Maladaptive cardiac hypertrophy results in phenotypic changes in several genes that are thyroid hormone responsive, suggesting that thyroid hormone receptor (TR) function may be altered by cellular kinases, including protein kinase C (PKC) isozymes that are activated in pathological hypertrophy. To investigate the role of PKC signaling in regulating TR function, cultured neonatal rat ventricular myocytes were transduced with adenovirus (Ad) expressing wild-type (wt) or kinase-inactive (dn) PKCα or constitutively active (ca) PKCδ and PKCε. Overexpression of wtPKCα, but not caPKCδ or caPKCε, induced a 28-fold increase (P < 0.001) in TRα1 protein in the nuclear compartment and a smaller increase in the cytosol. Furthermore, TRα1 mRNA was increased 55-fold (P < 0.001). This effect of PKCα was dependent on its kinase activity because dnPKCα was without effect. Phorbol 12-myristate 13-acetate (PMA) induced nuclear translocation of endogenous PKCα and Ad-wtPKCα concomitantly with an increase in nuclear TRα1 protein. In contrast, PMA-induced nuclear translocation of dnPKCα resulted in a decrease of TRα1. The increase in TRα1 protein in Ad-wtPKCα-transduced cardiomyocytes was not the result of a reduced rate of protein degradation, nor was the half-life of TRα1 mRNA prolonged, suggesting a PKCα-mediated effect on TRα transcription. Although phosphorylation of ERK1/2 was increased in Ad-wtPKCα-transduced cells, inhibition of phospho-ERK did not change TRα1 expression. PKCα overexpression in cardiomyocytes caused marked repression of triiodothyronine (T3)-responsive genes, α-myosin heavy chain, and the sarcoplasmic reticulum calcium-activated adenosinetriphosphatase SERCA2. Treatment with T3 for 4 h resulted in significant reductions of PKCα in nuclear and cytosolic compartments, and decreased TRα1 mRNA and protein, with normalization of phenotype. These results implicate PKCα as a regulator of TR function and suggest that nuclear localization of PKCα may control transcription of the TRα gene, and consequently, affect cardiac phenotype.
- extracellular signal-regulated kinase
- sarcoplasmic reticulum calcium-activated adenosinetriphosphatase
- myosin heavy chain
maladaptive cardiac hypertrophy results in phenotypic changes that include genes that are known to be thyroid hormone responsive, such as α- and β-myosin heavy chain (MHC), sarcoplasmic reticulum calcium-activated ATPase (SERCA2), phospholamban, and β-adrenergic receptors (reviewed in Refs. 13, 25). This observation has provided support of the hypothesis that thyroid hormone effects on the myocardium may be impaired in the failing heart, which may include alterations in nuclear thyroid hormone receptors (TRs) and their transcriptional coregulators, or alternatively, a decrease in serum or tissue levels of thyroid hormones. Studies of patients with ischemic and dilated cardiomyopathy have reported contradictory results of both increases and decreases in the TR mRNA isoforms, and TR protein data are lacking (7, 23, 39). An experimental animal model of pathological cardiac hypertrophy has shown that mRNA levels of three TR isoforms were downregulated, and it was concluded that this resulted in the hypertrophic phenotype (24). Recent identification of specific thyroid hormone transporters in the myocardium raises the possibility that alterations in these transporters under pathological conditions may affect hormone uptake and thus cardiac-specific thyroid hormone function (14). Alternatively, tissue content of thyroid hormones may be decreased by induction of thyroid hormone-degrading deiodinases as has been shown in a rodent model of cardiac hypertrophy and failure (43). Thus the role of thyroid hormones in the setting of heart disease has received considerable attention (20, 33).
Recent studies have provided evidence that thyroid hormones activate intracellular signaling pathways, including the ERK (9) and phosphatidylinositol 3-kinase (PI3K)-Akt pathways (5). Additionally, a role for TRα1 in the cytoplasmic compartment has been recently reported, supporting extranuclear functions for TRs (8, 22). Therefore, it was of interest to consider an interactive role for thyroid hormones, TRs, and the protein kinase C (PKC) family of serine-threonine kinases that have been implicated in most intracellular signaling pathways regulating cardiomyocyte hypertrophy, contractile function, and gene expression (12, 29, 36). Studies of cardiac pathological hypertrophy and failure in humans and animal models have reported increased activity of calcium-dependent isotypes, PKCα and PKCβ (2, 42). Although PKCα has recently been associated with myocyte hypertrophic growth (3, 41), in vivo studies using transgenic models have shown that PKCα may play a potentially more important role in modifying systolic and diastolic contractile function than inducing hypertrophy (4, 16). Calcium-independent PKCε and PKCδ have been identified as downstream mediators of Gq protein-coupled receptor-initiated signals; however, each isotype activates distinct intracellular responses (6). PKCε appears to selectively activate the ERK cascade that is implicated in myocyte growth responses and cell survival, whereas PKCδ preferentially activates JNK and p38MAPK pathways, resulting in apoptosis and contractile dysfunction (17, 19, 31, 38). Thyroid hormone has been reported to cause subcellular redistribution and activation of PKCα (1), whereas another study reported repression of PKCα and PKCε expression in response to thyroid hormone treatment in cultured cardiomyocytes (35). Thus, on the basis of a potentially significant interaction of PKCs and thyroid hormone, we proposed to examine the effects of PKCα, -δ, and -ε isozymes on TR function.
Using adenoviral gene transfer to overexpress PKCα in cultured cardiomyocytes, we show for the first time that nuclear localization of PKCα induced the expression of TRα1 in both nuclear and cytosolic compartments. This effect was not observed with overexpression of constitutively active PKCδ or PKCε. The potential physiological effects of high levels of unliganded TRα1 within the pathological myocardium of patients with heart disease, which is often associated with low circulating levels of thyroid hormones (33), may induce the “failing” phenotype and also alter cytoplasmic nongenomic activities of TRs, including activation of the ERK, p38MAPK, and PI3K/AKT pathways (5, 9, 22). Thus therapeutic approaches directed at cardiac TRα1 function or activity of the PKCα signaling pathway may prove efficacious in attenuating the dysfunction associated with pathological cardiac hypertrophy.
MATERIALS AND METHODS
Cell culture reagents were obtained from GIBCO-BRL (Grand Island, NY). l-Triiodothyronine (T3), phorbol 12-myristate 13-acetate (PMA), cyclohexamide (CHX), and actinomycin D were obtained from Sigma (St. Louis, MO). MEK1 inhibitor PD-98059 was from Calbiochem (San Diego, CA). Anti-TRα1 (PA1–211A) antibodies were from Affinity BioReagents (Golden, CO). Antibodies to TRβ1 (J52), PKCα, PKCδ, and PKCε, and TFIID (TATA box binding protein, TBP) were from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish peroxidase-conjugated goat anti-mouse and anti-rabbit secondary antibodies were purchased from Bio-Rad Laboratories (Hercules, CA). Chemiluminescent reagents for Western blot analysis and radioactive chemicals were from PerkinElmer (Boston, MA). Real-time quantitative PCR reagents were from Eurogentec (Belgium). All other reagents, including protease and phosphatase inhibitors, were of the highest available quality from Sigma, Calbiochem, or Pierce (Rockford, IL).
Isolation, culture, and viral transduction of neonatal rat ventricular myocytes.
Animals were treated in accordance with National Institutes of Health Guide for the Care and Use of Laboratory Animals (DHHS Publication No. 85–23), and study protocols were approved by the Institutional Animal Care and Use Committee. Ventricular myocytes were isolated from hearts of 2-day-old rats by collagenase digestion as we have previously described (21). Myocytes were plated at ∼1.5 × 104/cm2 on collagen-coated six-well plates (RNA analysis) or 60-mm dishes (protein analysis) and cultured for the first 20 h in DMEM/F12 medium containing 10% FBS, l-glutamine, cytosine β-d-arabinofuranoside (10 μM), and antibiotics. After the first 20 h, the neonatal rat ventricular myocyte (NRVM) cultures were washed with HBSS, and in select experiments, cells were exposed to adenovirus (at a multiplicity of infection as indicated in results) in DMEM/F12 for 1 h in the cell culture incubator (5% CO2, 37°C). Subsequently, the cells were washed and maintained in serum-free medium containing transferrin (5 mg/l), selenium (5 μg/l), l-glutamine, and antibiotics for 24 or 40 h before experimentation as indicated. Reagents used in cell culture were added to the final concentrations as follows: T3 (10−8 M), PMA (2 × 10−7 M), PD-98059 (10−5 M), actinomycin D (5 μg/ml), and CHX (10 μg/ml).
Replication-defective adenoviruses containing coding sequences of wild-type PKCα (Ad-wtPKCα), kinase-inactive mutant of PKCα (Ad-dnPKCα), constitutively active PKCδ (Ad-caPKCδ) and PKCε (Ad-caPKCε) were kindly provided by Dr. Allen Samarel (Loyola Univ., Maywood, IL) (17, 41). Adenovirus for nuclear-encoded β-galactosidase (Ad-neβgal) was used to control for nonspecific effects of adenoviral transduction of cultured cells. The recombinant adenoviruses were amplified by sequential infection of HEK-293 cells and purified by CsCl gradient ultracentrifugation. The viral titer of each preparation of replication-defective adenovirus was determined by dilution assay in HEK-293 cells.
Cell fractionation and immunoblot analysis.
For cytosolic and nuclear protein analysis, cells were washed and then scraped into homogenization buffer containing 20 mM HEPES, pH 7.9, 10 mM KCl, 1 mM EDTA, 10% glycerol, 0.2% NP-40 plus protease and phosphatase inhibitors. Cells were lysed by Dounce homogenization and fractionated by centrifugation at 12,000 g for 1 min at 4°C. The resulting supernatant was used as the cytosolic fraction. The nuclear fraction was prepared by resuspension of the pellet in buffer containing 20 mM HEPES, pH 7.9, 0.42 M NaCl, 0.2 mM EDTA, 1.5 mM MgCl2, 25% glycerol, and protease inhibitor cocktail, and incubated with agitation for 30 min at 4°C before centrifugation for 10 min at 12,000 g. The resulting supernatants were collected and used for nuclear protein analysis.
For studies of PKC translocation in response to PMA or T3 treatment, a commercially available extraction kit (Pierce) was used to isolate cytosolic, membrane, and nuclear fractions. Cells were prepared essentially as described by Vijayan et al. (41) in which cells were initially lysed by freeze-thaw in a homogenization buffer containing 20 mM Tris, pH 7.4, 2 mM EDTA, 2 mM EGTA plus protease and phosphatase inhibitors. The cell lysate was fractionated by centrifugation for 2 min at 500 g, 4°C, and the resulting supernatant was used as the cytosolic fraction. The pellet was resuspended in buffers provided by the manufacturer and further fractionated to obtain membrane and nuclear fractions. Protein concentrations in all fractions were determined by Micro-BCA assay (Pierce). After normalization of sample protein content, proteins were resolved by electrophoresis on 5% stacking/10% polyacrylamide-SDS gels, and the resolved proteins were transferred to nitrocellulose membrane (Bio-Rad) and probed with primary antibodies and horseradish peroxidase-conjugated secondary antibodies as previously described (21). Primary antibodies used were polyclonal anti-TRα1, and monoclonal antibodies against TBP, TRβ1, PKCα, PKCδ, and PKCε. Secondary antibodies used were either goat anti-rabbit or goat anti-mouse IgG conjugated with horseradish peroxidase. Protein bands were detected using chemiluminescence reagent and visualized by exposure to X-ray film. Protein band intensity was quantified by laser scanning densitometry (GS-800 Calibrated Densitometer; Bio-Rad). Results are expressed as arbitrary densitometric units (du).
Real-time quantitative RT-PCR.
RNA samples were prepared from ∼106 cells plated in six-well plates using a commercially available kit (RNeasy Mini Kit; Qiagen, Valencia, CA) and were treated with deoxyribonuclease during purification. The integrity of the RNA was verified before analysis. Real-time quantitative reverse-transcriptase-PCR utilized TaqMan chemistry and an ABI Prism 7700 Sequence Detector (Perkin Elmer-Applied Biosystems, Foster City, CA). Primers and TaqMan probes for TRα1, TRβ1, and GAPDH mRNAs were synthesized using Applied Biosystems 394 DNA Synthesizer in the Institutional Core Facility as we have previously published (21). GAPDH mRNA was used for normalization of each sample, and analysis of each specific mRNA was conducted in duplicate. Relative expression of the mRNAs was calculated as previously described (21), and results are expressed as fold change with respect to the corresponding experimental control.
