A Novel Interaction of cGMP-dependent Protein Kinase I with Troponin T*

cGMP-dependent protein kinase (cGK) is a major intracellular receptor of cGMP and is implicated in several signal transduction pathways. To identify proteins that participate in the cGMP/cGK signaling pathway, we employed the yeast two-hybrid system with cGK Iα as bait. cDNAs encoding slow skeletal troponin T (skTnT) were isolated from both mouse embryo and human skeletal muscle cDNA libraries. The skTnT protein interacted with cGK Iβ but not with cGK II nor cAMP-dependent protein kinase. The yeast two-hybrid and in vitro binding assays revealed that the N-terminal region of cGK Iα, containing the leucine zipper motif, is sufficient for the association with skTnT. In vivoanalysis, mutations in cGK Iα, which disrupted the leucine zipper motif, were shown to completely abolish the binding to skTnT. Furthermore, cGK I also interacted with cardiac TnT (cTnT) but not with cardiac troponin I (cTnI). Together with the observations that cTnI is a good substrate for cGK I and is effectively phosphorylated in the presence of cTnT in vitro, these findings suggest that TnT functions as an anchoring protein for cGK I and that cGK I may participate in the regulation of muscle contraction through phosphorylation of TnI.

cGMP plays important roles in the regulation of smooth muscle relaxation, platelet aggregation, intestinal secretion, and endochondral ossification mediated through the activation of cGMP-dependent protein kinase (cGK) 1 (1)(2)(3)(4)(5)(6)(7). The increase in intracellular cGMP levels is consistent with the increased stimulation of guanylate cyclases that are acceptors for nitric oxide (NO) and natriuretic peptides. However, the physiological effects of cGMP through the activation of cGKs have been little studied because the identification of the physiological significance of cGKs is further complicated by the existence of two forms of cGK, cGK I and cGK II, which are encoded by distinct genes, and of two different isoforms of cGK I (designated I␣ and I␤), produced by alternative splicing (8). cGK I␣ and cGK I␤ were shown to be dimeric enzymes that contain several domains with well defined functions. The regulatory compartment located in the N-terminal re-gion contains the leucine zipper motif and autoinhibitory domain, followed by two cGMP binding and catalytic domains. Within the protein kinase superfamily, cGK is most closely related to cAMP-dependent protein kinase (cAK). cGK and cAK share similar structural properties (9), showing that cGK and cAK have many similar and overlapping consensus phosphorylation motifs and that substrates identified for cGK are often efficiently phosphorylated by cAK in vitro. In addition to that cGKs are localized to few tissues relative to the ubiquitous cAKs, the relatively low cytosolic concentration of cGK in most cells (10) suggests that mechanisms other than recognition of consensus phosphorylation motifs contribute to the selective cGMP-mediated protein phosphorylation in cells. Recent studies indicate that specific anchoring proteins located at various sites in the cell compartmentalize the kinase to sites of action and that the location of anchoring proteins provides some of the specificity of the responses mediated by each kinase (11,12). Indeed, cAK binds to specific anchoring proteins, which are termed AKAPs (for A kinase anchoring proteins), through the regulatory subunits (13,14). Binding of cAMP to cAK holoenzyme releases the catalytic subunits, enabling them to phosphorylate their substrates. Therefore, the localization of cAK near its substrates may ensure the rapid phosphorylation of specific substrates in response to increases in the intracellular concentration of cAMP. Other proteins reported to be localized in the cell with substrates include protein kinase C (PKC) (15) and Raf (16).
To assess the potential functions of cGK Is and to determine the mechanism by which these functions are carried out, we isolated components of the cGMP/cGK signaling pathway such as substrates and regulatory proteins of cGK using the yeast two-hybrid system. We report evidence that cGK Is directly interact with skeletal and cardiac troponin Ts (TnT) and phosphorylate cardiac troponin I (TnI) in vitro and in vivo. Our findings suggest that the interaction of cGK I with TnT is important for regulating the muscle contraction through the phosphorylation of TnI.

