Abstract

Disruption of cell contact sites during ischemia contributes to the loss of organ function in acute renal failure. Because prior heat stress protects cell contact sites in ATP-depleted renal epithelial cells in vitro, we hypothesized that heat shock protein 72 (HSP72), the major inducible cytoprotectant in mammalian cells, interacts with protein kinases that regulate cell-cell and cell-matrix interactions. ATP depletion increased the content of Tyr416 Src, the activated form of this kinase. c-Src activation was associated with an increase in the tyrosine phosphorylation state of β-catenin, paxillin, and vinculin, three c-Src substrate proteins that localize to and regulate cell contact sites. Prior heat stress inhibited c-Src activation and decreased the degree of tyrosine phosphorylation of all three Src substrates during ATP depletion and/or early recovery. HSP72 coimmunoprecipitated with c-Src only in cells subjected to heat stress. ATP depletion markedly increased the interaction between HSP72 and c-Src, supporting the hypothesis that HSP72 regulates Src kinase activity. These results suggest that alterations in the tyrosine phosphorylation state of proteins located at the cell-cell and cell-matrix interface mediate, at least in part, the functional state of these structures during ATP depletion and may be modulated by interactions between HSP72 and c-Src.

  • ischemia
  • heat shock protein 70
  • cytoskeleton
  • Yes kinase
  • β-catenin
  • paxillin
  • vinculin
  • Triton X-100

transient ATP depletion of renal epithelial cells induces dramatic and potentially reversible changes in cell morphology and function that mimic renal ischemia in vivo. Collapse of the cytoskeleton results from the depolymerization of F-actin into macromolecular aggregates composed of actin fragments (5, 7, 28, 55). Either as a cause or as a consequence of cytoskeletal collapse, cell-cell and cell substratum contacts are disrupted (8, 55). Transepithelial cell resistance decreases and transcellular fluid flux increases, suggesting that impairment of tight junction integrity is an early event during ATP depletion. In these epithelial cells, disruption of the tight junction, a key component of junctional complex (13), compromises cell polarity and vectoral solute transport (5, 9,36, 55). As a consequence of ATP depletion, integrin-mediated cell-matrix interaction is reduced, decreasing the adherence of both live and dead cells to the substratum (28, 42, 56, 57,62). Importantly, these events are likely to mediate the acute loss of renal function by promoting backleak of glomerular filtrate and intratubular obstruction from cells that detach from the substratum.

Prior heat stress, sufficient to induce heat shock protein 72 (HSP72), minimizes many of the morphological and functional consequences of ATP depletion. Heat stress stabilizes the actin cytoskeleton and reduces the formation of large actin aggregates derived from F-actin (7,22). Heat stress also ameliorates the loss of function observed at cell contact sites in renal epithelial cells subjected to ATP depletion (7). Although the mechanism by which heats stress prevents stress-induced injury is unknown, HSP72, the major inducible protein member of the HSP70 family, has been closely linked with cytoprotection. Selective expression of HSP72 has been shown to protect against ischemic injury in vitro (34) and in vivo (31), suggesting that HSP72 mediates, at least in part, the protection afforded by heat stress. Although the molecular chaperone function of HSP72 has been studied extensively (6,20), few protein-binding partners or cellular target sites of protection have been identified.

Protein tyrosine kinases regulate the phosphorylation of tyrosine residues of structural and regulatory proteins located at both cell-cell and cell-matrix contact sites. c-Src, one of nine members of the Src family, is a tyrosine kinase with multiple identified protein substrates (10, 14, 49). Several of these protein substrates localize to cell contact sites, including β-catenin, paxillin, and vinculin (10, 12, 29, 35, 44, 48, 50, 52). Changes in the degree of tyrosine phosphorylation regulate protein function at cell-cell and cell-matrix adhesion sites. Cell stressors such as hypoxia (51) or ATP depletion (12, 48,55) perturb protein tyrosine phosphorylation and alter the functional state of cell contact sites.

