ORAI1 Limits SARS-CoV-2 Infection by Regulating Tonic Type I IFN Signaling

This article is featured in Top Reads, p.

Type I IFN (IFN-I) response provides the first line of defense against viral infection (1) and is also important for the development of adaptive immunity (2). IFN-I response is ubiquitous in almost all nucleated cells, whereas type III IFNs (IFN-IIIs) are restricted to anatomic barriers and specific immune cells (3). This selectivity is due to the receptor expression patterns: IFN-I binds IFN-α receptor 1 (IFNAR1) and IFNAR2, which are ubiquitously expressed, while IFN-III binds IFN-λ receptor 1 and L10RB, which are preferentially expressed in epithelial cells (3). Despite using different receptors, IFN-I and IFN-III activate the same downstream signaling pathways and induce transcription of hundreds of IFN-stimulated genes (ISGs), with IFN-III inducing lower levels of ISG expression than IFN-I (4). The importance of the IFN signaling pathway is underscored by the fact that most viruses have developed mechanisms to inactivate this pathway (5).

The coronavirus disease 2019 (COVID-19) pandemic is one of the worst crises of our times, prompting an urgent need to uncover host defense mechanisms to its causative agent, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). SARS-CoV-2 uses multiple approaches to evade host IFN response, including suppression of IFN-I production and IFN-I signaling (6). The clinical manifestations and pathology of COVID-19 are similar to those of SARS-CoV-1, but the severities of the diseases differ (7). Unlike severe SARS-CoV-1 infection, SARS-CoV-2 infection shows a wide range of clinical features, ranging from asymptomatic to mild and moderate to severe and critical. The clinical manifestations of virus infection depend on virus–host interactions. Asymptomatic outcomes of SARS-CoV-2 infection may thus be attributed to strong host innate antiviral defense. In clinics, patients with COVID-19 with circulating Abs to IFN-I or loss-of-function mutations in genes necessary to mount the IFN-I response developed life-threatening pneumonia (8, 9). Recent studies have found that unlike original coronaviruses, SARS-CoV-2 is highly sensitive to pretreatment with type I and III IFNs. Pretreatment of immortalized or primary human airway epithelial cells with IFN-I or IFN-III imparts resistance to SARS-CoV-2 infection (10–12). Further, recent studies aiming at boosting IFN-I signaling, via pharmacological activation of stimulator of IFN genes (STING), robustly inhibited infection of multiple coronaviruses, including SARS-CoV-2 in vitro and in vivo (13–16). Taken together, a picture emerges that tonic IFN-I signaling mediated by the baseline levels of IFN-I, which is known to prime the host antiviral status (17), can also provide a significant barrier to SARS-CoV-2 infection.

Store-operated Ca²⁺ entry (SOCE), induced by depletion of the endoplasmic reticulum (ER) Ca²⁺ stores after activation of G protein–coupled receptors or receptor tyrosine kinases, is a ubiquitous mechanism of elevation of intracellular Ca²⁺ concentration ([Ca²⁺]_i) in most cell types. Ca²⁺ release-activated Ca²⁺ (CRAC) channels are a specialized class of store-operated Ca²⁺ channels that play a primary role in the elevation of [Ca²⁺] in immune cells (18, 19). CRAC channels consist of two major components: the plasma membrane–localized pore subunit, ORAI1; and an ER-resident Ca²⁺ sensor, stromal interaction molecule 1 (STIM1). STIM1 senses depletion of the ER Ca²⁺ stores and interacts with ORAI1 to open the pore. Ca²⁺ signaling mediated by CRAC channels is essential for induction of transcriptional programs via the NFAT pathway (18, 19). Severe combined immune deficiency caused by mutations in ORAI1 or STIM1, and the widespread clinical use of inhibitors of this pathway, cyclosporine A (CsA) and FK506, as immunosuppressants underscore the importance of therapeutic targeting of the Ca²⁺-NFAT pathway (18, 19).

Although the current understanding of the role of ORAI1 is limited to Ca²⁺ signaling, a Ca²⁺ signaling–independent role of STIM1 in regulating the IFN-I response has been shown (20). Biochemical studies identified STIM1 as an interacting partner of STING, an ER-resident adaptor protein that plays a central role in cellular response to cytosolic DNA (21). It was shown that STIM1 forms a complex with STING on the ER membrane to prevent its aberrant activation in an SOCE-independent manner. Myeloid cell–specific Stim1 knockout (KO) mice and a patient lacking STIM1 expression were shown to have elevated serum IFN-β and increased expression of various ISGs (20, 22). Accordingly, myeloid-specific Stim1 KO mice showed enhanced resistance to DNA virus infection, which was not observed in Orai1 KO cells and mice (20). In this study, we examined the role of ORAI1 and STIM1 in host resistance to infection with an RNA virus, SARS-CoV-2. Loss of ORAI1 markedly reduced the host resistance to SARS-CoV-2 by abolishing the baseline IFN-β levels and impairing tonic IFN-I signaling. On the contrary, STIM1 KO cells showed remarkable resistance to SARS-CoV-2 infection, supporting previous studies of enhanced baseline IFN-I signaling (20). Our data suggest that tonic IFN-I signaling regulated by ORAI1 and STIM1 determines the cellular antiviral state and host resistance to SARS-CoV-2 infection.

