Introduction

Ca2+ release-activated Ca2+ (CRAC) channels mediate a sustained increase in cytoplasmic Ca2+ concentration that is essential for T cell activation. T cell receptor stimulation induces the depletion of the endoplasmic reticulum (ER) Ca2+ stores that activates CRAC channels via a process called store-operated Ca2+ entry (SOCE). STIM1 and ORAI1 are two essential components of CRAC channels. STIM1 is an ER-resident regulatory subunit that senses depletion of the ER Ca2+ stores, and the plasma membrane (PM)-resident ORAI1 is the pore subunit of CRAC channels. Upon store depletion, STIM1 multimerizes and translocates from the ER to the ER-PM junctions (a space of 10–15 nm) via passive diffusion and interacts with ORAI1 to induce its opening1,2,3,4,5. The C-terminal poly(K) residues of STIM1 interact with phosphatidylinositol 4,5-bisphosphate (PIP2) in the PM for its recruitment to the ER-PM junctions. Further, protein interactions also play a crucial role for STIM1 recruitment into the junctions. We recently identified junctional proteins, including junctate and junctophilins, as important components of the ER-PM junctions in T cells6,7. Junctate-junctophilin complex is localized at the ER-PM junctions, and after sensing ER Ca2+ depletion via its ER-luminal Ca2+-binding motif, junctate recruits STIM1 into the junctions. Other molecules such as TMEM110 (alternatively, STIM-activating enhancer) and SARAF (SOCE-associated regulatory factor) have also been identified as components of the ER-PM junctions in T cells8,9,10. All these molecules interact with STIM1 to modulate its function, including translocation into the junctions (e.g., junctate and junctophilin-4) and induction of conformational changes (e.g., TMEM110 and SARAF). Other than this, our current understanding of the structural and regulatory components of the ER-PM junctions involved in Ca2+ signaling in T cells is limited.

The ER-PM junctions are ubiquitous structures essential for intermembrane communications, including lipid transfer and Ca2+ dynamics11. Yeast proteins, tricalbins (Tcb1p, Tcb2p, and Tcb3p) are selectively concentrated in the cortical ER and play an essential role in ER-PM tethering. Their mammalian homologs are E-Syt1, E-Syt2, and E-Syt3. E-Syts contain an ER membrane integration segment, a cytosolic synaptotagmin-like mitochondrial lipid-binding protein (SMP) domain followed by multiple PIP2 and Ca2+-binding C2 domains. SMP domains of E-Syts have a specialized function in lipid exchange12,13. E-Syt2 and E-Syt3 are ER-resident proteins that constitutively interact with the PM while E-Syt1 requires elevation of cytosolic Ca2+ (low micromolar range) for interaction with the PM23,24,25. Interestingly, we found that STIM1: ORAI1 ratio in Jurkat T cells was lower than that in HeLa cells due to abundant expression of ORAI1 (Fig. 4D). Therefore, it is possible that low STIM1: ORAI1 ratio in Jurkat T cells makes them dependent on accessory factors, including E-Syts, that regulate translocation of STIM1 to the junctions.

Figure 4
figure 4

E-Syts regulate recruitment of STIM1 to the ER-PM junctions. (A) Representative confocal images of endogenous STIM1 in control and DKO Jurkat T cells expressing GFP-tagged F-actin under resting conditions (left) and store depletion with TG (right) (Scale bars: 5 μm). Bar graphs show normalized cluster area (top) and the number of STIM1 clusters (below) (Ctrl: n = 9, DKO: n = 8). ***p < 0.001. (B) Representative confocal images of endogenous STIM1 in control and DKO Jurkat T cells stably expressing low levels of ORAI1-YFP in resting and store-depleted conditions (Scale bars: 5 μm). Bar graph shows normalized STIM1 (top) and ORAI1 (middle) clustering area, STIM1 and ORAI1 cluster intensity (below—left two panels) and STIM1/ORAI1 intensity ratio (below—right panel) (Ctrl: n = 5, DKO: n = 5). *p < 0.05. N.S. not significant. (C) Expression of ORAI1 and STIM1 in control and DKO Jurkat T cells. The bar graph shows densitometry analysis of expression levels of ORAI1 and STIM1 (normalized with β-actin) from two independent experiments. N.S. not significant. (D) Expression of ORAI1 and STIM1 in HeLa S3 and Jurkat T cells. Representative immunoblots showing levels of endogenous ORAI1 and STIM1 proteins in HeLa S3 and Jurkat T cells. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a loading control. The bar graph on the right shows densitometry analysis of the ratios of STIM1 to ORAI1 (normalized to that of GAPDH) from three independent cellular lysates. *p < 0.05.

