Cross-talk between Rho-associated Kinase and Cyclic Nucleotide-dependent Kinase Signaling Pathways in the Regulation of Smooth Muscle Myosin Light Chain Phosphatase

Background: Myosin phosphatase is regulated by multisite phosphorylation of its myosin targeting subunit of MLCP (MYPT1) subunit.

Results: Multisite phosphorylation at Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵ has no effect on activity; phosphorylation at Thr⁶⁹⁷ and Thr⁸⁵⁵ is inhibitory.

Conclusion: Cross-talk between cyclic nucleotide and RhoA signaling pathways occurs at both MYPT1 inhibitory sites.

Significance: Knowledge of MYPT1 phosphorylation is key to understanding the molecular basis of smooth muscle contractile regulation.

Abstract

Ca²⁺ sensitization of smooth muscle contraction depends upon the activities of protein kinases, including Rho-associated kinase, that phosphorylate the myosin phosphatase targeting subunit (MYPT1) at Thr⁶⁹⁷ and/or Thr⁸⁵⁵ (rat sequence numbering) to inhibit phosphatase activity and increase contractile force. Both Thr residues are preceded by the sequence RRS, and it has been suggested that phosphorylation at Ser⁶⁹⁶ prevents phosphorylation at Thr⁶⁹⁷. However, the effects of Ser⁸⁵⁴ and dual Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ phosphorylations on myosin phosphatase activity and contraction are unknown. We characterized a suite of MYPT1 proteins and phosphospecific antibodies for specificity toward monophosphorylation events (Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵), Ser phosphorylation events (Ser⁶⁹⁶/Ser⁸⁵⁴) and dual Ser/Thr phosphorylation events (Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵). Dual phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ by cyclic nucleotide-dependent protein kinases had no effect on myosin phosphatase activity, whereas phosphorylation at Thr⁶⁹⁷ and Thr⁸⁵⁵ by Rho-associated kinase inhibited phosphatase activity and prevented phosphorylation by cAMP-dependent protein kinase at the neighboring Ser residues. Forskolin induced phosphorylation at Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵ in rat caudal artery, whereas U46619 induced Thr⁶⁹⁷ and Thr⁸⁵⁵ phosphorylation and prevented the Ser phosphorylation induced by forskolin. Furthermore, pretreatment with forskolin prevented U46619-induced Thr phosphorylations. We conclude that cross-talk between cyclic nucleotide and RhoA signaling pathways dictates the phosphorylation status of the Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ inhibitory regions of MYPT1 in situ, thereby regulating the activity of myosin phosphatase and contraction.

Introduction

Smooth muscle is responsible for the involuntary contraction and relaxation of hollow organs within the body, e.g. blood vessels and gastrointestinal tract (1). Contractile force is driven by the phosphorylation status of Ser¹⁹ of the 20-kDa myosin regulatory light chain (LC₂₀),⁴ which facilitates formation of the actomyosin complex and cross-bridge cycling (reviewed in Refs. 2–4). The extent of phosphorylation of LC₂₀ at Ser¹⁹ is primarily dependent on the relative activities of myosin light chain kinase (MLCK) and myosin light chain phosphatase (MLCP). Although MLCK is a Ca²⁺/calmodulin-dependent protein kinase, MLCP activity can be regulated independently of changes in cytosolic free [Ca²⁺] ([Ca²⁺]_i).

External stimuli such as hormones or neurotransmitters can activate smooth muscle membrane receptors and initiate a cascade of events resulting in an increase in [Ca²⁺]_i via release from intracellular stores (sarcoplasmic reticulum) or entry from the extracellular space (2). Relaxation occurs as [Ca²⁺]_i is restored via re-uptake into the sarcoplasmic reticulum and extrusion to the extracellular space. The decrease in [Ca²⁺]_i leads to inactivation of MLCK and dephosphorylation of LC₂₀ by MLCP (3).

Smooth muscle contraction has frequently been observed in the absence of a change in [Ca²⁺]_i. This phenomenon, commonly called Ca²⁺ sensitization, involves an increase in the MLCK:MLCP activity ratio (3), which is achieved primarily through the activities of Ca²⁺-independent kinases that can directly phosphorylate LC₂₀ (5–8) and/or inhibit MLCP activity (9–11). MLCP is composed of three subunits: a 38-kDa catalytic subunit (PP1cδ), a 115–130-kDa myosin targeting subunit (MYPT1), and an additional 20-kDa subunit (M20). Although a variety of regulatory mechanisms have been documented (reviewed in Refs. 12–14), inhibition of MLCP activity has been shown to result from direct phosphorylation of the MYPT1 regulatory subunit at two key residues: Thr⁶⁹⁷ and Thr⁸⁵⁵ (rat sequence numbering) (9, 11, 15, 16). Thr⁶⁹⁷ phosphorylation inhibits MLCP activity by a competitive inhibition mechanism, as MYPT1 phosphorylated at Thr⁶⁹⁷ becomes a substrate for PP1cδ (17). Phosphorylation at Thr⁸⁵⁵ has been shown to cause dissociation of MYPT1 from phosphorylated LC₂₀, resulting in an apparent decrease in MLCP activity (18), and has also been shown to inhibit MLCP activity without dissociation (16). Additional studies have described intracellular translocations of MYPT1 following Thr⁶⁹⁷ phosphorylation (19). Regardless of their molecular mechanisms, Thr⁶⁹⁷ and/or Thr⁸⁵⁵ phosphorylations are associated with decreased MLCP activity in smooth muscle tissue, serving to increase the ratio of MLCK:MLCP activity and resulting in increased LC₂₀ phosphorylation and contractile force. It is important to note, however, that an increase in Thr⁸⁵⁵ phosphorylation has been observed much more frequently than Thr⁶⁹⁷ phosphorylation in intact smooth muscle preparations (20).

Cyclic nucleotides (cAMP and cGMP) are the principal second messengers involved in smooth muscle relaxation and vasodilatation (21). By activating their cognate protein kinase pathways, i.e. PKAc and PKG, these messengers can elicit smooth muscle relaxation via Ca²⁺-dependent and Ca²⁺-independent pathways. PKAc and PKG can act to lower [Ca²⁺]_i by inhibiting both influx of extracellular Ca²⁺ and release of Ca²⁺ from intracellular stores (21, 22). In addition, PKAc and PKG can regulate MLCP activity (23–25). The two inhibitory Thr residues of MYPT1 are surrounded by similar protein sequences (Fig. 1) and each is immediately preceded by a Ser residue that matches PKAc and PKG phosphorylation consensus motifs (26). Wooldridge and co-authors (25) provided evidence that PKAc could phosphorylate MYPT1 at Ser⁶⁹⁶ in vitro and disinhibit MLCP in ileal smooth muscle by preventing phosphorylation at Thr⁶⁹⁷. Similar results have been described for gastric smooth muscle cells (27) and rabbit femoral artery smooth muscle (28).

FIGURE 1.

Amino acid sequences surrounding the phosphorylation sites in MYPT1. MYPT1 contains four principal phosphorylation sites located in highly conserved regions. In rat numbering, the phosphorylation sites are: Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵, highlighted in bold italics. The human and chicken sequences are shown for comparison and to indicate the slightly different locations of the phosphorylation sites within these sequences. GenBank^TM ID numbers are as follows: human, 219842212; rat, 41017251; chicken, 45384106. Sequence numbering includes the initiating Met residues. These species were selected because these are the most common sources of MYPT1 for experimental purposes. Rat sequence numbering is used throughout this paper.

Although various studies have recently investigated the effects of PKAc and PKG on MYPT1 phosphorylation and Ca²⁺ desensitization (17, 25, 27, 28), it is still unclear from these reports if: (i) phosphorylation of MYPT1 at Ser⁸⁵⁴ occurs in smooth muscle; (ii) phosphorylation of Ser⁸⁵⁴ can prevent Thr⁸⁵⁵ phosphorylation; and (iii) Ser⁸⁵⁴–Thr⁸⁵⁵ dual phosphorylation occurs in tissue and has any functional effect on smooth muscle contraction. In this study, we provide a comprehensive validation of the specificity of a panel of phosphospecific antibodies to enable the investigation of MYPT1 phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵. The data presented demonstrate the ability of PKAc to phosphorylate MYPT1 in vitro at all four sites: Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵. Furthermore, phosphorylation at Ser⁶⁹⁶ and Ser⁸⁵⁴ prevents subsequent phosphorylation at Thr⁶⁹⁷ and Thr⁸⁵⁵, respectively. In rat caudal arterial smooth muscle, phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ was induced by application of the phosphatase inhibitor microcystin to demembranated tissues or of the adenylyl cyclase agonist forskolin to intact tissues. These dual phosphorylation events were associated with disinhibition of Thr⁶⁹⁷ and Thr⁸⁵⁵ phosphorylation as well as smooth muscle relaxation.

EXPERIMENTAL PROCEDURES

Materials

Microcystin LR was purchased from Alexis Biochemicals (San Diego, CA), [γ-³²P]ATP was from ICN Biomedical Inc. (Aurora, OH) and antibodies to LC₂₀ and actin were from Santa Cruz Biotechnology (Santa Cruz, CA) and Cell Signaling (Danvers, MA), respectively. Anti-rabbit IgG coupled to horseradish peroxidase (HRP) was purchased from Chemicon (Temecula, CA), PreScission Protease and the Enhanced Chemiluminescence Kit were from GE Healthcare (Piscataway, NJ), One Shot BL21(DE3)pLysS chemically competent Escherichia coli cells from Invitrogen and the QuikChange Site-directed Mutagenesis Kit from Stratagene (La Jolla, CA). LC₂₀ was purified from chicken gizzard as previously described (5). The full-length clone of chicken MYPT1 (FL-MYPT1; PPP1R12A, NP_990454.1) was a gift from Dr. David Hartshorne (University of Arizona). PKAc was purified from bovine heart as previously described (29). All other chemicals were reagent grade unless otherwise indicated and were obtained from Sigma or VWR (Mississauga, ON, Canada).

Phosphospecific MYPT1 Antibodies

Antibodies specific for MYPT1 phosphorylated at Thr⁶⁹⁷ or Thr⁸⁵⁵ were purchased from EMD Millipore (Billerica, MA). Anti-[phospho-Ser⁶⁹⁶]MYPT1 was obtained from Santa Cruz Biotechnology. Polyclonal antibodies for dual phosphorylations, anti-[phospho-Ser⁶⁹⁶–Thr⁶⁹⁷]MYPT1 and anti-[phospho-Ser⁸⁵⁴–Thr⁸⁵⁵]MYPT1 were raised in rabbits by injection of synthetic peptides coupled to keyhole limpet hemocyanin (SRRpSpTQ GVTL and PREKRRpSpTGVSFWTQDSD, respectively). A pan-MYPT1 polyclonal antibody recognizing all forms of MYPT1, phosphorylated and unphosphorylated, was also used. Peptides were synthesized by Genemed Synthesis (San Antonio, TX) and confirmed by MALDI-TOF-MS. In addition, a polyclonal antibody recognizing MYPT1 phosphorylated at Ser⁶⁹⁶ and/or Ser⁸⁵⁴ (anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1) was produced by New England Peptide (Gardner, MA) following injection of rabbits with the synthetic peptide SRRpSTQGVTL. All antibodies were affinity purified as previously described (30).

Generation and Expression of MYPT1 Mutants

Glutathione S-transferase (GST)-MYPT1 (residues 667–1004 of the chicken protein) was amplified by PCR as previously described (8). Single residue mutations (S696A, T697A, S854A, and T855A) and dual residue mutations (S696A/T697A, S854A/T855A, T697A/T855A, and S696A/S854A) were generated in the base sequence of GST-MYPT1(667–1004) using the QuikChange Site-directed Mutagenesis Kit according to the manufacturer's instructions. MYPT1 proteins were expressed in E. coli BL21(DE3)pLysS and purified with glutathione-Sepharose as previously described (8). The GST moiety was cleaved from the recombinant proteins with PreScission Protease.

Circular Dichroism

CD spectra were recorded with a Jasco J-810 spectropolarimeter (Jasco, Inc., Easton, MD) to determine conservation of secondary structure of truncated MYPT1 mutant proteins. Samples consisted of 300 μl of protein (at 5 μm in 10 mm Tris, pH 7.0), and spectra were recorded at room temperature in a 1-mm path-length cuvette. Far-UV spectra were collected between 260 and 190 nm using a 0.5-nm step size and a scanning speed of 50 nm/min. The bandwidth was set to 1 nm and the response time to 0.5 s. A blank spectrum, lacking protein, was subtracted in the final analysis. A final spectrum was obtained by accumulating the average of ten replicate scans.

Kinase Assays for the Phosphorylation of MYPT1

All kinase assays (whether including PKAc, PKG, or ROK) were initiated by addition to MYPT1 substrate and protein kinase of 5× ATP solution (1.5 mm ATP, 10 μCi/μl of [γ-³²P]ATP (only for radioactive assays), 5 mm MgCl_2, 125 mm HEPES, pH 7.4). For sequential kinase assays, i.e. phosphorylation of the substrate by a second kinase after an initial incubation with a primary kinase, the 5× ATP solution was replenished upon addition of the second kinase. Reactions were terminated by several distinct methods, depending on the subsequent use of the reaction components. If the product was to be analyzed by SDS-PAGE and Western blotting, the reaction was terminated by addition of 4× SDS-PAGE loading buffer (140 mm Tris, 55% glycerol (v/v), 70 mm SDS, 0.01% bromphenol blue (w/v)). To determine the stoichiometry of ³²P incorporation (as described previously (31)), reactions were terminated by the addition of 1 ml of 25% TCA (w/v) and substrates were precipitated in the presence of 25 μl of 25 mg/ml of BSA.

Myosin Phosphatase Assays

Phosphatase assays were performed to assess the effects of MYPT1 phosphorylation on the catalytic activity of the PP1cδ-MYPT1 complex. MLCP assays were carried out with ³²P-labeled LC₂₀ substrate as previously described (32). Phosphatase assays consisted of PP1cδ (6.25 nmol), FL-MYPT1 (10 nmol), and 5× phosphatase reaction buffer (25 mm Tris, 1 mm DTT, 1 mm EDTA) in a 50 μl volume. Reactions were initiated by addition of ³²P-LC₂₀ (40 μm). For termination, 130 μl of ice-cold 25% (w/v) TCA and 20 μl of 20 mg/ml of BSA were added. The mixture was vortexed and incubated on ice (15 min) to facilitate protein precipitation prior to centrifugation (13,000 × g, 15 min). Following centrifugation, the [³²P]phosphate was measured in the supernatant by scintillation counting. From this value the moles of phosphate released were calculated.

Tissue Preparation and Force Measurements

Caudal arteries were removed from male Sprague-Dawley rats (300–350 g) that had been anesthetized and euthanized according to protocols approved by the University of Calgary Animal Care and Use Committee. The arteries were cleaned of excess adventitia, de-endothelialized, and cut into helical strips (1.5 × 6 mm). Muscle strips were mounted on a Grass isometric force transducer (FT03C), and force was recorded. Intact tissues were treated with U46619 (1 μm; synthetic thromboxane A₂ receptor agonist) and/or forskolin (10 μm; adenylyl cyclase activator). Some experiments utilized Triton X-100-demembranated (skinned) arterial strips for the examination of Ca²⁺-free, microcystin-induced contractions, as previously described (33). The Ca²⁺-free solution (pCa 9) contained: 4 mm K₂EGTA, 5.83 mm MgCl₂, 75.6 mm potassium propionate, 3.9 mm Na₂ATP, 16.2 mm phosphocreatine, 30 units/ml of creatine kinase and 20 mm TES, pH 6.9. Ca²⁺-induced contractions were elicited with pCa 4.5 solution. At selected sampling points, muscle strips were flash-frozen in 10% (w/v) TCA, 10 mm DTT in acetone followed by 3 × 10-s washes in 10 mm DTT in acetone. Tissues were then lyophilized overnight. Protein was extracted from each arterial strip by incubation (16 h, 5 °C) in 0.5 ml of SDS-PAGE sample buffer.

Western Blot Analysis

For the analysis of MYPT1 phosphorylation, samples of in vitro kinase assays or tissue homogenates were resolved by 10% SDS-PAGE. Proteins were transferred to 0.2-μm nitrocellulose membranes in a Tris glycine transfer buffer containing 10% methanol. Nonspecific binding sites were blocked with 5% (w/v) I-Blok in TBST (25 mm Tris, 150 mm NaCl, 2.7 mm KCl, 0.05% Tween 20). Membranes were washed and incubated overnight (5 °C) with primary antibody at 1:1,000 dilution in 1% (w/v) nonfat dry milk in TBST. Membranes were incubated for 1 h with HRP-conjugated secondary antibody (1:10,000 dilution) in TBST and developed with SuperSignal West Femto Chemiluminescence reagent. Pan-MYPT1 or α-actin levels were quantified to ensure equal protein loading and to normalize the signal obtained with phospho-MYPT1 antibodies.

For analysis of LC₂₀ phosphorylation, samples were resolved by Phos-tag SDS-PAGE, as previously described (8). Proteins were transferred to 0.2 μm polyvinylidene difluoride membranes at 25 volts for 16 h at 4 °C and fixed on the membrane with 0.5% glutaraldehyde in phosphate-buffered saline. Nonspecific binding sites were blocked with 5% (w/v) nonfat dry milk in TBST. Membranes were washed with TBST and incubated overnight with anti-LC₂₀ at 1:500 dilution in 1% (w/v) nonfat dry milk in TBST. Membranes were incubated for 1 h with HRP-conjugated secondary antibody (1:10,000 dilution) and developed with ECL reagent.

All Western blots were visualized with an LAS4000 Imaging Station (GE Healthcare), ensuring that the representative signal occurred in the linear range. Quantification was performed by densitometry with ImageQuant TL software (GE Healthcare).

HPLC Fractionation and Phosphoamino Acid Analysis of MYPT1 Phosphopeptides

Phosphorylated MYPT1 (3 μg) was subjected to trypsin digestion (0.5 μg in 50 mm ammonium bicarbonate, 100 μl) overnight at 37 °C. For fractionation of ³²P-MYPT1 phosphopeptides by high-performance liquid chromatography (HPLC), tryptic digests were acidified with 0.5% (v/v) trifluoroacetic acid (TFA) and applied to a reversed-phase C18 column (Acclaim 300, 3 μm, 4.6 × 250 mm) equilibrated in 0.1% TFA. The column was washed with 0.1% (v/v) TFA and peptides were eluted with a linear gradient of acetonitrile (0–60% (v/v) over 100 min, 0.5 ml/min). Fractions (0.5 ml each) were collected and phosphopeptides were identified by scintillation counting. Fractions containing the major peaks of radioactivity were subjected to phosphoamino acid analysis. Lyophilized tryptic peptides were resuspended in 6 m HCl (250 μl), incubated for 2 h at 110 °C, dried, and resuspended in 5–10 μl of electrophoresis buffer (water:acetic acid:pyridine, 94.875:5:0.125) containing phospho-Ser, phospho-Thr, and phospho-Tyr standards (40 ng of each). Samples were spotted onto cellulose thin-layer plates (10 × 20 cm). Amino acids were electrophoresed at 1200 volts at 10 °C in electrophoresis buffer for 35 min. The plates were dried for 10 min in a fume hood, baked at 110 °C for 10 min, and amino acid standards were visualized by spraying the plates with 0.25% (v/v) ninhydrin dissolved in acetone and baking at 110 °C for 15 min. ³²P-Labeled amino acids were visualized by autoradiography.

Data Analysis

Values are presented as the mean ± S.E., with n indicating the number of independent experiments. Data were analyzed by one-way ANOVA followed by Bonferroni's or Dunnett's post hoc analysis, with p < 0.05 considered to indicate statistically significant differences.

RESULTS

Recombinant MYPT1 Protein Generation

A truncated version of chicken MYPT1, amino acids 667–1004 containing the phosphorylation sites of interest, was generated. This truncation not only improved the stability of the recombinant protein but also increased overall expression levels in the bacterial system. Site-directed mutation was performed at the phosphorylation sites of interest, and the corresponding proteins were expressed and purified (supplemental Fig. S1A). To confirm that the introduction of the mutations into the MYPT1 sequence did not induce major conformational changes in protein secondary structure, circular dichroism (CD) was performed (supplemental Fig. S1B). The data suggest that several of the introduced mutations caused slight variations in secondary structure, primarily the S854A and S696A/S854A proteins. Both proteins demonstrated a loss of α-helical structure with a comparable increase in random coil. Due to issues with expression, only the WT version of the full-length MYPT1 protein could be obtained in sufficient quantities for experiments (supplemental Fig. S1A). The phosphorylation-null mutants were designed and transformed into bacteria; however, the expression and purification of the various mutants proved to be problematic in a variety of bacterial systems, including BL21(DE3)pLysS, Rosetta-gami-B(DE3)pLysS, and Origami2(DE3)pLysS.

Biochemical Analysis of in Vitro MYPT1 Phosphorylation by PKAc

MYPT1 can be phosphorylated by PKAc/PKG at Ser⁶⁹⁶ in vitro and in situ (25, 28). Moreover, it has been suggested that PKAc/PKG may phosphorylate MYPT1 at Ser⁸⁵⁴; this is supported by the distinct similarities in sequence surrounding the sites (25) (Fig. 1). MYPT1(667–1004) phosphorylation by PKAc was assessed with radioactive [³²P]ATP kinase assays. PKAc-catalyzed phosphorylation of WT-MYPT1 increased with time, reaching a plateau between 60 and 90 min under the assay conditions used (1:50 molar ratio of PKAc to MYPT1) (Fig. 2A). The maximum stoichiometry achieved was 4.2 ± 0.6 mol of phosphate/mol of MYPT1, consistent with four phosphorylation sites. Tryptic peptides from ³²P-labeled WT and mutant MYPT1 species were separated by reversed-phase HPLC. Notably, two discrete elutions of ³²P (Fig. 2B, labeled Peak 1 and Peak 2) were obtained in each case, with the corresponding peaks originating from S696A/S854A eluting later than the other MYPT1 variants. This delay in elution may have resulted from the introduction of two hydrophobic residues in place of polar residues or from the difference in secondary structure indicated by CD spectroscopy (supplemental Fig. S1B).

FIGURE 2.

