The C-terminal Kinase Domain of the p34cdc2 -related PITSLRE Protein Kinase (p110C) Associates with p21-activated Kinase 1 and Inhibits Its Activity during Anoikis*

The PITSLRE protein kinases are parts of the large family of p34cdc2 -related kinases. During apoptosis induced by some stimuli, specific PITSLRE isoforms are cleaved by caspase to produce a protein that contains the C-terminal kinase domain of the PITSLRE proteins (p110C). The p110C induces apoptosis when it is ectopically expressed in Chinese hamster ovary cells. In our study, similar induction of this p110C was observed during anoikis in NIH3T3 cells. To investigate the molecular mechanism of apoptosis mediated by p110C, we used the yeast two-hybrid system to screen a human fetal liver cDNA library and identified p21-activated kinase 1 (PAK1) as an interacting partner of p110C. The association of p110C with PAK1 was further confirmed by in vitro binding assay, in vivo coimmunoprecipitation, and confocal microscope analysis. The interaction of p110C with PAK1 occurred within the residues 210–332 of PAK1. Neither association between p58 PITSLRE or p110 PITSLRE and PAK1 nor association between p110C and PAK2 or

The PITSLRE protein kinases are parts of the large family of p34cdc2-related kinases. During apoptosis induced by some stimuli, specific PITSLRE isoforms are cleaved by caspase to produce a protein that contains the C-terminal kinase domain of the PITSLRE proteins (p110C). The p110C induces apoptosis when it is ectopically expressed in Chinese hamster ovary cells. In our study, similar induction of this p110C was observed during anoikis in NIH3T3 cells. To investigate the molecular mechanism of apoptosis mediated by p110C, we used the yeast two-hybrid system to screen a human fetal liver cDNA library and identified p21-activated kinase 1 (PAK1) as an interacting partner of p110C. The association of p110C with PAK1 was further confirmed by in vitro binding assay, in vivo coimmunoprecipitation, and confocal microscope analysis. The interaction of p110C with PAK1 occurred within the residues 210 -332 of PAK1. Neither association between p58 PITSLRE or p110 PITSLRE and PAK1 nor association between p110C and PAK2 or PAK3 was observed. Anoikis was increased and PAK1 activity was inhibited when NIH3T3 cells were transfected with p110C. Furthermore, the binding of p110C with PAK1 and inhibition of PAK1 activity were also observed during anoikis. Taken together, these data suggested that PAK1 might participate in the apoptotic pathway mediated by p110C.
Apoptosis is a form of altruistic cell suicide, which is involved in many physiological processes, including tissue homeostasis, embryonic development, and immune response (1,2). It is becoming increasingly clear that cell cycle regulators such as the p34cdc2 gene family could influence apoptosis (3)(4)(5)(6)(7). The PITSLRE protein kinases, which are coded by the gene localized to human chromosome 1p36.3 and a syntenic region of mouse chromosome 4, are parts of the large family of p34cdc2related kinases (8 -14). There are at least 20 PITSLRE protein kinase isoforms, which are differentially expressed in mamma-lian tissues and regulate diverse cellular functions, including the cell cycle control, tumorigenesis, the regulation of RNA splicing or transcription, and so on (8, 10 -11,13-26). Studies indicated that several of the PITSLRE protein kinase isoforms might serve as effectors in apoptotic signaling pathways (15)(16)(17)(18).
During apoptosis induced by Fas or tumor necrosis factor, the action of p110C, a novel processed PITSLRE isoform resulting from the proteolysis of specific larger PITSLTR isoforms, significantly increased (15,18). Previous studies of our group and others (15,26) showed that apoptosis was increased when p110C was ectopically expressed in Chinese hamster ovary cells or SMMC 7721 hepatocarcinoma cells. These data suggested that p110C might play an important role in mediating apoptosis. In this study, we also detected the expression of p110C in NIH3T3 cells during anoikis, which is a form of apoptosis induced by the disruption of cell-matrix interaction. In agreement with the previous studies, a similar induction of p110C was observed during anoikis in NIH3T3 cells. It seems that p110C could participate in apoptosis induced by different factors. However, little is known about the mechanism of p110C-mediated apoptotic signaling pathway. Previous studies of the p58 PITSLRE or p110 PITSLRE isoforms showed that they could interact with different proteins and mediate different cellular functions. For example, p110 PITSLRE isoform interacted with the RNA-binding protein RNPS1, RNA polymerase II, and multiple transcriptional elongation factors, regulating some aspects of RNA splicing or transcription in proliferating cells (21,22), whereas the p58 PITSLRE isoform interacted with cyclin D3 in the G 2 /M phase of the cell cycle, regulating G 2 /M phase cell cycle progression (24,25). Thus, the identification of cellular proteins that do interact with p110C, physically and/or functionally, should be a useful approach for defining the cellular functions and regulatory mechanisms of p110C in apoptotic signaling pathways. To investigate this issue, the yeast two-hybrid system with p110C as bait was used to screen a human fetal liver cDNA library. As a result, PAK1 1 (p21activated kinase 1) was identified as a p110C-associated pro-tein via the domain (aa 210 -332). This association of p110C with PAK1 was specific, as verified by the inability of PAK2 or PAK3 to associate with p110C and the inability of PAK1 to associate with p58 PITSLRE or p110 PITSLRE . Anoikis was increased and PAK1 activity was inhibited when NIH3T3 cells were transfected with p110C. Furthermore, the association of p110C, with PAK1 and inhibition of PAK1 activity, was also observed when anoikis was induced in NIH3T3 cells. Taken together, the data suggested that PAK1 might play a role in the regulation of p110C-mediated apoptotic signaling pathway.

EXPERIMENTAL PROCEDURES
Cell Lines and Reagents-The NIH3T3 cells and SMMC 7721 cells were obtained from the Institute of Cell Biology, Academic Sinica. The mammalian expression vector pcDNA3 was from Invitrogen (Carlsbad, CA). MATCHMAKER LexA two-hybrid system and human fetal liver MATCHMAKER LexA cDNA library were products of Clontech (Palo Alto, CA). The rabbit polyclonal anti-PITSLRE antibody and the goat polyclonal anti-␥-PAK antibody were purchased from Santa Cruz Biotechnology. The rabbit polyclonal anti-PAK1 and the anti-rabbithorseradish peroxidase secondary antibodies were purchased from New England BioLabs. The rabbit polyclonal anti-PAK3 antibody was from Calbiochem. Protein G-agarose, glutathione-Sepharose beads, and the mouse monoclonal anti-HA (12CA5) antibody were purchased from Roche Applied Science. Leupeptin, aprotinin, PMSF, poly-HEMA, and MBP were purchased from Sigma Chemical Co. [␥-32 P]ATP (Ͼ3000 Ci/mM), [ 35 S]methionine, Hybond PVDF membrane, and the enhanced chemiluminescence (ECL) assay kit were purchased from Amersham Biosciences. Annexin V-FITC recombinant protein was purchased from Bender MedSystems. Other reagents were commercially available in China.
Cell Culture, Transient Transfection, and Induction of Anoikis-NIH3T3 cells were cultured with DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. SMMC 7721 cells were cultured with RPMI 1640 supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. They were maintained in a 37°C humidified atmosphere containing 5% CO 2 in air. For transient transfection assays, the cells (5 ϫ 10 5 ) were grown on 60-mm tissue culture dishes. The following day, the cells were transfected with relevant plasmids. The transfection was performed with LipofectAMINE TM according to the manufacturer's recommendations. After 48 h, the cells were harvested for further analysis. For induction of anoikis (27), NIH3T3 cells were seeded in 3 ϫ 10 6 cells/100-mm tissue culture dish. After culturing for 6 h, the medium was changed to serum-free DMEM. Following a further culturing for 18 h, the cells were trypsinized and plated in cell suspension dishes previously coated with 10 mg/ml poly-HEMA. After indicated time, the cells were collected for further investigation. Apoptosis was evaluated by staining with propidium iodide (PI) to detect subdiploid DNA content and by using annexin V to detect cell surface phosphatidylserine (PS). For PI staining, the cells (10 6 ) were washed twice, collected, and then fixed with 70% ethanol at Ϫ20°C overnight. After staining with 10 g/ml propidium iodide, the cells were analyzed on a BD Biosciences fluorescence-activated cell sorter. The percentage of apoptosis was quantitated from sub-G 1 events. To use annexin V to detect cell surface PS, the cells were washed twice in PBS and then resuspended in binding buffer (10 mM Hepes/ NaOH, pH 7.4, 140 mM NaCl, 2.5 mM CaCl 2 ). 195 l of cell suspension containing 10 5 cells was taken, and 5 l of annexin V-FITC was added, mixed, and incubated for 10 min in the dark. After that, the cells were wash twice in PBS and resuspended in 190 l of binding buffer. 10 l of 20 g/ml PI was added and then analyzed by fluorescence-activated cell sorting. Apoptotic cells could be stained with annexin V but not PI.
