Heart-specific Small Subunit of Myosin Light Chain Phosphatase Activates Rho-associated Kinase and Regulates Phosphorylation of Myosin Phosphatase Target Subunit 1

Abstract

Phosphorylation of myosin regulatory light chain (MLC) plays a regulatory role in muscle contraction, and the level of MLC phosphorylation is balanced by MLC kinase and MLC phosphatase (MLCP). MLCP consists of a catalytic subunit, a large subunit (MYPT1 or MYPT2), and a small subunit. MLCP activity is regulated by phosphorylation of MYPTs, whereas the role of small subunit in the regulation remains unknown. We previously characterized a human heart-specific small subunit (hHS-M₂₁) that increased the sensitivity to Ca²⁺ in muscle contraction. In this study, we investigated the role of hHS-M₂₁ in the regulation of MLCP phosphorylation. Two isoforms of hHS-M₂₁, hHS-M₂₁A and hHS-M₂₁B, preferentially bound the C-terminal one-third region of MYPT1 and MYPT2, respectively. Amino acid substitutions at a phosphorylation site of MYPT1, Ser-852, impaired the binding of MYPT1 and hHS-M₂₁. The hHS-M₂₁ increased the phosphorylation level of MYPT1 at Thr-696, which was attenuated by Rho-associated kinase (ROCK) inhibitors and small interfering RNAs for ROCK. In addition, hHS-M₂₁ bound ROCK and enhanced the ROCK activity. These findings suggest that hHS-M₂₁ is a heart-specific effector of ROCK and plays a regulatory role in the MYPT1 phosphorylation at Thr-696 by ROCK.

Introduction

Phosphorylation of myosin regulatory light chain (MLC)² plays pivotal roles in activation of actomyosin, regulation of cell shape, cell motility, and cytokinesis in eukaryotic cells (1–4). MLC phosphorylation is regulated by a balanced activity of two key enzymes: MLC phosphatase (MLCP) and MLC kinase (5). MLCP dephosphorylates the phosphorylated MLC, resulting in inactivation of myosin to cross-link actin filaments (6). It is well known that the MLC dephosphorylation induces Ca²⁺ desensitization of contractility, especially in the smooth muscle (6).

Many earlier works focused on MLCP in smooth muscle, and the current model for MLCP is based on the chicken gizzard MLCP (6). MLCP in smooth muscle is composed of a catalytic type 1 δ isoform (PP1cδ) subunit of ∼38 kDa, a myosin phosphatase target subunit (MYPT) of ∼110 kDa, and a small subunit of ∼20 kDa (sm-M20) (6). PP1cδ required MYPT to interact with specific substrates at precise cellular localizations, and there are two isoforms of MYPT, MYPT1 and MYPT2. MYPT1 distributes ubiquitously in various tissues (7), whereas MYPT2 is mainly expressed in striated muscle and brain (8). MYPT1 and MYPT2 have common structural features as follows: ankyrin repeats in the N-terminal half and leucine zipper (LZ) motif in the C-terminal part. It was reported that both MYPT1 and MYPT2 formed a complex with PP1cδ and other molecules, such as moesin, microtubule-associated protein Tau, and MAP2, at the ankyrin repeats (9, 10). The C-terminal part of MYPT functions as a scaffold of substrates for several kinases and signal proteins, including small GTPase RhoA (11).

There are several mechanisms proposed for the regulation of MLCP activity, and the phosphorylation and dephosphorylation of MYPT were recognized as a major regulatory mechanism (12). Rho-associated kinase (ROCK), which belongs to a family of serine/threonine kinase and is activated by small GTPase RhoA, is known as a primary regulator of MYPT (13), and ROCK phosphorylates MYPT1 at threonine 696 (Thr-696) and threonine 853 (Thr-853) (numbering according to the human isoform) (12). Because MYPT2 contains equivalent phosphorylation sites (Thr-646 and Thr-808), it may also be regulated by the phosphorylation similar to MYPT1 (14). In addition, several other kinases, such as zipper-interacting protein kinase (ZIPK) and p21-activated kinase, are able to phosphorylate Thr-696 of MYPT1 (15, 16). Phosphorylation of MYPT1 at either Thr-696 or Thr-853 inhibited the MLCP activity and increased the sensitivity to Ca²⁺ of muscle contraction in smooth muscle (17). On the other hand, only limited information is available for the MLCP small subunit. The sm-M20 was characterized only in chicken gizzard, and it could bind the C-terminal part of MYPT1. However, it was reported that the binding of sm-M20 and MYPT1 did not affect the MLCP activity (6). It was demonstrated that sm-M20 increased the Ca²⁺ sensitivity of the contractile apparatus in vascular smooth muscle, the effect of which was conferred by the N-terminal half of sm-M20 (18). However, little is known about the functional roles of the small subunit in the regulation of MLCP activity, especially in muscles other than smooth muscle, such as cardiac muscle.

We have previously isolated heart-specific 21-kDa isoforms of the small subunit from human cardiac muscle, designated as hHS-M₂₁ (7). The hHS-M₂₁ is encoded by the same gene for MYPT2, involving exons 14–25, and there are two isoforms, hHS-M₂₁A and hHS-M₂₁B, which differed by structures of LZ motif at the C-terminal end, generated by an alternative splicing of exon 24 (7). It was revealed that hHS-M₂₁ augmented the contraction of permeabilized porcine renal artery and rat cardiomyocyte at a constant Ca²⁺ concentration and that the N-terminal half of hHS-M₂₁ conferred this action. In addition, hHS-M₂₁ bound the C-terminal one-third region of MYPT1 with high affinity, and it bound to the corresponding region of MYPT2 only to a small extent. These findings implied that hHS-M₂₁ played a role in the regulation of MYPT, although it has been established that the sensitivity to Ca²⁺ concentration in cardiac muscle contraction is mainly regulated by the troponin complex.

We demonstrate here that hHS-M₂₁ is a positive regulator of ROCK and involved in the MYPT phosphorylation. This is the first study suggesting a regulatory mechanism by which hHS-M₂₁ activates ROCK and modulates MYPT1 phosphorylation at Thr-696 by ROCK.

EXPERIMENTAL PROCEDURES

Mammalian Two-hybrid (M2H) Assay

The cDNA fragments for MYPT1 corresponding to amino acids (aa) 1–344 (N-terminal one-third of MYPT1; MYPT1-A), aa 345–688 (middle one-third of MYPT1; MYPT1-M), and aa 689–1030 (C-terminal one-third of MYPT1; MYPT1-P) and for MYPT2 corresponding to aa 1–328 (N-terminal one-third of MYPT2; MYPT2-A), aa 329–656 (middle one-third of MYPT2; MYPT2-M), and aa 657–982 (C-terminal one-third of MYPT2; MYPT2-P) were obtained by PCR from total heart cDNA. The C-terminal one-third regions of MYPTs were further divided into two parts as follows: MYPT1-Pt1 (aa 689–877 of MYPT1) and MYPT1-Pt2 (aa 878–1030 of MYPT1) for MYPT1, and MYPT2-Pt1 (aa 657–783 of MYPT2) and MYPT2-Pt2 (aa 784–982 of MYPT2) for MYPT2. The cDNA fragments for hHS-M₂₁ were hHS-M₂₁A and hHS-M₂₁B (two isoforms of hHS-M₂₁; aa 1–208 and 1–224, respectively), hHS-M₂₁-t1 (deletion of the LZ motif from hHS-M₂₁), hHS-M₂₁-t2 (N-terminal half of hHS-M₂₁), hHS-M₂₁-t3 (deletion of N-terminal 56 residues and LZ motif from hHS-M₂₁), hHS-M₂₁-t4 (C-terminal half of hHS-M₂₁A), hHS-M₂₁-t5 (C-terminal half of hHS-M₂₁B), hHS-M₂₁-o1 (N-terminal 56 residues of hHS-M₂₁), and hHS-M₂₁-o2 (deletion of N-terminal 56 residues from C-terminal half of hHS-M₂₁). These cDNAs for MYPT1, MYPT2, and hHS-M₂₁ were obtained by PCR from the MYPT1, MYPT2, and hHS-M₂₁ plasmids, respectively (7). A full-length cDNA fragment (aa 1–186) of human sm-M20 was amplified by RT-PCR from human uterus cDNA. The PCR products were cloned into pCRII (Invitrogen) and sequenced to confirm that no PCR errors were introduced, and the insert cDNAs were excised by digestion with BamHI and SalI (for hHS-M₂₁, sm-M20 and MYPT2, MYPT1-A, and MYPT1-M constructs) or SalI and NotI (for MYPT1-P, -Pt1, and -Pt2 constructs). Excised cDNA fragments were then cloned into pACT vector containing VP16 as a prey (for hHS-M₂₁ and sm-M20 constructs) and the pBIND vector containing GAL4 as a bait (for MYPT constructs) (CheckMate mammalian two-hybrid system, Promega). Sequence of primers and PCR conditions are available upon request. M2H assays were performed using Dual-Luciferase reporter assay system (Promega) as described previously (19). Arbitrary units (AU) were expressed as Firefly luciferase activities corrected by Renilla reniformis luciferase activities to normalize the transfection efficiency.

Pulldown Assay

The cDNA fragments for hHS-M₂₁ and sm-M20 were generated by PCR from VP16-tagged hHS-M₂₁ and sm-M20 plasmids described above and cloned into the pcDNA4/HisMax vector (Invitrogen) to obtain hexahistidine (His₆)-tagged hHS-M₂₁ and sm-M20. The cDNA fragments for MYPT1-Pt3 (aa 822–1002) and MYPT2-Pt3 (aa 775–954), which lacked the LZ motifs and contained the phosphorylation sites at Ser-852/Thr-853, were generated by PCR and cloned into the pACT vector. Pulldown assay was performed using a HIS buffer kit (GE Healthcare) according to the manufacturer's instructions. Briefly, COS-7 cells (2 × 10⁶) were seeded onto 150-mm dishes 1 day before the transfection, and each construct (10 μg) was individually transfected into the cells with 12 μl of COSFectin lipid reagent (Bio-Rad) according to the manufacturer's instruction. After 72 h of transfection, cells were harvested in binding buffer (20 mm sodium phosphate, 500 mm NaCl, 20 mm imidazole, 1% Nonidet P-40) containing a protease inhibitor mixture (Sigma). Total protein concentration was measured by BCA protein assay (Pierce). To evaluate the expression levels of transfected VP16-tagged MYPT, aliquots of the cell lysates were subjected to SDS-PAGE and transferred to a membrane for Western blotting (WB) with primary mouse anti-VP16 monoclonal antibody (mAb) (1:100, Santa Cruz Biotechnology) and secondary rabbit anti-mouse IgG HRP-conjugated Ab (1:1000, Dako A/S). Equivalent amounts of VP16-tagged proteins were used in the experiments. Cell lysates containing His₆-tagged hHS-M₂₁/sm-M20 or VP16-tagged MYPT1/MYPT2 were mixed with each other, and incubated at 37 °C for 2 h to form a binding complex. Subsequently, the admixture was incubated with nickel-Sepharose^TM 6 Fast Flow (GE Healthcare) overnight at 4 °C. After washing three times with binding buffer, bound proteins were eluted by binding buffer containing 500 mm imidazole. The eluted samples were subjected to the SDS-PAGE followed by the WB analysis. The membrane of WB was incubated with primary mouse anti-VP16 mAb or rabbit anti-His₆ polyclonal antibody (pAb) (1:100 each, Santa Cruz Biotechnology) and with secondary rabbit anti-mouse (for mAb) or mouse anti-rabbit (for pAb) IgG HRP-conjugated Ab (1:1000, Dako A/S). Signals were visualized by Immobilon Western chemiluminescent HRP substrate (Millipore) and analyzed in Luminescent Image Analyzer LAS-3000mini (Fujifilm). The densitometric intensities were measured by MultiGauge version 3.0 (Fujifilm).

To mimic the phosphorylation and dephosphorylation states at Ser-852 and Thr-853 of MYPT1, cDNA fragments for wild-type (WT) and mutant MYPT1 (mMYPT1s, aa 689–1002) containing TCT to GAT (Ser-852 to Asp-852), TCT to GCT (Ser-852 to Ala-852), ACA to GAT (Thr-853 to Asp-853), or ACA to GCA (Thr-853 to Ala-853) were prepared where the substitutions were generated by the primer-directed mutagenesis method. Each mMYPT1 cDNA fragment was cloned into pACT to obtain VP16-tagged protein in transfection experiments. COS-7 cells (1 × 10⁶) were seeded onto 100-mm dishes and co-transfected with a combination of pcDNA4/HisMax-hHS-M₂₁A constructs (4 μg) and pACT-mMYPT1 constructs (4 μg) with 9.6 μl of COSFectin lipid reagent. After 72 h of the transfection, cells were harvested, and the pulldown assay was performed as described above.

Phosphorylation Assay for MYPT1 and Ezrin/Radixin/Moesin (ERM)

The cDNA fragments for hHS-M₂₁ were cleaved from pACT-hHS-M₂₁A, -hHS-M₂₁B, and -hHS-M₂₁-t1 plasmids and inserted into pCMV-Tag3 vector (Stratagene) to obtain Myc-tagged hHS-M₂₁ in the transfection experiments. The cDNA fragments of full-length MYPT1 (major isoform in human heart; spliced out of exon 13) and constitutively active form of ROCK2 (ROCK2-act, aa 1–557) were amplified by RT-PCR from human heart cDNA and inserted into pcDNA3.1/Hygro(+) vector and pcDNA3.1/Zeo(+) vector, respectively. As for the MYPT1 phosphorylation assay, COS-7 (4 × 10⁵) and HEK293 (8 × 10⁵) cells were seeded onto 60-mm dishes and co-transfected with a combination of pcDNA3.1/Hygro full-length MYPT1 (2 μg) and pCMV-Tag3-hHS-M₂₁ (0.1, 0.5, 1, and 2 μg) with 4.8 μl of COSFectin lipid reagent (for COS-7 cell) or 9 μl of TransFectin lipid reagent (for HEK293 cells, Bio-Rad). COS-7 cells were also co-transfected with pcDNA3.1/Hygro full-length MYPT1 (2 μg) alone, pcDNA3.1/Hygro full-length MYPT1 (2 μg) plus pcDNA3.1/Zeo-ROCK2-act (2 μg), or pcDNA3.1/Hygro full-length MYPT1 (2 μg) plus pcDNA3.1/Zeo-ROCK2-act (2 μg) plus pCMV-Tag3-hHS-M₂₁ (2 μg), with 7.2 μl of COSFectin lipid reagent. As for the ERM phosphorylation assay, COS-7 cells (4 × 10⁵) were seeded onto 60-mm dishes, and pCMV-Tag3-hHS-M21 (2 μg) or pcDNA3.1/Zeo-ROCK2-act (2 μg) was transfected using 2.4 μl of COSFectin lipid reagent. In some experiments, COS-7 cells were treated with a ROCK-specific inhibitor, Y-27632 (Sigma), or fasudil hydrochloride (Wako) at a final concentration of 20 μm before the transfection, when it was needed.

After 48 h of the transfection, cells were subjected to brief sonication in TNE buffer (1% Nonidet P-40, 1 mm EDTA, 150 mm NaCl, and 10 mm Tris-HCl, pH 7.8) containing a protease inhibitor mixture and phosphatase inhibitors (10 mm NaF and 2 mm Na₃VO₄). After measuring the protein concentration, equal amounts of proteins were subjected to SDS-PAGE and WB experiments. The following primary Abs were used: mouse anti-c-Myc mAb (1:100, Santa Cruz Biotechnology), rabbit anti-phospho-MYPT1-Thr-696 pAb (1:500), rabbit anti-phospho-MYPT1-Thr-853 pAb (1:350, Millipore), rabbit anti-MYPT1 pAb (1:100, Santa Cruz Biotechnology), rabbit anti-ROKα/ROCK2 pAb (1:500, Millipore), rabbit anti-ezrin/radixin/moesin pAb (1:500, Santa Cruz Biotechnology), and rabbit anti-phospho ezrin (Thr-567)/radixin (Thr-564)/moesin (Thr-558) pAb (1:250, Santa Cruz Biotechnology).

Inhibition Assay of ROCK Activity

COS-7 cells (4 × 10⁵) were seeded onto 60-mm dishes, and 24 h later, a ROCK-specific inhibitor, Y-27632 (Sigma) or fasudil hydrochloride (Wako), was added at a final concentration of 1, 10, or 20 μm. Then pcDNA3.1/Hygro full-length MYPT1 (2 μg), pcDNA3.1/Hygro full-length MYPT1 (2 μg) plus pcDNA3.1/Zeo-ROCK2-act (2 μg), or pcDNA3.1/Hygro full-length MYPT1 (2 μg) plus pCMV-Tag3-hHS-M₂₁ (2 μg) were transfected using COSFectin lipid reagent. Cells were cultured with Y-27632 or fasudil hydrochloride for 48 h and harvested in TNE buffer containing a protease inhibitor mixture and phosphatase inhibitors. Cell lysates were subjected to SDS-PAGE and WB analysis, as described above.

Silencing of Endogenous ROCK by Small Interfering RNA (siRNA)

Pre-designed siRNAs for human ROCK2 (siRNA ID, s18162 and s225145) and a nonsilencing siRNA as a negative control were obtained from Ambion (Austin). COS-7 cells (2 × 10⁵) were seeded onto 60-mm dishes and transfected with siRNAs at a final concentration of 5 nm using Lipofectamine 2000 (Invitrogen). After 24 h of the treatment with siRNAs, pcDNA3.1/Hygro full-length MYPT1 (2 μg) and pCMV-Tag3-hHS-M₂₁ (2 μg) were co-transfected into the siRNA-transfected cells. After 48 h of the initial siRNA transfection, each siRNA (5 nm) was transfected into the cells. After an additional 24 h, cells were harvested and subjected to the SDS-PAGE and WB analysis. The primary Abs used for WB were as follows: mouse anti-c-Myc mAb (1:100), rabbit anti-phospho-MYPT1 Thr-696 pAb (1:500), rabbit anti-MYPT1 pAb (1:100), rabbit anti-ROKα/ROCK2 pAb (1:500), and mouse anti-GAPDH mAb (1:100, Santa Cruz Biotechnology).

Co-immunoprecipitation (Co-IP) Assay

The cDNA fragment for full-length MYPT1 was inserted into pCMV-Tag2 vector (Stratagene) to obtain FLAG-tagged full-length MYPT1. COS-7 cells (1 × 10⁶) were seeded onto 100-mm dishes, and either pCMV-Tag3-hHS-M₂₁ (4 μg) or pCMV-Tag2-full-length MYPT1 (4 μg) alone or both were transfected using COSFectin lipid reagent. After 48 h of the transfection, cells were harvested, and co-IP assay was performed as described previously (20). Cell lysates were incubated with 4 μg of control mouse IgG (Caltag Laboratories), mouse anti-c-Myc mAb, or mouse anti-FLAG mAb (Sigma) for co-IP. The primary Abs used for WB were as follows: mouse anti-c-Myc mAb (1:100), mouse anti-FLAG mAb (1:250), and rabbit anti-ROKα/ROCK2 pAb (1:500).

Statistical Analysis

Numerical data were expressed as means ± S.E. Statistical differences were analyzed using Tukey's multiple comparison tests. Statistical analyses were performed using the R statistical computing environment version 2.6.1. p values less than 0.05 were considered to be statistically significant.

RESULTS

Analysis of Binding Domains in hHS-M₂₁ and MYPTs

We first characterized a human equivalent isoform of sm-M20 that had been characterized only in chicken gizzard (21). It was suggested that the N-terminal end of human sm-M20 would be encoded within intron 18 of the gene for MYPT2, whereas the C-terminal residues were translated from the following exons (7, 22). Because we found a region homologous to 5′ side of avian sm-M20 cDNA in intron 18, which was 18.6-kbp upstream of exon 19, this information was used to design a 5′-sided primer to isolate a full-length sm-M20 by RT-PCR. It was demonstrated that human sm-M20 consisted of 186-aa residues, and the N-terminal side (aa 1–34) differed from the N-terminal side of MYPT2 and hHS-M₂₁ (Fig. 1). It was found that exon 24 was not alternatively spliced in encoding the cDNA for sm-M20 (human sm-M20 cDNA sequence was submitted to DDBJ and given an accession no. AB588820). Expression analysis of sm-M20 in various human tissues by RT-PCR showed a high level expression in the tissues abundant with smooth muscle, including uterus, small intestine, and colon, whereas a weak expression and no expression were observed in skeletal muscle and heart, respectively (data not shown).

FIGURE 1.

Schematic representation of constructs used in M2H and pulldown assays. Full-length and deletion mutants for large and small regulatory subunits of MLCP are schematically indicated along with their covering residues. A, constructs for myosin phosphatase target subunits (MYPT1 and MYPT2). Diagonal hatched boxes in the N terminus of MYPTs represent ankyrin repeats. Shaded and open boxes in the C terminus of MYPTs indicate LZ motifs of MYPT1 and MYPT2, respectively. B, constructs for MLCP small subunits (hHS-M₂₁ and sm-M20). Open and dotted boxes in the C terminus of hHS-M₂₁ indicate LZ motifs in hHS-M₂₁A and hHS-M₂₁B, respectively. C terminus of MYPT2 has identical sequences to that of hHS-M₂₁A and sm-M20. N-terminal 34 residues of sm-M20 (vertical hatched box of sm-M20) are different from that of hHS-M₂₁.

To determine the binding domains in hHS-M₂₁ and MYPTs, we first performed M2H assay. A bait construct containing each one-third of MYPTs was co-transfected with a prey construct for hHS-M₂₁ or sm-M20 (Fig. 1), and luciferase activities in co-transfectants were measured. When the luciferase activity in the transfectants with hHS-M₂₁A plus MYPT1-P was arbitrarily defined as 1.00 AU, that of hHS-M₂₁A plus MYPT2-P showed a significant binding (0.40 ± 0.01 AU, p < 0.001, as compared with negative controls containing either MYPTs or hHS-M₂₁ construct alone), and those of hHS-M₂₁B plus MYPTs were 0.21 ± 0.01 or 0.39 ± 0.01 AU for MYPT1-P or MYPT2-P, respectively (p < 0.001 in each case), indicating that both isoforms of hHS-M₂₁ could interact with the C-terminal one-third of MYPTs (MYPT1-P and MYPT2-P) (Fig. 2A), as reported previously by using an overlay assay (7). In addition, hHS-M₂₁A preferentially bound MYPT1-P, whereas hHS-M₂₁B preferentially bound MYPT2-P (Fig. 2A). It should be noted that sm-M20, which differed from hHS-M₂₁A only at the N-terminal part (aa 1–34), bound both MYPT1-P and MYPT2-P, and the luciferase activities in the transfectants with sm-M20 plus MYPTs were similar to that of hHS-M₂₁ plus MYPTs (Fig. 2A).

FIGURE 2.

M2H assay in evaluating the binding of hHS-M₂₁/sm-M20 with MYPTs. Luciferase activities obtained in the M2H assay. A, binding pairs were deletion mutants of MYPT1 or MYPT2 with hHS-M₂₁A, hHS-M₂₁B, or sm-M20. pACT and pBIND indicate transfection with pACT or pBIND vectors, respectively, as controls (no VP16- and GAL4-tagged proteins). B, binding pairs were deletion mutants of hHS-M₂₁A or hHS-M₂₁B with MYPT1-P or MYPT2-P. Data for MYPT1-P with hHS-M₂₁A were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 3 for control or n = 4 for each case). ***, p < 0.001 versus controls.

Next, the hHS-M₂₁-binding site in the C-terminal one-third of MYPTs was analyzed using MYPT1-Pt1, MYPT1-Pt2, MYPT2-Pt1, and MYPT2-Pt2 (Fig. 1A). As shown in Fig. 2A, luciferase activities in the transfectants containing hHS-M₂₁A or hHS-M₂₁B with MYPT1-Pt2 (1.61 ± 0.03 or 0.24 ± 0.01 AU, respectively) or with MYPT2-Pt2 (0.29 ± 0.01 or 0.18 ± 0.01 AU, respectively) were significantly higher than those of the negative controls (p < 0.001 for each case), demonstrating that the C-terminal end of MYPTs (aa 878–1030 of MYPT1 and aa 784–982 of MYPT2) possessed a binding site for hHS-M₂₁. In addition, we analyzed the MYPT-binding site in the deletion mutant of hHS-M₂₁ (Fig. 1B). Luciferase activities in the transfectants containing hHS-M₂₁-t4 plus MYPT1-P and hHS-M₂₁-t5 plus MYPT2-P were significantly higher (0.37 ± 0.01 and 0.28 ± 0.01 AU, respectively, p < 0.001 for each case) than the negative controls (Fig. 2B), suggesting that the C-terminal half of hHS-M₂₁A preferentially bound MYPT1, whereas that of hHS-M₂₁B preferentially bound MYPT2.

To confirm and further investigate the binding between hHS-M₂₁ and MYPTs, we performed a pulldown assay. Consistent with the M2H data showing the interaction between the C-terminal one-third of MYPT1 (MYPT1-P) and MLCP small subunits (hHS-M₂₁A, hHS-M₂₁B, and sm-M20), WB analysis of pulldown products from the mixture of cell lysates prepared from transfectants of VP16-MYPT1-P in combination with His₆-hHS-M₂₁A, -hHS-M₂₁B, or -sm-M20 revealed that the C-terminal one-third of MYPT1 could bind all isoforms of the MLCP small subunit (Fig. 3A). To map the binding site in MYPT1, VP16-MYPT1-Pt2 (aa 878–1030) and VP16-MYPT1-Pt3 (aa 822–1002) were tested for pulldown with His₆-hHS-M₂₁A, -hHS-M₂₁B, or -sm-M20. As shown in Fig. 3A, the MYPT1 series were pulled down, and the interaction of the small subunits with MYPT1-Pt3 was higher than that with MYPT1-Pt2, suggesting that the LZ motif of MYPT1 was not a major binding motif, whereas the region around the phosphorylation site of MYPT1 at Thr-853 might be a binding domain for the small subunits. WB analysis of pulldown products from lysates of His₆-tagged small subunits with VP16-MYPT2-P, -Pt2, or -Pt3 also confirmed the binding of the small subunits with the C-terminal one-third of MYPT2. As was shown in the M2H assay, hHS-M₂₁B bound MYPT2-P stronger than hHS-M₂₁A and sm-M20 (Fig. 3B).

FIGURE 3.

Pulldown assay for binding of hHS-M₂₁/sm-M20 with C terminus of MYPT1 or MYPT2. Binding of hHS-M₂₁A, hHS-M₂₁B, or sm-M20 with MYPT1 (A) or MYPT2 (B) was analyzed. VP16-tagged MYPT1-P or -MYPT2-P was pulled down (PD) with His₆-tagged hHS-M₂₁A, -hHS-M₂₁B, or -sm-M20 and detected by using anti-VP16 Ab. Expression levels of VP16-tagged MYPTs and His₆-tagged small subunits were evaluated by immunoblotting of whole-cell lysates. Densitometric data obtained in the pulldown assays are shown in lower panel. Bars indicate the amounts of pulldown products normalized to the amount of input VP16-MYPTs. Data for VP16-MYPT1-P with His₆-hHS-M₂₁A (A) or VP16-MYPT2-P with His₆-hHS-M₂₁B (B) were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 5 for each case). C, VP16-tagged MYPT1s pulled down with His₆ protein alone or VP16 protein alone pulled down with His₆-tagged small subunits were not detected.

Binding of MYPT1 with hHS-M₂₁ under Mimicking Phosphorylation/Dephosphorylation Status at Ser-852/Thr-853

Because the possible binding domain of MYPT1 for the small subunits contains two phosphorylation sites at Thr-853 and Ser-852, which is phosphorylated by protein kinase A/G and/or ROCK (23, 24), we investigated whether the phosphorylation status of MYPT1 at these sites would affect the binding between the C-terminal MYPT1 and hHS-M₂₁A. We generated VP16-tagged MYPT1 (aa 689–1002) of WT or mutants (mMYPT1), in which Ser-852 and/or Thr-853 phosphorylation sites were replaced by nonphosphorylatable (Ala) or phosphorylation-mimicking (Asp) residues (Fig. 4A) to be used in the pulldown assay. As shown in Fig. 4B, binding of hHS-M₂₁A with mMYPT1–853A or -853D was not significantly different from that with mMYPT1-WT. In clear contrast, mMYPT1–852A/853A, -852D/853D, -852A, and -852D bound hHS-M₂₁A significantly less (0.39 ± 0.02, 0.24 ± 0.02, 0.28 ± 0.01, and 0.25 ± 0.02 AU, respectively, p < 0.01 for each case) than WT (defined as 1.00 AU) (Fig. 4B). These data indicated that the substitutions at Ser-852, irrespective of mimicking the phosphorylation/dephosphorylation status, affected the binding of mMYPT1 and hHS-M₂₁A, whereas the phosphorylation status of mMYPT1 at Thr-853 was suggested not to change the binding.

FIGURE 4.

Binding of hHS-M₂₁ and MYPT1 in mimicking phosphorylation or dephosphorylation status at Ser-852/Thr-853. A, schematic structures of wild-type (WT) and mutant MYPT1 (mMYPT1) truncated the LZ motif from MYPT1-P. Nonphosphorylatable (Ala) and/or phosphorylation-mimicking (Asp) substitutions at Ser-852 and Thr-853 within mMYPT1 are indicated. B, binding of hHS-M₂₁A and MYPT1 with or without the substitutions at Ser-852/Thr-853. VP16-mMYPT1s pulled down (PD) with His₆-hHS-M₂₁A were detected by using anti-VP16 Ab. Expression levels of VP16-MYPTs and His₆-hHS-M₂₁A were evaluated by immunoblotting of whole-cell lysates. Densitometric data obtained in the pulldown assay are shown in the lower panel. Bars indicate the amounts of pulldown products normalized to the amounts of VP16-mMYPT1s. Data for VP16-mMYPT1-WT with His₆-hHS-M₂₁A were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 5 for each case). **, p < 0.01; ***, p < 0.001 versus WT.

Regulation of Phosphorylation at Thr-696 of MYPT1 by hHS-M₂₁

To investigate the functional role of hHS-M₂₁ in the regulation of the phosphorylation/dephosphorylation status of MYPT1, nontagged MYPT1 was co-transfected with Myc-tagged hHS-M₂₁A, hHS-M₂₁B, or hHS-M₂₁-t1 into COS-7 cells. WB analyses of the lysates from the co-transfected cells using antibodies against MYPT1-Thr(P)-696 or -Thr(P)-853 demonstrated a 1.8-fold increase of phosphorylation at Thr-696 (p < 0.001) in the presence of hHS-M₂₁A, whereas Thr(P)-853 level was not altered by the co-expression of hHS-M₂₁A (Fig. 5, A and B). In addition, the phosphorylation level of MYPT1 at Thr-696 was increased depending on the amounts of co-expressed hHS-M₂₁A (Fig. 5, C and D). In clear contrast, expression of hHS-M₂₁B or hHS-M₂₁-t1 did not change the phosphorylation level of MYPT1 at both Thr-696 and Thr-853. Because the increased phosphorylation of MYPT1 at Thr-696 regulated by hHS-M₂₁A might be specific to COS-7 cells, we examined HEK293 cells co-transfected with nontagged MYPT1 plus Myc-tagged hHS-M₂₁A. We obtained essentially the same results of increased phosphorylation of MYPT1 at Thr-696 in the presence of hHS-M₂₁A (data not shown). These findings suggested that hHS-M₂₁A specifically regulated the phosphorylation of MYPT1 at Thr-696 and that the LZ motif of hHS-M₂₁A played a role in the regulation, because such effects were not observed for hHS-M₂₁B or hHS-M₂₁-t1 lacking the LZ motif.

FIGURE 5.

MYPT1 phosphorylation at Thr-696 and Thr-853 in the presence of hHS-M₂₁. A, amounts of MYPT1 phosphorylation level in COS-7 cells transfected with nontagged MYPT1 alone (2 μg) or in the combination of nontagged MYPT1 (2 μg) and Myc-tagged hHS-M₂₁ (2 μg) (referred to in Fig. 1B) were measured by using anti-phospho-MYPT1-Thr-696 (pThr696) or -Thr-853 (pThr853) Abs. Expression levels of MYPT1 and Myc-hHS-M₂₁ were evaluated by immunoblotting of whole-cell lysates. B, densitometric analysis of MYPT1-Thr(P)-696 blotting data in A. Bars indicate the Thr(P)-696 phosphorylation levels normalized to the amounts of total MYPT1. Data for MYPT1 without Myc-hHS-M₂₁ was arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 5 for each case). ***, p < 0.001 versus MYPT1 without Myc-hHS-M₂₁. C, MYPT1 phosphorylation at Thr-696 in the cells transfected with nontagged MYPT1 (2 μg) in combination of Myc-tagged hHS-M₂₁A (0, 0.1, 0.5, 1, and 2 μg). D, densitometric analysis of MYPT1-Thr(P)-696 blotting data in C. Bars indicate the Thr-696 phosphorylation levels normalized to the amounts of total MYPT1. Data for MYPT1 without Myc-hHS-M₂₁ were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 5 for each case). *, p < 0.05; ***, p < 0.001 versus MYPT1 without Myc-hHS-M₂₁.

Effect of ROCK on the Phosphorylation of MYPT1 Regulated by hHS-M₂₁

It has been reported that ROCK phosphorylates MYPT1 at two different phosphorylation sites, Thr-696 and Thr-853 (24, 25), which lead to inhibition of MLCP activity (17, 26). More recently, it was demonstrated that aa 683–866 of MYPT1 was a minimal binding domain of ROCK (27). Because the binding region of MYPT1 with hHS-M₂₁A identified in this study appeared similar to that with ROCK, we hypothesized that ROCK might functionally interact with hHS-M₂₁A in the phosphorylation of MYPT1. We investigated the effect of two ROCK-specific inhibitors, Y-27632 and fasudil, on the phosphorylation levels of MYPT1 at Thr-696 and Thr-853 in the presence of hHS-M₂₁A. COS-7 cells transfected with nontagged MYPT1 alone or co-transfected with nontagged MYPT1 plus Myc-tagged hHS-M₂₁A were treated with Y-27632 or fasudil at the concentrations of 1, 10, and 20 μm, and subjected to the WB analyses. It was demonstrated that the treatments with Y-27632 and fasudil inhibited the increased phosphorylation of MYPT1 at Thr-696 regulated by Myc-hHS-M₂₁A and that the inhibition was in a dose-dependent manner and specific to the phosphorylation status, because the treatment with Y-27632 or fasudil showed no significant effect on the expression levels of MYPT1, hHS-M₂₁A, and endogenous ROCK (Fig. 6, A and B).

FIGURE 6.

Effect of ROCK-specific inhibitors on MYPT1 phosphorylation in the presence of hHS-M₂₁. A, COS-7 cells co-transfected with nontagged MYPT1 alone (left panel) or nontagged MYPT1 plus Myc-tagged hHS-M₂₁A (right panel) were treated with ROCK-specific inhibitor, Y-27632 (0, 1, 10, and 20 μm; upper panel) or fasudil (0, 1, 10, and 20 μm; lower panel). MYPT1 phosphorylation level was measured by using anti-phospho-MYPT1-Thr-696 (pThr696) or -Thr-853 (pThr853) Abs. Expression levels of MYPT1, Myc-hHS-M₂₁A, and endogenous ROCK were evaluated by immunoblotting of whole-cell lysates. B, densitometric analysis of MYPT1-Thr(P)-696 blotting data in A. Bars indicate the Thr-696 phosphorylation levels normalized to the amounts of total MYPT1. Data for without Y-27632 or fasudil were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 5 for each case). *, p < 0.05; **, p < 0.01; ***, p < 0.001 versus no treatment with Y-27632 or fasudil.

To evaluate the involvement of ROCK in the regulation of MYPT1 phosphorylation at Thr-696 by hHS-M₂₁A, ROCK was silenced by using two pre-designed siRNAs (s18162 and s225145) in COS-7 cells co-transfected with nontagged MYPT1 and Myc-tagged hHS-M₂₁A. The amount of targeted ROCK protein was decreased by ∼72% or 86% with s18162 or s225145, respectively, as compared with that with control siRNA (Fig. 7A). Silencing of ROCK by the siRNA, s18162 or s225145, caused significant reduction in the phosphorylation status of MYPT1 at Thr-696 (p < 0.001 for each case versus control siRNA) regulated by hHS-M₂₁A, whereas the knocking down of endogenous ROCK did not affect the expression of MYPT1 and Myc-hHS-M₂₁A (Fig. 7B). These data suggested that hHS-M₂₁A regulated the phosphorylation at Thr-696 of MYPT1 through ROCK.

FIGURE 7.

Effect of ROCK silencing on MYPT1 phosphorylation in the presence of hHS-M₂₁. A, COS-7 cells co-transfected with nontagged MYPT1 and Myc-hHS-M₂₁A (right panel) were transfected with siRNA against endogenous ROCK. Immunoblotting showed specific silencing of endogenous ROCK by double-stranded RNA oligonucleotides in COS-7 cells co-transfected with nontagged MYPT1 and Myc-tagged hHS-M₂₁A. GAPDH is shown as a loading control. B, MYPT1 phosphorylation level was measured by using anti-phospho-MYPT1 Thr-696 (pThr696) Ab. Expression levels of MYPT1 and Myc-hHS-M₂₁A were evaluated by immunoblotting of whole-cell lysates (upper panel). Results of densitometric analysis for MYPT1-Thr(P)-696 blotting data are shown (lower panel). Bars indicate the Thr-696 phosphorylation levels normalized to the amounts of total MYPT1. Data for treatment with control siRNA were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 6 for each case). ***, p < 0.001 versus treatment with control siRNA.

Interaction among ROCK, MYPT1, and hHS-M₂₁

It was reported that MYPT1 bound both ROCK (27) and hHS-M₂₁A (7). To further investigate the interaction among them, FLAG-tagged MYPT1 and Myc-tagged hHS-M₂₁ were individually immunoprecipitated using anti-FLAG or -Myc Ab from COS-7 cells transfected with FLAG-MYPT1 or Myc-hHS-M₂₁A alone. WB analysis of immunoprecipitates from the transfectants revealed that endogenous ROCK bound both FLAG-MYPT1 and Myc-hHS-M₂₁ (Fig. 8). We also co-transfected FLAG-MYPT1 and Myc-hHS-M₂₁A into COS-7 cells. It was found that the binding of endogenous ROCK with FLAG-MYPT1 or Myc-hHS-M₂₁A was not altered in the presence of Myc-hHS-M₂₁A or FLAG-MYPT1, respectively (Fig. 8), suggesting that hHS-M₂₁A and MYPT1 formed a trimetric complex with ROCK.

FIGURE 8.

Binding of ROCK with hHS-M₂₁ and MYPT1 in co-IP assay. Binding of endogenous ROCK with Myc-tagged hHS-M₂₁ and FLAG-tagged MYPT1 was analyzed by co-IP assay by using anti-FLAG and -Myc Abs. Endogenous ROCK co-immunoprecipitated with Myc-hHS-M₂₁ and FLAG-MYPT1 was detected by using anti-ROCK Ab. Expression levels of FLAG-MYPT1, Myc-hHS-M₂₁A, and endogenous ROCK were evaluated by immunoblotting of whole-cell lysates.

Effect of hHS-M₂₁ on the Phosphorylation of MYPT1 Regulated by ROCK

To evaluate the effect of hHS-M₂₁ in the regulation of MYPT1 phosphorylation at Thr-696 and Thr-853 by ROCK, we examined COS-7 cells transfected with nontagged MYPT1 alone, co-transfected nontagged MYPT1 plus nontagged ROCK2-act, or nontagged MYPT1 plus nontagged ROCK2-act plus Myc-hHS-M₂₁A. WB analyses of lysates from co-transfected cells using antibodies against MYPT1-Thr(P)-696 or -Thr(P)-853 showed that the phosphorylation at both Thr-696 and Thr-853 was significantly increased in the presence of ROCK2-act (Fig. 9). The phosphorylation level of MYPT1 at Thr-696 was further increased by the presence of Myc-hHS-M₂₁A with ROCK2-act (Fig. 9, A and B). Interestingly, the increased phosphorylation level of MYPT1 at Thr-853 by ROCK2-act was significantly suppressed by Myc-hHS-M₂₁A (Fig. 9, A and C), suggesting that Myc-hHS-M₂₁A had an inhibitory effect on the phosphorylation at Thr-853 but not at Thr-696 of MYPT1 by ROCK.

FIGURE 9.

Effect of hHS-M₂₁ on phosphorylation of MYPT1 in the presence of ROCK. A, amounts of MYPT1 phosphorylation level in COS-7 cells co-transfected with nontagged MYPT1 alone (2 μg), nontagged MYPT1 (2 μg) plus nontagged ROCK2-act (2 μg), or nontagged MYPT1 (2 μg) plus nontagged ROCK2-act (2 μg) plus Myc-tagged hHS-M₂₁ (2 μg) was measured by using anti-phospho-MYPT1-Thr-696 (pThr696) or -Thr-853 (pThr853) Abs. Expression levels of MYPT1, ROCK2-act, and Myc-hHS-M₂₁ were evaluated by immunoblotting of whole-cell lysates. B, densitometric analysis of MYPT1-Thr(P)-696 blotting data in A. C, densitometric analysis of MYPT1-Thr(P)-853 blotting data in A. Bars in B and C indicate the Thr-696 and Thr-853 phosphorylation levels, respectively, normalized to the amounts of total MYPT1. Data for MYPT1 alone were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 8 for each case). *, p < 0.05; ***, p < 0.001 versus MYPT1 plus ROCK2-act.

Regulation of ERM Phosphorylation by hHS-M₂₁

Because hHS-M₂₁ might increase ROCK activity, we investigated the phosphorylation level of another substrate of ROCK, endogenous ERM (28). WB analyses of lysates from transfected COS-7 cells using antibodies against phospho-ERM demonstrated a 1.9-fold increase of ERM phosphorylation in the presence of hHS-M₂₁A, and the phosphorylation was specifically attenuated by the ROCK inhibitor (Fig. 10), suggesting that ROCK was activated by hHS-M₂₁A.

FIGURE 10.

Endogenous ERM phosphorylation in the presence of hHS-M₂₁. Top panel, immunoblotting showed amounts of endogenous ERM phosphorylation level in COS-7 cells transfected with Myc-tagged hHS-M₂₁ (2 μg) or nontagged ROCK2-act (2 μg) by using anti-phospho-ezrin(Thr-567)/radixin(Thr-564)/moesin(Thr-558) pAb. The cells were treated with a ROCK-specific inhibitor, fasudil (20 μm), before the transfection, when it is needed. Expression levels of Myc-hHS-M₂₁A, nontagged ROCK2-act, and endogenous ERM were evaluated by immunoblotting of whole-cell lysates. GAPDH is shown as a loading control. Densitometric analysis of ERM blotting data is shown in the lower panel. Bars indicate the endogenous phospho-ERM levels normalized to the amounts of total ERM. Data for whole-cell lysates only without transfection were arbitrarily defined as 1.00 AU. Data are represented as means ± S.E. (n = 9 for each case). *, p < 0.05; **, p < 0.01 versus whole-cell lysates lysate without transfection.

DISCUSSION

Ca²⁺ plays a central role in the regulation of the muscle contractile process mediated by the interaction of myosin with actin. In the smooth muscle, the phosphorylation status of MLC is correlated with the Ca²⁺ sensitivity of muscle contraction (13). By contrast, in the cardiac muscle, it is well known that the Ca²⁺ sensitivity of muscle contraction is mainly regulated by the troponin complex, and the significance of MLC phosphorylation in Ca²⁺ sensitization is not fully elucidated. It has been suggested that MLC phosphorylation by MLC kinase is involved in the regulation of cardiac function (29, 30), and recent reports identified a novel heart-specific MLC kinase involved in the MLC phosphorylation (31, 32). Because the MLC phosphorylation is regulated by the balance between MLC kinase and MLCP, the function of MLCP in the cardiac muscle should be unraveled for better understanding the regulation of muscle contractility in the heart. In this study, we showed that the presence of hHS-M₂₁ increased the phosphorylation at the phosphorylation site, Thr-696, in MYPT1, implying that hHS-M₂₁ was a positive regulator of MYPT1 phosphorylation. In addition, we previously reported that hHS-M₂₁ increased the Ca²⁺ sensitivity of muscle contraction (7). Because it was reported that phosphorylation of MYPT1 at Thr-696 inhibited MLCP activity (12), it was likely that the augmented Ca²⁺ sensitivity by hHS-M₂₁ was a reflection of the inhibition of MLCP activity, leading to the increased phosphorylation of MLC. These observations suggested that hHS-M₂₁ played a key role in the regulation of MLCP in the heart.

As shown in our previous report (7) and in this study, hHS-M₂₁ bound both MYPT1 and MYPT2. Interestingly, we revealed in this study that the C-terminal half of hHS-M₂₁A, which contains an identical LZ motif to that of MYPT2 (7), preferentially interacts with the C-terminal one-third of MYPT1, whereas hHS-M₂₁B, which contains another LZ motif (7), preferentially interacts with the C-terminal one-third of MYPT2. These observations are in good agreement with the finding that hHS-M₂₁A, but not hHS-M₂₁B, increased the phosphorylation of MYPT1 at Thr-696, and the functional domain of the phosphorylation was mapped in the LZ motif of hHS-M₂₁A. These findings suggest that the binding of hHS-M₂₁A with MYPT1 plays a crucial role in the MYPT1 phosphorylation. In addition, because the mRNA expression of hHS-M₂₁A was more abundant as compared with that of hHS-M₂₁B and the mRNA level of MYPT1 was not different from that of MYPT2 in the heart (7), the majority of the MLCP activity may be conferred by MYPT1 and hHS-M₂₁A in the heart.

The binding domain of MYPT1 for hHS-M₂₁A included the phosphorylation sites, Ser-852 and Thr-853. The amino acid substitutions at a phosphorylation site of MYPT1, Ser-852, impaired the binding with hHS-M₂₁, suggesting that the region around the phosphorylation sites was important in the binding between MYPT1 and hHS-M₂₁A. Quite interestingly, it was reported that phosphorylation of Thr-853 impaired the binding with myosin (12). One could hypothesize that hHS-M₂₁A acts as a competitor to dissociate MYPT1 from myosin, leading to an altered functional distribution of the MLCP activity. Further studies will be required to demonstrate whether the molecular complex of myosin and MYPT1 would be dissociated or impaired by hHS-M₂₁A.

The most important findings in this study were that the phosphorylation level of MYPT1 at Thr-696 was regulated by hHS-M₂₁, which was attenuated by the ROCK inhibitors or siRNA for ROCK, and that hHS-M₂₁ was capable of binding to ROCK. These observations suggested that ROCK was involved in the hHS-M₂₁A-mediated phosphorylation at Thr-696 of MYPT1. On the other hand, Thr-696 in MYPT1 could also be phosphorylated by other protein kinases, including ZIPK (33), and Hagerty et al. (34) showed that ROCK phosphorylated to activate ZIPK. Therefore, it was speculated that hHS-M₂₁A would indirectly modulate the ZIPK activity to phosphorylate Thr-696 of MYPT1 via activation of ROCK signaling. We could speculate another possibility that the binding of hHS-M₂₁ with MYPT1 concealed the Thr-853 residue in the hHS-M₂₁-binding region, which is the major ROCK phosphorylation site (26), resulting in the phosphorylation only at Thr-696, because increased MYPT1 phosphorylation at Thr-853 by the constitutively active form of ROCK2 was suppressed in the presence of hHS-M₂₁A.

As for the molecular mechanisms of hHS-M₂₁-mediated regulation of MYPT1, activation of ROCK by binding to hHS-M₂₁ was suggested, because the phosphorylation level of ERM, which is another substrate of ROCK, was increased. These observations suggested that hHS-M₂₁ was a heart-specific activator of ROCK. Further studies to identify the other targets of the ROCK-hHS-M₂₁ complex in cardiac muscle will shed light on the novel regulatory mechanism of cardiac function and pathogenesis of cardiac diseases such as cardiomyopathy.

In our previous study using an overlay assay, the binding domain of hHS-M₂₁ for MYPT1 was mapped to the N-terminal half-region of hHS-M₂₁, although the C-terminal half of hHS-M₂₁ showed a weak binding with MYPT1 (7). In clear contrast, it was revealed by using M2H and pulldown assays in this study that the C-terminal half-region of hHS-M₂₁A had the major binding activity for MYPT1. This apparent discrepancy in the binding domain might be due to the difference in methodology. Because the M2H and pulldown assays had an advantage over the overlay assay to demonstrate the protein-protein interaction under the intracellular physiological condition, the binding domain of hHS-M₂₁ should be mainly mapped to the C-terminal half-region. Overall, our results implied that the functional domain in Ca²⁺ sensitization of hHS-M₂₁ was different from its binding domain to MYPT1.

In summary, we investigated the functional role of hHS-M₂₁ in regulation of MYPT phosphorylation. It was revealed that hHS-M₂₁ interacted with MYPT1 and ROCK, which regulate the phosphorylation of MYPT1 at Thr-696 via ROCK. These observations provide new insights into the regulation of MLCP activity in the cardiac muscle.