Dynamics of myosin light chain phosphorylation at Ser19 and Thr18/Ser19in smooth muscle cells in culture

Katsuhiko Sakurada, Minoru Seto, Yasuharu Sasaki


Using the specific antibodies pLC1 and pLC2 for mono- and diphosphorylated 20-kDa myosin light chain (MLC20) at Ser19 and at both Thr18 and Ser19, respectively, we visualized the dynamics of the MLC20phosphorylation in rabbit aortic smooth muscle cells (cell line SM-3) stimulated with PGF. In the resting state, the diphosphorylated form was located in the peripheral region of the cell, such as the leading edge or the adhesion plaque, and the monophosphorylated form was located not only in the peripheral region but also on a discontinuous fibrillary structure along the long axis of the cell. After stimulation with 30 μM PGF, although localization of the monophosphorylated form changed little, the content of the diphosphorylated form increased and the distribution spread along the fibrillary structure to an extent the same as or similar to that of the monophosphorylated form, which colocalized with actin filament bundles. The diphosphorylation of MLC20 was more sensitive to protein kinase inhibitors, HA-1077, HA-1100, staurosporine, wortmannin, and ML-9, than was the monophosphorylation. In light of these observations, we propose that MLC20 diphosphorylation and monophosphorylation are regulated by different mechanisms.

  • visualization
  • monophosphorylation
  • diphosphorylation

regulatory 20-kDa myosin light chain (MLC20) of smooth muscle and nonmuscle myosin has five phosphorylatable sites; Thr18 and Ser19 for myosin light chain kinase (MLCK) (11) and Ser1, Ser2, and Thr9 for both protein kinase C (PKC) and cdc2 kinase (13, 28). Phosphorylation at Ser19 by MLCK regulates activity of actin-activated myosin ATPase (11, 12, 14) and myosin filament formation (14). Ser19 is dominantly phosphorylated in vitro, and Thr18 is phosphorylated only at a high concentration of MLCK (11, 12). The resulting diphosphorylation at Thr18 and Ser19(Thr18/Ser19) further increases actin-activated myosin ATPase activity over that obtained with monophosphorylation at Ser19 (11, 12) and stabilizes myosin filament (14). We reported that contraction of rabbit aortic strips stimulated with PGF, with diphosphorylation of MLC20 at unconfirmed sites, showed a more rapid force development (31) and that the content of diphosphorylated MLC20 increased in a thickened intimal artery, a state producing a higher maximum force than that seen in a normal artery (33). There are reports that not only the monophosphorylation of MLC20 at Ser19 but also the diphosphorylation at Thr18/Ser19occur in smooth muscle tissue such as carbachol-stimulated (5) and neurally stimulated (20) bovine trachea and PGF-stimulated rabbit aorta (31) and in actively growing cultured SM-3 smooth muscle cells (30). On the other hand, there are reports that diphosphorylation of MLC20 correlates with cellular shape change and exocytosis (4, 15). There are also data showing that monophosphorylation at Ser19generates a maximum force in permeabilized smooth muscle and that additional phosphorylation at Thr18 does not augment development of this force (8) and does not increase sliding speed of actin filaments produced by interaction with myosin in in vitro motility assays (36). Thus the function and generation mechanism of diphosphorylation of MLC20 is unknown.

To investigate in detail the mode and role of phosphorylation of MLC20 by MLCK in smooth muscle cells, it is necessary to achieve a higher resolution of the analysis of MLC20 phosphorylation than that obtained with the usual methods, such as Western blotting with anti-MLC20 antibody or two-dimensional gel electrophoresis. We developed phosphorylation site-specific antibodies for mono- and diphosphorylated MLC20 at Ser19 and Thr18/Ser19, respectively, which have facilitated studies of the dynamics of protein phosphorylation. We report here characteristics of the specific monoclonal antibody pLC1 for monophosphorylated MLC20 at Ser19 and the specific polyclonal antibody pLC2 for diphosphorylated MLC20 at Thr18/Ser19and the dynamics of phosphorylated MLC20, by application of these antibodies to Western blotting and immunohistochemistry of cultured SM-3 smooth muscle cells stimulated with PGF. We also confirmed the effects of protein kinase inhibitors on phosphorylation of MLC20 in the cells.



