Study populations

Peripheral blood was obtained from 97 patients with RA, 15 patients with PsA, 6 patients with gout, 7 patients with SLE and 119 HCs. RA patients fulfilled the ACR diagnostic criteria for RA and were positive for rheumatoid factor and/or anti-citrullinated protein Abs. Demographic and clinical data are given in Supplementary Table 1. RA and PsA patients were predominantly male reflecting the gender distribution at the Palo Alto Veterans Administration Rheumatology Clinic. Healthy individuals did not have a personal or family history of autoimmune diseases and no history of cancer. Chronic diseases such as hypertension or diabetes mellitus were included as long as controlled on oral medication. The protocol was approved by the Stanford University Institutional Review Board, and all participants gave written, informed consent.

Cell purification

PBMCs were purified from venous blood by density gradient centrifugation with Ficoll (STEMCELL Technologies, #07851). CD4⁺ naive T cells were purified from PBMC by a two-step selection procedure; first, CD4⁺ T cells were isolated with Human CD4⁺ T Cell Enrichment Cocktail (STEMCELL Technologies, #15062), then CD45RO⁺ cells were depleted to enrich for naive CD4⁺ T cells (STEMCELL Technologies, #19155). Subset purity of CD4⁺CD45RO^–CD45RA⁺ T cells monitored by fluorescence-activated cell sorting routinely exceeded 95%.

Reagents and antibodies

AA (ab120916), thapsigargin (ab120286), ORAI3 Ab (ab115558), and NUR77 Ab (ab109180) were purchased from Abcam. Abs recognizing p-CAMKII (Thr286, #12716), p-ERK (Thr202/Tyr204, #4377s), CAMKII (#4436), ERK (#4695), Myb (#12319), AP-2a (#3208), and IKAROS (#9034s) were from Cell Signaling. Abs to β-actin (#A5441) were obtained from Sigma, to p-CD3ζ (Y142) from BD Biosciences (558402) and to CD3ζ from BioLegend (644102). Abs for flow cytometry were purchased from BD Biosciences [CD4-V450 (560345), CD8-PE-Cy7 (557746), CD45RA-AF700 (560673), CD69-Percp-Cy5.5 (560738), p-ERK (Thr202/Try204)-AF670 (561992), and p-SLP76 (Y128)-AF647 (558438)] or BioLegend [CD62L-PE (304806), IKAROS-AF647 (368404), and CD3-AF488 (30045)]. The LIVE/DEAD^TM Fixable Aqua Dead Cell Stain Kit from Molecular Probes, Thermo Fisher Scientific was used to gate out dead cells; Fluo-8 (21080) and Fura Red^TM (F3021) for staining cytoplasmic Ca²⁺ was from AAT Bioquest, Inc., and Thermo Fisher Scientific, respectively.

Flow cytometry

Freshly purified PBMCs from RA patients and HCs were stained with Abs to CD3, CD4, CD45RA, and CD62L on ice in phosphate-buffered saline (PBS) containing 1% bovine serum albumin (BSA). Patient and HC samples were always run in parallel to reduce batch effects in fluorescence intensities. Cells were washed once with 1% BSA PBS and fixed with pre-warmed BD Cytofix buffer for 10 min (BD Biosciences, #554655). Cells were pelleted and resuspended in 1 ml of BD Phosflow Perm buffer (#558050) for 30 min on ice. After washing twice with 1% BSA PBS, cells were incubated with Abs to p-SLP76 (Y128) or p-ERK (Thr202/Tyr204) at room temperature for 30 min in the dark. Fluorescence data were obtained on an LSRII or a Fortessa Cell Analyzer (BD Biosciences).

Cytometric measurements of cytoplasmic Ca²⁺ concentrations

A total of 2 × 10⁶ purified CD4⁺ naive T cells from RA patients or HCs were resuspended in 37 °C Hank’s balanced salt solution (HBSS) with 1 µM Fura Red^TM and 0.01% Pluronic F-127 and incubated at 37 °C for 30 min. Cells were washed with HBSS buffer once, resuspended in 1 ml HBSS, and allowed to equilibrate for 10 min in 37 °C. Fura Red fluorescence was measured by excitation with the Violet laser (406 nm) and the Green laser (532 nm). The fluorometric ratio (406 nm/532 nm) was calculated as the signal increase stimulated by the Violet laser over the signal decrease induced by the Green laser using the kinetics tool in FlowJo software version 9.3.3⁶³.

