Regulation of Myosin-Bound Protein Phosphatase by Insulin in Vascular Smooth Muscle Cells: Evaluation of the Role of Rho Kinase and Phosphatidylinositol-3-Kinase-Dependent Signaling Pathways

Abstract

In this study, we examined the molecular mechanism of myosin-bound protein phosphatase (MBP) regulation by insulin and evaluated the role of MBP in insulin-mediated vasorelaxation. Insulin rapidly stimulated MBP in confluent primary vascular smooth muscle cell (VSMC) cultures. In contrast, VSMCs isolated from diabetic and hypertensive rats exhibited impaired MBP activation by insulin. Insulin-mediated MBP activation was accompanied by a rapid time-dependent reduction in the phosphorylation state of the myosin-bound regulatory subunit (MBS) of MBP. The decrease observed in MBS phosphorylation was due to insulin-induced inhibition of Rho kinase activity. Insulin also prevented a thrombin-mediated increase in Rho kinase activation and abolished the thrombin-induced increase in MBS phosphorylation and MBP inactivation. These data are consistent with the notion that insulin inactivates Rho kinase and decreases MBS phosphorylation to activate MBP in VSMCs. Furthermore, treatment with synthetic inhibitors of phosphatidylinositol-3 kinase (PI3kinase), nitric oxide synthase (NOS), and cyclic guanosine monophosphate (cGMP) all blocked insulin’s effect on MBP activation. We conclude that insulin stimulates MBP via its regulatory subunit, MBS partly by inactivating Rho kinase and stimulating NO/cGMP signaling via PI3-kinase as part of a complex signaling network that controls 20-kDa myosin light chain (MLC₂₀) phosphorylation and VSMC contraction.

INTRODUCTION

It is well known that insulin promotes relaxation and vasodilation and acutely decreases vascular tone (1–4). However, the molecular mechanism by which insulin regulates relaxation of vascular smooth muscle cells (VSMCs) is unclear. A detailed understanding of how insulin causes vasorelaxation, inhibits VSMC contraction, and decreases vascular tone in normal cells is necessary to explore the pathophysiology of the vascular complications associated with diabetes.

Smooth muscle contraction and relaxation are largely mediated by the phosphorylation and dephosphorylation of the 20-kDa regulatory myosin light chain (MLC₂₀) at threonine 18 and serine-19 by myosin light chain kinase (MLCK) and a myosin-bound serine/threonine specific protein phosphatase (MBP, Refs. 5, 6). Intracellular calcium levels ([Ca²⁺]i) are known to modulate the MLCK to MBP activity ratio and ultimately the degree of contractile force because MLCK activity depends on the amount of the Ca²⁺/calmodulin complex, which itself hinges on [Ca²⁺]i (7–9). However, recent studies with intact smooth muscle suggest the existence of other mechanisms that regulate contraction independent of changes in[ Ca²⁺]i. One such mechanism by which this can occur is the inhibition of MBP activity (10–12). Thus, MLC₂₀ phosphorylation and contractile force can be increased through Ca²⁺ sensitization, a G-protein-coupled Ca²⁺ independent process that inhibits MBP (10–12).

MBP holoenzyme consists of three subunits with molecular masses of 110/130 kDa, 38 kDa, and 20 kDa (13). The 38-kDa subunit is an isoform of the catalytic subunit of protein phosphatase-1 (PP-1C). The two other subunits (110/130 kDa and 20 kDa) are putative regulatory and targeting subunits that bind to myosin and regulate the catalytic activity of the phosphatase (13). Studies on purified preparations of MBP indicate that phosphorylation of the large 130-kDa regulatory myosin-bound subunit (MBS), by an associated kinase, results in an inhibition of phosphatase activity (14). Further studies revealed that an active GTP-bound Rho, a small guanosine triphosphatase, specifically interacts with MBS (15). The Rho-associated kinase phosphorylates MBS and consequently inactivates MBP (14, 15), resulting in an increase in MLC₂₀ phosphorylation and contraction of smooth muscle. Although these results clearly explain how activated Rho increases Ca²⁺ sensitivity of both MLC₂₀ phosphorylation and contraction, the pathway that leads from receptor, to G proteins, and then to the ultimate inhibition of the phosphatase is not known. Arachidonic acid has been suggested as a possible secondary messenger since it dissociates the trimeric structure of the phosphatase leading to its inhibition (5, 16). Moreover, arachidonic acid is known to be released in smooth muscle at concentrations and times consistent with a physiological role (16). In addition, Ikebe and Brozovich (17) reported that agonist-induced activation of protein kinase C can also increase VSMC contraction via an inhibition of MBP.

If phosphorylation of MBP does indeed have a regulatory function in VSMC contraction, then an effective dephosphorylation mechanism must exist and should occur within the time frame observed for physiological effects. Given that insulin’s vasodilatory effects are mediated via nitric oxide (NO) (18), and cGMP inhibits MLC₂₀ phosphorylation (19), it can be hypothesized that insulin may inhibit VSMC contraction by activating MBP via the NO/cGMP signaling pathway. Whether or not dephosphorylation of MBS via agonist-induced signal transduction pathways causes activation of MBP remains unknown.

In this study, we examined whether insulin promotes VSMC relaxation via dephosphorylation of MLC₂₀ by activating MBP and studied the mechanism of MBP regulation by insulin. The results of this study indicate that insulin rapidly stimulates MBP by decreasing the phosphorylation of the myosin-bound regulatory subunit, MBS. Furthermore, insulin-mediated MBP activation and decreased MBS phosphorylation are accompanied by Rho kinase inhibition and an increase in NO/cGMP signaling. Blocking signaling through phosphatidylinositol-3 kinase (PI3-kinase), NO, and cGMP inhibits insulin’s effects on MBP activation. Activation of MBP seems to be coordinated with inhibition of VSMC contraction due to decreased MLC₂₀ phosphorylation.

