Molecular Dissection of PINCH-1 Reveals a Mechanism of Coupling and Uncoupling of Cell Shape Modulation and Survival*

How cells couple and uncouple regulation of cellular processes such as shape change and survival is an important question in molecular cell biology. PINCH-1, a widely expressed protein consisting of five LIM domains and a C-terminal tail, is an essential focal adhesion protein with multiple functions including regulation of the integrin-linked kinase (ILK) level, cell shape, and survival signaling. We show here that the LIM1-mediated interaction with ILK regulates all these three processes. By contrast, the LIM4-mediated interaction with Nck-2, which regulates cell morphology and migration, is not required for the control of the ILK level and survival. Remarkably, a short 15-residue tail C-terminal to LIM5 is required for both cell shape modulation and survival, albeit it is not required for the control of the ILK level. The C-terminal tail not only regulates PINCH-1 localization to focal adhesions but also functions after it localizes there. These findings suggest that PINCH-1 functions as a molecular platform for coupling and uncoupling diverse cellular processes via overlapping but yet distinct domain interactions.

PINCH-1 siRNA and DNA Constructs-The sequence of the PINCH-1-specific siRNA was previously described (9). The 21-nucleotide synthetic siRNA duplex was prepared by Dharmacon Research. For rescue experiments we generated PINCH-1 expression vectors encoding transcripts that are resistant to the PINCH-1 siRNA. To do this we introduced silent mutations into the siRNA-targeted site (5Ј-AAGGTGAT-GTGGTCTCTGCTC-3Ј, in which the underlined T and G were replaced with C) in the PINCH-1 cDNA sequence using a QuikChange TM sitedirected mutagenesis system (Stratagene). The sequence of the resulting "siRNA resistant" PINCH-1 cDNA, which encodes a protein sequence that is identical to that of wild type PINCH-1, was confirmed by automated DNA sequence. The Q40A and R197A/R198A point mutations were introduced into the siRNA-resistant PINCH-1 expression vector using the QuikChange TM site-directed mutagenesis system (Stratagene). The LIM5 deletion mutation, the C-terminal tail deletion mutation, and the PINCH-1/PINCH-2 C-terminal tail swap mutation were generated by PCR using the siRNA-resistant PINCH-1 cDNA (and PINCH-2 cDNA for the swap mutation) as templates. The mutations were confirmed by automated DNA sequence.
DNA and siRNA Transfection-HeLa cells were transfected with the siRNA resistant expression vectors encoding the wild type or mutant forms of PINCH-1 as specified in each experiment using Lipofectamine PLUS (Invitrogen). Two days after DNA transfection, the cells were transfected with the PINCH-1 siRNA or an irrelevant 21-nucleotide synthetic RNA duplex as a control using Oligofectamine (Invitrogen) as previously described (9). Cell spreading, Akt phosphorylation, and apoptosis were analyzed 48 h after siRNA transfection.
Cell Spreading-Cell spreading was analyzed as previously described (9,31). Briefly, cells were plated in Opti-MEM I serum-free medium (Invitrogen) in fibronectin (10 g/ml)-coated 96-well plates. After incubation at 37°C under a 5% CO 2 , 95% air atmosphere for 30 min, cells were observed under an Olympus IX70 microscope and recorded with a digital camera. Unspread cells were defined as round cells, whereas spread cells were defined as cells with extended processes as previously described (9,(31)(32)(33). The percentage of cells adopting spread morphology was quantified by analyzing at least 300 cells from four randomly selected fields (Ͼ80 cells/field).
Apoptosis-Cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Apoptosis was analyzed 48 h after siRNA transfection using terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) and caspase-3 assays as previously described (34). The TUNEL assay was performed using an in situ cell death detection kit (Roche Applied Science). The caspase-3 activities were measured using fluorogenic caspase-3 substrate VII (Ac-DEVD-aminofluoromethylcoumarin) from Calbiochem following the manufacturer's protocol.
GFP-PINCH-1 Expression and Immunofluorescent Staining-Expression vectors encoding GFP-tagged wild type and mutant forms of PINCH-1 were generated by inserting the corresponding PINCH-1 cDNA sequences into the pEGFP-C2 vector (Clontech). HeLa cells were transfected with the GFP-PINCH-1 expression vectors using Lipofectamine PLUS (Invitrogen). The transiently transfected cells were plated on fibronectin-coated coverslips 1 day after DNA transfection. After incubation overnight at 37°C under a 5% CO 2 , 95% air atmosphere, the cells were fixed and stained with mouse primary antibodies against ILK (clone 65.1) or paxillin (clone 349). Mouse monoclonal antibodies were detected with Rhodamine Red TX -conjugated antimouse IgG antibodies.

