Novel calcium-dependent mechanisms of NF-B activation

T cell activation following antigen binding to the T cell receptor (TCR) involves the mobilization of intracellular calcium (Ca 2+ ) to activate the key transcription factors NFAT and NF-κB. The mechanism of NFAT activation by Ca 2+ has been determined; however, the role of Ca 2+ in controlling NF-κB signaling is poorly understood and the source of Ca 2+ required for NF-κB activation is unknown. We demonstrate that TCRbut not TNFinduced NF-κB signaling upstream of IκB kinase (IKK) activation absolutely requires the influx of extracellular Ca 2+ via STIM1-dependent CRAC/Orai channels. We further show that Ca 2+ influx controls phosphorylation of the NF-κB protein p65 on Ser536 and that this posttranslational modification controls its nuclear localization and transcriptional activation. Notably our data reveal that this role for Ca 2+ is entirely separate from its upstream control of IκBα degradation, thereby identifying a novel Ca 2+ dependent distal step in TCR-induced NF-κB activation. Finally, we demonstrate that this control of distal signaling occurs via Ca 2+ -dependent PKC-mediated phosphorylation of p65. Thus, we establish the source of Ca 2+ required for TCR induced NF-κB activation and we define a new distal Ca 2+ -dependent checkpoint in TCR-induced NF-B signaling that has broad implications for the control of immune cell development and T cell functional specificity. INTRODUCTION Activation of T cells following antigen binding to the T cell antigen receptor (TCR) induces diverse lineage and fate-specific pro-inflammatory and immune-modulatory responses. Central to these responses is the induction of quantitatively distinct intracellular Ca 2+ signals and their selective activation of the key transcription factors NFAT and NF-κB (1-6). The mechanism by which Ca 2+ controls NFAT activation in lymphocytes is well http://www.jbc.org/cgi/doi/10.1074/jbc.M115.713008 The latest version is at JBC Papers in Press. Published on January 29, 2016 as Manuscript M115.713008 Copyright 2016 by The American Society for Biochemistry and Molecular Biology, Inc. at D R E X E L U N IV L IB R A R IE S on Feruary 6, 2016 hp://w w w .jb.org/ D ow nladed from Novel calcium-dependent mechanisms of NF-B activation 2 established (7). In contrast, although Ca 2+ has been implicated in TCR-induced NF-κB signaling (8-10), how Ca 2+ regulates NF-B activity is largely unexplored and represents a significant gap in our understanding of transcriptional control of T cell development, activation and functional specificity. In resting T cells, classical NF-κB consists of hetero-dimers of p50/p65 or p50/c-Rel that are retained inactive in the cytosol by members of the inhibitory family of IB proteins (11,12). Following TCR engagement, IB kinase (IKK) mediated phosphorylation triggers the ubiquitination and proteasomal degradation of IκBα, releasing p50/p65 and p50/c-Rel, which localize to the nucleus to initiate transcription of crucial immune-regulatory, pro-inflammatory and pro-proliferative genes (13-30). Although TCRmediated Ca 2+ mobilization has been implicated in proximal steps of NF-B activation (8-10), the precise mechanisms and source of Ca 2+ that regulate nuclear localization and transcriptional activation of NF-B are poorly defined. It is well established that TCR signaling induces inositol 1,4,5-trisphosphate (IP3)-mediated depletion of Ca 2+ from the endoplasmic reticulum (ER). A resulting Ca 2+ dissociation from the ER membrane protein Stromal Interaction Molecule 1 (STIM1) triggers its oligomerization and re-localization to ER membrane domains juxtaposed to the plasma membrane (31-33), where STIM1 physically gates Orai (also known as Ca 2+ release-activated Ca 2+ (CRAC)) channels allowing extracellular Ca 2+ to enter the cell (34,35). However, it