Inhibitory Phosphorylation Site for Rho-associated Kinase on Smooth Muscle Myosin Phosphatase*

It is clear from several studies that myosin phosphatase (MP) can be inhibited via a pathway that involves RhoA. However, the mechanism of inhibition is not established. These studies were carried out to test the hypothesis that Rho-kinase (Rho-associated kinase) via phosphorylation of the myosin phosphatase target subunit 1 (MYPT1) inhibited MP activity and to identify relevant sites of phosphorylation. Phosphorylation by Rho-kinase inhibited MP activity and this reflected a decrease in V max. Activity of MP with different substrates also was inhibited by phosphorylation. Two major sites of phosphorylation on MYPT1 were Thr695 and Thr850. Various point mutations were designed for these phosphorylation sites. Following thiophosphorylation by Rho-kinase and assays of phosphatase activity it was determined that Thr695 was responsible for inhibition. A site- and phosphorylation-specific antibody was developed for the sequence flanking Thr695 and this recognized only phosphorylated Thr695 in both native and recombinant MYPT1. Using this antibody it was shown that stimulation of serum-starved Swiss 3T3 cells by lysophosphatidic acid, thought to activate RhoA pathways, induced an increase in Thr695 phosphorylation on MYPT1 and this effect was blocked by a Rho-kinase inhibitor, Y-27632. In summary, these results offer strong support for a physiological role of Rho-kinase in regulation of MP activity.


It is clear from several studies that myosin phosphatase (MP) can be inhibited via a pathway that involves
RhoA. However, the mechanism of inhibition is not established. These studies were carried out to test the hypothesis that Rho-kinase (Rho-associated kinase) via phosphorylation of the myosin phosphatase target subunit 1 (MYPT1) inhibited MP activity and to identify relevant sites of phosphorylation. Phosphorylation by Rho-kinase inhibited MP activity and this reflected a decrease in V max . Activity of MP with different substrates also was inhibited by phosphorylation. Two major sites of phosphorylation on MYPT1 were Thr 695 and Thr 850 . Various point mutations were designed for these phosphorylation sites. Following thiophosphorylation by Rho-kinase and assays of phosphatase activity it was determined that Thr 695 was responsible for inhibition. A site-and phosphorylation-specific antibody was developed for the sequence flanking Thr 695 and this recognized only phosphorylated Thr 695 in both native and recombinant MYPT1. Using this antibody it was shown that stimulation of serum-starved Swiss 3T3 cells by lysophosphatidic acid, thought to activate RhoA pathways, induced an increase in Thr 695 phosphorylation on MYPT1 and this effect was blocked by a Rho-kinase inhibitor, Y-27632. In summary, these results offer strong support for a physiological role of Rho-kinase in regulation of MP activity.
It is accepted that phosphorylation of the 20-kDa myosin light chain (MLC20) 1 is an essential mechanism in regulating contractile activity in smooth muscle and non-muscle cells (1)(2)(3). Following stimulation by several agonists an increased MLC20 phosphorylation and resultant contraction frequently occurs. Since this is effective at submaximal Ca 2ϩ levels it is referred to as increased Ca 2ϩ sensitivity (3). The extent of MLC20 phosphorylation depends on the relative activities of myosin light chain kinase (4) and myosin phosphatase (MP) (5). The mechanism for Ca 2ϩ sensitization is thought to occur via inhibition of MP via a G-protein-linked pathway (3,6,7).
There is strong evidence that an additional component of the Ca 2ϩ sensitization pathway is RhoA. GTP␥S induces Ca 2ϩ sensitization in permeabilized smooth muscle fibers and this effect was blocked by ADP-ribosylation of Rho using an exoenzyme from Clostridium botulinum, C3, or Staphylococcus aureus, named epidermal differentiation inhibitor (8,9). Addition of the GTP␥S-bound active form of RhoA (8) or a constitutively active mutant of RhoA (9) induced a Ca 2ϩ sensitization in various permeabilized smooth muscle preparations. A target for RhoA is Rho-kinase (10 -12) and introduction of a constitutively active recombinant fragment of Rho-kinase into Triton X-100-permeabilized smooth muscle caused a Ca 2ϩ -independent contraction via phosphorylation of MLC20 (13). A specific inhibitor of Rho-kinase, Y-27632, also was reported to inhibit the agonist-induced Ca 2ϩ sensitization of smooth muscle contraction (14). In non-muscle cells activation of RhoA promotes the formation of stress fibers and focal adhesion complexes (15). C3 blocked the formation of both structures in Swiss 3T3 cells treated with the Rho-activating stimulus, lysophosphatidic acid (LPA) (16). In addition, the expression of a constitutively active form of Rho-kinase induced stress fiber and focal adhesion formation in fibroblasts and an increase in the level of MLC20 phosphorylation (17). These results in smooth muscle and in non-muscle cells are attributed to an increase in MLC20 phosphorylation and this is thought to reflect the inhibition of MP.
