T Cell Receptor Mediated Calcium Entry Requires Alternatively Spliced Cav1.1 Channels

The process of calcium entry in T cells is a multichannel and multi-step process. We have studied the requirement for L-type calcium channels (Cav1.1) α1S subunits during calcium entry after TCR stimulation. High expression levels of Cav1.1 channels were detected in activated T cells. Sequencing and cloning of Cav1.1 channel cDNA from T cells revealed that a single splice variant is expressed. This variant lacks exon 29, which encodes the linker region adjacent to the voltage sensor, but contains five new N-terminal exons that substitute for exons 1 and 2, which are found in the Cav1.1 muscle counterpart. Overexpression studies using cloned T cell Cav1.1 in 293HEK cells (that lack TCR) suggest that the gating of these channels was altered. Knockdown of Cav1.1 channels in T cells abrogated calcium entry after TCR stimulation, suggesting that Cav1.1 channels are controlled by TCR signaling.


Introduction
Calcium ion entry across the plasma membrane is necessary for the initiation of T lymphocyte activation and proliferation following antigen encounter [1][2][3][4][5]. A typical calcium response occurs in two distinct steps. Initially, calcium is released from the intracellular stores, like the ER [6], which then triggers extracellular calcium entry through store-operated calcium (SOC) channels in the plasma membrane [7,8]. Activation of NFAT occurs upon elevation in cytosolic free calcium levels, which results in its retention in the nucleus and subsequent gene transcription [9,10]. This process is modulated by variations in the amplitude and/or duration of the calcium signal [11], which subsequently affect gene transcription and thus T cell activation and differentiation.
Apparently, a wide variety of calcium channels participate in calcium entry to T lymphocytes [12,13]. The most studied pathway for calcium entry in non-excitable cells is the CRAC (Calcium Release Activated Calcium Channel) pathway and its two key players, the stromal interaction molecule 1 (STIM1) and ORAI1 (also known as CRACM1 or TMEM142A) (reviewed in [14][15][16]). However, recent reports using deletion of ORAI or STIM proteins suggest that there are other pathways of calcium entry and that other plasma membrane calcium channels might be functionally involved [17][18][19].
Voltage gated calcium channels are known to mediate calcium entry in excitable cells [20]. The Ca v channel complex contains the pore-forming α1 subunit and the auxiliary subunits α2, δ, γ, and β subunits, which play a critical regulatory role [20]. A total of ten α 1 subunits have been identified and divided into 5 groups (L, P or Q, N, R, T) based on their properties [20]. The α 1 subunit compose (~190 kDa in molecular mass) the actual functional calcium selective pore. It is composed of four homologous domains (I-IV) each containing six transmembrane α-helices (S1-S6). The α 1 subunit also contains the voltage-sensing machinery (composed of the S4 helix from each domain). These channels are subject to rapid inactivation, which consist of two components: voltage-dependent (VDI) and calcium-dependent (CDI) [21]. The latter is mediated by the binding of calmodulin (CaM) to the channel [21].
Growing evidence suggests that these channels also contribute to calcium entry in nonexcitable cells. In fact, several studies have suggested the functional presence of Ca v channels in T lymphocytes (a non-excitable cell type), using pharmacological approaches [22][23][24][25][26]. We have examined the role of Ca v channels and associated proteins in T cells. We have shown that CD4 + T cells express α1 subunits of the Ca v 1 calcium channel family, but not Ca v 2 or Ca v 3 [27] and demonstrated the importance of the Ca v β3 and β4 regulatory subunits in TCR-triggered calcium response, NFAT nuclear translocation, and cytokine production [27][28][29]. More recently, we have demonstrated the importance of a scaffold protein AHNAK1 in regulating calcium signaling in peripheral CD4 + T cells. AHNAK1 is associated with the regulatory β2 subunit of Ca v channels and is required for normal expression of the Ca v 1.1 α1 subunit and calcium influx after TCR cross-linking in CD4 + T cells [30].
T cells from both β4 or AHNAK1 deficient mice (the latter was generated in our lab [31]) have reduced Ca v 1.1 channel membrane expression, deficient calcium entry and IL-2 production [27,30]. Altogether, these observations suggested that the Ca v 1.1 channels are functionally active in T cells in vivo.
An unresolved question is how are these T cell Ca v 1.1 channels gated? Several studies, including ours, have shown that in contrast to excitable cells, treatment of T cells with KCl, to induce an artificial depolarization, does not lead to calcium entry [26,27,32]. While Ca v 1.1 channels are likely functional in T cells, Ca v 1.2 channels are apparently inactivated by direct interaction with STIM1 [33,34]. In addition, another recent report suggests naïve CD44low T cells may have calcium currents dependent on Ca v 1.4 channels, possibly due to their inherent conductance in the hyper-polarized range (i.e. at/near T cell resting membrane potentials) [35]. Therefore unique subtleties for Ca v α1 subunit activation and gating appear to exist in T cells and possibly other non-excitable cell types.
With the functional effects of β3 and β4 Ca v accessory subunits and Ca v 1.4 channels already demonstrated in T cells [27,28,30,35], we hypothesized that Ca v 1.1 channels may be active in T cells, but with electrophysiological properties distinct from those expressed by excitable cells.
In this paper, we report the sequence of Ca v 1.1 channels in T cells and show that calcium entry after TCR stimulation is dependent on these channels. Together with prior studies [22][23][24][25][26]36], these data suggest a physiological function for Ca v 1.1 channels in T lymphocytes.

