Calcineurin Increases Cardiac Transient Outward K+ Currents via Transcriptional Up-regulation of Kv4.2 Channel Subunits*

Fast transient outward potassium currents (Ito,f) are critical determinants of regional heterogeneity of cardiomyocyte repolarization as well as cardiomyocyte contractility. Additionally, Ito,f densities are markedly down-regulated in cardiac hypertrophy and heart disease, conditions associated with activation of the serine/threonine phosphatase calcineurin (Cn). In this study, we investigated the regulation of Ito,f expression by Cn in cultured neonatal rat ventricular myocytes (NRVMs) with and without α1-adrenoreceptor stimulation with phenylephrine (PE). Overexpression of constitutively active Cn in NRVMs induced hypertrophy and caused profound increases in Ito,f density as well as Kv4.2 mRNA and protein expression and promoter activity, without affecting Kv4.3 or KChIP2 levels. The effects of Cn on hypertrophy, Ito,f, and Kv4.2 transcription were associated with NFAT activation and were abrogated by NFAT inhibition. Despite activating Cn and inducing hypertrophy in NRVMs, PE resulted in profound down-regulation of Ito,f densities as well as Kv4.2, Kv4.3, and KChIP2 expression. Although hypertrophy and NFAT activation were inhibited by the Cn inhibitory peptide CAIN, Ito,f and Kv4.2 expression were further reduced by CAIN, whereas Cn overexpression eliminated PE-induced reductions in Ito,f and Kv4.2 expression without affecting Kv4.3 or KChIP2 levels. We conclude that Cn increases cardiac Ito,f densities by positively regulating Kv4.2 gene transcription. Consistent with this conclusion, we found that Ito,f was increased in myocytes isolated from young mice overexpressing Cn prior to the development of heart disease. This positive regulation of Kv4.2 transcription by Cn activation is expected to minimize the reductions in Ito,f and Kv4.2 expression observed in hypertrophic cardiomyocytes.

results establish that Cn overexpression strongly enhanced I to,f and Kv4.2 transcription in an NFAT-dependent manner. In addition, Cn overexpression reversed I to,f reductions induced by ␣ 1 AR stimulation with phenylephrine (PE) via increases in Kv4.2 mRNA and protein, whereas Cn inhibition with the Cn inhibitory peptide CAIN (21) caused further reductions in I to,f and Kv4.2 transcription observed with PE treatment. These results demonstrate that Cn/NFAT is a positive regulator of I to,f via transcriptional up-regulation of Kv4.2 in myocytes and thus minimizes the degree of I to,f reductions induced in cardiac hypertrophy.

EXPERIMENTAL PROCEDURES
Isolation, Culture, and Intervention of NRVMs-Day 1 or 2 Sprague-Dawley NRVMs (Charles River Laboratories, Inc., Montreal, Canada) were isolated and cultured in medium containing 5% fetal bovine serum in the presence of 0.1 mmol/liter bromodeoxyuridine (Sigma) following removal of non-myocytes by pre-plating (37°C, 1 h) as described previously (25). After 24 h, the myocytes were cultured in the absence of serum (and supplemented with insulin-transferrin-selenium-X (Invitrogen), 25 g/ml ascorbic acid, and 1 nmol/liter LiCl) and infected with adenoviruses (green fluorescent protein (GFP), CAIN, constitutively active Cn, ⌬NFATc3, or VIVIT (a peptide consisting of Met-Ala-Gly-Pro-His-Pro-Val-Ile-Val-Ile-Thr-Gly-Pro-His-Glu-Glu designed to inhibit Cn-mediated NFAT activation (32); VIVIT adenoviruses were kindly provided by Dr. S. D. Kraner) at a multiplicity of infection of 5-10 for 16 h. Myocyte cultures were then washed and incubated in the presence or absence of the ␣ 1 AR agonist PE (100 mol/liter) for 48 h. Infection efficiency was typically Ͼ98%, with minimal cell death at these levels of multiplicity of infection. Myocytes were plated on laminincoated coverslips at a density of 1.5 ϫ 10 5 cells/ml for experiments involving immunofluorescence staining and electrophysiological recordings and at 5 ϫ 10 5 cells/ml for all other experiments.
