Protein-tyrosine Phosphatase-α and Src Functionally Link Focal Adhesions to the Endoplasmic Reticulum to Mediate Interleukin-1-induced Ca2+ Signaling*

Calcium (Ca2+) signaling by the pro-inflammatory cytokine interleukin-1 (IL-1) is dependent on focal adhesions, which contain diverse structural and signaling proteins including protein phosphatases. We examined here the role of protein-tyrosine phosphatase (PTP) α in regulating IL-1-induced Ca2+ signaling in fibroblasts. IL-1 promoted recruitment of PTPα to focal adhesions and endoplasmic reticulum (ER) fractions, as well as tyrosine phosphorylation of the ER Ca2+ release channel IP3R. In response to IL-1, catalytically active PTPα was required for Ca2+ release from the ER, Src-dependent phosphorylation of IP3R1 and accumulation of IP3R1 in focal adhesions. In pulldown assays and immunoprecipitations PTPα was required for the association of PTPα with IP3R1 and c-Src, and this association was increased by IL-1. Collectively, these data indicate that PTPα acts as an adaptor to mediate functional links between focal adhesions and the ER that enable IL-1-induced Ca2+ signaling.

The interleukin-1 (IL-1) 3 family of pro-inflammatory cytokines mediates host responses to infection and injury. Impaired control of IL-1 signaling leads to chronic inflammation and destruction of extracellular matrices (1,2), as seen in pathological conditions such as pulmonary fibrosis (3), rheumatoid arthritis (4,5), and periodontitis (6). IL-1 elicits multiple signaling programs, some of which trigger Ca 2ϩ release from the endoplasmic reticulum (ER) as well as expression of multiple cytokines and inflammatory factors including c-Fos and c-Jun (7,8), and matrix metalloproteinases (9,10), which mediate extracellular matrix degradation via mitogen-activated protein kinase-regulated pathways (11).
PTPs can be divided into two main categories: receptor-like and intracellular PTPs (33). Two receptor-like PTPs have been localized to FA (leukocyte common antigen-related molecule and PTP␣). Leukocyte common antigen-related molecule can dephosphorylate and mediate degradation of p130 cas , which ultimately leads to cell death (34,35). PTP␣ contains a heavily glycosylated extracellular domain, a transmembrane domain, and two intracellular phosphatase domains (33, 36). The amino-terminal domain predominantly mediates catalytic activity, whereas the carboxyl-terminal domain serves a regulatory function (37,38). PTP␣ is enriched in FA (23) and is instrumental in regulating FA dynamics (39) via activation of c-Src/Fyn kinases by dephosphorylating the inhibitory carboxyl tyrosine residue, namely Tyr 529 (22, 40 -42) and facilitation of integrindependent assembly of Src-FAK and Fyn-FAK complexes that regulate cell motility (43). Although PTP␣ has been implicated in formation and remodeling of FAs (44,45), the role of PTP␣ in FA-dependent signaling is not defined.
Ca 2ϩ release from the ER is a critical step in integrin-dependent IL-1 signal transduction and is required for downstream activation of ERK (13,46). The release of Ca 2ϩ from the ER depends on the inositol 1,4,5-triphosphate receptor (IP 3 R), which is an IP 3 -gated Ca 2ϩ channel (47). All of the IP 3 R subtypes (subtypes 1-3) have been localized to the ER, as well as other the plasma membrane and other endomembranes (48 -50). Further, IP 3 R may associate with FAs, enabling the anchorage of the ER to FAs (51,52). However, the molecule(s) that provide the structural link for this association has not been defined.
FA-restricted, IL-1-triggered signal transduction in anchorage-dependent cells may rely on interacting proteins that are enriched in FAs and the ER (53). Here, we examined the possibility that PTP␣ associates with c-Src and IP 3 R to functionally link FAs to the ER, thereby enabling IL-1 signal transduction.
Cell Culture-Human gingival fibroblasts were grown in minimal essential medium containing 10% fetal bovine serum. Rat2 cells were maintained in Dulbecco's modified Eagle's medium containing 5% fetal bovine serum. Wild-type (PTP␣ ϩ/ϩ ) and PTP␣-null (PTP␣ Ϫ/Ϫ ) fibroblasts were provided by Dr. Jan Sap (University of Copenhagen, Copenhagen, Denmark) (41). In some experiments, PTP␣ Ϫ/Ϫ cells were transfected with wild-type PTP␣ and designated as PTP␣Rescue. Genetically modified NIH3T3 fibroblasts that express HA-tagged wild-type PTP␣ (NIH3T3 PTP␣ ) and C433S/ C723S double mutant PTP␣ (NIH3T3 CCSS ) under control of a doxycycline sensitive repressor were obtained from David Shalloway (Department of Molecular Biology and Genetics, Cornell University, Ithaca, NY) and were generated as previously described (54). The latter cells were grown in Dulbecco's modified Eagle's medium containing 5% fetal bovine serum in the presence of 5 ng/ml doxycycline. Prior to experiments doxycycline was removed (14 -16 h before) to allow expression of recombinant PTP␣.
Plasmid Constructs and Transient Transfection-HA-tagged wild-type PTP␣, PTP␣ lacking the D2 domain (PTP␣ ⌬D2 ) and PTP␣ lacking the D1 and D2 domains (PTP␣ ⌬D1/D2 ) were kindly provided by Dr. J. den Hertog (Hubrecht Laboratory, Netherlands Institute for Developmental Biology, Utrecht, The Netherlands). The cells were seeded in six-well plates at a density of 1 ϫ 10 5 /well 24 h before transfection to yield a 30 -40% confluent culture on the day of transfection. Transient trans-fections were performed using FuGENE 6 transfection reagent (Roche Applied Science), according to the manufacturer's protocol. Briefly, the cells were incubated with DNA-FuGENE 6 reagent (1:3) complexes for 5-7 h. Within 48 h after transfection, the cells were subjected to further experiments.
Short Interfering RNA (siRNA)-Specific knockdown of PTP␣ expression was conducted with commercially available siRNA against PTP␣ (Qiagen). Human gingival fibroblasts were transfected with PTP␣ siRNA or GFP-siRNA (control) using X-tremeGENE siRNA transfection reagent (Roche Applied Science) according to the manufacturer's specifications. The cells were washed in PBS and lysed with SDS-lysis buffer. The lysates were collected, and measurement of the gene knockdown was preformed 24 -72 h after transfection by Western blotting.
Isolation of Focal Adhesions-The cells were grown to 80 -90% confluence on 60-mm tissue culture dishes and were cooled to 4°C prior to the addition of collagen or BSA-coated magnetite beads. Focal adhesion complexes were isolated from cells after specific incubation time periods as described (55). In brief, the cells were washed three times with ice-cold PBS to remove unbound beads and scraped into ice-cold cytoskeleton extraction buffer (0.5% Triton X-100, 50 mM NaCl, 300 mM sucrose, 3 mM MgCl 2 , 20 g/ml aprotinin, 1 g/ml leupeptin, 1 g/ml pepstatin, 1 mM phenylmethylsulfonyl fluoride, 10 mM PIPES, pH 6.8). The cell-bead suspension was sonicated for 10 s (output setting 3, power 15% Branson), and the beads were isolated from the lysate using a magnetic separation stand. The remainder of the lysates was used to assess the non-focal adhesion fraction. The beads were resuspended in fresh ice-cold cytoskeleton extraction buffer, homogenized with a Dounce homogenizer (20 strokes), and reisolated magnetically. The beads were washed in CSKB, sedimented with a microcentrifuge, resuspended in Laemmli sample buffer, and placed in a boiling water bath for 3-5 min to allow the collagen-associated complexes to dissociate from the beads. The beads were sedimented, and lysates were collected for analysis. Subcellular Fractionation-Subcellular fractionation was performed as previously described (56), and the cells were harvested, resuspended in an isotonic buffer (10 mM Tris, pH 7.6, 100 mM CaCl 2 , 200 mM sucrose), and disrupted by Dounce homogenization followed by 20 strokes. The homogenate was spun at 800 ϫ g for 10 min, and the supernatant was recovered and further centrifuged for 10 min at 8,000 ϫ g. The resulting supernatant was further spun for 1.5 h at 28,000 ϫ g. The resulting pellet constituted the microsomal ER fraction. The specificity of the ER fraction was confirmed by immunoblotting with the ER-specific protein calnexin.
Immunoblotting-The protein concentrations of cell lysates were determined by Bradford assay (Bio-Rad). Equal amounts of protein were loaded onto SDS-polyacrylamide gels (10% acrylamide), resolved by electrophoresis, and transferred to nitrocellulose membranes. The membranes were incubated for 1 h at room temperature in Tris-buffered saline solution with 5% milk or 0.2% BSA to block nonspecific binding sites. The membranes were incubated with the primary antibodies overnight at 4°C in Tris-buffered saline with 0.1% Tween 20. Horseradish peroxidase secondary antibodies were incubated for 1 h at room temperature in Tris-buffered saline with 0.1% Tween 20 and 5% milk or 0.2% BSA. Labeled proteins were visualized by chemiluminescence as per the manufacturer's instructions (Amersham Biosciences, Oakville, Canada).
