Smooth Muscle Myosin Kinase Requires Residues on the COOH-terminal Side of the Phosphorylation Site PEPTIDE INHIBITORS*

The COOH-terminal residue in peptide analogs of the phosphorylation site sequence in smooth muscle myosin light chains, Ly~"-Lys'~-Arg'~-Ala-Ala-Arg'~-Ala- Thr-Ser'9-(P)Asn20-Va12'-Phez2-Ala'3, were shown to have a strong influence on the kinetics of peptide phos- phorylation. Thepeptides 11-19,ll-20,ll-21, 11-22,and 11-23 were all phosphorylated by the myosin light chain kinase with similar apparent K , values in the range 11-17 p ~ . The V,,, varied 40-fold, with the peptides 11-19,ll-20,ll-21, 11-22, and 11-23 having V,,, values of 0.035, 0.045, 0.32, 1.74, and 1.43 pmol. min" . mg" respectively. These results indicated that Alaz3 was not essential whereas PheZz and Val" had a strong influence on the V,,, of peptide phosphorylation. This series of peptides competitively inhibited myosin light chain phosphorylation with Ki values sim- ilar to their respective K , values. Peptide 11-19 had a Ki value of approximately 10 p~ and a V,,, less than 0.1% of the value with myosin light chains and is therefore an effective inhibitor of the smooth muscle myosin kinase.

The myosin light chain kinase is thought to act as the principal regulatory enzyme in the control of smooth muscle contraction (1). This enzyme is responsible for phosphorylating Ser" in smooth muscle myosin light chains (2, 3). We have recently reported that smooth muscle myosin kinase phosphorylates synthetic peptides corresponding to the region around Serlg, the phosphorylation site in the myosin light chain (4, 5). From structure-function studies it was found that the four basic residues, 11,12,13, and 16 in the sequence

Lys11-Lys12-Arg'3-Ala-Ala-Arg16-Ala-Thr-Ser1g-Asn2o-Valz1-
PheZ2-AlaZ3 had a strong influence on the kinetics of peptide phosphorylation. Moreover, the spatial relationships between Serlg, Arg", and Arg13 were found to be important. Relocation of Arg" to position 15 caused a complete switch in specificity from the natural site of phosphorylation Serlg to Thr". In * This research was supported by grants from the National Health and Medical Research Council of Australia and by equipment grants from the Philip Bushell Foundation, the Sunshine Foundation, the Clive and Vera Ramaciotti Foundation, and the Ian Potter Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. addition to the basic residues, one of the other striking structural features of the known mysoin light chain phosphorylation site sequences (6) is that they all contain an invariant region, Ser(P)-Asn-Val-Phe, on the COOH side of the phosphorylated serine. In this paper we report the contribution of these residues to the kinetics of peptide phosphorylation and the development of peptide inhibitors of the myosin kinase with Ki values in the micromolar range.

EXPERIMENTAL PROCEDURES
All materials were reagent-grade unless otherwise indicated. [y-3P]ATP was obtained from New England Nuclear.
Peptide Synthesis and Purification-The synthetic peptides were synthesized as the COOH-terminal amide form by the Merrified solidphase synthesis procedure (7). Amino acid derivatives protected with the t-butyloxycarbonyl group in the a-amino position and benzyhydrylamine resin were obtained from the Protein Research Foundation (Osaka, Japan). The peptides were assembled manually with ninhydrin testing at each step (8). The completed peptides were simultaneously deprotected and cleaved from the resin in anhydrous HF (9) and purified by ion-exchange chromatography (10). Peptide purity was assessed by quantitative amino acid analysis, high-voltage electrophoresis at pH 1.9 and 6.4, and reverse-phase HPLC (Brownlee RP-300 column, 0.1% (v/v) trifluoroacetic acid with a gradient 0 to 60% (v/v) acetonitrile, 30 min, 1.0 ml/min) (data not shown).
Protein Purification-The purification of myosin light chains ( l l ) , calmodulin (12), and chicken gizzard myosin kinase (13) was by published procedures modified as described previously (4).
Protein Kinase Assay-Myosin kinase was assayed in a volume of 0.08 ml of 40 mM Hepes' buffer, pH 7.0, 5 mM magnesium acetate, 0.50 mM [y-32P]ATP (100-2000 cpm/pmol), 0.55 mM CaC12, 5 pg of calmodulin, 1 mg/ml bovine serum albumin, 0.1% (w/v) Tween 80, and myosin light chains or peptide as indicated. The enzyme was diluted in 25 mM Tris-HC1, 1 mM dithiothreitol buffer, pH 7.5, containing 0.1% (w/v) Tween 80 to prevent loss of enzyme through binding to plastic and to ensure an accurate measure of the enzymespecific activity (14). Incubations were carried out a t 30 "C and aliquots (0.03 ml) were taken at 3 and 6 min. The aliquots were applied to phosphocellulose ion-exchange paper and were washed and counted as described (5). In experiments to determine the apparent K,,, and Vmax of the peptide analogs, the parent peptide MLC 11-23 was used as an internal control for the Vm=. Inhibition of myosin light chain phosphorylation by synthetic peptide was assessed by trichloroacetic acid precipitation, 10% (w/v) on filter papers, as described by Reimann et al. (15). Under these conditions the synthetic peptide is not retained on the filter paper and does not contribute to the measured [32P] phosphate transferred to the myosin light chains.