Western Blot Analysis-Western blot experiments were used to measure certain proteins. Briefly, the cells were lysed in lysis buffer (120 mM Tris (pH 7.4), 135 mM NaCl, 1 mM EDTA, 1% Nonidet P-40, 0.1% SDS, 1 mM Na 3 VO 4 , 1 mM aprotinin, 1 mM PMSF). A total of 50 g of protein from each sample was electrophoresed by 12% SDS-PAGE and then transferred to a PVDF membrane. After blocking with PBS containing 5% nonfat milk and 0.1% Tween 20 overnight, the membrane was incubated with primary antibody at 4°C overnight. After washing with PBS containing 0.1% Tween 20 three times, each for 5 min, the membrane was then incubated with horseradish peroxidaselabeled secondary antibody for another 2 h at room temperature. The membrane was then developed using the enhanced chemiluminescent (ECL) detection system. Yeast Two-hybrid cDNA Library Screening and Protein-Protein Interaction Assay-MATCHMAKER LexA two-hybrid system was used to perform yeast two-hybrid screening according to the manufacturer's instruction. The p110C was used as bait to screen by the reporter genes LEU2 and lacZ. The transformed yeast was grown on plates lacking both histidine to select the bait plasmid and tryptophan to select the prey plasmid, respectively. The individual cDNA of interest was purified from the positive clones and re-transformed into yeast to confirm its specific interaction with p110C. 14 clones could activate both the LEU2 and lacZ reporter genes only when they contain pLexA-p110C. All of these clones were revealed to encode the fragment of p21-activated kinase 1 (PAK1) by sequence analysis. The quantitative ␤-galactosidase assay was performed to detect the protein-protein interaction on at least three independent colonies from the same clone, using o-nitrophenyl-␤-D-galactopyranoside as a substrate. Murine p53 fused with LexA as a bait and SV40 large T-antigen fused as a prey were used as a positive control.
In Vitro Binding Analysis-p110C was cloned into pcDNA3-GST vector. PAK1 and its mutants, PAK2, and PAK3 were cloned into pcDNA3 individually. Their transcripts were produced by using an in vitro transcription kit (Promega). Rabbit reticulocyte lysates from Promega were used to generate [ 35 S]Met-labeled proteins, which were used immediately for binding studies. The binding reactions were performed in protein binding buffer (20 mM Tris, pH 7.5, 50 mM NaCl, 10% glycerol, 10 mM NaF, 1% Nonidet P-40, 1 mM NaVO 4 , 10 g/ml aprotinin, 10 g/ml leupeptin, 1 mM PMSF) at 4°C for 2 h with constant mixing (28). The beads were washed three times with the same buffer, and the bound proteins were subjected to 12% SDS-PAGE analysis. The gel was then dried and autoradiographed.
Whole cell lysates were incubated with mouse normal IgG (Santa Cruz Biotechnology) or relevant antibody at 4°C for 2 h. Pre-equilibrated protein G-agarose beads (Roche Applied Science) were then added and collected by centrifugation after 2 h of incubation and then gently washed three times with the lysis buffer. The bound proteins were eluted by boiling in SDS sample buffer and resolved on a 12% SDS-PAGE gel. The proteins were transferred onto a PVDF membrane and analyzed using Western blots.