Recent evidence suggests that c-Src-mediated protein tyrosine phosphorylation is causally linked to both morphological and functional changes at cell contact sites. Specifically, c-Src regulates tight junction permeability (16, 29, 41, 52) and adhesion of cells to the substratum (12, 25, 38, 39, 42, 46, 56) and maintains the actin cytoskeleton (14, 19, 24, 25). In addition to regulating cell contact sites, c-Src also modulates cell proliferation (43, 44), an important determinant of organ recovery after an ischemic insult. Despite its potential role in mediating injury and repair of cell contact sites in epithelial cells, changes in c-Src activity after ATP depletion have not been previously reported.

Because c-Src alters the phosphorylation state of structural and regulatory proteins that localize to contact sites that are protected by prior stress, we hypothesized that HSP72, an important cytoprotectant, modulates the activation of c-Src in ATP-depleted renal epithelial cells. After ATP depletion, marked changes in the overall degree of protein tyrosine (but not serine) phosphorylation were observed. ATP depletion increased the steady-state content of c-Src but not Yes kinase, another member of the Src kinase family. ATP depletion and/or recovery increased the tyrosine phosphorylation of c-Src at Tyr416 (the activated form of the kinase) and enhanced protein tyrosine phosphorylation of three known Src substrates: β-catenin, paxillin, and vinculin. Prior heat stress inhibited the activation of Tyr416 Src associated with ATP depletion and decreased protein tyrosine phosphorylation of all three c-Src substrate proteins. In previously heated cells, HSP72 was detected in immunoprecipitates obtained with antibody directed against c-Src. ATP depletion markedly increased the binding between HSP72 and c-Src. These data support the hypothesis that HSP72 inhibits c-Src-mediated tyrosine phosphorylation of proteins that localize to cell contact sites. These observations may explain how prior heat exposure increases the resistance of cell contact sites to ATP depletion injury.

METHODS

Materials.

All reagents were obtained from Sigma Chemical (St. Louis, MO) unless otherwise indicated.

Cell culture.

Renal epithelial cells obtained from the opossum kidney (OK) were obtained from American Type Culture Collection (CRL-1840) and were grown in Dulbecco's modified Eagles's medium (DMEM; GIBCO BRL, Grand Island, NY) supplemented with 10% fetal calf serum. Cells were used within 72 h of achieving confluence as assessed by visual inspection with a phase-contrast microscope.

ATP depletion and induction of HSP72.

ATP content was reduced by exposing the cells to 1–1.5 h of glucose-free medium (DMEM, GIBCO BRL) containing cyanide (5 mM) and 2-deoxy-d-glucose (5 mM). This maneuver reduces ATP content to <10% of the baseline value within 10 min and sustains this low level of ATP content (58). Recovery from ATP depletion was performed by exchanging the medium with DMEM containing 10 mM glucose (without metabolic inhibitors). Parallel medium changes were made in controls using glucose-containing DMEM. To induce HSP72, OK cells were heated to 43 ± 0.5°C for 60 min in a temperature-regulated incubator, followed by incubation at 37°C for 16–18 h (58).

Whole cell lysate.

Harvested cells were resuspended in cell lysis buffer (described in the immunoprecipitation procedure below) that contained protease inhibitors [25 μM benzamidine, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), and 0.2 IU/ml aprotinin, as previously described (56)] as well as 1 mM sodium vanadate and 50 mM sodium fluoride, as previously reported (48). The cells were sonicated and then centrifuged at 10,000 g for 10 min at 4°C. The supernatant was designated as whole cell lysate.

Triton X-100 extraction.

Monolayers of cells in individual culture dishes were incubated for 30 min at 4°C with Triton X-100 containing extraction buffer (in mM: 50 NaCl, 300 sucrose, 10 PIPES, 3 MgCl2, 1.2 PMSF, 0.1 mg/ml DNase, 0.1 mg/ml RNase, and 0.5% vol/vol Triton X-100, pH 6.8; modified from Ref. 38). At the end of the incubation, the cells were scraped and then centrifuged at 10,000 g for 10 min. The supernatant (the Triton X-100-soluble protein fraction) was obtained. The Triton X-100-soluble protein fraction is primarily composed of cytosolic and some membrane-associated proteins that are released in the presence of detergent. The pellet (the Triton X-100-insoluble protein fraction) was sonicated in lysis buffer (described earlier). Cytoskeletal proteins are the major constituents of the Triton X-100-insoluble protein fraction (4, 22,55).