MP is composed of three subunits: a catalytic subunit of type 1 phosphatase ␦ isoform, PP1c␦; and two non-catalytic subunits, 110 and 20 kDa. The 110-kDa subunit is a targeting molecule and thus has been termed myosin phosphatase target subunit 1 (MYPT1). MYPT1 is the key molecule involved in regulation of MP (for reviews, see Ref. 5).
Purified Rho-kinase thiophosphorylated MYPT1 and induced an inhibition of MP activity (18,19). In permeabilized smooth muscle, thiophosphorylation of MYPT1 was associated with a decreased activity of myosin phosphatase (20). Thus, a potential mechanism for inhibition of MP is via the phosphorylation of MYPT1 by Rho-kinase. Phosphorylation of MYPT1 and inhibition of MP was first observed with a gizzard isoform (M133) in which the major phosphorylation site for an endogenous kinase was Thr 695 (21). Using purified Rho-kinase the inhibitory site(s) on MYPT1 was found to be in the C-terminal part of MYPT1, i.e. Ser 667 -Ile 1004 (19). Also, it is was suggested that the phosphorylation site(s) for Rho-kinase was in the sequence 753 to 1004 of gizzard MYPT1 (18).
In order to understand the regulation of MP it is essential to identify various functional regions of MYPT1. These include binding sites for PP1c␦ and for the substrate, phosphorylated myosin, and also the inhibitory phosphorylation site(s). In the present study, we examined the effect of thiophosphorylation of MYPT1 on the enzymatic properties of MP and identified two sites, Thr 695 and Thr 850 , on chicken MYPT1 as the major phosphorylation sites for Rho-kinase using point mutational analysis. Thr 695 , but not Thr 850 , was identified as the inhibitory phosphorylation site. Using a site-and phosphorylation statespecific antibody we found that Thr 695 of MYPT1 was phosphorylated in Swiss 3T3 cells stimulated by LPA, which is known to activate Rho/Rho-kinase pathways (16).
Phosphorylation of MP and MYPT1 Mutants-Phosphorylation of MP and MYPT1 mutants by Rho-kinase were carried out in the presence of 2 M GTP␥S⅐GST-RhoA as described previously (19). To measure the phosphorylation levels of MYPT1 or mutants, samples were subjected to SDS-PAGE (7.5-20%), and 32 P incorporation in the excised gel slices corresponding to the position of MYPT1 or mutants was determined. To determine the effects of phosphorylation on phosphatase activity, phosphorylation of MP and MYPT1 mutants were carried out under similar conditions without microcystin-LR using ATP␥S instead of ATP. The reactions were diluted (ϫ10) with a buffer containing 30 mM Tris-HCl, pH 7.5, 0.2 mg/ml bovine serum albumin, and 1 mM EDTA and assayed for phosphatase assay. Phosphorylation of MP, rG-MYPT1 1-1004 , and rG-MYPT1 T695A by PKA was performed as described (32). Phosphorylation of rG-MYPT1 1-1004 by protein kinase C was carried out at 30°C for 60 min with 10 g/ml protein kinase C in 20 mM Tris-HCl, pH 7.5, 100 mM KCl, 0.1 mM dithiothreitol, 1 mM MgCl 2 , 1 mM EGTA, 100 g/ml phosphatidyl-L-serine, 100 ng/ml phorbol 12myristate 13-acetate, 1 M microcystin-LR, 250 M [␥-32 P]ATP, and 30 g/ml rG-MYPT1 1-1004 . Phosphorylation of MP by endogenous kinase(s) was carried out at 30°C for 60 min in 20 mM Tris-HCl, pH 7.5, 100 mM KCl, 0.1 mM dithiothreitol, 1 mM MgCl 2 , 1 M microcystin-LR, 250 M [␥-32 P]ATP, and 30 g/ml MP holoenzyme containing the endogenous kinase (21).