Expression of Ca v 1.1 in Activated T Cells
The expression of Ca v channels mRNA was previously described in primary T cells [27]. During primary stimulation of T cells, Ca v 1.1 channels mRNA is constitutively expressed [30]. Therefore, we examined the expression of Ca v 1.1 channel protein in primary and activated T cells (Fig 1A). Using immunoblot analysis, we found that Ca v 1.1 channel protein is detected in undifferentiated primary CD4 + T cells, and is up-regulated in differentiated effector Th1 (incubated with IL-12 plus anti-IL-4) and Th2 (incubated with IL-4 plus anti-IFNγ) cells ( Fig 1A). Ca v 1.1 channels expression level and protein molecular weight were compared to other cell types by immunoblotting using protein extracts from brain (as negative control) and C2C12 cells (a skeletal muscle cell line known to express Ca v 1.1 channels) ( Fig 1B). Moreover, we previously showed that Ca v 1.1 channels are expressed in the plasma membrane of CD4 + T cells [30]. We found that T cells express Ca v 1.1 channels of identical molecular weight (180KDa) and at comparable levels to the C2C12 skeletal muscle cell line. Altogether, these data demonstrate that Ca v 1.1 channels are constitutively expressed by CD4 + T lymphocytes. In particular, these channels are probably required for activation and differentiation of CD4 + T lymphocytes, where they are highly expressed.
T Cells Express an Alternatively Spliced Variant of Ca v 1.1 Channels with Deletion of Extracellular Linker Region, S3-S4 in Domain 4 We hypothesized that modifications might exist in T cells that affect this channel gating. By immunoblot analysis, we found that the molecular weight of Ca v 1.1 channels was identical between muscle cells and T cells (Fig 1B). This suggests that only minor modifications occurred that affect the function of these channels in T cells. Since very little is known about the molecular structure of the Ca v 1.1 channels in lymphocytes, we sequenced the full-length Ca v 1.1 mRNA from DO11.10 mouse and Jurkat human T cells. To enable this, polymerase chain reaction (PCR) and 5' RACE primers were designed, based on the muscle sequence (NM_000069) in order to isolate the Ca v 1.1 α1 subunit cDNA from T cells. cDNA sequence analysis showed that the mouse and human sequences of the T cell and muscle Ca v 1.1 channels are identical, except that a novel splice variant of the Ca v 1.1 channels was expressed in T cells, which contained new exons, A through E, which replaced the original exons 1 and 2 of the muscle cells, and lacked exon 29 (57bps) due to alternative splicing (Fig 2A and 2B). Expression of exon 29 could not be detected in DO11.10, Jurkat cell lines, or primary C57BL/6 mouse spleen T cell cDNAs (not shown). The excision of exon 29 causes the deletion of the S3-S4 linker in motif IV of the T cell Ca v 1.1 channels based on the published sequence ( Fig 2B).