Immunoblot analysis.

HSP72 was detected with a mouse monoclonal antibody specific for this inducible member of the HSP70 family (Amersham, Arlington Heights, IL) as reported by our laboratory (59). Structural cell proteins, including β-catenin (Sigma), paxillin (ICN Biomedicals, Cosa Mesa, CA), and vinculin (Sigma), were detected using specific, commercially available antibodies. Specific antibodies were used to assess the content of Yes kinase (Upstate Biotechnology, Lake Placid, NY) and c-Src (Chemicon, Temecula, CA). The degree of tyrosine and serine phosphorylation in either whole cell lysates or in immunoprecipitated proteins obtained from whole cell lysates was determined using a PY20 (Transduction Laboratories, Lexington, KY) and anti-serine (Zymed Laboratories, San Francisco, CA) antibodies, respectively. Tyr416 Src, the activated form of the kinase (15, 18), was assessed using an antibody kindly provided by Dr. Lei Chen (New England Biolabs, Boston, MA). Specific protein bands were detected with an anti-IgG antibody linked to a horseradish peroxidase-based enzyme-linked chemiluminescence system (Lumigolo; Kirkegaard and Perry, Gaithersburg, MD) and were quantified using NIH Image Quant software after the blot was scanned with a densitometer (Hewlett-Packard, Desk Scan II).

Immunoprecipitation and coimmunoprecipitation.

Aliquots of samples obtained from both whole cell lysates or from the Triton X-100-soluble protein pool were subjected to immunoprecipitation (IP) as previously reported by our laboratory (59). Samples were diluted to 0.5–1 mg of protein/ml with IP buffer (150 mm NaCl, 10 mM Tris · HCl, 5 mM EDTA, 1 mM EGTA, 1% Triton X-100, 0.5% Nonidet P-40, 10 μg/ml PMSF, and 10 μg/ml aprotinin, pH 7.4). Apyrase (10 U/ml) was added to prevent ATP-mediated release of potential binding partners from HSP72 (59). Samples were “precleared” for 2 h with nonimmune serum (5 μl/mg protein) obtained from the same host species as the primary antibody. Supernatant was incubated overnight at 4°C with antibodies directed against specific proteins (1–2 μg · mg protein−1 · ml−1 IP buffer) for the initial immunoprecipitate. Protein A-coated beads were added to the solution during the final 2 h of incubation. A preclearing procedure was performed for each IP study. The pellet obtained by centrifugation was washed for 5 min with a high-stringency buffer [HS-B; 0.1% SDS, 1% deoxycholic acid, 0.5% Triton X-100, 20 mM Tris · HCl, 120 mM NaCl, 25 mM KCl, 5 mM EDTA, 5 mM EGTA, 0.1 mM dithiothreitol (DTT)] with 1 M sucrose, pH 7.5. Samples were then washed in a high-salt buffer (HS-B + 1 M NaCl). A final wash in a low-salt buffer (2 mM EDTA, 0.5 mM DTT, and 10 mM Tris · HCl, pH 7.5) was performed. In coimmunoprecipitation studies, the initial immunoblots were reprobed with specific primary antibodies to detect potential HSP72 binding partners.

Protein assay.

Protein concentrations were determined with a colorimetric dye-binding assay (BCA Assay; Pierce, Rockford, IL). Results are expressed in milligrams of protein per milliliter.

Statistical analysis.

Data are expressed as means ± SE. Results involving more than one group were compared using analysis of variance and were then analyzed with Fisher's post hoc test. A result was considered significant if the P value was <0.05.