Induction of p110C Protein during Anoikis in NIH3T3
Cells-Previous studies indicated that transfection of p110C in Chinese hamster ovary cells did not influence mitosis and cell growth but could induce apoptosis (15). When the apoptosis of cells was induced by treating with Fas antibody or tumor necrosis factor, p110C protein was induced, which was not due to new protein synthesis but processing of some PITSLRE isoforms (15,18). These suggested that p110C might have an important role in apoptotic signaling pathway. To test whether the induction of p110C isoform was in general in response to different apoptotic inducers, NIH3T3 cell apoptosis was induced by detachment from the cell matrix (also known as anoikis), and then the expression of the PITSLRE protein isoforms was detected. Cells undergoing apoptosis could be recognized by flow cytometry by detecting a sub-G 1 peak using PI staining and phosphatidylserine (PS) on the outer leaflet of the cell membrane using an annexin V label. About 26% of the NIH3T3 cells cultured in a poly-HEMA-coated dish (detached from the extracellular matrix) under serum-free condition for 10 h underwent apoptosis (Fig. 1A). Western blot analysis using anti-PITSLRE antibody showed that p110C was markedly induced when the cells were induced during anoikis. This result was consistent with the previous studies (15,18). In addition, it was found that the induction of p110C did not result from the processing of p110 PITSLRE isoform but largely results from cleavage of the p58 PITSLRE isoform when anoikis was induced in NIH3T3 cells (Fig. 1B). Therefore, our study also showed the similar induction of p110C during anoikis in NIH3T3 cells.
Identification of PAK1 as p110C-associated Protein-To identify proteins that interact with p110C, the yeast two-hybrid system was used with p110C as bait, and a human fetal liver cDNA library was screened. We plated a pool of cells that contained 5 ϫ 10 7 primary library transformants at a multiplicity of 10 onto selection induction medium. Approximately 40 clones were obtained whose LEU2 and lacZ reporter genes were activated. Among them 14 clones could activate both the LEU2 and lacZ reporter genes only when they contained pL-exA-p110C. The DNA of interest was isolated and sequenced. Sequence homology search identified the encoded proteins. All of these peptides were positioned in-frame to the p21-activated kinase (PAK1) and contained a partly N-terminal deletion portion of PAK1.
PAK1 was first identified as a Rac-interacting protein (30). There are at least six known PAK isoforms (PAK1-6) that are differentially expressed in mammalian tissues (31)(32)(33). Based on their conserved structure, PAK1-3 are classed together as the Group I PAKs (34 -36). To investigate if PAK1 was the unique one in the Group I PAKs that could interact with p110C, a direct two-hybrid assay was performed. The other two Group I PAKs, PAK2 and PAK3, were co-transformed, respectively, with p110C into EGY48 (p8op-lacZ). Subsequent transformation results showed that neither of them permitted activation of the reporter genes, whereas PAK1, in the presence of p110C, did activate the two report genes. It suggested that the association between PAK1 and p110C was specific in the yeast two-hybrid system (data not shown). To further test this interaction, a GST pull-down experiment was performed. GST-p110C, PAK1, PAK2, and PAK3 were synthesized, isotopically labeled, and tested for interaction in vitro. After GST-p110C was incubated with PAK1, PAK2, or PAK3, the protein mixtures were bound to glutathione-Sepharose beads, washed, and subjected to SDS-PAGE. The resulting gel was then exposed. The results revealed that only PAK1 could bind GST-p110C, suggesting that this interaction was specific in vitro (Fig. 2).
Mapping of the PAK1 Region That Interacted with p110C-To identify which domain of PAK1 was important for this interaction, we generated a number of proteins corresponding to different domains of PAK1 (designed as FL, N, C, and F1-F7) as shown in Fig. 3A. Direct two-hybrid tests were performed to map out the domain required for the interaction, and ␤-galactosidase activities were measured to determinate the interaction (Fig. 3B). As shown in Fig. 3B, the interaction was detected when amino acids 210 -332 were present and was disrupted when the PAK1-F3 (aa 1-332) was further deleted into the PAK1-F6 (aa 240 -332) and PAK1-F7 (aa 210 -300) mutants. To further confirm this interaction, GST pull-down assay was performed. As shown in Fig. 3C, p110C could bind with the PAK1-F3 (aa 1-332) and PAK1-F5 (aa 210 -332). No interaction was detected for the PAK1-F6 (aa 240 -332). This suggested that the region within PAK1 that directly interacted with p110C was between amino acids 210 and 332, which consisted of part regulatory domain and part kinase domain of PAK1 (Fig. 3C).  Binding of p110C with PAK1 in Vivo-Previous studies indicated that p110C was produced from the cleavage of some PITSLRE kinases isoforms, such as p110 PITSLRE and p58 PITSLRE (15)(16)(17)(18), and p58 PITSLRE and p110 PITSLRE were the best studied and described among the PITSLRE protein kinase family. To further determine whether p110C interacts with PAK1 in vivo and whether the interaction was specific, the HA-p110C, HA-p58 PITSLRE , or HA-p110 PITSLRE was transiently expressed in NIH3T3 cells, and pcDNA3 was also transiently expressed in NIH3T3 cells as control. The expression of HA-p110C, HA-p58 PITSLRE , or HA-p110 PITSLRE was confirmed by a monoclonal antibody against HA epitope. The whole cell lysates, with the equal amount of proteins, were immunoprecipitated with normal mouse IgG or anti-HA monoclonal antibody, followed by immunoblot analysis using a specific anti-PAK1 antibody, anti-PAK2 antibody, or anti-PAK3 antibody. As shown in Fig. 4B, PAK1 was co-immunoprecipitated with HA-p110C, whereas PAK1 could not be detected in the control mouse IgG immunoprecipitation or p58 PITSLRE and p110 PITSLRE immunoprecipitated with anti-HA monoclonal antibody. We also found that PAK2 or PAK3 could not coimmunoprecipitate with HA-PITSLRE kinases (data not shown). Therefore, this result further verified that PAK1 did associate with p110C in vivo, and this interaction was specific.