RESULTS

In the initial experiments, cellular protein phosphatase-1 and protein phosphatase 2A (PP-1 and PP-2A) activities were examined in control and insulin-treated VSMCs using[ ³²P]-labeled phosphorylase a as a substrate. PP-1 constitutes 85% of cellular phosphatase activities while the remaining 15% of activity is due to PP-2A (Fig. 1A). Physiological concentrations of insulin rapidly increase PP-1 activity in a time- and dose-dependent manner (Fig. 1, B and C). A half-maximal stimulation of PP-1 was observed with 0.1 nm insulin with a maximal effect seen between 10 and 100 nm insulin after 5 min (Fig. 1C). In contrast to PP-1, insulin treatment caused a decrease in PP-2A activity (Fig. 1A).

Fig. 1.

Effect of Insulin on Phosphatase Activities in VSMCs A, Cellular distribution and effect of insulin on PP-1 and PP-2A. Confluent serum-starved VSMCs were treated with insulin (10 nm) for 10 min and extracted in phosphatase assay buffer. Equal amounts of proteins (1 μg) were assayed for phosphatase activities in the presence and absence of 1 nm okadaic acid. Results are the mean ± sem of three separate experiments performed in duplicate. B, Kinetics of PP-1 activation by insulin. VSMCs were exposed to 10 nm insulin for the indicated time points and assayed for PP-1 activity with[ ³²P]-labeled phosphorylase a as a substrate. Results are the mean ± sem of four separate experiments performed in duplicate. *, P < 0.05 vs. basal levels. C, Dose response of PP-1 activation by insulin: VSMCs were exposed to various concentrations of insulin for 10 min and assayed for PP-1 activity as detailed in Fig. 1B. Results are the mean ± sem of three separate experiments performed in duplicate.* , P < 0.05 vs. basal levels.

Insulin Activates MBP in VSMCs

Since VSMCs express MBP, we next examined whether the observed effect of insulin on cellular PP-1 activation was due to an increase in MBP activity. Myosin-enriched fractions were prepared by extraction with a high salt buffer followed by sedimentation of myosin by reducing the salt concentration (20). The resulting pellet, containing more than 95% of total myosin, was dissolved in high-salt buffer and MBP activity was measured using [³²P]-labeled MLC and [³²P]-labeled phosphorylase a as substrates. As reported earlier (21), MBP activity constituted 10% of total cellular PP-1 activity, and the myosin-enriched fraction contained only PP-1 and essentially no PP-2A activity (20).

Treatment with 100 nm insulin for 10 min caused a 80% increase in MBP activity. The effect of insulin was sustained for the 20-min incubation period studied (Fig. 2A). Dose-response analyses revealed a maximal increase in MBP activity with 10 nm insulin after 10 min incubation (Fig. 2B). Similar results were obtained using[ ³²P]-labeled phosphorylase a substrate (basal MBP activity = 1.440 ± 0.156 nmol Pi released/mg protein/min; maximal insulin-stimulated MBP activity = 2.4 ± 0.2 nmol Pi released/mg protein/min). PP-1 activity in myosin-depleted supernatants was comparable between control and insulin-treated VSMCs (data not shown). Thus the time and dose-dependent increase observed in cellular PP-1 activity in response to insulin in VSMCs shown in Figs. 1, B and C, was entirely due to an increase in MBP activity seen in Fig. 2, A and B.

Fig. 2.

Effect of Insulin on MBP Activation A, PP-1 activation by insulin in VSMCs is due to stimulation of MBP activity: VSMCs were treated with 100 nm insulin for the indicated times. Myosin-enriched fractions were prepared as detailed in the text. Equal amounts of proteins from myosin-enriched fractions were assayed for MBP activity using[ ³²P]labeled MLC as a substrate. Results are the mean ± sem of three separate experiments performed in duplicate. *, P < 0.05 vs. basal levels. B, Dose-response of MBP activation by insulin: VSMCs were treated with various concentrations of insulin for 10 min. MBP activity was assayed in myosin-enriched fractions. Results are the mean of two separate experiments. C, Insulin does not alter the contents of MBS and PP-1Cδ in the myosin-enriched fractions. VSMCs were treated with insulin (100 nm × 10 min), thrombin (1U/ml × 3 min), insulin + thrombin (100 nm insulin × 10 min followed by 1U/ml thrombin for 3 min). Equal amounts of proteins (20 μg) from myosin-enriched fractions were separated on a 7.5% SDS polyacrylamide gel. The proteins were transferred to PVDF membrane. The membrane was cut into two halves across the 50-kDa marker. Top portion was probed with anti-MBS antibody and the bottom portion was used to analyze PP-1Cδ content. A representative autoradiogram is shown. Similar results were obtained in multiple experiments. D, Insulin does not increase the association of PP-1Cδ with the MBS. Equal amounts of proteins from control and insulin-treated VSMC lysates were immunoprecipitated with the anti-MBS antibody as detailed in the text. The immunoprecipitates were separated by SDS-PAGE followed by Western blot analysis as detailed in Fig. 2C. A representative autoradiogram is shown.

Western blot analysis indicated that insulin treatment did not alter the contents of either MBS or PP-1Cδ examined in the myosin-enriched fraction (Fig. 2C) as well as the MBS immunoprecipitates (Fig. 2D). Furthermore, thrombin, which is known to inhibit MBP activity, did not alter the content of PP-1Cδ in the myosin-enriched fraction. This observation suggests that insulin may be activating the enzyme already bound to the MBS.