Over Expression Analysis of T cell Ca v 1.1 Channels in Non-Excitable Cells
Similar Ca v 1.1 channel variants, lacking exon 29, were previously detected in muscle cells, where a modified voltage sensor activity was observed [37,38]. However, compared to this muscle variant, the T cell variant contains five additional alternatively spliced exons (A-E) in the N-terminus of the channel, as reported above.
To test the function of this Ca v 1.1 channel variant in non-excitable cells, we cloned the Ca v 1.1 channel cDNA from T cells and fused a green fluorescent protein to its N-terminus ( Fig  3A; Henceforth designated as T cell Ca v 1.1). This clone lacks exon 29 but contains exons A-E. Surprisingly, over-expression of these channels in Jurkat T cells was lethal, thus no GFP positive cells were detected. Instead, we expressed these channels in Human Embryonic Kidney cells (HEK), which proved to be more resilient to over-expression. GFP positive cells were detected within 24 hrs post transfection of T cell Ca v 1.1 channels and cell death was observed within 48 hrs. Resting intracellular calcium levels were then measured in these transfected HEK cells using an InCyt dual wavelength ratiometric fluorescence imaging system and Fura-2 AM. This system allows gating on GFP positive cells while measuring their calcium content simultaneously. At 24 hours post transfection, we found that high GFP expression is correlated with high intracellular  To test if indeed the transfected Ca v 1.1 channels conduct the calcium leading to the high intracellular calcium levels, we transfected T cell Ca v 1.1 channels containing restored exon 29. Indeed, in this case, transfected GFP positive cells show low intracellular calcium levels similar to control GFP negative cells in the dish (Fig 3C and 3D). This suggests that restoration of exon 29 probably restored the gating activity that closes the pore at rest, thereby preventing cell death. In addition, this observation suggests that Ca v 1.1 channels probably directly transmit calcium through the T cell membrane.
To further test the conductance of T cell Ca v 1.1 channels, we mutated the channels pore (Glutamates of the P loops -E278Q, E600Q, E1000Q, E1309Q). Such pore mutations, where negatively charged glutamic acids are replaced with neutral glutamines, are expected to abrogate the calcium conductance of any calcium channels. Cells expressing these mutant channels show low intracellular calcium levels compared to control GFP negative cells in the dish ( Fig  3C and 3D), consistent with the notion that these mutant channels do not constitutively conduct calcium. In conclusion, T cell Ca v 1.1 channels (lacking exon 29) are likely to conduct calcium themselves.
Lastly, the addition of the Ca v channel blocker-nifedipine to the medium of HEK cells transfected with T cell Ca v 1.1 channels, significantly reduced calcium flux through T cell Ca v 1.1 channels (Fig 4). Taken together, the evidence above demonstrates that calcium is directly transmitted through T cell Ca v 1.1 channels in non-excitable cells.
As in Fig 3, the results presented in the montages and graph are independent experimental repeats. Results are representative of 5 independent experiments.

