Activation of the protein kinase C (PKC) family is a potential signaling mechanism by which high ambient glucose concentration modulates the phenotype and physiological function of cells. Recently, the cardiac renin angiotensin system (RAS) has been reported to promote PKC translocation in the diabetic heart via the angiotensin (ANG) II type 1 receptor (AT-1R). To evaluate the molecular events coupled with high glucose−induced PKC translocation and to examine the role of endogenously released ANG II in myocyte PKC signaling, primary cultures of adult rat ventricular myocytes were exposed to normal (5 mmol/l) or high (25 mmol/l) glucose for 12–24 h. Western blot analysis indicated that adult rat ventricular myocytes coexpress six PKC isozymes (α, β1, β2, δ, ε, and ζ). Translocation of five PKC isozymes (β1, β2, δ, ε, and ζ) was detected in response to 25 mmol/l glucose. Inhibition of phospholipase C with tricyclodecan-9-yl-xanthogenate blocked glucose-induced translocation of PKC-β2, -δ, and -ζ. Inhibition of tyrosine kinase with genistein blocked glucose-induced translocation of PKC-β1 and -δ, whereas chelation of intracellular Ca2+ with 1,2-bis(2-aminophenoxy)ethane N,N,N,’N′-tetraacetic acid blocked translocation of PKC-β1 and -β2. Enzyme-linked immunosorbent assay performed on culture media from myocytes maintained in 25 mmol/l glucose detected a twofold increase in ANG II. Addition of an AT-1R antagonist (losartan; 100 nmol/l) to myocyte cultures blocked translocation of PKC-β1, -β2, -δ, and -ε. Phosphorylation of troponin (Tn) I was increased in myocytes exposed to 25 mmol/l glucose. Losartan selectively inhibited Tn I serine phosphorylation but did not affect phosphorylation at threonine residues. We concluded that 1) 25 mmol/l glucose triggers the release of ANG II by myocytes, resulting in activation of the ANG II autocrine pathway; 2) differential translocation of myocyte PKC isozymes occurs in response to 25 mmol/l glucose and ANG II; and 3) AT-1R−dependent PKC isozymes (β1, β2, δ, and ε) target Tn I serine residues.
Hyperglycemia dominates the pathophysiology and clinical course of type 1 and type 2 diabetes. Compelling evidence from the Diabetes Control and Complications Trial has indicated that rigorous control of blood glucose reduces the risk of long-term microvasculature complications (1). Although the cardiac myocyte is not a principal target cell for insulin, several lines of investigation support the notion that a primary myocardial defect exists in diabetes (2,3,4). An important question concerns the signals used by high concentrations of extracellular glucose to alter the biochemical and mechanical properties of cardiac muscle cells. Recruitment of the protein kinase C (PKC) family of serine-threonine kinases is an integral component of the signaling events that direct the cardiac phenotype expressed during postnatal cardiac development (5) and in response to pathological stimuli (6,7,8,9,10).
Adult rat ventricular myocytes express multiple PKC isozymes, which are redistributed from the cytosol to the membrane in response to a diverse array of extracellular stimuli. In the presence of an activating signal, PKC isozymes bind to specific anchoring proteins, collectively known as receptors for activated C kinase (11). Translocation of cardiac PKC isozymes in response to chronic hyperglycemia has been reported in the streptozotocin (STZ) model of diabetes (12,13,14). De novo synthesis of diacylglycerol (DAG) has been proposed as the mechanism of sustained PKC activation in diabetic tissue (15,16,17,18). More recently, ligand activation of the angiotensin (ANG) II type-1 receptor (AT-1R) on cardiac myocytes has been linked to the activation of PKC-ε in the diabetic rat heart (14), implicating the renin angiotensin system (RAS) in glucose-induced PKC translocation. However, the inherent constraints of the in vivo model limits the ability to establish potential substrates for activated PKC isozymes and the role of local or circulating ANG II in the diabetic heart. Given the ample body of evidence implicating PKC in hyperglycemia-related cellular injury (19,20,21), the upregulation of the myocyte RAS in the diabetic heart (22,23), and the well-documented cytotoxic effects of ANG II (24,25), such information may prove critical to preventing or arresting the progression of diabetic cardiomyopathy.