Insulin Decreases the MBS Phosphorylation and Prevents Thrombin-Induced Increase in MBS Phosphorylation

Recent studies suggest that the large 110/130 kDa MBS of myosin-associated phosphatase (MBP) gets phosphorylated by agonists that stimulate smooth muscle contraction (14, 15). Furthermore, phosphorylation of MBS results in a decrease in MBP activity (14, 15). Our observations that insulin rapidly stimulates MBP suggest that insulin may be activating the phosphatase either by altering MBS phosphorylation status and/or activating the catalytic subunit bound to MBS via another mechanism. Therefore, we examined the effect of insulin on MBS phosphorylation status. As shown in Fig. 3A, treatment with 100 nm insulin for 2 min caused a rapid 53% decrease in[ ³²P] incorporation into MBS (Fig. 3A, compare lane 2 vs. lane 1). A 72% decrease in MBS phosphorylation was observed after 5 min insulin incubation (Fig. 3A, compare lane 3 vs. lane 1). The observed decrease in MBS phosphorylation was sustained for 20 min of insulin treatment (Fig. 3A, compare lane 5 vs. lane 1).

Fig. 3.

Effect of Insulin on MBS Phosphorylation A, Insulin decreases the phosphorylation status of MBS.[ ³²P]-labeled VSMCs were treated with 100 nm insulin for 2–30 min. Equal amounts of cell lysate proteins (250 μg) were immunoprecipitated with anti-MBS antibody. The immunoprecipitates were subjected to SDS-PAGE, transferred to PVDF followed by autoradiography. The blot was probed with anti-MBS antibody to estimate the quantity of MBS protein immunoprecipitated. A representative autoradiogram is shown. Data from multiple experiments were quantitated by densitometric scanning and normalized for variations in proteins by dividing the intensity of ³²P signal with that of protein signal. The ³²P-specific activity of controls was assigned a value of 100, and the insulin effect was calculated relative to control. Results are the mean ± sem of four separate experiments. *, P < 0.05 vs. control (0 min). B, Insulin prevents thrombin-mediated increase in MBS phosphorylation. [³²P]-labeled VSMCs were treated with insulin (100 nm) for 10 min followed by the addition of thrombin (1 U/ml) for 5 min. MBS immunoprecipitates were subjected to SDS-PAGE and autoradiography. A representative autoradiogram is shown. Data from two separate experiments were quantitated by densitometric scanning and normalized for variations in proteins by dividing the intensity of ³²P signal with that of protein signal. The ³²P-specific activity of controls was assigned a value of 100, and the insulin and thrombin effect was calculated relative to control. Results are the mean of two separate experiments. C, Insulin prevents thrombin-mediated inactivation of MBP in WKY. VSMCs were treated with and without insulin followed by thrombin as detailed in Fig. 3B. MBP activity was measured in myosin-enriched pellets using [³²P]-labeled MLC as a substrate. Results are the mean ± sem of four independent experiments performed in duplicate. *, P < 0.05 vs. control; **, P < 0.05 vs. thrombin.

Since thrombin inactivates MBP by increasing the phosphorylation status of MBS (22), we next examined whether insulin prevents the thrombin-induced increase in MBS phosphorylation. As shown in Fig. 3B, thrombin caused a 70% increase in MBS phosphorylation over basal levels (Fig. 3B, compare lane 3 vs. lane 1). Pretreatment with insulin for 10 min prevented the thrombin-induced increase in MBS phosphorylation (Fig. 3B, compare lane 4 vs. lane 3). However, insulin did not reverse the effect of thrombin on MBS phosphorylation when added to cells prestimulated with thrombin (Fig. 3B, compare lane 5 vs. lane 3). Thrombin-induced MBS phosphorylation was accompanied by marked reductions in MBP activity (Fig. 3C). Insulin pretreatment prevented thrombin-induced MBP inactivation and restored the activity to levels observed with insulin alone (Fig. 3C).

High Concentrations of Okadaic Acid Inhibit Insulin-Mediated MBS Phosphorylation

The above results raise the interesting possibility that insulin may be dephosphorylating MBS via another phosphatase. To test this possibility, [³²P]-labeled VSMCs were pretreated for 30 min with low (10 nm) and high concentrations (1 μm) of okadaic acid (OA) to inhibit the cellular pool of PP-2A and PP-1 activities. These cells were treated with and without insulin and examined for MBS phosphorylation status. Insulin caused a 57% decrease in MBS phosphorylation. Low concentrations of OA that specifically inhibit PP-2A activity did not prevent insulin-mediated decrease in MBS phosphorylation. In contrast, OA at a higher concentration of 1 μm completely blocked insulin’s inhibitory effect on MBS phosphorylation and increased MBS phosphorylation by 44% over the basal levels. OA alone did not appreciably alter basal levels of MBS.

Insulin Inhibits Rho Kinase Activity

Numerous reports (15, 23) suggest that MBS is phosphorylated by Rho kinase, which is activated upon stimulation with agonists such as thrombin or angiotensin II (AII). We tested the possibility that insulin may be inhibiting Rho kinase activity and thereby decreasing MBS phosphorylation. To test this hypothesis, Rho kinase activity was assayed in anti-Rok-α immunoprecipitates using MBP as a substrate. As shown in Fig. 4A, insulin treatment for 10 min caused a 40% decrease in Rho kinase activity when compared with control lysates. More importantly, pretreatment with insulin effectively prevented the thrombin-mediated increase in Rho kinase activity (Fig. 4A).

Fig. 4.