TCR Induced Calcium Entry Is Dependent on Ca v 1.1 Channel
Previously, we reported the expression pattern for Ca v 1.1 channel mRNA, as well as other Ca v 1 calcium channel family members, in primary T cells [27]. These studies showed that during primary stimulation of T cells, Ca v 1.1 channel mRNA is constitutively expressed, even in naïve CD4 T cells [30]. Other studies in our lab have also shown that Ca v 1.1 channels are critical for TCR induced calcium entry into CD4 + T cells. First, we showed that loss of function of Ca v 1 β4 subunits results in reduced expression of Ca v 1.1 channel expression, deficient calcium influx after TCR cross-linking and subsequent impairment in CD4 + T cell activation [27]. Second, we showed that loss of function of AHNAK1, a scaffold protein, which binds β subunits in CD4 + T cells, results in reduced plasma membrane expression of Ca v 1.1 channels, reduced calcium influx after TCR cross-linking and subsequent low proliferation and IL-2 production [30].
To study the functional role of the Ca v 1.1 channel in T lymphocyte calcium entry, we used a lentivirus-based RNAi approach [39]. The DO11.10 T cell hybridoma was transduced with the appropriate viral preparations to knockdown Ca v 1.1 gene expression. Subsequently, GFP + cells were sorted and then tested for Ca v 1.1 channel expression. We successfully reduced the expression of the Ca v 1.1 gene using two independent siRNA sequences directed to different locations in the Ca v 1.1 mRNA (Fig 5A and 5B) and thus named 2184 and 3549. We performed three independent viral transductions to generate knockdown cells using 2184 and 3549 siR-NAs and similar inhibition of Ca v 1.1 channel expression was observed. Immunoblot analysis revealed a typical and significant reduction of Ca v 1.1 protein by 58% and 67% for 2184 and 3549 siRNAs, respectively (Fig 5A and 5B). Our negative control was a third sequence, named 1271, which yielded no inhibition of Ca v 1.1 gene expression, suggesting that the inhibition by 2184 and 3549 was sequence specific (Fig 5A and 5C). In addition, the RNAi effect was specific to Ca v 1.1 channel since the expression of other genes, including Ca v 1.2, Ca v 1.3 channels or STAT1α and STAT1β was unaffected (Fig 5B). Immunoblotting for actin was used to confirm that equal amounts of protein were loaded on the gel (Fig 5A).
We next investigated the role of Ca v 1.1 channels in the sustained calcium influx following specific T cell activation. Using Fura-2/AM and a ratiometric calcium probe, we detected, under physiological calcium concentration conditions (~1 mM in the extracellular media), a significant reduction in the TCR-induced calcium response in both 2184 and 3549 Ca v 1.1 channel knockdown cells relative to the control cells (Fig 5D and S2 Fig), which mounted a robust calcium response as shown by the absolute calcium concentration reached. Both the initial peak response and the plateau were inhibited (Fig 5D and S2 Fig). Interestingly, both 2184 and 3549 siRNA Ca v 1.1 channel knockdown cells showed normal release of calcium from intracellular stores in response to thapsigargin stimulation, an inhibitor of calcium-ATPase pumps that depletes intracellular stores [40][41][42], in a calcium free medium (S1A and S1B Fig, see w/o calcium). Similar results were seen when TCR stimulation was performed in calcium free medium (data not shown). Together, these observations, as well as the normal PLCγ1 activation following TCR cross-linking in Ca v 1.1 channel knockdown cells (S1C Fig), indicate that calcium release from intracellular stores is independent and temporally distinct from that of Ca v 1.1 channel activity.