MLCK (37) and MLC20 (39) were prepared from chicken gizzard. Calmodulin (CaM) was prepared from porcine brain, by the method of Yazawa et al. (38). Inhibition constant (K i) values of HA-1100 for PKC and cAMP-dependent protein kinase (PKA) were determined by the method of Hidaka et al. (9). According to the method of Ikebe and Hartshorne (11), MLC20 (0.4 mg/ml) was mono- and diphosphorylated by incubation for 60 min at 25°C with 1.2 μM CaM, 1 mM ATP, 3 mM MgCl2, 1 mM CaCl2, and 25 mM Tris ⋅ HCl, pH 7.5, and 0.5 or 50 μg/ml MLCK, respectively, and after placement on ice the reactions were halted. The mono- and diphosphorylated MLC20were loaded on a Mono Q column (Pharmacia) and eluted by an NaCl linear gradient from 0 to 0.5 M. Production and characterization of polyclonal anti-MLC20 antibody was described by Sasaki et al. (27). MLC20phosphatase was a kind gift from Dr. M. Ito (Mie University, Japan). 1–5-(Isoquinolinesulfonyl)homopiperazine HCl (HA-1077) was synthesized from 5-isoquinolinesulfonic acid (2). 1-Hydroxy-HA-1077 (HA-1100) was synthesized by the method of Hidaka and Sone (10). ML-9 was obtained from Seikagaku Kogyo. Staurosporine and wortmannin were obtained from Wako Pure Chemical. All other chemicals were of reagent grade or better.

Production and characterization of anti-phosphopeptide antibodies.

Peptide monophosphorylated at Ser19 (pLC1 peptide; Lys11-Lys-Arg-Pro-Gln-Arg-Ala-Thr-phosphoSer-Asn-Val-Phe22-Cys) was prepared enzymatically from the nonphosphorylated peptide (LC peptide) by the method of Sakurada et al. (26). Peptide diphosphorylated at Thr18 and Ser19 (pLC2 peptide; Pro14-Gln-Arg-Ala-phosphoThr-phosphoSer-Asn-Val-Phe-Ala-Met24-Cys) was synthesized by the Peptide Institute (Osaka, Japan). pLC1 and pLC2 peptides were coupled with keyhole limpet hemocyanine (KLH), as described by Nishizawa et al. (23). Each 2 mg of pLC1 and pLC2 peptides coupled with 6 mg of KLH were emulsified with the Ribi adjuvant system (RIBI ImmunoChem Research, Hamilton, MT). The emulsified pLC1 peptide-KLH conjugate (50 μg) was injected into BALB/c mice. At 2-wk intervals, a booster injection was given and serum-positive mice were given a further tail vein injection. Three days after the final booster, the spleen cells were fused with mouse myeloma cells, P3U1, using polyethylene glycol 1500 (Boehringer Mannheim, Indianapolis, IN). Resulting hybridomas were screened in the following three tests: enzyme immunoassay (EIA) tests, Western blotting, and immunostaining. We selected hybridomas that showed specificity to pLC1 peptide in EIA tests using 96-well plates (Nunc) coated with pLC1 peptide and LC1 peptide. Next, to check whether the selected antibodies showed specificity and immunoreactivity to monophosphorylated MLC20, we did EIA tests and Western blotting of glycerol-PAGE (13%) (31) using non-, mono-, and diphosphorylated MLC20. To check specificity of these antibodies to monophosphorylated MLC20 in homogenates of cultured SM-3 rabbit aortic smooth muscle cells, Western blotting of SDS-PAGE (15%) (27) was done using the homogenate. Antibodies exhibiting specificity for monophosphorylated MLC20 and immunoreactivity were further tested by immunostaining of SM-3 cells. Quenching tests were also done. When the antibodies were preincubated with non-, mono-, and diphosphorylated MLC20 (0.1 mg/ml) for 60 min at room temperature, only monophosphorylated MLC20 quenched the reactivity of the antibodies. The hybridomas that passed through these screenings were subscreened by limited dilution. With four trials of fusion, we obtained three hybridomas and named one of the three pLC1. The isotype of the antibody produced by hybridoma pLC1 was determined using an isotyping kit (Amersham) and was termed IgG1. One milligram of the emulsified pLC2 peptide-KLH conjugate was injected into 10 New Zealand White rabbits, and a booster injection was given at 2-wk intervals. Specificity and immunoreactivity of resulting antisera were checked using similar EIA tests, Western blotting, and immunostaining. We obtained only one specific antiserum for diphosphorylated MLC20 at Thr18/Ser19that was sufficient to pass these tests and quenching tests from 7 wk after the initial injection, and this was antibody pLC2.

Cell culture and measurements of MLC20phosphorylation.

