Calcineurin Aβ gene knockdown inhibits transient outward potassium current ion channel remodeling in hypertrophic ventricular myocyte

Abstract It has been shown that the activation of calcineurin is involved in regulating ion channel remodeling in hypertrophic cardiomyocytes. But the precise role of calcineurin in the regulation of transient outward potassium current (I to), an ion channel associated with fatal arrhythmia, remains controversial. This study aimed to examine the effects of calcineurin Aβ (CnAβ) gene knockdown on I to channel remodeling and action potential duration (APD) in the hypertrophic ventricular myocytes of neonatal rats. Results showed that phenylephrine stimulation caused hypertrophy of ventricular myocytes, upregulation of CnAβ protein expression, downregulation of Kv4.2 mRNA and protein expression, a decrease in I to current density, and prolongation of APD. CnAβ gene knockdown significantly inhibited the effects of phenylephrine stimulation. Our data indicate that CnAβ gene knockdown can inhibit I to channel remodeling and APD prolongation in hypertrophic neonatal rat ventricular myocytes. This finding suggests that calcineurin may be a potential target for the prevention of malignant ventricular arrhythmia in a hypertrophic heart.


Introduction
Ventricular remodeling caused by pathological cardiac hypertrophy is chronic congestive heart failure's most important pathophysiological mechanism [1,2]. Sudden cardiac death is the dominant reason of death in patients with congestive heart failure [3]. Ion channel remodeling in ventricular myocytes is the main pathophysiological basis leading to changes in the action potential duration (APD) of ventricular myocytes, resulting in malignant ventricular arrhythmia [4,5]. Congenitally, dysregulation of transient outward potassium current (I to ) has also been demonstrated to play a pivotal role in the Brugada syndrome [6,7].
I to is a rapidly activated and inactivated outward potassium current, mainly involved in phase 1 of action potentials. The activity of I to channels influences the activation of voltage-gated Ca 2+ channels and the balance of inward and outward currents during the plateau, thereby mediating the duration and the amplitude of phase 2. In cultured hypertrophic ventricular myocytes, or the ventricular myocytes from a myocardial infarction heart, downregulated or dysfunctional I to channels of ventricular myocytes lead to delayed repolarization and prolonged APD, which may easily cause fatal arrhythmia [8,9].
In transgenic mice, overexpression of calcineurin results in cardiac hypertrophy and the downregulation of I to , which was reversed by the calcineurin inhibitor cyclosporine [13]. In rats after myocardial infarction, cyclosporine significantly attenuated the decreases in mRNA levels of Kv4.2 and Kv4.3, the components of α subunit in I to channel, and I to density in the left ventricular [14]. In cultured adult canine left ventricular cardiomyocytes, rapid pacing reduced I to density and Kv4.3 mRNA and protein expression, which was markedly prevented by inhibiting calcineurin with cyclosporine [15]. Those results indicated that the activation of calcineurin may lead to I to downregulation. Conversely, there was also evidence in cultured neonatal rat ventricular myocytes that the overexpression of constitutive calcineurin upregulates Kv4.2 expression without affecting Kv4.3 [9].
Thus, the precise role of calcineurin in the regulation of I to remains unclear.
The purpose of this study is to clarify the regulatory effect of calcineurin on I to channel remodeling and APD alterations in the hypertrophic ventricular myocytes of rats by way of knockdown of calcineurin-related genes.

Identification of ventricular myocytes
Cells were cultured on fibrin-coated glass slides for 48 h and α-SCA was detected by immunofluorescence staining. Neonatal rat cardiac fibroblasts were used as a negative control ( Figure A).

Ad-CnAβshRNA sequence screening
Primary ventricular myocytes were cultured for 48 h. Ad-CnAβshRNA1 (A1, interference base sequence 5′-3′: CAG AAAGGGTCTATGAAGCTTGTAT), Ad-CnAβshRNA2 (A2, interference base sequence 5′-3′: CCGCCAGTTTAACTGTT CTCCACAT), Ad-CnAβshRNA3 (A3, interference base sequence 5′-3′: GCAAGATGGCAAGAGTCTTCT), and null at a multiplicity of infection (MOI) of 50 were selected to infect cultured ventricular myocytes for 48 h. CnAβ protein expression in the ventricular myocytes of each group was detected by western blotting. The Ad-CnAβshRNA corresponding to the lowest CnAβ protein expression was regarded as the optimal Ad-CnAβshRNA. A1 caused the most obvious decrease in CnAβ protein expression after its infection of ventricular myocytes at 48 h, and thus it was used in subsequent experiments ( Figure A2).