RESULTS

To characterize changes in protein tyrosine and serine phosphorylation, renal epithelial cells were subjected to Triton X-100 extraction before, during, and after transient ATP depletion. The Triton X-100 fraction is primarily composed of cytosolic as well as membrane-associated proteins that are released in the presence of detergent. In contrast, Triton X-100-insoluble proteins are largely derived from the cytoskeleton (4, 22, 55). Before ATP depletion, most tyrosine phosphorylated proteins localized to the Triton X-100-soluble fraction (Fig.1 A, right,lane C). After 60 min of ATP depletion (time 0), the majority of tyrosine residues in the Triton X-100-soluble fraction was dephosphorylated. During recovery from ATP depletion, a marked increase in tyrosine phosphorylation of selected protein bands was observed (15- and 30-min recovery from ATP depletion). Although much less marked, increased protein tyrosine phosphorylation was also observed in the Triton X-100-insoluble fraction during recovery from ATP depletion (Fig. 1 B, left).

Fig. 1.

A and B: protein tyrosine phosphorylation during ATP depletion and recovery. Samples were obtained using Triton X-100 extraction (described inmethods) and divided into detergent-soluble and -insoluble pools. Cells were maintained at 37°C (control, A) or were subjected to 43°C (heat stress, B). Samples were obtained before ATP depletion (C) or 0, 15, and 30 min after transient ATP depletion. Proteins (40 μg total/lane) were separated by 7.5% SDS-PAGE. Immunoblots were probed with antibody directed against phosphotyrosine (PY20). Most tyrosine-phosphorylated proteins reside in the Triton X-100-soluble pool in both control and previously heated cells. A marked increase in tyrosine phosphorylation of some protein bands was observed in control cells during recovery from ATP depletion (A, arrowheads). Heat stress inhibited the increase in phosphotyrosine content associated with recovery (B, arrowheads). Virtually identical results were observed when proteins in whole cell lysates were separated using a 4–20% gradient SDS gel (data not shown). MW, molecular weight.

Compared with controls (Fig. 1 A), prior heat stress did not substantially alter the basal state of tyrosine phosphorylation in either the Triton X-100-soluble or -insoluble protein fractions (Fig.1 B, lane C, left and right vs. Fig.1 A). In addition, the overall content of tyrosine-phosphorylated proteins was similar in control and previously heated cells after 60 min of ATP depletion (time 0, heat stress vs. control). However, heat stress ameliorated the relative hyperphosphorylation of select Triton X-100-soluble protein bands during recovery from ATP depletion at both the 15- and 30-min time points (Fig. 1 B, right). In contrast to the relatively large changes in protein tyrosine phosphorylation caused by ATP depletion, minimal changes in the overall degree of serine phosphorylation were seen in protein fractions obtained from ATP-depleted cells with or without prior heat stress (data not shown).

To evaluate the potential contribution of Src kinases in mediating tyrosine phosphorylation during ATP depletion and recovery, the steady-state content of c-Src and Yes kinase was assessed. Although c-Src and Yes are both members of the Src kinase family and associate with structural cell proteins, the two kinases bind distinct ligands (37) and are differentially activated (51, 53, 54,61). In control cells, c-Src content increased approximately twofold at 1.5 and 3 h post-ATP depletion (Fig.2, top, left). In contrast, the content of Yes kinase markedly decreased in ATP-depleted control cells. The loss of Yes kinase content was also evident during recovery from ATP depletion. Prior heat stress prevented the increase in steady-state c-Src content during late recovery from ATP depletion (Fig. 2, top, right). These data suggest that c-Src is more likely than Yes kinase to mediate protein tyrosine phosphorylation in ATP-depleted renal epithelial cells and show that prior heat stress ameliorates changes in c-Src content during recovery.

Fig. 2.

Effect of ATP depletion and recovery on total c-Src and Yes kinase content. Opossum kidney (OK) cells were subjected to 1.5 h of ATP depletion, followed by either 1.5 or 3 h of recovery. Immunoreactive protein content was evaluated in whole cell lysates obtained from control and previously heated cells before, during, and after ATP depletion using antibodies directed against either c-Src or Yes kinase. Each lane contains 80 μg of total protein.