Rabbit aortic smooth muscle cells (cell line SM-3) were isolated from rabbit aorta and cultured in our laboratory (27). Characterization was on the basis of biochemical, histochemical, and pharmacological factors, as a reversible cell line between contractile and synthetic phase (30). We used 0.5% FCS-MEM for preparation of growth-arrested cells and 10% FCS-MEM for preparation of growing cells on plastic culture dishes (Iwaki). The growth-arrested cells were prepared according to our methods (30); the 5-day culture of SM-3 cells after confluence in 10% FCS-MEM was gently dispersed and replated at a density of 7 × 105cells/60-mm-diameter dish containing 0.5% FCS-MEM for 5 days, and then the cells were stimulated with 30 μM PGF. To investigate the effects of the protein kinase inhibitors HA-1077, staurosporine, wortmannin, and ML-9 on phosphorylation of MLC20 in the cells, the cells were preincubated with these compounds for 10 min and then stimulated without removal of these compounds. Beforehand, we examined preincubation time of the inhibitors (5, 10, 15, 20, and 30 min), and 10-min preincubation was sufficient to maximally inhibit MLC20 phosphorylation (data not shown). At appropriate periods after the stimulation, cellular reaction was terminated by addition of 5% trichloroacetic acid followed by scraping with a rubber policeman. The precipitate was washed three times with acetone containing 10 mM dithiothreitol to remove the trichloroacetic acid and then dried. The dried cell powder was placed in 50 μl of glycerol-PAGE sample buffer containing 20 mM Tris base, 22 mM glycine, 10 mM dithiothreitol, 8 M deionized urea, and 0.1% bromphenol blue, and the preparation was then incubated for 30 min at room temperature. The samples were then passed through a 0.45-μm membrane filter (Millipore) to remove undissolved materials. For applyication to SDS-PAGE, we added 1% SDS and 2-mercaptoethanol to the samples and then boiled them for 5 min. The samples (3 μl, 4.2 × 104 cells) were subjected to Western blotting of glycerol-PAGE (13%) (31) and SDS-PAGE (15%) (27). The regions of MLC20 were visualized using diaminobenzidine or the enhanced chemiluminescence method (Amersham). The contents of non-, mono-, and diphosphorylated MLC20 separated on glycerol-PAGE were measured using a densitometer (Densitron PAN-FV, Jooko, Japan). Although both the major faster migrating and the minor slower migrating bands were recognized by anti-MLC20 antibody, we did not confirm that the minor bands were for an isoform of MLC20. Therefore, we did not include the minor bands in the densitometric analysis. We confirmed that the intensities of bands recognized with anti-MLC20 antibody were linear in a range of cell number between 1 × 104 and 6 × 104 cells/lane. As there were no differences in the resulting values of Western blotting between cells cultured on plastic dishes and those cultured on glass dishes, Western blotting of cultured cells on plastic culture dishes was done.

Histochemical immunostaining of SM-3 cells.

The SM-3 cells cultured on nontreated glass coverslips were stimulated with PGF and fixed in 1% formaldehyde-PBS for 10 min at room temperature, followed by permeabilization in 0.1% Triton X-100-PBS for 2 min. After gentle washing in PBS, the permeabilized cells were incubated in blocking buffer (1% nonfat dry milk-PBS) for 60 min to minimize nonspecific binding of immunoglobulin. Subsequently, cells were incubated with antibody pLC1, which was the hybridoma supernatant diluted at 1:10 with blocking buffer, or antibody pLC2, which was the antiserum diluted at 1:500 with blocking buffer for 60 min. After washing in PBS, secondary antibody staining was carried out with anti-mouse IgG or anti-rabbit IgG conjugated with FITC or Texas red (1:50 dilution for both; Amersham) for 60 min. Actin filament bundle was visualized with phalloidin-tetramethylrhodamine isothiocyanate (TRITC) label (1:1,000 dilution; Sigma), as described by Sasaki et al. (27). After being washed in PBS, coverslips were mounted with 50% glycerol-PBS containing 5% triethylenediamine. Quenching assays were done as follows: antibodies pLC1 and pLC2 preincubated for 60 min with MLC20 (0.1 mg/ml) that had been non-, mono-, or diphosphorylated by MLCK were subjected to immunostaining of SM-3 cells.