Fura 2-AM (F1221) was purchased from Thermo Fisher Scientific. Thapsigargin and cyclosporin A were purchased from EMD Millipore. SARS-CoV Spike (S) Ab, polyclonal anti-SARS coronavirus (BEI Resources: NR-10361 antiserum, Guinea Pig), was used for immunoblotting, and monoclonal anti-SARS-CoV S protein Ab (BEI Resources: NR-616, similar to 240C) was used for immunocytochemistry (ICC). Abs for detection of human STIM1 (5668S) were purchased from Cell Signaling Technologies, Abs for detection of IFNAR1 and IFN-β receptor 1 were purchased from Leinco Technologies (I-400), Abs for detection of ORAI1 (ACC-060) were purchased from Alomone Labs, and Abs for detection of β-actin (sc-47778) were obtained from Santa Cruz Biotechnology. BTP2 and ionomycin were purchased from Millipore Sigma.

The human angiotensin-converting enzyme 2 (ACE2) coding sequence was cloned into a lentiviral vector as described previously (23). Single guide RNAs (sgRNAs) targeting human ORAI1 (guide RNA target sequence 5′-GATCGGCCAGAGTTACTCC-3′) and human STIM1 (5′-TGAGGATAAGCTCATCAGCG-3′) genes were subcloned into pLentiguide puro (Addgene). sgRNAs targeting human IFNAR1 and IFN-β receptor 1 subcloned into pLentiCRISPR v2 were purchased from GenScript (catalog no. IFNAR1 crRNA 1; guide RNA target sequence 5′-GGCGTGTTTCCAGACTGTTT-3′). Vero E6, HEK293, and A549 cells were obtained from the American Type Culture Collection center (ATCC, Manassas, VA). Vero cells were cultured in Eagle’s MEM growth media (ATCC) supplemented with 10% (v/v) FBS (Hyclone) and penicillin/streptomycin (Pen/Strep; Mediatech) at 37°C and 5% CO₂. HEK293 cells were cultured in DMEM (Mediatech) supplemented with 10% (v/v) FBS (Hyclone), 2 mM l-glutamine (Mediatech), 10 mM HEPES (Mediatech), and Pen/Strep (Mediatech) at 37°C and 5% CO₂. A549 cells were cultured in Ham’s F12-K (Kaighn’s) medium (Thermo Fisher) supplemented with 10% (v/v) FBS (Hyclone), 10 mM HEPES (Mediatech), and Pen/Strep (Mediatech) at 37°C and 5% CO₂.

SARS-CoV-2 (USA-WA1/2020) was obtained from BEI Resources of National Institute of Allergy and Infectious Diseases. All the studies involving live virus were conducted in the UCLA BSL3 high-containment facility. SARS-CoV-2 was passaged once in Vero E6 cells, and viral stocks were aliquoted and stored at −80°C. Virus titer was measured in Vero E6 cells by established plaque assays.

To generate HEK293-ACE2 or A549-ACE2 cells, we transfected HEK293T cells with plasmid encoding ACE2 and packaging vectors (pMD2.G and psPAX2; Addgene) using calcium phosphate transfection method. Culture supernatants were harvested at 48 and 72 h posttransfection, filtered using a 0.45-μm polyvinylidene difluoride filter, and used for infection of HEK293 or A549 cells together with polybrene (8 µg/ml) using the spin-infection method. Transduced cells were selected with puromycin (1 μg/ml) 48 h after a second infection. For generation of ORAI1^−/− and STIM1^−/− cells, HEK293T cells were transfected with plasmids encoding sgRNA and packaging vectors as described earlier. Lentiviruses encoding Cas9 were generated using the same technique. Culture supernatants harvested at 48 and 72 h posttransfection were used for infection (50% of Cas9-encoding virus + 50% of sgRNA-encoding virus) of HEK293-ACE2 or A549-ACE2 cells together with polybrene (8 µg/ml) using the spin-infection method. Cells were selected with puromycin (1 µg/ml) and blasticidin (5 µg/ml) 48 h after the second infection.

For total endogenous ORAI1 detection, control, STIM1 KO, and ORAI1 KO HEK293 (0.5 × 10⁶) were fixed with 4% paraformaldehyde at room temperature for 15 min, followed by permeabilization with PBS + 0.5% Igepal CA-630, and incubated with 1 μg unlabeled anti-ORAI1 Ab (ACC-060; Alomone laboratories) in PBS + 2% FBS + 0.5% Igepal CA-630. Negative control sample did not contain the primary Ab. Cells were subsequently treated with FITC-conjugated secondary Ab, washed, and data were acquired on a BD Fortessa flow cytometer (BD Biosciences) and analyzed using FlowJo software (Tree Star).