Full size image

E-Syt2S recruits STIM1 to the ER-PM junctions via direct interaction

STIM1 is recruited into the ER-PM junctions via both intrinsic and extrinsic mechanisms; by binding of its C-terminal poly(K) residues to PIP2 at the PM, and via interaction with ORAI1 and other junctional proteins, including junctate and junctophilin-46,7,26,27,28,29. PIP2 needs to be replenished upon its depletion by phospholipase C (PLC)-mediated hydrolysis, in which E-Syts play a major role. Consistently, several reports have validated a role of E-Syt1 in STIM1 recruitment into the PIP2-enriched ER-PM junctions in various cell types15,16,17,18. Hence, it is likely that E-Syts regulate PIP2 enrichment at the junctions in T cells as well. However, since we identified E-Syts as interacting partners of STIM1 from affinity protein purification, we examined possible interaction between E-Syts and STIM1 by co-immunoprecipitation (co-IP). Under resting conditions there was detectable binding of STIM1 with both E-Syt2L and E-Syt2S, while E-Syt1 did not show any interaction (Fig. 5A). After store depletion, STIM1 binding to E-Syt2S was enhanced, whereas that with E-Syt2L was reduced. Under similar conditions, E-Syts did not show any interaction with ORAI1 (Fig. S3A). Interestingly, although E-Syt2S did not directly interact with ORAI1, it influenced STIM1-ORAI1 interaction. Addition of E-Syt2S, but not E-Syt2L to cellular lysates containing ORAI1 and STIM1, significantly increased STIM1-ORAI1 association (Fig. 5B). These results suggest that E-Syt2S, not E-Syt2L, recruits STIM1 to the junctions and enhances its interaction with ORAI1.

Figure 5
figure 5

Interaction between E-Syt2S and STIM1. (A) Interaction of E-Syts with STIM1. FLAG-immunoprecipitates from lysates of HEK293T cells expressing FLAG-tagged E-Syt1, E-Syt2L, or E-Syt2S together with His-tagged STIM1 were blotted for detection of the indicated proteins. Cells were treated with thapsigargin before lysis (1 μM TG for 10 min) in Ringer solution containing 2 mM Ca2+. Bar graph shows densitometry analysis of fold changes in STIM1 band intensity (normalized to individual immunoprecipitates) from three independent experiments. *p < 0.05, **p < 0.005, N.D not detected. (B) Interaction of ORAI1 and STIM1 in the presence of E-Syt2L or E-Syt2S. FLAG-immunoprecipitates from lysates of HEK293T cells expressing FLAG-tagged ORAI1 together with His-tagged STIM1 were blotted for detection of the indicated proteins. Lysates of HEK293T cells expressing E-Syt2L or E-Syt2S were included in the binding reaction separately (see “Materials and methods" section). The bar graph shows densitometry analyses of fold changes in STIM1 band intensity (normalized to individual immunoprecipitates) from four independent experiments. *p < 0.05, N.S. not significant. (C) Schematic showing domain structure of E-Syt1, E-Syt2L and E-Syt2S. Grey boxes indicate transmembrane (TM) segments that span the ER membrane. Cytoplasmic region contains an N-terminal short fragment, a synaptotagmin-like mitochondrial lipid-binding protein (SMP) domain, and multiple C2 domains involved in targeting proteins to the plasma membrane. E-Syt2S lacks the first 48 amino acids present in E-Syt2L. (D) Measurement of SOCE in Jurkat T cells expressing N-terminus of E-Syt2L (E2LN) or E-Syt2S (E2SN). Intracellular stores were passively depleted with thapsigargin (1 μM, TG) in Ca2+-free solution, and SOCE was measured by perfusion with 2 mM Ca2+-containing solution. Traces show averaged SOCE responses from 30 to 50 cells, and bar graph shows averaged SOCE response ± S.E.M. from three independent experiments. *p < 0.05, N.S. not significant.