Immunofluorescence Staining of NRVMs-Immunoconfocal fluorescence was used to assess myocyte size and numbers. To perform these experiments, cultured cells were fixed and permeabilized with methanol (Ϫ20°C, 15 min). After washing with phosphate-buffered saline, cells were incubated with 1:200 anti-␣-actinin monoclonal antibody (sarcomeric; Sigma) in phosphate-buffered saline with 1% bovine serum albumin at room temperature for 1 h. Phosphate-buffered saline-washed cells were subsequently incubated at room temperature for 30 min with 1:400 Alexa Fluor 488-labeled goat anti-mouse IgG antibody (H ϩ L; Molecular Probes) and viewed under a confocal microscope. Typically, Ͼ98% of the cells were ␣-actininpositive 96 h after NRVM isolation, confirming the dominance of myocytes in the culture.
[ 3  Cells were then incubated with 10% trichloroacetic acid at 4°C for 2 h, followed by cell lysis with 1 N NaOH. Each of the lysed samples was added to 7 ml of ScintiSafe TM 30% liquid scintilla-tion mixture (Fisher) in a 20-ml polyethylene scintillation vial (Fisher) and subjected to liquid scintillation counting.
Chord conductance (G) was calculated according to the following equation: G ϭ I/(V m Ϫ E K ), where I is I to,f recorded in response to step depolarization to V m and E K represents the Nernst potential for K ϩ ions. Fits of the relationship between G and V m to the Boltzmann equation allow estimation of the maximal chord conductance (G max ) and the voltage for half-maximal activation of I to,f (i.e. V1 ⁄ 2 ). Slope conductance (G slope ) was estimated by linear regression of the relationship between peak I to,f and V m at step voltages ranging from ϩ10 to ϩ60 mV.
Western Immunoblot Analysis-Cells were collected and lysed in Laemmli buffer. Equal amounts of protein (determined using the Bio-Rad RC DC assay) were loaded, separated on 12% SDS-polyacrylamide gels, transferred onto polyvinylidene difluoride membranes, and probed using rabbit polyclonal antibodies against Kv4.2 or Kv4.3 (Chemicon) or KChIP2 (Affinity BioReagents). GAPDH was detected by a mouse monoclonal antibody (Research Diagnostics, Inc.). Anti-rabbit or antimouse antibodies conjugated with horseradish peroxidase (Amersham Biosciences) were used as secondary antibodies to allow protein visualization with enhanced chemiluminescence (Amersham Biosciences).
Isolation of the KChIP2 Promoter and Cloning into the pGL3-Basic Vector-Rat genomic DNA was isolated from Sprague-Dawley rat livers using previously described methods (27). The KChIP2 promoter region (Ϫ1524 to ϩ312) was isolated from rat genomic DNA by Advantage TM -GC genomic PCR (BD Biosciences) using primers designed according to the nucleotide sequence of NW_043419. The primers used were 5Ј-TAAGC-TAGCTGTGGGGTGCATCATCTCTATTCA-3Ј (forward) with the addition of an NheI site and 5Ј-AACCCCGGGAAC-TATCACCGACTCACCCGTAAG-3Ј (reverse) with the addition of an SmaI site at the 5Ј-end. Following purification, the amplified DNA was digested with NheI and SmaI and subcloned into the pGL3-Basic vector (Promega Corp.). Positive clones containing the KChIP2 promoter were confirmed by nucleotide sequencing and Dual-Luciferase assay.