Phosphatase Assays-Phosphatase assays were carried out at 30°C using 32 P-labeled MLC20, 32 P-labeled calponin, or 32 P-labeled glycogen phosphorylase as substrates. The concentration of substrates used is indicated in each experiment. For the reconstituted phosphatase assay, various amounts of each mutant were mixed with 1 nM PP1c␦ and preincubated for 5 min at 30°C, then the phosphatase activity was assayed using 32 P-labeled MLC20 (final concentration 5 M) as substrate. Assay conditions were: 30 mM Tris-HCl, pH 7.5, 100 mM KCl, 1 mM EDTA, 0.2 mg/ml bovine serum albumin in a final volume of 50 l as described previously (22).
Cell Culture and Measurement of MYPT1 Phosphorylation-Swiss 3T3 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin in an atmosphere of 5% CO 2 and 95% humidified air at 37°C. Subconfluent cells were incubated in serum-free medium for 24 h, then 1 M LPA or vehicle (phosphate-buffered saline) was added to the medium. In some experiments, serum-starved cells were pretreated for 30 min with 10 M Y-27632 before stimulation with LPA. At each time indicated, reactions were terminated by addition of icecold trichloroacetic acid to a final 10% (w/v). Trichloroacetic acid was removed by centrifugation at 15,000 ϫ g for 10 min and the precipitate was washed 3 times with cold acetone and dried. The dried cell powder was lysed in SDS sample buffer and analyzed by Western blotting using either pM133 T695 or anti-MYPT1 antibody. The blots were exposed to films (Fuji RX) for various times (from 5 to 60 s) to obtain a linear response by the enhanced chemiluminescence method (Amersham Pharmacia Biotech). For pM133 T695 the scans of Western blots were linear in the range of 10 to 500 ng of phosphorylated rG-MYPT1 1-1004 . The amount of MYPT1 in the pertinent immunoreactive bands was estimated by densitometry (Densitograph, Atto, Japan).
Other Procedures-SDS-PAGE was carried out with the discontinuous buffer system (34). Western blotting (35) and two-dimensional phosphoamino acid analysis (36) were performed as described previously. Protein concentrations were determined by the BCA procedure (Pierce) using bovine serum albumin as a standard. Gel filtration was carried out with a Superdex 200 PC 3.2/1.6 column attached to a SMART system (Amersham Pharmacia Biotech) as described previously (22).

Effects of Thiophosphorylation of Myosin Phosphatase on
Enzymatic Properties-To confirm and extend the earlier observation that phosphorylation of MYPT1 by Rho-kinase inhibited MP activity (18,19), the influence of phosphorylation on kinetic parameters and substrate specificity were examined. ATP␥S rather than ATP was used because of the resistance of the thiophosphate group to phosphatase (there is some phosphatase activity toward phosphorylated MYPT1 in the holoenzyme (21)). The effects of thiophosphorylation of MYPT1 by Rho-kinase are shown in Table I. The dominant effect is on V max which is decreased by about 70% following thiophosphorylation. There is a slight, but probably insignificant, effect on K m . The effects of thiophosphorylation also was investigated using various substrates. Two additional substrates for PP1 were chosen and compared with MLC20, namely glycogen phosphorylase (31) and calponin (30). Thiophosphorylation of MP by Rho-kinase resulted in inhibition of phosphatase activity for each substrate tested (Table II). The extents of inhibition were approximately the same (about 75%).