HEK293-ACE cells were grown overnight on coverslips and loaded with 1 μM Fura 2-AM for 40 min at 25°C for imaging. [Ca²⁺]_i measurements were performed using essentially the same methods as previously described (24). In brief, microscopy was performed using an Olympus IX2 illumination system mounted on an Olympus IX51 inverted microscope using previously described methods (25). Acquisition and image analysis were performed using Slidebook (Intelligent Imaging Innovations) software, and graphs were plotted using Origin2020 (OriginLab). For immunofluorescence analysis, uninfected or SARS-CoV-2–infected cells were fixed for 20 min with ice-cold methanol, permeabilized in buffer containing PBS + 0.2% Triton X-100, and blocked with the same buffer containing 2.5% goat serum, 2.5% donkey serum, and 2% BSA. Cells were stained overnight in the blocking buffer with primary Abs washed and treated with secondary Ab for 1 h, stained for DAPI in permeabilization solution, visualized using 40× oil immersion lens, and imaged using Slidebook 6.0 software (Intelligent Imaging Innovations). Images were processed for enhancement of brightness or contrast using Slidebook 6.0 software.

Control, ORAI1^−/−, or STIM1^−/− HEK293-ACE2 cells or A549-ACE2 cells were seeded at 1 × 10⁵ cells/well in 0.4 ml vol in a 48-well plate, or 5 × 10⁴ cells/well in 0.2 ml vol in a 96-well plate. The following day, viral inoculum (multiplicities of infection [MOIs] of 0.1 or 1.0 as indicated) was prepared using serum-free media. The spent media from each well were removed, and 100 µl of prepared inoculum was added onto cells. For mock infection, serum-free media (100 µl/well) alone were added. The inoculated plates were incubated for 1 h at 37°C with 5% CO₂. The inoculum was spread by gently tilting the plate sideways every 15 min. At the end of incubation, the inoculum was replaced with serum-supplemented media. At 20 h postinfection, the cells were either fixed with methanol for ICC analysis, harvested in TRIzol reagent (Thermo Fisher) for RNA sequencing (RNA-seq) and RT-PCR analysis, or lysed in radioimmunoprecipitation assay lysis buffer for immunoblotting as described later.

Viral titers were determined by plaque assay on Vero E6 cells. Vero E6 cells were seeded in 12-well plates for 24 h. Virus-containing supernatants were 10-fold serially diluted in PBS. The growth medium was removed from the cells, cells were washed once with PBS, and diluted supernatants were added (150 μl/well). After 30-min inoculation, an overlay medium (double-concentrated MEM [supplemented with 2% l-glutamine, 2% Pen/Strep, 0.4% BSA], mixed 1:1 with 2.5% Avicel solution [prepared in ddH₂O]) was added to the cells (1.5 ml/well). Then, cells were incubated for 72 h at 37°C. After 72 h, the overlay medium was removed from the cells, and after a washing step with 1× PBS, the cells were fixed with 4% paraformaldehyde for 30 min at 4°C. Subsequently, the cells were counterstained with crystal violet solution to visualize the virus-induced plaques in the cell layer. The number of plaques at a given dilution was counted to calculate the viral titers as PFUs (PFUs/ml).

Total RNA from cells harvested in TRIzol Reagent (Thermo Fisher) was isolated using the Direct-zol RNA isolation kit (Zymo Research). RNA quantity and quality were confirmed with a NanoDrop ND-1000 spectrophotometer. cDNA was synthesized using 0.5–1.0 μg of total RNA using random hexamers and oligo(dT) primers and qScript cDNA Supermix (Quantabio). Real-time quantitative PCR was performed using iTaq Universal SYBR Green Supermix (Bio-Rad) and an iCycler IQ5 system (Bio-Rad) using gene-specific primers. In brief, amplification was performed using 20-μl vol reactions in a 96-well plate format with the following conditions: 95°C for 30 s for polymerase activation and cDNA denaturation, then 40 cycles at 95°C for 15 s and 60°C for 1 min, with a melt-curve analysis at 65°C–95°C and 0.5°C increments at 2–5 s/step. Threshold cycles (C_T) for all the candidate genes were normalized to those of 36B4 to obtain ΔC_T. The specificity of primers was examined by melt-curve analysis and agarose gel electrophoresis of PCR products. Primer sequences are as follows: SARS-CoV-2 (forward: 5′-GACCCCAAAATCAGCGAAAT-3′, reverse: 5′-TCTGGTTACTGCCAGTTGAATCTG-3′), human ATF2 (forward: 5′-GCACAGCCCACATCAGCTATT-3′, reverse: 5′-GGTGCCTGGGTGATTACAGT-3′); human FOS (forward: 5′-GGGGCAAGGTGGAACAGTTAT-3′, reverse: 5′-CCGCTTGGAGTGTATCAGTCA-3′); human MEF2C (forward: 5′-GTATGGCAATCCCCGAAACT-3′, reverse: 5′-ATCGTATTCTTGCTGCCTGG-3′), and human 36B4 (forward: 5′-AGATGCAGCAGATCCGCAT-3′, reverse: 5′-GTTCTTGCCCATCAGCACC-3′).

