| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Anesthesiology, Perioperative and Pain Medicine, Center for Experimental Therapeutics and Reperfusion Injury, Brigham and Women’s Hospital, Boston, MA 02115;
Cell Signaling Technology, Beverly, MA 01915; and
Goldman School of Dental Medicine, Boston University, Boston, MA 02118
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
15 s) phosphorylation at Ser20, Ser192/197, and Thr402 in neutrophils stimulated with fMLP. Phosphorylation at Ser192/197 and Thr402 were highly transient events, whereas that at Ser20 was more persistent. In contrast, Pak 2 was constitutively phosphorylated at Ser141 in unstimulated neutrophils and phosphorylation at this site was less sensitive to cell stimulation than at other residues. Studies with selective inhibitors suggested that a variety of phosphatases might be involved in the rapid dephosphorylation of Pak 2 at Thr402 in stimulated neutrophils. This was consistent with biochemical studies which showed that the activation loop of GST-Pak 3, which is homologous to that in Pak 2, was a substrate for protein phosphatase 1, 2A, and a Mg2+/Mn2+-dependent phosphatase(s) which exhibited properties different from those of the conventional isoforms of protein phosphatase 2C. The data indicate that Pak 2 undergoes a complex pattern of phosphorylation in neutrophils and that dephosphorylation at certain sites may involve multiple protein phosphatases that exhibit distinct modes of regulation. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B in macrophages (10), and NFAT in T cells (11). Recent studies have also implicated Pak in Ras-mediated transformation (12), Fas-mediated apoptosis (e.g., Refs. 13 and 14), invasiveness of breast cancer cells (15), and the pathogenesis of HIV (e.g., Refs. 16, 17, 18). In addition to activated Rac/Cdc42, Pak 2 also contains binding sites for the 
subunits of complex G-proteins, certain adaptor proteins (Nck), and Pak-interacting guanine nucleotide exchange factors (PIXs) (for review, see Ref. 6). Thus, Paks are capable of receiving messages from a variety of signal transduction pathways. Considerable effort has been made to understand the regulation of Pak at the molecular level. The kinase domain (KD) of Pak is located in the C-terminal region of this enzyme, whereas the N-terminal region contains a Cdc42/Rac interactive binding motif and an autoinhibitory domain (ID; e.g., Ref. 19). Biochemical and biophysical studies have demonstrated that the ID interacts with KD to maintain Pak in an inactive/low affinity state (e.g., Refs. 20 and 21). Pak 1 and 2 can form homodimers in vivo (21). Dimerized Pak adopts a trans-inhibited conformation with the N-terminal domain of one Pak molecule inhibiting the catalytic domain of the other (21). Binding of one activated Rac or Cdc42 to this dimer disrupts these interactions and promotes autophosphorylation of the activation loop (AL) (Thr402) of both kinases (21). Phosphorylation at Thr402 results in a marked increase in the activity of Pak 2 toward certain substrates that persist even in the absence of the small GTPase (e.g., Ref. 22). In addition to the Thr402, peptide mapping studies have revealed that Pak 2 can also undergo autophosphorylation in vitro at several other residues including Ser20, Ser141, Ser192, and Ser197 (e.g., Ref. 23). Serines 20, 141, and 192/197 are located within the Nck binding site, the ID and PIX binding site, respectively. Phosphorylation within the ID (Ser141) partially activates Pak (24), whereas phosphorylation at Ser20 and Ser192 blocks binding to Nck and PIX, respectively (25). To date, few studies have focused on the phosphorylation of Pak 2 in vivo, particularly on residues outside of the AL (e.g., Ref. 26).
Pak 2 undergoes a rapid but transient activation in chemoattractant-stimulated neutrophils (1, 4) with activation/inactivation corresponding to phosphorylation and subsequent dephosphorylation at Thr402 (27). Distinct signals appear to be required for the activation and inactivation of Pak 2 in neutrophils (4). Thus, phosphatase(s) active against Pak are likely to play major roles in regulating this enzyme. Studies on cells other than neutrophils have shown that Pak can exist in complexes with protein phosphatase (PP) 2A (28) and/or with a highly specific phosphatase termed FEM-2 (29). The phosphatases active against Pak 2 in neutrophils have not been studied until now. In this paper, we have investigated the phosphorylation of Pak 2 at Ser20, Ser141, Ser192/197, and Thr402 in stimulated neutrophils along with the PPs which are active at some of these sites. Ancillary data on Pak 1 are also provided. The data indicate that Pak 2 undergoes multisite phosphorylation in neutrophils in a manner that is considerably different from that predicted from in vitro studies, and highlight the complex regulation of this kinase in vivo.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A phosphospecific Ab (pAb) that only recognized Pak 1, 2, or 3 when phosphorylated in the AL (pPak(AL) Ab) (Thr402 in Pak 2) was generated in rabbits (Research Genetics, Huntsville, AL) and purified/characterized in our laboratory (27). A similar affinity purified pAb to the AL of Pak 1, 2, and 3 along with affinity purified rabbit pAbs that react with Pak 2 when phosphorylated at Ser20 (pPak2(Ser20) Ab), Paks 1 and 2 when phosphorylated at Ser144 and Ser141, respectively (pPak1/2(Ser144/141) Ab) and Paks 1 and 2 when phosphorylated at Ser199 and/or Ser204, and Ser192 and/or Ser197, respectively (pPak1/2(Ser199 and Ser204/Ser192 and Ser197) Ab) were obtained from Cell Signaling Technology (Beverly, MA). The sequences of the phosphopeptide Ags used to generate these pAbs are provided in Fig. 1. The phosphopeptide Ags used to generate the pAbs described above and a rabbit polyclonal Ab that recognized both the phosphorylated and nonphosphorylated forms of Pak 2 (Pak 2 Ab) were also obtained from Cell Signaling Technology (Beverly, MA). Goat polyclonal Abs raised to peptides derived from the N terminus of human Pak 2 (-Pak(N-19) Ab) and human Pak 3 (-Pak(N-19) Ab) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Recombinant calmodulin produced in Escherichia coli, PP 2B/calcineurin isolated from bovine brain, recombinant human PP 2C
, PP 2A purified from human RBCs, PP 1 purified from rabbit skeletal muscle, recombinant PP 1 from rabbit muscle produced in E. coli and a sheep Ab to PP 2C generated to full-length human PP 2C were obtained from Upstate Biotechnology (Lake Placid, NY). Mouse mAbs generated to human PP 2A, human PP 1, and human calcineurin were obtained from BD Transduction Laboratories (San Diego, CA). Okadaic acid, calyculin A, ascomycin, FK-506, and cyclosporin A were purchased from Calbiochem (La Jolla, CA). Plasmid pGST-mPak 3 carrying mouse fibroblast mPak 3 fused to GST in pGEX-KG was a gift from Dr. A. Mak (Queen’s University, Kingston, Ontario, Canada). Immobilon-P transfer membranes (0.45 µm) were obtained from Millipore (Bedford, MA). Sources of all other materials are described elsewhere (1, 3, 4, 27).
|
Preparation of neutrophils and the neutrophil soluble fraction. Guinea pig neutrophils were prepared as described previously (30). These preparations contained >90% neutrophils with viabilities always >95%. The studies described herein were performed in compliance with all institutional policies and federal guidelines governing the use of animals in research.
Neutrophils (1.0 x 108) were suspended in 1.0 ml of ice-cold phosphatase buffer (50 mM HEPES, pH 7.5, 100 mM KCl, 0.20 mM EDTA) containing 0.025 ml of protease inhibitor mixture (Sigma-Aldrich, St. Louis, MO) and freeze-thawed three times with a mixture of dry ice and ethanol. The homogenate was centrifuged at 100,000 x g for 45 min and the resulting supernatant was used immediately as the source of Mg2+/Mn2+-dependent PP(s) that has certain properties that are different from those of conventional PP 2Cs.
Immunoblotting/detection of phosphorylated Pak 2 in neutrophils.
Neutrophils (1.9 x 106/ml) were stimulated in disposable 1-cm plastic cuvettes at 37°C. The standard assay mixture consisted of a modified Dulbecco’s PBS medium (135 mM NaCl, 2.7 mM KCl, 16.2 mM Na2PO4, 1.47 mM KH2PO4 0.90 mM CaCl2, and 0.50 mM MgCl2, pH 7.35) containing 7.5 mM D-glucose. Cells were generally incubated in this medium for 10.0–30.0 min with or without inhibitors before stimulation. Concentrations of inhibitors and stimulation times are presented in the figure legends. At the appropriate time, the cells were rapidly lysed by adding 0.25 ml of 5x concentrated solubilization buffer (SDS-B) to 1.0 ml of the reaction mixture and the samples were boiled for 4.0 min. The final composition of SDS-B after mixing was 2.3% (w/v) SDS, 62.5 mM Tris-HCl, pH 6.8, 5.0 mM EDTA, 10.0% (v/v) glycerol, 5.0% (v/v) 2-ME, and 0.002% (w/v) bromophenol blue. Aliquots of these samples were separated by SDS-PAGE (35 µg/lane) on 9.0% (v/v) polyacrylamide slab gels and transferred electrophoretically to Immobilon P-membranes (1). Specific phosphorylation of Pak was assayed by Western blotting with the appropriate pAb (e.g., Ref. 27). Membranes were incubated with the pAb (
1/1,000 dilution) overnight at 4°C. Phosphorylated Pak 2 was visualized with a luminol-ECL detection system (Pierce, Rockford, IL). Conditions for Western blotting are detailed elsewhere (4). At the end of these experiments, both the immunodetection system and the bound Ab were removed from the blot by incubating the membrane with ImmunoPure elution buffer (Pierce) for 30–60 min at room temperature followed by two washes in TBST (20 mM Tris-HCl, 150 mM NaCl, and 0.01% (v/v) Tween 20, pH 7.4). The blots were then stained with an Ab that recognized both the phosphorylated and nonphosphorylated forms of Pak 2 or 3.
The concentrations of inhibitors and conditions used in these studies did not affect cell viability, as measured by the exclusion of trypan blue or by the release of lactate dehydrogenase (data not shown; Ref. 31).
Preparation of phosphorylated GST-Pak 3 and in vitro dephosphorylation of this protein catalyzed by PP 1, PP 2A, PP 2B, and PP 2C.
The GST-Pak 3 (-Pak) fusion protein was expressed in E. coli, purified on a glutathione-sepharose affinity column and eluted with reduced glutathione (GSH) (e.g., Refs. 32 and 33).
Dephosphorylation of GST-Pak 3 by PP 1 or PP 2A was carried out at 30°C for 15 min in a 0.10-ml reaction mixture containing 30 µg/ml GST-Pak 3 in 20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 2.0 mM MnCl2, 1.0 mM EDTA, 0.50 mM DTT and 0.10 mg/ml BSA. The assay conditions for PP 2B (calcineurin) were identical to those described above but also included 0.20 mM CaCl2 and 70 µg/ml calmodulin (34). The assay conditions for PP 2C were identical to those for PP 1 and PP 2A but in this case MnCl2 was replaced by 5.0 mM MgCl2 (e.g., Ref. 35). The assay mixtures were allowed to equilibrate for 10 min at 30°C before the reactions were initiated by the addition of phosphatase. The units of phosphatase activity used are provided (see Fig. 7). Reactions were terminated by the addition of 20 µl of 5x concentrated SDS-B and the samples boiled for 4.0 min. Aliquots of these samples were subjected to SDS-PAGE on 7.0% (w/v) slab gels and dephosphorylation of Pak at a particular site was monitored by Western blotting with the appropriate pAb as described above.
|
|
Unless otherwise noted, all of the autoradiographic observations were confirmed in at least three separate experiments performed on different cell preparations. The loss of phosphate from Pak was estimated by densitometry by comparing the peak heights of the Pak bands with a Hoeffer GS 300 scanning densitometer. The numbers of observations (n) are based on different cell preparations.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have previously reported that Thr402 in the AL of Pak 2 always undergoes rapid and transient phosphorylation in fMLP-stimulated neutrophils (Ref. 27 ; Fig. 2A). In contrast, phosphorylation of Thr423 in the AL of Pak 1 was only observed in some of these experiments (e.g., Fig. 2B). Phosphorylation at these residues was monitored with a pAb that only recognized Pak 1/2 when phosphorylated at these sites (27). Maximal phosphorylation at Thr402 occurred within 15 s of cell stimulation followed by significant dephosphorylation at 2.0–3.0 min (Fig. 2A; Ref. 27). Identical kinetics were observed for the activation of Rac and Pak 2 in these cells (e.g., Refs. 1 and 31).
|
A variety of data indicate that the pAbs used in these experiments were specific for the designated residues. For example, these pAbs reacted with only one or two major proteins when blotted against a lysate of neutrophils with the major bands corresponding to the masses of Pak 2 and/or 1. When the blots shown in Fig. 2A were washed and restained with a regular Ab to Pak 2, the resulting bands overlapped the bands observed with the pAbs (data not shown). Most importantly, the bands observed with each pAb were specifically blocked by the phosphopeptide Ag (5.0 µg/ml) used to generate the pAb but not by phosphopeptides corresponding to the other phosphorylation sites in Pak 1 or 2 (Fig. 2B). Finally, the unique kinetics of phosphorylation observed with three of four of these pAbs along with the different sensitivities of the various sites to phosphatases and inhibitors monitored with these reagents (see below) strongly suggest that each pAb reacts with a distinct site in Pak.
It is noteworthy that the pPak(AL) Ab recognized only one form of Pak 2 in stimulated neutrophils, whereas the pPak 2(Ser20) Ab recognized both slow and fast migrating bands (Fig. 2A). One explanation for these data is that the slow migrating band results from phosphorylation at Thr402. Thus, the pPak(AL) Ab would only recognize this form of Pak 2. Under these circumstances, the mobility shift observed with the pPak(Ser20) Ab would be due to phosphorylation at Ser20 occurring both during and after phosphorylation at Thr402 such that this pAb would bind to both the slow and fast migrating forms of Pak 2. In contrast, the rapid dephosphorylation of Pak 2 at Ser192/197 would preclude detection of the lower band with the pPak1/2(Ser199/204 and Ser192/197) Ab. Consistent with this interpretation, a proteolytic fragment of Pak 1 (residues 201–491) which contains only the AL phosphorylation site (Thr423) exhibits a mobility shift when phosphorylated in vitro (20).
Interestingly, evidence of a possible mobility shift and/or dephosphorylation of Pak 2 with the pPak 1/2(Ser144/141) Ab was only observed in stimulated cells at 15 s (Fig. 2A). These data may indicate that any Pak 2 which undergoes phosphorylation at both Ser141 and Thr402 is particularly sensitive to dephosphorylation, and/or that Pak 2 phosphorylated at Ser141 does not undergo significant phosphorylation at Thr402.
Effects of inhibitors of PPs on the phosphorylation of Pak 2 at Thr402 in stimulated neutrophils
The effects of inhibitors of PP 2A (100 nM okadaic acid) (e.g., Refs. 35 and 37) and/or PP 2B (200 nM ascomycin) (e.g., Ref. 38) on the phosphorylation of Pak 2 at Thr402 were evaluated to obtain information on the PPs that were active at or controlled phosphorylation at this residue (Fig. 3). Neutrophils were treated with these inhibitors for 30 min at 37°C and then stimulated with 1.0 µM fMLP for 15 s. Both okadaic acid and ascomycin increased phosphorylation at Thr402 in stimulated cells (Fig. 3). The combination of these inhibitors also increased the basal phosphorylation at Thr402 in unstimulated neutrophils (Fig. 3B). Similar results were observed with two other, structurally distinct inhibitors of PP 2B (1.0 µM cyclosporin A, 100 nM FK-506) (e.g., Ref. 38) (data not shown). Okadaic acid directly inhibits PP 2A, whereas cyclosporin A, FK-506, and ascomycin induce conformational changes in certain proteins (immunophilins) which then enable these proteins to inhibit PP 2B (38). In addition to PP 2A, okadaic acid also inhibits PP 1 but at substantially higher concentrations than that needed to block PP 2A (e.g., Refs. 35 and 37). Okadaic acid inhibits dilute solutions of PP 1 and PP 2A in vitro at concentrations of 1.0–2.0 µM and 1.0–2.0 nM, respectively (35, 37). However, the effective concentrations of this inhibitor towards PP 1 and PP 2A in vivo are considerably higher due to the high concentrations of these phosphatases in cells (e.g., Refs. 35 and 37). Thus, the ability of 100 nM okadaic acid to enhance phosphorylation at Thr402 in stimulated neutrophils is consistent with the involvement of PP 2A in this reaction. Inhibitors of PP 1 (e.g., okadaic acid >1.0 µM, calyculin A) (37) could not be used in these studies because these compounds are known to reduce phosphorylation at Thr402 and block activation of Pak 2 in neutrophils (1, 3, 27). Selective inhibitors of PP 2C are not yet available.
|
-Pak(N-19) Ab) demonstrated that this loss of phosphate was not due to a proteolytic event (Fig. 4). These data indicate that neutrophils are likely to contain phosphatase(s) in addition to PP 2A and/or PP 2B that are active against Thr402.
|
Dephosphorylation of GST-Pak 3 by PPs
Many serine/threonine PPs can be broadly divided into different categories (e.g., PP 1, PP 2A, PP 2B, PP 2C) on the basis of their substrate specificity, requirement for divalent cations, and sensitivity to various tumor promoters (e.g., okadaic acid, calyculinA) (e.g., Ref. 35). The presence of PP 1, PP 2A, PP 2B, and PP 2C in neutrophils was demonstrated with a panel of Abs that specifically reacted with the catalytic subunits of these enzymes (Fig. 5). Each of these Abs reacted with a major protein when blotted against a lysate of unfractionated neutrophils with the resulting bands corresponding to the expected molecular masses of these phosphatases.
|
-D-thiogalactopyranoside (IPTG) and purified with GSH-sepharose beads as referenced in Materials and Methods (Fig. 6). As reported previously, GST-Pak 3 is susceptible to degradation (33). The resulting fusion protein reacted with the pPak(AL) Ab and the pPak 1/2(Ser144/141) Ab (Figs. 6 and 7) which indicted that a portion/population of this protein was phosphorylated at Thr421 and Ser139. Perturbation of the N-terminal region of Pak by fusion to GST results in activation of a portion of this kinase in the absence of activated Rac or Cdc42 (32). GST-Pak 3 was used in these studies because this isoform is easily expressed in E. coli (Fig. 6) and the sequences surrounding the critical phosphorylation site in the AL of Pak 2 and Pak 3 are completely homologous (i.e., Thr402 in Pak 2 is analogous to Thr421 in Pak 3; Fig. 1). Similarly, there is very strong homology between the sequence within the ID surrounding Ser141 in Pak 2 and Ser139 in Pak 3 (Fig. 1). Thus, this fusion protein may serve as a substrate to gain information on potential phosphatases that recognize phosphorylated residues in the AL and ID of Pak.
|
, PP 2A, PP 2B, and PP 2C for 15 min at 37°C under conditions known to be effective for each of these phosphatases (34) and phosphorylation at Thr421 and Ser139 was monitored with the pAbs described above (Fig. 7). PP 1 and PP 2A, but not PP 2B or PP 2C, were effective against Thr421 under the conditions used. In contrast, all four phosphatases were capable of catalyzing dephosphorylation at Ser139 which established that these phosphatases were active under the assay conditions used (Fig. 7). The PP 2B catalyzed dephosphorylation at Ser139 exhibited a strong activation by calmodulin (Fig. 7C). The insensitivity of the AL phosphorylation site to PP 2B suggests that the ability of ascomycin to enhance phosphorylation at Thr402 in stimulated neutrophils (Figs. 3 and 4) may occur indirectly provided that the in vitro experiments summarized in Fig. 7 adequately reflect the situation for native Pak 2 in vivo (see Discussion). Effects of neutrophil lysates on the phosphorylation of GST-Pak 3; dephosphorylation of Pak by novel phosphatase(s)
Studies were undertaken to determine if neutrophils contain novel phosphatase(s) active against the AL of Pak (Fig. 8). Incubation of GST-Pak 3 with the undialyzed, soluble fraction of neutrophils resulted in an increased detection of phosphate at Thr421 and Ser139 but less so at Ser200/205 when monitored with the pAbs described above (Fig. 8, lanes a and b). However, addition of 10.0 mM Mg2+ to the reaction mixture 30 min after the addition of the soluble fraction resulted in a pronounced, time-dependent loss of phosphate from Thr421 but not from the Ser139 or Ser200/205 sites. A similar loss of phosphate from Thr421 was also observed when 10.0 mM MnCl2 was added to the assay mixture. Because the Thr421 site was highly phosphorylated at the time the divalent cation was added, these data are consistent with Mg2+/Mn2+ stimulating the activity of highly specific phosphatase(s) that recognize the AL of Pak. It is unlikely that this activity is due to PP 1, PP 2A, or a typical isoform of PP 2C because inhibitors of these phosphatases were present in the assay system at optimal concentrations (100 nM calyculin A, 1.0 µM orthovanadate) (e.g., Refs. 36 and 37). Moreover, a typical isoform of PP 2C from humans was not active against the AL in GST-Pak 3 but did recognize the Ser139 site (Fig. 7). Finally, sodium fluoride (100 mM) partially inhibited the Mg2+/Mn2+-dependent phosphatase(s) from neutrophils that were active against the AL of GST-Pak 3 (Fig. 8, compare lanes f and h). Sodium fluoride (100 mM) does not inhibit the typical isoforms of PP 2C (39). The effects of Mg2+ and F- were estimated by densitometry by comparing the heights of bands b, f, and h in Fig. 8. Treatment of the fusion protein for 30 min with 10.0 mM MgCl2 alone or 10.0 mM MgCl2 plus 100 mM NaF reduced the signal with the pPak(AL) Ab by 79 ± 7% and 36 ± 15% (SD, n = 3–4), respectively. Thus, a variety of data suggest that this specific Mg2+/Mn2+-AL phosphatase present in neutrophils is not a typical form of PP 2C. This Mg2+/Mn2+-dependent activity may be responsible for the dephosphorylation of the AL of Pak 2 that occurred in stimulated neutrophils in the presence of okadaic acid and ascomycin (Fig. 4).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Incubation of baculovirus expressed Pak 2 or Pak 2 purified from rabbit reticulocytes with Mg · ATP alone results in autophosphorylation at Ser19, Ser20, Ser55, Ser192, and Ser197. Addition of activated Cdc42 to the reaction mixture results in additional phosphorylation of Pak 2 at Ser141, Ser165, and Thr402 (23). In contrast, Pak 2 was not phosphorylated on Ser20 and Ser192/197 in unstimulated neutrophils but was phosphorylated on Ser141. Stimulation of neutrophils with fMLP resulted in rapid phosphorylation of Pak 2 at Ser20, Ser192/197, and Thr402 (Fig. 2A). Elicited guinea pig neutrophils and human blood neutrophils stimulated with fMLP exhibit rapid activation of Cdc42 and Rac with kinetics consistent with these GTPases functioning upstream of Pak (e.g., Refs. 31 and 40). This rapid accumulation of activated Cdc42/Rac in stimulated neutrophils would promote autophosphorylation of Pak 2 on Thr402 as predicted by the in vitro studies. Why phosphorylation of Pak 2 on Ser20, Ser192/197, and Ser141 in neutrophils does not conform to the in vitro pattern is not known, but may reflect the activities of phosphatases and kinases that are present in neutrophils and active at these sites. The peritoneal cells used here are likely to be primed to exhibit enhanced responses upon stimulation (e.g., Ref. 41). Whether this priming also effects the phosphorylation of Pak is not known. Thus, the phosphorylation/regulation of Pak 2 in stimulated neutrophils is considerably different from that predicted by in vitro studies.