Determining the effectiveness of intervention on hypertrophy in cultured cells
Cell hypertrophy was identified by the measurement of brain natriuretic peptide (BNP) mRNA expression and cell size after 48 h of intervention. Upon completion of cell grouping and intervention, real-time reverse transcriptionpolymerase chain reaction (RT-PCR) was conducted to determine BNP mRNA expression in ventricular myocytes. Cells were cultured on glass slides. Crystal violet staining assay was performed after grouping and intervention. Three fields of view were randomly selected in each group. The surface area of the cells was assessed using Image J software.

Western blotting
Total protein (40 μg) was loaded and then transferred to a nitrocellulose membrane after electrophoresis. The membrane was blocked with 5% skim milk for 1 h. Rabbit anti-rat GAPDH antibody (1:1,000), rabbit anti-rat rabbit anti-CnAβ antibody (1:500), and rabbit anti-rat rabbit anti-rat Kv4.2 antibody (1:1,000) were added before overnight incubation at 4°C. After thorough rinsing, horseradish peroxidase-labeled secondary antibody was added for incubation at room temperature for 1 h. After thorough rinsing again, color development, photography, and quantitative measurement of the gray scale were performed.

Whole-cell patch clamp detection
A glass microelectrode formed a high resistance seal with the cells and ruptured the membrane. I to was recorded under the voltage clamp mode. Current density analysis was used (current density [pA/pF] = current intensity/ capacitance) to avoid errors caused by cell size. The action potential of the individual cells was recorded under the current clamp mode. The current signal was guided by an Ag/AgCl electrode, amplified by a patch clamp AXON 700B amplifier through an AD/DA converter board, and stored in a computer hard disk. During the experimental procedure, stimulation discharge and signal acquisition were controlled by pCLAMP 10.0 software. In the I to depolarization step, the clamping voltage was set to −80 mV with an −40 mV to +70 mV pulse stimulation series, with a step voltage of 10 mV, wave width of 300 ms, and frequency of 0.2 Hz. The I to steady-state activation curve stimulation protocol was as follows. The clamping voltage was set to −80 mV with a −40 to + 70 mV pulse stimulation series, with a step voltage of 10 mV and a wave width of 300 ms. The I to current was then recorded. The I to steady-state inactivation curve stimulation protocol was as follows. The clamping voltage was set to −80 mV with a −40 mV to +50 mV pulse stimulation series, with a step voltage of 10 mV and a wave width of 1,000 ms. The residual current was then recorded. The steady-state activation and inactivation curves were created using the normalized current values obtained by the two above-mentioned stimulation schemes as the ordinate and the stimulation pulses under different voltages as the abscissa. The semi-activated voltage (V 1/2,act ) and semi-inactivated voltage (V 1/2,inact ) were calculated by curve fitting using the Boltzmann equation The action potential recording method was similar to the voltage clamp mode. After membrane sealing, membrane breaking, and compensation, the recording was switched to the current clamp mode. A 1 nA current pulse was then applied, with a wave width of 2.5 ms, which induced the action potential in the ventricular myocytes. The APDs at 20, 50, and 90% repolarization (APD20, APD50, and APD90) were recorded and analyzed.

Statistical analysis
Statistical analysis was performed using the Graphpad Prism 6 software. All data are expressed as mean ± SD. Differences among groups were compared by one-way analysis of variance, and the q test was used for comparison of the two groups. A P value of <0.05 was considered statistically significant.

Effectiveness of stimulation on BNP
mRNA expression and the surface area of ventricular myocyte In the null + PE group, PE treatment for 48 h significantly upregulated BNP mRNA expression by an average of 2.24 times (P < 0.01), and significantly increased the surface area ((1,360 ± 90) µm 2 vs (700 ± 40) µm 2 , P < 0.01) of ventricular myocytes compared with the null group. Therefore, PE intervention led to hypertrophy of ventricular myocytes. The BNP mRNA expression in the A1 + PE group was markedly attenuated compared with that in the null + PE group (1.15 ± 0.09 vs 2.24 ± 0.48, P < 0.05), as well as the surface area of cells ((680 ± 180) µm 2 vs (1,360 ± 90) µm 2 , P < 0.01), indicating that the cellular hypertrophy induced by PE was significantly inhibited by Ad-CnAβshRNA intervention (Figures 1 and 2, and Table 1).