PINCH-1 Functions in the Maintenance of the ILK Protein
Level via Its Interaction with ILK-Consistent with the previous studies (9), suppression of PINCH-1 expression in HeLa cells (Fig. 1A, compare lane 2 with lane 1) significantly reduced the protein level of ILK (Fig. 1C, compare lane 2 with lane 1). To test whether PINCH-1 functions in the maintenance of the ILK protein level via binding to ILK, we expressed the wild type PINCH-1 (Fig. 1A, lane 4) and an ILK binding-defective PINCH-1 mutant in which Gln-40 located at the ILK binding interface was replaced with alanine (Fig. 1A, lane 3) in PINCH-1 knockdown cells. The Q40A mutation ablates the ILK binding but does not alter the overall structure of LIM1 (21, The percentage of the cells that adopted spread morphology was quantified as described under "Experimental Procedures" (H). Data represent the means Ϯ S.D. 24). Western blotting analyses showed that expression of the wild type PINCH-1 but not that of the ILK binding-defective point mutant restored the level of ILK (Fig. 1C, lanes 3 and 4).
In parallel experiments, an equal amount of actin was detected in all cell lysates (Fig. 1B). These results confirm the specificity of the ILK down-regulation induced by the PINCH-1 siRNA. Furthermore, they demonstrate that PINCH-1 functions in the maintenance of the ILK level through its interaction with ILK.
Expression of the Wild Type PINCH-1, but Not the ILK Binding-defective Point Mutant, in PINCH-1 Knockdown Cells Rescues the Defects in Cell Spreading and Survival-We next sought to identify the PINCH sites that are involved in the regulation of cell spreading and survival signaling. To this end, we first compared the effects of the wild type PINCH-1 and the ILK binding-defective point mutant on cell spreading. Unlike cells expressing a normal level of PINCH-1 (Fig. 1, D and H), PINCH-1 knockdown cells exhibited impaired ability to spread (Fig. 1, E and H). Re-expression of the wild type PINCH-1 (Fig.  1, G and H) but not that of the ILK binding-defective point mutant (Fig. 1, F and H) significantly improved the ability of cells to spread in response to cell-ECM adhesion. These results indicate that the interaction with ILK is essential for PINCH-1-mediated cell spreading.
To assess the significance of the ILK binding in cell survival signaling, we analyzed the activating phosphorylation of Akt, a key signaling intermediate that protects cells from apoptosis (35)(36)(37). Consistent with previous studies (9), knockdown of PINCH-1 (Fig. 2D, lanes 9 and 10) markedly inhibited the insulin-like growth factor 1-induced phosphorylation of Akt at Ser-473 ( Fig To further analyze this, we measured the activity of caspase-3, a key mediator of apoptosis, in PINCH-1 knockdown cells as well as in cells that re-express the wild type or the ILK binding-defective point mutant of PINCH-1. As expected, the level of caspase-3 activity was elevated in PINCH-1 knockdown cells (Fig. 3A, Vector/PINCH-1(Ϫ)). Transient expression of the wild type PINCH-1 (Fig. 3A, WT/PINCH-1(Ϫ)) but not that of the ILK binding-defective Q40A point mutant (Fig. 3A, Q40A/PINCH-1(Ϫ)) significantly reduced the caspase-3 activity. Consistent with this, the percentage of apoptotic cells measured by TUNEL was significantly increased in the PINCH-1 knockdown cells (Fig. 3B). Transient expression of the wild type PINCH-1 but not that of the ILK binding-defective Q40A point mutant in the PINCH-1 knockdown cells significantly reduced the percentage of apoptotic cells (Fig. 3B). These results suggest that the interaction of PINCH-1 with ILK is crucial for protection of cells from apoptosis.
