Calmodulin-binding and Autoinhibitory Domains ofAcanthamoeba Myosin I Heavy Chain Kinase, a p21-activated Kinase (PAK)
2001; Elsevier BV; Volume: 276; Issue: 50 Linguagem: Inglês
10.1074/jbc.m108957200
ISSN1083-351X
AutoresHanna Brzeska, Rachel Young, Cristina A. Tan, Joanna Szczepanowska, Edward D. Korn,
Tópico(s)Erythrocyte Function and Pathophysiology
ResumoThe sequence homology betweenAcanthamoeba myosin I heavy chain kinase (MIHCK) and other p21-activated kinases (PAKs) is relatively low, including only the catalytic domain and a short PAK N-terminal motif (PAN), and even these regions are not highly homologous. In this paper, we report the expression in insect cells of full-length, fully regulatedAcanthamoeba MIHCK and further characterize the regulation of this PAK by Rac, calmodulin, and autoinhibition. We map the autoinhibitory region of MIHCK to its PAN region and show that the PAN region inhibits autophosphorylation and kinase activity of unphosphorylated full-length MIHCK and its expressed catalytic domain but has very little effect on either when they are phosphorylated. These properties are similar to those reported for mammalian PAK1. Unlike PAK1, MIHCK is activated by Rac only in the presence of phospholipid. However, peptides containing the PAN region of MIHCK bind Rac in the absence of lipid, and Rac binding reverses the inhibition of the MIHCK catalytic domain by PAN peptides. Our data suggest that a region N-terminal to PAN is required for optimal binding of Rac. Also unlike mammalian PAK, phospholipid stimulation ofAcanthamoeba MIHCK and Dictyostelium MIHCK) (which is also a PAK) is inhibited by Ca2+-calmodulin. In contrast toDictyostelium MIHCK, however, Ca2+-calmodulin also inhibits Rac-induced activity of Acanthamoeba MIHCK. The basic region N-terminal to PAN is essential for calmodulin binding. The sequence homology betweenAcanthamoeba myosin I heavy chain kinase (MIHCK) and other p21-activated kinases (PAKs) is relatively low, including only the catalytic domain and a short PAK N-terminal motif (PAN), and even these regions are not highly homologous. In this paper, we report the expression in insect cells of full-length, fully regulatedAcanthamoeba MIHCK and further characterize the regulation of this PAK by Rac, calmodulin, and autoinhibition. We map the autoinhibitory region of MIHCK to its PAN region and show that the PAN region inhibits autophosphorylation and kinase activity of unphosphorylated full-length MIHCK and its expressed catalytic domain but has very little effect on either when they are phosphorylated. These properties are similar to those reported for mammalian PAK1. Unlike PAK1, MIHCK is activated by Rac only in the presence of phospholipid. However, peptides containing the PAN region of MIHCK bind Rac in the absence of lipid, and Rac binding reverses the inhibition of the MIHCK catalytic domain by PAN peptides. Our data suggest that a region N-terminal to PAN is required for optimal binding of Rac. Also unlike mammalian PAK, phospholipid stimulation ofAcanthamoeba MIHCK and Dictyostelium MIHCK) (which is also a PAK) is inhibited by Ca2+-calmodulin. In contrast toDictyostelium MIHCK, however, Ca2+-calmodulin also inhibits Rac-induced activity of Acanthamoeba MIHCK. The basic region N-terminal to PAN is essential for calmodulin binding. myosin I heavy chain kinase CDC42/Rac interactive binding polyacrylamide gel electrophoresis, PAK, p21-activated kinase, PAN, PAK N-terminal motif p21-binding domain polymerase chain reaction inhibitory switch base pair(s) kilobase(s) nitrilotriacetic acid MIHCK1 phosphorylates a single Ser or Thr in the head domain of each of the three class-I myosins from Acanthamoeba castellanii substantially increasing their actin-activated MgATPase activities (for reviews, see Refs. 1Brzeska H. Korn E.D. J. Biol. Chem. 1996; 271: 16983-16986Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar and 2Barylko B. Binns D.D. Albanesi J.P. Biochim. Biophys. Acta. 2000; 1496: 23-35Crossref PubMed Scopus (49) Google Scholar). When the amino acid sequence of its catalytic domain is compared with the catalytic domains of other kinases (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar), MIHCK is most similar to PAK1, a member of the PAK-I family of the mammalian Ste20 group kinases (according to the classification system of Dan et al. (4Dan I. Watanabe N.M. Kusumi A. Trends Cell Biol. 2001; 11: 220-230Abstract Full Text Full Text PDF PubMed Scopus (508) Google Scholar)). The PAK-I family (for reviews, see Refs. 5Manser E. Lim L. Prog. Mol. Subcell. Biol. 1999; 22: 115-133Crossref PubMed Scopus (57) Google Scholar, 6Sells M.A. Chernoff J. Trends Cell Biol. 1997; 7: 162-167Abstract Full Text PDF PubMed Scopus (265) Google Scholar, 7Daniels R.H. Bokoch G.M. Trends Biochem. Sci. 1999; 24: 350-355Abstract Full Text Full Text PDF PubMed Scopus (227) Google Scholar, 8Bagrodia S. Cerione R.A. Trends Cell Biol. 1999; 9: 350-355Abstract Full Text Full Text PDF PubMed Scopus (343) Google Scholar) share an homologous C-terminal catalytic domain and a conserved autoregulatory region (9Lei M. Lu W. Meng W. Parrini M.C. Eck M.J. Mayer B.J. Harrison S.C. Cell. 2000; 102: 387-397Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar) in the N-terminal half, also called PAN for PAK N-terminal motif (5Manser E. Lim L. Prog. Mol. Subcell. Biol. 1999; 22: 115-133Crossref PubMed Scopus (57) Google Scholar,10Zhao Z.-S. Manser E. Chen X.-Q. Chong C. Leung T. Lim L. Mol. Cell. Biol. 1998; 1: 2153-2163Crossref Google Scholar). This region is responsible for autoinhibition of mammalian PAK1, -2, and -3 (members of the PAK-I family) and for binding of Rac and Cdc42, which reverses autoinhibition (9Lei M. Lu W. Meng W. Parrini M.C. Eck M.J. Mayer B.J. Harrison S.C. Cell. 2000; 102: 387-397Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar, 10Zhao Z.-S. Manser E. Chen X.-Q. Chong C. Leung T. Lim L. Mol. Cell. Biol. 1998; 1: 2153-2163Crossref Google Scholar, 11Thompson G. Owen D. Chalk P.A. Lowe P.N. Biochemistry. 1998; 37: 7885-7891Crossref PubMed Scopus (121) Google Scholar, 12Frost J.A. Khokhlatchev A. Stippec S. White M.A. Cobb M.H. J. Biol. Chem. 1998; 273: 28191-28198Abstract Full Text Full Text PDF PubMed Scopus (169) Google Scholar, 13Stevens W.K. Vranken W. Goudreau N. Xiang H. Xu P. Ni F. Biochemistry. 1999; 38: 5968-5975Crossref PubMed Scopus (16) Google Scholar, 14Zenke F.T. King C.C. Bohl B.P. Bokoch G.M. J. Biol. Chem. 1999; 274: 32565-32573Abstract Full Text Full Text PDF PubMed Scopus (203) Google Scholar). As described by Leiet al. (9Lei M. Lu W. Meng W. Parrini M.C. Eck M.J. Mayer B.J. Harrison S.C. Cell. 2000; 102: 387-397Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar), the ∼80-residue autoregulatory region of mammalian PAK1 consists of an N-terminal p21-binding domain (PBD) of ∼44 residues, an overlapping inhibitory switch (IS) domain of ∼50 residues that includes the C-terminal ∼27 residues of the PBD and a C-terminal ∼14-residue kinase inhibitory domain. A CRIB (Cdc42/Rac interactive binding) motif (15Burbelo P.D. Drechsel D. Hall A. J. Biol. Chem. 1995; 270: 29071-29074Abstract Full Text Full Text PDF PubMed Scopus (557) Google Scholar) of ∼16 residues near the N terminus of the autoregulatory region is an essential component of the PBD. PAK4 (16Abo A. Qu J. Cammarano M.S. Dan C. Fritsch A. Baud V. Belisle B. Minden A. EMBO J. 1998; 17: 6527-6540Crossref PubMed Scopus (316) Google Scholar) and other PAK-II family members also have a C-terminal catalytic domain and an N-terminal PBD but lack a recognizable autoinhibitory domain (4Dan I. Watanabe N.M. Kusumi A. Trends Cell Biol. 2001; 11: 220-230Abstract Full Text Full Text PDF PubMed Scopus (508) Google Scholar). Acanthamoeba MIHCK has a C-terminal catalytic domain, a region of sequence homology to the PAK1 PBD, including the CRIB motif, and an IS domain, but MIHCK does not have a region with sequence homology to the kinase inhibitory domain of PAK1. Therefore, by sequence alone it is difficult to determine whether MIHCK more closely resembles the PAK-I or PAK-II family. Following the region of homology with PAK1, the MIHCK sequence becomes extremely Pro-rich. We will use the terms PAN or putative autoregulatory region when referring to the segment of MIHCK that starts with the CRIB motif and ends just before the Pro-rich region. The C-terminal catalytic domain and the N-terminal PAN ofAcanthamoeba MIHCK have relatively low sequence homology to most PAKs (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar), and the remainder of the Acanthamoeba MIHCK sequence has no homology to either mammalian PAKs orDictyostelium MIHCK, which is also a PAK (17Lee S.-F. Egelhoff T.T. Mahasneh A. Côté G.P. J. Biol. Chem. 1996; 271: 27044-27048Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar). However, the substrate specificities and regulation of the activities of PAK1, the best characterized mammalian PAK, and the two MIHCKs are surprisingly similar but with interesting differences (18Brzeska H. Knaus U.G. Wang Z.-Y. Bokoch G.M. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 1092-1095Crossref PubMed Scopus (70) Google Scholar, 19Wu C. Lee S.-F. Furmaniak-Kazmierczak E. Côté G.P. Thomas D.Y. Leberer E. J. Biol. Chem. 1996; 271: 31787-31790Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar). For example, although all three kinases are activated in vitro by autophosphorylation, Rac or lipids are required for autophosphorylation of PAK1 (20Manser E. Leung T. Salihuddin H. Zhao Z.-S. Lim L. Nature. 1994; 367: 40-46Crossref PubMed Scopus (1300) Google Scholar, 21Bokoch G.M. Reilly A.M. Daniels R.H. King C.C. Olivera A. Spiegel S. Knaus U.G. J. Biol. Chem. 1998; 273: 8137-8144Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar), but Acanthamoeba MIHCK (22Brzeska H. Lynch T.J. Korn E.D. J. Biol. Chem. 1990; 265: 3591-3594Abstract Full Text PDF PubMed Google Scholar) can be fully autophosphorylated and activated and Dictyostelium MIHCK (23Lee S.-F. Mahasneh A. DeLaRoche M. Côté G.P. J. Biol. Chem. 1998; 273: 27911-27917Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar) partially activated in the absence of lipids and Rac. Also, the lipids that activate mammalian PAK1 (21Bokoch G.M. Reilly A.M. Daniels R.H. King C.C. Olivera A. Spiegel S. Knaus U.G. J. Biol. Chem. 1998; 273: 8137-8144Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar) differ from those that activate Acanthamoeba (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar, 24Brzeska H. Lynch T.J. Martin B.M. Corigliano-Murphy A. Korn E.D. J. Biol. Chem. 1990; 265: 16138-16144Abstract Full Text PDF PubMed Google Scholar) and Dictyostelium(23Lee S.-F. Mahasneh A. DeLaRoche M. Côté G.P. J. Biol. Chem. 1998; 273: 27911-27917Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar) MIHCK. Furthermore, Ca2+-calmodulin inhibits lipid-stimulated activation of Acanthamoeba (25Brzeska H. Kulesza-Lipka D. Korn E.D. J. Biol. Chem. 1992; 267: 23870-23875Abstract Full Text PDF PubMed Google Scholar) andDictyostelium (23Lee S.-F. Mahasneh A. DeLaRoche M. Côté G.P. J. Biol. Chem. 1998; 273: 27911-27917Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar) MIHCK but not of PAK1. Thus, although the sequence homology between the two MIHCKs is no greater than the sequence homology between either MIHCK and PAK1, there are more common elements in the regulation of the activity of the two MIHCKs, i.e. the biochemical properties of at least these three PAKs are not easily predictable from their protein sequences. Therefore, it is important to identify the elements of the primary structure that are responsible for the particular properties of each member of the PAK family. In this paper we further characterize some aspects of the regulation of the activity of Acanthamoeba MIHCK using fully regulated, full-length kinase and its catalytic domain expressed in insect cells and Escherichia coli-expressed kinase peptides corresponding to regions outside of the catalytic domain. We show that Rac-induced activity of Acanthamoeba MIHCK is inhibited by Ca2+-calmodulin, as shown previously for lipid-induced activity (25Brzeska H. Kulesza-Lipka D. Korn E.D. J. Biol. Chem. 1992; 267: 23870-23875Abstract Full Text PDF PubMed Google Scholar), and localize the calmodulin-binding region to a segment N-terminal to PAN. We show also that the putative autoregulatory region inhibits the activity of unphosphorylated full-length Acanthamoeba MIHCK and its catalytic domain and that residues N-terminal to PAN are required for optimal binding of Rac. To construct a full-length cDNA clone, clones 43 (which lacks 23 bp at the 5′ end) and 45 (which has a 3-bp deletion, bp 613–615) (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar) were digested with KpnI and XmnI; the large digestion fragment of clone 45 (4.8 kb) was ligated to the small fragment of clone 43 (2 kb) producing plasmid pCB1 containing the complete sequence of MIHCK in pBK-CMV vector. For ease of cloning into several E. coliexpression vectors, plasmid pCB1 was than digested with NdeI and KpnI and subcloned into pET30a digested with the same enzymes producing pCB2 with several unique restriction sites after the stop codon. The NdeI/HindIII fragment of pCB2, containing the MIHCK gene, was then subcloned intoNdeI/HindIII-digested pET23a, which contains an f1 origin of replication, allowing production of single-stranded DNA for mutagenesis and a T7 promoter for E. coli expression. Site-specific mutagenesis using the Kunkel method (26Kunkel T.A. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 488-492Crossref PubMed Scopus (4900) Google Scholar) and primer 5′-TAT TCT CCA TAT GTC TAA TTC A-3′ was performed to introduce anNdeI site at the initiator ATG. The resulting plasmid (pCB4), containing two NdeI sites about 700 bp apart, was digested with NdeI and the 6.2-kb fragment ligated to itself to give pCB6 with one NdeI site at the initiator methionine. pCB6 was used for protein expression in E. coli, and the MIHCK gene from pCB6 was subcloned into a variety of E. coliexpression vectors, including ones with T7lac and PLpromoters. PCB6, whose backbone is pET23a, has a unique HindIII site 25 bp before the fusion sequence of a polyhistidine tag. The MIHCK stop codon in pCB6 was mutated to a HindIII site using 5′-AAG GAA GGA AAA GCT TGA AGA ACA TC-3′ as primer. The resulting plasmid (which has twoHindIII sites) was digested with HindIII and the largest fragment, which contained the MIHCK gene, was ligated to itself (pCB14). The expressed protein will have 13 extra amino acids (KLAAALEHHHHHH) at the C terminus. pCB14 was digested withNdeI and BlpI and filled-in using the Klenow procedure according to the manufacturer's suggestion (Stratagene, La Jolla, CA). The fragment that contained the gene was subcloned into pBlueBac 4.5 (Invitrogen Corp., Carlsbad, CA), which was digested withSacI and filled-in by the Klenow (26Kunkel T.A. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 488-492Crossref PubMed Scopus (4900) Google Scholar) procedure. The resulting plasmid, pCB15, was transfected to SF9 cells using the Bac-N-Blue transfection kit (Invitrogen). Plaque assay was done on the P0 stock, and 24 different plaques were shown by PCR to contain the recombinant viral DNA. One of the plaques was amplified (P2-stock) to a volume of 300 ml. SF9 cells in suspension culture (100 ml, cell density of 2 × 106 cells/ml) were infected with 1.5 ml of P2-stock and harvested after 3 days by centrifugation at 1,500 rpm in a Sorvall SS-34 rotor for 15 min. The cell pellet was washed once with ice-cold binding buffer (20 mm Tris, pH 7.9, 500 mmNaCl, 10 mm imidazole) and centrifuged at 1,500 rpm in a Sorvall SS-34 rotor for 10–15 min. The cells were then resuspended in 4 ml of binding buffer containing protease inhibitors (0.25 mm phenylmethylsulfonyl fluoride, 5 μg/ml leupeptin, 5 μg/ml pepstatin, 10 mm 2-mercaptoethanol), homogenized with 20–24 strokes in a Dounce homogenizer, and spun at 15,000 rpm in Sorvall SS-34 rotor for 10–20 min. The supernatant was loaded onto an 0.5-ml Ni-NTA (Qiagen, Valencia, CA) column equilibrated with binding buffer containing 10 mm 2-mercaptoethanol, and the column was washed with 20 column volumes of binding buffer containing 10 mm 2-mercaptoethanol and eluted with elution buffer (20 mm Tris, pH 7.9, 500 mm NaCl, 80 mmimidazole, 10 mm 2-mercaptoethanol). Fractions containing MIHCK were pooled and dialyzed against kinase storage buffer (20 mm Tris, pH 7.5, 50 mm KCl, 10 mm2-mercaptoethanol, 50% glycerol, 0.01% NaN3). The yield varied between 0.4 and 0.6 mg of MIHCK/100 ml of culture. The catalytic domain was expressed in SF9 cells and purified as described previously (27Brzeska H. Szczepanowska J. Korn E.D. J. Biol. Chem. 1996; 271: 27056-27062Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar). Incubation of full-length kinase and the catalytic domain with type III acid phosphatase was as described previously (28Szczepanowska J. Zhang X. Herring C.J. Qin J. Korn E.D. Brzeska H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 8503-8508Crossref PubMed Scopus (19) Google Scholar). Before treatment with λ-phosphatase, 0.1–0.5 mg of MIHCK eluted from Ni-NTA column was dialyzed against 50 mmTris, pH 7.5, 10 mm 2-mercaptoethanol, 50 mmKCl. After dialysis MnCl2, to final concentration 2 mm, and 2–10 μl of λ-phosphatase (catalog number: 753S, New England Biolabs Inc., Beverley, MA) were added, and the reaction was allowed to proceed for 30 min at 30 °C. Alternatively, kinase in storage buffer was diluted twice with 80 mm Tris, pH 7.5, and phosphatase and MnCl2 were added as above. Neither longer incubation times (up to 2 h) nor higher phosphatase concentrations resulted in further dephosphorylation (as judged by kinase activity). Kinase was separated from phosphatase by chromatography on a Ni-NTA column. The same procedures were used for dephosphorylating the expressed catalytic domain. For complete phosphorylation (and, thus, complete activation) full-length kinase (50–100 μg/ml) and catalytic domain (40 μg/ml) were incubated at 30 °C in the activity assay buffer (minus PC9) for 60 and 30 min, respectively. DNAs for expression of MIHCK peptides (see Fig. 1) were produced by PCR using full-length MIHCK cDNA as template. All of the cDNAs, except N4 cDNA, had 5′-HindIII and 3′-BglII sites introduced during PCR and were cloned between these sites into pRESET B vector (Invitrogen) creating an N-terminal poly-His-tag in the expressed peptides. N4 cDNA had 5′-BglII and 3′-HindIII sites introduced during PCR and was cloned between these sites into pFLAG-2 vector (Sigma) resulting in an N-terminal FLAG sequence in the expressed N4 peptide. Expression of these cDNAs in BL-21(DE3) cells was induced with isopropylthio-β-d-galactoside. The peptides, except N4, were purified on Ni-NTA columns as described for purification of full-length kinase. The two proline-rich fragments (M1 and M2) were additionally purified by diluting Ni-NTA column eluates with 40 mm Tris, pH 7.9, to lower the NaCl concentration to 25 mm, and loading them onto S-support cation exchange columns (Bio-Rad) and eluting with an 0.05–1 m KCl gradient in 40 mm Tris, pH 7.9. N4 was purified on a FLAG-affinity column according to the manufacturer's (Sigma) protocol. Bacterial extract containing N4 in FLAG-column buffer (containing the same protease inhibitors as for MIHCK extracts) was loaded onto a FLAG-affinity column equilibrated with 20 mm Tris, pH 7.5, 200 mm KCl, 1% Triton. The column was washed with the same buffer, and M1 and M2 peptides were eluted with the same buffer containing 0.1 mg/ml of FLAG peptide. All peptides except N1 and N2 were dialyzed against kinase storage buffer and stored at −20 °C. N1 and N2 (which precipitated in low salt buffers) were dialyzed against 20 mm Tris, pH 7.5, 300 mm KCl and frozen in small aliquots at −70 °C. Unless otherwise stated, kinase activity was assayed as described previously (24Brzeska H. Lynch T.J. Martin B.M. Corigliano-Murphy A. Korn E.D. J. Biol. Chem. 1990; 265: 16138-16144Abstract Full Text PDF PubMed Google Scholar) using synthetic peptide PC9 (200 μm), which corresponds to the phosphorylation site ofAcanthamoeba myosin IC (29Brzeska H. Lynch T.J. Martin B.M. Korn E.D. J. Biol. Chem. 1989; 264: 19340-19348Abstract Full Text PDF PubMed Google Scholar), as substrate at 20 °C in activity buffer containing 50 mm imidazole, pH 7.0, 2.5 mm [γ-32P]ATP (30,000 cpm/nmol), 3.5 mm MgCl2, 0.2 mg/ml bovine serum albumin, 1 mm EGTA. Dephosphorylated MIHCK (1.3–3 μg/ml) and catalytic domain (1.1 μg/ml) were incubated for 6–10 min when determining the inhibitory properties of kinase peptides and for 10–120 s when characterizing the properties of the expressed enzyme (the basal activity of MIHCK increases with a time of assay, because kinase autophosphorylates during assay (22Brzeska H. Lynch T.J. Korn E.D. J. Biol. Chem. 1990; 265: 3591-3594Abstract Full Text PDF PubMed Google Scholar, 24Brzeska H. Lynch T.J. Martin B.M. Corigliano-Murphy A. Korn E.D. J. Biol. Chem. 1990; 265: 16138-16144Abstract Full Text PDF PubMed Google Scholar)). Phosphorylated kinase (1.3–3 μg/ml) and catalytic domain (1.1 μg/ml) were incubated for 1–3 min. To assay calmodulin binding, peptides were incubated at room temperature with calmodulin covalently bound to agarose beads in 20 mm Tris, pH 7.5, 60 mm KCl, 25% glycerol, 0.4 mg/ml bovine serum albumin in the presence of 1 mmCaCl2 or 5 mm EGTA. Samples were pelleted and pellets were washed three times with buffer containing 50 mm Tris, pH 7.5, 200 mm KCl, and 1 mm CaCl2 or 5 mm EGTA, resuspended in SDS sample buffer, and equivalent volumes of original supernatants and washed pellets were run and analyzed by SDS-PAGE. All restriction enzymes were purchased from New England Biolabs, Inc. The pET vectors, DNA ligation kit, and competent cells were from Novagen, Inc. (Madison, WI). The pLEX expression system was from Invitrogen. M13KO7 phage and CJ 236 for mutagenesis were from New England Biolabs. The Phagemid site-specific mutagenesis kit was purchased from Bio-Rad. All materials for insect cell expression were from Invitrogen. Bovine calmodulin and calmodulin-agarose were from Calbiochem. Human GST-Rac1 was expressed in E. coli and purified on glutathione-agarose. Nucleotide exchange for Rac was performed as described previously (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar). Protein concentrations were determined by the Bradford assay, and molecular weights calculated from protein sequences were used for calculating molar concentrations. Protein sequence analyses were performed using gcg and Expasy analysis tools. All attempts to express full-length kinase in E. coli using different plasmids and different bacterial cell lines were unsuccessful with only very little or no protein expressed. However, MIHCK was readily expressed in and purified from SF9 insect cells by the procedures described under “Experimental Procedures.” Purified MIHCK was partially, and variably, active, and activity was substantially reduced by treatment with phosphatase (indicating that the expressed enzyme was partially phosphorylated) and fully activated by subsequent autophosphorylation (Table IA). As with native MIHCK purified from Acanthamoeba, activation of the expressed kinase was accelerated by addition of phosphatidylserine to the assay mixture (Table IB) and Ca2+-calmodulin inhibited the phosphatidylserine stimulation (Table IC). Ca2+-calmodulin does not directly affect the activity of Acanthamoeba MIHCK but, rather, inhibits lipid-stimulated activity by competing with lipid binding (25Brzeska H. Kulesza-Lipka D. Korn E.D. J. Biol. Chem. 1992; 267: 23870-23875Abstract Full Text PDF PubMed Google Scholar). Also, and in contrast to mammalian PAKs, activation ofAcanthamoeba MIHCK by Rac occurs only in the presence of lipids (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar). Calmodulin, in a Ca2+-dependent manner, also inhibited phosphatidylserine-dependent Rac activation of MIHCK (Table IC), presumably by inhibiting lipid binding that is required for Rac activation. To our knowledge,Acanthamoeba MIHCK is the first member of the PAK family to be expressed on a preparative scale as a fully regulated enzyme with properties indistinguishable from those of the native enzyme.Table ICharacterization of the activity of expressed Acanthamoeba myosin I heavy chain kinasePretreatment of kinaseAddition to assayActivityμmol/min·mgA. NoneNone3.29.46.3 PhosphataseNone0.20.20.7 Phosphatase and autophosphorylationNone7.28.28.4B. PhosphataseNone0.10.91.8 PhosphatasePS0.94.510.2C. PhosphataseCa2+0.2 PhosphatasePS, Ca2+0.8 PhosphatasePS, CaM, Ca2+0.1 PhosphatasePS, Rac, Ca2+12.5 PhosphatasePS, Rac, CaM, Ca2+4.8 PhosphatasePS, Rac, EGTA10.5 PhosphatasePS, Rac, CaM, EGTA11.3A, expressed MIHCK was assayed as isolated, after treatment with phosphatase, and after phosphatase treatment followed by autophosphorylation. Three different MIHCK preparations were incubated for 1 min at 20 °C, 10 s at 30 °C, or 1 min at 30 °C, left to right. B, three different preparations of expressed, dephosphorylated MIHCK were assayed with 50, 100, and 200 μm phosphatidylserine (PS), left to right, for 1 min at 30 °C, 2 min at 20 °C and 1 min at 30 °C, respectively. C, expressed MIHCK (37 nm) was dephosphorylated and its activity assayed for 1 min at 20 °C with addition of 100 μm PS, 1.5 μm GTP-Rac, and 12 μm calmodulin (CaM) in the presence of 0.1 mmCa2+ or 2 mm EGTA as shown. Open table in a new tab A, expressed MIHCK was assayed as isolated, after treatment with phosphatase, and after phosphatase treatment followed by autophosphorylation. Three different MIHCK preparations were incubated for 1 min at 20 °C, 10 s at 30 °C, or 1 min at 30 °C, left to right. B, three different preparations of expressed, dephosphorylated MIHCK were assayed with 50, 100, and 200 μm phosphatidylserine (PS), left to right, for 1 min at 30 °C, 2 min at 20 °C and 1 min at 30 °C, respectively. C, expressed MIHCK (37 nm) was dephosphorylated and its activity assayed for 1 min at 20 °C with addition of 100 μm PS, 1.5 μm GTP-Rac, and 12 μm calmodulin (CaM) in the presence of 0.1 mmCa2+ or 2 mm EGTA as shown. We had shown previously that Acanthamoeba MIHCK is activated by proteolytic removal of N-terminal segments (30Brzeska H. Martin B.M. Kulesza-Lipka D. Baines I.C. Korn E.D. J. Biol. Chem. 1992; 267: 4949-4956Abstract Full Text PDF PubMed Google Scholar, 31Brzeska H. Martin B.M. Korn E.D. J. Biol. Chem. 1996; 271: 27049-27055Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar). This, and the sequence homology between MIHCK residues 93–148 and the autoregulatory region of mammalian PAKs, suggested that the PAN region of MIHCK might be responsible for autoinhibition and Rac binding. We had also shown (25Brzeska H. Kulesza-Lipka D. Korn E.D. J. Biol. Chem. 1992; 267: 23870-23875Abstract Full Text PDF PubMed Google Scholar,30Brzeska H. Martin B.M. Kulesza-Lipka D. Baines I.C. Korn E.D. J. Biol. Chem. 1992; 267: 4949-4956Abstract Full Text PDF PubMed Google Scholar) that neither calmodulin nor phosphatidylserine bind to the C-terminal 90-kDa proteolytic fragment of MIHCK, residues 101–753 (3Brzeska H. Young R. Knaus U. Korn E.D. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 394-399Crossref PubMed Scopus (37) Google Scholar). This, and the fact that basic α-helical regions are typically the site of Ca2+-dependent calmodulin binding in other proteins (32Zhang M. Yuan T. Biochem. Cell Biol. 1998; 76: 313-323Crossref PubMed Scopus (109) Google Scholar), made the basic region, residues 53–71, of MIHCK a good candidate for the calmodulin-binding site. For these reasons, we expressed the following peptides (Fig. 1): N1, residues 1–157, which begins at the N terminus and extends through the basic region and includes the entire putative PAN autoregulatory region; N2, residues 51–157, which is 50 residues shorter than N1 but still contains the basic region and putative autoregulatory region; and N3, residues 91–157, which is limited to the putative autoregulatory region. The middle of Acanthamoeba MIHCK, residues 149–151, is Pro-rich (37%) and basic (pI = 12.1). Although similar regions occur in other proteins, for example, tyrosine kinase ACK1 (33Manser E. Leung T. Salihuddin H. Tan L. Lim L. Nature. 1993; 363: 364-367Crossref PubMed Scopus (260) Google Scholar) and WASP-interacting protein (WIP) (34Ramesh N. Anton I.M. Hartwig J.H. Geha R.S. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 14671-14676Crossref PubMed Scopus (305) Google Scholar), they are absent from other PAKs, and it was of interest to determine whether this region was directly involved in autoregulation of MIHCK. Therefore, we also expressed (Fig. 1): peptides N4, residues 1–198, which includes the entire basic region and the putative PAN autoregulatory region and extends into the Pro-rich region; M1, residues 198–459, which contains just the Pro-rich region not included in N4; and M2, residues 80–459, which includes the putative autoregulatory region and the entire Pro-rich region but not the basic region. Peptides N1 and N2 bound to calmodulin-agarose beads in the presence of Ca2+ but M2 did not (Fig. 2), demonstrating that residues 51–80, the basic region that precedes the PAN region (Fig. 1), are essential for calmodulin binding. The sequence of this region predicts an α-helical structure in agreement with the secondary structure of other known Ca2+-dependent binding sites and different from the structure of Ca2+-independent binding sites (for reviews, see Refs. 32Zhang M. Yuan T. Biochem. Cell Biol. 1998; 76: 313-323Crossref PubMed Scopus (109) Google Scholar and 35Rhoads A.R. Friedberg F. FASEB J. 1997; 11: 331-340Crossref PubMed Scopus (744) Google Scholar). In contrast to mammalian PAKs, full-length Acanthamoeba MIHCK has significant activity without addition of Rac. Therefore, inhibition of MIHCK by peptides can be assayed in the absence of Rac, thus avoiding the possibility that any inhibition might be due simply to sequestering of Rac by the peptides. Peptides N1, N2, N3, and M2 but not M1 inhibited the activity of full-length dephosphorylated MIHCK (Fig. 3 A). N3 was a more effective inhibitor than either N1, N2, or M2, suggesting that residues N- and C-terminal to the autoregulatory region do not contribute to inhibition (Fig. 1). Similarly, N3 was a more effective inhibitor of the expressed MIHCK catalytic domain than either N1 or N2 (M2 was equivalent to N3) (Fig. 3 B), confirming that residues 1–80, which contain the calmodulin/lipid-binding site, do not contribute to inhibition and may slightly weaken inhibition, by the autoregulatory domain, possibly by partially blocking the inhibitory region. Furthermore, the concentrations of N3 required for 50% inhibition of full-length MIHCK and the catalytic domain were similar, 50–100 nm (Fig.A and B), indicating that the inhibitory peptides do not interact with any sites in MIHCK outside of the catalytic domain. The concentration of N3 that inhibited MIHCK was significantly lower than the concentration of a corresponding peptide required for 50% inhibition of PAK1, 1,200 nm (14Zenke F.T. King C.C. Bohl B.P. Bokoch G.M. J. Biol. Chem. 1999; 274: 32565-32573Abstract Full Text Full Text PDF PubMed Scopus (203) Google Scholar). Prior autophosphorylation of full-length MIHCK and the catalytic domain greatly reduced the ability of the peptides to inhibit their activity (Fig. 3 C), suggesting that the peptides inhibit MIHCK kinase activity by inhibiting autophosphorylation. In agreement with this supposition, N1, N2 and M2 substantially inhibited autophosphorylation of full-length kinase (Fig. 4) under the same conditions in which they inhibited kinase activity. Similar results were reported for peptide inhibition of PAK1 (14Zenke F.T. King C.C. Bohl B.P. Bokoch G.M. J. Biol. Chem. 1999; 274: 32565-32573Abstract Full Text Full Text PDF PubMed Scopus (203) Google Scholar). Because the activity of the MIHCK catalytic domain is unaffected by Rac, reversal by Rac of the inhibition of catalytic domain activity by the inhibitory peptides can be used to determine the affinity of Rac for those peptides. Because their inhibitory potency differs, different concentrations of peptides were used in the experiment described in Fig. 5 so that inhibition, in the absence of Rac, was within the range of 30–50%. Rac (1.46 μm) almost completely reversed the inhibition of catalytic domain activity by N4 (0.3 μm) and N1 (0.88 μm) but reversed only slightly inhibition by N3 (0.16 μm) and M2 (0.5 μm) (Fig. 5), indicating that Rac has a higher affinity for N1 and N4 than for N3 and M2. This suggests (Fig. 1) that residues in the region N-terminal to the CRIB motif are required for effective Rac binding but that residues C-terminal to PAN are not required. Also, binding of Rac to N1 and N4 does not require lipids, which were not present in these assays. We have shown that, despite their significant sequence differences, the mechanism of regulation of AcanthamoebaMIHCK and mammalian PAK1 are quite similar and, in this respect as well as by sequence, MIHCK more closely resembles the mammalian PAK-I family than the PAK-II family; more specifically, MIHCK has an autoinhibitory domain. The PAN region of MIHCK (residues 92–157) is sufficient for inhibition of the activity of full-length MIHCK and the catalytic domain, and inhibition is abolished by autophosphorylation just as for PAK1 (10Zhao Z.-S. Manser E. Chen X.-Q. Chong C. Leung T. Lim L. Mol. Cell. Biol. 1998; 1: 2153-2163Crossref Google Scholar, 14Zenke F.T. King C.C. Bohl B.P. Bokoch G.M. J. Biol. Chem. 1999; 274: 32565-32573Abstract Full Text Full Text PDF PubMed Scopus (203) Google Scholar). Lei et al. (9Lei M. Lu W. Meng W. Parrini M.C. Eck M.J. Mayer B.J. Harrison S.C. Cell. 2000; 102: 387-397Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar) (see Hoffman et al.(36Hoffman G.R. Cerione R.A. Cell. 2000; 102: 403-406Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar) for commentary) recently reported the crystal structure of a complex between peptides corresponding to the catalytic domain and the autoregulatory region of PAK1 and identified the residues of the autoregulatory region that interact directly with the catalytic domain: Phe96, Leu106, Leu107, Ile112, Leu128, and Lys141 (Fig. 6). Mutations in Leu107, Glu110, and Asp126 result in constitutively active kinase. Phe96 to Ile112 lie within the PBD, Glu116, Asp126, and Leu128 are C-terminal to the PBD in the IS domain, as defined by Lei et al. (9Lei M. Lu W. Meng W. Parrini M.C. Eck M.J. Mayer B.J. Harrison S.C. Cell. 2000; 102: 387-397Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar), and Lys141 is in the kinase inhibitory domain. Acanthamoeba MIHCK has identical residues in positions corresponding to Phe96, Leu107, and Glu116 of PAK1 and conserved substitutions Met, Leu, Glu, and Met at positions corresponding to positions 106, 112, 116, and 128, respectively, of PAK1 (Fig. 6). However, MIHCK has a Pro at the position that corresponds to Lys141 of PAK1. As Lys141 in PAK1 interacts with the activating loop of the catalytic domain and blocks the active site (9Lei M. Lu W. Meng W. Parrini M.C. Eck M.J. Mayer B.J. Harrison S.C. Cell. 2000; 102: 387-397Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar), the absence of this residue in MIHCK may be related to the “looser” autoinhibition of MIHCK, i.e. Acanthamoeba MIHCK can be fully autophosphorylated and fully activated in the complete absence of both Rac and lipids. It is likely that no other residue in MIHCK fulfills the role of Lys141 in PAK1, since M2 peptide, which contains the 310 amino acids C-terminal to the autoinhibitory region of MIHCK, inhibits MIHCK activity no more than N3 (Figs. 1 and6). Residues important for interaction of the PBD region of mammalian PAKs with small GTP-binding proteins were recently identified by nuclear magnetic resonance (37Morreale A. Venkatesan M. Mott H.R. Owen D. Nietlispach D. Lowe P.N. Laue E.D. Nat. Struct. Biol. 2000; 7: 384-388Crossref PubMed Scopus (159) Google Scholar, 38Gizachew D. Guo W. Chohan K.K. Sutcliffe M.J. Oswald R.E. Biochemistry. 2000; 39: 3963-3971Crossref PubMed Scopus (38) Google Scholar): Ile75, Ser76, Pro78, Phe81, His83, His86, Met99, Trp103, Leu106, and Leu107 (37. Most of these amino acids also occur at the corresponding positions inAcanthamoeba MIHCK (Fig. 6), but the generally highly conserved His83 of PAK1 is replaced by Arg101in MIHCK, and Met99 and Leu106 of PAK1 are replaced by Leu117 and Met124 in MIHCK. Possibly these differences account for the fact that the region that includes basic residues N-terminal to the autoregulatory domain of MIHCK increases the affinity of the autoregulatory region to Rac (Fig. 5); however, Lys residues N-terminal to the CRIB motif have also been shown to enhance Rac binding to and activation of PAK1(39). Both Acanthamoeba (25Brzeska H. Kulesza-Lipka D. Korn E.D. J. Biol. Chem. 1992; 267: 23870-23875Abstract Full Text PDF PubMed Google Scholar) and Dictyostelium (23Lee S.-F. Mahasneh A. DeLaRoche M. Côté G.P. J. Biol. Chem. 1998; 273: 27911-27917Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar) MIHCK bind Ca2+-calmodulin, which inhibits lipid-stimulated activation of both kinases. However, in contrast toAcanthamoeba MIHCK (this paper), Ca2+-calmodulin does not inhibit Rac-stimulated activation of DictyosteliumMIHCK (23Lee S.-F. Mahasneh A. DeLaRoche M. Côté G.P. J. Biol. Chem. 1998; 273: 27911-27917Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar). The probable explanation of this difference is that, unlikeAcanthamoeba MIHCK, Rac activation ofDictyostelium MIHCK does not require lipids, and calmodulin inhibits Acanthamoeba MIHCK by competing with lipid binding. Basic regions immediately N-terminal to the CRIB motif similar to that which is required for calmodulin to bind to AcanthamoebaMIHCK occur not only in Dictyostelium MIHCK but also in other members of the PAK family such as Ste20. Therefore, it would be of interest to test the effect of calmodulin on their activities.
Referência(s)