EXPERIMENTAL PROCEDURES
Materials-The MATCHMAKER II two-hybrid system and the mouse 17-day embryo and human skeletal muscle MATCHMAKER cDNA libraries were obtained from CLONTECH. Restriction endonucleases and DNA-modifying enzymes were purchased from Takara Shuzo (Kyoto, Japan). COS-7 cells were from Dainippon Pharmaceutical Co. (Osaka, Japan). Dulbecco's modified Eagle's medium and fetal calf serum were obtained from Life Technologies, Inc. [␥-32 P]ATP was a product of Amersham Pharmacia Biotech.
Plasmid Construction-The full-length bovine cGK I␣ cDNA was a gift from Dr. Thomas M. Lincoln (The University of Alabama at Birmingham). A cDNA encoding full-length cGK I␣ was subcloned into the BamHI (made blunt with T4 DNA polymerase) site of the yeast expression vector pAS2-1 (CLONTECH) fused in frame with the DNA binding domain of yeast transcriptional activator GAL4, which generated pAS2-1 cGK I␣. Deletion mutants of cGK I␣ were constructed by * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Yeast Two-hybrid Screening-The yeast two-hybrid screening was performed according to the instructions of the manufacturer. In brief, the yeast strain CG1945 was cotransformed with pAS2-1 cGK I␣ and the mouse 17-day embryo and human skeletal muscle cDNA libraries in pGAD10 using the lithium acetate method. Transformants were selected on synthetic dropout agar plates lacking tryptophan, leucine, and histidine but containing 5 mM 3-aminotriazole. Yeast colonies were transferred onto a nylon membrane and processed by the ␤-galactosidase filter assay. Plasmid DNAs were isolated from positive colonies and re-transformed into the yeast strain Y190 with either pAS2-1 empty vector or pAS2-1 cGK I␣. The ␤-galactosidase assay was again conducted to ensure that the activity was cGK I␣-dependent. The cDNA inserts from true positive clones were sequenced by dideoxy chain termination methods using an Applied Biosystems model 373A automated sequencer and a dye terminator cycle sequencing kit (PE Biosystems).

Production of Recombinant Proteins and in Vitro Binding Analysis-
The glutathione S-transferase (GST) expression vector pGEX-5X-3 (Amersham Pharmacia Biotech) and GST fusion protein construct pGEX-cGK I␣-(1-66) carrying cDNA encoding the bovine cGK I␣-(1-66) were introduced into the bacterial strain JM109 cells (TOYOBO). An overnight culture in LB medium was diluted 1:100 into 100 ml of fresh LB and incubated at 37°C in a shaking incubator for 2 h. Isopropyl-1-thio-␤-D-galactopyranoside was then added to the culture to a final concentration of 0.2 mM, and the whole was incubated for an additional 2 h. The cells were washed once with ice-cold soluble buffer (50 mM Tris-HCl at pH 8.0 and 1 mM EDTA) and resuspended in 5 ml of ice-cold soluble buffer containing 10 g/ml aprotinin, 10 M leupeptin, and 1 mM dithiothreitol. After freezing and thawing, suspended cells were sonicated on ice in short bursts. The lysate was cleared by centrifugation at 16,000 ϫ g for 15 min at 4°C. The supernatant was then incubated with glutathione-Sepharose 4B (Amersham Pharmacia Biotech) for 1 h at 4°C. The beads were settled by centrifugation at 700 ϫ g, washed five times with ice-cold soluble buffer, and incubated with 10 mM reduced glutathione for 10 min at 4°C to elute the GST or GST-cGK I␣-(1-66) fusion protein from the beads. After centrifugation at 700 ϫ g, the supernatant was dialyzed against phosphate-buffered saline and used for in vitro binding analysis. Similarly, skeletal TnT (skTnT) was expressed as maltose-binding protein (MBP) fused protein in Escherichia coli and purified by maltose affinity chromatography (NEB).
Immunoblotting-The proteins were subjected to SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Millipore). After being blocked with Block Ace (Snow Brand Milk Products) overnight, the membrane was incubated with goat anti-GST antibody for 2 h at room temperature. Bound primary antibodies were detected using horseradish peroxidase-conjugated anti-goat IgG (Jackson) and visualized by the ECL Western blotting detection system (Amersham Pharmacia Biotech).