Effect of Ad-CnAβshRNA intervention on PE-induced CnAβ protein expression in ventricular myocytes
In the null + PE group, PE treatment significantly upregulated CnAβ protein expression by an average of 2.29 times (P < 0.01) when compared with the null group. In contrast, CnAβ protein expression was significantly lower in the A1 + PE group than in the null + PE group (0.90 ± 0.12 vs 2.29 ± 0.24, P < 0.01) ( Figure 3 and Table 1).

Effect of Ad-CnAβshRNA intervention on I to in ventricular myocytes
At a stimulation voltage of +20 to +70 mV, I to current density in the null + PE group was significantly lower than that in the null group ( Table 2), the averaged current-voltage (I-V) curve relations of I to remarkably shifted downward, and the peak current density was decreased by 49% (P < 0.05; Figure 5). The A1 + PE group had significantly higher I to current density than the null + PE group (P < 0.05; Figure 5 and Table 2). The shape and distribution of the activation curves of I to were similar among the different groups (Figure 6a), and the V 1/2,act showed no significant difference among the groups (P > 0.05; Figure 6b and Table 3). The inactivation curve of I to was significantly shifted to the left in the null + PE group when compared with the null group and was significantly shifted to the right in the A1 + PE group compared to the null + PE group (Figure 6c). The V 1/2,inact was significantly lower in the null + PE group than in the null group (P < 0.05) and significantly higher in the A1 + PE group than in the null + PE group (P < 0.05; Figure 6d and Table 3). Therefore, stimulation of ventricular myocytes with PE accelerated the inactivation of I to , whereas Ad-CnAβshRNA1 intervention inhibited such an effect.

Effect of Ad-CnAβshRNA intervention on APD in ventricular myocytes
APD20, APD50, and APD90 were significantly longer in the null + PE group than in the null group and were  The expression of mRNAs and proteins were exhibited as relative data. Compared with null + PE group, *P < 0.05, **P < 0.01. PE: phenylephrine. A1: Ad-CnAβshRNA. BNP: brain natriuretic peptide. CnAβ: calcineurin Aβ.
significantly shorter in the A1 + PE group than in the null + PE group (Figure 7 and Table 4). Therefore, PE intervention led to significant prolongation of APD in ventricular myocytes, whereas Ad-CnAβshRNA intervention attenuated such an effect.

Discussion
The reduction of I to slows the repolarization of the action potential in the first phase and reduces the depth of the phase 1 notch, thus affecting the activity of other ion channels. The Kv4.3 channel is expressed in human left ventricular muscle and shows a gradient of protein across the ventricular wall, thus forming electrophysiological transmural heterogeneity [16,17]. In the hypertrophic ventricular myocardium in animal models of cardiac hypertrophy or patients with organic heart disease, especially in patients with co-existing heart failure and myocardial injury, the expression of Kv4.3 and Kv4.2 in ventricular myocytes is downregulated and the activity of I to channels is reduced. This leads to abnormal early repolarization, repolarization delay, and APD prolongation, which may easily cause fatal ventricular arrhythmias [8,9,18,19].  The molecular structure of the I to ion channel includes a pore-forming α subunit and an auxiliary β subunit. The α subunit of I to channel has two functionally distinct components. The fast component, I to,fast (I to,f ), for example, recovers from inactivation very rapidly with time constants in the range of 60-100 ms [20,21]. In contrast, the slow component, I to,slow (I to,s ), recovers from inactivation slowly with time constants on the order of seconds [22,23]. Usually, I to is referred to the I to,fast . The fast component of α subunit is formed by assembly of Kv4.2 subunits, Kv4.3 subunits, or a combination of the two, which shows the heterogeneity of the species and the different regions of the same heart [23][24][25][26][27].
Selective gene silencing using antisense oligonucleotides (AsODNs) targeted against Kv4.2 and Kv4.3 reduced I to,f in cultured rodent ventricular myocytes [28,29]. In rat atrial myocytes, AsODNs targeted against Kv4.2, but not Kv4.3, attenuated I to,f [30], whereas in human atrial myocytes, I to,f was significantly attenuated by Kv4.3, but not by Kv4.2 and AsODNs [31]. In addition, targeted gene deletion of Kv4.2 in mice eliminates ventricular I to,f , further revealing the critical role of Kv4.2 in the generation of I to,f channels in rodents [32]. So, in this study, Kv4.2 was chosen as the research component of I to channel other than Kv4.3.
Ventricular hypertrophy is an effective compensation for chronic heart overload, but eventually develops into congestive heart failure because of decompensation following ventricular remodeling [1,2]. Remodeling of numerous ion channels in ventricular myocytes during this process is an important basis for malignant ventricular arrhythmias. Many cell signaling factors are involved in regulating ion channel remodeling in hypertrophic cardiomyocytes. Promoting calcineurin mRNA and protein expression and enhancing calcineurin activity can promote hypertrophy of cardiomyocytes and participate in regulating multiple ion channel remodeling in cardiomyocytes [10,12,13,29]. In rat models of myocardial infarction, the calcineurin inhibitor cyclosporin A can significantly inhibit ventricular remodeling and hypertrophy, improve diastolic function, inhibit a decrease in I to current density in ventricular myocytes, and downregulate Kv4.2 and Kv4.3 mRNA and protein expression [14]. In ventricular myocytes of mouse myocardial infarction models, I to current density is decreased, and Kv4.2 and Kv4.3 mRNA and protein expression is downregulated [8]. Furthermore, the β-blocker metoprolol, the calcineurin inhibitor cyclosporin At a stimulation voltage of 20-70 mV: null vs null + PE, P < 0.05; A1 + PE vs null + PE, P < 0.05. PE: phenylephrine. A1: Ad-CnAβshRNA. A, and knockout of calcineurin-specific downstream signaling factors (i.e., NFAT4 gene, a member of the nuclear factor of activated T cells [NFAT] family) can significantly inhibit a decrease in I to current density and inhibit changes in Kv4.2 and Kv4.3 mRNA and protein expression. In hypertrophic neonatal rat cardiomyocytes, activation of the calcineurin-NFAT signal upregulates the transcriptional expression of Kv4.2 mRNA and protein and increases I to current density [9]. In canine ventricular myocytes with simulated ventricular tachycardia, Kv4.3 mRNA and protein expression is downregulated, and I to current density is significantly reduced, which manifests as significant inhibition of cyclosporine [15]. The semi-activated voltage (V 1/2,act ) of I to . There were no significant differences of V 1/2, act among groups (P > 0.05; the cell numbers are 14, 12, 12, and 9 in null, null + PE, A1 + PE, and A1 groups, respectively). (c) The steady-state inactivation curve of I to . (d) The semi-inactivated voltage (V 1/2,inact ) of I to . PE stimulation accelerated inactivation of I to , which was inhibited by Ad-CnAβshRNA1 intervention (*P < 0.05; n = 10 in each group). PE: Phenylephrine. A1: Ad-CnAβshRNA. In this study, the knockdown of the CnAβ gene, which encoded the main functional unit of calcineurin, completely inhibited CnAβ protein expression, which defected in the substance base that enables calcineurin to function. Therefore, we evaluated the effect of completely suppressed calcineurin activity on I to . When the stimulation voltage was between +20 and +70 mV, intervention with the conventional α 1 adrenergic receptor agonist PE significantly reduced I to current density in the ventricular myocytes of neonatal rats. The peak current density decreased by 49%, and the I-V curves of I to remarkably shifted downward. Furthermore, APD20, APD50, and APD90 were significantly prolonged. Knockdown of the CnAβ gene significantly inhibited the effect of PE intervention on I to current density and APD. Moreover, PE stimulation did not affect the activation of I to , but PE reduced I to current density by accelerating its inactivation. Additionally, the knockdown of the CnAβ gene inhibited the effect of PE on I to inactivation.
The results of earlier studies showed that the precise role of calcineurin in the regulation of I to remains unclear [9,[13][14][15]. In those studies, calcineurin activity was promoted by the agonist or transgenic method, and was The analysis results of the APD20, APD50, and APD90 in groups. PE stimulation prolonged the APD, which was inhibited by Ad-CnAβshRNA1 intervention. **P < 0.001, *P < 0.05; the cell numbers are 12, 12, 12, and 11 in null, null + PE, A1 + PE, and A1 groups, respectively. PE: Phenylephrine. A1: Ad-CnAβshRNA. depressed by cyclosporine [9,[13][14][15]. As per our knowledge, this study is the first one to assess the role of calcineurin in the regulation of I to by way of calcineurin gene silence. Our results potently showed that calcineurin is a negative regulator of I to activity in ventricular myocytes from neonatal rats. In summary, our findings indicated that CnAβ gene knockdown can inhibit PE-induced I to channel remodeling and APD prolongation in hypertrophic neonatal rat ventricular myocytes. This finding suggested that calcineurin may be a potential target for the prevention of malignant ventricular arrhythmia in hypertrophic hearts. However, the study had some limitations. Many ion channels are involved in the formation of action potential in ventricular myocytes. However, this study only detected the I to ion channel. Moreover, as one of the coding genes of I to ion channel, the expression of Kv4.3 was not detected. Data availability statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.