Full size image

E-Syts contain an ER membrane integration segment, a cytosolic SMP domain followed by multiple PIP2 and Ca2+-binding C2 domains (Fig. 5C). SMP domains have specialized function in lipid exchange12,13. E-Syt2 and E-Syt3 are ER-resident proteins that constitutively interact with the PM, while E-Syt1 requires elevation of cytosolic Ca2+ for interaction with the PM8,9. Among these molecules, the depletion of only TMEM110 decreased the density of ER-PM junctions. The current study suggests that E-Syts are also the structural components of the ER-PM junctions as well as positive regulators of SOCE in T cells, together with TMEM110. The functional redundancy between E-Syts and TMEM110 in ER-PM tethering can be a topic of future investigation; however, they show apparent differences in their localization. While TMEM110 is uniformly distributed throughout the ER membrane, two members of the E-Syt family, E-Syt2 and E-Syt3, constitutively localize to the ER-PM junctions. These observations suggest that E-Syts can act as independent membrane-tethering factors, while localization of TMEM110 to the ER-PM junctions requires support from additional proteins, including STIM1 or potentially E-Syts.

Interestingly, although deletion of ESYT1 and ESYT2 significantly decreased the density of the ER-PM junctions in T cells, we found that their membrane-tethering function does not play a significant role in SOCE. This conclusion is based on the observation that the density of the junctions was reduced in both HeLa and Jurkat T cells, but SOCE was decreased only in Jurkat T cells. Also, artificial expansion of the junctions by membrane cross-linking did not rescue the reduced SOCE. Since STIM1 itself can expand the ER-PM junctions after store depletion4,22, it is possible that the decrease in the number of ER-PM junctions observed after ESYT deletion may not affect SOCE. The function of E-Syts in SOCE was more closely related to the regulation of STIM1 since STIM1 expression almost entirely rescued the decrease in SOCE and also, STIM1 recruitment was significantly decreased in ESYT DKO Jurkat T cells. There can be two possible explanations for the unique role of E-Syts in regulation of SOCE, specifically in human T cells. First, we found that the STIM1: ORAI1 ratio is lower in Jurkat T cells than HeLa cells, due to increased expression of ORAI1. Therefore, it is likely that SOCE in Jurkat T cells is sensitive to the reduced function of STIM1, in the absence of E-Syts. Second, T cells predominantly express E-Syt2S, while E-Syt2L is the major isoform in HeLa cells. Protein interaction studies showed that E-Syt2S-STIM1 interaction was enhanced when ER Ca2+ stores are depleted. These data suggest that E-Syt2S may be involved in recruitment of STIM1 to the junctions, similar to junctate and junctophilin-4, independently from PIP2 binding of STIM16,7. Our data suggest that direct regulation of STIM1 is the predominant role of E-Syts in T cells, compared to other known functions of E-Syts including glycolipid exchange, because proximal TCR signaling (e.g., phosphorylation of ZAP70) and downstream signaling pathways were not influenced in ESYT DKO cells.

Analysis of phenotypes of Jurkat T cells and human primary T cells demonstrated the physiological importance of E-Syts in regulation of the Ca2+-NFAT pathway and production of inflammatory cytokines. The major difference between T cells and HeLa cells was in the expression of E-Syt2S. While E-Syt2S was the predominant isoform in human T cells, HeLa cells predominantly express E-Syt2L. Mechanistically, both E-Syt2S and E-Syt2L were able to interact with STIM1, but E-Syt2L, with extra 48 amino acids in its N terminus, engages in additional intramolecular interactions. We surmise that increased intramolecular interaction due to this longer N terminus may restrict the function of E-Syt2L in regulation of SOCE. In support of this hypothesis, we observed increased interaction between ORAI1 and STIM1 upon store depletion in the presence of E-Syt2S but not E-Syt2L. Therefore, it is possible that E-Syt2S interacts with the regulatory region of STIM1 to facilitate its recruitment to the junctions and conformational change for ORAI1 interaction, whereas the E-Syt2L N terminus inhibits this function.