The purpose of the present study was to determine the mechanism(s) by which high ambient concentrations of glucose induce PKC translocation in cardiac myocytes and whether PKC isozymes redistributed in response to high glucose target specific intracellular proteins linked to the biochemical and mechanical alterations of the diabetic myocardium. Primary cultures of adult rat ventricular myocytes were exposed to 5 or 25 mmol/l glucose for 12–24 h, and translocation of PKC isozymes from cytosol to membrane was examined by Western blot analysis, using isozyme-specific antibodies. To identify signal transduction pathways coupled to the subcellular redistribution of PKC isozymes, studies were performed with the intracellular Ca2+ chelating agent, 1,2-bis(aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA/AM), and specific inhibitors of phospholipase C and tyrosine kinase signaling pathways. To determine whether high glucose provokes the release of ANG II from cardiac myocytes and whether endogenously released ANG II is coupled with the activation of AT-1R signaling pathways that promote PKC translocation, ANG II was measured in culture media, and Western blot analysis of PKC isozymes was performed in the presence and absence of selective AT-1R and ANG II type 2 receptor (AT-2R) antagonists. Finally, we demonstrated phosphorylation of the myofilament regulatory protein troponin I (Tn I) by AT-1R−dependent PKC isozymes, documenting a role for the autocrine RAS in the biochemical and mechanical abnormalities of the diabetic myocardium.
RESEARCH DESIGN AND METHODS
SD rats, 275–300 g body wt, were anesthetized with a cocktail mixture containing ketamine and xylazine (80 and 10 mg/kg body wt, i.p.). Hearts were removed from the chest cavity, rinsed with minimum essential media (Sigma), and used for isolation of cardiac myocytes.
Isolation of adult rat ventricular myocytes and cell culture.
Cardiac myocytes were isolated by enzymatic dissociation with collagenase, precisely as described in previous publications (14,26). Based on their characteristic morphologic shape, striations, and trypan blue staining, ∼80–90% of cells were viable myocytes. Freshly isolated myocytes were plated in laminin-coated Petri dishes at a density of 2 × 104 cells/cm2 and incubated in serum-free medium (SFM) at 37°C in an atmosphere containing 5% CO2. Primary cultures of adult rat ventricular myocytes were exposed to SFM with 5 or 25 mmol/l glucose for 12 h, unless indicated otherwise. For time course experiments, myocytes were plated in SFM containing 5 or 25 mmol/l glucose and incubated for 15 min, 1 h, 12 h, or 24 h. Inhibition studies were performed with four concentrations of tricyclodecan-9-yl-xanthogenate (D609) and genistein (1, 10, 25, and 100 μmol/l), BAPTA/AM (25 μmol/l), losartan (100 nmol/l), and PD123,319 (100 nmol/l). Inhibitors were added to freshly plated cultures of cardiac myocytes maintained in SFM containing 5 mmol/l glucose for 1 h. The media was then replaced by SFM containing 25 mmol/l glucose plus inhibitors for 12 h. Myocytes were harvested and processed for immunoblots and biochemical studies. To control for the effect of osmolarity, separate studies were performed with cardiac myocytes maintained in SFM containing 5 mmol/l glucose and 20 mmol/l mannitol. For phorbol 12-myristate 13-acetate (PMA) activation of PKC, myocytes were cultured in SFM (25 mmol/l glucose) for 12 h. Then, PMA (100 μmol/l) was added to the culture for 30 min.
PKC fractionation in cardiac myocytes.
Extraction and partial purification of PKC from cardiac myocytes was performed as described previously (14). Briefly, cells were homogenized in 0.5 ml buffer A containing 20 mmol/l Tris-HCl, 0.33 mol/l sucrose, 2 mmol/l EGTA, 2 mmol/l EDTA (pH 7.5), 0.1 mmol/l sodium vanadate, 20 mmol/l NaF, 20 μmol/l leupeptin, 200 μmol/l phenylmethylsulfonyl fluoride (PMSF), protease inhibitor cocktail set (type I and III; Calbiochem), 5 mmol/l dithiothreitol (DTT), and the phosphatase inhibitors calyculin A (1 nmol/l), and okadaic acid (5 nmol/l). The cytosolic fraction (supernatant) was separated by centrifugation of the crude homogenate (2 mg) at 25,000g for 30 min. The pellet was resuspended in homogenization buffer A without sucrose containing 0.1% Triton X-100, followed by centrifugation and collection of the particulate fraction (supernatant). To quantitate the immunoreactivity of total PKC isoenzymes, myocytes were homogenized in buffer A in the presence of 0.1% Triton X-100, and the supernatant was collected. The samples were processed for SDS-PAGE and immunoelectrophoresis.
