses.Thawed PBMCs were stimulated as indicated in 500 μl media in 48-well flat-bottom plates at 37°C, 5% CO 2 and 95% humidity according to experimental conditions.Cells were treated as described above and analyzed (online supplemental table 2 and online supplemental method 2).
Multicolor flow cytometry was used to quantify the phosphorylation status of pSTATs in gated cell subset populations (supplemental table 2 and supplemental method 1,2).Cells were acquired using a FACS Symphony (BD Biosciences) flow cytometer and analyzed with FlowJo™ v10.8 (BD Life Sciences).

RNA isolation and RT-qPCR
RNA was isolated using Monarch Total RNA miniprep Kit (New England BioLabs).Standard Taqman assays (supplemental table 3) were performed in technical triplicates with at least biological duplicates to measure relative gene expression using the delta-delta Ct method 14 on a QuantStudio™ 6 Pro Real-Time PCR System (Thermo Fisher).

Western Immunoblot Analysis
Extracted total protein was resolved using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (4% stacking, 12% resolving, Invitrogen) and transferred onto Polyvinylidene fluoride (PVDF) membranes.Then, membranes were incubated with primary and secondary antibodies (online supplemental Table 2).The signal was detected by ChemiDoc MP Imaging System (BIO-RAD), and the density of the bands was analyzed using Fiji. 15

LDH assay and AnnexinV staining
The potential for cytotoxicity or apoptosis induced by tofacitinib was measured using the plate-based CyQUANT™ LDH Cytotoxicity Assay kit (Invitrogen) and the flow-cytometry-based Dead Cell Apoptosis Kit with Annexin V Alexa Fluor 488 & Propidium Iodide (Invitrogen) following the manufacturer's protocol on a FACS Symphony flow cytometer.

Statistical Analysis for Difference in Immune Cell Populations/Cytokines
Statistical methods were employed using GraphPad Prism (GraphPad), matlab, or -R as described, and the type and nature of the data were considered when assessing differences in mean values and variances across biological and experimental replicates.Generally, a p-value of <0.05 was considered statistically significant, unless otherwise noted (e.g., where adjusted p-values to compensate for multiple comparisons).Specific statistical methods are reported in specific methodological sections and/or figure legends.For the analysis of pseudobulk RNAseq from scRNAseq data, the voom pipeline from the limma package (Bioconductor) was used determine DEG from the pseudo-bulk scRNAseq expression profile of individual clusters with cut-offs of adjusted p-value of <0.05 and fold-change >2-forld or <0.5-fold expressed.

Supplemental Figure 1 :Supplemental Figure 3 :Supplemental Figure 4 :Supplemental Figure 5 :Supplemental Figure 6 :PBMCSupplemental Figure 7 :
of Z-scores Type I Sum of Z-scores R 2 = 0.21; p=0.0064Bulk sequencing of minor salivary gland and IFN signature (A) Differential expression of IFN score in SjD and HV.Kruskal-Walls test.(B) Gene graph enrichment analysis showed consistent and direct regulation of the JAK-STAT pathway signaling through IL7/IL15/IL21 via JAK3 (solid red connections).(C) Differential expression of JAK-STAT related genes in SjD and HV.(D) Differential expression of IFNG showing increased IFNG in SjD (~3-fold, p=0.0258) and the correlation between Type I and II sum of Z-scores.P value was calculated using Mann-Whitney test and Spearman correlationmicroscopy (A, B) Unbiased segmentation and quantitation of per cell expression in SjD and non-SjD.(C) Cell subset population difference in SjD and non-SjD MSG from IF.The cellular proportion was changed to less epithelial and more immune cells in the SjD glands.(D) tSNE figures in SjD and non-SjD MSG from IF. JAK1 and JAK3 expressions were clearly elevated in immune cells in SjD, but both expression change in the SjD epithelial cells were mild.Single cell RNAseq of MSG (A-D) Differential gene expressions in SjD and non-SjD MSG.JAK1 was the most ubiquitously expressed in MSGs with appreciably less expression of JAK2, JAK3, and TYK2.Seromucous acinar cells showed increased expression of all JAKs, while ductal cells had increased expression of JAK3 and TYK2; APCs and plasma cells exhibited increased expression of JAK1, JAK3, TYK2.STATs gene expression showed increased expression of STAT1 in all cell clusters in MSGs from patients with SjD, while others were expressed in only specific cell clusters.Single cell RNAseq of PBMC (A-C) Differential gene expressions in SjD and non-SjD MSG.DEG analysis showed upregulation of many ISGs (e.g., IFI44L, IFIT3, ISG15, MX1, and IFI6) in SjD PBMCs.Basal pSTATs frequencies in PBMCs(A-D) Basal pSTAT frequencies in SjD and HV PBMCs.The frequency of pSTAT1, pSTAT3(Ser727), and pSTAT6 were higher in SjD patients compared to HV, but not for pSTAT3(Tyr705).* p<0.05, **p<0.01,P value was calculated using Welch's test.Treatment effects of tofacitinib in PBMCs (A, C) Effects of tofacitinib on with or without IFN-β induced pSTATs in PBMCs.Tofacitinib treatment significantly downregulated pSTAT levels in SjD.This trend was not seen in HV which showed lower levels even without tofacitinib treatment.P value was calculated using Mann-Whitney test.(B) Tofacitinib abolished the IFNβ-induced IFN score to baseline level on other cell subsets.(D) 5mM of tofacitinib did not induce necrosis in PBMC under these experimental conditions.Treatment effects of tofacitinib in pSGECs (A, B) 5µM of tofacitinib did not induce apoptosis and necrosis in pSGEC under these experimental conditions.(C, D) Effects of tofacitinib on with or without IFN-β induced pSTATs in pSGEC.

Table 1 :
Cohort descriptionDescriptions of the cohorts used for each analysis.For race, values are numbers of white/black/Asian or other study subjects included in each analysis.Significant differences in age in the cohort are in comparison to the HV or SjD cohort.* p<0.05,P value was calculated using Mann-Whitney test.UK: unknown.