Calpain Inhibition Reduces Amplitude and Accelerates Decay of the Late Sodium Current in Ventricular Myocytes from Dogs with Chronic Heart Failure

Calpain is an intracellular Ca2+ -activated protease that is involved in numerous Ca2+ dependent regulation of protein function in many cell types. This paper tests a hypothesis that calpains are involved in Ca2+ -dependent increase of the late sodium current (INaL) in failing heart. Chronic heart failure (HF) was induced in 2 dogs by multiple coronary artery embolization. Using a conventional patch-clamp technique, the whole-cell INaL was recorded in enzymatically isolated ventricular cardiomyocytes (VCMs) in which INaL was activated by the presence of a higher (1μM) intracellular [Ca2+] in the patch pipette. Cell suspensions were exposed to a cell- permeant calpain inhibitor MDL-28170 for 1–2 h before INaL recordings. The numerical excitation-contraction coupling (ECC) model was used to evaluate electrophysiological effects of calpain inhibition in silico. MDL caused acceleration of INaL decay evaluated by the two-exponential fit (τ1 = 42±3.0 ms τ2 = 435±27 ms, n = 6, in MDL vs. τ1 = 52±2.1 ms τ2 = 605±26 control no vehicle, n = 11, and vs. τ1 = 52±2.8 ms τ2 = 583±37 ms n = 7, control with vehicle, P<0.05 ANOVA). MDL significantly reduced INaL density recorded at –30 mV (0.488±0.03, n = 12, in control no vehicle, 0.4502±0.0210, n = 9 in vehicle vs. 0.166±0.05pA/pF, n = 5, in MDL). Our measurements of current-voltage relationships demonstrated that the INaL density was decreased by MDL in a wide range of potentials, including that for the action potential plateau. At the same time the membrane potential dependency of the steady-state activation and inactivation remained unchanged in the MDL-treated VCMs. Our ECC model predicted that calpain inhibition greatly improves myocyte function by reducing the action potential duration and intracellular diastolic Ca2+ accumulation in the pulse train. Conclusions Calpain inhibition reverses INaL changes in failing dog ventricular cardiomyocytes in the presence of high intracellular Ca2+. Specifically it decreases INaL density and accelerates INaL kinetics resulting in improvement of myocyte electrical response and Ca2+ handling as predicted by our in silico simulations.


Introduction
The role of the late sodium current (I NaL ) in electrophysiological remodeling and arrhythmias in chronic heart failure (HF) has been extensively studied during the last decade. It has been shown that I NaL is augmented and its decay slowed in failing human and dog ventricular cardiomyocytes (VCMs)(see for review [1]). A remarkable contribution of I NaL into HF mechanisms has been demonstrated in experiments where ''correction'' of I NaL in failing VCMs resulted in: 1) rescue of normal repolarization, 2) decrease beat-to-beat action potential (AP) duration variability, and 3) improvement of Ca 2+ handling and contractility [1]. Accordingly, I NaL has emerged as a novel target for cardioprotection to treat the failing heart [1,2] The new approaches may involve: 1) discovery new drugs that directly and specifically target I NaL , 2) targeting intracellular signaling pathways (for example Ca 2+ -dependent signaling) that are altered in HF and may have modulatory effect on I NaL , 3) modulation of altered Na + channel (NaCh) microenvironment, such as different expression of auxiliary b-subunits and sub-sarcolemmal cytoskeleton that, in turn, may be responsible for the augmented slowed I NaL in HF, 4) combination of two latter mechanisms. The new drug, ranolazine (RAN) that was developed as an antianginal agent, has been demonstrated to specifically inhibit I NaL [3,4]. RAN reduced arrhythmias in the immediately post-MI patients in the recent MERILIN-TIMI trial [5] confirming the clinical relevance of I NaL . Ca 2+ , calmodulin and CaMKII and this Ca 2+ signaling pathway can significantly amplify I NaL in HF affecting both contractile and electrical performance [6,7]. As to NaCh microenvironment, it has been shown that alterations in membrane phospholipids composition and/or in sub-sarcolemmal cytoskeleton, which consists of ankyrin, actin, spectrin (fodrin), can affect NaCh gating in heart in the way that the late openings may occur [1,8,9]. Recently we have shown that silencing SCN1B but not SCN2B, the genes that are responsible for expression of the b 1 and b 2 NaCh subunits, could be a plausible mechanism to modulate I NaL in HF with the aim to improve both contractility and rhythm [10].
Calpain is an intracellular Ca 2+ -activated protease and an important mediator of the actions of the intracellular Ca 2+ in heart. Cleavage by calpain is critical in a variety of calciumregulated cellular processes such as muscle contraction, neuronal excitability, secretion, signal transduction, cell proliferation, differentiation, cell cycle progression, and apoptosis [11,12]. Deregulation of calpain caused by impaired Ca 2+ homeostasis during cardiac pathologies such as atrial fibrillation, heart failure, hypertrophy, or ischemia reperfusion, is critically involved in the myocardial damage. One of the intracellular targets of calpain is fodrin, a dynamic structure that is altered under a variety of pathological conditions featuring poor Ca 2+ handling (e.g. ischemia or heart failure [13,14,15,16]). In the present study we tested the hypothesis that the membrane-permeant calpain inhibitor MDL-28170 (MDL) can prevent, in part, Ca 2+ -related I NaL modulation in VCMs from dogs with chronic HF. We found that MDL reduces density of whole-cell I NaL and makes I NaL decay faster in the failing VCMs. Using the excitation -contraction coupling (ECC) numerical model [17] we also assessed physiological significance of the MDL effects. We show that these MDLinduced I NaL alterations: 1) reduce AP duration, and 2) prevent diastolic intracellular Ca 2+ accumulation during the excitation pulse train in silico.