Because changes in kinase activity may be independent of fluctuations in c-Src content, a specific antibody was used to detect tyrosine phosphorylation at Tyr416, the activated form of the enzyme (15, 39). Compared with baseline conditions (“ATP replete”), ATP depletion significantly increased Tyr416phosphorylation (Fig. 3, A andB; P < 0.05, n = 3). A significant increase in Tyr416 Src phosphorylation was also observed after 15 min of recovery (P < 0.05). Prior heat stress prevented the increase in Tyr416 c-Src during both ATP depletion and recovery. Importantly, prior heat stress per se did not significantly alter the content of phospho-Tyr416c-Src compared with control (998.7 ± 224.6 vs. 1,062.7 ± 214.9 relative density units in heat stress and control groups, respectively; P > 0.05, n = 3). The observed increases in Tyr416 c-Src were independent of changes in Src expression, since the steady state of level of c-Src did not change during ATP depletion or early recovery (Fig. 3 A,bottom).

Fig. 3.

A and B: effect of ATP depletion on c-Src activation. A: steady-state content of total c-Src (bottom) and activated Src kinase (Tyr416 Src; top) in whole cell lysates obtained from control and previously heated OK cells before (ATP replete), immediately after ATP depletion (ATP deplete), and after 15 or 30 min of recovery (Recovery). Each lane contains 60 μg of total protein.B: densitometric analysis of Tyr416 Src content in cells subjected to ATP depletion followed by recovery. Data expressed are means ± SE of the relative density compared with either control or previously heated cells at baseline (i.e., before ATP depletion). Tyr416 content was similar in control and heat-stressed cells before ATP depletion (see text). *P< 0.05 control (open bar) vs. heat stress (filled bar); †P < 0.05 vs. ATP replete; n = 3. NS, not significant.

To determine whether c-Src activation resulted in protein tyrosine phosphorylation, the phosphorylation state of β-catenin, paxillin, and vinculin, three Src substrates that localize to cell contact sites, was examined before, during, and after ATP depletion. Both ATP depletion and recovery were associated with an increase in the content of tyrosine-phosphorylated β-catenin, a regulatory component of the intercellular junction (Fig. 4, control,top, lanes 2 and 3 vs. lane 1). Prior heat stress inhibited the increase in the content of tyrosine-phosphorylated β-catenin during ATP depletion and recovery (Fig. 4, heat stress, top, lanes 2 and3 vs. control, lanes 2 and 3). Importantly, total β-catenin content was similar in both control and heat stress groups in each experimental condition (Fig. 4,bottom). In contrast to phosphotyrosine, the degree of serine phosphorylation of β-catenin was minimally affected by ATP depletion or recovery in either control or heat-stressed cells (Fig. 4,middle).

Fig. 4.

Effect of ATP depletion on β-catenin phosphorylation. β-Catenin was immunoprecipitated from the Triton X-100-soluble protein fraction in samples obtained from control (ATP replete), ATP depleted (ATP deplete), and 15 min after ATP depletion (Recovery) cells. After SDS-PAGE, nitrocellulose membranes were probed with anti-p-tyrosine (top), anti-p-serine (middle), and anti-β-catenin (bottom) antibodies. The membrane was stripped after exposure to each antibody. There was no difference between the phosphorylation state of β-catenin in control and heat stress cells before ATP depletion. Results in each lane were obtained from immunoprecipitation of 100 μg of total protein in the initial sample.

To assess the content of tyrosine-phosphorylated paxillin, a cell adhesion protein, the Triton X-100-soluble protein fraction, was immunoprecipitated with an anti-tyrosine antibody (PY20). In control cells, ATP depletion markedly decreased the content of tyrosine-phosphorylated paxillin (Fig.5, top, left). During recovery, however, the content of tyrosine-phosphorylated paxillin exceeded the level observed before injury (lane 3vs. lane 1, top). In previously heated cells, the content of tyrosine-phosphorylated paxillin was less than observed in controls during ATP depletion and recovery (Fig. 5,lanes 4 and 5 vs. lanes 2 and3, top).

Fig. 5.

Effect of ATP depletion on paxillin phosphorylation. Tyrosine-phosphorylated paxillin (p-paxillin) was immunoprecipitated from the Triton X-100-soluble protein pool using a PY20 antibody (top). Samples were obtained from cells as described in the legend to Fig. 4. Cells were permitted to recover from ATP depletion for 15 min. Paxillin content was then determined by immunoblot analysis (bottom). Minimal differences between the phosphorylation state of paxillin in control and heated stress cells before ATP depletion was observed. Results in each lane were obtained from immunoprecipitation of 100 μg (top) or 40 μg (bottom) of total protein in the initial sample.

Both ATP depletion and recovery were associated with an increase in the tyrosine phosphorylation of vinculin, a structural protein that localizes to both the cell-cell and cell-matrix interfaces (Fig.6, top, left). In contrast, the content of serine-phosphorylated vinculin did not change either during or after ATP depletion (Fig. 6, middle,left). Prior heat stress ameliorated the increase in tyrosine-phosphorylated vinculin during ATP depletion and recovery (Fig. 6, top, right) but did not affect the content of serine-phosphorylated vinculin (Fig. 6,middle, right). Within each group (control and heat stress), total vinculin content in the immunoprecipitates was virtually identical during all three experimental conditions (bottom).

Fig. 6.

Effect of ATP depletion on vinculin phosphorylation. Samples obtained from the Triton X-100-soluble protein fraction were prepared from cells as described in the legend to Fig. 4. Samples were obtained after 15 min of recovery from ATP depletion. After immunoprecipitation with an anti-vinculin antibody, the immunoblot was successively probed with anti-p-tyrosine (top), anti-p-serine (middle), and anti-vinculin (bottom) antibodies. There was no difference in the phosphorylation state of vinculin in control and heat-stressed cells before ATP depletion. Results in each lane were obtained from immunoprecipitation of 100 μg of total protein in the initial sample.

To evaluate potential interaction between HSP72 and c-Src, the degree of coimmunoprecipitation of the two proteins was assessed. IP was performed using antibody directed against c-Src. Although c-Src was relatively more abundant in immunoprecipitates obtained from control compared with previously heated cells (Fig.7, bottom, lanes 2–5 vs. lanes 6–9), in each experimental condition, coimmunoprecipitation of HSP72 with c-Src was observed only in cells subjected to heat stress (Fig. 7, top,lanes 6–9). ATP depletion markedly increased the interaction between HSP72 and c-Src (Fig. 7, top, lane 7vs. lane 6). Similarly, interaction between the two proteins persisted during early recovery (top, lane 8).

Fig. 7.

c-Src coimmunoprecipitates with heat shock protein 72 (HSP72). Anti-c-Src antibody was used to obtain immunoprecipitates from samples derived from whole cell lysates before ATP depletion (ATP replete), immediately after 1 h of ATP depletion (ATP deplete), and after 15 and 30 min of recovery (Recovery). These immunoprecipitates were separated by 7.5% SDS-PAGE; immunoblots were then probed with a monoclonal antibody specific for HSP72. Results in each lane were obtained by immunoprecipitating 500 μg of total protein in the initial sample.

DISCUSSION

In the present study, renal epithelial cells were subjected to pharmacological ischemia, a maneuver that rapidly reduces ATP content to <10% of the baseline value within 10 min in both control and previously heated OK cells (58). Despite the reduction in ATP content, some intracellular proteins sustain or increase their basal level of tyrosine phosphorylation (Fig. 1). During ATP depletion, marked and reversible alterations in transepithelial cell resistance (TER), a marker of tight junction integrity in renal epithelial cells, parallel changes in ATP content (7, 48). In a recent study by Schwartz and colleagues (48), loss of TER correlated with tyrosine phosphorylation of β-catenin and plakoglobin, proteins that localize to the adherens junction. Inhibition of protein phosphatases increased the tyrosine phosphorylation of β-catenin and plakoglobin and simulated the effect of ATP depletion on TER. In contrast, a tyrosine kinase inhibitor prevented the tyrosine phosphorylation of β-catenin and plakoglobin and partially reversed the defect in TER caused by the phosphatase inhibitor. This previous report demonstrated that changes in the degree of protein tyrosine phosphorylation correlate with functional changes at contact sites in ATP-depleted renal epithelial cells.