Effect of Insulin on Rho Kinase Activation A, Insulin inhibits Rho kinase activity and prevents thrombin-mediated increase in Rho kinase. VSMCs were treated with insulin (100 nm) for 10 min followed by 1 U/ml of thrombin for 5 min. Equal amounts of cell lysate proteins (100 μg) were immunoprecipitated with anti-Rok-α antibody. Rho kinase activity in the immunoprecipitates was assayed using MBP as a substrate along with γ-[³²P]ATP. Results are the mean ± sem of four separate experiments performed. *, P < 0.05 vs. control; **, P < 0.05 vs. control/insulin; ***, P < 0.05 vs. thrombin. B, Time course of insulin-mediated inactivation of Rho kinase: VSMCs were treated with 100 nm insulin for 0–20 min. Equal amounts of proteins were immunoprecipitated as detailed in Fig. 4A followed by assay of Rho kinase activity. Results are expressed as percent of control activity in the absence of insulin. C, Insulin inhibits thrombin-mediated translocation of Rho to membrane fraction. VSMCs were exposed to insulin and thrombin as detailed in Fig. 4A. Equal amounts of membrane proteins were subjected to SDS-PAGE followed by Western blot analysis with anti-Rho a antibody, a representative autoradiogram. Similar results were obtained in four different experiments.

The kinetics of Rho kinase inhibition by insulin revealed a rapid 40% decrease in Rho kinase activity after 2 min, which was sustained for the 20-min time period examined (Fig. 4B). The kinetics of Rho kinase inhibition upon insulin treatment correlate very well with the decrease observed in MBS phosphorylation in insulin-treated VSMCs.

The effect of insulin on Rho kinase inactivation was accompanied by an inhibition of thrombin-mediated translocation of Rho from cytosol to the membrane fraction (Fig. 4C). As evident from Fig. 4C, a considerable amount of Rho was present in the membrane fraction in untreated VSMCs (Fig. 4C, lane 1). Treatment with 100 nm insulin for 10 min decreased thrombin-mediated translocation of Rho to the membrane fraction (Fig. 4C, compare lane 4 vs. lane 3). Insulin alone caused a very small change in the amount of Rho in the membrane fraction (Fig. 4C, lane 2).

Impact of NOS/cGMP Signaling on Insulin-Mediated Rho Kinase Inactivation and MBP Activation

Given that insulin’s vasodilatory effects are mediated via NO (18), we next examined the contribution of NOS and cGMP signaling pathways in insulin-mediated Rho kinase inactivation as well as MBP stimulation. VSMCs were pretreated with 1 mm N^Gmonoethyl l-arginine acetate (l-NMMA, a synthetic inhibitor of NOS) and RpcGMP (100μ m, a cGMP antagonist) for 30 min followed by treatment with and without 100 nm insulin for 10 min. Rho kinase activity was measured in the immunoprecipitates, and MBP activity was assayed in myosin-enriched pellets. Both l-NMMA and RpcGMP prevented insulin’s inhibitory effect on Rho kinase and restored Rho kinase activity to control levels [Rho kinase activity (% of control value): control, 100%; insulin, 60 ± 6%; l-NMMA + insulin, 98 ± 6%; RpcGMP + insulin, 101 ± 10%] and also blocked insulin’s effect on MBP activation (Fig. 5). Furthermore, sodium nitroprusside (SNP), a NO donor, and 8-bromo cGMP, a cGMP agonist, both mimicked insulin’s effects on MBP activation (Fig. 5) and Rho kinase inactivation. Combined treatment with insulin and SNP or 8-bromo cGMP did not further increase MBP activation (data not shown). These results suggest that insulin-induced NOS/cGMP signaling pathway may participate in insulin-mediated Rho kinase inactivation and MBP activation. In our earlier studies, we have demonstrated that insulin rapidly induces iNOS protein expression and cGMP generation in VSMCs from WKY while VSMCs isolated from spontaneous hypertensive rats (SHR) exhibit resistance to insulin in terms of iNOS protein induction and cGMP generation (24).

Fig. 5.

Inhibitors of NOS and cGMP Signaling Block Insulin-Mediated MBP Activation Serum-starved VSMCs were treated with l-NMMA (1 mm), RpcGMP (0.1 mm) for 30 min followed by 100 nm insulin for 10 min or treated with 100 nm SNP or 8-bromo cGMP (1 mm) for 10 min. MBP activity was measured in myosin-enriched pellets using MLC as a substrate. Results are the mean ± sem of four independent experiments performed in duplicate. *, P < 0.05 vs. control; **, P < 0.05 vs. insulin.

Effect of Wortmannin on Insulin-Mediated Rho Kinase Inactivation and MBP Activation

In an attempt to understand the upstream signaling components that mediate Rho kinase inactivation, iNOS induction, and MBP activation, we blocked PI3-kinase signaling by treatment with 100 nm wortmannin, a synthetic inhibitor of PI3-kinase and examined insulin’s effect on PI3-kinase activation, Rho kinase inactivation, and MBP activation. Pretreatment of VSMCs with 100 nm wortmannin prevented insulin’s stimulatory effect on IRS-1-associated PI3-kinase activity (Fig. 6A but did not alter insulin’s inhibitory effect on Rho kinase (Fig. 6B). In contrast to its lack of effect on Rho kinase, wortmannin completely prevented insulin-induced MBP activation (Fig. 6C). In our recent studies we have shown that wortmannin blocks insulin’s effects on iNOS protein induction and cGMP generation (24). Pretreatment with rapamycin, a synthetic inhibitor of ribosomal S6 kinase, did not affect insulin-mediated Rho kinase inactivation (data not shown).

Fig. 6.

Inhibition of PI3-Kinase Signaling Blocks InsulinMediated MBP Activation without Affecting Rho Kinase Activity A, Wortmannin inhibits PI3-kinase activation by insulin. VSMCs were pretreated with wortmannin (100 nm) for 30 min followed by treatment with and without insulin (100 nm) for 10 min. Equal amounts of precleared lysate proteins (100 μg) were immunoprecipitated with an antirabbit IRS-1 antibody followed by the assay of PI3-kinase activity in the immunoprecipitates. Top, A representative autoradiogram from TLC performed on PI3-kinase products is shown. Bottom, Phosphatidylinositol phosphate (PIP) spots were scraped and radioactivity was counted. Results are the mean ± sem of three separate experiments. *, P< 0.05 vs. basal, ** P, 0.05 vs. insulin. B, Wortmannin does not prevent insulin’s inhibitory effects on Rho kinase activity. VSMCs were pretreated with wortmannin (100 nm) for 30 min followed by treatment with and without insulin (100 nm) for 10 min. Equal amounts of precleared lysate proteins (100 μg) were immunoprecipitated with anti-ROK-α antibody followed by Rho kinase assay in the immunoprecipitates. Details are given in Fig. 4C, Wortmannin inhibits insulin’s stimulatory effect on MBP activation. Myosin-enriched fractions prepared from control, insulin, wortmannin, ± insulin VSMCs were assayed for MBP activity as detailed in Fig. 2. Results are the mean ± sem of three separate experiments performed in duplicate. *, P < 0.05 vs. control;** , P < 0.05 vs. insulin.

Insulin Causes Dephosphorylation of MLC₂₀

Several studies indicate that phosphorylation of MLC₂₀ is a major event in the contraction of VSMCs. MLC₂₀ phosphorylation status reflects a balance between the activities of MLCK and MBP. To further examine the physiological consequences of MBP activation by insulin, we analyzed the effect of insulin on MLC₂₀ phosphorylation. As shown in Fig. 7, insulin activation of MBP in VSMCs was accompanied by a marked reduction in AII-mediated MLC₂₀ phosphorylation examined in cells pretreated with 100 nm insulin for 10 min followed by treatment with AII (100 nm) for 5 min (Fig. 7).

Fig. 7.

Insulin Causes Dephosphorylation of MLC₂₀ VSMCs were treated with and without insulin (10 nm) for 10 min followed by the addition of AII (100 nm) for 5 min followed by glycerol PAGE and immunoblotting with anti-MLC₂₀ antibody. A representative autoradiogram is shown.

Insulin Inhibits Thrombin-Induced VSMC Contraction

We next examined the ability of insulin to inhibit thrombin-induced VSMC contraction. In these experiments VSMCs (2× 10⁴ cells/cm²) were grown to confluence on collagen-coated polyethylene cell culture inserts in medium containing 0.5% FBS. Serum-starved VSMCs were treated with 100 nm insulin for 30 min followed by treatment with and without thrombin (0.5 U/ml) for 15 min. Transvascular diffusion of horseradish peroxidase (HRP) due to contraction-induced increase in cell permeability was determined spectrophotometrically as detailed in Materials and Methods. Pretreatment with insulin caused a 60% decrease in thrombin-induced contractions (Fig. 8).

Fig. 8.

Insulin Inhibits Thrombin-Induced VSMC Contraction Serum-starved VSMCs were treated with and without insulin(100 nm) for 30 min followed by treatment with thrombin (0.5 U/ml × 15). Cell contraction was measured as detailed in the text. Results are the mean ± sem of three experiments.

Diabetes and Hypertension Are Accompanied by Marked Reductions in Insulin-Mediated MBP Activation

To further understand the importance of MBP in VSMC function, we examined the activity of this enzyme in VSMCs isolated from diabetic GK rats, a model for non-insulin-dependent diabetes mellitus (NIDDM) (25) and spontaneous hypertensive rats. As seen in Fig. 9, both diabetes and hypertension resulted in a 30–40% decrease in basal MBP activity in VSMCs. In addition, these VSMCs were resistant to insulin in terms of MBP activation. While 100 nm insulin caused a 60% increase in MBP activity in VSMCs isolated from control (WKY) rats, VSMCs from Type II diabetic GotoKakizaki (GK) rats and SHR exhibited only a 13% increase in MBP activity over the basal values.

Fig. 9.

Diabetes and Hypertension Are Accompanied by Impaired MBP Activation by Insulin Serum starved VSMCs were treated with and without insulin (100 nm) for 10 min. Myosin-enriched fractions were assayed for MBP activity using [³²P]-labeled MLC as a substrate. Results are the mean ± sem of three separate experiments in duplicate. *, P < 0.05 vs. control; **, P < 0.05 vs. WKY control; ***, P < 0.05 vs. WKY insulin.

DISCUSSION

The results presented in this study clearly indicate that insulin rapidly stimulates MBP in VSMCs isolated from control rats. The kinetics of MBP activation by insulin parallels the time course of the insulin-mediated decrease in MBS phosphorylation as well as Rho kinase activity. Thus, insulin appears to stimulate MBP in part by decreasing the phosphorylation of its regulatory subunit, MBS, via Rho kinase inhibition. Insulin activation of MBP is accompanied by the inhibition of MLC₂₀ phosphorylation as well as the inhibition of VSMC contraction.

Several lines of evidence presented in this study suggest that MBP activation by insulin in VSMCs is mediated by multiple inputs from the Rho kinase and PI3-kinase/NO/cGMP signaling pathways. First, insulin rapidly inactivates Rho kinase, and this inactivation is accompanied by a marked time-dependent reduction in MBS phosphorylation and MBP activation. Second, insulin pretreatment prevents thrombin-mediated Rho kinase activation, MBS phosphorylation, and inactivation of MBP and restores MBP enzymatic activity by partially blocking thrombin’s stimulatory effects on the translocation of Rho to the membrane fraction. Support for our observations comes from a recent study by Feng et al. (26). Using site- and phosphorylation state-specific antibodies, these authors have identified two major inhibitory Rho kinase phosphorylation sites on MBS: Thr⁶⁹⁵ and Thr⁸⁵⁰ (26). Direct phosphorylation of recombinant MBS by Rho kinase was accompanied by inhibition of MBP activity. Furthermore, agonists that cause Rho kinase activation in intact Swis 3T3 cells also induced an increase in Thr⁶⁹⁵ phosphorylation on MBS, and this effect was blocked by a Rho kinase inhibitor, Y-27632 (26). Earlier studies have also shown that treatment of smooth muscle cells with C3 exoenzyme, which ADP ribosylates and inactivates the Rho family members, results in MBP activation because of Rho kinase inhibition (27) supporting the notion that inactivation of RhoA by insulin may cause MBP activation via Rho kinase inhibition.

Unlike the other effects of insulin, which are mediated via the PI3-kinase signaling pathway, insulin’s effect on Rho kinase inhibition does not appear to involve PI3-kinase as wortmannin, the PI3-kinase inhibitor, did not prevent insulin-mediated Rho kinase inactivation although it completely abolished insulin-mediated PI3-kinase activation. Earlier studies by Kureishi et al (28) have also shown that high concentrations of wortmannin (100μ m), which inhibit myosin light chain kinase, did not abolish contractions mediated by the constitutively active form of Rho kinase nor its catalytic activity in permeabilized smooth muscle of rabbit portal vein. However, in our study, inhibition of PI3-kinase with wortmannin effectively blocks insulin-mediated MBP activation by attenuating insulininduced NO and cGMP generation.

From the inhibitor studies, it appears that the NOS/cGMP signaling pathway may mediate insulin’s inhibitory effects on Rho kinase inactivation, thereby decreasing MBS phosphorylation as well as MBP activation. For example, blocking NOS activity with l-NMMA attenuates the effect of insulin on Rho kinase inhibition. Furthermore, treatment with RpcGMP, a cGMP antagonist, prevents insulin-mediated Rho kinase inactivation as well as MBP activation. More importantly, treatment with a cGMP agonist, 8-bromo-cGMP, and SNP, a NO donor, mimics the effect of insulin on Rho kinase inactivation, MBS phosphorylation, and MBP activation. It is not clear, at present, why wortmannin failed to block insulin’s inhibitory effect on Rho kinase. Further studies with transfected VSMCs overexpressing constitutively active and dominant negative mutants of p85^PI3-kinase and p110^PI3-kinase, NOS, and protein kinase G (PKG) are needed to clearly establish whether insulin inactivates Rho kinase via a signal generated by NOS signaling independent of the PI3-kinase pathway. Nonetheless, our results, though indirect, suggest a complex cross-talk between two major contraction/relaxation signaling pathways, Rho kinase and NO/cGMP, to mediate insulin’s stimulatory effect on MBP activation.

Regarding the role of the NO/cGMP in the insulin-mediated MBP activation, studies performed by others (29), as well as our laboratory (24), have shown that insulin stimulates the induction of iNOS protein leading to the generation of NO and causes an elevation in the cGMP levels (24). Also, it is well known that cGMP signaling pathway inhibits the contraction of smooth muscle by directly activating MBP (30). It is not exactly clear how cGMP activates MBP. Our preliminary studies suggest that in addition to its directs effects on MBP, cGMP may be inhibiting Rho kinase, thereby decreasing MBS phosphorylation leading to MBP activation. Studies by Wu et al. (30) on smooth muscle from rabbit ileum suggest that cGMP-mediated activation of PKG may phosphorylate MBS on a specific site, resulting in MBP activation. However, recent studies by Nakamura et al. (31) indicated that phosphorylation of MBS by cGMP-dependent PKG did not affect the phosphatase activity toward MLC₂₀ but phosphorylation of the MBP holoenzyme decreased the binding of MBP to phospholipid. Thus, phosphorylation of MBS by PKG is not a direct mechanism in activating MBP.

Our results do not exclude the possibility that decreased MBS phosphorylation and phosphatase activation by insulin may be due to some unknown mechanism other than Rho kinase inhibition. For example, insulin may activate another phosphatase which dephosphorylates MBS causing MBP activation. This possibility was tested with OA. We observed that higher concentrations of OA completely prevented insulin-mediated MBS dephosphorylation. Similar results were obtained in an in vitro study on purified preparations of myosin phosphatase reported by Ichikawa et al. (14). Although these results suggest the involvement of a type 1 phosphatase, we cannot rule out the possibility that MBP holoenzyme itself may be catalyzing the dephosphorylation of its regulatory subunit, MBS. A detailed analysis of the MBS phosphorylation site(s) that are dephosphorylated by insulin is needed. However, at present, these experiments are not feasible due to the nonavailability of a site-specific phosphoantibody that would react with the phosphothreonine 695 on MBS, which is believed to be involved in inhibition of MBP (27).

In addition to these potential mechanisms for the regulation of MBP activity, recent studies have reported the presence of heat-stable inhibitors of MBP that are activated via phosphorylation by protein kinase C (PKC) and presumably by Rho kinase (32). Therefore, it is plausible that inhibition of Rho kinase by insulin will inhibit the heat-stable inhibitors of MBP, resulting in the activation of the phosphatase.

In contrast to WKY, VSMCs isolated from diabetic GK rats and SHR exhibit marked insulin resistance in terms of MBP activation. In our recent studies we have demonstrated that VSMCs from SHR exhibit impaired iNOS induction in response to insulin (24). To our knowledge, this is the first in vivo study demonstrating MBP activation by insulin in VSMCs via reductions in the MBS phosphorylation status and its abnormal regulation in the insulin-resistant states associated with hypertension and NIDDM. Other studies have demonstrated an inhibitory regulation of MBP by a G protein-dependent mechanism (10–12). Thus, an inhibition of MBP activity by thrombin (22), arachidonic acid (16), PKC (17), and GTPγS (10–12) have been reported. This study adds a new dimension to the above observations by demonstrating that insulin’s stimulatory effect on MBP activation is due, in part, to a reduction in the MBS phosphorylation state in addition to the concomitant activation of the enzyme bound to MBS via cGMP signaling, thereby causing MLC₂₀ dephosphorylation and insulin-mediated vasorelaxation. Therefore, the impaired vasorelaxation observed in patients with diabetes and hypertension may be due to inherent reductions in MBP activity resulting from defective regulation of MBP activation in response to insulin. Given the knowledge that PKC levels are elevated in VSMCs isolated from diabetic rat aortas (33) and the fact that PKC can activate Rho and inhibit MBP, it is tempting to speculate that the impaired MBP activation by insulin observed in diabetes and SHR may be due to an elevation in the PKC activity via excessive release of arachidonic acid by phospholipase A2 (23). Arachidonic acid could increase Rho kinase activity as well as interact directly with MBS, causing dissociation of the holoenzyme, thereby reducing MBP activity (23). Alternatively, arachidonic acid could activate a kinase that phosphorylates MBS and inhibits MBP. Our results complement the above observations by documenting that MBS is phosphorylated under basal conditions and the enzymatic activity of MBP is low. Insulin treatment relieves the inhibition and restores the enzymatic activity of MBP by decreasing MBS phosphorylation.

In summary, the results of this study suggest that insulin utilizes the Rho kinase and NO/cGMP signaling pathways to reduce the phosphorylation of MBS and activate MBP which causes vasorelaxation by the dephosphorylation of MLC₂₀.

MATERIALS AND METHODS

Cell culture reagents, FBS, phosphorylase kinase, and phosphorylase b were purchased from Life Technologies, Inc. (Gaithersburg, MD). [γ-³²P]-ATP (specific activity ≥3000 Ci/mmol) and[ ³²P]orthophosphoric acid were purchased from New England Nuclear Corp. (Boston, MA). 8-bromo-cGMP, l-NMMA, wortmannin, and RpcGMP were purchased from BIOMOL Research Laboratories, Inc. (Plymouth Meeting, PA). Okadaic acid was from Moana Bioproducts (Honolulu, Hawaii). Porcine insulin was a kind gift from the Eli Lilly & Co. (Indianapolis, IN). Type-1 Collagenase was from Worthington Biochemical Corp. (Freehold, NJ). SDS-PAGE and Western blot reagents were from Bio-Rad Laboratories, Inc. (Hercules, CA). Antibody against the 160-kDa ROK-α was purchased from Transduction Laboratories, Inc. (Lexington, KY). Antibody against the plekstrin homology domain of IRS-1 was obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Monoclonal antibody against Rho A was purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Myosin light chain (MLC₂₀) antibody, antimouse IgG-Agarose, Protein A-Sepharose CL-4B, protease inhibitors, calmodulin, sodium orthovanadate, angiotensin II, thrombin, and all other reagents were purchased from Sigma (St. Louis, MO). Anti-MBS antibody and MLCK were a kind gift from Dr. Hartshorne (Tucson, AZ). Myosin light chains were prepared from chicken gizzards according to the published protocol (34).

Culture of VSMCs and Treatment with Insulin

VSMCs in primary culture were obtained by enzymatic digestion of the aortic media of male Wistar Kyoto (WKY) rats with body weights of 200–220 g, as described in our recent publications (24, 35). Unless otherwise indicated, primary cultures of VSMCs were maintained inα -MEM containing 10% FBS, and 1% antibiotic/antimycotic mixture. VSMCs isolated from diabetic Goto-Kakizaki (GK) rats, a model for NIDDM, were maintained in medium containing 20 mm glucose to mimic a hyperglycemic condition. Subcultures of VSMCs at passage 5 were used in all experiments. All experiments on MBP activation, MLC₂₀ phosphorylation, and Rho kinase were performed on highly confluent cells (9–11 days in culture) at identical passages. Before each experiment, cells were serum starved for 24 h in serum free α-MEM containing 5.5 mm glucose and 1% antibiotics. The next day, cells were exposed to insulin (0–100 nm) for 0–30 min. In some experiments, VSMCs were pretreated with various inhibitors for 30 min followed by exposure to insulin as detailed in the figure legends.

Preparation of Myosin-Enriched Fractions

Myosin-enriched fractions of VSMCs were prepared by extraction with a high-salt buffer as described previously (20). This fraction contains only PP-1 activity and essentially no PP-2A activity.

Measurement of Myosin-Bound Phosphatase Activity

Phosphatase activity in myosin-enriched fractions was assayed using [³²P]-labeled phosphorylase a as well as[ ³²P]-labeled MLC as substrates (21). Briefly, equal amounts of proteins (1 μg) were diluted with assay buffer (50 mm Tris-HCl, pH 7.5, 0.1 mm EDTA, 28 mm β-mercaptoethanol, and 30 mm KCl). The reaction was initiated by the addition of[ ³²P]-labeled substrates and stopped after 5 min incubation at 30 C by the addition of 20% trichloroacetic acid (TCA). The radioactivity released in the TCA supernatants was counted as detailed in our recent publications (36, 37).[ ³²P]-labeled phosphorylase ’a’ was prepared by incubating [γ-³²P]-ATP with purified phosphorylase kinase and phosphorylase b (38).[ ³²P]-labeled MLC was prepared according to the published protocol (39) by incubating MLC with purified MLCK and 50μ m [γ-³²P]ATP.

Metabolic Labeling of VSMCs and Measurement of MBS Phosphorylation by Immunoprecipitation and Western Blot Analyses

Serum-starved VSMCs labeled with[ ³²P]orthophosphoric acid (0.3 mCi/ml for 4 h) were exposed to various agonists as detailed in the figure legends. The cells were lysed in a buffer containing 50 mm HEPES, pH 7.5, 2 mm EDTA, 1% Triton X-100, 100 mm NaCl, 50 mm β-glycerophosphate, 100 mm NaF, 100 mm sodium pyrophosphate, 2 mm sodium orthovanadate, 2 μm microcystin, and a cocktail of protease inhibitors (36). Precleared lysates with equal amounts of proteins were immunoprecipitated with anti-MBS antibody prebound to Protein A Sepharose for 3 h at 4 C followed by separation of the immunoprecipitates by SDS-PAGE and autoradiography. The level of MBS phosphorylation was measured by densitometric scanning of the autoradiograms. To overcome variations in proteins due to immunoprecipitation, the membranes were probed with anti-MBS antibody followed by incubation with HRP-conjugated secondary antibodies and detection by enhanced chemiluminescence (ECL). The extent of MBS phosphorylation was quantitated by dividing the intensity of radioactive signal with the protein signal.

Immunoprecipitation and in Vitro Assay of PI3-Kinase Activity in the IRS-1 Immunoprecipitates

Equal amounts of precleared lysate proteins (100 μg) were immunoprecipitated with rabbit anti-IRS-1 antibody. PI3kinase activity was assayed in the IRS-1 immunoprecipitates as detailed in our recent publication (24).

Immunoprecipitation and in Vitro Assay of Rho Kinase Activity in the Immune Complexes

Rho kinase was immunoprecipitated by incubating equal amounts of precleared lysate proteins (100 μg) with anti-ROK-α antibody (6μ g/tube) at 4 C with constant shaking. Kinase activity in the immunoprecipitates was assayed using myosin-enriched fraction as a substrate (40). After incubation at 30 C for 10 min (enzyme concentration was adjusted to ensure first-order kinetics), 25 μl aliquots of the reaction mixture were spotted on phosphocellulose paper followed by extensive washing of the paper and ³²P incorporation determined by liquid scintillation spectroscopy.

Analyses of Agonist-Induced Rho Translocation to the Membrane Fraction

Cytosolic and membrane fractions were prepared by differential centrifugation according to previously published protocols (41). Equal amounts of membrane proteins were subjected to SDS-PAGE, transferred to polyvinylidene fluoride (PVDF) membrane, and probed with mouse anti-Rho A antibody followed by incubation with HRP-labeled secondary antibody and subsequent detection with ECL.

MLC Phosphorylation

MLC₂₀ phosphorylation was analyzed by urea, glycerol-PAGE separation of the mono and diphosphorylated forms of MLC₂₀ as detailed earlier (42). MLC₂₀ was detected by immunoblot analysis with MLC₂₀ antibody (Sigma) followed by treatment with HRP-conjugated secondary antibodies and detection by ECL. The autoradiograms were scanned and quantified, and the percent maximal MLC₂₀ phosphorylation was determined by dividing the sum of fast migrating diphosphorylated MLC₂₀ area and the monophosphorylated MLC₂₀ area by the total of phosphorylated and nonphosphorylated areas.

Measurement of VSMC Contraction

VSMC contraction was measured by analyzing the insulin-mediated decrease in thrombin-stimulated transvascular HRP diffusion according to the published protocol (22). Briefly, VSMCs plated on collagen-coated polyethylene trephtalate cell culture inserts (3-μm pore size, Becton Dickinson and Co.) were treated in quadruplicate with and without insulin (100 nm) for 30 min followed by the addition of 500 μl serum free medium containing thrombin (0.5 U/ml). After 15 min, the lower compartment was filled with 500 μl serum-free medium and the medium in the upper compartment was replaced with fresh medium containing HRP (0.34 mg/ml). After 1 min, 60 μl of medium from the lower compartment were transferred to a tube and mixed with 860 μl of reaction buffer containing 50 mm NaH₂PO₄, 5 mm Guaiacol and freshly made 0.6 mm H₂O₂. The absorbance was measured at 470 nm after 15 min incubation at room temperature.

Protein Assay

Proteins in the cellular extracts and lysates were quantitated by the bicinchoninic acid (43) or by the Bradford technique (44).

Statistics

The results are presented as means ± sem of four to six independent experiments each performed in triplicate at different times. Paired Student’s t test was used to compare the basal vs. insulin-treated preparations. Unpaired t test or ANOVA was used to compare the mean values between treatments. A P value of <0.05 was considered statistically significant.

Acknowledgments

We thank Dr. David Hartshorne for helpful suggestions and interpretations of results.

This work was supported in part by an Established Investigator Award (N.B.), a Grant-in-Aid New York State affiliate (L.R.) from the American Heart Association, a research award from the American Diabetes Association (N.B.) and funds from the medical eduction and research, Winthrop University Hospital.