In Vivo Binding Analysis-COS-7 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 units/ml penicillin, and 100 g/ml streptomycin at 37°C in 5% CO 2 . The full-length bovine cGK I␣ or human cGK I␤ cDNA in the expression vector pFLAG-CMV-2 (Kodak) and the full-length mouse skeletal TnT or human cardiac TnT cDNA in pcDNA3.1/His C (Invitrogen) were transiently expressed in COS-7 cells using LipofectAMINE PLUS regents according to the instructions of the manufacturer (Life Technologies, Inc.). Cells were washed with ice-cold phosphate-buffered saline 24 h after transfection and scraped in an ice-cold TNE buffer. Cell extracts were centrifuged at 16,000 ϫ g for 15 min at 4°C to remove cellular debris. The supernatants were immunoprecipitated with either anti-Xpress polyclonal antibody (Invitrogen) or anti-FLAG M5 monoclonal antibody (Kodak) with protein G-Sepharose overnight at 4°C by rotation. The beads were washed five times with TNE buffer, and immune complexes were eluted by heating at 95°C in 2 ϫ SDS sample buffer, subjected to SDS-PAGE, and analyzed by immunoblotting as described above.
In Vitro Kinase Assay-For the analysis of cGK activity, the immunoprecipitated samples were used. The in vitro kinase reaction was performed in kinase reaction buffer (50 mM Tris-HCl at pH 7.5, 20 mM magnesium acetate, 0.2 mM [␥-32 P]ATP, 2 M protein kinase A inhibitor (5-24) (PKI (5-24), Calbiochem), 5 mM glycerophosphoric acid, and 1 mM sodium orthovanadate) in the presence or absence of cGMP (5 M final concentration). The samples were incubated at 30°C for 30 min and centrifuged at 16,000 ϫ g and 4°C. The beads were mixed with an equal volume of 2 ϫ SDS sample buffer and heated at 95°C for 5 min, and the denatured proteins were subjected to SDS-PAGE. Gels were dried and subjected to autoradiography at Ϫ80°C.
In Vivo Kinase Assay-COS-7 cells were co-transfected with pFLAG-cGK I␣ and pFLAG-cTnI as described above. For [ 32 P]phosphate labeling, transiently transfected COS-7 cells were washed and preincubated with phosphate-free medium containing 1 mCi/ml [ 32 P]phosphate (500 Ci/ml, NEN Life Science Products) for 3 h and either left untreated or treated with 8-CPT-cGMP (Sigma) for 1 h. Cells were scraped in an ice-cold cell extract buffer (10 mM Tris-HCl at pH 7.5, 0.1% Triton X-100, 0.15 M NaCl, 1 mM EDTA, 10 g/ml aprotinin, 10 M leupeptin, 5 mM glycerophosphoric acid, and 1 mM sodium orthovanadate). The cell extracts were centrifuged at 16,000 ϫ g for 15 min at 4°C to remove cellular debris, and the supernatants were immunoprecipitated with anti-FLAG M5 monoclonal antibody with protein G-Sepharose for 4 h at 4°C by rotation. The beads were washed twice with cell extract buffer, and immune complexes were eluted by heating at 95°C in 2 ϫ SDS sample buffer, subjected to SDS-PAGE, and autoradiographed at Ϫ80°C.

cGK I Interacts with Troponin T in the Yeast Two-hybrid
System-To determine the molecular mechanism of cGMP/cGK signaling, and identify proteins that interact with cGK, the yeast two-hybrid system was used. The full-length bovine cGK I␣ was used as bait to screen mouse 17-day embryo and human skeletal muscle cDNA libraries. Of the 2 ϫ 10 6 transformants from the mouse library screened, three colonies were positive for the ␤-galactosidase assay. Sequence analysis revealed that all of the isolated cDNAs encode mouse slow skTnT (24). Furthermore, we isolated several independent cDNAs encoding human skTnT from the human skeletal muscle cDNA library as true positive clones. Experiments using the yeast two-hybrid system were conducted to confirm that the interaction of cGK I␣ with skTnT is specific. As shown in Fig. 1, neither GAL4-BD alone nor the GAL4-BD p53 fusion protein interacted with skTnT. Only when the cGK I␣ cDNA was co-transformed with skTnT cDNA was expression of the lacZ reporter gene observed.