Discussion
We have previously observed in vivo that T cells from mice lacking β4 or AHNAK1 show reduced plasma membrane expression of Ca v 1.1 channels and reduced calcium entry after activation [13,27,30]. In the present study, we show that, similar to their roles in other tissues, such as muscle, Ca v 1.1 channels are required for transmission of calcium into T cells after activation.
The present study provides new evidence for the role of Ca v 1.1 channels during calcium entry into stimulated T cells. These channels are detected in naïve T cells and their expression is highly upregulated in activated T cells (at day 4 after activation in vitro). This suggests that these channels are employed during the immune response of T cells. These variant Ca v 1.1 channels lack an extracellular loop located adjacent to the S4 voltage sensor of the 4 th domain, and contain five new alternatively spliced N-terminal exons. We show that reduction in Ca v 1.1 channel expression in T cells results in reduced calcium entry after TCR stimulation. Over expression of cloned T cell Ca v 1.1 channels in HEK 293 cells, shows that calcium influx through these channels occurs at the normal resting potential of these cells, probably due to exon 29 deletion.
Previous studies have provided significant insights into the outcome of exon 29 deletion on Ca v 1.1 channel function in muscle cells, where a modification of voltage-dependence was observed [37,38]. Calcium currents through these variants showed a -30-mV left-shifted voltage dependence of activation and a substantially increased open probability, giving rise to an increased current density [38]. Thus, aberrant myopathic muscle cells containing Ca v 1.1 channels lacking exon 29 have increased calcium conductance [37,38]. Compared to muscle cells, which have membrane potentials of around (-90mV), T cells or HEK 293 cells, maintain a more positive resting potential of~-60mV [43][44][45]. Thus, in T cells or 293 HEK cells, such Ca v 1.1 channel variants would be expected to show even higher calcium conductance at rest. This may provide an explanation for the constitutive calcium influx through these channels at the resting potential of such non-excitable cells. Other, as yet uncharacterized factors may, however, also contribute to the increased calcium flux observed upon over expression of these Ca v 1.1 channels in 293 HEK cells. In addition, further studies will determine the specific function of the alternatively spliced exons A through E, found, so far, only in T cells.
It is important to consider these results in the context of the large body of work on other calcium conductance pathways in lymphocytes, in particular the CRAC pathway. In lymphocytes, calcium entry through calcium release activated calcium (CRAC) channels is the primary mechanism that has been characterized to be responsible for the increase of free intracellular calcium, which is in turn necessary for cell activation and cytokine production. Two novel molecules, STIM1 and ORAI1, have been identified as important for the CRAC current in lymphocytes (reviewed in [14][15][16]). Both CD4 + and CD8 + T cells express all STIM (STIM1-2) and ORAI (ORAI1-3) family members [46]. However, CD4 + T cell deficient in ORAI1 show minimal effects on calcium entry or proliferation after TCR cross-linking [17,18]. A recent report suggests that TRPV1 is required for the activation of CD4 + T cells [12].
Our studies on the components of the Ca v complex identify an alternative path of calcium influx in response to TCR stimulation. When the Ca v 1.1 channel membrane expression is reduced by either deficiency in AHNAK1 or β4 subunits there is a decrease in calcium entry even though the CRAC channel pathway remains intact [27,30]. We therefore believe that multiple channels, including CRAC and Ca v , are probably involved in calcium entry into T cells. It is possible that these channels function sequentially or independently, and further studies are required to resolve this issue. One step towards this goal was shown recently when the interactions between STIM1 and Ca v 1.2 channels were described, leading to silencing of the latter [33,34].
Clearly it will be of great interest to further unravel the relative contributions of the CRAC/ TRP currents and the Ca v 1.1 current for calcium entry in T cells.

Activation and Differentiation of CD4 T Cells In Vitro
CD4 T cells from spleens and/or lymph nodes were isolated from 6-to 8-week-old C57BL/6 mice purchased from the National Cancer Institute (Frederick, MD). CD4 cells were isolated by immunomagnetic negative selection using Abs against CD8, NK1.1, and MHC class II, followed by incubation with anti-mouse and anti-rat Ig-coated magnetic beads (PerSeptive Biosystems). Cells were cultured in Bruff's medium containing 10% FCS and were stimulated for various days in the presence of plate-bound anti-CD3 and anti-CD28 Abs. Cells were then harvested for further analysis. To generate Th1 cells, 3.5 ng/ml of IL-12 (a gift from Wyeth Research) and 2 μg/ml of anti-IL-4 mAb 11B11 were added to the culture, whereas 1000 units/ ml of IL-4 and 2 μg/ml of anti-IFNγ mAb XMG1.2 were added to induce Th2 cells. 20 units/ml of recombinant huIL-2 (a gift from Biogen, Cambridge, MA) were added in both conditions. The Yale University institutional animal care and use committee has approved this study. All mice were cared for in accordance with protocols established by the institutional animal care and use committee at the Yale University animal facility.