In the Fluo-8 cytometric experiments, PBMCs from RA patients and HCs were stained with Abs to CD3, CD4, CD45RA, and CD62L on ice in PBS containing 1% BSA. Cells were loaded with 2 µM Fluo-8 at room temperature (RT) for 20 min in HBSS buffer (Gibco, Thermo Fisher Scientific) containing both Ca²⁺ and Mg²⁺. Fluo-8 MFIs were recorded on an LSRII or a Fortessa Cell Analyzer (BD Biosciences) for indicated times; 0.1 µM AA was added at 30 s into the recording.

Functional read-outs of AA stimulation

A total of 5 × 10⁶ naive CD4⁺ T cells were stimulated with 0.3 µM AA; cells were assayed for CaMKII (Thr286) phosphorylation at 0, 0.5, 1, 2, and 5 min. Culture in 10 µM Ca²⁺ containing medium served as a positive control. To assess the effects of AA on T cell activation, 5 × 10⁶ naive CD4⁺ T cells were treated with 0.3 µM AA for 3 min. Cells were assayed for CD3ζ and ERK phosphorylation or NUR77 expression by Western blotting at indicated times after AA stimulation. Naive CD4⁺ T cells stimulated on 100 ng CD3/CD28 Ab-coated plates served as a positive control. Alternatively, PBMCs (1 × 10⁶/ml) from HC or RA patients or Jurkat cells were stimulated with 0.3 µM AA for 3 min and analyzed by flow cytometry for CD69 expression after 24 h culture in medium without additional stimulation.

AA measurements by enzyme-linked immunosorbent assay

Plasma AA from RA or PsA patients or HC was quantified using the Human AA Competitive ELISA (enzyme-linked immunosorbent assay) Kit following the manufacturer’s instructions (MyBioSources, #MBS703581). All samples were run in duplicates. Briefly, 50 µl of 1:50 diluted plasma was added to anti-AA Ab-coated wells, followed by the addition of 50 µl horseradish peroxidase enzyme AA conjugate. Plates were incubated for 40 min at 37 °C, washed five times and then incubated with 90 µl 3,3′,5,5′-tetramethylbenzidine substrate per well for an additional 20 min at 37 °C. Reactions were stopped with 50 µl stop solution, and plates were read at 450 nm within 5 min. Concentrations were calculated with the CurveExpert 1.3 professional software.

ATAC-seq

CD4 T cells were isolated using Human T cell Enrichment Cocktail (STEMCELL Technologies, Canada), followed by purification of naive (CD3⁺CD4⁺CD62L⁺CD45RA⁺CD28⁺), CM (CD3⁺CD4⁺CD62L⁺CD45RA⁻CD28⁺), and effector memory (CD3⁺CD4⁺CD62L⁻CD45RA⁻CD28⁺) subsets by fluorescence-activated cell sorting. ATAC-seq libraries were generated using 50,000 cells for each subset. Cells were first washed with cold PBS and RSB buffer (10 mM Tris-HCl (pH 7.4), 10 mM NaCl, 3 mM MgCl₂), followed by washing with RSB buffer containing 0.1% NP-40 and 0.1% Tween-20. Subsequently, cell pellets were resuspended in a transposase reaction mix (25 μl 2× TD buffer, 2.5 μl transposase (Illumina), and 22.5 μl nuclease-free water) and incubated at 37 °C for 30 min. DNA fragments were purified using Qiagen MiniElute Kit and the library was amplified with Nextera PCR primers. Library quality was checked on bioanalyzer and sequenced on Illumina NextSeq 500. The sequencing reads were processed by trimming adapters and low-quality reads using in-house scripts and aligned to human reference genome hg19. Peaks were identified for each sample using macs2 and a consensus peak set was determined consisting of peaks present in at least three samples. Reads were converted into bigwig format for visualization of the genomic tracks, which were normalized to total reads mapped within the consensus peak set⁶⁴.

RNA-seq

Naive CD4⁺ T cells were isolated from PBMC using EasySep human Naive CD4⁺ T cell Enrichment Kit (STEMCELL Technologies, #19555). Two million cells were transfected with IKAROS siRNA smart pool (Dharmacon: L-019092-02-0005) or control siRNA (Dharmacon: D-001810-10-05) using P3 Primary cell 4D-Nucleofector Kit and the Lonza 4D-Nucleofector System (Lonza). Transfected cells were cultured for 48 h. Libraries were generated as recently described⁶⁵ and sequenced using the Novogene Hiseq platform (Novogene). The fastq files generated from the sequencing runs were analyzed using the nf-core pipeline⁶⁶ to determine read counts mapped to genes in GRCh37 genome. The data were further analyzed using Bioconductor packages edgeR and conditional quantile normalization. The experimental design was modeled upon donor and condition (control/knockdown) (~donor + condition). The downstream analysis to identify differentially expressed genes was performed as described in Chen et al.⁶⁷ with the addition of offsets from conditional quantile normalization⁶⁸, followed by the application of gene-wise negative binomial generalized linear models⁶⁹.

Reporter gene assays

The region upstream of the TSS of ORAI3 (chr16:30,960,352–30,961,215) was amplified from T cell-derived genomic DNA using the primer pair 5ʹ-GGGGGTACCCCAAGTTGTTTTATATTTCCATG-3ʹ and 5ʹ-GGGCTCGAGTTAGCAAGCTCCATGAAGTCAAG-3ʹ. PCR products were purified and cloned into the PGL3-basic plasmid (Promega). Sequences of insertions were authenticated. Thirty nanograms of PGL3-basic control, ORAI3-TSS/PGL3 or ORAI3-TSS/PGL3 plus 0.5 μg pcDNA-IKFZ1 (Genescript: clone Hu28071D) together with 1 ng Renilla luciferase reporter pR-TK plasmids were cotransfected into HEK293 using Fugene HD transfection reagent (Promega, E2311). Cells were cultured for 48 h before being analyzed using the dual-luciferase reporter assay system (Promega, E1910).

ChIP assay

ChIP assays were performed on five million CD4⁺ T cells by using the ChIP-IT Kit (53040) from Active Motif. Oligonucleotide primers were designed to amplify the ORAI3 promoter sequence (chr16/30,961,071–137; forward: 5ʹ-TTGCTGTTATTCTGTGGTGAG-3ʹ and reverse: 5ʹ-CAAAATAAGGGATCCATCAGA-3ʹ). Normal IgG was used as control.

Lentiviral transduction

GFP-expressing lentiviral vectors expressing ORAI3-targeting shRNA (pGFP-lenti-Sh-ORAI3; OriGene: TL307587) and full-length ORAI3-expressing pCMV6-AC lentiviral vectors (OriGene: RC202325L1 NM_152288), respectively were packaged into HEK293T cells using the lenti-vpak lentiviral packaging kit (OriGene TR30037). Scrambled control shRNA and pCMV6-AC empty vector were used as negative controls. Jurkat cell lines stably expressing sh-ORAI3 or ORAI3 were established as per the manufacturer’s instructions (OriGene).

Transfection with siRNA

Two million CD4⁺ naive T cells were transfected with control siRNA, ORAI3 siRNA smart pool (Dharmacon: L-015896-00), or IKAROS siRNA smart pool (Dharmacon: L-019092-02) using the P3 Primary cell 4D-Nucleofector Kit and the Lonza 4D-Nucleofector System (Lonza). Transfected cells were cultured in RPMI-1640 media supplemented with 10% FBS for 48 h before use. For screening TFs regulating ORAI3 expression, control siRNA, TFAP2a siRNA (A-006348-14), IKAROS siRNA (L-019092-02), or MYB siRNA (M-003910-00) were transfected into HEK293T cells with Fugene HD transfection reagent (Promega, E2311).