In summary, our current work describes the distinct role of E-Syts in establishment of the ER-PM junctions and SOCE in T cells. The E-Syt family is evolutionarily conserved through species from yeasts to humans, and its role in membrane tethering and lipid transfer has been emphasized so far. This study is significant because it reveals a cell type-specific role of E-Syts in Ca2+ signaling. The current study also uncovers the underlying molecular mechanism by demonstrating a direct interaction of E-Syt2S with STIM1. We showed that T cells have a unique strategy to modulate Ca2+ signaling, one of the fundamental signaling pathways for their activation. Our findings suggest that although the E-Syt family has a highly conserved function in establishment of the structural platform for the ER-PM junctions, it has also developed a strategy to endow more specific features to modulate signaling pathways, depending on the cell types.

Materials and methods

Plasmids and cells

GST-tagged truncated STIM1 fragments have been previously described19. The cDNAs encoding E-Syt1 and E-Syt2S were a kind gift from Dr. Pietro De Camilli (Yale University) and E-Syt2L was cloned from cDNA of Jurkat cells. The plasmid encoding MAPPER was a kind gift from Dr. Jen Liou (UT Southwestern Medical Center). FLAG- and GST-fused full-length E-Syts and their fragments were subcloned into pMSCV-CITE-eGFP-PGK-Puro and pGEX4T-1 vectors, respectively, using primers described in Supplementary Table 1. HeLa S3 and Jurkat E6-1T cell lines were purchased from American Type Culture Collection center (ATCC, Manassas, VA).

Affinity protein purification

HeLa cells stably expressing FLAG-tagged STIM1 were used for affinity protein purification using previously described methods19.

Generation of E-Syt knockdown or knockout cells

For generation of knockdown cells7, HEK293T cells were transfected with plasmid(s) encoding shRNA (Supplementary Table 1) and packaging vectors (pMD2.G and psPAX2, purchased from Addgene), using the calcium phosphate transfection method. Culture supernatants were harvested at 48 and 72 h post transfection and used for infection of 2.5 × 106 Jurkat T cells together with polybrene (8 µg/ml) using the spin-infection method. Cells were selected with puromycin (1 µg/ml) 48 h post infection. For knockout cells34, HEK293T cells were transfected with plasmid(s) encoding sgRNA (Supplementary Table 1) and packaging vectors (pMD2.G and psPAX2, Addgene) using calcium phosphate transfection method. Lentiviruses encoding Cas9 were generated using the same technique. Culture supernatants were harvested at 48 and 72 h post transfection and used for infection (50% of Cas9-encoding virus + 50% of sgRNA-encoding virus) of HeLa or Jurkat T 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 post infection.

Single-cell Ca2+ imaging and confocal microscopy

Single-cell Ca2+ imaging of T Cells loaded with 1 μM Fura 2-AM was performed as previously described35. For each experiment, 30–50 individual T cells were analyzed using OriginPro8.5 (Originlab) software. Acquisition and data analysis were performed using Slidebook (Intelligent Imaging Innovations, Inc.) and OriginPro8.5 (Originlab) software. For depletion of stores, cells were treated with 1 µM thapsigargin or ionomycin in Ca2+-free Ringer’s solution (unless indicated) for 5 min. TCR stimulation of Jurkat T cells for Ca2+ imaging was done using 10 µg/ml of soluble α-CD3 antibodies (OKT3 clone, NCI preclinical repository. Confocal laser scanning microscopy was performed using Fluoview FV10i Confocal Microscope (Olympus), images were captured with a 60 × oil objective. Images were processed for enhancement of brightness or contrast using Fluoview software.