is not known whether Ca 2+ control of TCR-induced NF-κB signaling requires STIM1 and Orai1-mediated Ca 2+ influx or whether the initial release of Ca 2+ from the ER is sufficient for classical NF-κB activation. In this study, we sought to determine both the source and mechanism of Ca 2+ control of antigen receptor induced NF-κB activation in T cells. We show that influx of extracellular Ca 2+ via STIM1 and Orai is critical for TCRbut not TNF-induced IBα degradation and NF-κB activation. Importantly, we also demonstrate that Ca 2+ dependent PKC-mediated phosphorylation of p65 critically regulates its nuclear localization and transcriptional activation following TCR engagement. Thus, our findings define important new proximal and distal Ca 2+ -dependent checkpoints in TCR-induced NF-κB signaling that have broad implications for the control of immune cell development and functional specificity. METHODS Cells and Cell Culture Primary human T cells were obtained from the University of Pennsylvania Immunology Core facility. Jurkat T cells were from ATCC and Jurkat T cells stably expressing E106A Orai1 were a gift of Dr. Jonathan Soboloff (Temple University, Philadelphia, PA). All cells were cultured in RPMI 1640 medium (BioWhittaker, Walkersville, MD) supplemented with 10% heat inactivated fetal bovine serum (Hyclone, Thermoscientific, Logan, Utah), 2 mM Lglutamine, penicillin (50 U/ml), and streptomycin (50 U/ml). Antibodies and Reagents Antibodies recognizing p65-phosho-Ser536 (#3033S) and IκB (#4814S) were purchased from Cell Signaling Technology (Danvers, MA); anti-p65 (#372R/G) was purchased from Santa Cruz Biotechnology (Dallas, TX); anti-tubulin (#T1568) was from Sigma-Aldrich (St. Louis, MO); anti-STIM1 (#610954) was from BD Transduction Laboratories (Franklin Lakes, NJ); anti-human TCR (c305 clone) was a gift from Dr. Gary Koretzky (Cornell University, New York, NY) and anti-CD28 was from Invitrogen. Protein A horseradish peroxidase–conjugated antibody (#18160) was from Millipore (Danvers, MA). Alexa fluor 488 goat anti-rabbit (A11008), Alexa fluor 546 goat anti-rabbit (A11010) used for immunofluorescence were obtained from Invitrogen (Waltham, MA). Recombinant human TNF-α was purchased from R&D Systems (Minneapolis, MN), PMA and Ionomycin were from Sigma-Aldrich, and the Luciferase Reporter Assay System was obtained from Promega (Fitchburg, WI). Plasmids and Transfections A cDNA construct expressing full-length p65 N-terminally tagged with EGFP was obtained from Addgene (Cambridge, MA). Mutant p65 constructs were generated using a site directed mutagenesis kit (Stratagene, La Jolla, CA) to convert serine 536 to alanine (A) or aspartic acid (D). Short hairpin (sh) STIM1 suppression and rescue constructs were generated in the lab of Dr. Dan Billadeau (Mayo Clinic, Rochester, MN) as was the EGFP-shPKC and EGFP-pCMS2 (control vector). For transfection, Jurkat T cells were suspended at 20 million cells per mL in RPMI 1640 and 10 million cells were electroporated with 10 g at D R E X E L U N IV L IB R A R IE S on Feruary 6, 2016 hp://w w w .jb.org/ D ow nladed from Novel calcium-dependent mechanisms of NF-B activation 3 of DNA (for overexpression or mutant expression) or 40 g of DNA (for suppression assays) at 315 V for 10 ms using a BTX ECM 830 electroporator (Harvard Apparatus, Holliston, MA). STIM1 and PKC suppression assays were performed 48 hours post-transfection and EGFP-p65 and p65 mutant expression assays were performed 16-24 hours post transfection. Immunoblotting Cells were harvested and lysed using NP-40 lysis buffer consisting of 50 mM TrisHCl (pH 7.5), 20mM EDTA, 1% NP-40 and complete inhibitors (1mM Sodium Orthovanadate, 1mM PMSF, 10 g/ml Leupeptin, 5 g/ml Aprotinin). Protein concentrations in cell lysates were determined using the Bio-Rad reagent (BioRad