The Role of pp60c- src in the Regulation of Calcium Entry via Store-operated Calcium Channels*

In many cell types, G protein-coupled receptors stimulate a transient Ca2+ release from internal stores followed by a sustained, capacitative Ca2+ entry, which is mediated by store-operated channels (SOCs). Although it is clear that SOCs are activated by depletion of internal Ca2+stores, the mechanism for this process is not well understood. Previously, we have reported that inhibitors of tyrosine kinase activity block the bradykinin- and thapsigargin-stimulated Ca2+ entry in fibroblasts, suggesting that a tyrosine kinase activity may be involved in relaying the message from the empty internal Ca2+ stores to the plasma membrane Ca2+ channel (Lee, K.-M., Toscas, K., and Villereal, M. L. (1993) J. Biol. Chem. 268, 9945–9948). We also have demonstrated that bradykinin activates the nonreceptor tyrosine kinase c-src (Lee, K.-M., and Villereal, M. L. (1996) Am. J. Physiol. 270, C1430–C1437). We investigated whether c-src plays a role in the regulation of SOCs by monitoring capacitative Ca2+ entry in 3T3-like embryonic fibroblast lines derived from either wild type orsrc −/src −(Src−) transgenic mice. We report that Ca2+entry, following store depletion by either bradykinin or thapsigargin, is dramatically lower in Src− fibroblasts than in wild type fibroblasts. The level of capacitative Ca2+ entry in Src− cells is restored to nearly normal levels by transfecting Src− cells with chicken c-src. These data suggest that c-src may play a major role in the regulation of SOCs.

Most cultured fibroblasts respond to the peptide hormone bradykinin with a biphasic elevation of intracellular Ca 2ϩ concentration. The initial peak of the Ca 2ϩ response is due to Ca 2ϩ release from inositol trisphosphate-sensitive stores, whereas the longer duration, plateau phase is due to Ca 2ϩ influx from the extracellular medium. Our previous studies have demonstrated that the BK-stimulated 1 Ca 2ϩ influx is via a "capacita-tive" Ca 2ϩ pathway (1) similar to the one first described in pancreatic acinar cells by Putney (2). The physiological importance of capacitative Ca 2ϩ entry is suggested by the resulting primary immunodeficiency associated with defective T cell proliferation in patients whose lymphocytes have low capacitative Ca 2ϩ entry following T cell receptor stimulation (3,4). Putney hypothesized that the mechanism for opening this type of Ca 2ϩ channel involves a signal transduction process in which the "fill state" of the internal Ca 2ϩ stores is sensed and a message is sent to open plasma membrane Ca 2ϩ channels, if the stores are empty. Subsequent studies from a number of different laboratories have supported the basic concepts of this hypothesis, although the exact mechanism for regulation of these capacitative Ca 2ϩ channels, or store-operated channels (SOCs), is still an area of active investigation.
Based upon previous results from our laboratory, we have proposed that a tyrosine kinase activity is involved in relaying the message from the empty internal Ca 2ϩ stores to the plasma membrane Ca 2ϩ channel (5). This hypothesis is based on our observations that inhibitors of tyrosine kinase activity (such as genistein and tyrphostin) block the plateau phase of the BKinduced Ca 2ϩ response, whereas an inactive analog of genistein (diadzein) or inhibitors of serine/threonine kinases have no effect (5). The tyrosine kinase activity appears to lie downstream from the empty Ca 2ϩ pool because tyrosine kinase inhibitors also block the Ca 2ϩ entry stimulated by thapsigargin (5), an agent that directly empties the Ca 2ϩ pools by inhibiting the Ca 2ϩ -ATPase that pumps Ca 2ϩ into the internal stores.
To investigate the identity of the tyrosine kinase involved in regulating SOCs, we first had to identify the tyrosine kinases activated in response to BK stimulation. We recently reported (6) that in fibroblasts BK stimulates the tyrosine kinase activity of pp60 c-src (c-src). With this information in hand, we turned to the investigation of whether this tyrosine kinase is involved in the regulation of SOCs. To test for the involvement of c-src in the regulation of SOCs, we chose to utilize fibroblast lines that do not express the c-src protein tyrosine kinase as a result of gene disruption by homologous recombination (7). To investigate the regulation of SOCs by c-src, the intracellular Ca 2ϩ stores of either wt cells or Src Ϫ cells were emptied by various methods in a nominally Ca 2ϩ -free buffer, and the influx of Ca 2ϩ was monitored following addition of Ca 2ϩ to the external medium. The absence of c-src produced a significant decrease in the Ca 2ϩ influx in response to both bradykinin and thapsigargin. This could be reversed by expression of c-src in Src Ϫ cells.

EXPERIMENTAL PROCEDURES
Materials-The wt and src Ϫ /src Ϫ (Src Ϫ ) cell lines were derived by the spontaneous immortalization of mouse embryo fibroblasts prepared from either wild type mice or mice homozygous for a disruption in the c-src gene. These nonclonal cell populations were kindly provided by Philippe Soriano (Fred Hutchinson Cancer Center, Seattle, WA). The chicken c-src plasmid was kindly provided by David Shalloway (8). The plasmid carrying hygromycin resistance was obtained from Invitrogen (San Diego, CA). Avian-specific monoclonal anti-c-src antibodies were from Upstate Biotechnology, Inc. (Lake Placid, NY). Monoclonal antic-src antibodies (mAb 327) were obtained from Oncogene Sciences (Uniondale, NY). Horseradish peroxidase-labeled secondary antibodies were purchased from Promega (Madison, WI). Protein assay kits and ECL reagents were obtained from Pierce.
Cell Culture-Cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 2 mM glutamine, 50 units/ml penicillin, and 50 g/ml streptomycin in a 5% CO 2 incubator. For measurements of intracellular calcium concentration, cells were plated onto coverslips 2-5 days prior to experiments. For Western blotting, the cells were plated onto 100-mm dishes.
Western Blot Analysis-Cell lysates were prepared by treating cells with 0.5% SDS/8 M urea. The lysates were freeze-thawed three times to reduce viscosity. The protein concentration was determined by the BCA method (Pierce). The protein samples were mixed with equal volume of 2 ϫ SDS sample buffer (1 ϫ ϭ 62.5 mM Tris, pH 6.8, 1% SDS, 0.001% pyronin-Y, 10% glycerol, 5 mM 2-mercaptoethanol), boiled for 3 min, and subjected to SDS-polyacrylamide gel electrophoresis. Proteins were transferred to nitrocellulose membranes, and nonspecific binding sites were blocked by incubating the membranes in TBS (20 mM Tris, pH 7.4, 150 mM NaCl, 5 mM KCl) containing 3% bovine serum albumin and 0.2% Nonidet P-40 for 1 h at room temperature. Antibodies at a concentration of 1 g/ml were added, and the blot was incubated for 16 h at 4°C. The membrane was washed with TBS/0.2% Nonidet P-40 and incubated for 40 min with horseradish peroxidase-labeled secondary antibody. The blot was washed five times, and the immunostaining was detected by enhanced chemiluminescence.
Image Analysis-Cells were loaded with 5 M fura-2 acetoxymethyl ester in HEPES-buffered Hanks' balanced salt solution (HHBSS) ϩ 1 mg/ml bovine serum albumin ϩ 0.025% Pluronic F127 detergent for 30 min at room temperature and incubated without fura-2 acetoxymethyl ester in HHBSS for 30 min, and the intracellular [Ca 2ϩ ] was monitored as described previously (1). All traces represent the average response of 300 -400 cells on a nearly confluent coverslip. Although data shown here are from nearly confluent coverslips, measurements on lower density coverslips showed that there was no density dependence of the results. Cells perfused in "nominally" Ca 2ϩ -free HHBSS prepared as described previously (1) are still tightly adherent and can be vigorously perfused for extended time periods.
Establishment of Src Ϫ Fibroblasts Stably Expressing Chicken c-src-Fibroblasts deficient in c-src were cotransfected with chicken c-src and a plasmid carrying hygromycin resistance (pCEN4). After 2 days, cells were cultured in the presence of 250 g/ml hygromycin for 24 h. The selection pressure was removed, and cells were cultured in normal medium. A week later cells were reselected with 150 g/ml hygromycin for 2 days. Approximately 200 clones survived. All of the surviving clones were harvested, mixed together, and expanded to generate a heterogenous population of cells expressing chicken c-src. Cells were periodically put under selection pressure.

BK-stimulated Ca 2ϩ Entry in wt Fibroblasts and Src
Ϫ Fibroblasts-For the initial studies, cells were challenged with 100 nM BK in a nominally Ca 2ϩ -free external medium. The Ca 2ϩ peak under these conditions (Fig. 1) is due to the emptying of inositol trisphosphate-sensitive intracellular stores (1). The difference in the height of initial peaks in individual experiments generally is due to the fact that some fibroblasts in the field had delayed response to BK, thereby producing a wider, lower peak Ca 2ϩ response. Analysis of multiple coverslips (n ϭ 12) indicates that the areas under the initial BK-stimulated peaks were not significantly different in Src Ϫ and wt fibroblasts, indicating that the amounts of Ca 2ϩ mobilized in the two cell lines were approximately equal. This suggests that the loss of c-src does not disrupt the signaling between the BK receptor and the Ca 2ϩ stores.
When the release of internal Ca 2ϩ stores was complete and the [Ca 2ϩ ] i returned to base line, Ca 2ϩ was added externally in the continuous presence of BK. Under these conditions the observed rise in [Ca 2ϩ ] i is due to capacitative Ca 2ϩ influx; the initial slope of this Ca 2ϩ rise was used as a measure of Ca 2ϩ influx. As observed in Fig. 1, a dramatic Ca 2ϩ influx was observed in BKstimulated wild type cells. In cells that were incubated with Ca 2ϩ -free medium without BK, the influx of Ca 2ϩ initiated by Ca 2ϩ addition was only 5% of that observed in the presence of BK (data not shown). In contrast to results in wt cells, BK induced a significantly lower capacitative Ca 2ϩ entry in fibroblasts deficient in c-src, as judged by the initial slope of the Ca 2ϩ trace (28 Ϯ 10% of the slope observed in wt cells, n ϭ 12; this value was significantly different from the wt value based on a t test, p Ͻ 0.0001). Similar experiments utilizing Ba 2ϩ to monitor capacitative Ca 2ϩ entry also showed dramatic differences between wt and Src Ϫ cells (data not shown). Because Ba 2ϩ is not pumped by Ca 2ϩ -ATPases, the reduced Ca 2ϩ entry observed in Src Ϫ cells is not due to an increased Ca 2ϩ pump activity but is the result of decreased capacitative Ca 2ϩ influx. This observation suggests that c-src may be involved in the regulation of Ca 2ϩ entry in response to BK.
Although we have demonstrated that BK-induced Ca 2ϩ entry is significantly suppressed in Src Ϫ fibroblasts, there does appear to be capacitative Ca 2ϩ entry in Src Ϫ cells. This residual Ca 2ϩ influx may be due to either regulation by other tyrosine kinases or to regulation by a tyrosine kinase-independent mechanism. It has been shown that other c-src tyrosine kinase family members (e.g. c-fyn and c-yes) have some overlapping activities with c-src, which suggests the possibility that c-fyn or c-yes could also participate in the regulation of SOCs. This possibility is supported by experiments using genistein, a tyrosine kinase inhibitor that further suppressed the residual capacitative Ca 2ϩ entry in Src Ϫ cells (data not shown).
Thapsigargin-stimulated Ca 2ϩ Entry in wt Fibroblasts and Src Ϫ Fibroblasts-To exclude the possibility that c-src might regulate events in the BK signaling pathway that lie upstream of the intracellular Ca 2ϩ stores, we have also examined capacitative Ca 2ϩ influx in Src Ϫ cells stimulated with thapsigargin (TG), a potent Ca 2ϩ -ATPase inhibitor (Fig. 2). Following TGinduced Ca 2ϩ store depletion, Ca 2ϩ influx was initiated by the addition of Ca 2ϩ to the external media. Analysis of multiple wt and Src Ϫ coverslips (n ϭ 7) indicates that the total amount of Ca 2ϩ mobilized by thapsigargin is not statistically different in Src Ϫ and wt fibroblasts. As was the case for BK, thapsigargininduced Ca 2ϩ influx is significantly lower in fibroblasts deficient in c-src than that measured in wt fibroblasts (30 Ϯ 15% of the slope observed in wt cells, n ϭ 7; value significantly different from wt value based on a t test, p Ͻ 0.005), suggesting that the effect of c-src in the regulation of SOCs is downstream from the depletion of intracellular Ca 2ϩ stores. The fact that Ca 2ϩ entry is lower in Src Ϫ fibroblasts in response to both BK and TG could mean either that c-src is involved in regulation of Ca 2ϩ entry following store depletion or that the level of Ca 2ϩ entry was coincidentally lower in the independently derived Src Ϫ cell lines. To distinguish between these possibilities, we established Src Ϫ fibroblasts stably expressing chicken c-src, thus eliminating the possibility that the independently derived Src Ϫ and wt cells are different because they fortuitously express different levels of regulatory proteins or SOCs.
Stable Transfection of c-src into Src Ϫ Fibroblasts-To express c-src in Src Ϫ fibroblasts, we chose chicken c-src, because this src protein was shown to have little or no transforming activity when overexpressed in NIH3T3 fibroblasts, compared with v-src and mouse c-src (8 -11). We did not want to transform the cells, because transformation very often results in the overexpression or hyperphosphorylation of proteins involved in important signaling processes. The expression of c-src in Src Ϫ fibro-blasts posed some technical difficulties because these cells are difficult to transfect transiently. Therefore, we generated a population of cells that stably expresses the protein (see "Experimental Procedures"). The expression of chicken c-src was detected by immunostaining and also by Western blotting using the avian specific anti-c-src antibodies (Fig. 3). Western blots performed with mAb 327 demonstrated a 3-4-fold higher level of staining in the c-src transfected cells compared with wt cells (data not shown), although without knowing the relative specificity of mAb 327 for chicken versus mouse c-src, it is not clear whether this represents a 3-4-fold difference in protein expression.
BK-and Thapsigargin-stimulated Ca 2ϩ Entry in Src Ϫ and Fibroblasts Stably Expressing Chicken c-src-When capacitative Ca 2ϩ entry was monitored following Ca 2ϩ pool depletion with BK, it was observed that stable expression of chicken c-src significantly enhanced the capacitative Ca 2ϩ influx (Fig. 4). The expression of chicken c-src resulted in a 2.8 Ϯ 0.2-fold increase (n ϭ 7) in the initial slopes of the Ca 2ϩ traces at the addition of external Ca 2ϩ (value is significantly higher than Src Ϫ value based on a t test, p Ͻ 0.001). In addition, when intracellular Ca 2ϩ stores were depleted by treating cells with thapsigargin (1 M), the capacitative Ca 2ϩ influx was greatly enhanced in the cells expressing chicken c-src (2.7 Ϯ 0.6-fold increase, n ϭ 9; value is significantly higher than Src Ϫ value based on a t test, p Ͻ 0.01) (Fig. 5). Although the initial BK-stimulated release of Ca 2ϩ was frequently higher in the cells transfected with chicken c-src than in the Src Ϫ cells, this does not explain the higher Ca 2ϩ influx in the transfected cells. First, although measurements of the initial peak in chicken c-src transfected cells varied widely in height between individual coverslips, there was no correlation between first peak height and size of Ca 2ϩ influx. Second, the TG-stimulated peaks frequently had the opposite relationship (although this was not the case in Fig. 5, the first peak height was often higher in Src Ϫ fibroblasts than in the cells transfected with chicken c-src), and Ca 2ϩ fluxes were still higher than in Src Ϫ cells. Therefore, these observations strongly support our proposal that c-src is indeed involved in the regulation of calcium entry via store-operated calcium channels.
The BK-induced capacitative Ca 2ϩ entry in c-src transfected cells was 78 Ϯ 11% of the signal observed in the wild type cells. In the case of TG-induced Ca 2ϩ entry we observed 82 Ϯ 18% restoration. Immunostaining of hygromycin-selected cells showed that we could detect strong expression of avian c-src in 25% of the cells, intermediate expression in 30% of the cells, and little or no chicken c-src staining in 45% of the cells (data not shown). Thus, because of the selection method used, not all of the cells in this heterogenous population express chicken c-src. Cells were selected based on their expression of an antibiotic resistance gene marker; thus some percentage of cells express the antibiotic resistance gene and no c-src or low levels of c-src.
Even though these data support the hypothesis that c-src may play an important role in the regulation of Ca 2ϩ entry via store-operated calcium channels, we cannot exclude the possibility that c-src might be acting via a longer term mechanism, such as by controlling the level of expression of SOCs. However, our previous data (5) indicate that tyrosine kinase inhibitors reduce capacitative Ca 2ϩ influx within a matter of minutes of their addition, a finding that supports the involvement of tyrosine kinases in a short term, regulatory role. However, based on the observation that the SH2 and SH3 domains but not the kinase activity of c-src are important for the effect of c-src on the rate of cell spreading on fibronectin (12), we cannot rule out the possibility that the c-src effect we observe might be independent of the src kinase domain. In the future, we plan to express c-src constructs with mutations in the kinase, SH2, and SH3 domains in Src Ϫ cells to determine in more detail the importance of these domains in the regulation of Ca 2ϩ entry.
Although the exact mechanism for c-src regulation of capacitative Ca 2ϩ entry is not known, it appears that the decrease in Ca 2ϩ entry in Src Ϫ cells is not the result of a reduction in the membrane potential and therefore a decrease in the driving force for Ca 2ϩ entry. Hyperpolarization of Src Ϫ and wild type cells, by the addition of valinomycin in a 6 mM K ϩ medium, did not alter the fact that wild type cells had a dramatically higher level of Ca 2ϩ entry than those measured in Src Ϫ cells (data not shown).
There is considerable evidence that protein phosphorylation can regulate ion channels. Ion channels are known to be regulated and directly phosphorylated by a number of serine/threonine kinases, such as protein kinase C, calmodulin-dependent kinase, and cyclic AMP-dependent protein kinase (13). Some recent studies suggest that tyrosine phosphorylation can also regulate channel activities. For example, tyrosine phosphorylation regulates the channel activities of N-methyl-D-aspartate receptor (14), a brain-and heart-specific delayed rectifier-type potassium channel (15), and a voltage-dependent n-type K ϩ channel (Kv1.3) (16). A recently cloned focal adhesion kinase family member, PYK2, directly tyrosine phosphorylates a potassium channel (Kv1.2) and inhibits currents elicited by phorbol myristyl acetate (17). In regard to the regulation of SOCs, several studies suggest that serine and threonine phosphorylation may inhibit capacitative Ca 2ϩ entry (18,19), and a number of studies utilizing tyrosine kinase inhibitors (5, 20 -23) support our initial proposal (5) for a role of tyrosine kinases in the activation of SOCs. Further studies are required to determine whether the effect of c-src on the regulation of capacitative Ca 2ϩ entry is a direct effect on the store-operated Ca 2ϩ channel, as seen for hKv1.5 and N-methyl-D-aspartate channels, or an indirect effect via other c-src substrates.