Immunoprecipitation and immunoblotting

For immunoprecipitation between E-Syts and STIM1, cDNA encoding FLAG-tagged E-Syt1, E-Syt2L or E-Syt2S were transfected separately with 6 × His-tagged STIM1 into HEK293T cells. For immunoprecipitation between ORAI1 and STIM1 in the presence of E-Syt2L or E-Syt2S, cDNA encoding full-length FLAG-tagged ORAI1 and 6 × His-tagged STIM1 was transfected into HEK293T cells7,19. Separately HEK293T cells were also transfected with GFP-tagged E-Syt2L or E-Syt2S. Transfected cells (2 × 107) were lysed in lysis buffer (20 mM Tris–Cl, 2 mM EDTA, 135 mM NaCl, 10% (vol/vol) glycerol, 0.5% Igepal CA-630, Complete Protease Inhibitor Cocktail [Sigma-Aldrich], pH 7.5) and centrifuged at 100,000×g for 1 h before preclearing the supernatant with protein G-Sepharose. Pre-cleared lysates of HEK293T cells expressing E-Syt2L or E-Syt2S were mixed with those of HEK293T cells expressing ORAI1 and STIM1 (1:1 ratio v/v) and immunoprecipitated with anti-FLAG antibody-conjugated resin for 6 h. Immunoprecipitates were washed five times in lysis buffer and analyzed by immunoblotting. For immunoblot analyses, cells were lysed in RIPA buffer (10 mM Tris–Cl, 1% Triton X-100, 0.1% SDS, 140 mM NaCl, 1 mM EDTA, 0.1% sodium deoxycholate, and Complete Protease Inhibitor Cocktail [Sigma-Aldrich], pH 8.0) and centrifuged to remove debris. Samples were separated on 8–10% SDS-PAGE. Proteins were transferred to nitrocellulose membranes and subsequently analyzed by immunoblotting with relevant antibodies.

Purification of recombinant proteins from E. coli

Full-length and fragments (a.a. 1–249, 250–400, 324–448, 400–600, and 600–685) of STIM1 and fragments of E-Syt2 (2NL, 2NS, SMP, C2A/B, linker, and C2C) were subcloned into pGEX4T-1 plasmid7,19. GST fusion protein expressing transformants were grown in liquid cultures and induced with isopropyl-1-thio-β-d-galactopyranoside (IPTG, 0.2 mM) at 18 °C overnight. Subsequently, cells were harvested and resuspended in lysis buffer (50 mM NaH2PO4, 500 mM NaCl, 10% glycerol, pH 8.0) containing protease inhibitors and 0.5% Triton X-100. Lysates were sonicated, centrifuged to remove debris and incubated with glutathione sepharose 4B beads for 2 h. After washing 8 times with lysis buffer, the beads were stored in lysis buffer without Triton X-100 at − 20 °C.

GST pulldown analysis

cDNAs encoding E-Syt1-FLAG, E-Syt2L-FLAG, E-Syt2S-FLAG, 2NL-GFP (amino acids 1–80 of E-Syt2L), or 2NS-GFP (amino acids 1–30 of E-Syt2S or 49–80 of E-Syt2L) were transfected into HEK293T cells7,19. Transfected cells (2 × 107) were lysed in lysis buffer (20 mM Tris–Cl, 2 mM EDTA, 135 mM NaCl, 10% (vol/vol) glycerol, 0.5% Igepal CA-630, protease inhibitor mixture, pH 7.5) and centrifuged at 100,000×g for 1 h before preclearing with protein G-Sepharose. Lysates were incubated with 20 µg of GST or GST-tagged fragments of STIM1 for 6 h in binding buffer (0.5% Igepal CA-630, 20 mM Tris–HCl, 100 mM NaCl, 2 mM EDTA, 10% glycerol, protease inhibitors, pH 7.5). Precipitates were washed five times in lysis buffer and analyzed by immunoblotting.