Next, 15–20 μl (2 mg/ml) of the protein samples were separated by SDS-PAGE (Bio-Rad Apparatus) gel using an 8% (wt/vol) acrylamide gel. Proteins were transferred to nitrocellulose using a semi-dry transfer cell (Bio-Rad). Nonspecific sites were blocked with 6% nonfat milk. Then, membrane was incubated with primary rabbit polyclonal antibodies against PKC-α, -ε, and -ζ (Santa Cruz Biotechnology) or PKC-δ (GIBCO-BRL) and mouse monoclonal antibodies for PKC-β1 and -β2 (Sigma) isoforms at a dilution of from 1:500 to 1:1,000 in the blocking solution, washed, and treated with the horseradish peroxidase (HRP)−linked secondary antibodies (1:5,000 to 1:10,000 dilution; Boehringer Mannheim). The bound antibody was detected by autoradiography using an enhanced chemiluminesce kit (Amersham). The specificity of the antibody was confirmed by incubation with the appropriate blocking peptide. Densitometric analysis for the translocation of PKC isozymes was performed using the computerized image analysis software, Un-Scan-IT (Automated Digitizing System).
Glucose-induced Tn I phosphorylation.
Phosphorylation of Tn I serine and threonine residues was determined by Western blot analysis. Freshly isolated myocytes were plated in SFM containing 5 or 25 mmol/l glucose for 12 h. In a separate set of studies, losartan (100 nmol/l) was added. Cells were collected and homogenized with ice-cold 0.5–0.7 ml of buffer containing 50 mmol/l Tris-HCl (pH 7.4), 5 mmol/l EDTA, 250 mmol/l NaCl, 0.1% Triton X-100, 50 mmol/l NaF, 0.1 mmol/l Na3VO4, 0.5 mmol/l PMSF, 5 mmol/l DTT, protease inhibitor cocktail set (type I and III; Calbiochem), and phosphatase inhibitors (1 nmol/l calyculin A and 5 nmol/l okadaic acid). The samples were prepared (2 mg/ml) for SDS-PAGE. Proteins (30–40 μg) were subjected to 10% SDS gels (Bio-Rad) and transferred electrophoretically to nitrocellulose, as described earlier. The membranes were probed with primary antibodies (monoclonal cTn I antibody, dilution 1:3,000) (27,28), a generous gift from the laboratory of Professor S. Schiaffino (Padova, Italy); monoclonal anti-phosphoserine, 1:700; and monoclonal anti-phosphothreonine, 1:700 (Sigma). For phosphoserine and phosphothreonine detection, 2% bovine serum albumin was used as a blocking reagent followed by HRP-linked secondary antibody at a dilution of 1:10,000 to 1:15,000. Phosphorylated Tn I was identified in the myocyte homogenates by anti-phosphoserine and anti-phosphothreonine antibodies using Tn I as a molecular weight marker (29–31 kDa).
To confirm the identity of the protein band at 29–31 kDa as Tn I, the blots were stripped and reprobed for Tn I using monoclonal anti−Tn I antibody. In a separate set of experiments, the protein band at 29–31 kDa was excised from the SDS gel and the extracted protein subjected to 10% SDS-PAGE. The identity of the protein was confirmed by probing with a monoclonal anti−Tn I antibody.
Enzyme-linked immunosorbent assay for ANG II.
Cardiac myocytes were plated in SFM containing 5 or 25 mmol/l glucose for 10 or 24 h. The culture media was collected, and ANG II was quantitated by enzyme-linked immunosorbent assay (ELISA) (29). ANG II in SFM was partially purified by extraction through Sep-Pak C18 cartridge (Waters, Milford MA). The samples were assayed using primary antibody (anti−ANG II; Peninsula Laboratories), a tracer (biotinylated ANG II; Peninsula), and streptavidin-HRP (Amersham Pharmacia Biotech). The concentration of ANG II was calculated from the standard absorbance curve for ANG II.
Data are expressed as means ± SE. Comparisons between two values were performed by unpaired Student’s t test. For multiple comparisons among different groups of data, the significant differences were determined by the Bonferroni method (30). Significance was defined as P < 0.05.
Alterations in the subcellular distribution of PKC isozymes by glucose.
In preliminary investigations, we detected the presence of six PKC isozymes (α, β1, β2, δ, ε, and ζ) in adult rat ventricular myocytes. PKC-α, -ε, and -ζ were found to be preferentially partitioned in the cytosolic fraction, whereas PKC-β1, -β2, and -δ were associated with the membrane compartment. To determine the effect of high ambient glucose concentration on the subcellular distribution of the above PKC isozymes, freshly isolated adult rat ventricular myocytes were plated and cultured in SFM containing 5 or 25 mmol/l glucose.
