Molecular identity of the late sodium current in adult dog cardiomyocytes identified by Na_v1.5 antisense inhibition

Abstract

Late Na⁺ current (I_NaL) is a major component of the action potential plateau in human and canine myocardium. Since I_NaL is increased in heart failure and ischemia, it represents a novel potential target for cardioprotection. However, the molecular identity of I_NaL remains unclear. We tested the hypothesis that the cardiac Na⁺ channel isoform (Na_v1.5) is a major contributor to I_NaL in adult dog ventricular cardiomyocytes (VCs). Cultured VCs were exposed to an antisense morpholino-based oligonucleotide (Na_v1.5 asOligo) targeting the region around the start codon of Na_v1.5 mRNA or a control nonsense oligonucleotide (nsOligo). Densities of both transient Na⁺ current (I_NaT) and I_NaL (both in pA/pF) were monitored by whole cell patch clamp. In HEK293 cells expressing Na_v1.5 or Na_v1.2, Na_v1.5 asOligo specifically silenced functional expression of Na_v1.5 (up to 60% of the initial I_NaT) but not Na_v1.2. In both nsOligo-treated controls and untreated VCs, I_NaT and I_NaL remained unchanged for up to 5 days. However, both I_NaT and I_NaL decreased exponentially with similar time courses (τ = 46 and 56 h, respectively) after VCs were treated with Na_v1.5 asOligo without changes in 1) decay kinetics, 2) steady-state activation and inactivation, and 3) the ratio of I_NaL to I_NaT. Four days after exposure to Na_v1.5 asOligo, I_NaT and I_NaL amounted to 68 ± 6% (mean ± SE; n = 20, P < 0.01) and 60 ± 7% (n = 11, P < 0.018) of those in VCs treated by nsOligo, respectively. We conclude that in adult dog heart Na_v1.5 sodium channels have a “functional half-life” of ∼35 h (0.69τ) and make a major contribution to I_NaL.

slowly inactivating late sodium current (I_NaL) has been identified in ventricular cardiomyocytes (VCs) of a large variety of mammalian species, including humans (20, 22, 23), dogs (23, 41, 44, 51), guinea pigs (1, 15, 34), rabbits (48), and rats (3). In failing hearts I_NaL was found to be increased and was implicated in repolarization abnormalities as well as Na⁺-related Ca²⁺ overload (for review, see Ref. 24). Furthermore, augmented I_NaL is also reportedly involved in Na⁺ and Ca²⁺ overload in ischemic heart, and inhibition of Na⁺-related Ca²⁺ overload was suggested as a new cytoprotective principle (42, 46). Accordingly, I_NaL has recently resurged as a plausible target for cardioprotection (22, 24, 30). Despite its great fundamental and clinical importance, the molecular identity of I_NaL is still poorly understood.

Both tetrodotoxin (TTX)-sensitive (neuronal type) and TTX-resistant (cardiac type) Na⁺ channel (NaCh) isoforms have been suggested to contribute I_NaL. The idea that the neuronal-type NaCh may contribute to the action potential (AP) plateau was first brought up in 1979 (5). It has been shown that TTX shortened the AP in Purkinje fibers of dog heart in doses two orders of magnitude lower than needed to reduce the upstroke velocity (5). More recent studies claimed that at least some part of the late/persistent Na⁺ current is highly TTX and saxitoxin (STX) sensitive in rat cardiomyocytes (13, 33). Indeed, different transcripts of the main α-subunits of highly TTX-sensitive neuronal isoforms (Na_v1.1, 1.3, 1.6) have recently been detected in mouse (18) and dog (11) heart (Na_v1.1, 1.2, 1.3, 1.6). These neuronal NaCh isoforms were responsible for 10–20% of the peak transient Na⁺ current (I_NaT) responsible for the AP upstroke velocity in dog myocardial and Purkinje cells, respectively (10, 11). On the other hand, previous pharmacological studies of I_NaL suggested that this current is likely produced by the cardiac Na_v1.5 NaCh isoform in human and dog cardiomyocytes, because it is blocked by cadmium ions and has relatively low sensitivities to TTX and STX (20, 24, 44). Furthermore, four splice variants of SCN5A gene encoding the cardiac-specific Na_v1.5 with different biophysical properties have been reported in humans (40), indicating that the question about the exact molecular identity of I_NaL is even more complex than just separation of contributions from TTX-sensitive and -resistant NaCh isoforms.

The present study approaches this problem by direct molecular targeting of the mRNA encoding Na_v1.5. More specifically, we studied both I_NaL molecular identity and the functional channel protein turnover, using antisense oligonucleotide (asOligo) inhibition of the mRNA encoding Na_v1.5 in adult canine VCs. We found that Na_v1.5 is a major I_NaL contributor in these cells.

MATERIALS AND METHODS

Cardiomyocyte isolation and culture.

This study conforms to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health and was approved by the Institutional Animal Care and Use Committee of the Henry Ford Health System. The five mongrel dogs were used for training purposes by the Department of Bioresources and served as heart donors for the present study. VCs were enzymatically isolated from slices of the apical third of the left ventricle from the midmyocardium as previously described (20). The yield of viable, Ca²⁺-tolerant, rod-shaped cells varied from 20% to 80%. The suspension of VCs was cultured in serum-free medium (M199, with Earle's salts, 25 mM HEPES, and bicarbonate, without l-glutamine) as previously described (21). The culturing M199 was supplemented with 2 mM carnitine, 5 mM creatine, 5 mM taurine, 2 g/l albumin (bovine, fraction V), 0.1 μM insulin, 100 U/ml penicillin, 100 μg/ml streptomycin, and 25 μg/ml gentamicin [CCT medium (47)]. The cells were incubated at 37°C in a humidified atmosphere of 95% O₂-5% CO₂. Under these conditions ∼80% of cells survived up to 5 days of incubation, similar to the survival reported previously for adult rabbit cardiomyocytes (28). During this time VCs undergo some changes in cell shape and morphology manifested in slightly rounded edges of the cultured cells. Nonsense oligonucleotide (nsOligo) or Na_v1.5 asOligo did not further affect morphology of VCs (see Fig. 2A).

Fig. 2.

Cell culture model and intracellular delivery of oligonucleotides. A: only slight changes in shape (rounded edges) of dog ventricular cardiomyocytes (VCs) were evident after 5 days of culture. 10 μm/div, distance between calibration dots. B: fluorescein-tagged oligo uptake by cardiomyocytes was monitored by confocal microscopy. Confocal images of live cardiomyocytes loaded with fluorescein-tagged oligo are shown. Optical slices were 0.5 μm (Zeiss Axiovert 100, Bio-Rad MRC 1024, excitation/emission wavelength 488/522 nm).

Heterologous expression of Na_v1.5 and Na_v1.2.

In this study we also used an HEK293 cell line (αHEK) stably expressing Na_v1.5 or Na_v1.2 as previously described (43). The wild-type human heart Na_v1.5 clone (formerly hH1a) was kindly provided by Drs. H. A. Hartmann and A. M. Brown (Baylor College of Medicine, Houston, TX) (9).

Antisense oligonucleotide: design and delivery.

We chose morpholino-based 25-mer oligos over phosphorothioate (sugar based) oligos because of their superior properties for our experimental approach (38). Design of oligo sequences was based on high homology between mRNAs encoding Na_v1.5 of human (8) and dog NaCh (GenBank AF017428 AY126477, AJ555547, AY126477). The Na_v1.5 asOligo was designed to be around the start codon. The asOligo for Na_v1.5 used in this study (5′-ccgaggtaataggaagtttgccatc; GenBank M77235) was aligned (BLAST 2 software) with gene transcripts encoding known human voltage-activated NaCh α-subunit isoforms (Na_vs): neuronal-type Na_v1.7 (GenBank X82835), skeletal muscle Na_v1.4 (GenBank M81758), brain-type 1 Na_v1.1 (GenBank X65362), and type 2 Na_v1.2 (GenBank M94055). Furthermore, we tested atypical Na_vs: Na_v2.1 (GenBank M91556) as well as mouse Na_v2.3 (GenBank L36179) and Norway rat SCL-11 (GenBank Y09164). No significant similarity was found. We used the nsOligo 5′-cctcttacctcagttacaatttaata as a control. We also did not find any significant similarity of the nsOligo with known α- and β-subunits of NaCh.

Given the well-known difficulty of using conventional nonviral methods to introduce exogenous DNA into terminally differentiated cells like cardiac myocytes (36), we chose a special delivery (SD) system based on ethoxylated polyethylenimine (EPEI) (29). The rationale of the SD system is that nonionic oligo is paired to the complementary DNA. The DNA is then bound electrostatically to weakly basic EPEI. This oligo-DNA-EPEI complex is efficiently endocytosed (29). The effect of asOligos (2.55 and 0.98 μM) on NaCh expression was tested both in stably transfected αHEK cells and VCs, respectively. VCs were cultured for 24 h before oligo delivery. Both αHEK cells and VCs were exposed to oligo-DNA-EPEI for 3 h and then washed out with CCT medium. About 98% of VCs cells were positive for the oligo, as monitored by fluorescence of the fluorescein-tagged oligo (see Fig. 2B). Other conventional gene transfection delivery systems were ineffective for the adult VCs, and limitations of our approach are discussed in Study limitations.

Patch-clamp technique and data analysis.

I_Na was recorded with a conventional whole cell patch-clamp technique (pCLAMP9, Axopatch 200A patch-clamp amplifier, Axon Instruments). The resistance of the borosilicate glass patch pipettes was 1.2–2.1 MΩ for HEK cells and 0.6–0.8 MΩ for VCs. The currents were low-pass filtered at 5 kHz and digitized at a sampling rate of 20 kHz. The quality of the voltage clamp was controlled in each cell as previously described (20, 44). Experiments were performed at 22–24°C.

I_NaL was elicited by a 2-s membrane depolarization to various potentials from a holding potential (V_h) of −130 mV applied with a pacing frequency of 0.1 Hz (see Table 1 for solutions). The “zero” current was determined after TTX (25 μM) application (20). The amplitude of I_NaL was determined from the averaged current measured at 200–220 ms after the onset depolarization to avoid contamination with I_NaT (43). I_NaT was measured at a symmetrical Na⁺ concentration of 5 mM (Table 1). All measurements were performed 10–25 min after the membrane rupture to allow for complete cell dialysis with intracellular recording solutions (26, 31, 32). Currents were normalized to the membrane electric capacitance (C_m) measured by a voltage ramp pulse with a slope (dV/dt) of −10 V/s from −80 mV to −100 mV (21). The fine structure of the I_Na time course was approximated by a double-exponential fit to I_NaT or I_NaL decay (19, 25):

(1)

where τ₁ and τ₂ are the time constants and I₀ is the amplitude 0.2 ms after peak of I_NaT or the instant value of I_NaL at 40 ms after membrane depolarization; k₁ and k₂ are the contributions of each exponent (k₁ + k₂ = 1), respectively. Five to fifteen experimental traces were averaged to improve the quality of analysis for I_NaL. The decay of I_NaL was evaluated at membrane potential (V_m) = −30 mV (20), and τ₁ and τ₂ correspond to the burst (BM) and late scattered (LSM) modes of the late NaCh gating, respectively (19, 25).

Table 1.

Extracellular (bath) and intracellular (pipette) solutions used in study

	Pipette Solution INaT, INaL	Bath Solutions
INaT for VCs		INaL for VCs, INaT for αHEK
NaCl	5	5	140
KCl		5.4
CsCl	133	133	5.4
CaCl2		1.8	1.8
MgCl2		2	2
MgATP	2
TEA	20
Nifedipine		0.002	0.002
EGTA	10
HEPES	5	5	5
pH	7.3 CsOH	7.3 NaOH	7.3 NaOH

Concentrations are expressed in mM. I_NaT, peak transient Na⁺ current; I_NaL, late Na⁺ current; VCs, ventricular cardiomyocytes; αHEK, HEK293 cell line stably expressing Na_v1.5 or Na_v1.2; TEA, tetraethylammonium.

The steady-state availability (or inactivation, SSI) parameters [the midpoint (V_½A) and the slope (k_A) of the relationship] were measured by a standard double-pulse protocol with the 2-s-duration prepulse (V_p) ranging from −130 mV to −40 mV. The data points of I_Na normalized to I_Na measured at −30 and −130 mV prepulse were fitted to a Boltzmann function A(V_p):

(2)

The data points of the normalized current in the current-voltage relationships were fitted (by Origin 7.0 software, Microcal Software) to the following function (21):

(3)

where G_max is a normalized maximum Na⁺ conductance, V_r is a reversal potential, V_t is testing voltage, and V_½G and k_G are the midpoint and the slope of the respective Boltzmann function underlying the NaCh steady-state activation (SSA). The time course of both I_NaT and I_NaL reduction by the asOligo was evaluated by the best single-exponential fit to the data points:

(4)

where τ is the time constant and I_o and I_i are the amplitudes and asymptotic levels, respectively. The functional half-life of NaCh protein was assessed as t_1/2 = 0.69τ.

Statistical analysis.

Data are reported as means ± SE, with n representing the number of cells. Multiple comparisons between treatment groups were made with one-way analysis of variance (ANOVA) followed by Bonferroni's post hoc test. The significance of SSA and SSI changes was evaluated with an F-test (StatMost, Data-Most, Salt Lake City, UT) for tabulated values predicted by the model (Eqs. 2 and 3) at a confidence level of 0.95. Differences for both experimental data and model predictions were considered statistically significant at P < 0.05.

Chemicals.

Enzymes protease type XXIV and hyaluronidase type IV-S were purchased from Sigma (St. Louis, MO) and collagenase type II (291 U/mg) from Worthington (Freehold, NJ). asOligo and nsOligo as well as the EPEI-based SD system were purchased from Gene Tools (Philomath, OR; www.gene-tools.com). All other chemicals were purchased from Sigma.

RESULTS

Probing efficacy of Na_v1.5 antisense oligonucleotide in a heterologous expression system.

In αHEK cells we tested the efficacy of our Na_v1.5 asOligo to reduce functional expression of Na_v1.5 but not Na_v1.2. Treatment with Na_v1.5 asOligo caused a significant decrease of I_NaT density without changes in midpoint potential of activation and inactivation (Fig. 1, A–C; Table 2) or the decay kinetics (Fig. 1D). Analysis of the time course of I_NaT reduction by Na_v1.5 asOligo revealed an exponential decrease (Eq. 4) of the I_NaT density with a time constant τ of 52 h (Fig. 1E) and I_i ∼27% of the initial I_NaT. The t_½ for NaCh evaluated as 0.69τ was found to be 36 h. Control cells loaded with nsOligo did not show any significant changes in I_NaT over this time (Fig. 1E), indicating stable expression of NaCh. We performed additional tests with the HEK cells stably transfected with the brain NaCh isoform Na_v1.2 to demonstrate specificity of our Na_v1.5 asOligo. Indeed, Na_v1.5 asOligo did not affect the maximum Na⁺ current density [14.2 ± 3.1 (n = 16) vs. 15.1 ± 2.5 (n = 13) pA/pF, nsOligo vs. asOligo, respectively, at V_m = −10 mV], SSI and SSA (Table 2). Thus these control experiments in αHEK cells proved the effectiveness of the Na_v1.5 asOligo to selectively silence (by ∼70%) functional expression of Na_v1.5.

Fig. 1.

Potency of antisense oligonucleotide (asOligo) to silence Na_v1.5 expression in HEK293 cell line stably expressing Na_v1.5. A: representative raw traces of transient Na⁺ current (I_NaT) were recorded at different membrane potentials (V_m) in control conditions [nonsense oligonucleotide (nsOligo)] and in the presence of Na_v1.5 asOligo. B: average data for peak I_NaT-voltage (V) relationship 5 days after nsOligo or Na_v1.5 asOligo delivery. Solid lines show theoretical I-V curves fitted in accordance with Eq. 3. asOligo reduced the maximum I_NaT conductance (G_max) from 12.5 × 10⁻³ to 5.1 × 10⁻³ nS/pF (P < 0.05, F-test) without changes in steady-state activation (SSA) parameters (see Table 2). *Statistically significant (P < 0.05) differences in data points. C: average experimental data of steady-state inactivation (SSI) along with their theoretical Boltzmann fits (Eq. 2, solid and dashed lines). D: asOligo did not affect I_NaT decay time course evaluated by the double-exponential fit (Eq. 1). Shown are data points of 2 time constants (τ_fast, τ_slow) at different V_m, pooled from 9 or 10 cells. E: I_NaT decreases in exponential manner in response to asOligo delivery. Shown are data points along with the linear regression (solid line for nsOligo) and a single-exponential fit (asOligo, Eq. 4). Exponential time constant (τ) value is indicated. Data points in B–E are means ± SE. A and C, insets: voltage-clamp protocols. V_h, holding potential; V_p, prepulse. Statistically significant difference (P < 0.05) in B and E was evaluated by ANOVA followed by Bonferroni's post hoc test. Detailed statistics for all SSA and SSI parameters of the theoretical fits shown in B and C are presented in Table 2. Equations 1–4 are given in materials and methods.

Table 2.

Steady-state activation and inactivation for I_Na in heterologously expressed NaCh isoforms and in dog ventricular cardiomyocytes

Conditions	SSI Parameters			SSA Parameters
	V1/2A, mV	kA, mV	n	V1/2G, mV	kG, mV	n
Heterologously expressed Nav1.2, INaT
nsOligo, 5 days	−70.0±2.6	−7.7±0.9	7	−21.4±1.5	5.1±0.8	5
asOligo, 5 days	−68.9±2.1	−8.0±0.9	7	−24.1±1.1	4.5±0.3	5
Heterologously expressed Nav1.5, INaT
nsOligo, 5 days	−81.4±1.4	−5.8±0.4	9	−39.7±1.2	5.1±0.5	7
asOligo, 5 days	−82.9±0.8	−5.9±0.3	8	−41.7±1.2	5.1±0.3	9
Dog ventricular cardiomyocytes, INaT
Freshly isolated	−80.6±0.7	−5.7±0.3	8	−38.8±1.5	5.1±0.3	9
5 Days in culture	−81.7±0.9	−6.2±0.4	7	−36.3±1.5	6.3±0.2	9
nsOligo, 5 days	−82.9±0.9	−6.2±0.2	11	−38.3±1.2	5.9±0.2	13
asOligo, 5 days	−82.8±1.2	−6.9±0.6	7	−37.9±1.1	6.3±0.3	11
Dog ventricular cardiomyocytes, INaL
Freshly isolated	−80.0363±0.9	−6.3±0.5	19	−39.4±0.6	7.1±0.1	6
5 Days in culture	−81.2822±	−5.9±0.6	8	−39.5±0.7	7.3±0.5	6
nsOligo, 5 days	−82.0061±1.2	−5.0±0.8	8	−36.8±1.2	6.3±0.3	11
asOligo, 5 days	−80.3534±	−5.1±0.7	8	−37.5±0.7	6.4±0.9	9

Data are means ± SE; n is cell number. Steady-state inactivation (SSI) and activation (SSA) parameters were obtained from I_Na data fit to Eqs. 2 and 3 (materials and methods), respectively. We did not find statistically significant differences between parameters within the experimental groups as evaluated by ANOVA followed by Bonferroni's post hoc test. nsOligo, nonsense oligonucleotide; asOligo, antisense oligonucleotide; V_1/2A and k_A, midpoint and slope of Boltzmann function A curve; V_1/2G and k_G, midpoint and slope of conductance curve.

I_NaL and I_NaT in cardiomyocytes in culture.

We first addressed how culturing conditions may affect cell morphology and levels of NaCh expression. VCs underwent well-known changes in cell shape (28), mainly observed as slightly rounded cell edges during the first 5 days of culture (Fig. 2A). Delivery of nsOligo or Na_v1.5 asOligo did not further affect the appearance of VCs (Fig. 2A). Moreover, these culturing conditions did not affect NaCh expression as judged from I_NaT and I_NaL over the time in culture (Figs. 3 and 4, Table 2). Thus these control experiments of the time course of Na⁺ current density confirmed the usefulness of cultured adult dog cardiomyocytes for experiments to further address the molecular identity of I_NaL.

Fig. 3.

I_NaT remains unchanged in cultured adult dog cardiomyocytes. A: representative raw current trace recorded in freshly isolated VCs (left) and after 5 days in culture (right) at different V_m. B: peak I_NaT-V relationship obtained in freshly isolated cells and cells cultured for 5 days. Theoretical curves fit to I-V (Eq. 3, solid and dashed lines) were not statistically different. C: SSI data points along with the fit to a Boltzmann function (Eq. 2, solid and dashed lines, respectively). No statistically significant difference was found when SSI curves were compared. D: decay time constants of I_NaT (double-exponential fit, Eq. 1) were evaluated at the different V_m. E: maximum density of I_NaT remained unchanged in these conditions. Data in B–D are means ± SE and were pooled from 5–8 cells. Detailed statistics for all SSA and SSI parameters of the theoretical fits shown in B and C are presented in Table 2. B and C, insets: voltage-clamp protocols.

Fig. 4.

Cultured adult dog cardiomyocytes represent a useful model to study I_NaL. A: data points of I_NaL-V relationships for freshly isolated and cultured cells along with their theoretical fits (Eq. 3, solid and dashed lines). Inset: typical examples of raw I_NaL traces recorded in fresh and cultured cells are shown superimposed. There was no statistical difference between the data points or theoretical fit of the I-V curves. B and C: density and SSI of I_NaL remain unchanged in cultured cells. C: data points of SSI are shown along with their theoretical fits (Eq. 2, solid and dashed lines). Inset: our voltage-clamp protocol. D: decay time course of I_NaL was evaluated by the 2-exponential fit (Eq. 1) and remained unchanged in the cultured cells. All data are means ± SE. Detailed statistics for all SSA and SSI parameters of the theoretical fits shown in A and C are presented in Table 2.

Effects of antisense oligonucleotide on I_NaT and I_NaL in cultured cardiomyocytes.

We tested the effects of asOligo in cultured VCs, assuming high homology of mRNAs encoding cardiac human and canine NaCh. Indeed, alignment of Na_v1.5 with partial sequence of the dog heart NaCh (GenBank AF017428; Ref. 49) revealed 92% identity with cardiac human Na_v1.5 NaCh. Visualization of Na_v1.5 asOligo delivery was based on fluorescence of fluorescein-tagged oligo. Typical examples of confocal images of cardiomyocytes loaded with the nsOligo or Na_v1.5 asOligo are shown in Fig. 2B. After 3 h of incubation, we found diffused distribution of the OLIGO in VC cytoplasm. Using fluorescence microscopy (not shown), we estimated that 98% of VCs take up oligos. However, it is important to note that treatment with oligos was toxic to the cells, and <50% of treated cells survived.

Voltage-clamp experiments showed that Na_v1.5 asOligo substantially reduced density of I_NaT in contrast to that of the control nsOligo-treated VCs (Fig. 5, A and B). Importantly, the decay kinetics (Fig. 5D) and SSA or SSI (Fig. 5, B and C) remained unchanged. Data on the SSA and SSI parameters are summarized in Table 2. Similar to αHEK cells, in VCs treated with asOligo I_NaT decreased exponentially, with τ = 46 h and I_i = 62.8% of the initial I_NaT (Fig. 5E). The evaluated functional half-life was t_½ = 32 h. There was moderate (∼1%/day) and negligible change of I_NaT in the control pool of cells treated by nsOligo (Fig. 5E). The data confirmed efficacy of the asOligo to reduce Na_v1.5 expression in adult dog cardiomyocytes.

Fig. 5.

asOligo effectively knocks down Na_v1.5 expression in adult dog cardiomyocytes. A: asOligo caused decrease in peak I_NaT. Representative raw traces of I_NaT were recorded at different membrane potentials in control conditions (nsOligo, right) and with asOligo (left). B: peak I_NaT-V relationship obtained in cells treated with nsOligo (control) and asOligo. *Statistically significant differences in data points (P < 0.05) compared at different V_m. Solid (nsOligo) and dashed (asOligo) lines show theoretical curves of I-V (Eq. 3) fitted to data points. asOligo significantly reduced G_max from 1.39 to 1.01 pS/pF (P < 0.05, F-test). C: asOligo did not affect the SSI of I_NaT. Shown are data points of SSI along with their theoretical fits (Eq. 2; solid lines nsOligo and dashed lines asOligo; see Table 2 for SSI parameters). D: asOligo did not affect the decay time course of I_NaT. The two time constants, τ_fast and τ_slow (double-exponential fit, Eq. 1), vs. membrane and the relative contribution of τ_slow (k_slow) are given at top and bottom, respectively. E: time course of I_NaT decrease in response to Na_v1.5 knockdown by asOligo. Data points are shown along with the linear regression (solid line for nsOligo) and exponential fit (asOligo). Time constant is indicated; n is the number of cardiomyocytes. The initial level of I_NaT (100%) was 47.3 ± 9 pA/pF (n = 25). Statistically significant difference (P) in A and E was evaluated by ANOVA followed by Bonferroni's post hoc test. Detailed statistics for all SSA and SSI parameters of the theoretical fits shown in A and C are presented in Table 2. Bars in B and D and points in A and C represent means ± SE.

asOligo affected I_NaL in a similar way as observed for I_NaT. Indeed, Na_v1.5 asOligo resulted in a substantial and statistically significant decrease of I_NaL density (Fig. 6A) without changes in the ratio of I_NaL to I_NaT (Fig. 6B), decay time course (Fig. 6C), and SSA or SSI (Table 2). The data on the time course of the Na_v1.5 asOligo effect on I_NaL are summarized in Fig. 6D. In VCs treated with Na_v1.5 asOligo but not with nsOligo, the amplitude of I_NaL decreased over time in culture and are well described by a single-exponential decay (Eq. 4) with τ = 56 h and an I_i ∼58.3% of the initial I_NaL. The estimated functional half-life t_1/2 for the pool of NaCh underlying I_NaL was 39 h. Thus the time course and degree of Na_v1.5 asOligo-induced density decrease in VCs were similar for both I_NaT and I_NaL (compare Figs. 5E and 6D).

Fig. 6.

Molecular identity of I_NaL in dog ventricular cardiomyocytes assessed by asOligo specific to mRNA encoding Na_v1.5. A: I_NaL density is statistically significantly reduced by the asOligo (P < 0.001). Inset: superimposed representative raw I_NaL traces recorded in the presence of nsOligo and asOligo, respectively. B: knockdown of SCN5A gene by asOligo leads to the parallel decrease of both I_NaT and I_NaL, as I_NaL-to-I_NaT ratio remains unchanged. C: asOligo does not affect the fine structure of I_NaL decay evaluated by the 2-exponential fit (Eq. 1). D: statistical analysis of the time course of I_NaL density changes in response to asOligo. Data points are shown along with the linear regression (solid line for nsOligo) and the exponential fit (asOligo). The time constant is indicated; the initial level (100%) of I_NaL current density was 0.223 ± 0.025 pA/pF (n = 18). Statistically significant difference (P) in A and E was evaluated by ANOVA followed by Bonferroni's post hoc test. Detailed statistics for all SSA and SSI parameters of the theoretical fits shown in A and C are presented in Table 2. Bars in A–C and points in D represent means ± SE.

Theoretical evaluation of I_NaL that is contributed by Na_v1.5.

Approximation of the components of the entire I_NaL time course by a double-exponential function (Eq. 1) allows interpretation of the whole cell data in terms of gating modes of late NaCh openings (19). According to a numerical model (25) of Na⁺ channel gating based on the single-channel data of late openings, the entire I_NaL time course is well described as a sum of two exponentials: BM generates the fast, exponentially decaying I_NaL component (I_BM), whereas LSM generates a relatively slow I_NaL decaying component (I_LSM):

(5)

(6)

where I₄₀, k₁, k₂, τ₁, and τ₂ are from Eq. 1 and correspond to the whole cell currents (I_BM and I_LSM) produced by BM and LSM, respectively. The results of the theoretical approximation of the idealized, average late currents that are generated by these modes are shown in Fig. 7. The analysis of I_NaL in cells treated with Na_v1.5 asOligo and nsOligo thus revealed that both BM (Fig. 7A) and LSM (Fig. 7B) of the late NaCh openings are produced by the Na_v1.5 channel. The total average I_NaL in adult dog cardiomyocytes determined by Na_v1.5 asOligo is shown in Fig. 7C.

Fig. 7.

Theoretical evaluation of the fine structure of I_NaL determined by Na_v1.5 in adult dog cardiomyocytes. Parameters (time constants and densities) for the idealized, 2-exponential decay time course of I_NaL were assigned from the averaged experimental data measured in cultured cardiomyocytes treated with nsOligo (density 0.276 pA/pF at 200 ms, τ₁ = 48.6, τ₂ = 502 ms, k₂ = 0.268) or asOligo (0.132 pA/pF, τ₁ = 38.8, τ₂ = 497 ms, k₂ = 0.305). Currents generated by the burst (I_BM in A) and late scattered openings (I_LSM in B) were obtained from Eqs. 5 and 6 (results), respectively, and the total I_NaL (C) was obtained as the sum of these two. The difference currents between nsOligo and asOligo-treated cells are shown as shaded areas.

DISCUSSION

Result summary, conclusions, and possible importance.

We used Na_v1.5 asOligos to determine molecular identity of the I_NaL in adult dog VCs. We demonstrated that culturing of these cells (up to 5 days) does not significantly alter either I_NaT or I_NaL. This fact allowed the use of the cell culture as a model for the experiments with gene expression knockdown by the Na_v1.5 asOligo. The asOligo that was designed with the sequence of human mRNA encoding cardiac Na_v1.5 was also effective in dog cardiomyocytes. Knockdown of cardiac Na_v1.5 in dog VCs resulted in exponential simultaneous decrease of I_NaT and I_NaL (to ∼60% of their initial levels after 100 h; Figs. 5E and 6D) without changes in I_NaL-to-I_NaT ratio, decay kinetics (Figs. 5D and 6D), and the parameters of SSA and SSI (Table 2). The t_½ for I_NaL was similar to that for I_NaT (32–36 h). Our data suggest that the cardiac Na_v1.5 isoform is the major contributor to I_NaL in cardiomyocytes of the adult dog, and we believe that this finding is important in the context of recent interest in I_NaL as a therapeutic target in heart failure and ischemia (4, 24, 30). It is also important to note that our previous studies of I_NaL (see Ref. 24 for review) and the fact that our asOligo was highly effective for human and dog Na_v1.5 (this study) indicate that functional and molecular properties of cardiac NaCh (both early and late currents) are very close in humans and dogs so that canine models of heart failure or ischemia are helpful in studying the roles of I_NaT and I_NaL in human heart disease.

A new paradigm in heart failure is that I_NaL is increased while I_NaT is reportedly decreased (45, 50). With this in mind, further blockade of I_NaT is proarrhythmic as it worsens existing conduction problems of the failing myocardium. The results of the present study showing that the same NaCh likely produces I_NaT and I_NaL indicate that specific pharmacological targeting of I_NaL will be challenging because many drugs will block both I_NaT and I_NaL. One approach to solving the problem is based on properties of specific drugs to change specific gating properties of NaCh (see Ref. 23 for review). For example, amiodarone and the new antianginal drug ranolazine interact with specific gating states of NaCh and thereby predominantly block I_NaL vs. I_NaT (22, 41). Another approach is based on indirect targeting of I_NaL via its multiple known modulator factors of NaCh environment [affecting late NaCh gating modes (25, 43)] such as cytoskeleton, phospholipids, auxiliary subunits, and Ca²⁺ (see Refs. 19, 24 for review). In any case, an established molecular identity of I_NaL is critical to approach the unresolved problem.

Functional half-life: comparison with other plasmalemmal proteins.

asOligos have been used previously to study turnover of plasmalemmal proteins, such as Na⁺/Ca²⁺ exchanger (NCX). For example, a relatively short t_1/2 (<12 h) for the NCX has been reported in some studies that utilized relatively high asOligo concentrations (3–10 μM), resulting in inhibition of NCX function by from 30% (2) up to 80% (16). The apparent fast turnover of NCX at higher doses of oligo could be due to either side effects of the oligo or the existence of a fraction of NCX1 protein with a relatively fast turnover (6). Alternatively, with lower doses of asOligo (0.5–2.0 μM), slower NCX protein turnover rates (t_½ > 30 h) were reported in neonatal and adult cardiomyocytes (7, 37), which is in line with t_½ of 20–40 h for several other plasmalemmal transport proteins (14, 39) and NaCh (this study and Ref. 35).

Cultured adult cardiomyocytes as a model to study NaCh function.

Primary cultures of cardiomyocytes are a useful model because they produce a homogeneous population of terminally differentiated cells free of extracellular matrix and neurohumoral factors. This is an attractive model for studies that utilize electrophysiological (including patch clamp) and molecular biology approaches. On the other hand, numerous studies have shown that over the culturing period adult cardiomyocytes undergo changes in their morphology and protein expression levels including those of ion channels (see Ref. 28 for review). Accordingly, results gathered over prolonged periods (≥6 days) in these cells should be treated with caution. These morphological and functional changes can be minimized by the design of an appropriate culturing medium (so-called CCT medium used in the present study) that does not contain serum and glutamate and is supplemented with creatine, l-carnitine, and taurine. In CCT medium the survival rate increased up to 14 days for rod-shaped myocytes. The loss of some contractile and cytoskeleton proteins may be linked to the absence of an appropriate humoral factor (47). Reduction of contractile function over the period of culture was linked to the atrophy of contractile proteins (in rat and rabbit but not in feline cardiomyocytes) and a reduced number of T tubules (27). Fairly rapid internalization of the intercalated disks (within 24 h) was reported in rat VCs (12). The present study clearly demonstrates that dog VCs in culture represent a relatively stable and useful model to study cardiac NaCh (using electrophysiological and/or molecular biology techniques), because of the stable intrinsic functional expression of these channels during a relatively long period of time (up to 100 h of cell culture).

Molecular identity of I_NaL.

Culture-specific changes of cells can affect ion channel expression, particularly for proteins that are localized within T tubules and intercalated disks. Indeed, it has been shown that brain-type isoforms Na_v1.1, Na_v1.3, and Na_v1.6 are located in T tubules and Na_v1.5 at the intercalated disks of rat VCs (17). In dog VCs Na_v1.1 was localized at the intercalated disks and Na_v1.2 was present at the Z lines (11). Unlike other ion channels (see Ref. 28 for review), sodium ion channels underlying both I_NaT and I_NaL did not change during the 5 days of culture (Figs. 3 and 4). Simultaneous inhibition of both I_NaT and I_NaL in response to the Na_v1.5 asOligo knockdown (Figs. 5 and 6) supports the idea that Na_v1.5 generates both I_NaT and I_NaL. This idea is also supported by previous pharmacological studies using NaCh-specific toxins, cadmium ions, and other experimental conditions to distinguish between NaCh isoforms (20, 23). Furthermore, the present study demonstrates that Na_v1.5 is responsible for both burst and late scattered gating mode components determining the fine structure of the total I_NaL time course (Fig. 7).

Study limitations.

The delivery of oligos to VCs has limitations that are common to all antisense methods [discussed in detail elsewhere (36)]. A major common limitation is a significant reduction in the number of viable VCs, although surviving cells demonstrate inclusion of the oligo (in ∼98% of cells in our experiments). The low number of surviving cells was suitable for our patch-clamp experiments but makes biochemical analysis (i.e., Western blotting) much more challenging. While we conclude that Na_v1.5 is the major contributor to I_NaL, a possible role for neuronal isoforms especially in disease states could not be completely excluded based on the data presented in this study.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants HL-53819 and HL-074238 (A. Undrovinas) and RO1-HL-65661 (J. W. Kyle) and by American Heart Association Grant-in-Aid 0350472Z (A. Undrovinas).