Confocal Microscope Analysis of the p110C and PAK1-To address the subcellular interaction of p110C with PAK1, we co-transfected the SMMC 7721 cells, a human hepatocarcinoma cell line, with EGFP-p110C and DsRed-PAK1. 48 h after transfection, the cells were fixed and analyzed under confocal microscopy. Merging the separate projection images as green only (Fig. 5A) and red only (Fig. 5B) emission detection, we observed that the double-transfected cells contained yellow granules indicating co-localization of p110C and PAK1 (Fig. 5C).
Increase of Anoikis and Decrease of PAK1 Activity in Cells Transiently Transfected with p110C-We next tested whether p110C had an effect on anoikis and PAK1 activity. p110C or pcDNA3 as control were transfected to NIH3T3 cells. 24 h after transfection, anoikis was induced in the cells. The cells were analyzed sub-G 1 content with a FACScan flow cytometer to detect apoptosis. As shown in Fig. 6A, the apoptotic rates of NIH3T3/p110C cells (18.2% at 2 h, 32.1% at 4 h, and 51.1% at 6 h) were much higher than those of the mock transfected cells (2% at 2 h, 4.9% at 4 h, and 16.7% at 6 h). Then an immune complex kinase assay was used to detect PAK1 activity. The whole cell lysates from NIH3T3/pcDNA3 or NIH3T3/p110C cells containing equal amounts of total proteins were immunoprecipitated with an anti-PAK1 antibody and assayed for PAK1 kinase activity in which myelin basic protein (MBP) was used as a substrate. As shown in Fig. 6B, PAK1 activity was markedly inhibited with the presence of p110C. Thereby, the data suggested that p110C could enhance anoikis and inhibit the activation of PAK1 in mammalian cells.
Association of p110C with PAK1 and Inhibition of PAK1 Activity in Apoptotic NIH3T3 Cells-We have observed the induction of p110C during anoikis and the association of p110C with PAK1 during co-expression of p110C in NIH3T3 cells. However, the ectopic expression of p110C was not in a normal physiological situation, because p110C was only induced during apoptosis. To investigate whether p110C could interact with PAK1 during anoikis, the model mentioned above was used. 10 h after the cells were detached from cell matrix, a significant amount of p110C was found (Fig. 1B). Cell lysates from the cells treated for different time were subjected to immunoprecipitation with anti-PAK1 antibody followed by immunoblot analysis using a rabbit anti-PITSLRE polyclonal antibody. The rabbit polyclonal anti-PITSLRE antibody used for immunoblotting was raised against PITSLRE peptide, which was localized in C-terminal and conserved in all the PITSLRE isoforms, so that it could recognize all the PITSLRE isoforms in the NIH3T3 cells. As shown in Fig. 7A, the unique protein that co-immunoprecipitated with PAK1 was p110C. Thereby, it demonstrated that only p110C could interact with PAK1. These results were further verified when the cell lysates were subjected to immunoprecipitation with anti-PITSLRE antibody followed by immunoblot analysis using anti-PAK1 antibody. It was found that the binding of PAK1 with p110C was increased in an anoikis-dependent manner, whereas the level of PAK1 protein was not changed (Fig. 7B). Next, we detected PAK1 activity during anoikis. The in vitro kinase assay of PAK1 revealed that PAK1 activity was decreased after anoikis was induced in the cells, whereas this decrease was smaller than that in p110C-transfected cells (Fig. 7C). Together, these observations suggested that PAK1 could bind to the caspaseprocessed p110 PITSLRE isoform and its activity was inhibited when the cells were induced during anoikis. DISCUSSION PITSLRE protein kinase could be cleaved and activated under certain apoptotic stimulation (15)(16)(17)(18). However, little is know about the mechanism of p110C-inducing