HF model and cardiomyocyte isolation
The study conforms to the Guidelines for Care and Use of Laboratory Animals published by the US National Institutes of Health and was approved by the Animal Care and Use Committee (IACUC protocols 0816 and 0777) of the Henry Ford Health System. Chronic heart failure that is similar by vast array of functional and pathophysiological parameters [18] to that in  humans was produced in 2 dogs by multiple sequential coronary artery microsphere embolizations as previously described [19]. At the time of harvesting the heart (,3 months after last embolization), left ventricular (LV) ejection fraction was approximately ,25%. Ventricular cardiomyocytes (VCMs) were enzymatically isolated from the apical LV mid-myocardial slices as previously reported [20]. The yield of viable rod-shaped, Ca 2+ -tolerant VCMs varied from 40 to 70%. 2.2. Patch clamp technique and data analysis I NaL was measured using a whole-cell patch-clamp technique [20]. I NaL was assessed by 2 s-long membrane depolarizations to various potentials from a holding potential of 2130 mV applied with a stimulation frequency of 0.2 Hz. The bath solution contained (in mM):140 NaCl, 5.0 CsCl, 1.8 CaCl 2 , 2.0 MgCl2, 5 glucose, 0.002 nifedipine, and 5 HEPES-CsOH buffer (Ph 7.4). The pipette solution contained (in mM): 5 NaCl, 133 CsCl, 0.9 CaCl 2 (free [Ca 2+ ] = 1 mM) MgATP, 20 Tetraethylammonium chloride, 1.0 EGTA, and 5.0 HEPES-CsOH buffer. The free [Ca 2+ ] of 1 mM was set in the pipette (and hence inside the cell) to exaggerate abnormal effect of Ca 2+ on I NaL in HF [21]. Experiments were performed at room temperature (22-24uC). A stock solution of the cell-permeant MDL 28170 (MDL) was prepared in DMSO. MDL was then diluted in the bath solution to a final concentration of 0.2 mM and 2.6 mM DMSO [21,22]. Cells suspensions were exposed to MDL from 1-2 hours prior to patch-clamp experiments, and MDL was also added to the pipette solution [23,24]. All measurements were made in the presence of MDL in the bath solution and 8-25 min after the membrane rupture to complete cell dialysis with intracellular recording solutions [25,26].
The time course of I NaL decay has been approximated by a double exponential fit to I NaL starting at 40 ms after the onset of depolarization to -30 mV as previously suggested [27]: where t 1 and t 2 are the time constants, I 40 is I NaL instant value 40 ms after membrane depolarization, k 1 and k 2 are the contributions of each exponents (k 1 + k 2 = 1), respectively. 5-15 experimental traces were averaged to improve the quality of analysis.
Original I NaL recordings were also analyzed to assess the current density (pA/pF), i.e. I NaL = (whole cell I NaL )/C m , where C m is cell electric capacitance that was measured by a voltage ramp (16) in each cell. The I NaL data points in the current-voltage relationships were measured as the averaged current density within 200-220 ms after depolarization onset (vertical bar in Fig. 1A). The steady-state activation (SSA) parameters were determined from the currentvoltage relationships by fitting data points of the normalized current with the function [27]: Where G max is a normalized maximum Na + conductance, V r is a reversal potential; V KG , and k G are the midpoint and the slope of the respective Boltzmann function underlying the steady-state Na + channel activation.
The steady-state inactivation (SSI) was evaluated by a doublepulse protocol with 2 s-duration pre-pulses (V p ) ranging from -130 mV to 240 mV followed by a testing pulse to 230 mV. I Na amplitudes were normalized to that measured at V p = 2130 mV and the data points were fitted to a Boltzmann function A(V p ):