The degree of protein tyrosine phosphorylation is, in part, regulated by protein kinases. Recent studies emphasize the role of specific stress-activated kinases in mediating protein phosphorylation during cell injury (2, 17, 40, 50). c-Src, one of the first protooncogenes to be described, is a tyrosine kinase with multiple protein substrates (10, 50). c-Src phosphorylates β-catenin, paxillin, and vinculin (10), three proteins that regulate cell contact site function. β-Catenin, a member of the catenin family, resides at the intracellular face of the adherens junction and is juxtaposed between E-cadherin and the actin cytoskeleton (1). In addition to its structural role, β-catenin is also involved in nuclear signaling that results in cell proliferation (1, 48). c-Src-mediated protein tyrosine phosphorylation of the catenins regulates the adherens junction, the component of the tight junction that determines TER (3, 24,49, 52). Paxillin, a component of the focal adhesion complex, interacts directly with pp125 focal adhesion kinase (FAK), a key regulatory protein for cell-matrix attachment (12, 40, 43, 44,47). Tyrosine phosphorylation of FAK by c-Src promotes the subsequent tyrosine phosphorylation of paxillin, permitting the focal adhesion complex to be assembled (25, 47). Vinculin, an actin-binding protein, modulates both integrin-mediated cell adhesion and tight junction integrity at the cell-cell interface (1, 10,35). On the basis of their established role in regulating cell contact sites, we assessed the potential for ATP depletion to activate c-Src and examined the phosphotyrosine content of these three known c-Src substrates.

ATP depletion resulted in the rapid activation of c-Src as detected by an antibody specific for the activated form of the kinase (Fig. 3). Phosphorylation of Tyr416 activates c-Src, permitting it to phosphorylate tyrosine residues on substrate proteins (15,38). Importantly, the degree of Src tyrosine phosphorylation positively correlates with a direct, in vitro assay of Src-kinase activity in OK cells (45). In the present study, changes in protein tyrosine phosphorylation appear to be specific, since the phosphoserine content of β-catenin and vinculin were minimally affected by ATP depletion or recovery (Figs. 4 and 6). Although the effects of ischemia or ATP depletion on c-Src have not been previously reported, insults such as hypoxia (51), oxidant stress (2), and ultraviolet light exposure (17) also activate this protein kinase. In the current study, c-Src activation was associated with an increase in the phosphotyrosine content of β-catenin, paxillin, and vinculin during ATP depletion and/or recovery. Although some degree of tyrosine phosphorylation of these substrate proteins is required for normal cell-cell and cell-substrate interactions (26, 32), increased tyrosine phosphorylation of structural proteins by Src disrupts cell contacts (7, 39, 44, 48).

Src kinase activation did not result in simultaneous or equivalent tyrosine phosphorylation of the three Src substrate proteins evaluated in this study. This suggests that additional factors may regulate tyrosine phosphorylation during ATP depletion and recovery. These factors could include redistribution of c-Src and its substrates in a manner that alters their physical proximity, changes in phosphatase activity (responsible for protein dephosphorylation), and protein phosphorylation by non-Src protein tyrosine kinases [e.g., FAK (12, 40)]. In a recent study, Meriin and colleagues (33) showed that HSP72 modulates protein phosphatase activity after a hyperthermic insult. Despite an incomplete understanding of the factors that regulate protein phosphorylation, the present data support the hypothesis that c-Src activation and tyrosine phosphorylation of substrate proteins mediate disruption of contact sites in ATP-depleted cells.