apoptosis. Recent studies showed that the caspase-processed p110 PITSLRE could interact with eIF3 during apoptosis (37). In this report, we found there was similar induction of p110C during the anoikis in NIH3T3 cells. To explore how p110C participated in apoptosis, we used p110C as bait in yeast two-hybrid system screens of a human fetal liver cDNA library and identified p21-activated kinase 1 (PAK1) as a new partner of p110C. The p110C-binding site within the PAK1 domain was located within residues 210 -332, which consisted of part regulatory domain and part kinase domain. The interaction of PAK1 and p110C was specific, because neither interaction between PAK1 and other PITSLRE kinases nor interaction between p110C and other Group I PAKs (PAK2 and PAK3) was detected. Anoikis was increased and PAK1 activity was inhibited in NIH3T3 cells transfected with p110C. Furthermore, this association and inhibition of PAK1 was observed in NIH3T3 cells during anoikis.
PAK1 was identified as a Rac-interacting protein from rat brain, which is the first mammalian member of PAKs (30). There are six known PAK isoforms (PAK1-6) that are differentially expressed in mammalian tissues (31)(32)(33). Based on their conserved structure, PAK1-3 are classed together as the Group I PAKs, whereas PAK4 -6 are classed together as the Group II PAKs (34 -36). PAKs are regulated by the small GTPases, Cdc42 and Rac, and could regulate diverse cellular functions, including gene expression, cytoskeletal actin assembly, mitogen-activated protein kinase pathways, neurite outgrowth, cell cycle control, control of phagocyte NADPH oxidase, and cell apoptosis (29, 38 -50). Structure analysis revealed that there are several conserved domains in Group I PAKs: (i) the N terminus of PAK1, which plays an important role in various FIG. 6. Increase of anoikis and decrease of PAK1 activity in NIH3T3 cells transiently transfected with p110C. A, elevated expression of p110C enhances anoikis. The NIH3T3 cells (5 ϫ 10 5 ) were grown on 60-mm tissue culture dishes to transfect with p110C or empty vector. 24 h after transfection, anoikis was induced for the indicated times, and then the cells were analyzed with FACScan flow cytometer to detect the apoptotic rate. The error bars indicate Ϯ S.E. of values pooled from three independent experiments. B, PAK1 activity was inhibited in cells transiently transfected with p110C. The whole cell lysates from NIH3T3/pcDNA3 or NIH3T3/p110C cells containing equal amounts of total proteins were immunoprecipitated with an anti-PAK1 antibody and assayed for PAK1 kinase activity in which myelin basic protein (MBP) was used as a substrate. The lower band shows that the amount of PAK1 used in the reaction was constant, as assayed by Western blot analysis. The figure is representative of three independent experiments performed. C, relative kinase activity of PAK1 was determined by quantitation of the labeled MBP bands with the ImageQuaNT software. The activity toward MBP in each experiment was normalized relative to that in NIH3T3 cells transfected with pcDNA3. cellular events such as transport, folding of proteins, and signal transduction; (ii) residues 70 -160 of PAK1, which include the CRIB domain that is responsible for binding of GTPase; (iii) residues 170 -230 of PAK1, which include glycoprotein E precursor signal transduction, the acidic domain, and proline-rich extension; and (iv) residues 370-end of PAK1, which include a kinase domain. In the present study, we identify that p110C interacted exclusively with PAK1 rather than with PAK2 or PAK3. The p110C-binding site was located in residues 210 -332 of PAK1, which correspond to the low homologous domain of the PAKs. Therefore, it might account for the specific interaction between p110C and PAK1.