Electron microscopy

After transient transfection with HRP-ER plasmid (a kind gift from Dr. Richard Lewis, Stanford), Jurkat T cells were fixed with 2% glutaraldehyde (Electron Microscopy Sciences) in 0.1 M Na cacodylate buffer (Electron Microscopy Sciences). Fixed cells were amplified with a TSA-biotin system (PerkinElmer) and ABC kit (Vector Laboratories) for 30 min each before being pre-reacted with 1 mg/ml of 3, 3′-Diaminobenzidine (Sigma-Aldrich) in Tris-buffered saline for 10 min and then reacted with DAB with 0.01% H2O2 for 30 min. After post-fixation with 1% OsO4 and en bloc stain with 1% uranyl acetate, cells were further processed as previously described5 before embedding in Embed 812 (Electron Microscopy Sciences). Cells located by light microscopy were punched out, and 50–90-nm sections were cut and mounted on formvar-coated grids and viewed with an 80-kV transmission electron microscope (T12 Quick CryoEM) equipped with a slow-scan cooled CCD camera (Gatan 2kX2k). The number or length of HRP-containing tubules located within 50 nm of the plasma membrane were measured with selected EM images where the nucleus and entire cell circumference were visible (taken at × 8,000–× 25,000 magnification) to restrict our analysis to sections cut through the middle rather than the edges of cells. Total junctional tubule length/section means percentage of summation of each length of HRP-tubules located within 50 nm of the plasma membrane per length of cell circumference.

Knockdown in human PBMCs

Mononuclear cells were prepared from buffy coats from healthy, unidentified adult donors, obtained under federal and state regulations from the UCLA CFAR Gene and Cellular Therapy Core Laboratory. Naïve CD4+ T cells were isolated using MagniSort Human CD4 naïve T cell enrichment kit (Thermo-Fisher). For T cell differentiation under non-polarizing conditions (ThN), cells were activated with 10 µg/ml of plate-coated anti-CD3 antibodies and soluble anti-CD28 antibodies in T cell medium (DMEM containing 20% FBS and 1% Pen-Strep) supplemented with 20 U/ml of IL-2. For Th17 cell differentiation, cells were activated with 10 µg/ml of plate-coated anti-CD3 antibodies and soluble anti-CD28 antibodies in T cell medium (described above) supplemented with 10 ng/ml of IL-1β and 10 ng/ml of IL-23. For generation of shRNA-encoding lentivirus, HEK293T cells were transfected with plasmid(s) encoding shRNA and packaging vectors (pMD2.G and psPAX2, Addgene) using the calcium phosphate transfection method. Culture supernatants were harvested at 48 and 72 h post transfection and used for infection of T cells together with polybrene (8 µg/ml) using the spin-infection method on days 1 and 2. Three hours post infection, virus-containing medium was replaced with fresh medium. On day 2, cells were detached from plate and on day 3, cells were selected with puromycin (2.5 µg/ml) for 18 h. The antibiotic was washed away and cells cultured for 2 more days in T cell medium (supplemented with IL-2 for ThN cells). On day 6, differentiated T cells were re-stimulated with 80 nM of PMA and 1 µM of ionomycin for 5 h for cytokine analysis. Brefeldin A (1 µg/ml) was added for the last 2 h of stimulation. Subsequently, cells were fixed/permeabilized and stained using the Transcription Factor Staining Buffer Set (eBioscience) for indicated cytokines. On day 6, around 1 million cells were harvested in TriZol (Thermo Fisher Scientific) for RNA preparation to check depletion efficiency by reverse transcription and quantitative PCR.

Real-time quantitative PCR

For real-time quantitative PCR7,34, total RNA from HeLa, Jurkat T cells, or primary cells was extracted using the Direct-zol RNA Prep Kit (Zymo Research) as per the manufacturer’s instructions. cDNA was synthesized from total RNA using qScript cDNA SuperMix Kit (Quantabio). Quantitative real-time PCR was performed using iTaq Universal SYBR Green Supermix (Bio-Rad) on an iCycler IQ5 system (Bio-Rad) using the primers described in Supplementary Table 1. Threshold cycles (CT) for all of the candidate genes were normalized to the CT values for GAPDH to obtain ΔCT and further normalized to the values obtained for control samples to obtain ΔΔCT. The specificity of primers was examined by melt-curve analysis and agarose gel electrophoresis of PCR products.

Statistical analyses

Statistical comparisons were performed using the Mann–Whitney U test to assess the significance between EAE-induced groups and two-tailed Student t test for other analyses36. Differences were considered significant when p values were < 0.05.