Determination of Phosphorylation Sites on MYPT1 for Rhokinase-From previous studies (19) it was established that the phosphorylation sites for Rho-kinase were located in the Cterminal part (Ser 667 to Ile 1004 ) of MYPT1. To identify the phosphorylated residues, the MYPT1 in the kinase-free MP holoenzyme was phosphorylated to about 2 mol of P/mol of MYPT1 by Rho-kinase, as described previously (19). Phosphoamino acid analysis of the phosphorylated MYPT1 indicated that Thr was the major site of phosphorylation (Fig. 1A). Phosphoserine was detected as a minor component. Two recombinant proteins, full-length MYPT1 (rG-MYPT1 1-1004 ), and a C-terminal fragment (rG-MYPT1 667-1004 ), were phosphorylated by Rho-kinase and showed the same phosphoamino acid pattern (data not shown).
A synthetic peptide representing the sequence around the inhibitory phosphorylation site on MYPT1 by endogenous kinase, Lys 689 to Asp 702 also was checked as a substrate for Rho-kinase. This peptide (at 0.1 mg/ml, ϳ60 M) was phosphorylated to 0.7 mol of P/mol of peptide and specific activities for Rho-kinase were 210 Ϯ 7 and 413 Ϯ 16 nmol/min/mg (mean Ϯ S.E., n ϭ 3) in the absence and presence of GTP␥S⅐GST-RhoA (2 M), respectively. These values are about half the specific activity for Rho-kinase with the C-terminal fragment of MYPT1 (19).
The site on gizzard MYPT1 (M133 isoform) for the endogenous kinase is Thr 695 and the N-terminal flanking sequence is Arg 692 -Arg-Ser. This sequence is found also at Arg 847 -Arg-Ser-Thr 850 . Thus two potential sites were Thr 695 and Thr 850 and mutants were constructed to test this hypothesis. Wild-type full-length MYPT1 (rG-MYPT1 1-1004 ) was phosphorylated by Rho-kinase to 1.78 Ϯ 0.14 mol of P/mol of MYPT1 (Fig. 1B). Substitution of Thr 695 to Ala (rG-MYPT1 T695A ) resulted in decreased phosphorylation to 0.72 Ϯ 0.10 mol of P/mol of mutant (Fig. 1B). The mutant targeted for Thr 850 , rG-MYPT1 T850A , was phosphorylated by Rho-kinase to a stoichiometry of 0.91 Ϯ 0.08 mol of P/mol of mutant (Fig. 1B). Also a double mutant was tested in which both Thr 695 and Thr 850 were changed to Ala (rG-MYPT1 T695A/T850A ) and the level of phosphorylation was reduced to 0.35 Ϯ 0.08 mol of P/mol of mutant (Fig. 1B). In contrast, a random control mutant (rG-MYPT1 T716A ) yielded phosphorylation levels similar to wild-type MYPT1. These results suggest that two major sites on MYPT1 for phosphorylation by Rho-kinase are Thr 695 and Thr 850 . Minor sites might also exist (accounting for about 20% of the total phosphoryla-tion) and these presumably contribute to the phosphoserine detected in the phosphoamino acid analysis.
Identification of Thr 695 on Gizzard MYPT1 as the Inhibitory Site-Previously it was shown that mixtures of PP1c␦ with MYPT1 or its N-terminal fragments increased phosphatase activity (22). Using this assay, the effects of phosphorylation by Rho-kinase of wild-type MYPT1 and various mutants were examined. Each recombinant protein was thiophosphorylated by Rho-kinase, mixed with native (chicken) PP1c␦ at the indicated molar ratios, and phosphatase activity was determined using 32 P-MLC20. The non-thiophosphorylated forms of MYPT1 each activated PP1c␦ about 7-fold at a 3-fold molar excess of MYPT1, or mutant, to PP1c␦ (Fig. 2). Thiophosphorylation of wild-type MYPT1 (rG-MYPT1 1-1004 ) and the Thr 850 to Ala point mutant (rG-MYPT1 T850A ) decreased dramatically the extent of PP1c␦ activation from 7.4 Ϯ 1.1-fold to 2.3 Ϯ 0.4-fold ( Fig. 2A) and 7.2 Ϯ 0.5-fold to 2.2 Ϯ 0.6-fold (Fig. 2C), respectively. However, thiophosphorylation of the Thr 695 to Ala point mutant (rG-MYPT1 T695A ) did not induce such a marked inhibition (7.4 Ϯ 0.5-fold for the non-phosphorylated and 5.9 Ϯ 0.3-fold for the thiophosphorylated form; Fig. 2B). Addition of the third MP subunit, M20 in either its nonphosphorylated or thiophosphorylated states, did not influence phosphatase activity for any of the combinations used (data not shown).
One mechanism to account for the inhibition of PP1c␦ activity would be a reduced binding of PP1c␦ to MYPT1. Gel filtration was used to check this possibility (22). A mixture of PP1c␦ and thiophosphorylated rG-MYPT1 1-1004 (at a 0.1 molar ratio for PP1c␦) was applied to a Superdex 200 column and phosphatase activity of the eluate was monitored. All phosphatase activity eluted at an apparent mass of 190 kDa, i.e. a peak corresponding to isolated PP1c␦ (at 37 kDa) was not detected (data not shown). These data indicate that PP1c␦ can bind to thiophosphorylated rG-MYPT1 1-1004 .
Finally, the effects of thiophosphorylation by Rho-kinase of the complex of MYPT1 and PP1c␦ was tested (previously, MYPT1 or its mutants was thiophosphorylated prior to the addition of PP1c␦). Thiophosphorylation of the complexes containing rG-MYPT1 1-1004 and rG-MYPT1 T850A decreased phosphatase activity to about 20% (Fig. 2D). With rG-MYPT1 T695A plus PP1c␦, thiophosphorylation induced only a small decrease in activity to about 84% (Fig. 2D). Again, these results support the contention that the inhibition of phosphatase activity is due primarily to the phosphorylation of Thr 695 by Rho-kinase.  1. Determination of phosphorylation sites on MYPT1 by Rho-kinase. A, phosphoamino acid analysis of MYPT1. MP was phosphorylated by Rho-kinase to 1.7 mol of P/mol of MYPT1. 32 P-MYPT1 was eluted from excised gel slices and subjected to two-dimensional phosphoamino acid analysis. The position of free phosphate (P i ), phosphoserine (P-Ser), phosphothreonine (P-Thr), and partially hydrolyzed residues (arrowhead) are indicated on the right. B, stoichiometry of phosphorylation of MYPT1 mutants by Rho-kinase. rG-MYPT1 1-1004 (E), rG-MYPT1 T716A (q), rG-MYPT1 T695A (OE), rG-MYPT1 T850A (‚), and rG-MYPT1 T695A/T850A (f) were phosphorylated by Rho-kinase and 32 P incorporation was determined as described under "Experimental Procedures." Error bars indicate mean Ϯ S.E. (n ϭ 3).

Production and Characterization of a Site-and Phosphorylation-dependent Antibody for Thr 695 -
The potential importance of Thr 695 phosphorylation has been documented above. If phosphorylation of this residue could be monitored (for example, in cell extracts) it would provide a valuable tool for studying Rho-kinase activity relative to MYPT1. To this end a rabbit polyclonal antibody was prepared that recognized phosphorylation of MYPT1 at Thr 695 . This antibody is referred to as pM133 T695 .
The specificity of the antibody was evaluated by Western blots. As shown in Fig. 3 8 and 11). These results suggest that pM133 T695 recognizes phosphorylation at Thr 695 of MYPT1. The kinases responsible for this phosphorylation are Rho-kinase and the endogenous kinase in the MP holoenzyme.
LPA-induced in Vivo Phosphorylation of MYPT1 at Thr 695 -To determine if the antibody could be used to detect MYPT1 phosphorylation under in vivo conditions the effect of LPA on serum-starved Swiss 3T3 cells was monitored. It was reported that in this system LPA stimulates stress fiber and focal adhesion formation and involves activation of RhoA (16).
The serum-starved cells were incubated with 1 M LPA for different times, as indicated, and cell lysates were analyzed by Western blots using pM133 T695 (see "Experimental Procedures"). Even in the absence of LPA some phosphorylation of Thr 695 was detected (Fig. 4A). Following LPA stimulation, the extent of Thr 695 phosphorylation increased and reached a maximum at 10 min (about 4-fold increase over basal levels). A relatively high level of phosphorylation at Thr 695 was retained for the 30-min duration of the experiment (Fig. 4A). As shown in Fig. 4B, the LPA-induced phosphorylation of Thr 695 was almost completely suppressed by pretreatment with 10 M Y-27632, an inhibitor for Rho-kinase. It is interesting, however, that this kinase inhibitor did not eliminate the basal levels of Thr 695 phosphorylation. DISCUSSION There is considerable evidence that Ca 2ϩ sensitization in smooth muscle reflects inhibition of myosin phosphatase (for reviews, see Refs. 3 and 5). This occurs at low, but finite, levels of Ca 2ϩ where the activity of myosin light chain kinase does not override the phosphatase activity. However, the mechanism of MP inhibition is still not established. Initially, it was thought that the Ca 2ϩ -sensitizing effect of arachidonic acid was due to dissociation of the MP complex (37) or, as subsequently suggested, the activation of an atypical protein kinase C isoform or an unknown kinase (38). Other theories include: phosphorylation of MYPT1 by an endogenous kinase (21) or, by Rho-kinase (18,19); and inhibition by a phosphorylation-dependent protein, CPI17 (39,40). However, there is compelling evidence that RhoA is involved in Ca 2ϩ sensitization (8,9) and any proposed mechanism should include consideration of this point. A logical link to RhoA is Rho-kinase (10 -12) and thus an attractive hypothesis is that phosphorylation via Rho-kinase is an important component of the Ca 2ϩ -sensitization mechanism (18). This does not exclude other kinases/phosphatases but suggests that if such are involved they may be linked to the RhoA cascade. The data presented in this study are consistent with direct phosphorylation of MYPT1 by Rho-kinase as an inhibitory mechanism.
MYPT1 is known to interact with several components and some of the regions of interaction have been identified. The N-terminal one-third of MYPT1 can bind to both PP1c␦ and phosphorylated myosin and activate phosphatase activity (22,29,41). The ankyrin repeats region (residues 39 -295) is important in this regard, but also the N-terminal 38 residues are involved in the PP1c interaction. Residues 35 to 38, Lys-Val-Lys-Phe, and flanking charged residues form the PP1c consensus-binding motif and these are critical for the binding of PP1c (42,43). In addition, residues 1 to 16 are thought to be involved in activation of PP1c (22). The C-terminal half of MYPT1 contains binding sites for M20 (44,45), RhoA (18), acidic phospholipids (32), and myosin (45). Inhibition of MP activity by phosphorylation of Thr 695 is due mainly to a decrease in V max . If PP1c is bound to the N-terminal end of MYPT1 it is difficult to rationalize the inhibition that results from phosphorylation in the C-terminal half of the molecule. Obviously long-range effects are possible and folding of the MYPT1 molecule may facilitate interaction with PP1c. A tentative suggestion is that phosphorylation at Thr 695 negates activation of PP1c by residues 1 to 16, however, this must be tested experimentally. Another point that was of interest was whether phosphorylation and inhibition of MYPT1 altered substrate specificity. Interaction of the N-terminal fragment of MYPT1 with PP1c enhances phosphatase activity toward MLC20 (see above) but suppresses activity with glycogen phosphorylase (46). From our results and those with moesin (47) and adducin (48) it appears that phosphorylation of MYPT1 causes a general inhibition of phosphatase activity. The influence of the phosphorylated Thr, therefore, is suggested to be focused toward PP1c rather than at the substrate-binding regions. The phosphatase activity toward glycogen phosphorylase also was inhibited by phosphorylation. This is additional evidence that phosphorylation at Thr 695 does not induce dissociation of PP1c and MYPT1. If dissociation had occurred the phosphatase activity with glycogen phosphorylase would increase.
In previous studies it was reported that phosphorylation of Thr 695 by an endogenous kinase caused inhibition of MP activity. This kinase was not identified but is not Rho-kinase. 2 The above data also suggest that Thr 695 is a major functional site of phosphorylation by Rho-kinase. The sequence around Thr 695 (numbering based on the chicken M133 isoform of MYPT1 (44)) is among the most conserved regions in the MYPT1 family. Gly 653 to Thr 747 in M133 is 100% identical to human (49) and porcine (50) MYPT1 and is 97% identical to rat MYPT1 (M 110 ) (51). A corresponding sequence in human MYPT2 (a separate family of MYPT and a product of a different gene (52)), i.e. 632 to 660, is 93% identical to M133 and contains two conservative substitutions, Ser to Thr and Lys to Arg, at positions 642 and 659, respectively. The putative phosphorylation site in MYPT2 is Thr 646 . The M133 sequence of Arg 692 to Thr 699 also is conserved in the MYPT homologue in Caenorhabditis elegans, i.e. MEL-11 (53). The presence of this conserved sequence argues in favor of a functional role, presumably via phosphorylation, but this has not been demonstrated in MYPT2 or in the C. elegans MEL-11. The flanking sequences of Thr 850 are conserved in the MYPT1 and MYPT2 (around Thr 808 ) but not in MEL-11. The two sites identified on MYPT1 for Rho-kinase are Thr 695 and Thr 850 and the sequences N-terminal to the phosphorylated Thr are identical, i.e. Arg-Arg-Ser-Thr. This sequence is a putative target for PKA, although Ser is preferred to Thr (54). This probably accounts for the low extent of phosphorylation by PKA of rG-MYPT1 1-1004 at Thr 695 , although the MP holoenzyme was not phosphorylated. In addition, it was shown previously that phosphorylation of MP by PKA did not affect the phosphatase activity (32). Thus, it is proposed that the phosphorylation of Thr 695 in vivo is not achieved by PKA. Protein kinase C may also be eliminated. Based on the limited evidence provided here it is suggested that the consensus sequence for Rho-kinase, and the endogenous kinase, is similar to that for PKA and requires basic residues at the Ϫ3 and Ϫ2 positions.
Several earlier studies carried out under "physiological" conditions (␣-toxin permeabilized smooth muscle (20), platelets (55,56); and NIH 3T3 cells (18)) had suggested a correlation between phosphorylation of MYPT1 and inhibition of MP activity, but the site(s) of phosphorylation were not identified. The only study to identify the inhibitory phosphorylation site, i.e. Thr 695 , was done with isolated proteins (21). The objective of the LPA-Swiss 3T3 cell experiment was to determine if Thr 695 was phosphorylated under in vivo conditions with a regime that activated RhoA. This was achieved. LPA induced about a 4-fold increase in Thr 695 phosphorylation and this stimulation was blocked by a Rho-kinase inhibitor, Y-27632 (14). These results add confidence to accepting the idea that Rho-kinase under physiological conditions may modify phosphatase activity via phosphorylation of MYPT1 at Thr 695 . However, there are several aspects that remain to be resolved. For example, which kinase is responsible for phosphorylation of MYPT1 in Swiss 3T3 cells in the non-stimulated state (Fig. 4) and what is the role of this phosphorylation? In addition, it was proposed that phosphorylation of MYPT1 may induce activation as well as inhibition of phosphatase activity. It was shown that MYPT1 was phosphorylated by a mitosis-specific kinase and activated MP activity (57). Also the activation of MP in vascular smooth muscle by elevated cGMP levels has been suggested (58,59). Thus, several features of MYPT1 and possibly MYPT2, must be considered to obtain a more complete picture of its regulation of PP1c activity.
In summary, the results presented above are supportive of a physiological role for Rho-kinase in the regulation of MP activ- ity. A direct phosphorylation of the MYPT1 subunit by Rhokinase is indicated and the major site involved in inhibition of phosphatase activity is the Thr 695 . This site is located in the C-terminal one-third of the MYPT1 molecule and is separated, at least in terms of the primary structure, from the PP1cbinding sites at the N-terminal end of MYPT1. Theoretically, inhibition of PP1c activity could result from modification of PP1c or its interaction or, from changes in substrate binding. Phosphorylation of Thr 695 decrease V max , rather than changing K m and also phosphatase activities with different substrates are inhibited. Thus, it is likely that the inhibitory effect is focused on PP1c and its interactions with MYPT1. A specific antibody was developed for phospho-Thr 695 and its flanking sequence and using this was demonstrated that phosphorylation of Thr 695 occurred under in vivo conditions where the RhoA pathway is thought to be activated, i.e. following LPA stimulation of Swiss 3T3 cells. The Rho-kinase inhibitor, Y-27632, blocked the LPA-induced phosphorylation of Thr 695 .