Transfection and Promoter Activity Measurements-For promoter experiments, myocytes were transfected 16 h after viral infection with pGL3 with a firefly luciferase reporter gene driven by the Kv4.2 (kindly provided by Dr. K. Takimoto) (28), Kv4.3 (kindly provided by Dr. E. S. Levitan) (29), or KChIP2 promoter using Lipofectamine 2000 (Invitrogen). pRL-TK encoding Renilla luciferase was cotransfected as a control to correct for transfection efficiency. Cell lysates were obtained 48 h after transfection to measure Dual-Luciferase activity (Promega Corp.). In the experiments to measure the activation of Cn/NFAT, NRVMs were transfected with pGL3 with a firefly luciferase reporter gene driven by a promoter containing three NFAT cis-elements (kindly provided by Dr. A. Rao).
Statistical Analysis-All data are expressed as the means Ϯ S.E. Statistical significance was calculated using Student's t test to compare two groups and analysis of variance to compare multiple groups. p values Ͻ0.05 were considered statistically significant.

Cn Increases I to,f via NFAT-dependent Increases in Kv4.2
Transcription-To examine the regulation of I to,f by Cn, we overexpressed constitutively active Cn in NRVMs. As reported previously (30,31), Cn overexpression induced marked myocyte hypertrophy as determined by cell microscopy, cell capacitance data, and [ 3 H]leucine uptake experiments (Fig. 1A). Rather than decreasing I to,f , as might be expected from previous studies (17,18), Cn caused a 2-fold increase in I to,f density, which was quantified as the difference in the peak current minus the sustained current (I sus ) at the end of the pulse divided by the cell capacitance. I to,f density recorded at ϩ60 mV ( Fig. 1, B and C; and Table 1) was increased in myocytes overexpressing Cn compared with control myocytes. As expected, G slope , which gives a more direct measure of the extent of I to,f enhancement, was also increased in myocytes overexpressing Cn compared with control myocytes, along with increased G max of I to,f ( Table 1). The increases in I to,f density and conductance were not accompanied by shifts (p ϭ 0.57) in the voltage dependence of I to,f activation as assessed by estimates of the voltages required for I to,f to reach 50% of the maximal conductance (V1 ⁄ 2 ) (data not shown). In addition, there were no changes in the time course of I to,f inactivation (data not shown). The effect of Cn overexpression on I to,f appeared to be specific because the current density remaining at the end of the voltage step (i.e. I sus ) B, representative families of outward potassium current density recorded in myocytes. Currents were recorded in response to step depolarization from Ϫ40 to ϩ60 mV (increments of 10 mV) for 450 ms from a holding potential of Ϫ80 mV in myocytes. A brief pre-pulse (50 ms) to Ϫ40 mV was applied to eliminate activated sodium current. I to,f density (pA/pF) was determined by dividing I to,f by the cell capacitance, where I to,f was estimated as the difference between the peak and the sustained outward currents. C, I to,f density plotted as a function of the step potential. D, chord conductance (G) plotted as a function of the step potential after fitting data to the Boltzmann equation. Error bars represent the means Ϯ S.E. of six cells/group. *, p Ͻ 0.05 versus GFP.
was not changed (p Ͼ 0.50) in myocytes by overexpression of Cn (Table 1).
To test whether the increases in I to,f induced by Cn were associated with changes in the expression of genes encoding I to,f , we measured the expression of Kv4.2, Kv4.3, and KChIP2. As shown in Fig. 2, Cn overexpression increased Kv4.2 mRNA and protein, without affecting Kv4.3 or KChIP2 expression compared with control cells. To investigate whether the increases in Kv4.2 mRNA were related to increased transcription activity (versus changes in mRNA stability), NRVMs were transfected with Kv4.2 (28), Kv4.3 (29), and KChIP2 promoter constructs. Fig. 3A shows that overexpression of Cn increased the activity of the Kv4.2EB promoter (Ϫ3162 to ϩ592). This increase in Kv4.2EB promoter activity, as well as the changes in I to,f , appeared to be mediated by NFAT dephosphorylation because I to,f recorded at ϩ60 mV in NRVMs upon overexpression of Cn was reduced to 8.71 Ϯ 2.25 pA/picofarad (pF) (n ϭ 6) when co-infected with adenoviral VIVIT (32), a selective inhibitory peptide of Cn-mediated NFAT dephosphorylation. VIVIT overexpression alone did not alter I to,f (11.52 Ϯ 1.86 pA/pF, n ϭ 5). Cn-mediated Kv4.2 promoter activity was abolished by coinfection with VIVIT, which alone had no effect on basal Kv4.2 promoter activity (Fig. 3A), and Cn overexpression induced a 6-fold increase in NFAT activity (Fig. 3B), whereas overexpression of ⌬NFATc3 (i.e. constitutively active NFATc3) (33) stimulated Kv4.2EB promoter activity by similar amounts (Fig. 3A). By contrast, neither Cn nor ⌬NFATc3 overexpression altered the activity of the Kv4.3 (Ϫ2337 to ϩ54), Kv4.3 (Ϫ663 to ϩ54), or KChIP2 (Ϫ1524 to ϩ312) promoter (data not shown), consistent with mRNA and protein results (Fig. 2). These results establish that Cn increases I to,f via NFAT-dependent increases in Kv4.2 transcription.