qPCR

Total RNA was extracted with the RNeasy plus Mini Kit (Qiagen), cDNA was synthesized with the High Capacity RNA-to-cDNA Kit (Applied Biosystems, 4387406). qPCR was performed in duplicates in 384-well plates using the ABI 7900HT System with Taqman Universal Master Mix II (Thermo Fisher, 4440040) and the following Taqman gene expression primers: ORAI3 (hS00743683-s1), 18S rRNA (Hs99999901_s1), IKAROS (HS00958474_m1 IKZF1), and β–actin (HS99999903_m1). For all other transcript analysis, qPCR was carried out using PowerUp™ SYBR™ Green Master Mix (Thermo Fisher, A25776) with primers listed in Supplementary Table 2.

Immunoblotting

CD4⁺ naive T cells (2 × 10⁶) were lysed on ice with RIPA buffer (Thermo Scientific) supplemented with phosphatase (sodium orthovanadate) and proteinase inhibitor cocktails (Roche, 33576300) and phenylmethylsulfonyl fluoride (Santa Cruz Biotechnology). Total protein (10 µg) was separated on 4–15% gradient SDS gels and then transferred to polyvinylidene difluoride membranes. ORAI3, NUR77, and IKAROS were detected with the respective Abs. Membranes were stripped with stripping buffer (Invitrogen) and re-probed with anti-β-actin Abs. For TCR signaling experiments, CD4⁺ naive T cells or Jurkat cells stably transduced with ORAI3 shRNA, control shRNA, or ORAI3 were stimulated with AA or on immobilized 100 ng anti-CD3/CD28 Abs, lysed at indicated time points, and probed for p-CD3ζ (Y142), CD3, p-CAMKII, CAMKII, p-ERK (Thr202/Tyr204), ERK, and NUR77 by immunoblotting.

Flow cytometry of transcription factor expression

One million T cells were collected and washed with 2% BSA PBS and stained with Aqua, anti-CD4, anti-CD45RA, and anti-CD62L for 30 min on ice. Cells were washed once and stained for IKAROS as recommended by the manufacturer (BioLegend, #424401). In brief, cells were fixed with 1 ml Transcription Factor 1× Fix solution for 30 min at RT, 2 ml of the Transcription Factor 1× Perm Buffer was added, centrifuged at 300–400 × g at room temperature for 5 min and wash again with 2 ml 1× Perm Buffer. Cell pellets were resuspended in 100 µl of the Transcription Factor 1× Perm Buffer; Alexa Fluor® 647-labeled anti-human IKAROS Ab (1:50 dilution, BioLegend, #368404) was added and stained for 30 min at RT.

Immunohistochemical staining

Tissues were snap frozen in OCT and stored at −80 °C. Five micromole sections were air-dried and fixed with acetone at 4 °C for 10 min for staining or shock frozen for RNA extraction. Synovial T cells were identified by immunohistochemistry with mouse anti-human CD3 Abs (1:50), and IFN-γ production was detected with rabbit anti-human IFN-γ Abs (1:100) as described³⁸. Sections were analyzed with an Olympus BX41 microscope and CellSense software.

Human synovial tissue-NSG mice chimera

NSG male mice (Jackson Laboratory, Bar Harbor, ME) were kept in pathogen-free facilities and used at the age of 8–12 weeks as described³⁶. Pieces of human synovial tissue were placed into a subcutaneous pocket. On day 7 after engraftment, mice were reconstituted with 20 million CD45RO⁻ PBMCs derived from RA patients transfected with 20 µM ORAI3 or control siRNA or HDs transfected with 20 µM IKAROS or control siRNA. All experiments were approved by the Palo Alto Veterans Administration Healthcare System Animal Care and Use Committee.

Statistical analysis

Statistical analysis was performed using PRISM 8.4.1 (GraphPad Software Inc., La Jolla, CA). Results are given as mean + SEM. Unpaired two-tailed and paired two-tailed Student’s t test or one-way analysis of variance, followed by Tukey’s multiple comparison test were used as appropriate. P < 0.05 was considered significant.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.