The Nck-2 Binding Is Dispensable for PINCH-1-mediated Cell Survival Signaling-In addition to interacting with ILK through the N-terminal-most LIM1 domain, PINCH-1 interacts with Nck-2, an SH2/SH3-containing adaptor protein, via its C-terminal LIM4 domain (25,26). It is increasingly clear that the interaction between PINCH-1 and Nck-2 is involved in regulation of actin cytoskeleton remodeling (25,26,28,38,39). It has not been determined, however, whether or not the binding of PINCH-1 to Nck-2 plays a role in PINCH-1-mediated cell survival signaling. To test this we expressed a Nck-2 binding-defective PINCH-1 mutant (RRAA) in which Arg-197-Arg-198, located at the Nck-2 binding interface (26), were replaced with alanine residues in PINCH-1 knockdown cells (Fig. 2D, lanes 5 and 6). The R197A/R198A mutations ablate the Nck-2 binding but do not alter the overall structure of the LIM4 domain (26,39). Transient expression of the Nck-2 binding-defective mutant (Figs. 2, A and B, lane 6), like that of the wild type PINCH-1 (Figs. 2, A and B, lane 8), significantly enhanced phosphorylation of Akt. Furthermore, it substantially reduced the activity of caspase-3 that was induced by the depletion of PINCH-1 (to a level that was comparable with what was seen in cells transiently expressing the wild type PINCH-1) (Fig. 3A, RRAA/PINCH-1(Ϫ)). TUNEL analyses confirmed that the Nck-2 binding-defective mutant effectively rescued the survival defect induced by the depletion of PINCH-1 (Fig. 3B). These results suggest that the interaction of PINCH-1 with Nck-2, unlike that with ILK, is dispensable for cell survival signaling.
The PINCH-1 C-terminal Region Is Crucial for Survival Signaling and Cell Spreading but Not Maintenance of the ILK Level-We next expressed a PINCH-1 mutant (⌬LIM5) in which the C-terminal LIM5 and non-LIM tail were deleted in PINCH-1 knockdown cells. The wild type PINCH-1 was expressed in PINCH-1 knockdown cells as a positive control. The expression of ⌬LIM5 (Fig. 4A, lane 4) and the wide type PINCH-1 (Fig. 4A, lane 3) in the PINCH-1 knockdown cells (Fig. 4A, lane 2) was confirmed by Western blotting with an anti-PINCH-1 antibody. Probing the cells with antibodies specific for phospho-Akt (Ser-473) or phospho-Akt (Thr-308) showed that the level of Ser-473-and Thr-308-phospho-Akt, as expected, was significantly reduced in PINCH In control experiments neither the total Akt protein level (Fig. 4F, lanes 1-4) nor the actin level (Fig. 4B, lanes 1-4) was altered, confirming the specificity of the rescue experiments. Consistent with the defect in Akt phosphorylation, expression of ⌬LIM5 failed to inhibit the activation of caspase-3 induced by the depletion of PINCH-1 (Fig. 3C). These results suggest that the C-terminal region of PINCH-1, like the N-terminal LIM1, is required for cell survival signaling. Importantly, although ⌬LIM5 failed to rescue the defect in survival signaling, expression of ⌬LIM5, like that of the wild type PINCH-1, restored the level of ILK protein (Fig. 4C, lanes 3 and 4).
To further analyze this we expressed a PINCH-1 mutant containing all five LIM domains but lacking the C-terminalmost non-LIM tail in PINCH-1-deficient cells (Fig. 4A, lane 5). The expression of the C-terminal-most tail deletion mutant (CD) restored the level of ILK protein (Fig. 4C, lane 5) but not that of Akt phosphorylation (Figs. 4, D and E, lanes 5). Consistent with this, expression of the CD mutant failed to significantly reduce the activation of caspase-3 (Fig. 3C) and the percentage of the TUNEL-positive (apoptotic) cells (Fig. 3D) induced by the depletion of PINCH-1.
We next assessed the role of the C-terminal region in cell spreading. The results showed that expression of ⌬LIM5 (Fig.  5, D and G) or CD (Fig. 5, E and G), unlike that of the wild type PINCH-1 (Fig. 5, C and G), failed to rescue the cell-spreading defect induced by the depletion of PINCH-1 (Fig. 5, B and G). Thus, despite their ability to restore ILK protein level, neither ⌬LIM5 nor CD was able to rescue the defect in cell shape modulation induced by the depletion of PINCH-1. This result together with those described earlier suggests that PINCH-1 not only functions in the control of the ILK level (which is mediated by its interaction with ILK) but also is directly involved in the cell shape modulation and survival signaling (which requires both the N-terminal LIM1 domain and the C-terminal region).