Numerical model simulation of calpain inhibition of effect in failing myocytes
We simulated effect of selective inhibition of calpain on AP shape and diastolic Ca 2+ accumulation in silico using our previously reported modification of EC coupling model of failing canine ventricular myocyte (originally developed by Winslow et al. [28]). In short, our model has introduced a new detailed formulation of I NaL lacking in the original model. This important model modification has allowed us to predict an important role of I NaL to alter AP shape (increase AP plateau duration) and to contribute to diastolic Ca 2+ accumulation in HF ventricular myocytes [17]. In short, in our in silico examinations we use I NaL data measured under voltage clamp at 24uC and then apply Q 10 factors to calculate model parameters for our full I NaL description at 37uC. The details of the model parameters calculations have been described in our previous publications [17]. Specific parameter values of the present study are given in Table 1. The stair case phenomenon was simulated by assigning a relatively low [Ca 2+ ] of 0.125 mM as an initial value in both network SR and junctional SR before application of stimulation pulse train (at 1 Hz or 1.5 Hz).

Statistical Analysis
Multiple comparisons between treatment groups were made using one-way analysis of variance (ANOVA) followed by Bonferroni's post hoc test or by the non-pared Student's t-test if appropriate. Data are reported as mean6SEM. The significance of SSA or SSI changes were evaluated using F-test (StatMost, DataMost Corp., Salt Lake City, UT) for tabulated values predicted by the model (Eqs. 2, 3) at a confidence level of 0.95. Differences for both experimental data and model predictions were considered statistically significant for P,0.05.

Chemicals
Collagenas type II (291 U/mg) was from Worthington (Freehold, NJ). All other chemicals and enzymes, including calpain inhibitor MDL was purchased from Sigma (St. Louis, MO).

Calpain inhibitor MDL makes I NaL decay faster in VCM from failing dog hearts
First we compare I NaL decay in VCMs exposed to MDL with that in control, in the absence of the drug, at the intracellular Ca 2+ = 1 mM (this intracellular Ca 2+ was used in all experiments presented in this study). As it is shown in Fig. 1A in VCM exposed to MDL the I NaL decay becomes faster than in control cell. Shown are raw traces along with the two-exponential fit (solid lines). Statistical data are given in Fig1. B. Note that decay time course on the I NaL became significantly faster in MDL-treated cells as it is obvious from the reduction of the time constants t 1 (upper panel) and t 2 (lower panel).

In VCMs from failing dog hearts calpain inhibitor
MDL decreases I NaL density in wide range of the membrane potential without affecting the steady-state activation voltage-dependency Treatment with MDL significantly reduced I NaL density in VCMs compared with control and vehicle-treated cells (Fig. 2 A). The density was measured as an average current at the membrane potential of -30 mV within 200-220 ms after the depolarization onset (shown by the vertical bar in Fig. 1A). Fig. B shows an effect of MDL on I NaL density in the wide range of the membrane potentials assessed in the IV relationship (dots). The solid lines represent theoretical fit to the Eq.2 (see Methods) with the aim to assess SSA parameters (Shown in the graph). The MDL does not affect the voltage-dependence of the SSA as it is evident of the mid-potential position and the slope of the curve, which we found to be unchanged. At the same time MDL reduced the maximum conductance, G max , for I NaL , which is expected because of the density reduction (See Fig. 2 Legend). Fig. 3A shows experimental points obtained by the two-pulse protocol along with the theoretical fit (solid lines, Eq.3 in Methods) for the SSI evaluation. There was no statistical difference (F-test) when the theoretical curves corresponding to a MDL, vehicle or control (no vehicle) were compared. Fig. 3B shows statistics for the SSI parameters, mid-point potential and slope coefficient K A . There was no significant difference for these parameters pointing to the absence of the effect on SSI by the MDL.

ECC model predictions of physiological importance
We tested if these experimentally measured I NaL changes caused by the calpain inhibitor are physiologically significant in HF myocytes. Using I NaL characteristics measured in the present study, we carefully calculated all I NaL parameters (Table 1) required for numerical modeling of I NaL effects as we previously established for VCMs of dog failing heart [17]. Specifically, I NaL,200ms is the density of I NaL measured at 200 ms after membarne depolarization onset; t BM and t LSM are t 1 and t 2 , respectively in Equation 1; k BM and k LSM are k 1 and k 2 , respectively in Equation 1; parameters k G , V 1/2G , and k A , V 1/2A are defined in Equations 2 and 3, respectively. We corrected I NaL decay constants and maximum amplitude density for 37C using Q 10 factors (2.2 and 1.5, respectively) as we described in our previous studies [29,30] as follows: t 37 = t 24 ?2.2 (24-37)/10 and I max_37 = I max_24 ?1.5 (37-24)/10 . Finally, I max = I max_BM + I max_LSM and G NaL_max are the maximal total I NaL peak current density and conductance, respectively. Fig. 4 upper panel shows results of the in silico test of how MDL-induced I NaL changes affect AP shape and duration at a physiological temperature of 37uC at a pacing rate of 1 Hz. Note that simulated APs are shorter with lower plateau in VCMs treated by MDL. Lower panel of Fig. 4 shows the prediction of I NaL dynamics (profile) during the AP in control and after MDL treatment. Our simulations show that the amplitude and duration of I NaL become substantially smaller in MDL-treated cells vs. control (untreated) cells. Fig. 5 upper panel shows in silico simulations of the intracellular [Ca 2+ ] dynamics in VCMs of failing heart. In response to the pulse train stimulation with the rate of 1.5 Hz, the diastolic Ca 2+ level gradually increases in control conditions. The MDL-induced changes in I NaL amplitude and decay kinetics almost completely eliminate this diastolic Ca 2+ accumulation pattern. Lower panel of Fig. 5 shows simultaneous AP simulations for this condition. Note shorter AP with the lower plateau similarly to that shown in Fig. 4.

Discussion
For the first time we demonstrate at the single cell level that I NaL alterations in amplitude and decay kinetics associated with chronic HF can be rescued by calpain inhibition. Our in silico simulations also demonstrate that the calpain modulation of I NaL is physiologically important in HF myocytes, specifically, calpain inhibition greatly improves the myocyte function by reducing the action potential duration, and intracellular diastolic Ca 2+ accumulation in the pulse train.
The calpain family is a group of cysteine proteases unique in their dependency on calcium to attain functionally active forms [31]. Calpain is involved in a wide range of Ca 2+ -regulated cellular processes such as signal trasduction, secretion, cell proliferation, differentiation and apoptosis [11]. Calpain deregulation resulting from the impaired Ca 2+ handling is one of the important mechanisms for the pathological processes such as apoptosis and necrosis, reperfusion-induced heart stunning, ischemia and hypoxia, hyperthrophy and heart failure, and atrial fibrillation [11,32]. Therefore calpain inhibition is considered a therapeutic strategy targeting multiple disease states [32].
Our findings thus suggest a novel cellular and molecular mechanism to modulate NaCh that could be targeted to prevent pathophysiological consequences related to the augmented I NaL in HF. There are some indications of the involvement of calpain into ion channel gating regulation, namely L-type Ca 2+ channels [33,34]. In this context the calpain inhibition may serve to improve Ca 2+ handling in failing heart and may be considered as a novel approach to modulate I NaL current its related arrhythmias, and improve contractility [1,2]. Below we discuss possible cellular and molecular mechanisms of calpain effect on I NaL .

Calpain and fodrin cytoskeleton
Fodrin-based cytoskeleton, an important element of the NaCh microenvironment in heart, is a dynamic structure that is altered under a variety of pathological conditions (e.g. ischemia or heart failure [13,14,35]). The role of the fodrin-based cytoskeleton in I NaL modulation has been confirmed in our previous studies [1]. It has also been shown that fodrin breakdown that occurs in some disease states featuring poor Ca 2+ handling can be mediated by calpain [15,35]. Therefore prevention of Ca 2+ -induced fodrin cytoskeleton degradation will likely improve Ca 2+ handling in HF.

Interplay between Ca 2+ , CAM/CaMKII cascade and calpain
It has been shown that I NaL depends on the [Ca 2+ ] i signaling cascade in the way that increased [Ca 2+ ] i binds to EF-hand motif on NaCh C-terminal domain [6,36,37] or via activating CaM/  CaMKII cascade resulting in the augmented and slowed I NaL [6,7]. This is very important mechanism of I NaL regulation in HF because Ca 2+ homeostasis is impaired in this disease stage. Inhibition of calpain results in reduced density of I NaL despite of the presence of high [Ca 2+ ] i that works in the opposite direction [6,7]. At the same time SSI and SSA parameters remain unchanged pointing to the fact that all channels are available for I NaL and that the parameters of SSI and SSA depend on [Ca 2+ ] i , rather than on calpain-dependent proteins. Indeed we have shown that the direct binding of Ca 2+ to NaCh [37] (likely to E-F hand domain of NaCh C-terminus) is responsible for shifts of the half membrane potential of SSI voltage dependence towards depolarizing potentials [6]. Therefore, reduction of I NaL density produced by MDL likely results from reduced probability of NaCh transitions into different modes (burst and late scattered modes) that are involved in I NaL formation [29]. The faster I NaL decay in the presence of calpain inhibition (Fig. 1) also indicates that gating of these modes is also affected by calpain.

Interplay between calpain and NaCh b-subunits
It has been shown that besides the main pore-forming a subunit of NaCh [37], the b 2 -subunit of NaCh is attached to the subsarcolemmal cytoskeleton [38]. Therefore prevention of cleavage of fodrin by calpain may stabilize the cytoskeleton and enhance the b 2 -subunit dependent modulation of I NaL that we have recently reported [10]. We have shown that reduction of b 2 expression by the siRNA increased I NaL density and delayed its decay in VCMs from dogs with HF, i.e. very similar to that caused by the increased [Ca 2+ ] i [6] (via activation of calpain) and opposite to that caused by the calpain inhibition by MDL shown herein.

4.4.
Physiological relevance of Ca 2+ -calpain signaling to modulate I NaL and to improve contractility and rhythm of failing heart It has been established that I NaL plays an important role in both electrical and contractile (via Ca 2+ handling) deficiencies caused by chronic HF [1,2]. Then an important question is whether the magnitude of the effect of calpain inhibition on I NaL reported in the present study is physiologically relevant. To address this question, we have carefully measured and analyzed specific characteristics of I NaL in control and in the presence of MDL (Table 1) and then integrated them into our recently published ECC numerical model for ventricular cardiomyocytes of the failing dog [10,17]. As it is evident from Figs. 4 and 5, MDL substantially reduces effects of I NaL on AP duration, which is known to increase in HF [1]. The resultant decrease of I NaL during AP plateau is observed as I NaL becomes scaled (decreased) by about a factor of two during most of the plateau. This substantial ''scaling'' contributes not only to in AP duration shortening but also in to that AP plateau becomes substantially lower. This insight does not directly follow from the voltage clamp data and even not from AP simulations, because of a complex interplay of many Na +and Ca 2+ -dependent mechanisms in ventricular cells reproduced by the dynamic ECC model.
Since shorter AP plateau and a smaller inward current during AP plateau are associated with less incidence of EADs [20,39], one expected beneficial effect of calpain inhibition is also to reduce the probability of the EADs [17], a major mechanism for the triggered arrhytmia. Recently we have demonstrated that the augmented and slowed I NaL in HF contributes to the diastolic [Ca 2+ ] i accumulation in VCM of failing hearts during the pulse train [17]. Reduction of I NaL by the MDL significantly reduces this accumulation [Ca 2+ ] i as it is predicted in silico (Fig. 5). Previously it has been demonstrated that delayed afterpotentials are linked to the diastolic Ca 2+ accumulation associated with I NaL , [17,40]. Therefore this predicted effect of MDL to prevent the diastolic [Ca 2+ ] i accumulation indicates, in turn, that calpain inhibition can reduce probability of occurrence of the DADs.

Conclusion
Based on our present results with the specific calpain inhibitor MDL in ventricular cardiomyocytes isolated from failing dog hearts, we conclude that Ca 2+ -dependent calpain activation is able to strongly modulate I NaL density and kinetics in failing myocardium. We illustrate in silico that the range of this modulation is physiologically relevant and remarkable as the calpain inhibition substantially improves (shortens) AP duration and prevents diastolic Ca 2+ accumulation. Therefore, this Ca 2+dependent signaling cascade may serve as a plausible target to regulate I NaL and its related electrical and contractile deficiencies in failing heart.