The PITSLRE homologues exist in human, mouse, chicken, Drosophila, and Xenopus (11,52,53). There are at least 20 PITSLRE protein isoforms whose functions appeared to be linked with cell cycle progression, apoptotic signaling, and others (8 -26). Among them p110 PITSLRE and p58 PITSLRE are the most studied and described. The p110 PITSLRE isoform could be detected in all phases of the cell cycle, participating in some aspects of RNA processing or transcription. The p58 PITSLRE isoform, which resulted from internal initiation of translation controlled by an internal ribosome entry site of p110 PITSLRE , was mainly expressed in G 2 /M phase and might play an important role in cell cycle progression control. In recent years p110C has increasingly been the focus of study. The p110C was proteolyzed from specific, larger PITSLTR isoforms during apoptosis and could induce apoptosis. The p58 PITSLRE isoform but not the p110 PITSLRE isoform or p110C could interact with cyclin D3 (24,25), whereas the p110C isoform rather than the p58 PITSLRE isoform or the p110 PITSLRE isoform could associate with PAK1. Thus, it seems that these three PITSLRE isoforms could interact with different proteins and mediate different functions. Importantly, the N-terminal ends of the p110C, p58 PITSLRE , and p110 PITSLRE molecules are different. Thus, it is likely that the N terminus of PITSLRE may interfere or block the conformation of the C terminus, so that the p58 PITSLRE sequence in p110 PITSLRE could not interact with cyclin D3 and the p110C sequence in p58 PITSLRE or p110 PITSLRE could not interact with PAK1 either. Collectively, these data might explain, at least in part, why p58 PITSLRE or p110 PITSLRE could not induce apoptosis.
The Group I PAKs have both pro-apoptotic and anti-apoptotic functions. In Xenopus oocytes, active PAK could prevent apoptosis (44). PAK2, another member of the PAK family, was cleaved and activated by caspase 3. Caspase-activated PAK2 could induce apoptosis (45). Some research groups revealed that activated PAK1 could protect cells from apoptosis, and this was mediated, at least in part, through the suppression of the pro-apoptotic activity of Bad. However, other research groups have provided evidence that activated PAK1 can promote induced in NIH3T3 cells as described in Fig. 1, and then cell lysates from each sample were immunoprecipitated with a specific anti-PAK1 antibody, followed by immunoblot analysis using anti-PITSLRE antibody. p110C was co-immunoprecipitated with PAK1. B, the binding of PAK1 with p110C was enhanced in an anoikis-dependent manner. Anoikis was induced in the cells for the indicated times, and then cell lysates from each sample were immunoprecipitated with anti-PITSLRE antibody, followed by immunoblotting with specific anti-PAK1 antibody. The lower band shows the amount of PAK1 during anoikis determined by using Western blot analysis. The figure is representative of three independent experiments performed. C, PAK1 activity was inhibited in apoptotic NIH3T3 cells. Cell lysates from each sample were immunoprecipitated with a specific anti-PAK1 antibody to isolate PAK1, followed by protein kinase assays using MBP as substrate. D, relative kinase activity of PAK1 was determined by quantitation of the labeled MBP bands with ImageQuaNT software. The activity toward MBP in each experiment was normalized relative to that in control NIH3T3 cells. FIG. 7. Association of p110C with PAK1 and inhibition of PAK1 activity in apoptotic NIH3T3 cells. A, association of p110C with PAK1 in NIH3T3 cells in which anoikis was induced. Anoikis was apoptosis through activation of the JNK1 pathway (46 -50). In this study it was found that PAK1 activity was inhibited in the cells transfected with p110C or the cells induced during apoptosis. This suggested a potential anti-apoptotic role of PAK1 in NIH3T3 cells undergoing anoikis. The cells transfected with p110C were prone to apoptosis, which was possibly mediated by a decrease of PAK1 activity. However, the decrease of PAK1 activity by anoikis was smaller than that caused by an overexpression of p110C. This difference might be caused by the different expression levels of p110C between anoikis and transfection, or other mechanisms might also be involved in the apoptotic pathway mediated by p110C. Our results also raised the following questions: How does the p110C reduce PAK1 activity: through directly dephosphorylating PAK1, changing its conformation, or alternating the binding of PAK1 with another molecule? What will the downstream factors be under the p110C⅐PAK1 complex? These issues remain to be studied.
In summary, in this study it was demonstrated that p110C could associate with PAK1, which is a member of the PAK family. Meanwhile, PAK1 activity was inhibited, which might partly explain how p110C could induce apoptosis. Further analysis of this interaction and search for the substrate molecules of p110C might result in a much better understanding of the regulation and function of p110C⅐PAK1 complex, thereby providing new avenues for the mechanism of apoptosis mediated by p110C.