Laboratories, Hercules, CA) and quantified in a Cary 50 Bio UV-visible Spectrophotometer. Proteins were resolved by SDS–polyacrylamide gel electrophoresis (4-15%, Bio-Rad, Hercules, CA) then transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Billerica, MA). Membranes were probed with respective primary anti-human antibodies and then incubated with Protein A HRP secondary antibodies. Blots were developed with enhanced chemiluminescence using pierce ECL Western Blotting Substrate (Pierce, Rockford, IL). All immunoblots presented are from a single experiment representative of at least three independent experiments. Luciferase Reporter analysis Luciferase based transcriptional analysis was performed on Jurkat T cells transfected with 2g of total DNA (PBXII κB firefly luciferase (FFL) and pRL TK Renilla luciferase (RL) in a 20:1 ratio) per transfection (5×10 6 cells in 500ul medium), using a square-wave BTX electroporator at 315V for 10 msec. Twentyfour hours after transfection, cells were treated with PMA (200nM), PMA (200nM) and ionomycin (1 g/ml), anti TCR ) and anti-CD28 (1:50), or TNF (10 ng/ml) for 4 hours. Cells were then lysed in Passive Lysis Buffer (Promega, Fitchburg, WI) and luciferase activity measured using a Luminoscan 96-well automated luminometer (Thermo Labsystems, Franklin, MA). Firefly/Renilla luciferase ratios were calculated using Ascent software (Thermo LabSystems), and the mean ratio from at least three independent experiments (3-4 replicates per experiment) for each condition was compared. Quantitative Real-Time PCR To quantify IB expression, cDNA was synthesized from RNA isolated (RNA MiniPrep, Zymo Research) from PMA or PMA and ionomycin stimulated cells with a high capacity cDNA reverse transcription kit (Thermo Scientific, Waltham, MA). cDNA was amplified with IB (forward: 5’-CCGAGAC TTTCGAGGA AATACC-3’, reverse: 5’-ACGTG TGGCCATTGTAGTT-3’) and GAPDH specific primers (forward: 5’-CACCGATTTATAGAAACC GGGGGCG-3’, reverse: 5’-AAACCGCCCCCGG TTTCTATAAAT C-3’) on a 7500 Fast Real-Time PCR System (Applied Biosciences, Warrington, United Kingdom) using Power SYBR Green PCR Master Mix (Applied Biosciences). Ct values were obtained in triplicate for each target and were analyzed with the instrument software v1.3.1 (Applied Biosystems, Warrington, United Kingdom). Microarray Analysis RNA was isolated using an RNeasy Plus kit (Qiagen, Hilden, Germany). Biotin labeled complementary RNA (cRNA) was generated using the Illumina TotalPrep RNA amplification kit and a Bioanalyzer (Agilent Technologies, Wilmington, DE) was used to assess total RNA and cRNA quality. Illumina HumanHT12 version-4 expression beadchips were hybridized with cRNA from two biological replicates per condition and scanned on an Illumina BeadStation 500GX. Scanned images were converted to raw expression values using GenomeStudio v1.8 software (Illumina). Data analysis was performed using the statistical computing environment, R (v3.2.3), the Bioconductor suite of packages for R, and RStudio (v0.98). Raw data were background subtracted, variance stabilized, and normalized by robust spline normalization using the Lumi package (36). Differentially expressed genes were identified by linear modeling and Bayesian statistics using the Limma package (37,38). Probe sets that were differentially regulated (≥1.5 fold change between all treatments, FDR ≦ 5% after controlling for multiple testing using the Benjamini-Hochberg method (39,40) were used for heatmap generation in R. Clusters of co-regulated genes were identified by Pearson correlation using the hclust function of the stats package in R. Differentially expressed NF-B dependent genes were identified using a list of validated and putative NF-B target genes curated at D R E X E L U N IV L IB R A R IE S on Feruary 6, 2016 hp://w w w .jb.org/ D ow nladed from Novel calcium-dependent mechanisms of NF-B activation 4 by Dr. Thomas Gilmore’s Lab at Boston University (http://bioinfo.lifl.fr/NF-KB/). All microarray data have been deposited on the Gene Expression Omnibus (GEO) database for public access. Chromatin Immunoprecipitation Jurkat T cells (20 x 10 6 ) transfected with EGFP-shPKC and control EGFP-pCMS2 vector (48 hours) were stimulated with PMA (200 nM) and ionomycin (1 M) for 30 minutes at 37°C. Chromatin was prepared using a Covaris truChIP chromatin shearing kit (Covaris Inc., Woburn, MA). Briefly, cells were fixed in 1% methanol-free formaldehyde for 5 minutes at RT, and then fixation was quenched with 0.125 M glycine at RT for 5 min. Cells were then washed twice with cold PBS and then were lysed at 4°C with rocking for 10 minutes. Nuclei were then washed and transferred to an AFA milliTUBE for ultra-sonication. Samples were sheared using a Covaris S220 Focused-ultrasonicator for 1500 seconds in a 6°C bath at a duty cycle of 5%, an intensity of 4, a peak incident power of 140 Watts, at 200 cycles per burst. p65 was precipitated from sheared chromatin (200-1000 bp) with anti-p65 (5 g, Santa Cruz, Rabbit #372X) or normal rabbit IgG (5 g, Cell Signaling, #2729) for 12-16 hours at 4°C. Immunoprecipitated chromatin was then incubated with protein G Dynabeads (Life Technologies, Frederick, MD) for 2 hours at 4°C and chromatin was eluted (50 mM Tris pH 8.0 and 10 mM EDTA) at 65°C on a thermomixer (1200 rpm) for 30 minutes. Crosslinking was reversed by incubating recovered chromatin at 65°C for 12 hours, followed by incubation with RNAse A for 2 h at 37°C, and proteinase K for 30 min at 55°C. DNA was then purified using a ChIP DNA Clean and Concentrator kit (Zymo Research, Irvine, CA) and qPCR was performed to quantify p65 binding to IB, CXCL8, and TNF promotors using IB (forward: 5’-TTGGGATCTCAGCAGCCGAC-3’ and reverse: 5’-GCCACTAGGGTCACGGACAG3’) CXCL8 (forward: 5’-CAGGTTTGCCCTGAG GGG ATG-3’ and reverse: 5’-GGAGTGCTCCG GTGG CTTTT-3’) and TNF (forward: 5’CCCGCGATGGAG AAGAAACC-3’ and reverse: 5’-GTCCTTGCTGAGGGAGCGTC-3’) specific primers. Quantitation of p65 Nuclear Translocation Jurkat T cells transfected with EGFP-shPKC or EGFPpCMS2 (48 hours) or untransfected cells suspended in medium containing 2 mM Ca 2+ or Ca 2+ free equivalent solution were adhered to Cell-Tak treated coverslips for 10-15 min and stimulated at 37 ̊C as indicated. For PKC suppression experiments, cells were stimulated in the presence of 2 mM Ca 2+ . At indicated times, cells were fixed in formaldehyde (3.7%) for 30 minutes, permeabilized with 0.2 % Triton-X-100 for 15 minutes, and blocked overnight in 2% BSA at 4 ̊C. Fixed and blocked cells were incubated with rabbit anti-p65 primary antibody (Santa Cruz Biotechnology, Cat #372, 1 g/ml) for 1 hour at 37 ̊C or overnight at 4 ̊C degrees, washed 3 x 5 minutes in 1% BSA in PBS, and incubated with Alexa 488 or 546 goat anti-rabbit secondary antibody (4 g/ml) for 1 hour at 37 ̊C. Nuclei were then labeled with Hoechst 33342 (Life Technologies, Cat #H3570, 4 g/ml), washed 3 x 5 minutes in 1% BSA in PBS, and mounted in Fluoromount (Fisher). Images of p65 localization were obtained with a Yokagawa spinning disk confocal system (Tokyo, Japan) mounted on a Leica DMI4000 microscope (Leica Microsystems, Wetzlar, Germany) and imaging parameters were optimized independently for each channel to maintain fluorescence within the linear range while maximizing intensity resolution. Images of p65 and Hoechst were overlaid and cytoplasmic/nuclear p65 localization was determined using MetaMorph (Molecular Devices, Downingtown, PA). Average nuclear and cytoplasmic p65 fluorescence intensities were quantified within cytoplasmic and nuclear compartments and intensity ratios were determined for each cell. Real Time Localization of WT and Mutant p65 Jurkat T cells expressing WT and p65 Ser536 mutants (16-24 hours) were adhered to Cell-Tak (BD Biosciences, Franklin Lakes, NJ) coated coverslips and maintained in culture medium (RPMI 1640, 10% FBS, 1% glutaMAX) in a temperature and CO2 controlled chamber for 1 hour during imaging. GFP-WT and GFP-p65 mutants were visualized every 10 seconds post stimulation with PMA (200 nM) with ionomycin (1 M) and/or PMA (200 nM) with and without the delayed addition of Ionomycin (1 M). Calcium Imaging Jurkat T cells (3 million cells/mL) were loaded with 3 M fura-2 acetoxymethyl ester (Molecular Probes, Eugene, at D R E X E L U N IV L IB R A R IE S on Feruary 6, 2016 hp://w w w .jb.org/ D ow nladed from Novel calcium-dependent mechanisms of NF-B activation 5 OR) in external solution containing 145mM NaCl, 4.5mM KCl, 2mM CaCl2, 1mM MgCl2, 10mM glucose, 10mM HEPES, 2mM glutamine, and 2% fetal bovine serum (Hyclone, ThermoScientific, Logan, Utah) for 10 minutes at 25°C. Cells were adhered to coverslips coated with Cell-Tak (BD Biosciences, Franklin Lakes, NJ), mounted on the stage of a Leica DMI6000 microscope configured with a Photometrics Evolve 512 Camera (Tucson, AZ) using an Olympus 40 x oil objective (Shinjuku, Tokyo, Japan), and images were acquired with MetaFluor software (Molecular Devices, Downingtown, PA). During imaging, cells were perfused with Ca 2+ free bath solution before activating with PMA (200 nM) and ionomycin (1 M), thapsigargin (1 M), anti-TCR (0.5 g/ml) and CD28 (1:50) antibodies, or TNF (10 ng/ml) to evaluate stimulus-dependent Ca 2+ release from the ER. The cells were then perfused with bath solution containing 2 mM Ca 2+ to assess Ca 2+ entry via activated Orai channels. In some experiments, cells were pretreated for 15 minutes with the Orai-1 inhibitor Synta66 (50 M, Aobious, Gloucester, MA) prior to stimulation. Ca 2+ mobilization was analyzed by plotting the emission ratio of 340/380nm excitation for each cell. Each plot is the averaged ratio from at least 30 cells. Statistical Analysis Significance for all statistical tests was determined at p-values <0.05 and is shown as * for p <0.05, ** for p <0.01, and *** for p < 0.001 in all figures. Average firefly/renilla luciferase ratios were calculated from 3-4 independent experiments and analyzed using twotailed Welch’s t-test. Western blot protein intensities were quantified using ImageJ (http://imagej.nih.gov/ij/) and average protein intensity values were compared using two-tailed Welch’s t-test. p65 nuclear to cytoplasmic fluorescence intensity ratios were assessed for normality using probability plots and the Kolmogorov-Smirnov test for normality. Normal distributions were compared using two-tailed Welch’s t-test and non-normal data were compared using Wilcoxon rank sum test. Quantitative PCR RQ values and % input values were compared using two-tailed Welch’s t-test. RESULTS Extracellular Ca 2+ is required for TCR-induced NF-κB signaling Ca 2+ regulates proximal TCR signaling upstream of IKK activation (8-10); however, the precise function of Ca 2+ , and the source of Ca 2+ required for NF-κB activation are unknown. To address these questions we first asked whether the initial release of Ca 2+ from ER stores was sufficient, or whether sustained influx of extracellular Ca 2+ is required for NF-κB activation in T cells. To distinguish between these pools of Ca 2+ , we activated T cells in the presence or absence of extracellular Ca 2+ with either anti-CD3 and antiCD28 (3/28) to co-engage the TCR and CD28, or with the DAG analog, PMA, together with Ionomycin (P/I), which activate PKCθ and release ER-stored Ca 2+ respectively, and we compared these responses to that induced by the proinflammatory cytokine TNF. In Ca 2+ -free medium, 3/28 and P/I, but not TNF, induced a transient rise in cytoplasmic Ca 2+ concentration due to release from the ER. Reintroduction of extracellular Ca 2+ led to a sustained secondary increase in Ca 2+ concentration via entry through activated Orai1/CRAC channels in 3/28 and P/I stimulated cells (35) (Fig.1A). Thus, stimulating cells in the absence of extracellular Ca 2+ allows us to specifically determine whether release from ER stores alone is sufficient for NF-κB signal activation. As shown in Figure 1B (top three panels), all three stimuli induced the expected degradation and re-synthesis of IBα in Jurkat T cells consistent with activation of the IKK complex and the classical NF-κB pathway. In contrast, neither 3/28 nor P/I induced IκBα degradation in Ca 2+ -free medium, whereas IBα degradation and resynthesis in response to TNF remained intact (Fig.1B, bottom three panels). Consistent with the effects on IκBα degradation, 3/28-stimulated NF-κB transcriptional activity was completely inhibited and P/I-induced transcriptional activation was significantly reduced in Ca 2+ -free medium (Fig. 1C). In contrast, TNF-induced NF-κB reporter activity was unaltered in the presence or absence of extracellular Ca 2+ (Fig.1C). Similar regulation of IκBα expression was observed in primary human CD4+ T cells (Fig.1D). Collectively, these findings reveal that transient release of Ca 2+ from ER stores at D R E X E L U N IV L IB R A R IE S on Feruary 6, 2016 hp://w w w .jb.org/ D ow nladed from Novel calcium-dependent mechanisms of NF-B activation 6 is not sufficient, and that extracellular Ca 2+ is required for TCR-induced NF-κB activation. TCR-induced NF-B activation requires STIM1 and Orai1. The extracellular Ca 2+ requirement for TCRinduced NF-κB activation implies a crucial role for STIM1-operated Orai1 channel-mediated Ca 2+ influx. To explore this, we expressed a STIM1 shRNA construct or a bi-cistronic variant for concomitant re-expression of shRNA resistant STIM1 to normal levels in Jurkat T cells (Fig.2A). STIM1 suppression inhibited 3/28 and P/I induced extracellular Ca 2+ influx and this was rescued by reexpression of STIM1 (Fig.2B). Consistent with the lack of TCR-induced NF-κB signaling in Ca 2+ -free medium (Fig.1, B and C), STIM1 suppression prevented 3/28 and P/I induced IκBα degradation in T cells (Fig.2C, left versus middle panels). In contrast, IκBα degradation and re-expression was normal in STIM1 rescued cells confirming that the inhibition was due to loss of STIM1 (Fig.2D, right panels). Furthermore, both 3/28and P/I-induced NF-κB transcriptional activity was reduced in STIM1 suppressed cells and this was again rescued by concomitant STIM1 re-expression (Fig. 2C). Notably, TNF-induced IκBα degradation and NFκB transcriptional activity were unaffected by suppressing STIM1 (Fig.2, C and D). To confirm the role of Orai in Ca 2+ dependent NF-κB activation, we examined the consequence of inhibiting Orai-mediated Ca 2+ influx with the Orai inhibitor Synta66 (Fig. 3A) and by expressing a mutant Orai1 (glutamic acid at position 106 mutated to alanine, E106A) that exerts a dominant negative effect on the Ca 2+ permeability of endogenous Orai channels (Fig. 3C). In the presence of Synta66, the SERCA inhibitor thapsigargin triggered Ca 2+ release from the ER, evident as a small transient increase in cytoplasmic Ca 2+ (in Ca 2+ free medium), but the subsequent sustained increase in cytoplasmic Ca 2+ observed in untreated cells (Fig. 3A, top panel) in the presence of Synta66 (Fig. 3A, bottom panel). Consistent with the effects of STIM1 suppression (Fig. 2), Synta66 (Fig. 3B) inhibited 3/28but not TNF-induced IB degradation. A similar block in stimulus induced Ca 2+ entry and IB degradation was observed in permeation defective Orai1-E106A cells (Fig. 3, C and D) Taken together these results reveal an obligate role for STIM1-operated Orai1-mediated Ca 2+ entry in TCRbut not TNF-induced IκBα degradation and