The C-terminal-most Tail Not Only Regulates PINCH-1 Localization to Cell-ECM Adhesions but Also Is Required for PINCH-1 Functions at Cell-ECM Adhesions-Proper subcellular localization is critical for normal protein functions. Thus, the failure of the PINCH-1 C-terminal-most tail deletion mu- tant to rescue cell survival and spreading defects could result from its deficiency in focal adhesion localization. To compare the ability to localize to cell-ECM adhesions, we expressed GFP-tagged wild type PINCH-1 (Fig. 6B, lane 2) and the CD mutant (Fig. 6B, lane 3) in HeLa cells and plated them on fibronectin-coated surface. Immunofluorescent staining of the cells with a monoclonal anti-ILK antibody showed that, as expected, GFP-PINCH-1 was readily co-clustered with ILK in cell-ECM adhesions (Fig. 6, C and D). By contrast, the ability of the C-terminal tail deletion mutant to localize to cell-ECM adhesions was compromised (Fig. 6, E and F). The reduced efficiency of the C-terminal tail deletion mutant to localize to cell-ECM adhesions was not caused by a general deficiency of cell-ECM adhesions, as abundant cell-ECM adhesions were detected in these cells (Fig. 6F). Similar results were obtained by immunofluorescent staining of the cells with a monoclonal antibody to paxillin, another marker of cell-ECM adhesions (data not shown). These results suggest that the C-terminal non-LIM tail plays an important role in the regulation of PINCH-1 localization to cell-ECM adhesions.
To test whether the C-terminal non-LIM tail is merely involved in the regulation of cell-ECM adhesion localization or it is also required for the functions of PINCH-1 after its arrival at cell-ECM adhesions, we replaced the PINCH-1 C-terminal non-LIM tail with that of PINCH-2 (40, 41) (Fig. 6A). HeLa cells were transfected with the expression vector encoding the GFPtagged C-terminal tail swap mutant (CS). The expression of GFP-CS was confirmed by Western blotting (Fig. 6B, lane 4). Immunofluorescent analyses showed that GFP-CS (Fig. 6G), like the wild type PINCH-1, readily localized to cell-ECM ad-hesions where abundant ILK (Fig. 6H) and paxillin (not shown) were detected.
To test whether the C-terminal tail-swap mutant could functionally substitute PINCH-1 in cell survival signaling and shape modulation, we expressed it in PINCH-1 knockdown cells (Fig. 4A, lane 6). Expression of the CS mutant, as expected, effectively relieved the suppression of the ILK level induced by the depletion of PINCH-1 (Fig. 4C, compare lane 6 with lane 2). However, despite its ability to restore the ILK level and to localize to cell-ECM adhesions, the CS mutant failed to rescue the defect in Akt phosphorylation induced by the depletion of PINCH-1 (Figs. 4, D and E, lane 6). Consistent with this, the caspase-3 activity level remained elevated (Fig. 3C), and the percentage of apoptotic (TUNEL positive) cells was not significantly reduced (Fig. 3D) after the expression of the CS mutant in the PINCH-1 knockdown HeLa cells. Comparison of cell spreading showed that the CS mutant, like the CD mutant, failed to rescue the cell spreading defect induced by the depletion of PINCH-1 in HeLa cells (Fig. 5, F  and G). These results suggest that the PINCH-1 C-terminalmost tail functions not only in the efficient localization of PINCH-1 to cell-ECM adhesions but also in cell shape modulation and survival signaling after PINCH-1 localizes to the adhesion sites. DISCUSSION One of the key questions in molecular cell biology is how cells couple and uncouple diverse cellular processes such as shape modulation and survival. PINCH-1 is a widely expressed focal adhesion adaptor that is required for multiple processes including regulation of the ILK level, modulation of cell shape, and transduction of survival signals (9,13). Using a structurebased reverse genetic rescue approach, we have molecularly dissected the functions of PINCH-1. The results reveal a mechanism in which PINCH-1, through its interactions with distinct binding partners, serves as a molecular platform for coupling and uncoupling diverse cellular processes via distinct domaindomain interactions.
By expression of an ILK binding-defective PINCH-1 point mutant (Q40A) in PINCH-1 knockdown cells, we have demonstrated that PINCH-1 function in the maintenance of the ILK protein level through its interaction with ILK. This result together with our previous finding that down-regulation of the ILK level is mediated at least in part by proteasome-mediated protein degradation (9) suggests that the binding of PINCH to the ankyrin domain of ILK probably helps to stabilize ILK and, hence, prevents its degradation. This may provide an explanation as to why the formation of the PINCH-1-ILK complex occurs early and precedes their localization to cell-ECM adhesions (24). Elevation of ILK levels is closely associated with and probably is an important causal factor for the progression of several human diseases including cancers and renal failure (for review, see Refs. 2,4,16,19,and 42). The importance of the PINCH-1 binding to the maintenance of the ILK level suggests that it could serve as a useful target for the therapeutic control of the ILK protein level and, hence, the progression of human diseases involving abnormal expression of ILK.