Antibodies
The anti-Ca v 1.1 Ab used against the dihydropyridine receptor (DHPR) α1S (Santa Cruz Biotechnology, Santa Cruz, CA), was an affinity-purified goat polyclonal Ab, which interacts with Ca v 1.1 α1 subunit from mouse, human, and rat origin. For usage in immunoblot, anti-Ca v 1.1 Ab was biotinylated as previously described [27]. The following antibodies were also used in this study: Rabbit polyclonal Ab against calnexin was a gift from Dr. Cresswell P.; Rabbit polyclonal Ab against Ca v 1.2 and rabbit polyclonal Ab against Ca v 1.3 purchased from Alomone labs, Israel. Rabbit phospho-PLCγ1 (Tyr783), rabbit PLCγ1 Abs were purchased from Cell Signaling, Danvers, MA.

Protein Extracts and Immunoblot Analysis
CD4 T cells (6 to 10 x 10 6 ) were resuspended in 100 μl of hypotonic solution (10 mM HEPES pH 7.9, 10 mM KCl, and 0.1 mM EDTA) containing protease and phosphatase inhibitors (cocktail tablets, Roche Diagnostics, Indianapolis, IN), and incubated for 10 min on ice. Then NP40 was added at 1% and the cells were centrifuged at 5000 rpm for 5 min. The supernatant was then recovered.

Analysis of Intracellular Calcium Concentration
Population based assay: Levels of intracellular calcium were measured using the ratiometric calcium-binding dye Fura-2/AM as previously described [27]. Cells were incubated with 5 μM of Fura-2/AM (Molecular Probes, Eugene, OR) for 30 min at 37°C. Subsequently, T cells were washed then incubated with Ab to CD3 for 30 min on ice. 50 μg/ml of anti-hamster IgG (GAH) was added at the indicated time to stimulate the cells by Ab cross-linking. Fluorescence was monitored in ratio mode using a fluorometer (Polarstar Galaxy, BMG labtechnologies, Offenburg, Germany). Collected data were analyzed using Fluostar Galaxy Software (BMG labtechnologies, Offenburg, Germany). Experiments were performed at room temperature (~20°C). At the end of each experiment, cells were treated with 5 μM ionomycin in calciumcontaining medium, then with calcium-free medium supplemented with 5 mM of EGTA. Experimental 340/380 ratio were converted to [Ca 2+ ] i as previously described [27] according to the equation described by Tsien [47].
Single cell based assay: Intracellular calcium concentration was measured with a ratiometric dye Fura-2, as previously described [34]. Coverslips with cells grown on them were placed in DMEM Media containing 2 μM fura-2/acetoxymethylester for 45 min at 20°C. Cells were then washed with the same medium and fura-2 trapped inside cells was allowed to deesterify for 30 min at 20°C. Cells were then placed in a solution containing: (mM) 130 NaCl, 5 KCl, 1 MgCl2, 2 CaCl2, 30 glucose and 25 Hepes-NaOH, pH = 7.2 and Calcium measurements were performed on an InCyt dual-wavelength fluorescence imaging system (Intracellular Imaging Inc.). Briefly, Fura-2 was excited at 340 and 380 nm (OMEGA Optical, XF1093, 1094 respectively) Intracellular calcium concentrations are represented by the ratio of the fluorescence intensities excited at 340 nm and 380 nm from groups of single cells. Data are shown of traces from groups of individual cells (~10-15 cells each), and are representative of three or more independent experiments.

Lentivirus-Based RNA Interference
RNAi was performed as previously described [39]. Briefly, Lentivirus was produced by cotransfecting pLL3.7 and packaging vectors into 293FT cells. The supernatant was collected 48 h later. DO11.10 T cell hybridomas were supplemented with lentiviral particles that have been

MEPSSPQDEGLRKKQPKKPVPEILPRPPRALFCLTLENPLRKACISIVEWKPFETIILLTI FANCVALAVYLPMPEDDNNSLNLGLEKLEYFFLIVFSIEAAMKII
Mutations in the P-loops of Ca v 1.1 (E278Q, E600Q, E1000Q, E1309Q) were performed using Quick change Site Directed Mutagenesis kit (Stratagene). The primers used are as follows: