Intrasteric Regulation of Myosin Light Chain Kinase
1995; Elsevier BV; Volume: 270; Issue: 28 Linguagem: Inglês
10.1074/jbc.270.28.16848
ISSN1083-351X
AutoresJoanna K. Krueger, Roanna C. Padre, James T. Stull,
Tópico(s)Cellular Mechanics and Interactions
ResumoCa2+/calmodulin activates myosin light chain kinase by reversal of an autoinhibited state. The effects of substitution mutations on calmodulin activation properties implicate 4 of the 8 basic residues between the catalytic core and the calmodulin-binding domain in maintaining autoinhibition. These residues are further amino-terminal to the basic residues comprising the previously proposed pseudosubstrate sequence and suggest involvement of the connecting region in intrasteric autoinhibition.The pseudosubstrate model for autoinhibition proposes that basic residues within the autoinhibitory region mimic basic residues in the substrate and bind to defined acidic residues within the catalytic core. Charge reversal mutations of these specific acidic residues, however, had little or no effect on the Km value for regulatory light chain. From a total of 20 acidic residues on the surface of the substrate binding lobe of the catalytic core, 7 are implicated in binding directly or indirectly to the autoinhibitory domain but not to the light chain. Only 2 acidic residues near the catalytic site may bind to the autoinhibitory domain and the arginine at P-3 in the light chain. Exposure of these 2 residues upon calmodulin binding may be necessary and sufficient for light chain phosphorylation. Ca2+/calmodulin activates myosin light chain kinase by reversal of an autoinhibited state. The effects of substitution mutations on calmodulin activation properties implicate 4 of the 8 basic residues between the catalytic core and the calmodulin-binding domain in maintaining autoinhibition. These residues are further amino-terminal to the basic residues comprising the previously proposed pseudosubstrate sequence and suggest involvement of the connecting region in intrasteric autoinhibition. The pseudosubstrate model for autoinhibition proposes that basic residues within the autoinhibitory region mimic basic residues in the substrate and bind to defined acidic residues within the catalytic core. Charge reversal mutations of these specific acidic residues, however, had little or no effect on the Km value for regulatory light chain. From a total of 20 acidic residues on the surface of the substrate binding lobe of the catalytic core, 7 are implicated in binding directly or indirectly to the autoinhibitory domain but not to the light chain. Only 2 acidic residues near the catalytic site may bind to the autoinhibitory domain and the arginine at P-3 in the light chain. Exposure of these 2 residues upon calmodulin binding may be necessary and sufficient for light chain phosphorylation. Activation of smooth/nonmuscle myosin light chain kinase by Ca2+/calmodulin results in phosphorylation of myosin regulatory light chain that plays important roles in initiation of smooth muscle contraction, endothelial cell retraction, secretion, and other cellular processes (32Stull J.T. Krueger J.K. Zhi G. Gao Z.-H. 1995 in International Symposium on Regulation of the Contractile Cycle in Smooth Muscle. April 26, 1995; (Mie, Japan, in press)Google Scholar). The smooth/nonmuscle myosin light chain kinase contains a catalytic core homologous to that of other protein kinases and a carboxyl-terminal regulatory domain consisting of both an inhibitory sequence and a calmodulin-binding sequence (11Hu S.-H. Parker M.W. Lel J.Y. Wilce M.C.J. Benian G.M. Kemp B.E. Nature. 1994; 369: 581-584Crossref PubMed Scopus (175) Google Scholar; 32Stull J.T. Krueger J.K. Zhi G. Gao Z.-H. 1995 in International Symposium on Regulation of the Contractile Cycle in Smooth Muscle. April 26, 1995; (Mie, Japan, in press)Google Scholar). Initially, inspection of the linear sequence within the regulatory domain revealed a similar number and sequential arrangement of 4 basic residues with those shown to be important substrate determinants in a synthetic peptide containing residues 11-23 of the myosin regulatory light chain (see Fig. 1). Thus, 26Pearson R.B. Wettenhall R.E.H. Means A.R. Hartshorne D.J. Kemp B.E. Science. 1988; 241: 970-973Crossref PubMed Scopus (101) Google Scholar proposed that the regulatory domain contained a pseudosubstrate inhibitory sequence whereby 4 specific basic residues in myosin light chain kinase mimic the basic substrate determinants in the light chain peptide substrate. Binding of the pseudosubstrate sequence to the active site inhibited activity. Intrasteric inhibition involves an autoinhibitory sequence that folds back on the catalytic site to inhibit kinase activity as opposed to an allosteric mechanism whereby a conformational change induced at a site distinct from the active site would be responsible for regulation of enzyme activity (17Kemp B.E. Pearson R.B. Biochim. Biophys. Acta. 1991; 1094: 67-76Crossref PubMed Scopus (122) Google Scholar). The sequence comprising the pseudosubstrate region was later expanded to include overlap with the complete amino terminus of the light chain (2Faux M.C. Mitchelhill K.I. Katsis F. Wettenhall R.E.H. Kemp B.E. Mol. Cell. Biochem. 1993; 128: 81-91Crossref Scopus (4) Google Scholar). However, these additional residues(1Blumenthal D.K. Stull J.T. Biochemistry. 1980; 19: 5608-5614Crossref PubMed Scopus (205) Google Scholar, 2Faux M.C. Mitchelhill K.I. Katsis F. Wettenhall R.E.H. Kemp B.E. Mol. Cell. Biochem. 1993; 128: 81-91Crossref Scopus (4) Google Scholar, 3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar, 4Gallagher P.J. Herring B.P. J. Biol. Chem. 1991; 266: 23945-23952Abstract Full Text PDF PubMed Google Scholar, 5Gallagher P.J. Herring B.P. Griffin S.A. Stull J.T. J. Biol. Chem. 1991; 266: 23936-23944Abstract Full Text PDF PubMed Google Scholar, 6Gallagher P.J. Herring B.P. Trafny A. Sowadski J. Stull J.T. J. Biol. Chem. 1993; 268: 26578-26582Abstract Full Text PDF PubMed Google Scholar, 7Gribskov M. Burgess R.R. Nucleic Acids Res. 1986; 14: 6745-6763Crossref PubMed Scopus (341) Google Scholar, 8Herring B.P. J. Biol. Chem. 1991; 266: 11838-11841Abstract Full Text PDF PubMed Google Scholar, 9Herring B.P. Fitzsimons D.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1990; 265: 16588-16591Abstract Full Text PDF PubMed Google Scholar, 10Herring B.P. Gallagher P.J. Stull J.T. J. Biol. Chem. 1992; 267: 25945-25950Abstract Full Text PDF PubMed Google Scholar) are not important for substrate binding and thus are not part of the consensus phosphorylation sequence (16Kemp B.E. Pearson R.B. Trends Biochem. Sci. 1990; 15: 342-346Abstract Full Text PDF PubMed Scopus (801) Google Scholar).Proteolysis studies have supported the hypothesis that myosin light chain kinase contains an autoinhibitory sequence (33Walsh M.P. Dabrowska R. Hinkins S. Hartshorne D.J. Biochemistry. 1982; 21: 1919-1925Crossref PubMed Scopus (56) Google Scholar; 12Ikebe M. Stepinska M. Kemp B.E. Means A.R. Hartshorne D.J. J. Biol. Chem. 1987; 262: 13828-13834Abstract Full Text PDF PubMed Google Scholar; 26Pearson R.B. Wettenhall R.E.H. Means A.R. Hartshorne D.J. Kemp B.E. Science. 1988; 241: 970-973Crossref PubMed Scopus (101) Google Scholar). Limited tryptic cleavage of the chicken smooth muscle myosin light chain kinase results in a nonactivatable form (64 kDa) that becomes constitutively active upon further digestion (61 kDa). Because of differences in the reported cleavage sites, there is disagreement on whether the inhibited form contains the pseudosubstrate sequence (26Pearson R.B. Wettenhall R.E.H. Means A.R. Hartshorne D.J. Kemp B.E. Science. 1988; 241: 970-973Crossref PubMed Scopus (101) Google Scholar; 13Ikebe M. Maruta S. Reardon S. J. Biol. Chem. 1989; 264: 6967-6971Abstract Full Text PDF PubMed Google Scholar).Recent mutational analyses have identified several acidic residues in the catalytic core that may bind the inhibitory sequence but not the light chain. When the charge of these residues was reversed by mutation, the respective KCaM1 1The abbreviations used are: KCaMconcentration of Ca2+/calmodulin required for half-maximal activation of myosin light chain kinase[Ca2+]concentration of Ca2+ required for half-maximal activation of myosin light chain kinasePKIinhibitor peptide for cAMP-dependent protein kinase. 1The abbreviations used are: KCaMconcentration of Ca2+/calmodulin required for half-maximal activation of myosin light chain kinase[Ca2+]concentration of Ca2+ required for half-maximal activation of myosin light chain kinasePKIinhibitor peptide for cAMP-dependent protein kinase. values decreased with no significant effect on the Km or Vmax values for the regulatory light chain (6Gallagher P.J. Herring B.P. Trafny A. Sowadski J. Stull J.T. J. Biol. Chem. 1993; 268: 26578-26582Abstract Full Text PDF PubMed Google Scholar). It was proposed that a lowered KCaM value reflected a weakened binding of the inhibitory region to the catalytic core. In this series of mutations, only 2 acidic residues, located near the catalytic site in the cleft between the two lobes of the kinase, were identified as binding to both the inhibitory region and the arginine residue at the P-3 position in the light chain substrate.A three-dimensional model has been proposed for the catalytic core of smooth muscle myosin light chain kinase, and a portion of the bound autoinhibitory region that includes the expanded pseudosubstrate sequence (22Knighton D.R. Pearson R.B. Sowadski J.M. Means A.R. Ten Eyck L.F. Taylor S.S. Kemp B.E. Science. 1992; 258: 130-135Crossref PubMed Scopus (90) Google Scholar). This model is based on the known crystal structure of the catalytic subunit of cAMP-dependent protein kinase in complex with its inhibitory peptide, PKI. A substrate binding groove has been defined utilizing the position of the PKI peptide in the crystal structure of this complex (20Kemp B.E. Parker M.W. Hu S. Tiganis T. House C. Trends Biochem. Sci. 1994; 19: 440-444Abstract Full Text PDF PubMed Scopus (108) Google Scholar). It is predicted that specific acidic residues lining this substrate-binding groove of the myosin light chain kinase catalytic core form salt bridges with basic residues in both the autoinhibitory sequence and the regulatory light chain substrate. Since the role of these acidic residues was not examined in a previous study (6Gallagher P.J. Herring B.P. Trafny A. Sowadski J. Stull J.T. J. Biol. Chem. 1993; 268: 26578-26582Abstract Full Text PDF PubMed Google Scholar), we have experimentally tested the predictions of this pseudosubstrate model using selected site mutagenesis to identify interacting residues. A modified molecular mechanism for autoinhibition is proposed.EXPERIMENTAL PROCEDURESOligonucleotide-directed MutagenesisA 1,660-base pair 5′-BamHI-XbaI-3′ cDNA fragment, representing the carboxyl-terminal half of rabbit smooth/nonmuscle myosin light chain kinase (5Gallagher P.J. Herring B.P. Griffin S.A. Stull J.T. J. Biol. Chem. 1991; 266: 23936-23944Abstract Full Text PDF PubMed Google Scholar), was subcloned into M13 bacteriophage. Mutagenesis was performed using the oligonucleotide-directed in vitro mutagenesis system (Amersham Corp.) and oligonucleotides designed to produce mutant cDNAs having the desired substitutions. For each mutant cDNA, the desired nucleotide substitutions were verified by DNA sequencing (28Sanger F. Nicklen S. Coulson A.R. Proc. Natl. Acad. Sci. U. S. A. 1977; 74: 5463-5467Crossref PubMed Scopus (52357) Google Scholar).Expression of Wild-type and Mutant Smooth Muscle Myosin Light Chain KinasesAll wild-type and mutant myosin light chain kinases were expressed in COS cells. In each case, two clones of each mutant were subcloned into a pCMV5 expression vector as described previously (5Gallagher P.J. Herring B.P. Griffin S.A. Stull J.T. J. Biol. Chem. 1991; 266: 23936-23944Abstract Full Text PDF PubMed Google Scholar). Subsequent expression was accomplished by transfection into COS cells using DEAE-dextran and chloroquine (9Herring B.P. Fitzsimons D.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1990; 265: 16588-16591Abstract Full Text PDF PubMed Google Scholar).Myosin Light Chain Kinase AssaysCOS cell lysates, prepared as described by 5Gallagher P.J. Herring B.P. Griffin S.A. Stull J.T. J. Biol. Chem. 1991; 266: 23936-23944Abstract Full Text PDF PubMed Google Scholar, were used to determine recombinant wild-type and mutant myosin light chain kinase activity (5Gallagher P.J. Herring B.P. Griffin S.A. Stull J.T. J. Biol. Chem. 1991; 266: 23936-23944Abstract Full Text PDF PubMed Google Scholar). The quantity of wild-type and mutant smooth muscle myosin light chain kinases present in COS cell lysates was determined by immunoblotting with purified smooth muscle myosin light chain kinase as a standard and probing with antiserum raised against the carboxyl-terminal telokin portion of rabbit smooth muscle myosin light chain kinase (Gallagher and Herring, 1991). The Ca2+/calmodulin-dependent activity of wild-type and mutant smooth muscle myosin light chain kinases was measured by 32P incorporation into myosin regulatory light chain purified from chicken gizzards (Blumenthal and Stull, 1980). Ca2+/calmodulin-independent background activity of the extracts was measured as radioactivity incorporated in the presence of 3 mM EGTA. Vmax and Km values were determined from Lineweaver-Burke double-reciprocal plots after performing kinase assays under varying regulatory light chain concentrations.To examine the calmodulin activation properties of mutant myosin light chain kinases in COS cell lysates, Ca2+ activation assays were performed as described by Herring(1991). The relative KCaM values of the mutant kinases within the COS cell extracts were measured at 1 μM calmodulin with varying Ca2+ concentrations from 75 nM to 100 μM with a Ca2+/EGTA buffer system (27Potter J.D. Gergely J. J. Biol. Chem. 1975; 250: 4628-4633Abstract Full Text PDF PubMed Google Scholar). The added calmodulin is in great excess of the endogenous calmodulin contributed by the COS cell lysate and therefore provides controlled conditions to establish the relative concentrations of Ca2+/calmodulin required for kinase activation. The free Ca2+ concentration was the determinant of the actual Ca2+/calmodulin concentration. To assess the quantitative changes in the calmodulin activation properties (KCaM) of mutant myosin light chain kinases, the ratio of activities at Ca2+ concentrations that resulted in less than maximal activity to the maximal activity measured at 100 μM Ca2+ was determined as described previously (25Miller J.R. Silver P.J. Stull J.T. Mol. Pharmacol. 1983; 24: 235-242PubMed Google Scholar; 3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar). The ratio of activities at a specific Ca2+ concentration that is less than that required for maximal activity increases quantitatively as the KCaM value for a mutant myosin light chain decreases relative to the wild-type enzyme. Although the ratio of activities does not allow determination of the absolute value of KCaM, it may be used to calculate the -fold change in KCaM relative to wild-type myosin light chain kinase that has an average KCaM value of 1 nM (25Miller J.R. Silver P.J. Stull J.T. Mol. Pharmacol. 1983; 24: 235-242PubMed Google Scholar; 31Stull J.T. Hsu L.-C. Tansey M.G. Kamm K.E. J. Biol. Chem. 1990; 265: 16683-16690Abstract Full Text PDF PubMed Google Scholar; 3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar). It is proposed that a decrease in the KCaM value (<1 nM) reflects a more easily activated kinase due to a weakening of the binding between the inhibitory region and the catalytic core.RESULTSProperties of Myosin Light Chain Kinase Mutated in the Connection RegionTo further define the boundaries of myosin light chain kinase inhibitory sequence, we created charge reversal and alanine substitution point mutations at basic residues amino-terminal to the proposed pseudosubstrate region of myosin light chain kinase (Fig. 1) (19Kemp B.E. Pearson R.B. Guerriero Jr., V. Bagchi I.C. Means A.R. J. Biol. Chem. 1987; 262: 2542-2548Abstract Full Text PDF PubMed Google Scholar). Full-length cDNA was constructed in the vector pCMV5 for transient expression in COS cells. Immunoblot analysis of COS cell lysates demonstrated that all of the mutant myosin light chain kinases were full-length and expressed at levels (5-20 μg/ml) similar to previously obtained results (3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar; 6Gallagher P.J. Herring B.P. Trafny A. Sowadski J. Stull J.T. J. Biol. Chem. 1993; 268: 26578-26582Abstract Full Text PDF PubMed Google Scholar). All active mutant myosin light chain kinases were dependent on Ca2+ for activity.The Ca2+/calmodulin activation properties of the wild-type and mutant kinases were determined by performing assays at a high calmodulin concentration with a Ca2+/EGTA buffer used to vary the free Ca2+ concentration and hence the Ca2+/calmodulin concentration (25Miller J.R. Silver P.J. Stull J.T. Mol. Pharmacol. 1983; 24: 235-242PubMed Google Scholar). Lys953 is 21 residues amino-terminal to basic residues (Arg974-Lys979) in the calmodulin binding domain that form the core of the proposed pseudosubstrate structure (Fig. 1). Both K953D and K953A mutants required significantly less Ca2+ for half-maximal activation (0.26 and 0.42 μM, respectively) relative to wild-type kinase (0.65 μM) with corresponding decreases in KCaM (0.16 and 0.53 nM from 1.0 nM) (Table I and Fig. 2A). These mutations that produced a 6.3- and 1.9-fold decrease in KCaM values were not accompanied by any significant changes in catalytic properties (Table I). In contrast to these results, mutant K956E had no significant effects on Ca2+ activation or catalytic properties (Table I).Table I:Kinetic properties of myosin light chain kinase containing point mutations of basic residues within the regulatory domain Open table in a new tab Figure 2:Calcium activation curves for mutant myosin light chain kinases. Myosin light chain kinase mutants were expressed in COS cells, and activity was measured in cell lysates at 1 μM calmodulin at different Ca2+ concentrations. The data were normalized to the percent maximal activity. Symbols represent a mean value of at least three independent assays, each performed in duplicate. Data are presented without error bars for clarity. A, mutations within the regulatory domain of myosin light chain kinase are as follows: ●, wild-type; ○, K953D; □, K956E; △, K961E; ▽, K962E; ◇, R967E. B, mutations within the catalytic core of myosin light chain kinase are as follows: ●, wild-type; ○, E858Q; □, D896K; △, E900R; ◇, D911K/D914K. [Ca2+] values are 0.65 ± 0.02, 0.13 ± 0.02, 0.57 ± 0.05, 0.57 ± 0.03, and 0.78 ± 0.09 nM, respectively.View Large Image Figure ViewerDownload Hi-res image Download (PPT)The K961E and K961A mutants required significantly less Ca2+ for half-maximal activation (0.30 and 0.31 μM) with 5- and 3.2-fold decreases in KCaM values, respectively (Table I). The charge reversal mutant K961E selectively changed the activation properties of the kinase with no change in Vmax or Km values (Table I). In contrast, the charge reversal and alanine substitution mutations of Lys962 had little effect on activation and catalytic properties. The Vmax and Km values were not changed for the charge reversal mutation. The Kmvalue for light chain was not measured for the alanine substitution, but the specific activity of the kinase was not different for wild-type enzyme. The charge reversal mutation K962E resulted in a modest decrease in [Ca2+] from 0.65 to 0.47 mM with only a 1.6-fold decrease in KCaM. By comparing these relative KCaM values with that previously reported (3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar) for the double mutant KK961/962EE (0.20 nM), it is evident that of these 2 residues, Lys961 is more sensitive to substitution.Individual substitution of basic residues Lys965 and Arg967 with oppositely charged acidic residues showed a marked decrease in KCaM value (0.03 and 0.09 nM, respectively) relative to the 1 nM value for wild-type kinase (Table I). These values can be compared with the previously reported values (3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar) for the individual alanine mutants K965A and R967A (0.11 and 0.25 nM) as well as the double mutants K965A/R967A and K965E/R967D (0.04 nM and inactive, respectively). None of the mutations caused significant changes in Vmax or Km values except for K965E in which the 33-fold decrease in KCaM was associated with lower Vmax and Km values (Table I) (3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar).Properties of Myosin Light Chain Kinase Mutated in the Catalytic CoreAcidic residues Glu858, Asp896, Glu900, and Asp911 (corresponding residues for myosin light chain kinase from chicken smooth muscle are Glu681, Asp719, Glu723, and Asp734, respectively) within the myosin light chain kinase catalytic core are predicted by the proposed pseudosubstrate model to be involved in binding both the inhibitory sequence in the kinase and the regulatory light chain (22Knighton D.R. Pearson R.B. Sowadski J.M. Means A.R. Ten Eyck L.F. Taylor S.S. Kemp B.E. Science. 1992; 258: 130-135Crossref PubMed Scopus (90) Google Scholar). This hypothesis was tested by charge reversal mutations of these residues. Three neighboring acidic residues, Asp898, Asp913, and Asp914, were chosen as controls for the charge reversal mutation studies. These latter residues are found in the vicinity of those proposed to bind to both the autoinhibitory sequence and the regulatory light chain but are not predicted to bind to either light chain or the autoinhibitory sequence.The resulting mutant kinases were characterized with respect to catalytic and activation properties. One of the charge reversal mutants, E858K, was catalytically inactive (Table II) even though it was expressed at high levels and comigrated with the 152-kDa wild-type recombinant myosin light chain kinase (data not shown). A more conservative mutation, E858Q, resulted in the expression of an active kinase with Vmax, Km, and Vmax/Km ratio values similar to those of the wild-type kinase. E858Q did, however, require significantly less Ca2+ (0.13 μM) for half-maximal activation relative to the wild-type kinase (0.65 μM) with a 7-fold decrease in the KCaM value (, Fig. 2B). These results are consistent with the binding of this residue to or near the inhibitory domain but not to the light chain.Table II:Kinetic properties of myosin light chain kinases containing catalytic core point mutations Open table in a new tab All of the other catalytic core mutants exhibited no significant changes in [Ca2+] and KCaM values from that of the wild-type kinase (, Fig. 2B). These results are consistent with the interpretation that these acidic residues do not bind to basic residues in the autoinhibitory domain.Minor yet statistically significant increases in the Km values for regulatory light chain were observed for D896K, D898K, E900R, D911K/D914K, D911R, D913K, and D914K mutants (19.8, 19.3, 11.6, 33.7, 10.2, 9.6, and 16.0 μM, respectively, versus 4.7 μM for wild-type kinase). In several cases (E900R, D911R, D913K, and D914K) the changes were only about 2-fold and could result from nonspecific electrorepulsion rather than perturbation of specific salt bridge bonds between the respective residues in the catalytic core and the light chain. The Vmax values were similar to wild-type kinase except for D913K, which was lower (8.7 μmol of 32P incorporated per min/mg). The changes in the Vmax/Km ratios for the charge reversal mutants were also modest (Table II). The largest change was noted in the double mutant D911K/D914K (5.8-fold). However, single mutations in these residues resulted in only 1.7- and 2.4-fold changes.DISCUSSIONAs a result of previous mutational and deletion analysis of the myosin light chain kinase regulatory domain, some residues immediately amino-terminal to the calmodulin-binding sequence were implicated in autoinhibition (29Shoemaker M.O. Lau W. Shattuck R.L. Kwiatkowski A.P. Matrisian P.E. Guerra-Santos L. Wilson E. Lukas T.J. Van Eldik L.J. Watterson D.M. J. Cell Biol. 1990; 111: 1107-1125Crossref PubMed Scopus (130) Google Scholar; 14Ito M. Guerriero Jr., V. Chen X. Hartshorne D.J. Biochemistry. 1991; 30: 3498-3503Crossref PubMed Scopus (40) Google Scholar; 3Fitzsimons D.P. Herring B.P. Stull J.T. Gallagher P.J. J. Biol. Chem. 1992; 267: 23903-23909Abstract Full Text PDF PubMed Google Scholar; 34Yano K. Araki Y. Hales S.J. Tanaka M. Ikebe M. Biochemistry. 1993; 32: 12054-12061Crossref PubMed Scopus (34) Google Scholar). Our data indicate that basic residues even further amino-terminal to the originally proposed pseudosubstrate sequence and close to the carboxyl terminus of the catalytic core could be involved in maintaining myosin light chain kinase in an inhibited state (Fig. 1). In each case, both single charge reversal and alanine substitution mutations of residues Lys953, Lys961, Lys965, and Lys967 resulted in kinases that were more easily activated by Ca2+/calmodulin (decreased [Ca2+] and KCaM values relative to wild-type). The results with Lys953 are most interesting since this residue is close to the junction between the catalytic core and the connecting region. These findings support the hypothesis that an intrasteric, autoinhibitory mechanism for myosin light chain kinase involves structures that extend beyond the proposed pseudosubstrate sequence. They also support the idea that the connecting region between the catalytic core and calmodulin binding domain may be bound to the surface of the catalytic core similar to the binding of the carboxyl-terminal tail of twitchin kinase to its catalytic core (20Kemp B.E. Parker M.W. Hu S. Tiganis T. House C. Trends Biochem. Sci. 1994; 19: 440-444Abstract Full Text PDF PubMed Scopus (108) Google Scholar). The multiple intramolecular contacts made by this connecting region may be involved in autoinhibition and, because of the extensive contacts, single mutations would not be expected to produce profound effects on Ca2+/calmodulin activation properties. However, 29Shoemaker M.O. Lau W. Shattuck R.L. Kwiatkowski A.P. Matrisian P.E. Guerra-Santos L. Wilson E. Lukas T.J. Van Eldik L.J. Watterson D.M. J. Cell Biol. 1990; 111: 1107-1125Crossref PubMed Scopus (130) Google Scholar showed that multiple charge reversal mutations in a portion of this autoinhibitory sequence could result in a constitutively active myosin light chain kinase.Many acidic residues in the catalytic core appear to have specific interactions with the autoinhibitory sequence, but not light chain. The conservative neutral mutation, E858Q, created a kinase with wild-type catalytic properties but a 7-fold lower KCaM value. This result is consistent with the suggestion that Glu858 is involved in maintenance of autoinhibition but not in binding to the regulatory light chain. As predicted, D898K, D913K, and D914K each displayed similar [Ca2+] and KCaM values as wild-type myosin light chain kinase, suggesting that these acidic residues do not bind to or near the autoinhibitory sequence. However, evidence was not obtained that other predicted acidic residues bind to the autoinhibitory sequence (Asp896, Glu900, and Asp911). Thus, the acidic residues on the surface of the catalytic core, which have been implicated from this and previous mutational analyses (6Gallagher P.J. Herring B.P. Trafny A. Sowadski J. Stull J.T. J. Biol. Chem. 1993; 268: 26578-26582Abstract Full Text PDF PubMed Google Scholar) in binding to or near the autoinhibitory sequence, are predicted to be mainly in the D α-helix (Fig. 3). The placement of these residues defines a binding pathway distinct from that of PKI binding to the G α-helix of cAMP-dependent protein kinase. By inference then, this pathway is also distinct from the purported substrate-binding groove that predicts pseudosubstrate as well as substrate binding contacts with residues in the G α-helix (Asp896, Glu900, and Asp911) (22Knighton D.R. Pearson R.B. Sowadski J.M. Means A.R. Ten Eyck L.F. Taylor S.S. Kemp B.E. Science. 1992; 258: 130-135Crossref PubMed Scopus (90) Google Scholar; 20Kemp B.E. Parker M.W. Hu S. Tiganis T. House C. Trends Biochem. Sci. 1994; 19: 440-444Abstract Full Text PDF PubMed Scopus (108) Google Scholar). In fact, none of these acidic residues nor any of those in the vicinity of the G α-helix (Asp898, Asp913 and Asp914) can be implicated by our mutational analyses in binding to or near the autoinhibitory domain. Additional proposed contacts include Glu858, which is found in a loop spatially situated between the D α-helix and the G α-helix, and Glu777, which is found in the active site cleft on the D α-helix (22Knighton D.R. Pearson R.B. Sowadski J.M. Means A.R. Ten Eyck L.F. Taylor S.S. Kemp B.E. Science. 1992; 258: 130-135Crossref PubMed Scopus (90) Google Scholar). However, Glu858 also does not appear to bind the light chain. These results suggest that the autoinhibitory sequence may bind on the surface that comprises the substrate binding pocket immediately surrounding the catalytic site, but it is likely to have more contacts with the D α-helix than with the G α-helix.Figure 3:Ribb
Referência(s)