HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 24, 18366-18374, June 16, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/24/18366    most recent M001339200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Goeckeler, Z. M. Articles by Wysolmerski, R. B. Search for Related Content PubMed PubMed Citation Articles by Goeckeler, Z. M. Articles by Wysolmerski, R. B. Social Bookmarking                      What's this?

Phosphorylation of Myosin Light Chain Kinase by p21-activated Kinase PAK2^*

Zoe M. Goeckeler, Ruthann A. Masaracchia^§, Qi Zeng^¶, Teng-Leong Chew^¶, Patricia Gallagher, and Robert B. Wysolmerski^¶**

From the Departments of ^¶ Pathology and  Anesthesiology, St. Louis University School of Medicine, St. Louis, Missouri 63104-1028, the ^§ Division of Biochemistry and Molecular Biology, Department of Biological Science, University of North Texas, Denton, Texas 76203-5018, and the  Department of Physiology and Biophysics, Indiana University School of Medicine, Indianapolis, Indiana 46202-5120

Received for publication, February 18, 2000, and in revised form, March 21, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Phosphorylation of myosin II regulatory light chains (RLC) by Ca²⁺/calmodulin-dependent myosin light chain kinase (MLCK) is a critical step in the initiation of smooth muscle and non-muscle cell contraction. Post-translational modifications to MLCK down-regulate enzyme activity, suppressing RLC phosphorylation, myosin II activation, and tension development. Here we report that PAK2, a member of the Rho family of GTPase-dependent kinases, regulates isometric tension development and myosin II RLC phosphorylation in saponin permeabilized endothelial monolayers. PAK2 blunts tension development by 75% while inhibiting diphosphorylation of myosin II RLC. Cdc42-activated placenta and recombinant, constitutively active PAK2 phosphorylate MLCK in vitro with a stoichiometry of 1.71 ± 0.21 mol of PO₄/mol of MLCK. This phosphorylation inhibits MLCK phosphorylation of myosin II RLC. PAK2 catalyzes MLCK phosphorylation on serine residues 439 and 991. Binding calmodulin to MLCK blocks phosphorylation of Ser-991 by PAK2. These results demonstrate that PAK2 can directly phosphorylate MLCK, inhibiting its activity and limiting the development of isometric tension.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The PAK¹ family of serine/threonine protein kinases have been implicated in a broad spectrum of signal transduction pathways leading to diverse physiological end points, including cytoskeletal reorganization, apoptosis, and Ras-mediated cell transformation (1, 2). All PAK isoforms are direct effectors of the Rho family GTP-binding proteins Rac and Cdc42, suggesting that cell-specific responses arise from either selective regulation of G protein activation in response to unique agonists or selective phosphorylation of tissue-specific protein kinase substrates.

In the initial studies of the relative roles of G protein activation and protein kinase activity in actin reorganization, Sells et al. (3) demonstrated that microinjection of activated PAK1 induced rapid formation of polarized filopodia in Swiss 3T3 cells. However, the relative importance of PAK-dependent phosphorylation of unique substrates during Cdc42/Rac-dependent cytoskeletal reorganization is incompletely understood. Transient transfection of HeLa cells with constitutively active PAK1 promoted cytoskeletal rearrangement, which was entirely analogous to that observed in Cdc42 and Rac transfected cells (4). In subsequent studies, a peptide that inhibited kinase activity in PAK blocked Cdc42, Rac, and constitutively active PAK-induced morphological changes in transfected HeLa cells (5). An important role for PAK-mediated phosphorylation in actin reorganization has also been supported by studies in permeabilized (6) and intact (7) endothelial cells as well as skinned smooth muscle cells (8). Permeabilized endothelial cells incubated with Cdc42-activated PAK2 or the catalytic domain of PAK2 underwent ATP-dependent retraction and actin reorganization (6). In addition, these studies using purified enzymes established that the regulatory nonmuscle and smooth muscle myosin II light chain is a substrate for PAK2 (9, 10). Recent studies have shown that the PAK family of kinases is involved in activation of myosin II in vivo. Zeng et al. (7) demonstrated that microinjection of constitutively active PAK2 into endothelial cells induced cell retraction and monophosphorylation of myosin II RLC. An increase in myosin light chain phosphorylation has also been shown by immunocytochemical staining in fibroblasts (11) and microvascular endothelial cells (12) transiently transfected with PAK1.

In contrast to the Ca²⁺/CaM-dependent MLCK, which catalyzes phosphorylation of RLC at both Ser-19 and Thr-18 (13, 14), PAK2 catalyzes only monophosphorylation of RLC (7) at Ser-19 (6). However, both PAK and MLCK phosphorylation of the RLC in intact nonmuscle myosin II activates the myosin ATPase (10). These results in nonmuscle cells are supported by observations of Van Eyk et al. (8) in smooth muscle cells, suggesting that PAK and MLCK phosphorylation of RLC both contribute to smooth muscle contraction. Calcium-independent contraction of triton-skinned smooth muscle cells occurred in response to incubation with both PAK and Rho kinase. These investigators confirmed that PAK catalyzed monophosphorylation of RLC at Ser-19, but in the skinned muscle preparation, contraction was better correlated with caldesmon and desmin phosphorylation and not RLC phosphorylation. Furthermore, recent studies (15) have shown a role for PAK1 in modulating Ca²⁺ sensitivity of smooth muscle contraction by phosphorylation of caldesmon. Sanders et al. (16) have reported that overexpression of Rac1 or constitutively active PAK1 results in inhibition of BHK-21 and HeLa cell spreading. These investigators present evidence that this effect is the result of a decrease in myosin RLC phosphorylation mediated by PAK1-catalyzed phosphorylation and down-regulation of MLCK. Collectively, these data strongly support a role for PAK-mediated protein phosphorylation in some cytoskeletal reorganization responses to Cdc42 and Rac.

Studies presented in this report characterize PAK2-catalyzed phosphorylation of MLCK in vitro. To further investigate the potential role for PAK-mediated phosphorylation in regulating cell retraction, we have also determined the effect of activated PAK2 on RLC phosphorylation and isometric tension in endothelial cells.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

MLCK Phosphorylation and Measurement of RLC Phosphorylation-- MLCK was phosphorylated by PAK2 (10:1 molar ratio MLCK:PAK2) in 50 µl of phosphorylation buffer (25 mM Tris-HCl, pH 7.5, 5 mM MgCl₂, 150 mM KCl, 1 mM DTT) containing 125 µM [-³²P]ATP at 30 °C for 10 min unless otherwise indicated. The reactions were stopped by the addition of an equal volume of ice-cold 20% trichloroacetic acid. Trichloroacetic acid-precipitated proteins were washed in ice-cold acetone and resuspended in Laemmli SDS sample buffer for analysis by 7.5% or 10% SDS-PAGE and autoradiography. In the case of mutant MLCK bound to beads, the beads were washed twice in PBS and resuspended in 100 µl of phosphorylation buffer containing 125 µM [-³²P]ATP. Phosphorylation reactions were initiated by addition of 100 nM PAK2 and incubated at 30 °C for 10 min. In experiments where MLCK was prephosphorylated by PAK2, PAK2 was removed from the reaction mixture by immunoprecipitation with a rabbit anti-PAK2 antibody preabsorbed to protein A beads.

To determine the stoichiometry of phosphate incorporation into MLCK by PAK2, a 10:1 (mol/mol) ratio of purified recombinant MLCK:PAK2 was incubated in the presence of [-³²P]ATP (specific activity = 200-300 cpm) for 30 min at 30 °C. Aliquots of the reaction mixtures were spotted onto Whatman P-81 filter papers and analyzed as described previously (17).

For analysis of RLC phosphorylation, endothelial cell populated collagen gels were snap-frozen, pulverized, and extracted with immunoprecipitation buffer (14). Undissolved collagen was spun down at 10,000 × g and myosin II immunoprecipitated as described previously (14, 18). For in vitro phosphorylation assays, endothelial cell myosin II was immunopurified as described previously (14). Myosin II was dissolved in urea sample buffer (19) and samples analyzed by glycerol-urea gel electrophoresis. Western blots were incubated using a rabbit anti-myosin II regulatory light chain antibody (6). RLC bands were visualized by ECL detection reagents (Amersham Pharmacia Biotech).

Identification of MLCK Phosphorylation Sites-- To determine the location of PAK2 phosphorylation sites, MLCK was phosphorylated by PAK2 at 30 °C for 30 min as outlined above. PAK2 was removed from the reaction mixture by applying the phosphorylated MLCK reaction mixture to a rabbit anti-PAK2 affinity column. Phosphorylated MLCK was precipitated by addition of an equal volume of 20% trichloroacetic acid, washed with 100% acetone, and air-dried. MLCK was denatured, reduced, and carboxymethylated prior to trypsin digestion. Urea (8 M) dissolved in 0.4 M NH₄HCO₃ (50 µl) was added to 500 µg of air-dried MLCK and the pH adjusted to between 7.5 and 8.5. 5 mM DTT (5 µl) was added and the reaction mixture incubated at 50 °C for 15 min. After cooling, the MLCK was carboxymethylated by addition of 5 µl of 100 mM iodoacetamide and incubated at room temperature for 15 min. H₂O was added to a final volume of 200 µl. TPCK-trypsin was added at a ratio of 1/25 enzyme to protein (w/w) and incubated at 37 °C for 24 h. The reaction was stopped by snap-freezing in liquid N₂, and the sample was lyophilized. The trypsin digest was adjusted to 0.1% trifluoroacetic acid and applied to a Bio-Rad RP Hi-Pore 318 column (250 × 10 mm) equilibrated with 0.1% trifluoroacetic acid. The column was washed for 10 min and developed with a gradient (120 ml at 1 ml/min) of 0-40% acetonitrile in 0.1% trifluoroacetic acid. Approximately 50% of the ³²PO₄ was eluted at 8.8% acetonitrile (fraction 1) and 50% of the ³²PO₄ was eluted at 14.4% acetonitrile (fraction 2). Both fractions were rechromatographed on a Gilson SynChropak RP-P column (250 × 4.6 mm) equilibrated with 0.1% trifluoroacetic acid. Fraction I was eluted with a 0-24% acetonitrile gradient (60 ml at 1 ml/min). ³²PO₄ was eluted at 7.2% acetonitrile. Fraction II was eluted with a 12-32% acetonitrile gradient (60 ml at 1 ml/min). ³²PO₄ was eluted at 16% acetonitrile. Fractions containing ³²PO₄ were pooled and dried to 100 µl in a vacuum centrifuge. Amino acid sequence was determined by Baylor College of Medicine Protein Chemistry Core Facility. Samples were covalently attached to Sequelon membranes; approximately 50% of the sample was used for ³²PO₄ determination at each round.

cDNA Constructs-- Site-specific mutants within the smooth muscle MLCK (20) were generated as described previously (21, 22). Mutant MLCK proteins with point mutations at residues Ser-439 (S439A), Ser-991 (S991A), Ser-992 (site A, S992A), Ser-1005 (site B, S1005A), or both Ser-991 and Ser-992 (S991A/S992A) and Ser-992 and Ser-1005 (S992A/S1005A) replaced with alanine, were subcloned into pCMV5 expression vector and expressed in rat embryo fibroblasts (REF-52) (21). Briefly, REF-52 cells were grown and maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 2 mM glutamine, 10% fetal calf serum, 50 units/ml penicillin, and 50 µg/ml streptomycin. REF-52 were plated onto 60-mm dishes at a density of 4 × 10⁶ and incubated at 37 °C for 24 h. Cells were transfected with 5 µg of DNA containing vector alone (pCMV5) or MLCK constructs (pCMV5-MLCK constructs) using Superfect reagent for 3 h following the manufacturer's protocol (Qiagen). REF-52 cells were harvested after 48 h and MLCK immunopurified as outlined below.

Protein Purification-- PAK2 was purified from human placenta (17) and activated by Cdc42-GTP or limited trypsin proteolysis (6, 17). Recombinant constitutively active PAK2 was expressed in Escherichia coli and purified over Talon Metal Affinity resin (CLONTECH, Palo Alto, CA). Recombinant MLCK was expressed in Sf-9 cells and purified as described (6, 23). MLCK site-specific mutants were immunopurified from REF-52 cell extracts as described below.

Immunopurification of Mutant MLCK Proteins-- REF-52 cells expressing mutant MLCK proteins were washed twice with Mg²⁺/Ca²⁺ PBS, flooded with 800 µl of MLCK extraction buffer (25 mM Tris-HCl, pH 6.8, 150 mM NaCl, 50 mM MgCl₂, 1.0 mM EGTA, 1.0 mM EDTA, 1% Nonidet P-40, 0.5% Triton X-100, 0.2 mM phenylmethylsulfonyl fluoride, 100 µg/ml benzamidine, 100 µg/ml soybean trypsin inhibitor, and 10 µg/ml each of TPCK, 1-chloro-3-tosylamido-7-amino-2-heptanone, aprotinin, leupeptin, and pepstatin at 4 °C), and immediately placed on ice. Cells were scraped up with a rubber policeman, culture dishes washed with an additional 200 µl of extraction buffer, and samples combined and extracted on ice for 20 min. The total cell extract was centrifuged at 132,000 × g for 10 min in a Beckman TL-100 ultracentrifuge. Pellets were extracted again by resuspending in 200 µl of extraction buffer, sonicating in a bath sonicator for two 5-s bursts, and incubating on ice for 20 min. The insoluble material was sedimented at 132,000 × g and supernatants combined. The supernatants were added to protein A beads pre-complexed with 15 µl of a polyclonal anti-rabbit smooth muscle MLCK IgG fraction (15 mg/ml) for 3 h at 4 °C. Immune complexes bound to protein A-Sepharose 4B were collected by centrifugation for 5 min at 12,000 × g at 4 °C. Pellets were first washed in 1 ml of extraction buffer and then once with a 1:1 dilution of extraction buffer/PBS and twice with PBS. Protein A beads containing the immunopurified mutant MLCK proteins were resuspended in PBS and equivalent aliquots were used for in vitro phosphorylation reactions as described above. ³²PO₄ incorporation into WT and mutant MLCK proteins was assessed by SDS-PAGE and autoradiography. The autoradiographs and corresponding Coomassie Blue-stained proteins were each quantitated by two-dimensional laser densitometry and MLCK phosphorylation expressed as a ratio of ³²PO₄ densitometric units to protein densitometric units. All values were normalized to WT MLCK.

One-dimensional Tryptic Peptide Mapping and Phosphoamino Acid Analysis-- ³²P-Phosphorylated MLCK was electrophoresed on 7.5% SDS-polyacrylamide gels, which were fixed in 50% methanol, 10% acetic acid, dried at 70 °C, and exposed to x-ray film. Labeled bands corresponding to the MLCK were cut from the SDS-polyacrylamide gel, rehydrated, and washed with five 1-ml changes of 10% methanol followed by two 1-ml changes of 50 mM ammonium bicarbonate, pH 8.8. The washed gel slices were cut into small pieces, placed in 800 µl of 50 mM ammonium bicarbonate, pH 8.8, and digested with 60 µg of TPCK-trypsin (1 mg/ml in 1 mM HCl) for 24 h at 37 °C. After the first 18 h, an additional 60 µg of trypsin was added. The solution was removed, gel pieces were washed with 500 µl of 50 mM ammonium bicarbonate, samples combined, lyophilized in a Savant SpeedVac, redissolved in 1 ml of distilled water, relyophilized, and then resuspended in 20 µl of 100 mM ammonium bicarbonate, pH 8.0. Tryptic peptides of MLCK were separated following the methods described by Daniel and Sellers (24). Apparent pI values of the MLCK phosphopeptides were determined as outlined previously (14). The pI values were not corrected for the presence of urea.

Phosphoamino acid analysis of phosphorylated MLCK was performed as described by Haeberle et al. (25) with slight modifications. ³²P-Labeled tryptic peptides of MLCK were acid-hydrolyzed in 6 N HCl at 100 °C under N₂. HCl was removed under reduced pressure, and the hydrolysates were electrophoresed on cellulose thin layer plates for 3 h at 1 kV at 4 °C in glacial acetic acid/formic acid (88% by volume)/H₂O, 80:20:900 (v/v). Phosphoamino acid standards (2 µg) were run side by side for comparison.

Isometric Tension Measurements of Saponin-permeabilized Endothelial Monolayers-- Isometric tension measurements were performed as described in detail previously (14, 26). Monolayers were permeablized in stabilization buffer (127 mM NaCl, 50 mM KCl, 1.1 mM NaH₂PO₄, 1 mM EGTA, 1 mM MgSO₄, 20 mM PIPES, 10 mM imidizole, 0.2 mM DTT, 5 µg/ml each aprotinin, leupeptin, pepstatin, and chymostatin, 10 µg/ml soybean trypsin inhibitor, 0.5 mM benzamidine, and 0.1 mM phenylmethylsulfonyl fluoride, pH 6.5) containing 25 µg/ml saponin for 10 min. Permeabilized monolayers were washed twice in contraction buffer (50 mM KCl, 25 mM PIPES, 10 mM imidizole, 1 mM EGTA, 1 mM MgSO₄, 0.2 mM DTT, 5 µg/ml each aprotinin, leupeptin, pepstatin, and chymostatin, 10 µg/ml soybean trypsin inhibitor, 0.5 mM benzamidine, and 0.1 mM phenylmethylsulfonyl fluoride, pH 7.0) (18, 23, 26) and allowed to stabilize for 10 min. Monolayers were stimulated to contract by addition of MgATP and Ca²⁺ as described previously (18, 23). The free Ca²⁺ concentration of contraction buffer, a Ca²⁺-EGTA buffer system, was determined as outlined previously (18).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

PAK2 Phosphorylation of MLCK-- MLCK phosphorylation by PAK2 in vitro using highly purified enzymes is shown in Fig. 1. Trypsin-activated (lane 1) or Cdc42-GTPS-activated (lane 2) placenta PAK2 as well as recombinant constitutively active PAK2 (lane 3) were incubated in the presence of purified MLCK for 10 min at 30 °C and MLCK phosphorylation assessed by SDS-PAGE and autoradiography. In the absence of CaM, both placenta PAK2 and recombinant constitutively active PAK2 catalyze efficient phosphorylation of MLCK. MLCK phosphorylation occurred with both Cdc42-GTPS activation and trypsin activation of placenta PAK2, demonstrating that the reaction was catalyzed by both holoenzyme and the catalytic domain of PAK2. Activated placenta and recombinant PAK2 both phosphorylated MLCK to the same extent. Using a molar ratio of MLCK:PAK2 of 10:1, a stoichiometry of 1.71 ± 0.21 mol of PO₄/mol of MLCK was measured. In the presence of unactivated PAK2 or in the absence of PAK2, there was no detectable MLCK phosphorylation (data not shown).

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Phosphorylation of MLCK by PAK2. Purified MLCK was phosphorylated by either trypsin-activated (lane 1), Cdc42-GTPS placenta PAK2 (lane 2), or by recombinant constitutively active PAK2 (lane 3) for 10 min at 30 °C. Phosphorylation was stopped by addition of equal volumes of 20% trichloroacetic acid, proteins analyzed on 7.5% SDS gels, and phosphorylated bands detected by autoradiography.

A one-dimensional tryptic phosphopeptide map and phosphoamino acid analysis were generated from the phosphorylated MLCK shown in Fig. 1 (lane 3). Phosphorylated MLCK was excised from the gel and digested for 24 h with TPCK-trypsin. Half of the sample was used for phosphopeptide mapping, and the remainder was acid hydrolyzed for phosphoamino acid analysis. Three major MLCK phosphopeptides were detected by autoradiography (Fig. 2A). These peptides are designated P1, P2, and P3 in the order of increasing acidity. The most acidic peptide (P3) has a pI of approximately 4.8, whereas the most basic peptide (P1) has a pI of approximately 7.5; P2 has an approximate pI of 5.8. Thin layer electrophoresis of the acid-hydrolyzed sample showed a single ³²P-labeled spot, which comigrated with the phosphoserine standard (Fig. 2B). On the basis of these results, we concluded that all three phosphopeptides were phosphorylated on serine residues.

View larger version (100K): [in this window] [in a new window]   Fig. 2.   One-dimensional tryptic peptide mapping and phosphoamino acid analysis of MLCK phosphorylated by PAK2. Protein band from Fig. 1 (lane 3) was excised from the gel and digested with TPCK-trypsin as outlined under "Experimental Procedures." Half of the sample was used for one-dimensional tryptic peptide mapping, while the remaining sample was subjected to acid hydrolysis in 6 N HCl for 3 h at 100 °C for phosphoamino acid analysis. A, phosphopeptide map of MLCK phosphorylated by PAK2 consists of three major phosphopeptides designated P1, P2, and P3 in order of increasing acidity. B, acid hydrolysates analyzed by TLC electrophoresis and autoradiography. Migration positions of phosphoserine, phosphothreonine, and phosphotyrosine standards are shown in lane 1. Lane 2 illustrates that only serine residues are phosphorylated by PAK2.

Identification of MLCK Phosphorylation Sites-- In order to identify the specific serines within MLCK phosphorylated by PAK2, phosphorylated MLCK was trypsin digested and the tryptic phosphopeptides isolated by C₁₈ reverse phase HPLC chromatography. Two major peaks were obtained. To achieve further purification, these fractions were rechromatographed on a Gilson SynChropak column as outlined under "Experimental Procedures." Two phosphopeptide peaks were isolated and are identified as fractions I and II. The amino acid sequences of these fractions were determined by automated degradation using an Applied Biosystems sequencer.

A single peptide with the amino acid sequence AIGRLSSMAMISGL was obtained from analysis of fraction II (Table I). This sequence corresponds to amino acids 986-999 in rabbit smMLCK (20). ³²PO₄ significantly above background was detected in rounds 6 (85%) and 7 (13%), indicating that Ser-991 is phosphorylated by PAK2.

Analysis of fraction I from multiple samples (n = 4) consistently gave a mixture of peptides containing the majority of ³²PO₄ in sequencing rounds 1 (17%) and 4 (52%). Alignment of amino acids identified in each round with the predicted smMLCK (20) tryptic peptides produced partial peptide sequences matching the smMLCK sequence R⁴³⁶PKSSLPPVLGTESD⁴⁵⁰ as shown in Table I. Although not conclusive, these data are consistent with phosphorylation of Ser-439 based on the occurrence of radioactivity. Alternative cleavage of MLCK by trypsin after residue Arg-435 or Lys-438 would produce the two phosphopeptides shown in Table I. Also of note is that the sequence GTES in sequencing rounds 8-11 does not occur at any other position in MLCK, further strengthening the conclusion that the phosphorylated residue at this phosphorylation site is Ser-439. The sequences at both phosphorylation sites are consistent with the specificity determinants identified in previous studies (27-29). Phosphorylation of synthetic peptides from the PAK substrates H4 (27), S6 (28), and RLC (10, 29) has demonstrated that optimum reactivity is observed with Arg at position P-1 or P-3. In addition, the presence of a second basic residue at P-1 to P-3 enhances substrate phosphorylation (10, 27). Several, but not all, substrates also contain viscinal Ser/Thr residues at the phosphorylation sites (28, 29).

To unambiguously identify the sites of MLCK phosphorylation by PAK2, site-directed recombinant MLCKs in which putative Ser phosphorylation sites were mutated to Ala were prepared. The mutant and WT MLCK proteins were expressed in REF-52 cells, immunopurified, phosphorylated by PAK2, and separated by SDS-PAGE as described under "Experimental Procedures." To quantitatively assess the extent of MLCK phosphorylation, MLCK phosphorylation is reported as the ratio of ³²PO₄ labeled MLCK to Coomassie Blue-stained MLCK with all values normalized to WT MLCK (1.0). Table II shows the results of PAK2-catalyzed phosphorylation of site-specific MLCK mutants. Mutation of Ser-439 or Ser-991 to Ala as well as the double mutation of Ser-991/Ser-992 to Ala resulted in a 35-50% reduction in ³²PO₄ incorporation in these mutant MLCKs (Table II).

Phosphorylated WT and mutant MLCK proteins were also analyzed by one-dimensional tryptic peptide mapping (Fig. 3, A and B). Three major phosphopeptide bands, designated as P1, P2, and P3 as in Fig. 2, were routinely observed in WT MLCK tryptic maps (Fig. 3, A and B, WT). ³²PO₄ incorporation into phosphopeptide P3 appeared to be consistent between tryptic preparations, while ³²PO₄ incorporation into phosphopeptides P1 and P2 was much more variable.

View larger version (63K): [in this window] [in a new window]   Fig. 3.   Phosphorylation of WT and mutant MLCK proteins by PAK2. Site-specific MLCK mutants in which putative PAK2 phosphorylation sites (serine residues) were replaced with alanine were expressed in REF-52 cells, immunopurified, phosphorylated by PAK2 for 30 min at 30 °C, and analyzed by one-dimensional tryptic peptide mapping. A, three major tryptic phosphopeptides of WT MLCK, P1, P2, and P3 as in Fig. 2 (WT). Substitution of Ser-991 with Ala (S991A) causes complete loss of P1 and partial loss of P2. Mutation of both Ser-991 and Ser-992 to Ala (S991A/S992A) results in complete loss of both P1 and P2. Mutation of Ser-992 (S992A) or Ser-1005 (S1005A) results in no change in the phosphopeptide map. The phosphopeptides P1and P2 result from alternative trypsin cleavage of MLCK and are designated as S991 and S991'. B, mutation of Ser-439 to Ala (S439A) results in a shift in mobility (*) of P3 and a decrease in 32PO4 incorporation.

Mutation of Ser-991 results in complete loss of phosphopeptide P1 and a marked decrease of phosphopeptide P2 (Fig. 3A, S991A) without altering the incorporation of ³²PO₄ into phosphopeptide P3, suggesting that both P1 and P2 are derived from alternatively cleaved MLCK tryptic peptides that contain Ser-991. Mutation of Ser-992 to Ala resulted in a tryptic map unchanged from WT MLCK while double mutation at Ser-991/992 (Fig. 3A, S991A/S992A) resulted in complete loss of both phosphopeptides (P1 and P2), confirming the conclusion that Ser-991 is a phosphorylation site for PAK2. Ser-992 appears to be an inefficient phosphorylation site for PAK2 when Ser-991 is mutated to Ala.

Mutation of Ser-439 resulted in significant reduction in ³²PO₄ incorporation into phosphopeptide P3 (Fig. 3B, S439A) as well as a slight shift in the electrophoretic mobility of the phosphopeptide. These results, together with the amino acid sequence analysis, provide strong evidence that Ser-439 is a PAK2 phosphorylation site. The residual phosphoryation seen in phosphopeptide P3 of the mutant MLCK could be explained by inefficient phosphorylation of Ser-440 when Ser-439 is mutated to Ala.

Cyclic AMP-dependent kinase (30, 31), protein kinase C (32), and Ca²⁺/CaM-dependent protein kinase II (33) phosphorylate MLCK at multiple sites in vitro. The predominant sites are Ser-992 and Ser-1005. Recombinant MLCK mutants that replace these residues with alanine were generated in order to demonstrate conclusively that PAK2 does not phosphorylate these residues. Replacement of either Ser-992 or Ser-1005 (Fig. 3A, S992A and S1005A) with Ala or double mutation of Ser-992/1005 to Ala (data not shown) results in no change in ³²PO₄ incorporation (Table II) or in phosphopeptide maps of these MLCK mutants.

Effect of PAK2 on MLCK Activity-- To determine if phosphorylation of MLCK by PAK2 modified MLCK activity, myosin II RLC phosphorylation was assayed. Immunopurified endothelial cell myosin II was incubated in the presence or absence of activated MLCK, activated PAK2, or MLCK prephosphorylated by PAK2. The extent of RLC phosphorylation was assessed by glycerol/urea gel electrophoresis as shown in Fig. 4. MLCK catalyzes incorporation of 1.8 mol of PO₄/mol of RLC (lane 2) and PAK2 catalyzes incorporation of 0.9 mol of PO₄/mol of RLC (lane 4). As previously shown (6), this stoichiometry corresponds to predominantly diphosphorylation and monophosphorylation of RLC by the respective enzymes. When MLCK is prephosphorylated by PAK2, PAK2 removed from the mix, and MLCK added to myosin II in the presence of ATP/Ca²⁺/CaM, there is no significant increase in RLC phosphorylation (0.5 mol of PO₄/mol of RLC, lane 3) when compared with the buffer control (0.4 mol of PO₄/mol of RLC, lane 1). In addition, when PAK2, MLCK, and myosin II are incubated together and then stimulated with ATP/Ca²⁺/CaM, RLC phosphorylation stoichiometrically is comparable to that of PAK2 alone (data not shown). This result clearly demonstrates that phosphorylation of MLCK by PAK2 inhibits MLCK-catalyzed phosphorylation of myosin II RLC.

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Effects of PAK2 on MLCK-induced RLC phosphorylation. Immunopurified myosin II was incubated for 10 min at 30 °C as follows: lane 1, buffer only; lane 2, 100 nM MLCK + Ca 2+/CaM; lane 3, 100 nM PAK2 + 100 nM MLCK were incubated in the presence of ATP for 10 min at 30 °C, PAK2 removed by immunoprecipitation, and the MLCK added to myosin II containing Ca2+/CaM; lane 4, 100 nM activated PAK2.The phosphorylated states of myosin II were analyzed on glycerol/urea gels, blotted to nitrocellulose, and probed with anti-myosin RLC antibody (6).

Effect of Ca²⁺/CaM on PAK2 Phosphorylation of MLCK-- Because activated MLCK is bound to Ca²⁺/CaM in vivo, the effect of CaM on the ability of PAK2 to phosphorylate MLCK was tested. Results are shown in Fig. 5A. MLCK (1 µM) was preincubated with 1 µM CaM/500 µM CaCl₂ for 15 min prior to the addition of 100 nM activated PAK2 and [-³²P]ATP (lane 3) or 1 µM CaM, 500 µM CaCl₂, 100 nM activated PAK2, and [-³²P]ATP were added simultaneously to MLCK (lane 5). After a 15-min incubation in the presence of all additions, half of each reaction mixture was removed and proteins precipitated by addition of trichloroacetic acid (lanes 1, 3, and 5). To the remaining reaction mixtures, EGTA was added to a final concentration of 2 mM and these were incubated for an additional 15 min before addition of equal volumes of 20% trichloroacetic acid (lanes 2, 4, and 6). In the absence of Ca²⁺/CaM (lane 1), PAK2-catalyzed MLCK phosphorylation and the addition of EGTA had no effect on PAK2 ability to phosphorylate MLCK (lane 2). In contrast, preincubation of MLCK with Ca²⁺/CaM (lane 3) or simultaneous addition of Ca²⁺/CaM (lane 5), significantly blocked PAK2-catalyzed phosphorylation of MLCK. Autophosphorylation of PAK2 was evident in these samples, demonstrating that PAK2 was not inactivated. Ca²⁺/CaM, in concentrations ranging from 1 to 12 µM (data not shown), consistently reduced PAK2-catalyzed MLCK phosphorylation by 80%, based on two-dimensional laser densitometric quantitation of autoradiographs. The inhibitory effect of Ca²⁺/CaM on PAK2-catalyzed MLCK phosphorylation is confirmed by the observation that chelation of Ca²⁺ with EGTA restored phosphorylation of MLCK by PAK2 (Fig. 5A, lanes 4 and 6).

View larger version (88K): [in this window] [in a new window]   Fig. 5.   Effect of Ca2+/CaM on PAK2-induced MLCK phosphorylation. A, purified MLCK was incubated with 100 nM PAK2 under the conditions described for each lane of the gel. After a 15-min incubation at 30 °C, half of each reaction mixture was removed and proteins trichloroacetic acid-precipitated (lanes 1, 3, and 5). EGTA was added to a final concentration of 2 mM to the remaining reaction mixture and incubated for an additional 15 min before proteins were precipitated (lanes 2, 4, and 6). Proteins were analyzed on 7.5% SDS gels and gels exposed to PhosphorImager plates. Lane 1, 1 µM MLCK incubated in the presence of 100 nM PAK for 15 min at 30 °C; lane 2, addition of 2 mM EGTA; lane 3, 1 µM MLCK preincubated in buffer containing 1 µM CaM and 500 µM Ca2+ at room temperature for 15 min before addition of 100 nM PAK2 and further incubated at 30 °C for 15 min; lane 4, addition of 2 mM EGTA; lane 5, simultaneous addition of 1 µM MLCK, 100 nM PAK2, 1 µM CaM, and 500 µM Ca2+ for 15 min at 30 °C; lane 6, addition of 2 mM EGTA. B, one-dimensional tryptic peptide map of MLCK phosphorylated by PAK2 in the presence or absence of Ca2+/CaM. MLCK (1 µM) was phosphorylated by 100 nM PAK2 either in the absence (lane 1) or the presence (lanes 2 and 3) of Ca2+/CaM; the proteins were trichloroacetic acid-precipitated and separated on 7.5% SDS gels. The MLCK bands were excised and digested for tryptic peptide mapping. Because of the low amount of 32PO4 incorporation into MLCK in the presence of Ca2+/CaM, approximately 10-fold more sample was loaded in lanes 2 and 3. In the absence of Ca2+/CaM, S439, S991, and S991' phosphopeptides are present. Ca2+/CaM, either prebound to MLCK or added simultaneously with PAK2, completely blocks the phosphorylation of Ser-991 as indicated by loss of phosphopeptide bands S991 and S991'. Lane 1, MLCK (1 µM) incubated with 100 nM PAK2 in the absence of CaM; lane 2, MLCK (1 µM) preincubated in buffer containing 1 µM CaM and 500 µM Ca2+for 15 min before addition of 100 nM PAK2 for 15 min at 30 °C; lane 3, simultaneous incubation of 1 µM MLCK, 100 nM PAK2, 1 µM CaM, and 500 µM Ca2+ for 15 min at 30 °C.

Tryptic phosphopeptide maps of MLCK phosphorylated by PAK2 in the presence or absence of Ca²⁺/CaM are shown in Fig. 5B. The phosphorylated MLCK bands from lanes 1, 3, and 5 (Fig. 5A) were excised from the gel, trypsin-digested, and analyzed by isoelectric focusing. Because of the low amount of MLCK phosphorylation in the presence of Ca²⁺/CaM, approximately 10-fold more sample was loaded (Fig. 5B, lanes 2 and 3) than the sample of MLCK phosphorylated in the absence of CaM (Fig. 5B, lane 1). When MLCK is phosphorylated by PAK2 in the presence of Ca²⁺/CaM, only Ser-439 is phosphorylated. Either preincubation of MLCK with Ca²⁺/CaM (Fig. 5B, lane 2) or addition of Ca²⁺/CaM simultaneously with PAK2 (Fig. 5B, lane 3) resulted in complete loss of phosphate incorporation into peptides P1 and P2, which correspond to phosphorylation at Ser-991.

Effect of PAK2 on Isometric Tension and RLC Phosphorylation-- To explore the physiological consequences of PAK2-catalyzed MLCK phosphorylation, tension development was monitored in permeabilized EC monolayers. In the absence of external Ca²⁺ (1 mM EGTA), addition of MgATP resulted in minimal change in isometric tension (Fig. 6A). In contrast, a rapid development of isometric tension occurred upon addition of 500 µM MgATP to permeabilized monolayers bathed in media containing 500 µM Ca²⁺ (Fig. 6B), indicating the retention of MLCK in permeabilized cell preparations (18, 23). Tension evoked by ATP and Ca²⁺ reached a maximal level of 190 dynes within 6 min and was maintained for the duration of the experiment. In contrast, monolayers preincubated in contraction buffer containing Ca²⁺ and 100 nM recombinant constitutively active PAK2 exhibited a slower rate and degree of tension production when stimulated by MgATP (Fig. 6C). Tension reached a peak force equivalent to 25% of that achieved by addition of MgATP alone. Additionally, similar experiments have shown that constitutively active PAK2 induces tension development in permeablized monolayers pretreated with KT5926, a known inhibitor of MLCK (data not shown), demonstrating that PAK2 alone causes tension development. These results are consistent with our recent studies demonstrating the ability of PAK2 to induce cell retraction (7). Under the same conditions, inactive placenta PAK2 in the presence of MgATP did not cause a reduction in tension (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effect of PAK2 on isometric tension in saponin-permeabilized EC monolayers. Monolayers were permeabilized with 25 µg/ml saponin as described previously (18). A, control monolayers incubated in contraction buffer (18, 23) containing 1 mM EGTA produce no tension upon addition of 500 µM MgATP. B, permeabilized monolayers incubated in contraction buffer containing 500 µM Ca2+ contract upon addition of MgATP. Addition of CaM had no effect on isometric tension development. C, saponin-permeabilized monolayers preincubated in contraction buffer containing 100 nM PAK2 for 10 min prior to addition of Ca2+ and MgATP exhibit a 75% reduction in tension compared with ATP treatment alone (B). Inset, after a 30-min incubation, monolayers were snap-frozen, mysoin II immunoprecipitated, and MLC phosphorylation analyzed by glycerol/urea electrophoresis. Control monolayer (lane 1; A) exhibited only unphosphorylated RLC. Incubation in Ca2+/ATP (lane 2; B) resulted in both mono- and diphosphorylated RLC. Diphosphorylation of RLC was almost completely inhibited in monolayers preincubated in PAK2 (lane 3; C).

In a parallel set of experiments in permeabilized EC monolayers, the extent of myosin RLC phosphorylation was determined by glycerol/urea gel electrophoresis in order to correlate tension development with myosin II activation. The inset in Fig. 6 shows the extent of RLC phosphorylation corresponding to tension tracings 6A, 6B, and 6C. In the absence of external Ca²⁺, RLC is exclusively in the unphosphorylated state (inset, lane 1). After addition of MgATP in the presence of Ca²⁺, there are 1.0 mol of PO₄ incorporated/mol of RLC with approximately equal amounts of mono- and diphosphorylated RLC (inset, lane 2). Monolayers preincubated with PAK2 showed 0.3 mol of PO₄ incorporated/mol of RLC with virtual absence of diphosphorylated RLC (inset, lane 3).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The novel observation that both the Ca²⁺/CaM-dependent MLCK and the Cdc42/GTP-dependent PAK2 promote phosphorylation of myosin II (6, 9) and the development of isometric tension in nonmuscle cells (this report), invites the hypothesis that cytoskeletal reorganization can occur by both a calcium-dependent and a calcium-independent pathway. MLCK is activated by the binding of Ca²⁺/CaM in response to a variety of agonists including thrombin and histamine in HUVE cells (14, 34), IgE in RBL-2H3 cells (35), and LPA in fibroblast cells (36). In nonmuscle cells, the active MLCK complex catalyzes diphosphorylation of myosin II RLC (14, 34-36) and subsequent development of myosin ATPase activity (37), increased filament formation (14, 38), and development of isometric tension (14, 36). Changes in cell physiology arising from this activation pathway include platelet aggregation (37), increased permeability (34), and cell secretion (35, 38).

The activation of PAK2 and the role of PAK2 in nonmuscle myosin-mediated contractile events is less completely understood. PAK2 is the predominant nonmuscle isoform of the PAK family protein kinases (1, 2). These enzymes are activated by binding Cdc42-GTP or rac-GTP (1, 2), and there is some suggestion that rac-GTP may be activated by the Cdc42-PAK complex so that both enzymes participate in the cellular response. No activation or regulation by calcium has been described for these protein kinases. Initial evidence that one substrate of PAK is the nonmuscle myosin II was reported over a decade ago (10). Since PAK2 activation by Cdc42/rac-GTP was not known at that time, activation in the initial studies was accomplished by in situ production of the PAK catalytic core by limited trypsin digestion. Using this method, Masaracchia and co-workers (10) established that PAK2 catalyzed phosphorylation of the nonmuscle myosin II light chain in myosin II and isolated myosin II light chains (10). Skeletal muscle myosin was not a substrate for PAK. Using synthetic peptides, the RLC phosphorylation site was identified as the same serine residue modified by the Ca²⁺/CaM-dependent MLCK (10). Studies have established that PAK2 activated by Cdc42/GTP also catalyzes RLC phosphorylation at the regulatory Ser-19 site in nonmuscle myosin II (6, 11, 12), isolated myosin II RLC (6, 9), and permeabilized bovine pulmonary artery endothelial cells.² Immunocytochemical studies using a Ser-19-specific RLC antibody have shown that transient transfection of PAK1 causes an increase in MLC phosphorylation at Ser-19 (11, 12). In addition, microinjection or osmotic loading of constitutively active PAK2 in endothelial cells induces monophosphorylation of myosin II (7). No diphosphorylation of myosin light chains is observed with this kinase. The deletion of PAK2 substrate specificity determinants either in synthetic peptides corresponding to the myosin light chain phosphorylated sequence (9, 10) or in recombinant light chains mutated at key basic residues (6) substantiate the direct data that the regulatory light chain Ser-19 is a target for Cdc42-dependent enzyme. Finally, the K_m (12 µM) observed for RLC phosphorylation by PAK2 activated by either trypsin (10) or Cdc42 (6, 10) is consistent with the conclusion that the RLC are the most likely in vivo substrate for PAK2 identified to date.

Since the identification of other specific cellular substrates of PAK2 has met with limited success, and since PAK2 has been shown to promote actin reorganization in PAK-transfected cells (1-5) and PAK-microinjected cells (3, 7), the identification of nonmuscle myosin II as a substrate for the PAK2 isoform is a significant observation. The data suggest that nonmuscle myosin II activation and cytoskeletal reorganization can be initiated by a Ca²⁺-independent mechanism. There are few agonist-dependent PAK2 data to draw upon in analyzing this hypothesis since agonists that activate PAK2 in vivo have not been extensively characterized. Nerve growth factor promotes PAK2-mediated neurite outgrowth and development in cultured PC12 cells (39), bradykinin initiates Cdc42-dependent microspike and filopodia formation in Swiss 3T3 fibroblasts (40), and in T lymphocytes both growth factors (41) and T cell activators are required for PAK2 activation (42). Other studies suggest that growth factor-initiated response cascades and PAK-regulated response cascades function in parallel with some cross-talk between the events at key points (43). Extending the original observations that STE20, the yeast PAK homologue, is activated by osmotic challenge (1, 2), several investigators have provided evidence that stress stimuli activate PAK, which in turn regulates the terminal stress kinases JNK and p38 (44). Finally, some data suggest that PAK2 may mediate apoptosis (45). Collectively, the agonists that are known to activate MLCK and the putative agonists for PAK2 are not overlapping with the possible exception of thrombin (46). Furthermore, the physiological end points observed after MLCK and PAK2 activation are not redundant since MLCK does not activate p38 or JNK. The lack of commonality among agonists and the different morphological responses seen in the Ca²⁺-dependent and Ca²⁺-independent mechanisms strongly suggest that these pathways mediate cytoskeletal reorganization in different physiological contexts.

In addition to the likelihood that different agonists activate the MLCK and PAK2 pathways, the modification of myosin II itself in Ca²⁺-dependent and Ca²⁺-independent responses probably differs. MLCK causes diphosphorylation of the myosin II RLC (6, 14, 36). The exact importance of the diphosphorylation has not been unambiguously determined. Both actin-activated myosin ATPase and myosin II filament stability are increased after diphosphorylation in some smooth muscle preparations (13, 47); however, in other laboratories, maximum smooth muscle force generation was observed with myosin II monophosphorylation (48). The correlation between myosin II diphosphorylation and cellular force generation might be more germane in nonmuscle cells; however, the data available to support the functional significance of diphosphorylation in vivo are scant. Using direct tension measurements, Goeckeler and Wysolmerski (14) have shown that sustained isometric tension development correlates with myosin II diphosphorylation. In both RBL-2H3 cells (35, 38) and human platelets (37), diphosphorylation of myosin II correlates with reorganization of the actin and association of myosin II filaments. No diphosphorylation of myosin is evident in PAK2-mediated myosin II phosphorylation, perhaps suggesting that the myosin II response when this pathway is invoked differs significantly from the maximal response observed in the MLCK pathway. We consistently observe that tension development in permeabilized bovine pulmonary artery endothelial monolayers induced by PAK2 is less forceful than in MLCK-mediated retraction, an observation that is correlated with the monophosphorylation and diphosphorylation patterns observed with the two enzymes, respectively. Finally, in our studies of permeabilized cell retraction in response to MLCK activation (18, 23) and PAK2 activation (6), we observed that the extent of endothelial cell retraction differed in the Ca²⁺-dependent and Ca²⁺-independent response, although we have not investigated the physiological significance of these observations.

The occurrence of two potential pathways for the initiation of actin/myosin II reorganization in somatic cells raises the question of potential intercommunications, or cross-talk, between these two pathways, an event that might be essential to segregate independent responses or coordinate and modulate cooperative responses. MLCK is phosphorylated by a variety of protein kinases including protein kinase C (30, 31), protein kinase A (32), and CaM kinase II (33). The effects of phosphorylation by protein kinase C and protein kinase A may or may not be regulatory in vivo; however, phosphorylation of MLCK by CaM kinase II in smooth muscle cells in vivo decreases MLCK's sensitivity to activation by Ca²⁺/CaM (49, 50). Evidence for cross-talk between the PAK2 signaling pathway and other protein kinase cascades is scant. Frost et al. (43) and others (1, 2) have suggested that there may be cross-talk between growth factor pathway intermediates and PAK2, and this is supported by the observation that insulin may activate both the MAP kinase and PAK2 pathways (51). In T lymphocytes, both the NFB pathway and Cdc42/JNK pathway are activated in response to tumor necrosis factor, suggesting the two pathways may have some points of conjuncture or coregulation (41). CD28-activated T cells appear to modulate the T cell receptor response via a PAK2-dependent mechanism although details of this signaling system have not been elucidated (42). At best, cross-regulation of other protein kinase cascades and the PAK2 signaling system is implied in these studies.

Results presented in this report and others (6, 9, 14) describe in detail for the first time cross-talk between PAK2 and another agonist-mediated protein phosphorylation cascade, i.e. MLCK-induced actin-myosin II regulated nonmuscle cell retraction. MLCK is actively phosphorylated on residues Ser-439 and Ser-991 by purified activated PAK2 in the absence of Ca²⁺/CaM. In the presence of Ca²⁺/CaM, phosphorylation is significantly blocked, providing a potential regulatory mechanism by which MLCK once activated cannot be inactivated by PAK2. In vitro activity studies demonstrate that phosphorylation of MLCK by PAK2 results in inhibition of MLCK activity, i.e. RLC phosphorylation. It remains to be established whether the inhibitory activity requires phosphorylation of both Ser-439 and Ser-991 or is conferred by phosphorylation of one of these sites. Current data support a primary role for Ser-991 phosphorylation in mediating PAK2 inhibition of MLCK because 1) phosphorylation of this site is completely blocked by Ca²⁺/CaM, the complex which activates MLCK and 2) its obvious proximity to Ser-992 (site A), a site phosphorylated by other kinases known to inhibit MLCK activity in vitro and in vivo. In contrast, the Ser-439 residue is positioned within a 48-amino acid region between motif II-1, a fibronectin type III domain, and motif II-2, an immunoglobin C-2 domain, which have been implicated in protein-protein interactions (49, 52). Elucidation of the physiological significance of these phosphorylation sites clearly awaits further study.

To test the potential significance of MLCK-PAK2 interactions in vivo, tension development in endothelial cell monolayers was monitored. As expected, MLCK activation by Ca²⁺/CaM promoted rapid and vigorous isometric tension development in permeabilized endothelial cells, and this response was correlated with the occurrence of diphosphorylated myosin II RLC. In the absence of Ca²⁺/CaM, the major portion of the myosin II RLC pool was unphosphorylated. Addition of PAK2 prior to Ca²⁺ or Ca²⁺/CaM inhibited the generation of diphosphorylated myosin II RLC, consistent with the in vitro observation that PAK2 inhibits MLCK activity. In addition, tension development was diminished by 75%, consistent with the observation that tension development by PAK2 is less than that observed with MLCK, although PAK2 alone consistently induces tension development above that observed in control cells.

Since MLCK and PAK2 appear to respond to a different pool of agonists, the cross-regulation of one enzyme by the other could provide a mechanism for insuring that the degree of cytoskeletal reorganization, and indeed the type of actin/myosin II reorganization, which occurs in response to a specific agonist is appropriate for the cell response. We propose that agonists which activate the PAK2 pathway mediate events in which a less than maximal contractile force is appropriate and the inhibition of MLCK by activated PAK2 provides a mechanism to ensure that physiological response, whereas activation by Ca²⁺ in the absence of PAK2 activation results in a maximal contractile response. This response cannot be reversed by PAK2 since this enzyme cannot inhibit activated MLCK. This hypothesis would predict that the Ca²⁺/CaM-mediated contractile response and the Cdc42/GTP-mediated contractile response are not redundant but function to permit graded cytoskeletal reorganization in response to a variety of physiological signals (Fig. 7).

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Proposed model for the interaction of Ca2+/CaM-independent and Ca2+/CaM-dependent activation pathways for myosin II. Activation of the PAK pathway by Ca2+/CaM-independent agonists results in monophosphorylation of myosin II RLC as well as phosphorylation of MLCK; the latter inhibits MLCK activation by Ca2+/CaM and thus MLCK-mediated RLC phosphorylation. Ca2+/CaM-dependent agonists stimulate activation of the MLCK pathway, which catalyzes myosin II RLC diphosphorylation. Once activated, MLCK/Ca2+/CaM cannot be inhibited by PAK-mediated phosphorylation (*). The ability to regulate myosin II RLC monophosphorylation versus diphosphorylation by these pathways would serve to effect specific cellular cytoskeletal reorganizations.

	FOOTNOTES

^* This work was supported in part by the Department of Anesthesiology Research Fund (St. Louis University School of Medicine) and National Institutes of Health Grants HL-45788, HL-54245, HL-61952 (to R. B. W.), AI-39690 (to R. A. M.), and HL-54118 (to P. J. G.).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.

^** To whom correspondence should be addressed: Dept. of Pathology, St. Louis University School of Medicine, 1402 S. Grand Blvd., St. Louis, MO 63104. Tel.: 314-577-8497; Fax: 314-268-5649; E-mail: wysolmer@slucare1.sluh.edu.

Published, JBC Papers in Press, March 28, 2000, DOI 10.1074/jbc.M001339200

² R. B. Wysolmerski, unpublished observation.

	ABBREVIATIONS

The abbreviations used are: PAK, p21-activated kinase; MLCK, myosin light chain kinase; CaM, calmodulin; RLC, myosin II regulatory light chain; REF-52, rat embryo fibroblast; TPCK, tosylphenylalanyl chloromethyl ketone; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; WT, wild type; DTT, dithiothreitol; JNK, c-Jun N-terminal kinase; PIPES, 1,4-piperazinediethanesulfonic acid; GTPS, guanosine 5'-3-O-(thio)triphosphate.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit