TRPC6 fulfills a calcineurin signaling circuit during pathologic cardiac remodeling (original) (raw)
Increased expression of TRPC6 in hypertrophic and failing hearts. To investigate whether changes in TRPC expression participate in calcineurin-dependent cardiac growth (12), we examined the expression of Trpc transcripts by real-time RT-PCR in hearts from Tg mice harboring an α-MHC–calcineurin transgene, which results in profound cardiac hypertrophy. We detected cardiac expression of Trpc1, -3, -4, and -6, and among these, Trpc6 mRNA was significantly upregulated in ventricles of calcineurin Tg mice (Figure 1A). As TRPC2 is a pseudogene in humans (8), we did not test its expression. Among other TRPC family members, Trpc5 and -7 transcripts were not detected in the heart, consistent with previous studies (27, 28).
Increased mRNA expression of TRPC6 in models of cardiac hypertrophy and human failing heart. (A) RNA was isolated from hearts of WT and calcineurin Tg mice (Cn-Tg) at 10 weeks of age, and mRNA levels of Trpc1, -3, -4, and -6 determined by real-time RT-PCR as described in Methods. Shown are mRNA levels relative to WT normalized by 18S RNA levels. *P < 0.05 versus WT. (B) The correlation between Trpc6 and ANP mRNA expression in ventricles was examined by linear regression analysis. r = 0.953. P < 0.0001. (C) Trpc6 mRNA levels were determined by real-time RT-PCR in RNA isolated from hearts of mice subjected to TAB or sham operation for 3 weeks. Shown are mRNA levels relative to sham-operated ventricles normalized by 18S RNA levels. *P < 0.05 versus sham. (D) Trpc6 mRNA levels were determined by real-time RT-PCR in RNA from primary neonatal rat ventricular myocytes treated with or without ET-1 for 24 hours. Shown are mRNA levels relative to vehicle-treated controls normalized by 18S RNA levels. *P < 0.05 versus vehicle. (E) TRPC6 mRNA levels were determined by real-time RT-PCR in total RNA isolated from human hearts with dilated cardiomyopathy (DCM) or normal human hearts. Shown are mRNA levels relative to normal hearts (assigned as 1.0) normalized by 18S levels. *P < 0.05 versus normal hearts.
Expression of Trpc6 mRNA correlated closely with that of atrial natriuretic peptide (ANP), a sensitive marker of cardiac stress and hypertrophy (Figure 1B). Cardiac expression of Trpc6 was also upregulated in mice subjected to TAB (Figure 1C). In addition, Trpc6 mRNA was increased in cultured rat ventricular myocytes in response to ET-1, a potent prohypertrophic agonist (Figure 1D). TRPC6 mRNA expression was also augmented in hearts of human patients with dilated cardiomyopathy compared with nonfailing hearts (Figure 1E). We conclude that TRPC6 is upregulated in response to diverse pathologic stimuli that promote cardiac hypertrophy and remodeling.
TRPC6 is directly regulated by calcineurin-NFAT signaling. To begin to define the mechanism that regulates TRPC6 expression during pathologic cardiac growth, we analyzed the 5′ flanking region of the mouse Trpc6 gene, which is highly conserved among different mammalian species. Within this region, we identified 2 sequences between –791 and –763 that resembled the NFAT consensus-binding site ([A/T]GGAAA[A/N][A/T/C]N), which were relatively conserved among humans and rodents (Figure 2A).