Cells were lysed in radioimmunoprecipitation assay lysis buffer (50 mM Tris [pH 7.4], 1% Nonidet P-40, 0.25% sodium deoxycholate, 1 mM EDTA, 150 mM NaCl, 1 mM Na₃VO₄, 20 mM NaF, 1 mM PMSF, 2 mg/ml aprotinin, 2 mg/ml leupeptin, and 0.7 mg/ml pepstatin) for 1 h with intermittent vortexing and centrifuged to remove debris. Samples were separated on 10% SDS-PAGE. Proteins were transferred to nitrocellulose membranes and subsequently analyzed by immunoblotting with relevant Abs. Chemiluminescence images were acquired using an Image reader LAS-3000 LCD camera (FujiFilm).

For ELISA measurements under resting conditions, 1 million cells were cultured in low media volume to allow for concentration of cytokines. ELISA was performed on cell culture supernatants from indicated cells for detection of IFN-β (#41410; PBL Assay Science) or IL-6 (#88-7066-22; Thermo Fisher Scientific) using the manufacturer’s instructions.

RNA was extracted with Direct-zol RNA Mini prep kit (Zymo Research). Libraries for RNA-seq were prepared with TruSeq Stranded mRNA Library Prep Kit (Illumina). The workflow consisted of mRNA enrichment and fragmentation. Cleaved RNA fragments were copied into first-strand cDNA using reverse transcriptase and random primers. Strand specificity is achieved by replacing dTTP with deoxyuridine triphosphate, followed by second-strand cDNA synthesis using DNA Polymerase I and RNase H. cDNA generation is followed by A-tailing, adaptor ligation, and PCR amplification. Different adaptors were used for multiplexing samples in one lane. Sequencing was performed on Illumina HiSeq 3000 for SE 1 × 50 run. Data quality check was done on Illumina SAV. Demultiplexing was performed with Illumina Bcl2fastq v2.19.1.403 software. Raw reads for Orai1KO and STIM1KO SARS-CoV-2–infected samples (Orai1KO-SARS-CoV-2-3_S18 and STIM1KO-SARS-CoV-2-1_S19) were sampled down to 45 million reads before mapping. Illumina reads from all samples were mapped to human and SARS-CoV-2 reference genomes by STAR v2.27a (26), and read counts per gene were quantified using human Ensembl GRCh38.98 and SARS-CoV-2 GTF file to generate raw reads counts. Differential expression analysis was performed using DESeq2 v1.28.1 in R v4.0.3 (27). Median of ratios method was used to normalize expression counts for each gene in all experimental samples. Each gene in the samples was fitted into a negative binomial generalized linear model. Differentially expressed gene (DEG) candidates were considered if they were supported by a false discovery rate (FDR) p < 0.01. Unsupervised principal component analysis was performed using pcaExplorer (28) in R v4.0.3 (27). Reactome pathway analysis was performed for DEGs using all human genes as reference dataset in the Reactome v65 (29) implemented in PANTHER. Reactome pathways were considered only if they were supported by FDR p < 0.05. The ggplot2 v3.3.2 in R and Prism GraphPad v8.4.3 were used to generate figures. The shinyheatmap web interface was used to prepare heatmaps (30). The genes in the DEGs directly regulated by transcription factors (FOS, ATF2, MEF2C, and JUN) were identified based on binding profiles of all public ChIP-seq data for particular gene loci from the ChIP-Atlas database (31). The regulated or target genes were accepted if the peak-call intervals of a given gene overlapped with a transcription start site ±1 kb. RNA-seq data have been deposited to the NCBI GEO under the accession number GSE173707 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE173707).

Statistical analysis was carried out using two-tailed Student t test. Differences were considered significant when p values were <0.05. The hypergeometric test in R v4.0.3 was used to measure the statistical significance of the enriched genes in the DEGs regulated by transcription factors (FOS, ATF2, MEF2C, and JUN) compared with regulated genes in the background gene sets of the human reference genome.

To examine the role of ORAI1 and STIM1 in host response to SARS-CoV-2, we generated HEK293 cells stably expressing the receptor for SARS-CoV-2, ACE2 (32). The resulting HEK293-ACE2 cells were transduced with lentiviruses encoding Cas9 and sgRNAs targeting either ORAI1 or STIM1. All three lines expressed similar levels of ACE2 as judged by ICC and immunoblotting (Supplemental Fig. 1A, 1B). Loss of STIM1 protein expression was validated by immunoblotting, while that of ORAI1 was validated by flow cytometry (Fig. 1A, 1B). Multiple studies have shown loss of SOCE in HEK293 cells deleted for expression of ORAI1/STIM1 (33, 34). To measure SOCE in ORAI1^−/− and STIM1^−/− HEK293-ACE2 cells, we used thapsigargin, a blocker of sarcoplasmic reticulum/ER Ca²⁺-ATPase pump, that allows for passive depletion of the ER Ca²⁺ stores, thereby activating SOCE. We observed almost complete abrogation of SOCE in both ORAI1^−/− and STIM1^−/− HEK293-ACE2 cells (Fig. 1C). Although control cells showed a uniform increase in SOCE after reintroducing Ca²⁺-containing Ringer’s solution, more than 95% of the polyclonal ORAI1^−/− and STIM1^−/− HEK293-ACE2 cells remained unresponsive. The Ca²⁺ release from the ER in those cells was similar. Next, we checked whether loss of ORAI1/STIM1 affected resting cytoplasmic [Ca²⁺] in unstimulated cells. Using our previously established protocols, we sequentially perfused the cells with 0 (Ca²⁺-free), 2, and 20 mM Ca²⁺-containing Ringer’s solutions without any store depletion, to measure the homeostatic Ca²⁺ levels in cells (25). Interestingly, in these measurements, only ORAI1^−/− HEK293-ACE2 cells showed significant reduction in baseline [Ca²⁺] in the presence of 2 or 20 mM Ca²⁺-containing Ringer’s solution, suggesting an additional role of ORAI1 in regulating homeostatic [Ca²⁺] in these cells (Fig. 1D). These data are in concurrence with previous observations where STIM2, a homolog of STIM1, was shown to be involved in maintaining resting [Ca²⁺]_i together with ORAI1 (35). In those experiments, STIM2-depleted cells showed decreased resting [Ca²⁺]_i, while those depleted for STIM1 had normal levels of resting [Ca²⁺]_i. Collectively, by stable expression of ACE2 and deletion of ORAI1/STIM1, we developed a system to examine the role of ORAI1 and STIM1 in host responses to SARS-CoV-2 infection and uncovered the role of ORAI1 in Ca²⁺ homeostasis in these cells.

Next, we infected control, ORAI1^−/−, and STIM1^−/− HEK293-ACE2 cells with SARS-CoV-2 at two different MOIs, – 0.1 and 1.0. Twenty hours postinfection, cells were harvested to examine expression of viral proteins, as well as quantification of viral genome. We observed high levels of viral protein expression in SARS-CoV-2–infected ORAI1^−/− HEK293-ACE2 cells when compared with control cells by immunoblotting (Fig. 2A), such that 20 h postinfection at high MOI (1.0), a significant fraction of the ORAI1^−/− cells was lysed (Fig. 2B). On the contrary, STIM1^−/− HEK293-ACE2 cells showed strong resistance to SARS-CoV-2 infection as judged by markedly reduced expression of viral protein by immunoblotting and fewer plaques by ICC when compared with control cells (Fig. 2A, 2B). Genome copy measurements at 5 h postinfection showed modestly increased and decreased susceptibility of ORAI1^−/− and STIM1^−/− cells, respectively, especially in cells infected at high MOI (Fig. 2C, left). Further, the increased and decreased susceptibility of ORAI1^−/− and STIM1^−/− cells, respectively, were obvious at 20 h postinfection, where ORAI1^−/− cells showed an ∼100-fold increase in viral genome copy numbers, whereas STIM1^−/− cells showed more than 100-fold reduction, when compared with control HEK293-ACE2 cells (Fig. 2C, right). Taken together, these data show seemingly opposite phenotypes of resistance and susceptibility to SARS-CoV-2 infection in STIM1^−/− and ORAI1^−/− cells, respectively.

Based on our previous data showing preactivation of the IFN-I pathway in Stim1^−/− fibroblasts and macrophages (20) and high sensitivity of SARS-CoV-2 to IFN-I (10–12), we surmised that the resistance of STIM1^−/− cells to SARS-CoV-2 infection was most likely derived from an increase in the baseline IFN-I response. As expected, STIM1^−/− HEK293-ACE2 cells indeed showed higher levels of baseline IFN-β levels, which remained elevated even after SARS-CoV-2 infection (Fig. 3A). We also observed enhanced IL-6 production by STIM1^−/− cells under resting conditions and after SARS-CoV-2 infection when compared with control cells (Fig. 3B). Notably, these results showed that SARS-CoV-2 infection did not trigger a robust IFN-I response but induced strong IL-6 expression consistent with previous observations (36).

To validate whether the resistance of STIM1^−/− cells to SARS-CoV-2 infection was due to elevated IFN-I signaling, we generated HEK293-ACE2 cells deleted for both STIM1 and IFNAR1 using the CRISPR-Cas9 genome editing system (Fig. 3C). Deletion of IFNAR1 abolished the increased baseline IFN-β expression in STIM1^−/− cells (Fig. 3D). In addition, codeletion of IFNAR1 abolished the resistance of STIM1^−/− cells to SARS-CoV-2, as demonstrated by enhanced expression of viral proteins and increased burden of viral genome copies (Fig. 3E). These results suggest that the strong resistance of STIM1^−/− cells to SARS-CoV-2 is predominantly derived from enhanced basal level of IFN-β. Considering that SARS-CoV-2 did not trigger robust IFN-I signaling, these results also emphasize the importance of the basal IFN-I levels in host resistance to SARS-CoV-2 infection.

The increased susceptibility of ORAI1^−/− cells to SARS-CoV-2 infection was surprising because ORAI1 deficiency did not influence host resistance to infection with a DNA virus, HSV type 1 (HSV-1), in a previous study (20). Because SARS-CoV-2 is highly sensitive to the baseline IFN-I levels as observed in our analysis of STIM1^−/− cells, we hypothesized that ORAI1 may influence the basal IFN-I levels. Culture supernatant analysis showed almost complete abrogation of baseline IFN-β levels in ORAI1^−/− cells when compared with control cells under resting conditions, as well as on infection with SARS-CoV-2 at low MOI (Fig. 4A). IL-6 levels in these cells were not influenced. In support of reduced IFN-β levels, ORAI1^−/− cells also showed reduced activation of the tonic IFN-I signaling pathway, including reduced basal levels of p-STAT1, as well as decrease in total STAT1 proteins (Fig. 4B), which has been proposed to be dependent on tonic IFN-β signaling (17).

To check whether the high susceptibility of ORAI1^−/− cells to SARS-CoV-2 infection was rescued by treatment with IFN-β, we pretreated these cells with low levels of IFN-β for ∼20 h before SARS-CoV-2 infection. IFN-β pretreatment reduced the viral genome by ∼10-fold in control cells, in agreement with other studies (10–12) (Fig. 4C). Importantly, pretreatment of ORAI1^−/− cells with IFN-β made the cells highly resistant, with more than 1,000-fold reduction in viral genome copy number, similar to the levels observed with control cells. These observations were validated by immunoblotting to check expression of viral proteins, where pretreatment with IFN-β abrogated expression of viral proteins equally in control and ORAI1^−/− cells (Fig. 4D). Because IFN-β pretreatment in ORAI1^−/− cells almost completely rescued the phenotypes to the level in control cells, these data suggest that reduced IFN-I baseline may have a significant contribution toward high susceptibility of ORAI1^−/− cells to SARS-CoV-2 infection. Because the NFAT family of transcription factors are well-known downstream effectors of ORAI1-mediated SOCE, we checked whether reduced IFN-β levels in ORAI1^−/− cells were due to loss of NFAT function. To check this possibility, we used CsA, which blocks calcineurin, a phosphatase necessary for dephosphorylation of NFAT in the cytoplasm. We pretreated control and ORAI1^−/− HEK293-ACE2 cells with CsA and examined its effect on the basal IFN-β production. Surprisingly, CsA treatment did not affect IFN-β levels in either of these cell types, suggesting a potential role of Ca²⁺-dependent transcription factors other than NFAT in this event (Fig. 4E).

To uncover the antiviral defense mechanisms impaired in ORAI1^−/− cells, we performed transcriptome analysis by bulk RNA-seq of control, ORAI1^−/−, and STIM1^−/− HEK293-ACE2 cells before and 20 h after SARS-CoV-2 infection (Supplemental Fig. 2). Among infected samples, viral count analysis showed uniform increase in expression of all the viral genes in ORAI1^−/− cells, whereas the same was uniformly reduced in STIM1^−/− cells, when compared with control cells (Fig. 5A). Importantly, the viral genome reads comprised ∼20% of the total reads for control cells, ∼77% for ORAI1^−/− cells, and <1% for STIM1^−/− cells, in concurrence with the observed phenotypes (Fig. 5B). Pathway enrichment analysis of ORAI1^−/− cells before infection showed downregulation of multiple antiviral defense mechanisms, including ISG15 antiviral mechanism and MyD88 signaling pathways (Fig. 5C–E). Transcriptome analysis further identified multiple transcription factors known to be involved in antiviral pathways, whose expression was altered by loss of ORAI1 (Fig. 5F). Among the transcription factors that were significantly downregulated in ORAI1^−/− cells, we identified MEF2C and members of the AP-1 family of transcription factors, FOS, ATF2, and JUN. All of these transcription factors are known to be Ca²⁺ dependent for their functions (37–40). The AP-1 family of transcription factors, including FOS, JUN, and ATF2, are known to bind the IFNB promoter to regulate its expression, and JUN is also known to mediate tonic IFN signaling (17). Remarkably, promoter analysis showed profound enrichment of FOS, ATF2, MEF2C, and JUN bindings sites among the DEGs in ORAI1^−/− cells compared with the reference genome (Fig. 5G). Quantitative RT-PCR analysis confirmed downregulation of these transcription factors in ORAI1^−/− cells under sterile conditions (Fig. 5H). Together, these data suggest that ORAI1 activity upregulates expression of the transcription factors MEF2C, FOS, ATF2, and JUN, leading to induction of multiple genes involved in antiviral defense mechanisms, including those involved in tonic IFN-I signaling, providing resistance to SARS-CoV-2 infection.

Recent studies using genome-wide CRISPR-Cas9 KO screens with SARS-CoV-2 infection have identified a plethora of host factors important for productive infection by SARS-CoV-2 (41–45). The list of necessary host factors required for SARS-CoV-2 propagation varied widely among the different screens, likely because of different cell types and infection conditions used for each screen. Not surprisingly, all the genome-wide screens identified ACE2, the receptor for SARS-CoV-2, as an essential factor for virus propagation. Our transcriptome analysis identified many of these host factors among the DEGs in ORAI1^−/− cells (Supplemental Fig. 3). However, many of these factors were downregulated in ORAI1^−/− cells, suggesting they may not play a significant role in high susceptibility of ORAI1^−/− cells to SARS-CoV-2 infection. Collectively, these results suggest that ORAI1-mediated Ca²⁺ signaling is mainly important for the regulation of tonic IFN-I levels, rather than host factors for SARS-CoV-2 infection.

Because SARS-CoV-2 infects the respiratory tract, we sought to validate our observations with human lung cells. We generated A549 cells stably expressing ACE2 and deleted for expression of ORAI1 using the CRISPR-Cas9 genome editing system (Fig. 6A). Transcript analysis confirmed reduced expression of IFNB in ORAI1^−/− A549-ACE2 cells (Fig. 6B). Next, we infected control and ORAI1^−/− A549-ACE2 cells with SARS-CoV-2 and examined expression of viral proteins, as well as quantified expression of viral genome. Similar to our observations with HEK293-ACE2 cells, loss of ORAI1 imparted susceptibility to SARS-CoV-2 infection even in A549 cells, albeit to a lesser extent (Fig. 6C, 6D).

Recent studies focusing on identifying therapies for COVID-19 have found beneficial effects of IFN administration, especially at early stages of the disease, to reduce the severity and mortality associated with SARS-CoV-2 infections (46, 47). To check whether pharmacological alteration of homeostatic [Ca²⁺]_i affects host resistance to SARS-CoV-2, we pretreated A549-ACE2 cells with various doses of a well-known blocker of ORAI1, BTP2 (chemical name: N-[4-[3,5-Bis(trifluoromethyl)-1H-pyrazol-1-yl]phenyl]-4-methyl-1,2,3-thiadiazole-5-carboxamide), for ∼24 h to reduce homeostatic [Ca²⁺]_i. Transcript analysis confirmed reduced expression of IFNB in blocker-treated cells (Fig. 6E). Further, infection with SARS-CoV-2 showed increased viral protein expression and replication in cells treated with the ORAI1 blocker, in concurrence with our observations with ORAI1^−/− cells (Fig. 6F, 6G). Importantly, enhancing baseline [Ca²⁺] by pretreatment with low doses of an ionophore, ionomycin, elevated IFNB expression and imparted resistance to subsequent SARS-CoV-2 infection (Fig. 6H–J). These data show that enhancing baseline [Ca²⁺]_i is sufficient to impart resistance to SARS-CoV-2 infection, most likely via boosting tonic IFN-I signaling.

The role of Ca²⁺ signaling mediated by ORAI1 and STIM1 in adaptive immune cells (e.g., T and B cells) is well established (18, 19). Further, an earlier report suggested that loss of SOCE imparted susceptibility to IFN-I–induced cell death in Jurkat T cells (48). However, the function of this pathway in the host–pathogen responses in innate immunity is poorly understood. A recent study had identified an important function of STIM1 in regulating the cytosolic DNA-sensing pathway via direct interaction with STING independently from its role in Ca²⁺ signaling (20). This study demonstrates a key role of ORAI1 in regulating tonic IFN-β levels and thereby the host response to SARS-CoV-2 infection by modulation of homeostatic cytoplasmic [Ca²⁺]. Collectively, these studies uncover novel functions of CRAC channel components in the innate immune system.

Loss of STIM1 provides strong resistance to infection with HSV-1 by enhancing IFN-I signaling (20). The observation of resistance to virus infections in STIM1^−/− cells has been extended to this study using SARS-CoV-2. Abrogation of IFN-I signaling by codeletion of IFNAR1 reduced baseline IFN-β secretion and rendered STIM1^−/− cells susceptible to SARS-CoV-2 infection, emphasizing the role of IFN-I in this resistance. Interestingly, loss of ORAI1 did not affect host resistance to HSV-1, most likely because the STING signaling pathway, which is predominantly important for sensing DNA viruses, was not impaired in ORAI1^−/− cells (20). However, loss of ORAI1 resulted in very high susceptibility, specifically to SARS-CoV-2 infection. In concurrence, pharmacological inhibition of ORAI1 to reduce homeostatic [Ca²⁺]_i or enhancing homeostatic [Ca²⁺]_i by pretreatment with ionomycin modulated host susceptibility to SARS-CoV-2 infection. SARS-CoV-2 shows exceptionally high sensitivity to pretreatment with IFN-I or IFN-III, which profoundly reduces virus replication (10–12). Hence reduction in tonic IFN-β levels may enhance the susceptibility of ORAI1^−/− cells, especially to SARS-CoV-2 infection. It has been shown that in the absence of priming amounts of IFN-β, mouse embryonic fibroblasts do not produce other IFN-I, suggesting that IFN-β is a master regulator for all IFN-I activities and presumably those mediated by IFN-III (49, 50). It has also been suggested that tonic IFN-β signaling is required to maintain adequate expression of STAT1 and STAT2; accordingly, the absence of tonic IFN-β signaling reduces basal STAT expression (17), similar to our observation with ORAI1^−/− cells (Fig. 4B). Besides, loss of ORAI1 was shown to downregulate multiple pathways, including ISG antiviral signaling, TLR7/8 signaling, and MyD88 signaling cascades involved in RNA sensing. Collectively, loss of tonic IFN-I signaling in combination with impaired function of other antiviral signaling cascades is likely to contribute toward the exquisite sensitivity of ORAI1^−/− cells to SARS-CoV-2.

The IFNB promoter contains four positive regulatory domains (PRDI–PRDIV), which are occupied by overlapping transcription factor complexes. IFN regulatory factor 3 (IRF3) and IRF7 bind PRDI and PRDIII, the NF-κB RelA-p50 heterodimer binds PRDII, and the AP-1 heterodimer of ATF-2 and c-Jun binds PRDIV. The binding of each of these components in the correct orientation and location results in activation of the IFNB promoter in response to virus infection (17). Tonic IFN-β expression is independent of IRF3 and IRF7 but instead depends on the c-Jun and NF-κB p50 subunit. ORAI1^−/− cells showed reduced expression of multiple AP-1 family transcription factors, including ATF-2, FOS, and JUN, as well as the NF-κB p50 subunit, which are likely to contribute to reduced baseline IFN-β levels in these cells. Notably, although SOCE was abolished in both ORAI1^−/− and STIM1^−/− cells, the baseline [Ca²⁺] was reduced in only ORAI1^−/− cells. Based on the previous finding that STIM2, which has a lower binding affinity to Ca²⁺ than STIM1, is crucial for regulation of homeostatic Ca²⁺ levels (35), it is expected that STIM2^−/− cells may have lower tonic IFN-I signaling, and thereby high susceptibility to SARS-CoV-2 infection similar to ORAI1^−/− cells. Another notable difference between ORAI1 and STIM1 deficiency is differential expression of IFN-β and IL-6. STIM1^−/− cells showed enhanced expression of both of these cytokines, consistent with the observation that activation of the STING pathway upregulates expression of both IFN-β and IL-6 (20, 21). However, ORAI1^−/− cells had normal expression of IL-6 but still showed high susceptibility to SARS-CoV-2 infection, suggesting that the protective effect of STIM1 deficiency is predominantly derived from increased levels of IFN-β. Therefore, these results obtained from the analysis of ORAI1^−/− and STIM1^−/− cells cohesively suggest that tonic IFN-I signaling is a crucial determinant for the degree of host resistance to SARS-CoV-2.

In summary, we examined the role of CRAC channel components, ORAI1 and STIM1, in host resistance to SARS-CoV-2 infection. STIM1^−/− cells showed remarkable resistance to SARS-CoV-2 infection, supporting previous studies of the role of elevated baseline IFN-I levels (20). Loss of ORAI1 severely impaired the tonic IFN-I levels, thereby reducing resistance to SARS-CoV-2 infection. Mechanistically, ORAI1-mediated Ca²⁺ signaling affects expression of various transcription factors, including FOS, JUN, ATF2, and MEF2C, that regulate multiple host defense pathways. The key finding of this study is that the baseline IFN-β level regulated by ORAI1-mediated Ca²⁺ signaling is critical for priming the cellular antiviral state, thereby determining host resistance to SARS-CoV-2 infection. A recent phase 2 clinical trial of inhaled, nebulized IFN-β to treat people with COVID-19 resulted in accelerated recovery, supporting localized delivery of IFN-I as a potential prophylactic and therapeutic agent against COVID-19 (47). The therapeutic effect of IFN-I targeting SARS-CoV-2 infection may be partly derived from boosting tonic IFN-I signaling. Our findings may help develop novel therapeutic methods to combat SARS-CoV-2 infection by demonstrating that pharmacological enhancement of baseline cytoplasmic [Ca²⁺] improved host resistance to SARS-CoV-2. Furthermore, the current study emphasizing the crucial role of tonic IFN-I signaling in resistance to SARS-CoV-2 infection may guide future identification of vulnerable populations and targeted therapies at boosting tonic IFN-I signaling.

We thank Barbara Dillon, UCLA High-Containment Program Director, for BSL3 work. We also thank Yousang Gwack (UCLA) for helpful suggestions and Shawn Cokus (UCLA Collaboratory Fellow) for suggestions about statistical analysis of transcription factor enrichment data.

The authors have no financial conflicts of interest.