Northern blot analysis was used for quantitation of expression of cardiac α- and β-MHC and SERCA2 mRNAs. Five micrograms of total RNA were fractionated by denaturing agarose gels and blotted onto synthetic membrane as we have previously published (34). Detection of α- and β-MHC mRNAs was accomplished using radiolabeled oligodeoxynucleotide probes, whereas a cloned SERCA2 cDNA fragment was radiolabeled by random-priming methodology to detect rat SERCA2 mRNA as described (34). Ribosomal RNA (18S rRNA), measured by radiolabeled oligodeoxynucleotide probe, was used for sample normalization. Specific mRNAs were detected by exposure to X-ray film.
All data are presented as means ± SE derived from three to six individual samples per experiment derived from a minimum of two separate cell preparations. One-way ANOVA was used for statistical analysis of mean values among experimental groups, with pairwise multiple comparisons conducted using the Holm-Sidak method. Differences between means were considered significant at P < 0.05. Data were analyzed using SigmaStat 3.1 (Systat Software, Richmond, CA).
PKCα overexpression significantly increases TRα1 protein.
In light of recent studies showing nuclear translocation of PKCα in response to PMA and insulin-like growth factor-1 (41), we explored the possibility that intracellular signaling through the PKCα pathway may alter gene expression by modulating expression or activity of thyroid hormone receptors TRα1 and TRβ1. As shown in Fig. 1A, transduction of cultured cardiomyocytes with replication-deficient Ad-wtPKCα resulted in a significant increase in both nuclear and cytosolic TRα1 protein compared with myocyte cultures transduced with control Ad-neβgal. Furthermore, by increasing the multiplicity of infection (moi) of Ad-wtPKCα from 5 to 25, immunoreactive PKCα was increased from 1.7- to 4.5-fold compared with Ad-neβgal (Fig. 1, B and C), concomitantly with an increase in TRα1 protein from 28-fold to 63-fold compared with Ad-neβgal (5 moi) (Fig. 1, B and D). In the present studies, we routinely used 5 moi of Ad-wtPKCα, which provided an increase of ∼2-fold in total immunoreactive PKCα compared with control cultures (Fig. 1C). We have previously published that expression of the TRβ1 isoform is not detectable in cardiomyocytes cultured in the absence of thyroid hormone (21). In the present studies, overexpression of wtPKCα in cardiomyocytes cultured in serum-free medium had no effect on TRβ1 expression (data not shown).
PKCα effect on TRα1 is dependent on its kinase activity and is isozyme specific.
To determine whether the effect of PKCα on TRα1 was dependent on its kinase activity, we overexpressed a kinase-inactive form of PKCα (Ad-dnPKCα) in which the ATP binding site had been mutated (41). At similar levels of expression as the wtPKCα, the kinase-defective mutant dnPKCα had no effect on expression of TRα1 protein (Fig. 2A). Shown for comparison is the level of expression of nuclear TRα1 in cardiomyocytes transduced with control Ad-neβgal. Immunoreactive TATA box binding protein (TBP) shows that equivalent amounts of protein were analyzed per sample.
To address the PKC isoenzyme-specificity of this effect, constitutively active PKCδ and PKCε, were overexpressed in the cardiomyocytes by adenovirus gene transfer. As shown in Fig. 2B, neither PKCδ nor PKCε had an effect on TRα1 protein content, suggesting that this effect was specific to the PKCα isozyme. Furthermore, constitutively active PKCδ, but not PKCε, localized to the nuclear fraction. Thus, despite the similarity in nuclear localization of PKCδ and PKCα, only PKCα increased TRα1 protein expression.
PMA stimulates nuclear translocation of PKCα and increases nuclear TRα1.
We sought to determine whether the increase in TRα1 protein seen in response to overexpression of wtPKCα occurred with the activation of endogenous PKCα. After treatment with PMA to activate PKCα, we used subcellular fractionation to show that endogenous PKCα translocated to the nucleus as had been previously reported (41). As shown by Western blot analysis in Fig. 3, both endogenous PKCα (cells transduced with Ad-neβgal) and adenoviral-expressed wtPKCα and dnPKCα translocated to the nucleus and the membrane fractions, concomitant with a marked decrease in the cytosolic fraction, when treated with PMA for 30 min. Four hours after PMA-induced PKCα nuclear translocation, TRα1 protein in the nucleus was increased in both Ad-wtPKCα- and control Ad-neβgal-transduced cells (Fig. 3). In contrast, TRα1 protein decreased in Ad-dnPKCα-transduced myocytes treated with PMA, suggesting that nuclear translocation of the kinase-inactive dnPKCα attenuated the effect of PMA-activated endogenous PKCα on TRα1 protein.
PKCα-induced TRα1 expression is mediated at the transcriptional level.
Several potential mechanisms could account for the PKCα-induced increase of TRα1 protein, including prolonged TRα1 protein half-life, increased TRα1 mRNA stability, or increased transcription and mRNA production. To address these possibilities we initially tested the hypothesis that the half-life of TRα1 protein was prolonged. Using CHX, an inhibitor of translation, we showed that the rate of degradation of TRα1 protein was not significantly different in cardiomyocytes transduced with Ad-wtPKCα compared with Ad-neβgal (Fig. 4). The estimated half-life of TRα1 was ∼4 h. As seen in Fig. 4, the decrease in TBP after 6 h of treatment with CHX reflects the indiscriminate effect of CHX on inhibition of translation of all proteins. Because TRα1 expression was ∼30-fold higher in Ad-wtPKCα-transduced cells compared with Ad-neβgal, TBP was not detectable under optimum conditions for TRα1 measurement and thus was not used for normalization.
Since we had previously reported that TRα1 protein content was regulated at the level of mRNA (21), we quantified TRα1 mRNA in response to wtPKCα overexpression using real-time PCR and found that it was increased by 55-fold (P < 0.001) (Fig. 5A). Furthermore, PMA stimulation for 4 h increased TRα1 mRNA by an additional 60 ± 2%. TRβ1 mRNA content was unaffected by PKCα overexpression (data not shown).
To determine whether PKCα increased the stability of TRα1 mRNA, we measured mRNA half-life by inhibiting transcription using the RNA polymerase II inhibitor, actinomycin D, as we have previously reported (21). As shown in Fig. 5B, the rate of decay of TRα1 mRNA in cardiomyocytes transduced with Ad-wtPKCα was not prolonged compared with that in control uninfected cells; therefore, increased stability of TRα1 mRNA could not explain the PKCα-induced increase in the mRNA. Itshould be noted that total TRα1 mRNA in Ad-wtPKCα-transduced myocytes was 55-fold higher than uninfected cells; therefore, each decay curve is compared with its own control value after 15-min pretreatment with actinomycin D. These data support a potential role of PKCα on TRα1 transcription either by directly targeting specific transcription factors or by activating other downstream kinases.
PKCα-induced TRα1 expression is not mediated by activated ERK1/2.
Published studies have reported that PKCα activates ERK1/2 in cardiomyocytes (3, 41), suggesting that ERK may have a potential effect on nuclear factor activation. Therefore, we questioned whether the observed PKCα-induced TRα1 expression could be the result of ERK activation. As shown in Fig. 6, significant activation of ERK1/2 and increased TRα1 expression were observed in Ad-wtPKCα- compared with Ad-neβgal-transduced cardiomyocytes. Treatment with a specific MAPK kinase (MEK1) inhibitor, PD-98059, for 4 h significantly reduced ERK1/2 phosphorylation but had no significant effect on nuclear TRα1 content, suggesting that activation of the ERK pathway was unlikely to be involved in PKCα-mediated induction of TRα1.
Effect of T3 treatment on subcellular content and distribution of PKCα.
Previous studies have reported that thyroid hormone treatment of cultured chick embryo hepatocytes caused translocation of PKCα to the membrane fraction and induced PKCα activity (1) and that T3 treatment caused a reduction in PKCα expression in cardiomyocytes (35). Therefore, it was of interest to determine whether T3 had a similar effect on PKCα in the Ad-wtPKCα-transduced cardiomyocytes. The Western blot in Fig. 7A illustrates the subcellular distribution of PKCα after T3 treatment for 30 min, 4 h, and 24 h. Quantitation of these data showed no significant redistribution of PKCα between the soluble and membrane or nuclear fractions; however, PKCα protein decreased significantly by 25% in the cytosolic fraction and 50% in the nuclear fraction after 4 and 24 h of T3 treatment (Fig. 7B). No significant differences in PKCα were observed in any fraction after 30 min of T3 exposure.
Effect of T3 treatment on PKCα-induced TRα1 expression.
Since we had previously published that T3 rapidly decreased TRα1 content in the cardiomyocyte (21), we sought to determine whether the elevated level of expression of TRα1 in Ad-wtPKCα-transduced cells was similarly regulated. As shown in Fig. 8A, T3 treatment for 30 min showed no effect on nuclear TRα1 protein content, but treatment for 4 and 24 h resulted in a 45% and 55% decrease of TRα1 protein, respectively. This decrease in TRα1 protein was paralleled by a significant decrease of TRα1 mRNA (56 ± 4%) after 4 h of T3 treatment (Fig. 8B). Shown for comparison are data from control cells transduced with Ad-neβgal in which T3 treatment for 4 h significantly reduced TRα1 mRNA, as we have previously published (21). In contrast, TRβ1 mRNA was increased by 40–60% after 4 h of T3 treatment in both Ad-wtPKCα- and Ad-neβgal-transduced cardiomyocytes (data not shown).
Cardiac phenotype in response to PKCα overexpression and T3 treatment.
To understand the potential biological consequences of PKCα-induced increase in TRα1 protein in the cardiac myocyte, we examined the expression of several known thyroid hormone-responsive genes, including α- and β-MHCs and SERCA2. As shown in Fig. 9, cardiomyocytes transduced with Ad-wtPKCα showed significant reductions in expression of the positively regulated T3-responsive genes α-MHC and SERCA2 compared with control Ad-neβgal-transduced cells, indicative of an increased repressive role of TRα1 in PKCα-activated myocytes. In contrast, the negatively regulated T3-responsive β-MHC gene was expressed to a similar extent as control. Furthermore, the cardiomyocytes exhibited a robust response to treatment with T3 (10 nM) for 24 h with significant increases in α-MHC and SERCA2 expression and a decrease in β-MHC mRNA expression (Fig. 9).
Genetic studies of mouse models harboring deletions of either TRα or TRβ isoform have provided insight into the specific effects of each receptor isoform on cardiocirculatory function and phenotype. The results largely support a predominant role of TRα in the contractile and electrophysiological functions in the heart (reviewed in Ref. 20). The prominent role of thyroid hormone and its receptors in regulating cardiac physiology is exemplified by both nuclear and extranuclear mechanisms that regulate ion channel activities, calcium transients, contractile function, and transcriptional regulation of genes encoding proteins regulating these functions (reviewed in Refs. 8, 25). Furthermore, because the phenotype associated with pathological hypertrophy and heart failure includes many thyroid hormone-associated functions, substantial interest has been generated in developing TR isoform-specific therapeutic strategies in these disease settings (32).
In the present study, we proposed that TR function in the myocardium could be altered by intracellular protein kinase pathways known to be activated in the diseased heart. PKC-α and PKC-β activity and expression have been shown to be increased in failing human hearts (2). In experimental animal models of cardiac infarction, PKCα, -β,- ε, and -ζ content and calcium-dependent and calcium-independent PKC activities have been reported to be significantly increased in the remodeled left ventricle (42). Furthermore, transgenic models utilizing gain- and loss-of-function approaches of specific PKC isotypes support critical roles for PKCs in intracellular signaling pathways regulating cardiac contractile function, hypertrophy, and phenotype (4, 19, 31). We focused on PKCα, -δ and -ε isoenzymes, reported to lie downstream of G protein-coupled receptors and activated in the pathological hypertrophic cardiomyocyte (11, 12), to interrogate the effects of each isotype on TR activity and cardiomyocyte phenotype.
Our initial observation that a twofold increase in cellular PKCα content resulted in an increase of TRα1 protein by 28-fold was unexpected. No effect was observed on the TRβ1 isoform. Additionally, by progressively increasing the expression of virally transduced PKCα, TRα1 content in both nuclear and cytosolic compartments was further increased. Although we and others have shown that TRα1 distributes in the cytoplasmic fraction when virally transduced into the cardiomyocyte (21, 22), the present data showing significant cytosolic compartmentalization of endogenous TRα1 support a potential role of TRα1 outside the nuclear domain. This effect on TRα1 was specific to the PKCα isotype, as supported by the lack of effect when virally transduced constitutively active PKCδ and PKCε were overexpressed in cardiomyocytes. Furthermore, the kinase activity of PKCα was necessary for the observed increase of TRα1 protein because overexpression of a kinase-inactive mutant of PKCα (Ad-dnPKCα) was ineffective.
We initially investigated several cytoplasmic mechanisms that could result in these observations, including a potential effect of PKCα on prolongation of the half-lives of TRα1 protein and mRNA. RNA analysis revealed a ∼50-fold increase in TRα1 mRNA equivalent to the increase in TRα1 protein, suggesting a potential effect of PKCα on mRNA stability. However, experiments with the RNA polymerase II inhibitor actinomycin D showed that the half-life of TRα1 mRNA in Ad-wtPKCα-transduced myocytes was not increased. Similarly, CHX experiments showed that the half-life of TRα1 protein in Ad-wtPKCα-transduced cells was not sufficiently different from control cultures to explain the increase in protein. Taken together, these data support a mechanism whereby transcription of the TRα gene locus is induced by overexpression of PKCα, suggesting an important role of nuclear localization of PKCα. A recent report has shown that PKCα translocates to the nucleus in response to the phorbol ester PMA (41). Our data support this observation of nuclear translocation of endogenous PKCα as well as adenovirus-expressed PKCα in response to PMA. The PMA-induced nuclear translocation of PKCα, whether endogenous or virally expressed PKCα, resulted in a significant increase of TRα1 protein. In contrast, PMA-induced translocation of virally transduced dnPKCα repressed expression of TRα1 protein to a level below that observed in unstimulated myocytes. Whether endogenous PKCα is present in the nucleus under basal, unstimulated conditions is not clear; however, overexpression of PKCα resulted in detectable amounts in the nuclear compartment, which we hypothesize induced TRα1 expression even in the absence of activation by PMA.
Alternatively, PKCα may indirectly induce TRα1 transcription by activating other downstream kinases, including ERK1/2, as published by others (3, 41). Activation and nuclear translocation of ERKs has been reported in response to thyroxine, supporting a potential role of the ERK pathway in thyroid hormone-mediated nuclear events (9). Alternatively, ERK-mediated phosphorylation of 90-kDa ribosomal S6 kinases (p90 RSK), for example, could potentially activate nuclear transcription factors, including cAMP response element-binding protein, activating transcription factor 1, and c-Fos as has been reported in other cell systems (reviewed in Ref. 15). However, the promoter region of the TRα gene locus has not been extensively characterized (18, 27, 28), and thus the identities of plausible transcription factors as targets of PKCα will require further investigation. To address the possibility that the ERK pathway may be involved in the current observations, we studied the effects of PKCα overexpression on ERK phosphorylation and the effect of ERK inhibition on TRα1 expression. Similar to published reports, phosphorylation of ERK1/2 was increased with PKCα overexpression (3, 41); however, inhibition of the MEK1-ERK1/2 signaling pathway had no effect on expression of TRα1, suggesting that the PKCα effects on TRα1 were unlikely to be mediated by this pathway. Thus identification of potential nuclear substrates of PKCα may shed light on the mechanisms by which transcription of TRα1 is induced and determine how activation of this important intracellular kinase, as occurs in the pathological myocardium, can affect cardiac phenotype and function.
Because our previous study had shown that TRα1 was degraded rapidly in response to T3 treatment, it was of interest to determine how these effects of T3 interacted with the PKCα-mediated increase in TRα1 expression. Additionally, previous reports had provided evidence that thyroid hormone regulated PKCα expression as well as its subcellular localization (1). Although we found no effect of T3 on the subcellular redistribution of PKCα, T3 treatment for 4 h significantly reduced PKCα protein content in the cytosolic and nuclear fractions. Therefore, these data further support the hypothesis that PKCα activates transcription at the TRα gene locus and that the T3-mediated decrease in nuclear PKCα may in part be responsible for the decrease in TRα1 expression.
As we show in this study, PKCα-induced expression of TRα1 in the cardiomyocyte resulted in altered expression of T3-responsive genes similar to that seen in the pathological myocardium. Importantly, treating these cardiomyocytes with thyroid hormone reversed the phenotypic changes due to PKCα overexpression by decreasing TRα1 protein expression. In heart failure in which PKCα has been shown to be activated (2), expression of TRα1 may be increased as a consequence, and with decreased circulating levels of T3, as is often observed in heart failure patients (33), the unliganded TRα1 would act to repress T3-responsive genes, including SERCA2 and α-MHC. Furthermore, with the recent identification of extranuclear or cytoplasmic actions of TRs (5, 9, 22), the effects of unliganded TRα1 in the cytosolic compartment may have as yet undetermined metabolic and functional effects on the cardiomyocyte. Thus results from our studies underscore the importance of the relative concentrations of circulating thyroid hormones and the cellular content of unliganded TRs in determining cardiomyocyte phenotype and function.
This work was supported by National Heart, Lung, and Blood Institute Grant HL-71623.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 by the American Physiological Society