We have previously reported that the rapid activation/inactivation of Pak 2 in fMLP-stimulated neutrophils is closely associated with phosphorylation/dephosphorylation of this kinase at Thr402 (27). A variety of data implicate PP 2A in the dephosphorylation/inactivation of Pak. PP 2A catalyzes dephosphorylation of the AL of Pak 2 immunoprecipitated from neutrophils (27) and an inhibitor of PP 2A (100 nM okadaic acid) increased phosphorylation at this site in stimulated cells (Fig. 3). PP 2A is present in neutrophils (Fig. 5) and can exist in complexes with Pak in certain cell types (28). However, it is likely that phosphatases in addition to PP 2A are also active against the AL of Pak 2 because 100 nM okadaic acid did not prevent the pronounced dephosphorylation at Thr402 that occurred in neutrophils at 1.0 and 3.0 min after stimulation (Fig. 4). Other potential phosphatases that can catalyze this reaction are PP 1 and FEM-2 (Fig. 7; Ref. 29). We have detected PP 1 in neutrophils (Fig. 5) along with a specific Mg2+/Mn2+-dependent AL phosphatase. This Mg2+/Mn2+-dependent phosphatase differs from conventional PP 2Cs in that it is active in the presence of 1.0 µM orthorvanadate and is partially inhibited by 100 mM NaF (Fig. 8; Refs. 36 and 39). Moreover, a typical isoform of PP 2C was not found to be active against the AL of GST-Pak 3 (Fig. 7). FEM-2 is a possible candidate for this Mg2+/Mn2+-dependent phosphatase because it is activated by Mg2+, catalyzes dephosphorylation of the AL of Pak, is partially inhibited by F- and insensitive to orthorvanadate and calyculin A (29, 36, 39). Chronic activation of Pak can promote viral replication, metastasis, and cellular transformation (e.g., Refs. 6, 12, 18). Perhaps the presence of multiple phosphatases active at the Thr421 site are a safeguard against these possibilities. Interestingly, the Mg2+/Mn2+-dependent AL phosphatase in neutrophils was not active against phosphorylation sites in the ID and PIX-binding region of Pak 3 (Fig. 8) which suggests that different sites in Pak can be recognized by distinct phosphatases that are susceptible to different modes of regulation.
PIX can function as a scaffold protein because it binds Pak and FEM-2 simultaneously (29). Thus, PIX may participate in the activation and subsequent inactivation of Pak 2 through its GTP/GDP exchange activity and ability to serve as a targeting subunit that colocalizes Pak and FEM-2 (29). As noted earlier, Pak 2 in fMLP-stimulated neutrophils undergoes a rapid and transient phosphorylation on serines 192/197 (Fig. 2A), and these modifications may promote dissociation of the PIX-FEM-2 complex from Pak (25). The physiological significance of these reactions is not yet known because isoforms of PIX have been described that have activating, inhibiting, and permissive effects on Pak (42). The isoforms of PIX in neutrophils remain to be characterized. PIX itself can undergo phosphorylation which does not affect its GTP/GDP exchange activity (43). It is interesting to note that while inhibitors of PP 2B enhanced phosphorylation of Pak 2 at Thr402 in stimulated neutrophils, isolated PP 2B did not catalyze dephosphorylation of the AL in the GST-Pak 3 fusion protein (Fig. 7). If these data with the GST-Pak 3 fusion protein adequately reflect the situation in vivo, they indicate that PP 2B regulates Pak indirectly, perhaps by catalyzing the dephosphorylation of PIX or some unknown protein that is also involved in the regulation of Pak.
As noted above, Pak 1 or 2 can form dimers with the ID of one Pak molecule inhibiting the catalytic domain of the other (21). Monomeric Pak possesses considerable intrinsic kinase activity even in the absence of activated Rac or Cdc42 (cf, Ref. 21). Dimerization severely restricts this intrinsic/basal activity of Pak (21). Serine 141 resides within the ID and phosphorylation at this site partially activates Pak 2 by preventing the ID from packing against the kinase domain (21, 24). The constitutive phosphorylation of Pak 2 at Ser141 (and Pak 1 at Ser144) may provide for a pool of monomeric, partially active Pak that may be necessary for certain cellular functions in unstimulated neutrophils. Constitutive phosphorylation at Ser141 may also provide a population of Pak that is primed for rapid activation during cell stimulation. This may be particularly important for chemoattractants where different biological responses are triggered by increasing concentrations of agonists (e.g., Ref. 44). Finally, it is noteworthy that phosphorylation at Ser141 persisted in unstimulated neutrophils despite the fact that PP 1, PP 2A, PP 2B, and PP 2C were readily detected in these cells (Fig. 5) and active at this site (Fig. 7). Perhaps this site associates with a protein that shields it from these and other phosphatases.
In summary, we report that Pak 2 undergoes phosphorylation in neutrophils on multiple sites in a manner that exhibits significant differences from that observed during activation of this kinase in vitro. Dephosphorylation at some sites may involve multiple protein phosphatases that are susceptible to different modes of regulation. These data highlight the complex regulation that is necessary for coordinating the many cellular functions of Pak 2.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. John A. Badwey, Department of Anesthesiology, Perioperative and Pain Medicine, Center for Experimental Therapeutics and Reperfusion Injury, Brigham and Women’s Hospital, Thorn Building, Room 703, 75 Francis Street, Boston, MA 02115. E-mail address: Badwey{at}zeus.bwh.harvard.edu
3 Abbreviations used in this paper: Pak, p21-activated kinase; PIX, Pak-interacting guanine nucleotide exchange factor; KD, kinase domain; ID, autoinhibitory domain; AL, activation loop; PP, protein phosphatase; pAb, phosphospecific Ab; IPTG, isopropyl -D-thiogalactopyranoside; GSH, reduced glutathione.
Received for publication December 11, 2002. Accepted for publication July 25, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B activity by multiple signaling pathways requires Pak 1. J. Biol. Chem. 275:19693.Pak: inhibition of Pak kinases reveals their morphological roles downstream of Cdc42 and Rac1. Mol. Cell. Biol. 18:2153.-Pak as a function of activation. J. Biol. Chem. 274:8022.-p21-activated kinase (-Pak) and related kinases (MSTs) in normal and stressed neutrophils. J. Immunol. 166:6349.1PIX, the Pak-interacting exchange factor, requires localization via a coiled-coil region to promote microvillus-like structures and membrane ruffles. J. Cell Sci. 114:4239.This article has been cited by other articles:
|
|
 |
|
  E. A. Kramar, L. Y. Chen, N. J. Brandon, C. S. Rex, F. Liu, C. M. Gall, and G. Lynch Cytoskeletal Changes Underlie Estrogen's Acute Effects on Synaptic Transmission and Plasticity J. Neurosci., October 14, 2009; 29(41): 12982 - 12993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Bakema, A. Bakker, S. de Haij, H. Honing, M. Bracke, L. Koenderman, G. Vidarsson, J. G. J. van de Winkel, and J. H. W. Leusen Inside-Out Regulation of Fc{alpha}RI (CD89) Depends on PP2A J. Immunol., September 15, 2008; 181(6): 4080 - 4088. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Chan, L. Lim, and E. Manser PAK Is Regulated by PI3K, PIX, CDC42, and PP2C{alpha} and Mediates Focal Adhesion Turnover in the Hyperosmotic Stress-induced p38 Pathway J. Biol. Chem., September 5, 2008; 283(36): 24949 - 24961. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Sheehan, Y. Ke, and R. J. Solaro p21-Activated kinase-1 and its role in integrated regulation of cardiac contractility Am J Physiol Regulatory Integrative Comp Physiol, September 1, 2007; 293(3): R963 - R973. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Reutershan, R. Stockton, A. Zarbock, G. W. Sullivan, D. Chang, D. Scott, M. A. Schwartz, and K. Ley Blocking p21-activated Kinase Reduces Lipopolysaccharide-induced Acute Lung Injury by Preventing Polymorphonuclear Leukocyte Infiltration Am. J. Respir. Crit. Care Med., May 15, 2007; 175(10): 1027 - 1035. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T.-J. Lu, W.-Y. Lai, C.-Y. F. Huang, W.-J. Hsieh, J.-S. Yu, Y.-J. Hsieh, W.-T. Chang, T.-H. Leu, W.-C. Chang, W.-J. Chuang, et al. Inhibition of Cell Migration by Autophosphorylated Mammalian Sterile 20-Like Kinase 3 (MST3) Involves Paxillin and Protein-tyrosine Phosphatase-PEST J. Biol. Chem., December 15, 2006; 281(50): 38405 - 38417. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Hoffert, T. Pisitkun, G. Wang, R.-F. Shen, and M. A. Knepper Quantitative phosphoproteomics of vasopressin-sensitive renal cells: Regulation of aquaporin-2 phosphorylation at two sites PNAS, May 2, 2006; 103(18): 7159 - 7164. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and presence of 100 nM okadaic acid plus 200 nM ascomycin (
) (means ± SD). Phosphorylation at Thr402 was estimated by densitometry with the 100% value being that exhibited by cells not treated with inhibitors but stimulated with 1.0 µM fMLP for 15 s. Conditions for cell stimulation and Western blotting are provided in Materials and Methods.