RESULTS AND DISCUSSION
Since all of the known regulatory light chain sequences (6) have an invariant sequence, Ser(P)-Asn-Val-Phe, on the COOH-terminal side of the phosphorylated serine, it was of interest to assess whether this region influenced the kinetics of peptide phosphorylation by the myosin kinase. In the study, all peptides were prepared as their carboxyl-terminal amides so that the results would not be influenced by the presence of a free carboxyl. The parent peptide   were withdrawn at intervals to verify that the rates of phosphorylation were linear with respect to time for all peptide concentrations tested. similar to the myosin light chains, the Vmax values are approximately 25-fold lower ( Table I).
Deletion of the COOH-terminal Alaz3 had only a minor effect on the V,,, and no effect on the apparent K , ( Table  I). In contrast deletion of the hydrophobic residues Phe2' and Val2' had a marked effect on the V,,, reducing it to 23 and 3%, respectively. Deletion of Amz0 caused only a minor reduction in the v,,, over that obtained with the 11-20 peptide.
The V,,, of the peptide 11-19, 0.035 pmol.min" .mg' (Fig.  l ) , was less than 0.1% of the V,,, obtained with the myosin light chains (5). In contrast to the strong effects of the removal of the hydrophobic residues on the V,,,, the apparent K,,, was not significantly effected. Replacement of the Asn-Val-Phe sequence by Ala-Ala-Ala resulted in a peptide with a low K , and very poor V,,, comparable to the peptide 11-19 (Table  I). These results indicate that the peptide chain length is not a dominant factor but rather that the hydrophobic residues Val-Phe have a strong influence on the Vmax.
Previously we have reported that the spatial location of the basic residues, particularly Argl', had a powerful influence on whether Ser" or Thr" was phosphorylated ( 5 ) . The peptides containing deletions of the carboxyl-terminal residues were phosphorylated predominantly on Serlg as follows: MLC 11-(96%), and MLC 11-22, Azo, AZ1, A'' (88%). The site of phosphorylation was assessed by partial acid hydrolysis (5.7 M HC1, 110 "C, 2 h), high voltage paper electrophoresis at pH 1.9, autoradiography, and liquid scintillation counting of the 0-phosphothreonine and 0-phosphoserine spots. The results were not corrected for the differences in the acid stability of the phosphoaminoacids (16) and therefore tend to underesti- The results obtained in this study are in marked contrast to those obtained previously with the CAMP-dependent protein kinase (17,18). Truncation of the carboxyl-terminal residues in the pyruvate kinase peptide Leu-Arg-Arg-Ala-Ser-Leu-Gly was accompanied by a dramatic increase in the apparent K,,, and relatively modest changes in the V,,,. These earlier studies on the CAMP-dependent protein kinase were done with peptides having a free COOH terminus. However, this is not an important factor since the COOH-terminal amide analogs Leu-Arg-Arg-Ala-Ser-Leu-NH2 and Leu-Arg-Arg-Ala-Ser-NHz also show the same trend with increasing apparent K,,,.' The very poor VmaX and relatively low apparent K,,, of the myosin light chain peptide 11-19 raised the possibility that this peptide may be a useful inhibitor of the myosin kinase. The capacity of this peptide to inhibit myosin light chain phosphorylation was studied (Fig. 2). The Ki obtained with the peptide 11-19 was 10 p~. This value was comparable to the apparent K,,, (17 p~) obtained with peptide as substrate.
Previously we had found that the peptide Ser-Ser-Lys-Thr-

Thr-Lys-Arg-Pro-Gln-Arg-Ala-Thr-Ser-Asn-Val-Phe-Ser
competitively inhibited myosin light chain phosphorylation with a Ki of 590 p~ compared with an apparent K,,, for phosphorylation of 90 PM (18). This peptide differs from the ones described in the present study in not having the tribasic recognition site Lys-Lys-Arg. Several studies with peptide substrates of the CAMP-dependent protein kinase (19., 20) have found that the Ki values are higher than the corresponding apparent K,,, values. Direct binding studies with the CAMP-dependent protein kinase and synthetic peptides have shown that the Kd for peptide binding is also higher than the apparent K,,, for phosphorylation, by almost two orders of magnitude. The results obtained with myosin light chain peptide 11-19 were therefore unexpected. For this reason we examined the capacity of each of the truncated peptides to inhibit myosin light chain phosphorylation. In all cases the K, values obtained were comparable to the apparent K,,, values for peptide phosphorylation (Tables I and 11). Inspection of the double reciprocal plots for peptide inhibition (see Fig. 2) indicated that the lines did not intersect precisely on the ordinate. This did not reflect experimental variation, as the double reciprocal plots for the other four peptides tested all had intersection points slightly to the right of the ordinate.
Thus nase cause an unexpected %fold increase in the Kd for ADP.
In the event of ADP release being rate-limiting for the myosin light chain kinase, a peptide inhibitor that increased the Kd for ADP would have the effect of increasing the apparent V,...
The results of this study demonstrate that it is possible to construct relatively potent synthetic peptide inhibitors of the smooth muscle myosin light chain kinase simply by removing the residues on the COOH-terminal side of the phosphorylated serine. Furthermore, the Asn-Val-Phe peptide that is homologous on all myosin light chains has an important influence on the kinetics of peptide phosphorylation by the myosin kinase.