GST Pulldown Experiments-GST-cPTP␣ (residues 167-793, full cytoplasmic domain), GST-PTP␣D1 (residues 167-503, domain 1), and GST-PTP␣D2 (residues 504 -793, domain 2) were kindly provided by Dr. J. den Hertog (Hubrecht Laboratory, Netherlands Institute for Developmental Biology, Utrecht, The Netherlands). The cells were isolated by scraping and lysed for 10 min on ice in radioimmune precipitation assay buffer. The lysed cells were centrifuged at 900 ϫ g for 3 min to remove insoluble debris. The supernatants were removed and stored at Ϫ80°C until use. The cell lysates were precleared with 50 l of 50% slurry of glutathione-Sepharose 4B (1ϫ PBS) and 25 g of GST for 2 h at 4°C. The Sepharose matrix was removed by centrifugation at 500 ϫ g for 5 min. The supernatants were subsequently incubated with 50 l of glutathione-Sepharose 4B, 5 mg of GST protein in PBS ϩ 1% Triton X-100 with gentle agitation at room temperature for 30 min. The matrix was recovered by centrifugation at 500 ϫ g for 5 min. The glutathione-Sepharose 4B pellet was washed four times with 1 ml of PBS. GST was eluted from the glutathione-Sepharose 4B matrix by incubating twice with 50 l of elution buffer (10 mM reduced glutathione in 50 mM Tris-HCl, pH 8.0) for 10 min at room temperature and isolated by centrifugation at 500 ϫ g for 5 min and pooling the supernatants. The samples were boiled for 5 min and analyzed by immunoblotting.
In Vitro Phosphorylation-For in vitro phosphorylation using Fyn, immunoprecipitates bound to protein A-Sepharose beads were incubated for 10 min at room temperature in 20 l of kinase buffer (25 mM HEPES, pH 7.1, 10 mM MgCl 2 , 5 mM MnCl 2 , 0.5 mM EGTA, 1 mM Na 3 VO 4 , 1 mM dithiothreitol, 100 M MgATP) in the presence of 5 units of active Fyn. The reaction products were analyzed by immunoblotting using antibodies against phosphotyrosine.
Calcium Signals-For measurement of whole cell [Ca 2ϩ ] i , cells on coverslips were loaded with 3 M fura-2/AM for 20 min at 37°C. For estimating [Ca 2ϩ ] ER , the cells were incubated with mag-fura-2/AM (4 M) for 150 min at 37°C in culture medium and measured by ratio fluorimetry as described (13). The nominally calcium-free buffer consisted of a bicarbonate-free medium containing 150 mM NaCl, 5 mM KCI, 10 mM D-glucose, 1 mM MgSO 4 , 1 mM Na 2 HPO 4 , and 20 mM HEPES at pH 7.4 with an osmolarity of 291 mOsm. For experiments requiring external Ca 2ϩ , 2 mM CaC1 2 was added to the buffer; for experiments requiring chelation of external Ca 2ϩ , 1 mM EGTA was added. After incubation with fura-2/AM, inspection of cells by fluorescence microscopy demonstrated no vesicular compartmentalization of fura-2, suggesting that the dye loading method permitted measurement of cytosolic [Ca 2ϩ ] i . Visual inspection of mag-fura-2-loaded cells showed fluorescent labeling of intra-cellular organelles. Whole cell [Ca 2ϩ ] i measurements and magfura-2 ratios were obtained with C⅐IMAGING SYSTEMS (Compix, Inc., Cranberry, PA) with excitation wavelengths of 340 and 380 nm and an emission wavelength of 520 nm. Changes in [Ca 2ϩ ] i were monitored by the ratio of fura-2 fluorescence at 340 and 380 nm.
Data Analysis-The means Ϯ S.E. were calculated for [Ca 2ϩ ] i measurements including base-line [Ca 2ϩ ] i , net change in [Ca 2ϩ ] i above base line, and the mag-fura-2 ratios. For continuous variables, the means Ϯ S.E. were computed, and when appropriate, comparisons between two groups were made with the unpaired Student's t test or with analysis of variance for multiple samples. Statistical significance was set at p Ͻ 0.05. For all of the experiments, n Ն 3 replicates were used.
Spatial Relationships between PTP␣ and FA-and ER-associated Proteins-The dependence of IL-1-induced Ca 2ϩ signaling on PTP␣ motivated us to determine whether IL-1 affects the association of PTP␣ with focal adhesion-and ER-associated proteins. After IL-1 treatment, the relative abundance of PTP␣ and IP 3 R1 (the ER calcium release channel isoform that is most abundant in fibroblasts (58)) was increased in focal adhesion preparations (which also contain substantial amounts of plasma membrane) and in ER fractions ( Fig. 2A). By contrast, the relative abundance of the loading control proteins ␣5␤1 integrin (in focal adhesion fractions) and calnexin (in ER fractions) was unchanged by IL-1 treatment. We also found marked, IL-1-induced increases of vinculin in focal adhesion preparations, indicating that IL-1 enhanced the maturation of focal adhesions (44). Vinculin was almost undetectable in the ER fraction.
We examined the localization of PTP␣ relative to focal adhesions and ER-associated proteins using total internal reflection fluorescence microscopy in cells plated on FN or BSA latex beads dispersed on poly-L-lysine substrates. Plating cells on poly-L-lysine coatings restricts focal adhesion formation to fibronectin bead sites in this system (58). Total internal reflection fluorescence microscopy analysis revealed that co-localization of IP 3 R1 with PTP␣ occurred only after IL-1 treatment and only around FN-coated (but not BSA-coated) beads (Fig. 2B).
The relative abundance of IP 3 R1 and PTP␣ in whole cell lysates, focal adhesion preparations or ER fractions was examined by immunoblotting lysates prepared from PTP␣ Ϫ/Ϫ cells or PTP␣ Ϫ/Ϫ cells transfected with wild-type PTP␣ (PTP␣ Rescue ). IL-1 treatment enhanced accumulation of IP 3 R1 in focal adhesions and ER fractions in PTP␣ Rescue cells but not in PTP␣ Ϫ/Ϫ cells (Fig. 3A). We next found that IL-1 promoted the recruitment of PTP␣ and IP 3 R1 to focal adhesions and ER fractions in NIH3T3 cells expressing wildtype PTP␣ (NIH3T3 PTP␣ ), but this effect was much smaller in NIH3T3 CCSS cells (Fig. 3B). Immunoblot analysis of PTP␣ immunoprecipitates from NIH3T3 PTP␣ cells showed that IP 3 R1 co-precipitated with PTP␣ derived from whole cell lysates, focal adhesion preparations, and ER preparations, and these associations were markedly increased after IL-1 treatment (Fig. 4A). By contrast, in NIH3T3 CCSS cells, almost no IP 3 R1 co-precipitated with PTP␣ in cells expressing the catalytically inactive PTP␣ (Fig. 4B).
The associations of PTP␣ with IP 3 R1 were examined with pulldown assays using purified GST-cPTP␣ (cytosolic domain), GST-PTP␣D1 or GST-PTP␣D2 fusion proteins incubated with cell lysates, focal adhesion proteins, or ER fractions prepared from PTP␣ Rescue cells. From immunoblots of IP 3 R1, we found that the cytoplasmic and D1 domains of PTP␣ associated with IP 3 R1 in the FA and ER fractions, more strongly than D2 domain (Fig. 5A). This association appeared to be direct because the purified, bacterially expressed GST fusion-cPTP␣ strongly bound to immunoprecipitated IP 3 R1, whereas immunoprecipitated IP 3 R1 showed no association with control GST beads (Fig. 5B).

IL-1 Regulates Tyrosine Phosphorylation of PTP␣ and Src to Mediate ER-Ca 2ϩ Release-The optimum
Ca 2ϩ conductance of the IP 3 R release channel in the ER requires the activity of SFK such as c-Src and Fyn (59,60), which in turn are regulated by PTP␣ (30,31,61). Accordingly, when we examined the IL-1induced Ca 2ϩ release from the ER of PTP␣ ϩ/ϩ cells, the pronounced, short term release of Ca 2ϩ was blocked in cells pretreated with the SFK inhibitors, SU6656, or PP2, but not in cells pretreated with the inactive structural analog PP3 (Fig. 6A).
The level of SFK activation can be estimated by the phosphotyrosine levels of inhibitory and stimulatory residues, Tyr 529 and Tyr 419 in rodents, respectively (22,40,62). First we examined the regulation of SFK activation using PTP␣ ϩ/ϩ cells plated either on fibronectin (FN) or on poly-L-lysine (PL), conditions that either facilitate or block FA formation, respectively (57). In otherwise unstimulated PTP␣ ϩ/ϩ cells, immunoblot analysis of whole cell lysates with antibody to phospho-Src (Tyr(P) 529 ) revealed very low levels of phosphorylation of the inhibitory tyrosine in cells plated on FN (focal adhesions are present) compared with cells plated on PL (focal adhesions are absent; Fig. 6B). These data are con-FIGURE 2. PTP␣ co-localizes with ER-associated proteins in focal adhesions. A, whole cell lysates, focal adhesion proteins, and ER fractions prepared from Rat2 cells previously treated with IL-1 (ϩ) or vehicle (Ϫ). The lysates were immunoblotted for PTP␣, IP 3 R1, calnexin, vinculin, and GAPDH. B, Rat2 cells plated for 3 h on PL-coated glass coverslips that were preincubated with FN-or BSA-coated latex beads. In untreated cells or after stimulation with IL-1, the cells were co-immunostained for PTP␣/IP 3 R1 and viewed by fluorescence microscopy. WB, Western blotting. FIGURE 3. PTP␣ is required for IP 3 R1 recruitment to focal adhesion and ER. Whole cell lysates, focal adhesion proteins, and ER fractions prepared from the following cells were immunoblotted for PTP␣ and IP 3 R1 following induction with (ϩ) or without (Ϫ) IL-1. A, PTP␣ Ϫ/Ϫ cells or PTP␣ Ϫ/Ϫ cells transfected with wild-type PTP␣ (PTP␣ Rescue ). B, NIH3T3 PTP␣ and NIH3T3 CCSS cells. The numbers beneath the blots in B represent the ratios of PTP␣ or IP 3 R1 to the ␣5␤1 integrin or calnexin controls. WB, Western blotting. sistent with the observation that Src becomes activated via dephosphorylation of Tyr 529 during adhesion and spreading on a fibronectin substrate, and this activation is dependent on and mediated by PTP␣ (24,42). Further, in PTP␣ Ϫ/Ϫ cells, there was abundant phosphorylation of the inhibitory residues (Tyr 529 ) of Src that was unchanged after IL-1 treatment (Fig.  6B), consistent with previous reports (40,63). In PTP␣ ϩ/ϩ cells, there was a modest increase in phosphorylation of Tyr 529 Src, after 15 min of IL-1 treatment (Fig. 6B). This IL-1-induced inhi-bition of Src, however, was not evident at early time points, because the phosphorylation levels of inhibitory (Tyr 529 ) and stimulatory (Tyr 419 ) residues were relatively unchanged (Fig. 6C).
We investigated whether the ability of PTP␣ to dephosphorylate Src diminishes over time after IL-1 treatment. The phosphotyrosine displacement model by Zheng et al. (54) proposes that phosphorylation of tyrosine 789 of PTP␣ selectively promotes its phosphatase activity toward Src by enabling association with the SH2 domain of Src. Accordingly, when whole cell lysates from PTP␣ ϩ/ϩ cells were treated with IL-1 or vehicle and then immunoblotted for phospho-PTP␣ (Tyr(P) 789 ), we observed that there was an initial IL-1-induced phosphorylation of PTP␣ on Tyr 789 , which then decreased in a time-dependent manner after IL-1 treatment (Fig. 6D).
Functional Interactions between PTP␣ and IP 3 R-Because the tyrosine phosphorylation state of the IP 3 R dictates its Ca 2ϩ conductance (64 -66), we examined the effects of IL-1 and PTP␣ on the phosphorylation state of IP 3 R1, which is the most prominent isoform in fibroblasts (58). For this assay IP 3 R1 was immunoprecipitated from whole cell lysates of PTP␣ ϩ/ϩ and PTP␣ Ϫ/Ϫ murine fibroblasts, as well as genetically modified NIH3T3 PTP␣ cells or cells with the catalytically inert PTP␣ mutant (NIH3T3 CCSS ) that had been treated with IL-1 or vehicle control and blotted with a phosphotyrosine antibody. In response to IL-1, phosphorylation of IP 3 R1 in PTP␣ ϩ/ϩ was sharply increased (maximal at 2 min) and declined thereafter (Fig. 7A). Further, the phosphorylation of IP 3 R1 was dependent on the presence of catalytically active PTP␣, because cells lacking PTP␣ or expressing the catalytically inactive (CCSS) PTP␣ showed no detectable IL-1-induced phosphorylation of IP 3 R1 (Fig. 7B).
We next considered whether IP 3 R1 immunopurified from wild-type PTP␣ cells could be directly dephosphorylated by PTP␣. Purified IP 3 R1 bound to protein A-Sepharose beads was exposed to active Fyn in a kinase buffer to promote its tyrosine phosphorylation. Next, the phosphorylated IP 3 R1 protein was incubated with the recombinant cytosolic domain of PTP␣. Active Fyn was able to phosphorylate IP 3 R1, but the recombinant cytoplasmic domain of PTP␣ was unable to dephosphorylate IP 3 R1 (Fig. 7C). As a control, recombinant cPTP␣ readily dephosphorylated recombinant SHP-2 (data not shown). Consequently, IP 3 R1 is not a direct substrate of PTP␣. Instead, we considered that the ability of PTP␣ to activate SFK (c-Src or Fyn) may be achieved by dephosphorylating the carboxyl-terminal inhibitory tyrosine residue (Tyr-(P) 529 of Src), thereby enhancing the catalytic activity of these kinases  GST pulldown experiments to show association between PTP␣ and IP 3 R1 in vitro are shown. A, glutathione-agarose (GST) beads bound to bacterially expressed cPTP␣, cPTP␣D1, or cPTP␣D2 fusion proteins or GST control beads were incubated with whole cell lysates (WL), FA proteins, and ER fractions from PTP␣ Rescue cells. The materials bound to the beads were analyzed by immunoblotting for IP 3 R1. Whole cell lysates, FA proteins, and ER fractions were also immunoblotted for IP 3 R1. Note that the protein products of the domain deleted constructs (cPTP␣D1 or cPTP␣D2) have lower molecular masses (ϳ60 kDa) compared with the cytosolic cPTP␣ (95 kDa) or full-length PTP␣ (130 kDa). B, GST-cPTP␣ or GST control fusion proteins were incubated with IP 3 R1 that was IP purified from whole cell lysates of PTP␣ Rescue cells. WB, Western blotting. (30,31,61). We examined this possibility by first pretreating wild-type PTP␣ cells with the SFK inhibitor, SU6656. This treatment blocked IL-1-induced phosphorylation of IP 3 R1 (Fig.  7D), indicating that the catalytic activity of SFK is required for IL-1-induced phosphorylation of the IP 3 R1. We confirmed the ability of SU6656 to reduce the catalytic activity of Src by examining the Tyr 529 inhibitory residue of Src. For this experiment, PTP␣ wild-type cells were treated with SU6656 for 1 h and stimulated with or without IL-1 for 30 min. When the whole cell lysates were immunoblotted for Tyr(P) 529 , increased levels of Tyr(P) 529 were observed after SU6656 treatment (Fig. 7E). We examined whether SFKs associate with IP 3 R1 in response to IL-1. Wild-type PTP␣ cells were treated with IL-1. IP 3 R1 was immunoprecipitated from cell lysates and immunoblotted with SFKs or c-Src specific antibodies. Time course experiments showed that SFKs associated with IP 3 R1 after IL-1 and that in particular, c-Src was specifically detected in association with IP 3 R1 (Fig. 7F). However, this association was dependent on the presence of PTP␣, because in PTP␣-null cells, the IL-1-induced association of Src with IP 3 R1 was abolished (Fig. 7F, middle  panel). In a similar experimental design, we prepared PTP␣ immunoprecipitates and immunoblotted for Src and c-Src (Fig.  7G). These data showed that PTP␣ associates with SFK (and specifically c-Src) in response to IL-1 in a time-dependent manner. Taken together with the earlier data showing that catalyt-ically active PTP␣ associates with IP 3 R1 in focal adhesions and ER preparations in response to IL-1 treatment (Fig. 4), we conclude that PTP␣ may mediate IL-1-induced Ca 2ϩ signaling by acting as an adaptor to link IP3R1 to c-Src.

DISCUSSION
Our major finding is that PTP␣ provides an important structural linkage between focal adhesions and the ER, in part through its interactions with IP 3 R1 and c-Src. In previous work (46,58) we found rapid, focal adhesion-dependent release of Ca 2ϩ from the ER in response to IL-1, suggesting a functional relationship between these different organelles, as well as spatial sequestration of signaling molecules involved in Ca 2ϩ signaling (14). In view of these findings, we have sought to define the proteins that mediate this spatial selectivity and that enable focal adhesion restriction of IL-1 signaling.
Src-We found that inhibition of Src kinase activity with SU6656 or PP2 effectively blocked IL-1-induced IP 3 R1 phosphorylation and ER Ca 2ϩ release, indicating the importance of SFKs in focal adhesion-dependent calcium signaling initiated by IL-1. SFKs are pivotal for integrin-mediated signaling during cell adhesion and spreading because of their kinase-dependent and kinase-independent activities (67). Among the SFKs, c-Src and Fyn are dephosphorylated by PTP␣ on their inhibitory carboxyl-terminal tyrosine residue, which enhances their kinase activity (31,54). Consistent with these reports we found that PTP␣ was essential for enabling signaling downstream of SFKs because PTP␣-null cells exhibited very high levels of phospho-rylation of the inhibitory Tyr 529 residue and showed no IL-1-induced Ca 2ϩ signaling. When the inhibitory carboxyl-terminal tyrosine residue was dephosphorylated in cells that have intact focal adhesions and express catalytically active PTP␣, SFK catalytic activity enabled direct phosphorylation of IP 3 R1 and, subsequently, Ca 2ϩ release from the ER. Further, our immunoprecipitation studies showed that c-Src was one of the SFKs that associated with IP 3 R1, contemporaneous with enhanced ER Ca 2ϩ release. Thus c-Src could be one of the SFKs that directly phosphorylate and activate IP 3 R1, although other SFKs such as Fyn and Lyn could also be involved in this process (59,60).
Importantly, the activation of SFK is only achieved in the presence of intact FA and PTP␣ and is required for the immediate Ca 2ϩ response in cells treated with IL-1. Eventually continued IL-1 treatment longer than 15 min led to the inactivation of Src and this correlated temporally with increased phosphorylation of inhibitory residue (Tyr 529 ) of Src and diminished phosphorylation of PTP␣ at Tyr 789 , consistent with the notion that Tyr(P) 789 of PTP␣ is important for regulating Src activation (24,54).
IP 3 Receptors-IP 3 Rs are Ca 2ϩpermeable channels located on the membranes of organelles with releasable Ca 2ϩ stores (48). Cell surface biotinylation studies have shown that ϳ5-14% of total IP 3 receptors are localized to the plasma membranes of several cell types (49). We have previously found abundant IP 3 R1 isoform in the ER of human gingival fibroblasts (58) and in the ER of murine fibroblasts used here. Notably, IP 3 R1 was also enriched in focal adhesion-associated proteins, suggesting that IP 3 R1 may link the ER to FAs at specific cellular sites. As a result of the critical importance of IP 3 receptors in Ca 2ϩ release from the ER (51) and their central role demonstrated here in IL-1-induced Ca 2ϩ signaling, we focused our studies on IP 3 R as a Ca 2ϩ release channel and as a candidate protein for linking the ER to FAs. FIGURE 7. Tyrosine phosphorylation of IP 3 R1: regulation by PTP␣ and Src. A, tyrosine phosphorylation of IP 3 R1 rapidly increases after IL-1 treatment and then declines subsequently. IP 3 R1 was immunopurified (IP) from whole cell lysates of PTP␣ ϩ/ϩ cells that had been stimulated with or without IL-1, and immunoblotted for phosphotyrosine (pTyr, upper panel). The lower panel shows protein levels of immunoprecipitated IP 3 R1. The line graph on the right shows the ratios of the blot densities of tyrosine-phosphorylated IP 3 R1 to IP 3 R1 over the full time course. B, tyrosine phosphorylation of IP 3 R1 after IL-1 stimulation requires catalytically active PTP␣. IP 3 R1 was immunopurified from whole cell lysates of PTP␣ ϩ/ϩ , PTP␣ Ϫ/Ϫ , NIH3T3 PTP␣ , and NIH3T3 CCSS cells that had been stimulated with (ϩ) or without (Ϫ) IL-1. Immunoprecipitates were immunoblotted for phosphotyrosine (pTyr, upper panel) or IP 3 R1 (lower panel). C, IP 3 R1 phosphorylation is not affected by PTP␣. In vitro phosphorylation was analyzed by incubating immunopurified IP 3 R1 from wild-type PTP␣ cells in presence or absence of active Fyn and then incubated with GST-cPTP␣ fusion protein. Phosphorylation was analyzed by immunoblotting with anti-phosphotyrosine antibody (pTyr, upper panel). The lower panel shows protein levels of immunopurified IP 3 R1. D, tyrosine phosphorylation of IP 3 R1 is dependent on catalytically active SFK. IP 3 R1 was immunoprecipitated from PTP␣ ϩ/ϩ cells that had been pretreated with SU6656 (5 M for 1 h) or vehicle and then treated with (ϩ) or without (Ϫ) IL-1. Immunoprecipitates were immunoblotted for phosphotyrosine (pTyr, upper panel) or IP 3 R1 (lower panel). E, cells treated with Src inhibitor exhibit high levels of Src (Tyr(P) 529 ). The cells were preincubated with SU6656 (5 M for 1 h) and then treated with IL-1 (ϩ) or vehicle (Ϫ) for 10 min. The cell lysates were immunoblotted for phosphorylation of the carboxyl-terminal inhibitory tyrosine residue of Src (Tyr(P) 529 ). F, IL-1 promotes association of IP 3 R1 with SFK and c-Src and PTP␣ is required for interactions between Src and IP 3 R1. Left panel, PTP␣ ϩ/ϩ cells were stimulated with IL-1 as indicated. Middle panel, PTP␣ Ϫ/Ϫ cells were stimulated with IL-1 as indicated. IP 3 R1 immunoprecipitates were immunoblotted for either SFK (left and middle panels, Src) or c-Src (right panel). G, IL-1 promotes association of PTP␣ with SFK and c-Src. PTP␣ ϩ/ϩ cells were stimulated with IL-1. PTP␣ immunoprecipitates were immunoblotted for either SFK (left panel, Src) or c-Src (right panel). After longer exposure to IL-1, the association dissipated. WB, Western blotting.
The precise function of IP 3 R is mediated by the coordinated actions of Ca 2ϩ , phosphorylation, and nucleotides. Ca 2ϩ exerts biphasic control over IP 3 R (68 -70): stimulation by positive feedback at physiological concentrations and inhibition at low micromolar [Ca 2ϩ ] (71,72). Phosphorylation of IP 3 R1 increases its sensitivity to Ca 2ϩ conductance at even low IP 3 levels (73). We found that IP 3 R1 was strongly tyrosine-phosphorylated in response to IL-1, particularly at early time periods (prior to 15 min); this phosphorylation diminished rapidly thereafter. Because our immunoprecipitation studies showed temporal correlations between increased IP 3 R1 phosphorylation and IP 3 R1 association with SFK and c-Src, it seems likely that SFKs promote Ca 2ϩ signaling not only by phosphorylating and activating phospholipases that generate IP3 but also by directly phosphorylating IP 3 R (Tyr 353 ) and augmenting Ca 2ϩ channel activity (59,73). Further, tyrosine phosphorylation renders the IP 3 R insensitive to rising inhibitory Ca 2ϩ levels (59), which would otherwise inhibit its open probability.
In addition to its role as a Ca 2ϩ channel in the ER membrane, IP 3 R has been implicated in many adaptor functions because of its association with various phosphatases (51). Further, the association of IP 3 R with the FA proteins vinculin, ␣-actinin, and talin (52) suggest that IP 3 R could interact with other critical focal adhesion signaling molecules. We found here that IL-1 up-regulated the relative abundance of IP 3 R1 in ER fractions as well as in FA-associated proteins, but this enrichment only occurred in cells expressing catalytically active PTP␣. By total internal reflection fluorescence microscopy imaging of subplasma membrane compartments, we observed that PTP␣ colocalizes with IP 3 R1. Consistent with these observations, immunoprecipitation and pulldown assays with purified proteins indicated that PTP␣ associated with IP 3 R1. However, we found that IP 3 R1 was not a substrate of PTP␣. Instead, SFKs and, in particular, c-Src associated with and likely phosphorylated IP 3 R in response to IL-1. Notably, the disruption of the enzymatic activity of PTP␣ impaired the ability of IP 3 R1 to associate with PTP␣ and Src. Apparently, the adaptor function of PTP␣ relies on the integrity of its catalytic domain because only point mutations in the D1 (C433S) and D2 (C723S) domains were required to disrupt IL-1-induced tyrosine phosphorylation of IP 3 R1, Ca 2ϩ signaling, and association of IP 3 R1 with PTP␣. Notably, the association of IP 3 R1 with PTP␣ in response to IL-1 may mediate physical approximation of FAs with the ER. By this mechanism PTP␣ may indirectly control IL-1-induced tyrosine phosphorylation and enhancement of IP 3 R1 Ca 2ϩ channel activity, possibly via the activation of c-Src or Fyn.
We conclude that PTP␣ plays two essential roles in IL-1induced Ca 2ϩ signaling. First, PTP␣ dephosphorylates and activates SFKs, in response to integrin stimulation, that are essential for enabling Ca 2ϩ release through IP 3 receptors in the ER. Second, PTP␣ provides a critical structural link between FAs and the ER as a result of its interactions with IP 3 R1 and c-Src. This trimolecular interaction is needed for IL-1-induced Ca 2ϩ release.