A second important finding of the current study is that the cell spreading defect induced by the loss of PINCH-1 can be uncoupled from the survival defect induced by the loss of PINCH-1 and vice versa. PINCH-1 is required for both cell shape modulation and survival signaling (9,13). We previously found that the interaction of PINCH-1 with Nck-2 functions in actin cytoskeletal regulation (39). The data presented in this paper demonstrate that the interaction with Nck-2 is not required for cell survival (Figs. 2 and 3). Thus, although PINCH-1 is intimately involved in the regulation of both the actin cytoskeleton and survival signaling, they can be uncoupled from each other. Based on these results, it is attractive to propose a model in which cells can either coordinately or separately regulate the morphological change and survival. Down-regulation of the interaction of PINCH-1 with ILK inhibits cell spreading as well as survival signaling, resulting in coordinated regulation of these two processes. On the other hand, down-regulation of the interaction of PINCH-1 with Nck-2 inhibits cell spreading and migration but not survival and, hence, the uncoupling of these two processes. The ability of PINCH-1 to either coordinately or separately regulate morphological change and survival through interactions mediated by distinct sites provides a versatile system for cells to control these processes.
The studies presented in this paper have identified a new site that is crucial for PINCH-1 function. By expression of a PINCH-1 mutant in which the short C-terminal tail is deleted in PINCH-1 knockdown cells, we have shown that it is indispensable for PINCH-1-mediated regulation of cell spreading and survival signaling. The functions of the C-terminal tail are 2-fold. It facilitates PINCH-1 localization to cell-ECM adhesions. Although there is no doubt that this is important for the functions of PINCH-1, the C-terminal tail apparently is also required after PINCH-1 localizes to cell-ECM adhesions, as replacing it with a sequence derived from the C terminus of PINCH-2 restores the PINCH-1 localization to cell-ECM adhesion but not its functions in cell spreading and survival signaling. The identification of the C-terminal tail as a key site that regulates both the localization and the post-localization signaling of PINCH-1 should facilitate future studies aimed at identifying additional PINCH-1 binding partners involved in these processes.
Although the PINCH-1 C-terminal tail participates in cell shape modulation and survival signaling, it is not required for the maintenance of the ILK protein level. Thus, among the three PINCH-1 functional sites (LIM1, LIM4, and the C-terminal tail), only LIM1 is involved in the maintenance of the ILK level. Cell survival signaling involves at least two sites (i.e. LIM1 and the C-terminal tail). PINCH-1-mediated modulation of cell morphological change appears to be most demanding among the three PINCH-1-mediated processes. It requires all three PINCH-1 functional sites. The studies described in this paper have firmly established the functional importance and specificity of each of these three sites. Recent studies have demonstrated that Ras suppressor 1 (RSU-1) interacts with the C-terminal region of PINCH-1 (11,12). The PINCH-1-RSU-1 interaction, like the PINCH-1-ILK and ILK-parvin interactions (for review, see Refs. 2-4, 22, and 23), is evolutionally conserved and functionally important for embryonic development (11). Interestingly, Dougherty et al. (12) have shown that PINCH-2 does not interact with RSU-1.
Bock-Marquette et al. (18) recently reported that PINCH-1 and ILK function as key regulators of cardiac cell migration, survival and repair. Furthermore, they have shown that PINCH-1, through its C-terminal region, interacts with thymosin ␤4 (18). Although the specific thymosin ␤4-binding site within the PINCH-1 C-terminal region remains to be determined, PINCH-1 could function in cardiomyocytes by linking thymosin ␤4 with ILK and other binding partners through its distinct binding sites and, hence, promotes cardiac cell migration, survival, and consequently, cardiac repair after myocardial infarction (18).
Protein-protein interactions provide the foundation for control of cellular architecture and signal transduction (43). With the arrival of the post-genomics era, it has become increasingly clearly that a large number of key protein-protein interactions are mediated by a relatively small number of scaffolding proteins, which through its multiple proteinbinding motifs, provide a hub or platform for protein interactions. The studies presented in this paper together with previous studies suggest that PINCH-1, through interactions mediated by LIM1, LIM4, the C-terminal tail, and perhaps other yet to be identified sites, functions as one of the key protein binding platforms at the cell-ECM adhesions controlling cell shape modulation and survival.