Representative immunoblots for the six PKC isozymes are shown in Fig. 1. The immunoreactivity detected was specific to each of these isozymes, as illustrated by its absence in the presence of the respective blocking peptides. To control for equivalency of loading conditions, proteins were stained with Commassie blue (data not shown). Myocytes exposed to 25 mmol/l glucose for 12 h exhibited a marked increase in the membrane-associated immunoreactivity for PKC-β1, -β2, -δ, -ε, and -ζ. In contrast, 25 mmol/l glucose failed to increase the abundance of PKC-α in the membrane fractions. Following exposure to 400 nmol/l PMA for 30 min, a rapid translocation of PKC-α was detected in cardiac myocytes maintained in 25 mmol/l glucose (Fig. 2). This finding confirmed that the high ambient glucose concentration did not alter the membrane-anchoring proteins for PKC-α and that 25 mmol/l glucose is not a stimulus for translocation of PKC-α.
Time course of glucose-induced PKC translocation.
To determine whether 25 mmol/l glucose induces rapid or delayed redistribution of PKC isozymes and whether translocation is transient or sustained, cardiac myocytes were exposed to 25 mmol/l glucose for 15 min, 1 h, 12 h, or 24 h. Immunoblots of membrane and cytosolic fractions were then prepared for each of the above time intervals. Membrane/cytosolic ratios of immunoreactivity were used as indexes of the extent of PKC translocation. As shown in Fig. 3, exposure of cardiac muscle cells to 25 mmol/l glucose for 15 min and 1 h failed to promote translocation of PKC isozymes. At 12 h, significant increases (P < 0.05) in the membrane/cytosolic ratio were detected for all isozymes, with the exception of PKC-α. These findings indicated a delayed rather than an acute subcellular redistribution of PKC isozymes in response to 25 mmol/l glucose. The absence of an early response may have reflected the time required for glucose-induced generation of phospholipids or for activation of other signaling molecules. Moreover, the sustained increase in the membrane/cytosolic ratios for PKC-β1, -β2, and -ε at 24 h suggested that these isozymes might be preferentially redistributed in response to a high ambient glucose concentration.
Role of phospholipase C, tyrosine kinase, and calcium in glucose-induced PKC translocation.
To determine whether 25 mmol/l glucose promotes PKC translocation by activating one or more signaling pathways, adult rat ventricular myocytes were preincubated for 12 h with specific inhibitors of phospholipase C (D609), tyrosine kinase (genistein), or intracellular Ca2+ (BAPTA/AM) (31,32). To examine the role of phospholipase C and tyrosine kinase in glucose-induced PKC translocation, studies were performed with D609 and genistein at concentrations of 1, 10, 25, and 100 μmol/l. To determine whether 25 mmol/l glucose promotes PKC translocation by mobilizing intracellular Ca2+ stores, studies were performed with BAPTA/AM at a concentration of 25 μmol/l.
D609 failed to block glucose-induced translocation of PKC-β1, -β2, -δ, -ε, and -ζ at concentrations of 1, 10, and 25 μmol/l (data not shown). As shown in Fig. 4, at the 100 μmol/l concentration, D609 blocked the translocation of PKC-β2, -δ, and -ζ. A small but insignificant decrease in the translocation of PKC-β1 was detected at this dosage of D609, with no measurable effect on PKC-ε. These results indicated that the subcellular redistribution of PKC isozymes in response to high ambient concentrations of glucose is only partially dependent on phospholipase C and DAG generation.
To determine whether 25 mmol/l glucose recruits PKC isozymes via tyrosine kinase−dependent pathways, the effect of four different concentrations of genistein on PKC signaling was examined. Genistein failed to block glucose-induced translocation of PKC-β1, -β2, -δ, -ε, and -ζ at concentrations of 1, 10, and 25 μmol/l (data not shown). At 100 μmol/l, genistein blocked glucose-induced translocation of PKC-β1 and -δ (Figs. 4B and D). These findings indicated that glucose-induced redistribution of PKC-β1 is dependent on the activation of tyrosine kinase pathways, whereas PKC-δ requires both tyrosine kinase signaling and phospholipase C activation. Genistein failed to block glucose-induced translocation of PKC-β2, -ε, and ζ. The aggregate data indicated that 25 mmol/l glucose activates tyrosine kinase pathways, which promote PKC redistribution in adult rat ventricular myocytes.