The antibodies evaluated by EIA tests were checked by Western blotting. The mixture (60 ng) of non-, mono-, and diphosphorylated MLC20 (by MLCK) was separated on a glycerol-PAGE gel, transferred to nitrocellulose paper, and then immunostained with anti-MLC20 antibody, antibody pLC1, and antibody pLC2 (Fig.1 A,lanes a, b, and c, respectively). Although anti-MLC20 antibody stained all three forms of MLC20, antibody pLC1 recognized only monophosphorylated MLC20 and antibody pLC2 recognized only diphosphorylated MLC20. The linearity of Western blotting was confirmed in the range from 5 to 300 ng of each MLC20 form (data not shown). Moreover, antibodies pLC1 and pLC2 did not react with MLC20 phosphorylated by PKC (data not shown).

Fig. 1.

Western blotting analysis with antibodies pLC1 and pLC2.A: a mixture of purified nonphosphorylated (0Pi), monophosphorylated (1Pi), and diphosphorylated (2Pi) 20-kDa myosin light chain (MLC20; 60 ng each) with myosin light chain kinase was separated on a glycerol-PAGE gel (13%), transferred to nitrocellulose paper, and then immunostained with anti-MLC20 antibody (lane a), antibody pLC1 (lane b), or antibody pLC2 (lane c).B andC: extract of SM-3 cells (3 × 104 cells) before and after stimulation with 30 μM PGFfor 5 min was applied to Western blotting of glycerol-PAGE (B; 13%) and SDS-PAGE (C; 15%) gels and immunostained with anti-MLC antibody (a), antibody pLC1 (b), and antibody pLC2 (c).Lanes 1 and3 are extracts of cells before and after stimulation, respectively, andlanes 2 and4 are samples oflanes 1 and3, respectively, treated with MLC20 phosphatase.

Next, we applied these antibodies to Western blotting of glycerol-PAGE (13%), using the extract of SM-3 cells (3 × 104 cells) before (Fig.1 B,lane 1) and after (lane 3) 5 min of stimulation with 30 μM PGF. Antibody pLC1 reacted only with the second band representing monophosphorylated MLC20 but not with other bands (Fig. 1 Bb), and antibody pLC2 reacted only with the third band representing diphosphorylated MLC20 (Fig.1 Bc). The intensities of the bands detected by antibodies pLC1 and pLC2 increased after the stimulation (Fig. 1 B,b andc,lanes 1 and3). To confirm that the bands recognized by antibodies pLC1 and pLC2 were phosphorylated forms, we made use of MLC20 phosphatase. The same extract of SM-3 cells (Fig. 1 B,lanes 1 and3) was treated with MLC20 phosphatase and then applied to Western blotting (Fig. 1 B,lanes 2 and4). Intensity of the band representing the nonphosphorylated form was increased and bands representing mono- and diphosphorylated MLC20 vanished (Fig.1 Ba), and antibodies pLC1 and pLC2 recognized no bands (Fig. 1 B,b andc). Figure1 C shows Western blotting of SDS-PAGE (15%), using the same samples as for Fig.1 B. Antibodies pLC1 and pLC2 reacted only with a single protein band located at the same position of the band recognized by the anti-MLC20antibody. Similar to results in Fig.1 B, the intensity of these bands recognized by antibodies pLC1 and pLC2 after stimulation became stronger. These results suggest that the immunoreactivities of antibodies pLC1 and pLC2 were sufficient to detect phosphorylated MLC20 in the extract of SM-3 cells.

Time course of MLC20phosphorylation in growth-arrested SM-3 cells on exposure to 30 μM PGF was detected by Western blotting of glycerol-PAGE, using anti-MLC20 antibody (Fig.2 Aa), and the contents of the three forms were determined by measuring the intensities of the blots (Fig. 2 B). Before the stimulation, MLC20 was already monophosphorylated at 36.1 ± 2.1% of the total MLC20, and diphosphorylated at 3.1 ± 0.6%. Total phosphorylation was 0.42 mol Pi/mol MLC20. The content of monophosphorylated MLC20 increased to a maximum value of 43.3 ± 0.9% (n = 4) 2 min after the stimulation and was sustained for up to 15 min, without any distinct decrease. The content of diphosphorylated MLC20and the total phosphorylation also increased to a maximum value of 23.4 ± 1.9% (n = 4) and 0.88 mol/mol at 5 min and were also sustained. These results are compatible with our previous results (30). This sustained phosphorylation was also observed in a human umbilical vein endothelial cell line (6). Thus the phosphorylation system in cultured cells may tend to shift to a well-phosphorylated state, compared with that in arterial tissue (31).

Fig. 2.

Time course of MLC20phosphorylation in SM-3 cells. A: extracts of cells stimulated with 30 μM PGF for 0 (lane 1), 1 (lane 2), 2 (lane 3), 5 (lane 4), 10 (lane 5), or 15 min (lane 6) in MEM at 37°C were used for Western blotting of glycerol-PAGE (13%) and immunostained with anti-MLC20 antibody (a), antibody pLC1 (b), and antibody pLC2 (c).B: time course of monophosphorylated MLC20 (○), diphosphorylated MLC20 (•), and total phosphorylated MLC20 (□) determined from immunoblotting pattern of Aa. Degree of phosphorylation is expressed as percent of total MLC20. Data are means ± SD from 4 experiments.

Increases in the contents of mono- and diphosphorylated MLC20 were also detected by Western blotting of glycerol-PAGE using antibodies pLC1 and pLC2 (Fig.2 A, band c). The time course of each phosphorylated form correlated well with that detected with the anti-MLC20 antibody (data not shown). Because antibodies pLC1 and pLC2 recognize only MLC20 phosphorylated at Ser19 and Thr18/Ser19, respectively, transitions of each phosphorylation of MLC20 shown in Fig.2 B can be attributed to the activity of MLCK but not to that of PKC in the cells. This consideration is supported by our recent results showing that stimulation with PGF induced MLC20 phosphorylation at Ser19 and Thr18/Ser19but not at Ser1, Ser2, and Thr9 (30).

We detected two bands for each form of nonphosphorylated and phosphorylated MLC20, a major faster migrating band and a minor slower migrating band, with Western blotting of glycerol-PAGE, using extracts of SM-3 cells (Figs.1, B andC, and2 A). The existence of MLC20 isoforms was demonstrated by cloning of cDNAs (19, 34). These isoforms maintain a high degree of homology and contain phosphorylation sites for MLCK and PKC. It was confirmed that the minor slower migrating band was recognized not only by anti-MLC antibody but also by antibodies pLC1 and pLC2. Therefore, the minor bands might be isoforms of MLC20.

To investigate the dynamics of mono- and diphosphorylated MLC20 at Ser19 and Thr18/Ser19, as well as the correlation between the localization of actin and phosphorylated MLC20, we attempted to immunostain SM-3 cells cultured in 0.5% FCS-MEM. In this case, SM-3 cells cultured on the coverslip retained their original shape during exposure to PGF, whereas cells on the plastic culture dish contracted.

Before the stimulation, staining with antibody pLC1 revealed fibrillary structures stretched along the long axis and the periphery of SM-3 cells (Fig.3 A), whereas diphosphorylated MLC20 was mainly located on part of the periphery of the cells, with little in the cytoplasm (Fig. 3 E). After stimulation, the intensity and distribution of the image stained with antibody pLC1 changed little (Fig. 3,B andC), a finding compatible with data obtained with Western blotting (Fig.2 Ab). In contrast, intensity and distribution of the image stained with antibody pLC2 were distinctly and drastically changed within 2 min after stimulation. One minute after stimulation, the distribution of diphosphorylated MLC20 spread in peripheral regions and the intensity strengthened (Fig.3 F). Five minutes after stimulation, the immunostained region reached the cytoplasm and localization was the same as for monophosphorylated MLC20 at Ser19 (Fig.3 G). The intensity of the image increased in a time-dependent manner, and the time courses were also compatible with findings in Fig. 2 Ac. Immunostaining structures of both phosphorylated forms were distributed on a distinct fibrillary structure but not on amorphous and diffusing structures in the cytoplasm before and after stimulation. Preincubation with monophosphorylated MLC20completely quenched the activity of antibody pLC1 (Fig.3 D), although the other forms, such as non- and diphosphorylated MLC20, showed no inhibition of activity of the antibody (data not shown). Similarly, only diphosphorylated MLC20 quenched the reactivity of antibody pLC2 (Fig.3 H).

Fig. 3.

Double-label immunofluorescence microscopy of control and PGF-stimulated SM-3 cells with primary antibody pLC1 (AC), pLC2 (EG), and pLC1 and pLC2 preincubated with 0.1 mg/ml mono- and diphosphorylated MLC20(D andH, respectively) for 60 min before application to cells and with second antibodies conjugated with FITC (AD; anti-mouse IgG antibody) and Texas red (EH; anti-rabbit IgG), respectively. Time course of distribution of mono- and diphosphorylated MLC20 on exposure to 30 μM PGF for 0 min (A andE), 1 min (B andF), or 5 min (C,D, G, and H). Before stimulation, monophosphorylated MLC20 at Ser19 is localized in a fibrillary structure and peripheral region (A), and diphosphorylated MLC20 at Thr18/Ser19is localized in peripheral region (E). On PGF stimulation, distribution of diphosphorylated MLC20 becomes same as that of monophosphorylated MLC20(F andG). Preincubation with mono- and diphosphorylated MLC20 led to absorption of activities of these antibodies (D andH, respectively). Bar, 20 μm.

Figure 4 shows the correlation between localization of phosphorylated MLC20 and actin detected by double staining with antibodies pLC1 (A) or pLC2 (B) and with TRITC-phalloidin (C andD) in cells stimulated with 30 μM PGF for 5 min. The actin filament was a distinct bundle and stretched along the long axis of SM-3 cells. Both monophosphorylated MLC20 and diphosphorylated MLC20 were present on the actin filament bundles. Even before stimulation, phosphorylated MLC20 also colocalized on the actin filament bundle (data not shown). Although the actin filament bundles existed as a continuous fibrillary structure in the cells, phosphorylated MLC20 existed as an assembly of abundant dots on the fibrillary structure. In particular, the discontinuous distribution was more remarkable than in the case of monophosphorylation.

Fig. 4.

Double-label immunofluorescence microscopy of PGF-stimulated (5 min) SM-3 cells with primary antibodies pLC1 (A) or pLC2 (B) and secondary antibodies conjugated with FITC (A andB) and with tetramethylrhodamine isothiocyanate-phalloidin (C andD). Both mono- and diphosphorylated MLC20 are located on an actin filament bundle. Bar, 8 μm.

To reveal the participation of protein kinases in generation and distribution of mono- and diphosphorylated MLC20 in PGF-stimulated SM-3 cells, we investigated the sensitivity of MLC20 phosphorylation against different types of protein kinase inhibitors. Some are MLCK inhibitors, when used at an appropriate concentration, such as wortmannin and ML-9, and others are nonselective inhibitors, such as HA-1077, HA-1100, and staurosporine, that act against MLCK, PKC, PKA, and phosphatidylinositol 3-kinase. ML-9 at 30 μM, much higher than theK i value for MLCK (3.8 μM; Ref. 25), was required to inhibit the diphosphorylation yet had no detectable effect on monophosphorylation (Fig.5). A similar tendency was observed in the inhibition by 3 μM wortmannin. Although the diphosphorylation was also much more sensitive to the nonselective inhibitors than was monophosphorylation, 10 μM HA-1077, 10 μM HA-1100, and 10 nM staurosporine, concentrations lower than eachK i value for MLCK (1, 3, 26), decreased the diphosphorylation to a basal level of 2–3%.

Fig. 5.

Effects of protein kinase inhibitors on phosphorylation of MLC20 in SM-3 cells stimulated with PGF. SM-3 cells were cultured in MEM containing 0.5% FCS (0 min), preincubated for 10 min with HA-1077, HA-1100, staurosporine (Str), wortmannin (WM), or ML-9, and then stimulated with 30 μM PGF for 5 min, without removing compounds. Data for 0 and 5 min (left) were obtained from cells without exposure to compounds, before and after stimulation. Contents of monophosphorylated (open bar) and diphosphorylated (solid bar) MLC20 were determined by Western blotting of glycerol-PAGE (see materials and methods). Data are means ± SD from 4 experiments.

Next, to give attention to the inhibitory effects of these compounds on phosphorylation of MLC20 at Ser19 and Thr18/Ser19in SM-3 cells, we attempted to visualize the effects using antibodies pLC1 and pLC2 (Fig. 6). As in Fig. 5, ML-9 and wortmannin showed weak inhibitory activity; 100 μM ML-9 and 3 μM wortmannin finally decreased the intensity of the image stained with antibody pLC2 yet only slightly affected the image stained with antibody pLC1 (data not shown). On the other hand, the nonselective inhibitors showed strong inhibition even underK i values for MLCK. Ten micromolar HA-1077 strongly decreased the intensity of the image of diphosphorylated MLC20(Fig. 6 F), although the image of monophosphorylation changed little (Fig.6 B). Thirty micromolar HA-1077 eliminated all evidence of diphosphorylation (Fig.6 G), and monophosphorylation disappeared partially (Fig. 6 C). Thirty micromolar HA-1100 showed the same pattern of inhibition seen with HA-1077 (Fig. 6, D andH). Ten nanomolar staurosporine inhibited diphosphorylation of MLC20 but diffused the structure of monophosphorylated MLC20 (data not shown). There are also differences among these compounds in inhibition of the pattern of MLC20phosphorylation. Wortmannin and ML-9 tended to inhibit MLC20 diphosphorylation in cytoplasm of the cells more strongly than in the peripheral region, whereas HA-1077 and staurosporine inhibited it in a uniform manner. Thus differences in the effects of each compound on the distribution could be confirmed only by immunostaining with antibodies pLC1 and pLC2.

Fig. 6.

Double-label immunofluorescence microscopy of control and PGF-stimulated SM-3 cells with primary antibodies pLC1 (AD) and pLC2 (EH) and secondary antibodies conjugated with FITC (anti-mouse IgG antibody) and Texas red (anti-rabbit IgG), respectively. Condition of each cell is same as for Fig. 5. Shown are cells 5 min after stimulation without inhibitors (A andE), with 10 μM HA-1077 (B andF), with 30 μM HA-1077 (C andG), and with 30 μM HA-1100 (D andH). Bar, 40 μm.


Diphosphorylation of MLC20(Thr18/Ser19) has been noted in intact smooth muscles such as carbachol-stimulated (5) and neurally stimulated (20) bovine trachea and PGF-stimulated rabbit aorta (31) and in actively growing cultured SM-3 smooth muscle cells (30). Additionally, in experimentally pathological models, intimal hyperplasia in rabbit carotid artery (33) and interleukin-1-treated segments of porcine coronary artery (16), diphosphorylation of MLC20 was increased or appeared, compared with control artery. We proposed that in such arteries the augmented force generation (spastic contraction) is associated with increase or appearance of MLC20diphosphorylation. In this present work, we investigated the dynamics of mono- and diphosphorylation of MLC20 using phosphorylated site-specific antibodies and various types of protein kinase inhibitors.

Ikebe and colleagues (11, 12) reported that Ser19 is 500-fold more readily phosphorylated by MLCK than is Thr18. On the basis of this property of MLCK, diphosphorylated MLC20 can be generated after monophosphorylation is almost finished and thus hardly coexists with nonphosphorylated MLC20 in vitro (11). On the other hand, as shown in Fig. 2, the extract of SM-3 cells contains three forms of MLC20: non-, mono-, and diphosphorylated forms make up 35.3, 41.3, and 23.4%, respectively, at 5 min after PGF stimulation. Therefore, we expected that the three forms of MLC20 would not coexist in the same cell at the same time and that there were two different states in the cell culture, one with non- and monophosphorylated MLC20 and the other with mono- and diphosphorylated MLC20. However, we could not confirm this from immunoblotting analysis of cell extracts because immunoblotting analysis, which determines the contents of phosphorylated MLC20 as only the average value of a large number of cells, is not sufficient to precisely determine the phosphorylation states and subcellular distribution of phosphorylated MLC20. We then made use of immunohistochemical staining with antibodies pLC1 and pLC2, in which all the SM-3 cells uniformly responded. Diphosphorylated MLC20 was observed in all SM-3 cells, and the distribution site was much the same as that seen in monophosphorylated MLC20. Taken together with the result of immunoblotting analysis, this result suggests that nonphosphorylated MLC20 probably coexists with the diphosphorylated form in the same cell. In an intact cell, MLC20 phosphorylation at Thr18 and Ser19 might result not only from properties of MLCK but also from systems of MLC20 phosphorylation involving MLCK, phosphatase, and their regulation systems.

Ikebe and colleagues (11, 12) reported that, although the Ser19 site in MLC20 is phosphorylated even at a low concentration of MLCK, additional phosphorylation at Thr18 requires a high concentration of MLCK. On the other hand, there are reports that MLCK has actin binding activity (29) and is associated with actin filament in muscle cells (7), suggesting that the actin filament is surrounded by a high concentration of MLCK that is sufficient to phosphorylate MLC20 not only at Ser19 but also at Thr18. These reports support our observations that phosphorylated MLC20, especially diphosphorylated MLC20, is located only on actin filaments (Fig. 4).

We reported that only treatment with 10–100 nM calyculin A, a protein phosphatase inhibitor, strongly induced diphosphorylation of MLC20 at Thr18/Ser19in SM-3 cells, without an increase in intracellular Ca2+ concentrations (30), suggesting that, although in a resting state MLCK activity is sufficient to phosphorylate MLC20not only at Ser19 but also at Thr18, protein phosphatase(s) activities inhibit production of diphosphorylated MLC20. Noda et al. (24) reported that, although a rise of Ca2+concentration induced only monophosphorylation of MLC20 in permeabilized pig aortic smooth muscle cells, additional treatment with 30 μM guanosine 5′-O-(3-thiotriphosphate) (GTPγS), which inhibited dephosphorylation of MLC20 in this system, induced diphosphorylation and that pretreatment withClostridium botulinumexotoxin C3, which specifically inactivates the Rho family of small-molecular-weight G proteins, completely abolished the effects of GTPγS. Therefore, not only the activation of MLCK by Ca2+/CaM but also the inactivation of protein phosphatase(s) would be necessary for the diphosphorylation of MLC20 at Thr18/Ser19; the phosphatase activities might be associated with Rho. As suggested by Katsuyama and Morgan (17), PGF stimulation would induce inactivation of protein phosphatase(s). Dynamics of the distribution of diphosphorylated MLC20 in SM-3 cells stimulated with PGF, which spread from the peripheral region to the central region along actin filaments after stimulation, may result from dynamics of protein phosphatase(s) activity on the actin filament.

These inhibitors fall into two groups: in one group are the MLCK inhibitors wortmannin and ML-9 (at an appropriate concentration), and in the other are the nonselective inhibitors HA-1077, HA-1100, and staurosporine. The former group showed only partial inhibition of MLC20phosphorylation in SM-3 cells, even over a concentration range 2–30 times as high as concentrations sufficient to inhibit purified MLCK activity in an in vitro assay system (Figs. 5 and 6) (22,25), suggesting that inhibitions of MLCK activity do not always result in a proportional inhibition of MLC20 phosphorylation. MLCK activity may be excessive, or there may be alternative pathways for MLC20 phosphorylation. On the other hand, HA-1077, HA-1100, and staurosporine inhibited MLC20 phosphorylation at both Thr18 and Ser19, at concentrations below theK i values for MLCK [36 μM (3), 140 μM (26), and 18.4 nM (1), respectively]. Particularly, HA-1100, which has nonselective inhibitory potential against protein kinases other than MLCK, such as PKC (K i, 18 μM) and PKA (K i, 2.5 μM) (data not shown), strongly inhibited MLC20 phosphorylation of both Ser19 and Thr18 to the same degree and manner at the same concentration as HA-1077. These results suggest that these compounds modulate not only MLCK but also another factor that plays an important role in MLC20phosphorylation. There are reports that PKC is involved indirectly in MLC20 phosphorylation at MLCK sites (21, 30). Because not only HA-1100 but also HA-1077 and staurosporine have potent inhibitory activity for PKC [K i, 3.3 μM for HA-1077 (32) and 4.4 nM for staurosporine (1)], the inhibition study in the present work supports these reports. Therefore, PKC might play an important role in regulation mechanism(s) of phosphorylation or dephosphorylation in SM-3 cells stimulated with PGF.

There are also differences among these inhibitors in the pattern of inhibition of MLC20phosphorylation. The former group tended to inhibit MLC20 diphosphorylation in the cytoplasm more strongly than in the peripheral region (data not shown), and the latter group inhibited uniformly without distinction, even at the peripheral region where diphosphorylation is facilitated, suggesting that the latter group would effectively modulate the factor(s) that regulates diphosphorylation. Because diphosphorylation of MLC20 seems to be associated not only with MLCK but also with protein phosphatase(s) (24, 30), the action point of the latter group might be modulation systems of protein phosphatase activities for MLC20rather than MLCK.

During review and revision of this article, Uehata et al. (35) reported that HA-1077 has potent inhibitory activity to Rho-associated kinase (K i, 0.33 μM), and Rho-associated kinase inhibitor (Y-27632) has the potential to attenuate blood pressure in several hypertensive rat models, probably by inhibiting contraction of resistant arterial vessels. On the other hand, Rho-associated kinase can phosphorylate the 130-kDa subunit of myosin light chain phosphatase and consequently inactivate the phosphatase activity (18). Therefore, the effect of HA-1077, and probably also of HA-1100, on inhibition of MLC20 phosphorylation in SM-3 cells might result not only from direct inhibition MLCK but also from prevention of inactivation of myosin phosphatase via inhibition of Rho-associated kinase.


We thank M. Ohara for helpful comments and C. Kato for secretarial services.


  • Address for reprint requests: Y. Sasaki, Frontier 21 Project, Life Science Research Center, Asahi Chemical Industry Co., Ltd., 2-1, Samejima, Fuji, Shizuoka 416-0934, Japan.


View Abstract