The calcineurin-NFAT pathway regulates Trpc6 gene transcription. (A) Schematic representation of the 5′ upstream region of the mouse Trpc6 gene. Sequences between –796 and –750 bp of the mouse Trpc6 gene are aligned with the corresponding sequences of rat Trpc6 and human TRPC6 genes. NFAT-like sites are shown in red and blue. (B) Gel mobility shift assay was performed using in vitro translated NFATc4Δ317 and 32P-labeled probes of NFAT consensus sequences or NFAT-like sites 1 or 2 from the Trpc6 promoter. (C) Rat neonatal ventricular myocytes were cotransfected with an expression plasmid of NFATc4Δ317 and –913TRPC6-luc, –617TRPC6-luc, or the Trpc6 promoter with NFAT site mutations (–913mutNFAT1-luc, –913mutNFAT2-luc, or –913mutNFAT1+2-luc). Percent increase in expression with NFATc4Δ317 compared with empty pGL3 luciferase reporter vector is shown. (D) Rat neonatal ventricular myocytes were cotransfected with a plasmid expressing CnAΔC and –913TRPC6-luc, –617TRPC6-luc, or the Trpc6 promoter with 2 NFAT site mutations. Percent increase in expression with CnAΔC compared with empty pGL3 luciferase reporter vector is shown. (E) Rat neonatal ventricular myocytes were cotransfected with either –913TRPC6 promoter region or luciferase reporter gene fused to the Trpc6 promoter with 2 NFAT site mutations in the presence or absence of 10 nM ET-1. Percent increase in luciferase expression by ET-1 with each construct is shown.
NFATc4 translated in vitro bound the labeled NFAT-like sequences from the TRPC6 promoter, and binding was completely eliminated by the presence of unlabeled NFAT consensus sequences (Figure 2B). Similarly, unlabeled oligonucleotides representing either of the Trpc6 NFAT-like sequences effectively competed for binding of NFATc4 to a canonical NFAT consensus sequence (Figure 2B). The NFATc4-DNA complex was supershifted by an NFATc4 antibody, confirming the presence of NFATc4 in the complex.
To determine whether NFAT can activate the TRPC6 promoter, a DNA fragment extending to –913 bp relative to the translation initiation site of the mouse Trpc6 gene was fused to a luciferase reporter (–913TRPC6-luc) and cotransfected with an expression vector encoding constitutively active NFATc4 (NFATc4Δ317) in cardiomyocytes. Transcriptional activity of the TRPC6 promoter was significantly increased in the presence of NFATc4, and deletion of the region containing the NFAT binding sites (deletion of –913 to –617 in the TRPC6 promoter; –617TRPC6-luc) reduced NFAT responsiveness. Site-directed mutations in either of the NFAT-like sites of the TRPC6 promoter partially reduced, and mutations of both sites nearly abolished, the response to NFATc4Δ317 (Figure 2C). Moreover, overexpression of constitutively active calcineurin (CnAΔC) activated the TRPC6 promoter, and deletion of the region comprising the NFAT sites or site-directed mutations of both NFAT sites abolished the responsiveness of the promoter to calcineurin (Figure 2D). The TRPC6 promoter was also activated by ET-1, which activates calcineurin signaling, and the NFAT sites were required for this response (Figure 2E).
Increased expression of TRPC6 enhances NFAT activity. Because the TRPC family participates in receptor-mediated Ca2+ entry, which activates the calcineurin-NFAT pathway, we examined whether increased expression of TRPC6 in cardiomyocytes was capable of activating NFAT-dependent transcription. Indeed, overexpression of TRPC6 resulted in redistribution of NFATc4-GFP from the cytoplasm to the nucleus in transfected COS-1 cells, as typically occurs in response to calcineurin activation (Figure 3, A and B). TRPC6 expression also activated a luciferase reporter controlled by the promoter region of exon 4 of the RCAN1 gene, which contains 15 NFAT sites (29), in transfected COS-1 cells (Figure 3C). RCAN1, a calcineurin inhibitor (16), inhibited activation of the RCAN1-luciferase reporter by TRPC6 (Figure 3C). TRPC6 overexpression also significantly increased RCAN1-luciferase activity in cardiomyocytes; ET-1 acted synergistically with TRPC6 to activate the reporter (Figure 3D), and RCAN1 blocked this activation (Figure 3E).
TRPC6 activates NFAT-dependent transcription in ventricular myocytes. (A) COS-1 cells were cotransfected with the expression vectors for NFATc4-GFP (green) and TRPC6 or empty vector, fixed, permeablized, and immunostained with anti-TRPC6 (red). Nuclei are stained blue (DAPI). Magnification, ×400. (B) Effects of TRPC6 on NFATc4 subcellular distribution were quantified with microscopic examination of greater than 100 cells per condition. N, NFATc4-GFP localized exclusively in nucleus; N>C, nuclear NFATc4-GFP localization exceeds cytoplasmic; N≤C, cytoplasmic NFATc4-GFP localization equals or exceeds nuclear; C, NFATc4-GFP localized exclusively in cytoplasm. (C) COS-1 cells were cotransfected with _RCAN1-_luciferase and TRPC6 and RCAN1 expression vectors. Fold activation over _RCAN1-_luciferase without expression plasmids is shown. (D) Myocytes were cotransfected with _RCAN1-luciferase reporter plasmid and a plasmid expressing rat TRPC6 at various doses in the presence or absence of ET-1. Fold activation over RCAN1-luciferase alone is shown. *P < 0.05 versus control; †_P < 0.05 versus control, ET-1, and TRPC6 (200 ng). (E) Myocytes were cotransfected with _RCAN1-_luciferase and plasmids expressing rat TRPC6 and RCAN1. Fold activation over _RCAN1_-luciferase without expression plasmids is shown. *P < 0.05 versus TRPC6 alone. (F) Rat smooth muscle cells transfected with rat TRPC6 siRNA. Quantitative RT-PCR for TRPC1–TRPC7 was performed on RNA. Trpc5 and -7 mRNA were not detectable. Percent change of Trpc gene expression in cells transfected with TRPC6 versus control siRNA is shown. (G) Myocytes were cotransfected with _RCAN1_-luciferase and TRPC6 siRNA alone or with ET-1 or PE. Fold activation over _RCAN1-_luciferase with vehicle alone is shown. Control values were assigned as 1.0.
To examine the potential involvement of endogenous TRPC6 in hypertrophic signaling, cardiomyocytes were cotransfected with RCAN1-luciferase and siRNAs for rat TRPC6 in the presence or absence of ET-1 or PE, an α-adrenergic hypertrophic agonist. Addition of TRPC6 siRNA specifically suppressed expression of TRPC6, but not the other TRPC transcripts (Figure 3F). Knockdown of endogenous TRPC6 significantly reduced the inducible expression of RCAN1-luciferase by ET-1 or PE (Figure 3G), suggesting that G protein–coupled receptor agonists activate the calcineurin-NFAT pathway through TRPC6.
Increased expression of TRPC6 induces pathologic hypertrophy in vivo. To investigate the potential consequences of increased cardiac expression of TRPC6 in vivo, we generated Tg mice expressing rat TRPC6 in a heart-specific manner using the α-MHC promoter. Three independent Tg mouse lines with high, intermediate, and low levels of transgene expression were obtained (Tg L23, Tg L16, and Tg L8, respectively; Figure 4, A–C). Immunocytochemistry using TRPC6 antibody showed that overexpressed TRPC6 protein was preferentially localized to the membrane of ventricular myocytes isolated from Tg hearts (Figure 4D). To confirm that functional TRPC6 channels were properly processed and targeted to the cardiomyocyte cell membrane, we measured cationic current across the membrane using electrophysiologic techniques (Figure 4, E and F). We found that current density was significantly larger in ventricular myocytes isolated from Tg L16 hearts compared with WT littermates. The peak inward current was 3.8 pA/pF in Tg versus 2.5 pA/pF in WT cardiomyocytes, and the peak outward current was 23.3 pA/pF in Tg versus 11.7 pA/pF in WT cardiomyocytes.
Expression of TRPC6 in hearts of Tg mice. (A) Rat Trpc6 mRNA levels, shown relative to Tg L8 mice and normalized by 18S RNA levels, were determined using RT-PCR. (B) Western blot analysis for TRPC6 and β-actin expression in hearts. (C) Relative TRPC6 protein expression in WT (assigned as 1.0) and TRPC6 Tg mice, determined by densitometry. (D) Immunocytochemistry of adult myocytes isolated from TRPC6 Tg L16 (12 wk) and WT littermates using TRPC6 antibody. Green, anti-TRPC6; red, anti–α-actinin. Magnification, ×400. (E) TRPC current was measured by electrophysiologic voltage-clamp in ventricular myocytes isolated from Tg L16 and WT hearts. Current-voltage relations revealed increased current density in Tg L16 myocytes, consistent with increased functional TRPC channel expression at the cell surface. Voltage ramp protocol imposed on cells is inset. (F) Mean values of peak TRPC current density recorded in Tg L16 versus WT myocytes. *P < 0.05. (G) Mean values of percent TRPC6 current increase on exposure to 10 nM ET-1. (H) Cytoplasmic (C) and nuclear (N) proteins immunoblotted with anti-NFATc3, anti-PCAF (nuclear marker), or anti-HSP90α/β (cytoplasmic marker). (I) RCAN1 (exon 4) mRNA levels were determined by real-time RT-PCR using RNA from hearts of 7- to 8-week-old WT, Tg L8, Tg L16, and Tg L23 mice. mRNA levels relative to RCAN1 mRNA in WT (assigned as 1.0) normalized by 18S RNA are shown.
To further evaluate the increased expression of functional TRPC6 channels in Tg myocytes, we activated the TRPC receptors by treating cells with ET-1 and performed electrophysiologic recordings (Figure 4G). In these experiments, ET-1–induced increases in both inward and outward currents in Tg cardiomyocytes were significantly greater than in WT cardiomyocytes. These data are consistent with increased expression of functional TRPC6 channels in Tg cardiomyocytes.
Further evidence that functional TRPC6 channels are expressed in Tg mice was provided by the observation that NFATc3 accumulated in the nucleus (Figure 4H) and was hypophosphorylated (data not shown) in TRPC6 Tg L16 ventricles compared with ventricles of WT littermates. Consistent with these findings, the expression of RCAN1 (exon 4) mRNA, a sensitive marker of calcineurin signaling, was significantly increased (Figure 4I), even in Tg L8, the mouse line with the lowest expression of TRPC6. The majority of mice derived from the high-expressing line Tg L23 died 5–12 days after birth (Figure 5A) and exhibited increased heart weight/body weight (HW/BW) ratios at 7 days after birth (Figure 5, B–D). Histologic examination of α-MHC–TRPC6 Tg hearts revealed enlarged atria and ventricles with heterogeneity of myocyte size (Figure 5E), suggesting severe cardiomyopathy as the cause of death. In contrast, Tg L16 mice, with intermediate TRPC6 expression, showed no increase in HW/BW ratios at 8 weeks of age, but they developed cardiomegaly and congestive heart failure around 30 weeks of age (Figure 5, F–H). Histologic analyses of these mice revealed cardiac dilatation with heterogeneity of myocyte size and interstitial fibrosis as well as congestion of the lungs, indicative of heart failure (Figure 5, I and J). We conclude that increased expression of TRPC6 in the heart can evoke pathologic cardiac remodeling leading to cardiomyopathy in vivo.
Cardiomyopathy in TRPC6 Tg mice. (A) Kaplan-Meier survival analysis of TRPC6 Tg L23 (n = 10 per group). (B and C) HW/BW ratios (B) and BW (C) in WT and TRPC6 Tg L23 mice (n = 6 per group). (D) Gross hearts of WT and TRPC6 Tg L23 mice at 7 days of age. (E) Histologic analysis of hearts from TRPC6 Tg L23 and WT mice at 7 days of age. Hearts were sectioned longitudinally and stained with H&E. (F) HW/BW ratios in WT and TRPC6 Tg L16 mice (n = 3 per group). (G) HW/tibia length (HW/TL) ratios in WT and TRPC6 Tg L16 mice at 30 weeks of age (n = 3 per group). (H) TRPC6 Tg L16 and WT mice at 30 weeks of age. (I) Histologic analysis of hearts from TRPC6 Tg L16 and WT mice at 30 weeks of age. Hearts were sectioned longitudinally and stained with H&E or Masson’s trichrome (bottom panels). Scale bar: 5 mm. (J) Histologic analysis of lungs from TRPC6 Tg L16 and WT mice at 30 weeks of age. Lungs were sectioned longitudinally and stained with H&E. *P < 0.05 versus WT. Magnification, ×40 (E, bottom panels; I, right panels; and J).
Tg L8, the mouse line with the lowest expression of TRPC6, showed no difference in HW/BW ratios compared with WT littermates until at least 20 weeks of age (Figure 6A). Echocardiography also demonstrated no significant difference in ventricular wall thickness or systolic function between these Tg mice and WT littermates (Table 1). To determine whether expression of TRPC6 sensitizes the heart to stress, we subjected these animals to pressure overload by TAB and found that they exhibited an exaggerated response to pressure overload, with a dramatic increase in HW/BW ratios and decreased systolic function compared with WT littermates (Figure 6, B and C, and Table 1). These results indicate that increased expression of TRPC6 sensitizes the heart to pathologic hypertrophic signaling, leading to cardiac dysfunction.
Accelerated pathologic remodeling in response to TAB in TRP6 Tg L8 mice. (A) HW/BW ratios in WT and TRPC6 Tg L8 mice at 8 and 20 weeks of age (n = 6 per group) and (B) subjected to TAB (n = 3 per group). *P < 0.05 versus WT. (C) Gross hearts of WT and TRPC6 Tg L8 mice after 3 weeks of TAB. Hearts were sectioned longitudinally and stained with H&E. Scale bars: 5 mm.
Echocardiographic analysis of TRPC6 Tg L8 and WT mice subjected to TAB
Regulation of β-MHC expression by the TRPC6-calcineurin-NFAT pathway. Although mouse line Tg L8 showed no obvious hypertrophic phenotype until 20 weeks of age in the absence of stress, β_-MHC_ mRNA expression was significantly increased and brain natriuretic peptide (BNP) mRNA expression was moderately increased, while expression of ANP, skeletal α-actin, α-MHC, and sarco/endoplasmic reticulum Ca2+-ATPase isoform2 (SERCA2) was not significantly altered in the hearts of these mice (Figure 7A). These results suggest that signaling pathways downstream of TRPC6 specifically control β_-MHC_ gene expression. Indeed, β-MHC mRNA expression correlated with the level of TRPC6 transgene expression in the different lines of α-MHC–TRPC6 Tg mice (Figure 7B). These results were corroborated by real-time RT-PCR, which showed β_-MHC_ mRNA expression to be upregulated compared with ANP and BNP in hearts of α-MHC–calcineurin Tg mice at different ages (Figure 7, C and D). These results suggest that TRPC6-calcineurin-NFAT signaling preferentially regulates β_-MHC_ gene expression.
Induction of β_-MHC_ gene expression in TRPC6 Tg hearts. (A) Expression of cardiac genes β_-MHC_, BNP, α_-MHC_, ANP, α_-skeletal actin_ (SKA), and SERCA2 in hearts isolated from in TRPC6 Tg L8, the lowest-expressing line. Percent change in relative mRNA levels normalized by 18S RNA levels of Tg mouse hearts compared with WT littermates is shown. n = 4 per group. *P < 0.001, #P < 0.05 versus WT. (B) Expression of β_-MHC_ gene in hearts of different Tg TRPC6 mouse lines. Shown are β_-MHC_ mRNA levels relative to WT (assigned as 1.0) normalized by 18S RNA levels. n = 4 per group, except for 8-wk Tg L23 (n = 1). (C) HW/BW ratios of WT and calcineurin Tg mice at 2 and 10 weeks of age. n = 4 per group. (D) Expression of β_-MHC_, ANP, and BNP in calcineurin Tg mice at 2 and 10 weeks of age. Fold increase in mRNA levels over WT (assigned as 1.0) normalized by 18S RNA levels is shown. n = 4 in each group.
To explore the mechanistic basis of the sensitivity of the β_-MHC_ gene to calcineurin-NFAT signaling, we fused a –396 bp promoter region of the mouse β_-MHC_ gene to a luciferase reporter (–396β-MHC–luc) (30). This promoter region contains multiple MCAT-binding sites, 1 GATA site, and 1 NFAT site (31, 32) (Figure 8A). The GATA and NFAT sites are conserved in β_-MHC_ promoters of different mammalian species, and GATA and NFAT factors have been reported to bind to these sites (31, 32). TRPC6, as well as constitutively active NFAT3 (NFATc4Δ317) and CnAΔC, activated the –396β-MHC–luc reporter in cardiomyocytes (Figure 8B). The stimulatory activity of TRPC6 or NFATc4Δ317 was reduced when either the GATA or NFAT sites in the promoter were mutated. In addition, double mutation of both GATA and NFAT sites abolished responsiveness to TRPC6 or NFATc4, demonstrating that the TRPC6-calcineurin-NFAT pathway directly activates β_-MHC_ gene transcription through these sites (Figure 8C). TRPC6 expression also enhanced ET-1–inducible activation of –396β-MHC–luc (Figure 8D). The TRPC6-, ET-1–, and PE-induced activation of –396β-MHC–luc was significantly inhibited in the presence of calcineurin inhibitors cyclosporin A (Figure 8, E and G) and RCAN1 (Figure 8, F and H), indicating that the calcineurin-NFAT pathway is involved in –396β-MHC–luc activation by multiple hypertrophic stimuli (Figure 8, E–H). Furthermore, the increase in β_-MHC_ gene expression in TRPC6 Tg L16 mice was significantly attenuated in double Tg mice expressing both TRPC6 and the calcineurin inhibitor RCAN1 in the heart (Figure 8I) (33).
The TRPC6-calcinerurin-NFAT pathway directly regulates β_-MHC_ gene transcription. (A) Sequence of the mouse β_–MHC_ promoter. MCAT, NFAT, and GATA binding sites are shown. (B and C) Rat neonatal ventricular myocytes were cotransfected with (B) –396β-MHC–luc and an expression plasmid of TRPC6, CnAΔC, or NFATc4Δ317 or (C) –396β-MHC–luc with NFAT or GATA mutants and an expression plasmid encoding NFATc4Δ317 or TRPC6 (bar graphs show percent increase or change in luciferase activity by NFATc4Δ317 or rat TRPC6 compared with control vector); (D–F) –396β-MHC–luc and an expression vector encoding TRPC6 in the presence or absence of (D) ET-1, (E) cyclosporin A, or (F) an expression plasmid of RCAN1; (G, H, and J) –396β-MHC–luc in the presence or absence of ET-1, PE, or (G) cyclosporin A, (H) an expression vector encoding RCAN1, or (J) control or TRPC6 siRNAs. (I) Relative expression of β-MHC mRNA was measured using real-time RT-PCR in hearts of WT, β-MHC–TRPC6 (Tg), and β-MHC–TRPC6;β-MHC–RCAN1 (double Tg) mice. Fold change over control –396β-MHC–luc activity (assigned as 1.0) is shown.
To examine the role of endogenous TRPC6 in hypertrophic stimuli–inducible β_-MHC_ gene transcription, cardiomyocytes were cotransfected with –396β-MHC–luc and siRNAs for rat TRPC6 in the presence or absence of PE or ET-1. As shown in Figure 8J, TRPC6 knockdown reduced the induction of –396β-MHC–luc activity by PE or ET-1. Collectively, these findings indicate that TRPC6 significantly contributes to the inducible expression of the β_-MHC_ gene in the cardiac hypertrophic response.