Cn Minimizes I to,f Reductions following Chronic ␣ 1 AR Stimulation-Because several previous studies have suggested that Cn activation in cardiac hypertrophy is responsible for reductions in I to,f (8,21), we investigated the regulation of I to,f by Cn in an in vitro hypertrophic model induced by ␣ 1 AR stimulation. To achieve this, we overexpressed either CAIN or constitutively active Cn in NRVMs in the presence of 100 mol/ liter PE for 48 h. As expected from a previous study (8), PE treatment induced marked myocyte hypertrophy based on cell microscopy, cell capacitance data, and [ 3 H]leucine uptake experiments (Fig. 4A) and increased NFAT activity by 4-fold compared with myocytes not treated with PE (Fig. 4B). In addition, overexpression of CAIN (21), a specific inhibitory peptide of Cn, inhibited NFAT activation and prevented hypertrophy following treatment with PE, confirming the essential role of Cn signaling in ␣ 1 AR-induced hypertrophy. Consistent with previous studies (8,34), Fig. 4 (C and D) and Table 1 show that I to,f densities along with G slope and G max were decreased in PEtreated myocytes compared with control myocytes without PE. However, contrary to expectations based on studies concluding that Cn is responsible for I to,f reductions (17,18), CAIN overexpression in PE-treated myocytes further reduced I to,f densi-

and I sus recorded in NRVMs
Data are the means Ϯ S.E. The values were calculated from the current recordings illustrated in Figs. 1 and 4. G slope is estimated from the slope conductance of I to,f (between ϩ10 and ϩ60 mV). G max is the maximal conductance estimated from plots of I to,f /(V m Ϫ E K ) (see "Experimental Procedures"). n indicates the number of cells studied.  (Fig. 6). Overexpression of CAIN had no effect on the basal mRNA and protein levels of Kv4.2, Kv4.3, and KChIP2 (data not shown), but further reduced Kv4.2 mRNA and protein expression induced by PE treatment, without affecting Kv4.3 and KChIP2 expression (Figs. 5 and 6). On the other hand, overexpression of Cn following PE treatment increased Kv4.2 mRNA and protein levels compared with the control while having no effect (p Ͼ 0.12) on Kv4.3 and KChIP2 mRNA and protein levels (Figs. 5 and 6). It is important to note that the differences in Kv4.3 mRNA between PE and CAIN/PE or between PE and Cn/PE as shown in Fig. 5B are not significant. At first glance, it may seem surprising that the level of I to,f in the PE-treated myocytes expressing Cn was not increased (p ϭ 0.53) above that in the control (i.e. myocytes not treated with PE) given that Cn induced a nearly 2-fold increase in Kv4.2 protein expression above the control when overexpressed in PE-treated myocytes. However, this lack of concordance between I to,f levels and Kv4.2 protein levels is consistent with the large reductions in Kv4.3 and KChIP2 expression because rodent I to,f channels are formed primarily as heterotetramers of Kv4.2 and Kv4.3 channels (1) and because I to,f channels require KChIP2 for membrane insertion of functional channels (35). Therefore, taken together, these results establish that Cn is a positive regulator of Kv4.2 expression and I to,f density when I to,f is reduced by PE.

GFP
The above results establish that I to,f is positively regulated by Cn in cultured neonatal myocytes. To explore whether Cn also regulates I to,f in intact myocardium, we measured I to,f in myo-  cytes isolated from mice overexpressing Cn. We performed these experiments in mice that were 3-4 weeks of age to reduce the complexities associated with cardiac hypertrophy and heart disease that become apparent in older mice (30). As summa-rized in Fig. 7 and Table 2, myocytes isolated from mice overexpressing Cn in the heart had higher I to,f densities and G slope than age-matched non-transgenic littermate control mice. These results demonstrate that I to,f is positively regulated by Cn in intact adult myocardium.

DISCUSSION
Our study has confirmed that overexpression of constitutively active Cn induces hypertrophy. Surprisingly, the Cn-induced hypertrophy in cultured NRVMs was associated with marked increases in I to,f that correlated strongly with increases in the mRNA, protein, and promoter activities of Kv4.2, but not the other molecular constituents of I to,f (i.e. Kv4.3 and KChIP2). The increases in Kv4.2 mRNA were associated with enhanced promoter activity, suggesting that Cn induces increases in Kv4.2 transcription activity versus altered mRNA stability. Positive regulation of I to,f by Cn was also observed in very young (3-4 weeks old) mice overexpressing Cn, suggesting that this positive regulation of I to,f is not specific to cultured neonatal myocytes. Sixteen hours after viral infection, NRVMs were treated with 100 mol/liter PE for 48 h. RNA extracted from these myocytes was subjected to reverse transcription and subsequent real-time PCR to measure mRNA levels by the standard curve method. GAPDH was used as an internal control. The mRNA level normalized to the GAPDH level in NRVMs infected with adenoviral GFP (control (CTL)) was set as 100. Data are presented as the means Ϯ S.E. (n ϭ 5). *, p Ͻ 0.05 versus CTL; ␥, p Ͻ 0.05 versus PE.  DECEMBER 15, 2006 • VOLUME 281 • NUMBER 50

JOURNAL OF BIOLOGICAL CHEMISTRY 38503
Consistent with previous studies concluding that the hypertrophy induced by Cn requires dephosphorylation of NFAT (probably NFATc3 and NFATc4) (30,36), which leads to nuclear translocation and transcriptional alterations in gene expression (30), NFAT inhibition (with VIVIT) blocked both the Cn-induced hypertrophy and the enhancement of I to,f and Kv4.2 promoter activity. On the other hand, changes in NFAT did not affect Kv4.3 or KChIP2 promoter activity. 5 These results establish that Cn activation increases I to,f via transcriptional enhancement of Kv4.2, but not Kv4.3 or KChIP2. Our findings also suggest that Cn/NFAT has no effect on Kv4.2 transcription or I to,f under basal conditions because VIVIT alone had no effect on I to,f or Kv4.2 promoter activity, whereas CAIN alone did not alter I to,f or Kv4.2 expression.
It is important to note that only the 3.5-kb Kv4.2EB promoter (Ϫ3162 to ϩ592) activity was increased by Cn and ⌬NFATc3, whereas the shorter promoters Kv4.2HB (Ϫ1094 to ϩ592) and Kv4.2SB (Ϫ432 to ϩ592) (28) were not activated by Cn or ⌬NFATc3. These observations are consistent with the existence of 13 consensus NFAT-binding sites (i.e. GGAAA cis-elements) between Ϫ3162 and Ϫ1095 in the Kv4.2 promoter versus only three NFAT-binding sites between Ϫ1094 and 0. Similarly, the inability of Cn or ⌬NFATc3 to increase Kv4.3 or KChIP2 transcription correlates with the presence of only six putative NFAT sites in the Kv4.3 promoter (Ϫ2337 to ϩ54) and five NFAT sites in the KChIP2 promoter (Ϫ1524 to ϩ312). Although the ability of Cn and NFAT to increase Kv4.2 promoter activity correlates with the number of NFAT-binding sites, NFAT activation can be modulated by other transcription factors (37), which could also conceivably contribute to or be responsible for the enhanced Kv4.2 transcription by Cn.
We further investigated the role of Cn in the regulation of I to,f in an in vitro hypertrophic model of cardiomyocytes induced by ␣ 1 AR stimulation. Our experiments confirmed that chronic ␣ 1 AR stimulation in cultured NRVMs decreased I to,f in association with reductions in Kv4.2, Kv4.3, and KChIP2 mRNA and protein levels (8,34). We have also demonstrated that Cn inhibition (with CAIN overexpression) blocked ␣ 1 AR-mediated hypertrophy in cultured NRVMs, as reported previously (21). Previous studies established that hypertrophy induced by ␣ 1 AR stimulation in cultured NRVMs is associated with Cn activation (8,21) and I to,f reductions (8). Similarly, Cn activation (19,38,39) and I to,f reductions (10,11,14,18) are commonly observed in various animal models of cardiac hypertrophy and heart disease as well as in heart disease patients (40,41). Collectively, these observations have led to the suggestion that Cn activation is responsible for the reductions in I to,f (17,18) as well as hypertrophy. However, rather unexpectedly (17,18), our results showed that, in cultured NRVMs chronically treated with ␣ 1 AR agonists, inhibition of Cn (using CAIN) caused further reductions in I to,f , despite blocking hypertrophy, whereas overexpression of constitutively active Cn induced large increases in I to,f density without inhibiting myocyte hypertrophy. These changes in I to,f were associated with changes in mRNA and protein expression of Kv4.2, but not Kv4.3 or KChIP2, a pattern identical to that observed in the absence of ␣ 1 AR stimulation. Thus, in the presence of cardiac hypertrophy, we conclude that Cn activation will minimize the degree of reductions in I to,f by driving Kv4.2 transcription (see the proposed model in Fig. 8).
In contrast to our findings, a previous study showed that inhibition of Cn with cyclosporin or knock out of NFATc3 prevents I to,f reductions induced by myocardial infarction in adult mice (18). The differences between our experiments using NRVMs and those using infarcted (18) or normal (42) adult mouse hearts may reflect age-dependent differences in signaling or differences between cultured isolated myocytes and intact myocardium. Alternatively, because the reductions in I to,f following myocardial infarction in adult mice also require ␤AR stimulation (18), it is plausible that ␤AR and ␣ 1 AR stimulation activates distinct signaling pathways capable of modulating the net effects of Cn. The potential differential Cn-mediated regulation of I to,f activated by ␣ 1 AR and ␤AR stimulation is in keeping with the complex regulation of Cn signaling in the myocardium. For example, Cn/NFAT-mediated tumor necro-FIGURE 7. Typical K ؉ currents obtained from cardiomyocytes dissociated from the ventricular apices of 3-4-week-old non-transgenic (NTG) and Cn-overexpressing transgenic (Cn TG) mice. A, representative families of outward potassium currents recorded in dissociated cardiomyocytes in response to step depolarization from Ϫ40 to ϩ60 mV (increments of 10 mV) for 4000 ms from a holding potential of Ϫ80 mV. I to,f was estimated by fitting the decay phase of outward K ϩ currents with bi-or triexponential function using Clampfit Version 9.0, and I to,f density (pA/pF) was determined by dividing I to,f by the cell capacitance (Cm). B, I to,f density plotted as a function of the step potential. G slope values (between ϩ10 and ϩ60 mV) were compared by Student's t test between groups. *, p Ͻ 0.05 versus NTG.

TABLE 2 Electrophysiological properties of I to,f recorded in cardiomyocytes dissociated from the ventricular apices of 3-4-week-old nontransgenic and Cn-overexpressing transgenic mice
Data are the means Ϯ S.E. The values were calculated from the current recordings illustrated in Fig. 7. G slope is estimated from the slope conductance of I to,f (between ϩ10 and ϩ60 mV). n indicates the number of cells studied. sis factor-␣ expression and cardiomyocyte hypertrophy are suppressed by interaction between HDAC4 and the HSP40 protein Mrj (33). The reductions in I to,f in failing hearts from Cn transgenic mice (23,43) also suggest that Cn activation reduces I to,f . However, because heart disease causes I to,f reductions (13)(14)(15) and because these Cn transgenic mice develop heart disease by 6 weeks of age (30), the interpretation of these studies requires caution. Consistent with I to,f being reduced as a consequence of heart disease induced by Cn in transgenic mice, Kv4.2 expression and I to,f were reported to be unchanged in younger mice (6 -7 weeks of age) despite marked hypertrophy (23). Consequently, we investigated I to,f in 3-4-week-old Cn transgenic mice. Consistent with our cultured myocyte results, I to,f is indeed elevated in young mice overexpressing Cn, establishing that Cn is also a positive regulator of I to,f in intact adult myocardium.
A key conclusion of our study is that Cn activation is not responsible for the reductions in I to,f or Kv4.2, Kv4.3, and KChIP2 expression induced by ␣ 1 AR stimulation in NRVMs. Moreover, Cn overexpression in PE-treated myocytes increased I to,f levels, but only back to the levels observed in control myocytes not treated with PE. This Cn-mediated increase in I to,f back to control levels occurred despite the fact that Cn increased Kv4.2 protein expression by 2-fold (and Kv4.2 mRNA transcription by ϳ5-fold) above the control in PE-treated myocytes. Returning the I to,f to control levels in PEtreated myocytes is not unexpected because Cn had no effect on expression of Kv4.3 or KChIP2, which are both required for co-assembly with Kv4.2 as heterotetramers (1) to form functional I to,f channels (35). Additional studies will clearly be required to determine the cellular mechanisms underlying the reductions in Kv4.2, Kv4.3, and KChIP2 expression induced by ␣ 1 AR stimulation, which activates numerous hypertrophic signaling pathways (44).
The enhanced I to,f by Cn observed in our study may represent an important feedback pathway operating under conditions of cardiac hypertrophy and heart disease. Specifically, as already mentioned, many hypertrophic stimuli, including ␣ 1 AR stimulation and heart disease, reduce I to,f as well as Kv4.2, Kv4.3, and KChIP2 expression (8,45,46). These conditions have also been linked to Cn activation (8,45,47), which is expected from our results to limit the extent of I to,f down-regulation. Preventing excessive I to,f reductions may be of critical importance because changes in I to,f can lead to altered electrical and contractile properties of the myocardium, which may contribute to impaired pump function and arrhythmias (6, 48 -51). For example, we (48,52) and others (53) have shown previously that not only is I to,f an important determinant of action potential profile changes occurring in heart disease and regional differences in action potential profile, but I to,f is also a major regulator of the amplitude and kinetics of cardiac contraction in rodents and larger mammals by modulating Ca 2ϩ entry via L-type Ca 2ϩ channels. Thus, Cn activation not only induces hypertrophy, but also modulates the extent of the electrical and contractile changes induced as a result of I to,f reduction caused by hypertrophic stimuli. In addition, a recent study found that increased I to,f by adenoviral overexpression of Kv4.3 can prevent hypertrophy (45), whereas increased I to,f in NRVMs treated with PE prevents both myocyte hypertrophy and Cn activation (8), suggesting that, at least in these models, increased I to,f by Cn could limit the degree of hypertrophy.
In conclusion, we have found that Cn/NFAT potently increases Kv4.2 transcription and I to,f levels in rodent myocardium. This regulation provides a potentially useful mechanism for modulating the cellular, electrical, and contractile responses of the myocardium to cardiac hypertrophy and heart disease. The relevance of Cn-mediated regulation of Kv4.2 to larger mammalian species such as humans and dogs is uncertain because, in these species, I to,f is encoded by Kv4.3 and KChIP2 subunits. Additional studies will be required to completely define the role of Cn in the regulation of I to,f in non-rodents.