To examine the role of Ca2+ in glucose-induced PKC translocation, freshly isolated cardiac myocytes were incubated with 25 μmol/l of BAPTA/AM or vehicle for 30 min, washed, and then placed in media containing 5 or 25 mmol/l glucose and 25 μmol/l BAPTA/AM for 12 h. BAPTA/AM inhibited translocation of PKC-β1 and -β2 (Figs. 4B and C) but not PKC-δ, -ε, and -ζ in myocytes maintained at 25 mmol/l glucose concentration. Taken together with the above results, our findings indicated that 25 mmol/l glucose activates multiple signaling pathways that promote redistribution of cardiac PKC isozymes. Translocation of PKC-ε, the major cardiac isozyme, appears to occur via novel signal transduction pathways, independent of phospholipase C, tyrosine kinase, or intracellular Ca2+ release.
Glucose-induced ANG II release and AT-1R−dependent PKC translocation.
Ligand activation of the AT-1R has been described in the diabetic heart (14,23). To determine whether glucose provokes the release of ANG II from adult rat ventricular myocytes under in vitro conditions, ELISA was performed on the culture media of myocytes maintained at 5 or 25 mmol/l glucose for 10 or 24 h. As shown in Fig. 5, the spontaneous release of ANG II, previously described for cardiac myocytes in culture (33), was significantly increased by exposure to 25 mmol/l glucose. ANG II content was twofold greater in the culture media of cardiac myocytes maintained at 25 mmol/l glucose. To determine whether glucose-induced release of ANG II results in the activation of surface AT-1R on myocytes, promoting PKC translocation, the above studies were repeated in the presence and absence of the selective AT-1R antagonist losartan. Western blot analysis indicated that AT-1R blockade completely reversed glucose-induced translocation of PKC-β1, -β2, -δ, and -ε (Fig. 6). Losartan did not affect glucose-induced translocation of PKC-ζ. An identical analysis was performed with the selective AT-2R antagonist, PD123,319. Blockade of the AT-2R did not alter glucose-induced translocation of PKC isozymes (data not shown). These results indicated that activation of the AT-1R by endogenously released ANG II promotes translocation of PKC-β1, -β2, -δ, and -ε in cardiac myocytes exposed to 25 mmol/l glucose.
Activation of PKC and glucose-induced phosphorylation of Tn I.
To determine whether the sarcomeric protein Tn I is a substrate for PKC isozymes activated in response to glucose- or AT-1R−dependent pathways, the phosphorylation status of Tn I was examined in cardiac myocytes maintained at 5 or 25 mmol/l glucose for 12 h. As shown in Fig. 7 (left), Tn I phosphorylation was markedly enhanced in cells exposed to 25 mmol/l glucose. In contrast, 5 mmol/l glucose + 20 mmol/l mannitol had no effect on Tn I phosphorylation (data not shown). The high ionic strength of the salt required to extract and solubilize cardiac Tn I precluded the identification of this protein by immunoprecipitation. To confirm the identity of the 29–31 kDa protein as Tn I (Fig. 7, center), the band was excised, separated by SDS-PAGE, then subjected to Western blot analysis using monoclonal Tn I antibody (Fig. 7, right). Quantitatively, glucose-induced phosphorylation of serine residues was proportionately greater than that of threonine residues (Fig. 8). These findings indicated that 25 mmol/l glucose activated a serine/threonine kinase that targets Tn I as a substrate. To determine whether 25 mmol/l glucose−induced phosphorylation of Tn I was mediated by an AT-1R−dependent mechanism, the phosphorylation status of the Tn I protein was examined in the presence and absence of losartan. In cardiac muscle cells exposed to 25 mmol/l glucose, losartan completely reversed glucose-induced Tn I serine phosphorylation (Fig. 8). In contrast, losartan had no effect on glucose-induced phosphorylation of Tn I threonine residues.
The selective inhibition of Tn I serine phosphorylation by losartan implies that an AT-1R−dependent PKC isozyme targets critical serine residues (34) required for the activation of this regulatory protein. The absence of an inhibitory effect on Tn I threonine phosphorylation suggests that these residues are targets for PKC isozymes activated in response to high glucose. Taken together, the aggregate results suggest that glucose-induced phosphorylation of Tn I is mediated by an AT-1R−dependent mechanism.
In the present study, we demonstrated that in vitro exposure of adult rat ventricular myocytes to 25 mmol/l glucose activates signal transduction pathways that serve as a potent stimulus for the subcellular redistribution of PKC isozymes. Glucose-induced translocation of PKC isozymes in this system is characterized by a delayed onset, activation of novel signal transduction pathways, and redistribution of multiple PKC isozymes. Direct evidence was also provided for glucose-induced release of ANG II from cardiac muscle cells and AT-1R−dependent PKC translocation. Finally, we demonstrated that activated PKC isozymes target the myofibrillar regulatory protein, Tn I, by an AT-1R−dependent mechanism, thus establishing a role for the autocrine RAS in the biochemical and mechanical abnormalities of the diabetic myocardium.
High glucose−induced PKC translocation.
Although in vivo models of human disease have served as paradigms to study pathophysiology, limitations exist with respect to establishing cause and effect. Adult cardiac myocytes coexpress multiple PKC isozymes (10,32), increasing the complexity of determining the in vivo functional implications of PKC activation. A well-documented feature of experimental diabetes is the development of progressive cardiomyopathy that is independent of the vasculopathy (2,3,4). In the present study, we demonstrated for the first time that 25 mmol/l glucose, independent of counterregulatory hormones, promotes the translocation of multiple PKC isozymes (β1, β2, δ, ε, and ζ) in a homogeneous population of adult rat ventricular myocytes. Glucose-induced PKC translocation was characterized by a delayed onset, requiring more than 1 h of exposure to 25 mmol/l glucose. The latter property may serve to protect cardiac myocytes from transient glucose elevations or may reflect the time required for glucose-induced activation of signaling pathways involved in PKC translocation.
Although hyperglycemia dominates the pathophysiology of diabetes, the application of an in vitro system to study the effects of high ambient glucose on isolated cardiac muscle cells has obvious limitations. However, the results of the present study provide the basis for an interesting comparison with those reported following in vivo exposure to hyperglycemia. Our findings differ somewhat from those reported in whole heart tissue from STZ-induced diabetic rats, in which an increase in the membrane association of one or several of the PKC isozymes α, β2, δ, and ε has been reported (13,14,35,36). Differences between our results and those reported in whole heart tissue from STZ diabetic rats may reflect the physiological changes associated with the in vivo diabetic state, such as increased PKC-α expression in the nonmyocyte compartment of the myocardium (36). It must also be acknowledged that chronic exposure to a high glucose concentration may result in downregulation in the expression of some, but not all, PKC isozymes. As a case in point, Malhotra et al. (14), using the STZ model, reported that only cardiac PKC-ε exhibited evidence of sustained translocation and activation. However, that study was limited to PKC-δ and -ε and did not attempt to characterize the response of other cardiac PKC isozymes to high glucose or the signaling events by which glucose promotes PKC translocation. Compelling evidence has also been presented for the activation of PKC-β2 as a determinant of cardiac phenotype in diabetic rats (21). The pathogenic role of PKC-β2 was confirmed by demonstrating prevention of the histological and functional lesions in mice treated with a selective PKC-β isoform inhibitor (21). Moreover, transgenic mice with targeted overexpression of the PKC-β2 isozyme in cardiac myocytes exhibit a cardiac phenotype characterized by left ventricular hypertrophy, myocyte necrosis, multifocal fibrosis, and impaired left ventricular performance (37). Taken together, the available evidence indicates that a high ambient glucose concentration is a potent stimulus for the redistribution of cardiac PKC isozymes and suggests a role for multiple isozymes in the signaling events that result in the expression of the diabetic cardiac phenotype.
High glucose−induced activation of signaling pathways and PKC translocation.
Multiple lines of evidence have implicated activation of phospholipase C and the generation of the phospholipid DAG as the mechanism of PKC translocation in response to hyperglycemia (15,16,17,18). This is the first report documenting that in adult rat ventricular myocytes exposed to 25 mmol/l glucose, subcellular redistribution of PKC isozymes may occur independent of phospholipase C. Glucose-induced translocation of PKC-β1 and -ε was not attenuated or blocked by the phospholipase C inhibitor D609. This result was somewhat unexpected, as DAG is a phospholipid activator of PKC-β1 and -ε, and suggests that 25 mmol/l glucose recruits these isozymes through alternate signaling pathways. The dosage of D609 (100 μmol/l) required to demonstrate a partial inhibitory effect on glucose-induced redistribution of PKC isozymes may have reflected the initial activation of phospholipase C in the presence of 25 mmol/l glucose. D609 has also been reported to exhibit a higher affinity for phosphatidylcholine-specific than for phosphatidylinositol-specific phospholipase C (31). However, the selectivity of D609 for the latter enzyme has been documented in vitro under conditions of agonist and stretch-dependent activation of phospholipase C (38,39). It is also possible that higher dosages of inhibitors may be required to block translocation of PKC isozymes in response to noxious external stimuli.
To further explore signaling pathways that may be involved in glucose-induced PKC translocation, next we considered the role of tyrosine kinase and Ca2+. Genistein, a tyrosine kinase inhibitor, blocked the translocation of PKC-β1 and -δ. The intracellular Ca2+ chelator, BAPTA/AM, blocked translocation of PKC isozymes known to possess a Ca2+ binding domain but had no effect on PKC-δ, -ε, and -ζ. The balance of the evidence would appear to suggest that when cardiac muscle cells are exposed to abnormal physiological stimuli (31,32), the recruitment of PKC isozymes, via alternative signaling pathways, is an integral component of the PKC signaling mechanism.
High glucose−induced ANG II release and AT-1R−dependent PKC translocation.
The present study provides the first documentation that exposure of adult rat ventricular myocytes to a high ambient glucose concentration provokes the release of ANG II, coupled with selective AT-1R−dependent redistribution of PKC isozymes. The application of physical forces, such as mechanical stretch, to in vitro cultures of cardiac myocytes, has been reported to be a stimulus for the release of ANG II (39,40). Our finding that 25 mmol/l glucose also provokes the release of ANG II implies that these discordant external stimuli may activate a common pathway. Alternatively, release of ANG II from cardiac myocytes may be coupled with increased intracellular generation of this peptide. Although beyond the scope of the present investigation, a high ambient glucose concentration has been reported to promote PKC-dependent phosphorylation of intracellular signaling molecules that may enhance gene expression (41). For example, the p53 protein contains consensus sites for PKC-dependent phosphorylation at the carboxyl terminus and phosphorylation of serine/threonine residues at this domain promotes transcriptional activity (42). Recent work has provided persuasive evidence for transcriptional regulation of the myocyte RAS by p53 (24). It seems reasonable to advance the hypothesis that a similar mechanism may be operating to provoke glucose-induced release of ANG II from cardiac myocytes. Future investigations directed at the phosphorylation status of p53 and the putative PKC isozyme(s) implicated in the activation of this transcription factor will be required to test the validity of this hypothesis.
To determine whether high glucose−induced release of ANG II from cardiac muscle cells is coupled with the activation of ANG II signal transduction pathways that promote PKC translocation, the selective AT-1R antagonist, losartan, was added to cultures of freshly isolated cardiac myocytes maintained at 25 mmol/l glucose concentration. Western blot analysis indicated that losartan completely reversed glucose-induced translocation of PKC-β1, -β2, -δ, and -ε. Conversely, the selective AT-2R antagonist, PD 123,319, did not alter the redistribution of PKC isozymes in response to 25 mmol/l glucose. Importantly, translocation of PKC-ε, which exhibited resistance to D609, genistein, and BAPTA/AM, was completely reversed by losartan, underscoring the dependence of this isozyme on signaling pathways coupled to the AT-1R.
The finding that 25 mmol/l glucose promotes translocation of PKC-ζ in adult rat ventricular myocytes is novel and raises questions concerning the signaling events that trigger redistribution of this isozyme. PKC-ζ is categorized as an atypical isozyme that lacks a Ca2+-binding domain and is not activated by DAG or phorbol esters (10). Activating molecules for PKC-ζ include phosphatidylserine, phosphatidylinositol triphosphate, and ANG II (32,43,44). Under conditions of high ambient glucose concentration, translocation of PKC-ζ appears to occur independent of signaling events coupled to the AT-1R.
Activation of PKC and high glucose−induced phosphorylation of Tn I.
Translocation and activation of PKC isozymes may not be concurrent events, and the extent and duration of PKC redistribution may not be an index of PKC activity (45). To determine whether exposure of cardiac muscle cells to 25 mmol/l glucose promotes both translocation and activation of PKC isozymes, we examined the phosphorylation status of Tn I. Although this sarcomeric protein is known to be a substrate for several PKC isozymes, PKC-β2 and -ε have been reported to preferentially target Tn I in the diabetic myocardium (14,41,46). PKC-ε, the predominant isozyme in adult rat ventricular myocytes, associates with the sarcomere on activation (47) and has been functionally linked to phosphorylation of Tn I, resulting in reduced myofibrillar ATPase activity in vitro (14). Alternatively, PKC-β2 has also been reported to phosphorylate cardiac Tn I, with a resultant decrease in myofilament Ca2+ sensitivity (46). In vitro studies performed with Tn I mutants (34) have identified several phosphorylation sites for PKC isozymes and have indicated that the critical residues responsible for reduced Ca2+ sensitivity and ATPase activity are located at Ser-43/Ser-45. The results of the present study indicated that adult rat ventricular myocytes exposed to 25 mmol/l glucose exhibit an upregulation of Tn I phosphorylation. Stoichiometrically, the Tn I protein was found to exhibit a comparatively greater increase in serine phosphorylation. Interestingly, losartan completely reversed glucose-induced serine phosphorylation but had no detectable effect on the phosphorylation of threonine residues. The latter observation implies that activated PKC isozymes differentially target serine/threonine residues on Tn I, with the AT-1R−dependent isozymes PKC-β1, -β2, -δ, and -ε exclusively phosphorylating Tn I serine residues, whereas Tn I threonine sites appear to be targeted by PKC-ζ or other unknown kinases.
Although we focused on the interaction between PKC and the Tn I protein, a high ambient glucose concentration promotes the redistribution of several PKC isozymes, potentially affecting multiple biological functions of the diabetic myocardium. For example, PKC-ε has been reported to confer protection against cell death by hypoxia (48) and to promote ischemic preconditioning of cardiac myocytes (49). The latter properties may prove to be extremely valuable in determining the balance between cell death and cell survival in the diabetic heart. Alternatively, glycosylation of proteins is well documented in diabetes and serves as an important mechanism for the posttranslational modification of proteins (50). Recently, a specific gene regulating enzymatic glycosylation, core 2 N-acetylglycosaminyltransferase, has been shown to be induced by diabetes and hyperglycemia. Moreover, in genetically engineered mice, selective overexpression of this transgene in the myocardium induces the hypertrophy phenotype (51). These emerging concepts emphasize the complexity of events that participate in the expression of the diabetic cardiac phenotype.
The present study had certain limitations. First, the duration of exposure to 25 mmol/l glucose was brief when compared with an in vivo model of chronic hyperglycemia. Second, although beyond the scope of this investigation, single-cell mechanics were not performed to document that glucose-induced covalent modification of the Tn I protein results in the anticipated decreases in myofilament Ca2+ sensitivity. Third, we did not determine whether one or all of the AT-1R−dependent PKC isozymes target Tn I serine residues. Finally, the application of an in vitro system to study the problem of diabetic cardiomyopathy may not mimic the in vivo condition. Future investigations directed at issues not addressed in the present study will be important in clarifying the relationship of glucose-induced signaling events to the initiation and progression of cardiac disease in diabetes.
In conclusion, a high ambient glucose concentration is a potent stimulus for the translocation/activation of cardiac PKC isozymes. In adult rat ventricular myocytes, 25 mmol/l glucose promotes PKC translocation via the activation of multiple intracellular signal transduction pathways and through the release of endogenous ANG II. The latter peptide, by means of an autocrine signaling pathway, selectively promotes the translocation of PKC-β1, -β2, -δ, and -ε, which target critical serine residues implicated in the activation of Tn I. This is the first demonstration that endogenous release of ANG II by cardiac muscle cells activates an AT-1R−dependent signaling pathway that phosphorylates a myofilament regulatory protein implicated in the biochemical and mechanical abnormalities that characterize the diabetic myocardium. Finally, the results of the present study suggest that AT-1R blockade may offer a selective advantage in the clinical management of diabetic cardiomyopathy.
This work was partially supported by an American Heart Association Grant-in-Aid (AHA 97–50856A to A.M.), a research grant from the Foundation of University of Medicine and Dentistry of New Jersey (UMDNJ) Annual Grants Program (to A.M.), and support from Summit Area Public Foundation through the generosity of Mrs. Elaine B. Burnett Fund.
The authors express their gratitude to Dr. W.G. Johanson, Chairman, Department of Medicine, UMDNJ-New Jersey Medical School, for his encouragement and support. We thank Merck and Du Pont (Rahway, NJ) for their generous gift of losartan.
Address correspondence and reprint requests to Leonard G. Meggs, M.D., Director, Division of Nephrology and Hypertension, MSB I-524, Department of Medicine, UMDNJ-New Jersey Medical School, 185 South Orange Ave., Newark, NJ 07103. E-mail: firstname.lastname@example.org.
Received for publication 5 October 2000 and accepted in revised form 23 April 2001.
ANG, angiotensin; AT-1R, ANG II type 1 receptor; AT-2R, ANG II type 2receptor; BAPTA/AM, 1,2-bis(aminophenoxy)ethane-N,N,N′,N′-tetraaceticacid; DAG, diacylglycerol; D609, tricyclodecan-9-yl-xanthogenate; DTT, dithiothreitol; ELISA, enzyme-linked immunosorbent assay; HRP, horseradish peroxidase; PKC, protein kinase C; PMA, phorbol 12-myristate 13-acetate; PMSF, phenylmethylsulfonyl fluoride; RAS, renin angiotensin system; SFM, serum-free medium; STZ, streptozotocin; Tn, troponin.