In vitro phosphorylation of MYPT1 by PKAc. A, wild-type (WT) chicken MYPT1 (residues 667–1004, 25 μg) was incubated with the catalytic subunit of cAMP-dependent protein kinase (PKAc; 0.1 μg) and Mg[γ-³²P]ATP as described under “Experimental Procedures.” Samples of the reaction mixture were withdrawn at the indicated times for SDS-PAGE and Coomassie Blue (CBB) staining, autoradiography, and quantification of ³²P incorporation into MYPT1. Values are mean ± S.E. (n = 4). B, wild-type and mutant MYPT1 proteins, phosphorylated for 60 min as in A, were digested with trypsin and the resultant phosphopeptides separated by reverse-phase HPLC. C, phosphoamino acid analysis (autoradiogram) of Peak 1 and Peak 2 peptides from B. The positions of phosphoserine (S) and phosphothreonine (T) standards are shown. D, phosphopeptide mapping (two-dimensional TLC) of Peak 2 [³²P]phosphopeptides from B. The locations of [³²P]phosphopeptide spots detected by autoradiography are indicated by dashed boxes.

To identify the amino acids phosphorylated by PKAc, [³²P]phosphoamino acid analysis was performed by thin-layer chromatography (TLC) (Fig. 2C). Peak 1 from Fig. 2B demonstrated Ser phosphorylation in all of the variants, including the S696A/S854A protein, suggesting that Peak 1 contained an additional Ser phosphorylation site. It has been suggested that PKAc can phosphorylate Ser⁶⁹³ (25). The ³²P associated with Peak 1 could, therefore, represent Ser⁶⁹³ phosphorylation as there are two tryptic cleavage sites between Ser⁶⁹³ and Ser⁶⁹⁶, which could discriminate the Ser⁶⁹³ phosphopeptide from the Ser⁶⁹⁶/Thr⁶⁹⁷ phosphopeptide. When analyzing the elution profiles for Peak 2, WT-MYPT1 was associated with both Ser and Thr phosphorylation (Fig. 2C). S696A/T697A and S854A/T855A also showed both Ser and Thr phosphorylation; however, S696A/S854A showed only Thr phosphorylation and T697A/T855A showed only Ser phosphorylation. Therefore, the [³²P]phosphoamino acid analysis-TLC results suggest that PKAc phosphorylates all four sites, i.e. Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵, in vitro.

During the chromatographic separation of ³²P-labeled tryptic peptides, we observed an apparent co-elution of phosphopeptides originating from both Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ regions. To confirm the co-elution of ³²P-labeled tryptic peptides containing Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵, digests of WT, S696A/T697A, and S854A/T855A proteins were separated by reverse-phase HPLC and then two-dimensional TLC. Again, two distinct peaks were obtained from the HPLC fractionation. Peak 2 was chosen and subjected to two-dimensional TLC, and the results showed two distinct ³²P-labeled peptides present in the WT fraction (Fig. 2D). However, only a single ³²P-labeled peptide was observed by two-dimensional TLC for the S696A/T697A as well as the S854A/T855A proteins.

Phosphospecific Antibody Validation and Confirmation of Phosphorylation at Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵

Several antibodies for phospho-MYPT1 are available from commercial sources, i.e. anti-[phospho-Ser⁶⁹⁶]MYPT1, anti-[phospho-Thr⁶⁹⁷]MYPT1, and anti-[phospho-Thr⁸⁵⁵]MYPT1. Although these reagents are widely used to examine the phosphorylation status of MYPT1, there has yet to be a comprehensive analysis of their specificity toward the Ser and Thr residues in both inhibitory regions. Furthermore, we have generated several additional polyclonal antibodies to examine MYPT1 phosphorylation status: anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1, anti-[phospho-Ser⁶⁹⁶–Thr⁶⁹⁷]MYPT1, and anti-[phospho-Ser⁸⁵⁴–Thr⁸⁵⁵]MYPT1. To assess antibody specificities, MYPT1 was phosphorylated by PKAc at Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵ as described above. As depicted in Fig. 3A, anti-[phospho-Ser⁶⁹⁶]MYPT1 antibody did not cross-react with phospho-Thr⁶⁹⁷, phospho-Ser⁸⁵⁴, or phospho-Thr⁸⁵⁵. There was no cross-reaction of anti-[phospho-Thr⁶⁹⁷]MYPT1 antibody with phospho-Ser⁶⁹⁶, phospho-Ser⁸⁵⁴, or phospho-Thr⁸⁵⁵ (Fig. 3B). Anti-[phospho-Ser⁶⁹⁶–Thr⁶⁹⁷]MYPT1 antibody did not cross-react with any phosphorylated MYPT1 species that did not contain both phospho-Ser⁶⁹⁶ and phospho-Thr⁶⁹⁷ (Fig. 3C). Moreover, anti-[phospho-Thr⁸⁵⁵]MYPT1 antibody showed little or no cross-reactivity with phospho-Ser⁶⁹⁶, phospho-Ser⁸⁵⁴, or phospho-Thr⁶⁹⁷ (Fig. 3D). It is also important to note that the antibodies against the monophosphorylated residues (Ser or Thr) also detected these individual phosphorylated residues in the dual phosphorylated state. During a phosphorylation time course, the monophosphorylation signals did not decrease as the dual phosphorylation signals increased (Fig. 4A). Thus, anti-[phospho-Thr⁶⁹⁷]MYPT1 detected both monophosphorylated Thr⁶⁹⁷ and dual phosphorylated Ser⁶⁹⁶–Thr⁶⁹⁷. None of the phosphospecific antibodies cross-reacted with unphosphorylated MYPT1 (Fig. 3, A–F).

FIGURE 3.

Characterization of phosphospecific MYPT1 antibodies and in vitro analysis of PKAc-catalyzed phosphorylation of MYPT1. Truncated WT and mutant MYPT1 proteins (3.0 μg) were phosphorylated with PKAc (60 ng) for 60 min, subjected to SDS-PAGE, and transferred to PVDF membranes for Western blotting. Phosphorylation-null MYPT1 mutants (Ser/Thr to Ala mutations) were used to determine antibody specificity and identify in vitro phosphorylation sites. Control reactions were carried out in the absence of PKAc to provide unphosphorylated WT-MYPT1. Signals from Western blots were confirmed to lie within the linear range of intensity and then quantified and the results are graphically expressed below the representative Western blots using the indicated phosphospecific antibodies. Values are mean ± S.E. (n = 3).

FIGURE 4.

Phosphorylation of MYPT1 at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ by PKAc in vitro is a sequential process with neither Ser nor Thr being the preferred initial target. The time course of phosphorylation of truncated WT-MYPT1 (3.0 μg) by PKAc (60 ng) was analyzed by Western blotting with the indicated phosphospecific antibodies and with pan-MYPT1 antibody used to verify comparable loading levels. A, representative Western blots of the phosphorylation time course are provided for each antibody. B and C, quantification of the time courses and rates of Ser⁶⁹⁶ (□), Thr⁶⁹⁷ (△), and dual Ser⁶⁹⁶–Thr⁶⁹⁷ (○) phosphorylation (B) as well as Thr⁸⁵⁵ (▴) and dual Ser⁸⁵⁴–Thr⁸⁵⁵ (●) phosphorylation (C). *, significant difference from the Ser⁶⁹⁶ phosphorylation rate (Student's t-test, p < 0.05).

Initially, the anti-[phospho-Ser⁸⁵⁴–Thr⁸⁵⁵]MYPT1 antibody showed a noticeable cross-reaction with the monophosphorylated, phospho-Thr⁸⁵⁵ protein. This reactivity toward phospho-Thr⁸⁵⁵ was successfully removed with an affinity resin generated by covalently coupling phospho-Thr⁶⁹⁷–Thr⁸⁵⁵ WT-MYPT1 protein (generated by phosphorylation with ROK) to activated-CH-Sepharose (supplemental Fig. S2). The resulting antibody showed no cross-reactivity with phospho-Thr⁸⁵⁵ or dual [phospho-Ser⁶⁹⁶–Thr⁶⁹⁷]MYPT1 (Fig. 3E).

In multiple attempts to generate a specific anti-[phospho-Ser⁸⁵⁴]MYPT1 antibody, we identified noticeable and specific cross-reaction with monophosphorylated, phospho-Ser⁶⁹⁶ MYPT1. This cross-reactivity could not be removed by additional affinity purification, so the problem seems to result from the ability of this antibody to recognize either phosphorylation site, Ser⁶⁹⁶ or Ser⁸⁵⁴. Therefore, this antibody was classified as an anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1 and was equally active toward either monophosphorylated Ser residue (Fig. 3F). In conclusion, the Western blot analysis data support the biochemical results, i.e. PKAc phosphorylates MYPT1 at Ser⁶⁹⁶, Thr⁶⁹⁷, Ser⁸⁵⁴, and Thr⁸⁵⁵. But perhaps most importantly, these data provide the first validation of a suite of phosphospecific MYPT1 antibodies available for the explicit interrogation of several MYPT1 phosphorylation sites.

The Rates of Phosphorylation of MYPT1 by PKAc in the Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ Regions

To determine whether the observed in vitro phosphorylation of the Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ regions of MYPT1 occurred sequentially, a series of time course kinase assays was carried out. The underlying question was whether the Ser residues were preferential targets over the Thr residues. The FL-MYPT1 protein was incubated with PKAc for 0–60 min, and phosphorylation sites were visualized by Western blotting (Fig. 4A). The initial velocities were determined (Fig. 4, B and C), and the events suggest a sequential phosphorylation process. The rates of phosphorylation for the various residues were: Ser⁶⁹⁶, 0.41 ± 0.05 response units/min; Thr⁶⁹⁷, 0.19 ± 0.06 response units/min; Ser⁶⁹⁶–Thr⁶⁹⁷, 0.17 ± 0.06 response units/min; Thr⁸⁵⁵, 0.31 ± 0.13 response units/min; and Ser⁸⁵⁴–Thr⁸⁵⁵, 0.21 ± 0.07. We could not analyze the velocity of Ser⁸⁵⁴ phosphorylation due to the lack of this phosphospecific antibody; however, Ser⁶⁹⁶ phosphorylation occurred more rapidly than Thr⁶⁹⁷ or dual Ser⁶⁹⁶–Thr⁶⁹⁷ phosphorylation. The accumulation of dual phosphorylation (either Ser⁶⁹⁶–Thr⁶⁹⁷ or Ser⁸⁵⁴–Thr⁸⁵⁵) appeared dependent upon the relative rate of Thr phosphorylation.

Additional in vitro kinase assays were carried out to examine the effect of Thr phosphorylation on the subsequent phosphorylation of the neighboring Ser residue. As previously demonstrated, PKAc phosphorylation of WT-MYPT1 at Ser⁶⁹⁶ reached saturation at around 60 min (Fig. 5A, panel i). However, when WT-MYPT1 was preincubated with ROK, resulting in maximal Thr⁶⁹⁷ phosphorylation (Fig. 5A, panel ii), the ability of PKAc to phosphorylate Ser⁶⁹⁶ was abolished (Fig. 5A, panel iii). A similar experiment was performed to assess the Ser⁸⁵⁴–Thr⁸⁵⁵ region. However, because the phospho-Ser⁸⁵⁴ antibody was not available, the S696A/T697A protein was used as substrate with the anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1 antibody to visualize the Ser⁸⁵⁴ phosphorylation status (Fig. 5B, panel i). As previously observed for the Ser⁶⁹⁶ site, pre-treatment with ROK, resulting in Thr⁸⁵⁵ phosphorylation (Fig. 5B, panel ii), completely inhibited the ability of PKAc to phosphorylate the neighboring Ser⁸⁵⁴ residue (Fig. 5B, panel iii).

FIGURE 5.

In vitro analysis of MYPT1 phosphorylation by PKAc with or without pre-incubation with ROK. A, time courses of phosphorylation by PKAc of truncated WT-MYPT1 at Ser⁶⁹⁶ and Thr⁶⁹⁷ with and without pre-incubation with ROK for 60 min. Panel i shows representative Western blots and panels ii and iii provide cumulative data for the quantification of Thr⁶⁹⁷ and Ser⁶⁹⁶ phosphorylation, respectively. B, time courses of phosphorylation by PKAc of truncated S696A/T697A-MYPT1 at Ser⁸⁵⁴ and Thr⁸⁵⁵ with and without pre-incubation with ROK for 60 min. Panel i shows representative Western blots and panels ii and iii provide cumulative data for the quantification of Thr⁸⁵⁵ and Ser⁸⁵⁴ phosphorylation, respectively. Values are mean ± S.E. (n = 3). *, significant difference between PKAc and ROK/PKAc treatments (ANOVA with Dunnett's post hoc test; p < 0.05). All phospho-MYPT1 signals were normalized to pan-MYPT1 Western blots (loading controls).

The Effect of MYPT1 Phosphorylation at Specific Sites on Myosin Phosphatase Activity

The effect of site-specific MYPT1 phosphorylation on myosin phosphatase activity was examined using ³²P-LC₂₀ as substrate. Assays were performed under three conditions: unphosphorylated FL-MYPT1, phospho-Thr⁶⁹⁷/Thr⁸⁵⁵ FL-MYPT1, and phospho-Ser⁶⁹⁶–Thr⁶⁹⁷/Ser⁸⁵⁴–Thr⁸⁵⁵ FL-MYPT1. To obtain the phosphorylated MYPT1 species, FL-MYPT1 was incubated with kinase (ROK for Thr-specific phosphorylation, PKAc for phosphorylation at Ser and Thr) prior to complex formation with PP1cδ. Consistent with previous results, phosphorylation at Thr⁶⁹⁷ (16, 25) and Thr⁸⁵⁵ (16) by ROK resulted in a 65% decrease in phosphatase activity (Fig. 6A). The rate of LC₂₀ dephosphorylation was reduced from 22.3 ± 8.8 nmol/min (control; no MYPT1 phosphorylation) to 5.9 ± 1.2 nmol/min (ROK; Thr⁶⁹⁷ and Thr⁸⁵⁵ phosphorylation). However, dual phosphorylation at both regions (Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵) had no effect on MLCP activity (22.8 ± 7.6 nmol/min) when compared with control (unphosphorylated) MYPT1.

FIGURE 6.

The effect of dual Ser/Thr phosphorylation of FL-MYPT1 on myosin phosphatase activity. A, myosin phosphatase (FL-MYPT1 + PP1cδ) activity was assayed over a 30-min time course. Initially, MYPT1 (10 nmol) was incubated in the absence of kinases (■; nonphosphorylated MYPT1 control), in the presence of ROK (●, phospho-Thr⁶⁹⁷/Thr⁸⁵⁵) or in the presence of PKAc (▴; phospho-Ser⁶⁹⁶–Thr⁶⁹⁷/Ser⁸⁵⁴–Thr⁸⁵⁵). Purified PP1cδ (6.25 nmol) was added to form the MLCP complex, and the phosphatase activity toward the ³²P-labeled LC₂₀ was determined. Values are mean ± S.E. (n = 3). a, significant difference from control; b, significant difference from PKAc-treated (ANOVA with Dunnett's post hoc test, p < 0.05). B, Western blot analyses of FL-MYPT1 phosphorylation pre- and post-assay (i.e. before and after incubation with PP1cδ). Phosphospecific antibodies were used to examine monophosphorylation at Thr⁶⁹⁷ and Thr⁸⁵⁵ as well as dual phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ (panel i). Cumulative quantitative data are presented in panel ii with phosphorylation levels normalized to pan-MYPT1 as a loading control. Results are expressed relative to the initial FL-MYPT1 phosphorylation measured in pre-assay samples. Values are mean ± S.E. (n = 4). *, significantly different from pre-assay value (Student's t test, p < 0.05).

To examine the effect of PP1cδ on the phosphorylation status of FL-MYPT1 during the MLCP assay, Western blotting was performed on samples collected pre- and post-assay (Fig. 6B, panel i). There was no significant effect of PP1cδ incubation on the levels of dual phosphorylation at either Ser⁶⁹⁶–Thr⁶⁹⁷ or Ser⁸⁵⁴–Thr⁸⁵⁵. However, significant decreases in the levels of phosphorylation at Thr⁶⁹⁷ and Thr⁸⁵⁵ were detected in the absence of phosphorylation at the neighboring Ser residues. Thr⁶⁹⁷ phosphorylation levels were decreased by ∼50%, and Thr⁸⁵⁵ phosphorylation levels by ∼20% (Fig. 6B, panel ii).

Analysis of MYPT1 Phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ in Triton-skinned Rat Caudal Artery following Microcystin Stimulation

Treatment of rat caudal arterial smooth muscle strips with the phosphatase inhibitor microcystin unmasks the activity of Ca²⁺-independent LC₂₀ kinases, ILK and ZIPK, and induces contraction (8, 33, 34). This effect replicates the original reports of microcystin administration on contractile force (35, 36). Typical force and LC₂₀ phosphorylation responses to microcystin administration are depicted in Fig. 7, A and B, respectively. As depicted in Fig. 7C, monophosphorylation of LC₂₀ is associated with the initial phase of the microcystin-induced contraction, whereas diphosphorylation becomes apparent during the sustained phase of the contraction. Western blots performed with the anti-[phospho-Ser⁶⁹⁶–Thr⁶⁹⁷]MYPT1 antibody showed consistent basal levels of dual phosphorylation at these two sites (Fig. 7, D and E), with a ∼2.5-fold increase after 60 min incubation with microcystin. On the other hand, the anti-[phospho-Ser⁸⁵⁴–Thr⁸⁵⁵]MYPT1 antibody revealed barely detectable levels of Ser⁸⁵⁴–Thr⁸⁵⁵ dual phosphorylation under resting conditions, which were significantly increased within 15 min of the application of microcystin and reached a 10-fold increase by 60 min (Fig. 7, D and E).

FIGURE 7.

Microcystin treatment of Triton-skinned rat caudal arterial smooth muscle induces dual phosphorylation of MYPT1 at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵. A, Triton-skinned smooth muscle strips were contracted with pCa 4.5 solution followed by relaxation in Ca²⁺-free (pCa 9) solution. Microcystin (MC; 10 μm) was added in pCa 9 solution and tissues were quick-frozen for analysis at the indicated times. B, LC₂₀ phosphorylation was analyzed by Phos-tag SDS-PAGE with detection of unphosphorylated, mono-, and diphosphorylated forms by Western blotting with anti-LC₂₀. C, bands were quantified by scanning densitometry. Data are expressed as percentages of total LC₂₀ for unphosphorylated (LC₂₀-0P; white bar), monophosphorylated (LC₂₀-1P; gray bar), diphosphorylated species (LC₂₀-2P; hatched bar), and phosphorylation stoichiometry (mol of P_i/mol of LC₂₀; black bar). D, dual phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ was determined by Western blotting with phosphospecific antibodies. To control for variations in loading levels, the amount of MYPT1 phosphorylation was normalized to actin levels. The level of MYPT1 dual phosphorylation in the pCa 9 control was set to 1. Values are mean ± S.E. (n = 5). *, significantly different from control (ANOVA with Dunnett's post hoc test, p < 0.05).

Analysis of MYPT1 Phosphorylation at Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ in Intact Rat Caudal Artery following Stimulation with U46619 and/or Forskolin

As a control for normalization of contractile force and to ensure tissue viability, intact rat caudal arterial smooth muscle strips were first treated with 87 mm KCl to induce Ca²⁺ entry via voltage-gated Ca²⁺ channels and trigger contraction via activation of MLCK. Following washout and relaxation, tissues were treated with vehicle (Fig. 8A, panel i), U46619 (1 μm; synthetic thromboxane A₂ receptor agonist known to elicit a ROK-dependent contraction (30)) (Fig. 8A, panel ii), forskolin (10 μm; adenylyl cyclase agonist) (Fig. 8A, panel iii), forskolin followed by U46619 (Fig. 8A, panel iv), or U46619 followed by forskolin (Fig. 8A, panel v). U46619-induced contractile force was ∼40% of that elicited by KCl (Fig. 8A, panels ii and vi). As expected, the addition of forskolin did not elicit a contractile response (Fig. 8A, panels iii and vi). However, pre-treatment with forskolin prevented U46619-induced contraction (Fig. 8A, panels iv and vi). Furthermore, U46619-induced contraction was completely reversed upon addition of forskolin (Fig. 8A, panels v and vi).

FIGURE 8.

Forskolin treatment of intact rat caudal arterial smooth muscle and its effect on U46619-induced phosphorylation of the Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ regions of MYPT1. A, representative contractile responses of intact rat caudal arterial smooth muscle strips to: (i) high-[K⁺] extracellular solution (KES; 87 mm KCl); (ii) thromboxane A₂ receptor agonist U46619 (1 μm); (iii) adenylyl cyclase activator forskolin (10 μm, 10 min); (iv) forskolin (10 μm, 10 min) followed by U46619 (1 μm, 10 min); and (v) U46619 (1 μm, 10 min) followed by forskolin (10 μm, 10 min). (vi) The force generated relative to the initial contraction with KES was calculated from the individual contractile profiles. Values are mean ± S.E. with the n value represented by the number beside each bar. *, significantly different from U46619 (ANOVA with Dunnett's post hoc test, p < 0.05). NES, normal extracellular solution. B, tissue strips were processed and examined for MYPT1 phosphorylation in the Ser⁶⁹⁶–Thr⁶⁹⁷ region. Phosphorylation levels of Ser⁶⁹⁶, Thr⁶⁹⁷, and Ser⁶⁹⁶–Thr⁶⁹⁷ were determined by Western blotting with phosphospecific antibodies. C, tissue strips were processed and examined for MYPT1 phosphorylation in the Ser⁸⁵⁴–Thr⁸⁵⁵ region. Phosphorylation levels of Ser⁶⁹⁶/Ser⁸⁵⁴, Thr⁸⁵⁵, and Ser⁸⁵⁴–Thr⁸⁵⁵ were determined by Western blotting with phosphospecific antibodies. Values in B and C are mean ± S.E. (n = 5). a, significantly different from control (NES, no treatment); b, significantly different from U46619 treatment; c, significantly different from forskolin treatment (ANOVA with Dunnett's post hoc test, p < 0.05).

The effects of U46619 and forskolin, alone or in combination, on site-specific MYPT1 phosphorylation were then examined (Fig. 8B). Forskolin induced a 4-fold increase in Ser⁶⁹⁶ phosphorylation over basal levels, whereas U46619 had no effect on Ser⁶⁹⁶ phosphorylation. Treatment with U46619 after forskolin did not affect the forskolin-induced increase in Ser⁶⁹⁶ phosphorylation. Consistent with the results obtained from in vitro sequential kinase assays (Fig. 5), pre-treatment with U46619 (which increased phosphorylation at Thr⁶⁹⁷ 2-fold and at Thr⁸⁵⁵ 3-fold) prevented subsequent Ser⁶⁹⁶ phosphorylation upon addition of forskolin. Forskolin pre-treatment prevented the U46619-induced Thr⁶⁹⁷ phosphorylation, a result that was previously observed in isolated rabbit ileal smooth muscle strips (25) and cultured gastric smooth muscle cells (27). Dual phosphorylation of Ser⁶⁹⁶–Thr⁶⁹⁷ was also significantly increased with forskolin pre-treatment (Fig. 8B). Therefore, with respect to the Ser⁶⁹⁶–Thr⁶⁹⁷ region and forskolin treatment, pre-phosphorylation at either Ser⁶⁹⁶ or Thr⁶⁹⁷ was able to suppress phosphorylation of the neighboring residue, and dual phosphorylation of Ser⁶⁹⁶–Thr⁶⁹⁷ was induced by forskolin treatment.

Phosphorylation within the Ser⁸⁵⁴–Thr⁸⁵⁵ region has not previously been investigated to the same degree as the Ser⁶⁹⁶–Thr⁶⁹⁷ region, possibly due to the lack of an effective phospho-Ser⁸⁵⁴ antibody. To examine MYPT1 phosphorylation within this region, we performed Western blots for phospho-Thr⁸⁵⁵, dual phospho-Ser⁸⁵⁴–Thr⁸⁵⁵, and phospho-Ser⁶⁹⁶/Ser⁸⁵⁴ (Fig. 8C). The anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1 antibody could not provide specificity for the Ser⁸⁵⁴ site; however, it was used in combination with the anti-[phospho-Ser⁸⁵⁴–Thr⁸⁵⁵]MYPT1 antibody to provide a better understanding of the phosphorylation dynamics of this region during contraction. When signals from Western blots performed with the anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1 antibody were analyzed, it was clear that the data resembled those obtained with the anti-[phospho-Ser⁶⁹⁶]MYPT1 antibody, i.e. elevated signals (∼2-fold) were observed with forskolin treatment alone and forskolin followed by U46619 treatment (Fig. 8B). However, it could not be determined whether the signals were entirely reflective of phospho-Ser⁸⁵⁴ because phospho-Ser⁶⁹⁶ was also demonstrated to increase ∼3.5-fold under these conditions (Fig. 8B). Analysis of Western blots probed with anti-[phospho-Thr⁸⁵⁵]MYPT1 antibody revealed a 3-fold increase in Thr⁸⁵⁵ phosphorylation following application of U46619 (Fig. 8C). This increase was prevented by forskolin pre-treatment, an effect mirroring that observed for Thr⁶⁹⁷. Forskolin treatment induced dual phosphorylation of Ser⁸⁵⁴–Thr⁸⁵⁵. The combination of forskolin and U46619, applied in either order, also induced dual phosphorylation of Ser⁸⁵⁴–Thr⁸⁵⁵. Therefore, we can conclude with respect to the Ser⁸⁵⁴–Thr⁸⁵⁵ region: U46619 treatment induced increases in Thr⁸⁵⁵ phosphorylation, pre-treatment with forskolin blocked the ability of U46619 to induce Thr⁸⁵⁵ phosphorylation, and dual Ser⁸⁵⁴–Thr⁸⁵⁵ phosphorylation was observed following administration of forskolin.

DISCUSSION

MLCP and its regulation by MYPT1 phosphorylation are essential to the coordinated control of smooth muscle contraction. The primary method for examination of MYPT1 phosphorylation has been to apply biochemical results from in vitro analyses to the production of phosphospecific antibodies for interrogation of tissue samples. To assess the in situ or in vivo regulation of MLCP, we must possess a reliable reporting system for the examination of MYPT1 phosphorylation as a foundation to test additional experimental hypotheses. Therefore, complete confidence in the specificity of the antibodies employed is critical to the reliability of data interpretation. Several phosphospecific MYPT1 antibodies are commercially available or have been generated by individual research labs. Although these resources have been used extensively by a variety of groups, the specificities have not been comprehensively examined by biochemical analysis of phosphorylation-null mutants, which may have lead to erroneous conclusions. For instance, a significant signal was detected when PKAc-treated MYPT1 was analyzed by Western blotting with an anti-[phospho-Thr⁶⁹⁷]MYPT1 antibody (25). This signal was attributed to antibody cross-reactivity without fully investigating the matter. Our data, however, clearly indicate that PKAc phosphorylates Thr⁶⁹⁷ and Thr⁸⁵⁵, in addition to Ser⁶⁹⁶ and Ser⁸⁵⁴, which has important implications regarding the regulation of MLCP activity by cyclic nucleotides and cross-talk with the ROK pathway. In addition, it is unclear if antibodies raised against monophosphorylation sites (e.g. anti-[phospho-Ser⁶⁹⁶]MYPT1, anti-[phospho-Thr⁶⁹⁷]MYPT1, and anti-[phospho-Thr⁸⁵⁵]MYPT1) recognize the phosphorylated residue when the neighboring Ser/Thr residue is also phosphorylated. Nakamura and colleagues (28) were the first to demonstrate dual phosphorylation of MYPT1 at Ser⁶⁹⁶–Thr⁶⁹⁷ by PKG in vitro and in situ. However, an additional consideration is whether the anti-[phospho-Ser⁶⁹⁶–Thr⁶⁹⁷]MYPT1 antibody used in their study was cross-reactive with the Ser⁸⁵⁴–Thr⁸⁵⁵ region.

With the use of phosphorylation-null MYPT1 mutants, we have characterized a library of phosphospecific antibodies for examination of MYPT1 phosphorylation events. Given our observations of significant antibody cross-reactivity in some instances, we would caution investigators to perform adequate quality control of phosphospecific MYPT1 antibodies prior to their use in functional studies. For example, our data suggest that the widely used anti-[phospho-Thr⁶⁹7]MYPT1 antibody reacts against the dual phosphorylated Ser⁶⁹⁶–Thr⁶⁹⁷ site. Furthermore, it is likely that the anti-[phospho-Thr⁸⁵⁵]MYPT1 antibody reacts against the dual phosphorylated Ser⁸⁵⁴–Thr⁸⁵⁵ site. However, it is difficult to reach any definitive conclusions for the later event given that we lack a specific antibody for phospho-Ser⁸⁵⁴ to complete an analysis of phosphorylation during a time course (as provided for the anti-[phospho-Thr⁶⁹⁷]MYPT1 antibody). In addition, the relative sensitivities of the phospho-Thr⁸⁵⁵ and dual phospho-Ser⁸⁵⁴–Thr⁸⁵⁵ antibodies may be quite different.

An obvious handicap of the phosphospecific MYPT1 antibody repertoire currently available to researchers is the lack of an anti-[phospho-Ser⁸⁵⁴]MYPT1 antibody. There is no commercial source for this reagent, and multiple attempts by our laboratories to develop this phosphospecific antibody have consistently resulted in a nonspecific product. The best result was represented in this study by the anti-[phospho-Ser⁶⁹⁶/Ser⁸⁵⁴]MYPT1 antibody. The antigen was a phospho-Ser⁸⁵⁴ peptide; however, the final antibody product cross-reacted equally with the phosphorylated Ser⁶⁹⁶ residue. Nevertheless, this phosphospecific antibody is useful because it can act as a reporter of total disinhibition status of MYPT1.

Based on our results, it is clear that PKAc (and PKG; data provided in supplemental Fig. S3) can direct the dual phosphorylation of both inhibitory Thr residues (697/855) and their preceding Ser residues (696/854) in vitro. Previous examinations of MYPT1 phosphorylation by PKAc and PKG suggested that these kinases could only phosphorylate Ser⁶⁹³, Ser⁶⁹⁶, and Ser⁸⁵⁴ (25). With our use of phosphorylation-null mutants and validated antibodies, it is clear that the phospho-Thr⁶⁹⁷ signal represents a true phosphorylation event. If the signal resulted from background cross-reactivity with the nonphosphorylated MYPT1 protein, one would expect to observe an equal signal in the nonphosphorylated WT-MYPT1 control as well as the phosphorylation-null mutants. It could be postulated that we detected Thr⁶⁹⁷/Thr⁸⁵⁵ phosphorylation with PKAc because the kinase concentration in the assays was high. However, additional kinase assays were performed at lower kinase concentrations (PKAc:MYPT1 molar ratios of 1:100, 1:200, and 1:1,000; supplemental Fig. S4) indicated Thr⁶⁹⁷ phosphorylation in all cases. Furthermore, as shown in Fig. 8, B and C, forskolin treatment of vascular smooth muscle tissue induced phosphorylation at Thr⁶⁹⁷ and Thr⁸⁵⁵ in addition to Ser⁶⁹⁶ and Ser⁸⁵⁴.

As PKAc and PKG are well known participants in smooth muscle relaxation (22, 24, 27), it is difficult to rationalize the ability of these kinases to direct the Thr⁶⁹⁷ and Thr⁸⁵⁵ phosphorylations that are associated with MLCP inhibition, Ca²⁺ sensitization, and enhanced contractile force. Although we observed dual phosphorylation of Ser/Thr residues during the sustained phase of microcystin-induced contraction of rat caudal arterial smooth muscle, it is unlikely that the dual phosphorylation events are involved in sustaining the tonic phase of contraction. Indeed, no increase in dual phospho-Ser⁶⁹⁶–Thr⁶⁹⁷ or dual phospho-Ser⁸⁵⁴–Thr⁸⁵⁵ was observed during Ca²⁺ sensitization of intact caudal artery induced by the thromboxane A₂ receptor agonist U46619 (Fig. 8, B and C). Our in vitro results suggest that dual phosphorylation of MYPT1 prevents the PP1c-dependent dephosphorylation of Thr residues, and dual phosphorylation events elicited by forskolin application prior to U46619 correlated with abrogation of contractile force. Taken together, these data support a role for dual Ser/Thr phosphorylation in modulating the binding of phosphorylated Thr⁶⁹⁷ and/or Thr⁸⁵⁵ residues by PP1cδ in the MLCP complex (17, 37). This may serve to restore MLCP activity to normal in smooth muscle tissue. Alternatively, dual phosphorylation may target MYPT1 to an additional, yet undefined phosphatase, as proposed by Nakamura and colleagues (28).

It has been reported that PKAc/PKG treatment of MYPT1 results in Ser⁶⁹⁶ phosphorylation that subsequently impedes the ability of ROK to phosphorylate Thr⁶⁹⁷ and elicit Ca²⁺ sensitization (25, 27, 28). We investigated whether the reverse effect was also possible, i.e. if phosphorylation of Thr⁶⁹⁷ or Thr⁸⁵⁵ could inhibit the subsequent phosphorylation of Ser⁶⁹⁶ or Ser⁸⁵⁴. Our in vitro results revealed that when Thr⁶⁹⁷ and Thr⁸⁵⁵ were phosphorylated to saturation (by ROK), PKAc was unable to phosphorylate the adjacent Ser residues. Additional data support a similar effect of ROK on subsequent PKG phosphorylations (supplemental Fig. S5). Moreover, analysis of MYPT1 phosphorylation events in caudal arterial tissue indicated that phosphorylation of Thr⁶⁹⁷ and Thr⁸⁵⁵ in response to U46619 prevented the phosphorylation of Ser⁶⁹⁶ and Ser⁸⁵⁴ upon application of forskolin. Steric hindrance or charge repulsion due to close proximity of multiple phosphate groups is a possibility, but perhaps not a likely explanation for this scenario because our in vitro work revealed the ability of PKAc to dual phosphorylate both Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵. Furthermore, other studies have demonstrated the ability of PKAc to phosphorylate neighboring Ser and Thr residues of other substrates, e.g. death-associated protein-3 (DAP3) at Ser¹⁸⁵–Thr¹⁸⁶ (38) and glial fibrillary acidic protein at Thr⁷–Ser⁸ (39)). We suggest, therefore, that the results reflect kinetic constraints on the protein kinase.

It is interesting that prior phosphorylation of Thr⁶⁹⁷–Thr⁸⁵⁵ inhibited the ability of PKAc to phosphorylate Ser⁶⁹⁶/Ser⁸⁵⁴ because the data also indicate the ability of PKAc to dual phosphorylate Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵. Consequently, the dual phosphorylation by PKAc must occur in a nonrandom sequential process (either a processive or distributive mechanism (40)) with phosphorylation of Ser preceding that of the neighboring Thr residue. In a processive mechanism, the kinase binds to the substrate and phosphorylates multiple residues before dissociating to enable phosphorylation of another substrate molecule. On the other hand, in a distributive process, the kinase binds to the substrate and phosphorylates a single residue before dissociating and then phosphorylating another substrate molecule (possibly another site on the same molecule, the same site on a different molecule, or a different site on a different molecule). In situ, it is likely that MYPT1 dual phosphorylation occurs as a distributive process with multiple kinases engaging the MYPT1 substrate, e.g. ZIPK, ROK, ILK, PKG, and PKAc. According to this scenario, however, the potential for PKAc/PKG-dependent amplification of the dual phospho-MYPT1 pool would ultimately depend upon the magnitude of Thr monophosphorylation. The dual phosphorylation results for Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ in caudal arterial smooth muscle appear to parallel the kinetics of the monophosphorylation of the Thr residues. This suggests that in smooth muscle tissues, Thr monophosphorylation events may act as the rate-limiting factor in the development of dual-phosphorylation of the Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ regions.

In summary: (i) dual phosphorylation of MYPT1 occurs in vitro and in situ at both Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵; (ii) pre-phosphorylation of Ser⁶⁹⁶ and Ser⁸⁵⁴ inhibits the ability of ROK to phosphorylate the neighboring Thr⁶⁹⁷ and Thr⁸⁵⁵, thereby preventing Ca²⁺ sensitization. This effect of Ser⁶⁹⁶ phosphorylation has been previously described (25, 27, 28); (iii) PKAc-dependent phosphorylation of Thr⁶⁹⁷ and/or Thr⁸⁵⁵ cannot occur in the absence of Ser⁶⁹⁶/Ser⁸⁵⁴ phosphorylation; and (iv) monophosphorylation of Thr⁶⁹⁷ and Thr⁸⁵⁵ precludes the ability of PKAc to phosphorylate the neighboring Ser residues, thereby preventing the disinhibition of MLCP activity. It appears that the regulation of MLCP activity is dependent on relative phosphorylation of inhibitory (Thr) and disinhibitory (Ser) sites in MYPT1, with phosphatase activity being dictated by the balance in phosphorylation status of the four sites. Basal levels of monophosphorylated Ser, monophosphorylated Thr, dual-phosphorylated Ser/Thr, and nonphosphorylated MYPT1 exist in both Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ regions. Relaxant stimuli induce an increase in Ser phosphorylation and/or dual-Ser/Thr phosphorylation, with enhancement of MLCP activity. The results also support a protective effect of monophosphorylated Thr⁶⁹⁷ or Thr⁸⁵⁵ against dual phosphorylation, ensuring its inhibitory potential such that any increase in Thr monophosphorylation inhibits MLCP in response to contractile stimuli. Therefore, a relationship between these events likely exists, with the relative gain of the system being dependent upon the global Ser and Thr phosphorylation status within the Ser⁶⁹⁶–Thr⁶⁹⁷ and Ser⁸⁵⁴–Thr⁸⁵⁵ regions.