To identify the region of cGK I␣ responsible for the interaction with skTnT, we constructed a series of cGK I␣ deletion mutants and tested then for their ability to associate with skTnT in the yeast two-hybrid system. As shown in Fig. 2, we found that the cGK I␣ domain required for interaction with skTnT is restricted to the N-terminal sequence between amino acids 1-66. Expression of the GAL4-BD cGK I␣ deletion mutants in yeast was confirmed by immunoblotting analysis with an antibody against GAL4-BD (data not shown).
Importantly, other cyclic nucleotide-dependent protein kinases, including cGK I␤, cGK II, and cAK regulatory subunit II␣ (cAK RII␣), are known to contain the dimerization domain and autoinhibitory region in their N terminus, respectively. To determine whether skTnT associates with these kinase isozymes, the full-length cGK I␤, cGK II, cAK RII␣, and cAK catalytic subunit ␣ (cAK C␣) were fused with GAL4-BD and co-transformed with skTnT. The yeast two-hybrid system demonstrated that only cGK I␤ can associate with skTnT (Fig. 3). To be sure that the interaction with skTnT is cGK I-specific, expression of the GAL4-BD cAK RII␣ fusion protein in yeast was confirmed as demonstrated by the interaction of GAL4-BD cAK RII␣ with GAL4-AD microtubular-associated protein 2 (25, 26; and data not shown). Expression of the other GAL4-BD fused proteins was confirmed by immunoblotting analysis using anti-GAL4-BD antibody (data not shown).

cGK I Interacts with Troponin T in Vitro and in Intact Mammalian Cells-To test whether the interaction between cGK I␣
and skTnT is direct or is mediated by a third unidentified yeast protein, we performed in vitro binding assay. Bovine cGK I␣-(1-66) and mouse full-length skTnT were expressed as a fusion protein with GST and MBP, respectively. GST-cGK I␣-(1-66) bound specifically to MBP-skTnT immobilized on amylose resin but not to MBP alone (Fig. 4).
To confirm this interaction between cGK I␣ and skTnT in intact cells, we co-expressed a FLAG-epitope-tagged cGK I␣ with Xpress-epitope-tagged skTnT in COS-7 cells. The ability of an antibody against either the FLAG or Xpress-epitope to precipitate a complex of cGK I␣ and skTnT suggests that these two proteins strongly interact in vivo (Fig. 5). When cGMP binds to cGMP-binding domains of cGK I, the kinase undergoes a conformational change which results in activation of its catalytic function (27,28). We examined the effect of the confor-  3. Selective interaction of cyclic nucleotide-dependent protein kinase families with skTnT in the yeast two-hybrid system. The Y190 yeast strain was co-transformed with GAL4-BD fusion plasmid expressing cGK I␣, cGK I␤, cGK II, cAK C␣, or cAK RII␣ in combination with GAL4-AD fusion plasmid expressing skTnT. Following selection on leucine/tryptophan-negative plates, two independent colonies were isolated and assayed for the activation of ␤-galactosidase.

FIG. 4. Interaction between cGK I␣ and skTnT in vitro.
Immobilized MBP or MBP-skTnT was incubated with the bacterially synthesized GST or GST-cGK I␣-(1-66). The bound proteins were eluted with maltose, separated by SDS-PAGE, and immunoblotted with anti-GST antibody. mational change produced by cGMP on the interaction between cGK I␣ and skTnT. The COS-7 cells, which were co-transfected with cGK I␣ and skTnT, were treated with 8-bromo-cGMP, a cell-permeable analogue of cGMP. However, exposure of the cells to 8-bromo-cGMP caused no change in the amount of the cGK I␣-skTnT complex.
cGK I Interacts with Troponin T via Its Leucine Zipper Motif-The cGK I␣-(1-66) contains the leucine zipper, a sequence of heptad repeats of leucines and isoleucines forming an ␣-helical structure, identified as the domain responsible for the dimerization of cGK I. To define whether cGK I␣ interacts with skTnT via the leucine zipper motif, we prepared a mutant (I19A/L40A) of cGK I␣ in which both isoleucine 19 and leucine 40 are changed to alanines. As shown in Fig. 6, cGK I␣ (I19A/ L40A), which disrupted the leucine zipper motif, was shown to completely abolish the binding to skTnT. Identical results were obtained using the yeast two-hybrid system (data not shown). These observations indicate that this leucine zipper motif in the N-terminal region of cGK I plays a critical role in the association with skTnT. Furthermore, to directly determine whether cGK activity is required for the interaction between cGK I␣ and skTnT, a kinase-deficient version of cGK I␣, cGK I␣ (D502A), in which aspartate 502 is changed to alanine (29), as a FLAG-tagged protein, was co-expressed with skTnT in COS-7 cells. No significant inhibitory nor stimulatory effect on the interaction was seen on the introduction of the point mutation in the kinase domain of cGK I␣.
Moreover, we examined the ability of skTnT protein to modulate the cGK activity in vitro. cGK activity was measured by immune complex kinase assay using a synthetic substrate, BPDEtide (30). However, overexpression of skTnT did not affect the cGK activity in COS-7 cells (data not shown), suggesting that the association with TnT does not stimulate nor inhibit the cGK activity. Recent studies have revealed that PKCphosphorylates two sites in TnT, leading to a slight increase in Ca 2ϩ -sensitivity without affecting the Mg 2ϩ -ATPase activity (31). However, we could not detect the phosphorylation of sk-TnT by cGK I␣, suggesting that skTnT is not a substrate for cGK I␣.
cGK I Phosphorylates Cardiac Troponin I via the Interaction between cGK I and Troponin T-TnT is a muscle-specific pro-tein of the myofibrillar apparatus which is involved in the Ca 2ϩ -dependent regulation of contraction in cardiac and skeletal muscles (32). There are three TnT isoforms, corresponding to slow skeletal, fast skeletal, and cardiac TnTs. We examined whether cGK I associated with another TnT, cardiac TnT (cTnT). In vivo analysis has shown that both cGK I␣ and cGK I␤ can associate with human cTnT, as well as skTnT (Fig. 7). The troponin complex consists of the three subunits, troponin C (TnC), TnI, and TnT. Because a recent report provided that TnT interacts with TnI via the leucine zipper motif (33), we assessed whether cGK I␣ would interact with TnI. The yeast two-hybrid system demonstrated that cGK I␣ did not associate with cTnI, indicating that cGK I␣ interacts specifically with TnTs (data not shown). It has been shown by several groups that cTnI is phosphorylated by cAK, and the phosphorylation of cTnI decreases myofilament sensitivity to calcium. Previous studies demonstrated that cGK purified from bovine lung, as well as cAK, phosphorylates cTnI in vitro. To determine whether cTnI is phosphorylated by both cGK I␣ and cGK I␤ in vitro and in vivo, cGK I and cTnI were co-expressed as FLAGtagged proteins in COS-7 cells. As expected, an in vitro kinase assay using the immunoprecipitation complex showed that cTnI was phosphorylated by both cGK I␣ and cGK I␤ in cGMPdependent manner (Fig. 8B), and the level of the phosphorylation of cTnI by cGK I␣ was equal to that by cAK (Fig. 8A). A kinase-defective mutant of cGK I␣, cGK I␣ (D502A), did not catalyze the phosphorylation of either itself or cTnI, confirming that cTnI is a good substrate for cGK I in vitro. Some substrates are shown to be good in vitro but not in vivo (34). Thus, we conducted an in vivo kinase assay using 32 P-labeled COS-7 cells, which were co-transfected with cGK I and cTnI. Similar results were obtained in vivo, suggesting that cTnI is a good substrate for cGK I both in vitro and in vivo (Fig. 8B). Interestingly, cGK I␣ (D502A) was phosphorylated in the in vivo but not in vitro kinase assay, suggesting that cGK I␣ is phosphorylated by endogenous kinases in intact cells.
TnT is supposed to function as an anchoring protein for cGK I. To investigate this possibility, we co-expressed FLAGepitope-tagged cGK I␣, FLAG-epitope-tagged cTnI, and Xpress-epitope-tagged cTnT in COS-7 cells. The cell lysates were immunoprecipitated with anti-Xpress antibody, and the immunocomplex kinase assay was performed in the presence of PKI. Only when cTnT is co-expressed, was cTnI phosphorylated in a cGMP-dependent manner, indicating that cTnI is phosphorylated by cGK I␣ via the interaction between cGK I␣ and cTnT (Fig. 9). Additionally, to be certain that cTnI is phosphorylated by cGK I␣, the absence of cAK in the TnT immunopellet was confirmed by immunoblotting analysis using the anti-cAK C antibody (data not shown).

DISCUSSION
Troponins are well characterized as muscle-specific proteins of the myofibrillar apparatus and are involved in the Ca 2ϩ -dependent regulation of contraction in cardiac and skeletal muscles (32). The troponin complex consists of three subunits, the Ca 2ϩ -binding TnC, the ATPase-inhibiting TnI, and the tropomyosin-binding TnT. The binding of Ca 2ϩ to TnC releases the TnI inhibition of actomyosin ATPase through protein-protein interactions among the troponin complex, tropomyosin and actin, and leads to muscle contraction. In addition to its direct activation by Ca 2ϩ through binding to TnC, actomyosin Mg 2ϩ ATPase can be further modulated by phosphorylation of contractile protein. cAK was previously demonstrated to efficiently phosphorylate human cTnI at Ser 22 and Ser 23 residues in vitro (35,36). Recent studies demonstrated that phosphorylation of cTnI by cAK reduces the affinity of TnC for Ca 2ϩ and enhances dissociation of the TnC/Ca 2ϩ complex (37). Furthermore, PKC isozymes (␣, ␤2, ␦, and ⑀) were shown to phosphorylate cTnI and reduce the maximal activity of actomyosin Mg 2ϩ -ATPase (31). These observations suggest that serinethreonine protein kinases are involved in the modification of cardiac contractility without a significant change of calcium transient in cardiac myocytes. In the human heart, recent reports demonstrated that activation of ␤ 2 -adrenergic receptors hastens ventricular relaxation and promotes phosphorylation of myofibrillar proteins, including cTnI (38). However, further study is required to elucidate the correlation of the degree of protein phosphorylation with the magnitude of inotropic effects and the relaxation process.
Membrane-permeable cGMP analogues and cGMP-elevating agents are well known to reduce cardiac contractility through cGK activation by a decrease in cardiac calcium current (39,40). Recent studies provided that natriuretic peptide inhibited L-type Ca 2ϩ channel activity via a cGMP-dependent pathway, and that the phosphorylation of L-type Ca 2ϩ channels by cGK caused a decrease in the sensitivity of the channels to cAK activated by ␤ 2 -adrenergic stimulus (41). However, Shah et al. (42) reported that cGMP analogues decrease the relative myofilament response to Ca 2ϩ in intact cardiac myocytes and that the negative inotropic effects of cGMP analogues were inhibited by KT-5823, a specific cGK inhibitor. Additionally, cGK is shown to induce a rightward shift of a tension/pCa relation in skinned single rat ventricular cells, compatible with the reduction in Ca 2ϩ sensitivity of TnC (43). There are reports by Lincoln and Corbin (44) and others (45) that cGK and cAK phosphorylate isolated cTnI at the same site. In the present study, cTnI was demonstrated to be a good substrate for cGK I FIG. 7. In vivo interaction of cGK I with cTnT. Expression vector for Xpress-skTnT or Xpress-cTnT was cotransfected into COS-7 cells with the empty vector as a negative control or expression vector for either FLAG-cGK I␣ or FLAG-cGK I␤. After whole cell lysates were immunoprecipitated with anti-FLAG antibody (IP: Flag), immunoblotting was carried out with anti-Xpress antibody (Blot: Xpress).

FIG. 8. Phosphorylation of cTnI by cGK I in vitro and in vivo.
A, phosphorylation of cTnI by cGK I and cAK in vitro. B, phosphorylation of cTnI by cGK I␣ and cGK I␤ in vitro. COS-7 cells were transfected with expression plasmid for FLAG-cTnI. Where indicated, plasmid expressing either FLAG-cGK I␣, FLAG-cAK C␣, kinase-defective mutant of FLAG-cGK I␣ (cGK I␣ D502A), or FLAG-cGK I␤ was included in the transfection. Whole cell lysates were immunoprecipitated with anti-FLAG antibody (IP: Flag), and cGK activity was measured in an in vitro kinase assay. Reactions were performed in the absence (Ϫ) or presence (ϩ) of 5 M cGMP. To monitor the expression level of each kinase, the immunoprecipitates were blotted with anti-FLAG antibody (Blot: Flag). C, phosphorylation of cTnI by cGK I in vivo. COS-7 cells cotransfected with expression vectors for either cGK I␣, cGK I␣ D502A, or cGK I␤ in combination with the expression plasmid for FLAG-cTnI were labeled with 32 P as detailed under "Experimental Procedures." 32 P-labeled cell lysates were immunoprecipitated with anti-FLAG antibody (IP: Flag), and the immunoprecipitates were subjected to SDS-PAGE and autoradiography. To monitor the expression levels of kinase and cTnI proteins, the immunoprecipitates were blotted with anti-FLAG antibody (Blot: Flag).
in intact cells, as well as in vitro. Because the concentration of cGK in the heart appears to be approximately 11-fold lower than that of cAK (10) and the V max value for the phosphorylation reaction of cTnI catalyzed by cAK is 12-fold greater than the value obtained for cGK (45), phosphorylation of cTnI by cAK in the intact heart may occur at a rate 130-fold greater than cGK if both kinases are fully activated. One possible mechanism for selective cGK phosphorylation is the localization of cGK with substrate proteins by interaction with the specific anchoring proteins located at various sites in the cell, leading to compartmentalization of the kinase to sites of action.
Mutation studies indicated that an intact leucine zipper structure of cGK I␣ is required for the interaction between cGK I␣ and skTnT. The function of the leucine zipper motif is to mediate homodimerization or heterodimerization with other proteins containing this domain. Whereas cGKs are shown to exist as homodimers of the subunits, previous studies indicated that vimentin, one of the intermediate filament proteins, is a high affinity and specific binding protein for cGK, and that the dimerization of cGK is likely to be necessary for binding to vimentin (46,47). The current study could not exclude the possibility that TnT interacts with cGK I␣ dimerized via the leucine zipper structure. Recently, myosin was identified as a cGK anchoring protein (GKAP), which binds to cGK II (48). Interestingly, cGK II interacted with myosin via the dimerization domain similar to the interaction between cGK I and TnT although it has not been made clear whether the leucine zipper motif in cGK II is required for the direct interaction or the dimerization. Despite that the leucine zipper motif is observed in the N-terminal region of all cGKs, TnT could interact with cGK Is but not cGK II. It is also notable that the primary structure of the cGK I␣ and I␤ differ only in the N-terminal ϳ100 residues, sharing 36% identity in this region.
The present study suggests that TnT functions as an anchoring protein for cGK I and serves to ensure preferential and rapid phosphorylation of TnI by cGK I in response to increases in the intracellular concentration of cGMP and that the cGMP/ cGK I signaling pathway may participate in regulation of muscle contraction. FIG. 9. Phosphorylation of cTnI by cGK I via the association with cTnT. The expression plasmids for FLAG-cGK I␣ and FLAG-cTnI were cotransfected into COS-7 cells with the empty plasmid as a negative control or plasmid for Xpress-cTnT. Whole cell lysates were immunoprecipitated with anti-Xpress antibody (IP: Xpress), and cGK activity was measured in an in vitro kinase assay. Reactions were performed with (ϩ) or without (Ϫ) 5 M cGMP. Protein kinase inhibitor peptide (2 M) was incubated in all assay tubes to inhibit endogenous cAK activity.