In addition to the early activation of c-Src, ATP depletion resulted in a marked increase in steady-state c-Src content during late recovery (Fig. 2). This observation contrasts with the marked decrease in Yes kinase content during the same time period (Fig. 2). Upregulation of Src has been associated with cell proliferation (10, 18, 38,48) as well as with apoptotic cell death (30). Apoptosis has also been reported in renal epithelial cells subjected to ATP depletion (58), suggesting that an increase in c-Src content may have untoward consequences on renal cell survival.

By altering phosphatase and/or kinase activity, heat exposure could modulate the degree of protein tyrosine phosphorylation and ameliorate the functional consequences of ATP depletion on cell contact sites. In fact, prior heat stress significantly inhibited the activation of c-Src during ATP depletion and early recovery (Fig. 3) and decreased the phosphotyrosine content of all three c-Src substrate proteins (Figs.4-6). Because prior heat stress protects cell contact sites during ATP depletion (7), we hypothesized that a HSP72 might interact with c-Src. Although binding between these two proteins has not been previously reported, their interaction might be overlooked, since HSP72 and c-Src do not coimmunoprecipitate under normal circumstances (Fig. 7). When HSP72 content is increased by heat stress, however, interaction with c-Src is clearly observed. Prior observations support a potential association between heat stress proteins and protein kinases. HSP90, a protein involved in the nuclear translocation of ligands, including the steroid receptor, also binds c-Src (11,60). In addition, HSP72 has been shown to regulate stress-activated protein kinase activity (23, 33).

Importantly, the present data suggest that upregulation of HSP72 content alone, in the absence of ATP depletion, is not sufficient to inactivate c-Src (Fig. 3) or to alter the content of tyrosine-phosphorylated substrate proteins (Figs. 4-6). Only when the interaction between c-Src and HSP72 is markedly increased (i.e., after ATP depletion) are Src activation and protein tyrosine phosphorylation reduced. These observations suggest that the interaction between HSP72 and c-Src may be causally linked to activation of the kinase. How could HSP72 prevent Src activation? By “masking” the Tyr416 residue, HSP72 could prevent tyrosine phosphorylation, a prerequisite for Src activation. Alternatively, HSP72 could compete for tyrosine phosphorylation with c-Src or its substrate proteins. The latter is an attractive hypothesis, since a tyrosine phosphorylation site that facilitates the translocation of HSP72 from the cytosol to the nucleus has recently been identified (27). However, a direct effect of HSP72 on c-Src activation may not be required to explain the observed decrease in protein tyrosine phosphorylation. By preventing c-Src from associating with its substrate proteins, HSP72 could inhibit tyrosine phosphorylation. Although the present study focuses on its kinase function, c-Src also acts as an adapter molecule that facilitates protein-protein interactions. After binding to c-Src, HSP72 could affect its nonkinase-dependent functions by altering the interaction between structural proteins and the actin cytoskeleton (21,49).

We suggest that disruption of cell contact sites during pharmacological ischemia may be amenable to modulation by endogenous cytoprotectant proteins such as HSP72. By inhibiting c-Src activation and the subsequent tyrosine phosphorylation of specific structural cell proteins, prior induction of HSP72 could prevent or attenuate the loss of cell-cell interactions caused by ATP depletion. Preservation of the actin cytoskeleton and the maintenance of cell-cell interactions inhibit apoptosis (44) and may explain the recent observation that prior heat stress prevented apoptosis in ATP-depleted renal epithelial cells (59). Identification of the binding partners of this endogenous cytoprotectant could produce novel strategies for preventing cell injury and/or for accelerating cell repair.

Acknowledgments

This research was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-53387 (to S. C. Borkan) and DK-5298 (to J. H. Schwartz) and a supplemental award from the American Society of Nephrology (to S. C. Borkan).

Footnotes

  • Address for reprint requests and other correspondence: S. C. Borkan, Evans Biomedical Research Center, Renal Section, Rm. 547, 650 Albany St., Boston, MA 02118-2518 (E-mail:sborkan{at}bu.edu).

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract