inhibition of the late sodium current slows t-tubule disruption during the progression of...

Inhibition of the late sodium current slows t-tubule disruption during theprogression of hypertensive heart disease in the rat

Gary L. Aistrup,1 Deepak K. Gupta,1 James E. Kelly,1 Matthew J. O’Toole,1 Amanda Nahhas,1

Nimi Chirayil,1 Sol Misener,1 Lauren Beussink,1 Neha Singh,1 Jason Ng,1 Mahendra Reddy,1

Thitipong Mongkolrattanothai,1 Nesrine El-Bizri,2 Sridharan Rajamani,2 John C. Shryock,2

Luiz Belardinelli,2 Sanjiv J. Shah,1 and J. Andrew Wasserstrom1

1Department of Medicine (Cardiology) and the Feinberg Cardiovascular Research Institute, Northwestern UniversityFeinberg School of Medicine, Chicago, Illinois; and 2Gilead Sciences, Incorporated, Fremont, California

Submitted 14 May 2013; accepted in final form 12 July 2013

Aistrup GL, Gupta DK, Kelly JE, O’Toole MJ, Nahhas A,Chirayil N, Misener S, Beussink L, Singh N, Ng J, Reddy M,Mongkolrattanothai T, El-Bizri N, Rajamani S, Shryock JC,Belardinelli L, Shah SJ, Wasserstrom JA. Inhibition of the latesodium current slows t-tubule disruption during the progression ofhypertensive heart disease in the rat. Am J Physiol Heart CircPhysiol 305: H1068 –H1079, 2013. First published July 19, 2013;doi:10.1152/ajpheart.00401.2013.—The treatment of heart failure(HF) is challenging and morbidity and mortality are high. The goal ofthis study was to determine if inhibition of the late Na� current withranolazine during early hypertensive heart disease might slow or stopdisease progression. Spontaneously hypertensive rats (aged 7 mo)were subjected to echocardiographic study and then fed either controlchow (CON) or chow containing 0.5% ranolazine (RAN) for 3 mo.Animals were then restudied, and each heart was removed for mea-surements of t-tubule organization and Ca2� transients using confocalmicroscopy of the intact heart. RAN halted left ventricular hypertro-phy as determined from both echocardiographic and cell dimension(length but not width) measurements. RAN reduced the number ofmyocytes with t-tubule disruption and the proportion of myocyteswith defects in intracellular Ca2� cycling. RAN also prevented theslowing of the rate of restitution of Ca2� release and the increasedvulnerability to rate-induced Ca2� alternans. Differences betweenCON- and RAN-treated animals were not a result of different expres-sion levels of voltage-dependent Ca2� channel 1.2, sarco(endo)plas-mic reticulum Ca2�-ATPase 2a, ryanodine receptor type 2, Na�/Ca2�

exchanger-1, or voltage-gated Na� channel 1.5. Furthermore, myo-cytes with defective Ca2� transients in CON rats showed improvedCa2� cycling immediately upon acute exposure to RAN. Increasedlate Na� current likely plays a role in the progression of cardiachypertrophy, a key pathological step in the development of HF. Early,chronic inhibition of this current slows both hypertrophy and devel-opment of ultrastructural and physiological defects associated with theprogression to HF.

t-tubule; ranolazine; late INa; heart failure; calcium handling

HEART FAILURE (HF) is commonly increasing and associatedwith significant morbidity and mortality (19). Once estab-lished, HF is difficult to treat. An alternative approach wouldtarget patients before the onset of overt HF, during the stagewhen pathological cardiac changes are accumulating. Directedtherapies during this vulnerable period may slow the progres-sion to HF.

Recently, the late component of the rapid Na� current (INa,L)has been shown to be increased in a number of models of HF

(30) as well as in ischemia (10). The increase in INa,L may beresponsible for the rise in intracellular Na� concentration([Na�]i) that has been reported in myocytes from nearly allanimal models of HF, and in myocytes from human failinghearts (22, 24, 31, 32). Through reverse mode Na�/Ca2�

exchange (NCX), the rise in [Na]i leads to increased diastolicintracellular Ca2� concentration ([Ca2�]i), which causes intra-cellular Ca2� overload. The effect of INa,L to increase diastolic[Ca2�]i may cause activation of Ca2�/calmodulin-dependentprotein kinase II (CaMKII) (40). INa,L-induced Ca2� overload andactivation of CaMKII are responsible for the reverse force-fre-quency relationship and ultimately for Ca2� waves and triggeredarrhythmias (22–24, 31, 32). An activated CaMKII may alsophosphorylate voltage-gated Na� channel 1.5 (Nav1.5), slow-ing Na� channel inactivation and further increasing INa,L (16).These interesting findings about the interaction between INa,L

and CaMKII raise the possibility that inhibition of INa,L couldbe beneficial for the treatment of HF.

T-tubule disruption is a cellular mechanism that contributesto the development of HF. It has been known for some timethat t-tubule organization is decreased in nearly all animal andhuman models of HF (5, 7, 14, 15, 35). More recently, it hasbeen suggested that the development of HF occurs as thenumber of myocytes exhibiting t-tubule disruption increases sothat cardiac function declines as more myocytes are affected(4, 35). These new studies suggest that t-tubule organizationplays a critical role not only in normal Ca2� cycling but also inthe development of HF, as progressively more myocytes havepoor t-tubule organization.

The goal of this study was to test the hypothesis that earlyintervention with an inhibitor of INa,L might slow the progres-sive pathological changes of the myocardium (e.g., t-tubuledisruption and Ca2� cycling defects) that lead to HF. Sponta-neously hypertensive rats (SHRs; 7 mo old) were randomlyassigned to control (CON; no drug) or ranolazine (RAN)treatment groups for 3 mo. The 3-mo treatment with the INa,L

inhibitor ranolazine (RAN) (3, 6) was intended to serve as aproof of concept that inhibition of INa,L at an early stage ofhypertension-induced hypertrophy could delay or prevent thedevelopment of HF.

METHODS

Male SHRs (7 mo of age, 2 groups of 8 each) were used in thisstudy. Ionic current was measured in an additional 6 SHRs (3 each at7 and 10 mo of age). Animal use protocols were approved by theInstitutional Animal Care and Use Committee according to NationalInstitutes of Health (NIH) guidelines. Each group of animals was

Address for reprint requests and other correspondence: J. A. Wasserstrom,310 E. Superior St., Tarry 12-723, Chicago, IL 60611 (e-mail: [email protected]).

Am J Physiol Heart Circ Physiol 305: H1068–H1079, 2013.First published July 19, 2013; doi:10.1152/ajpheart.00401.2013.

0363-6135/13 Copyright © 2013 the American Physiological Society http://www.ajpheart.orgH1068

mailto:[email protected]

mailto:[email protected]

studied using echocardiography and then either maintained on normalchow (CON) or started on chow containing 0.5% RAN. The meanplasma RAN concentration measured in RAN-treated rats at the endof the study was 5.4 � 0.5 �M, a value that is within the therapeuticrange (2–8 �M). After 3 mo, a repeat echocardiogram was performed.There were no significant differences in body weight, heart weight,heart weight-to-body weight ratio, heart rate, or systolic and diastolicblood pressures of rats in the CON and RAN groups at the time ofstudy termination. After completion of the echocardiogram, each heartwas then removed, perfused using the method of Langendorff to washout pretreatment drug or vehicle, and used for imaging studies.

Imaging of t-tubule organization and intracellular Ca2� cycling. ALangendorff-perfused heart was placed in an experimental chamberon the stage of a Zeiss LSM510 laser scanning confocal microscopeand then exposed to the membrane potential-sensitive dye 4-{�-[2-(di-n-butylamino)-6-naphthyl]vinyl}pyridinium (di-4-ANEPPs; 8–10�M) during recirculation with modified Tyrode solution at a temper-ature of 25 � 1°C. Cytochalasin-D (60 �M) and blebbistatin (20 �M)were then added to the perfusate to abolish contraction. Three-dimensional images of t-tubule organization were recorded in 20–25subepicardial regions of the left ventricular (LV) free wall using a�40 (numerical aperature, 1.2) water immersion objective. T-tubuleorganization was recorded in at least 100–200 myocytes for eachheart. The dye-containing solution was then washed out by perfusionof the heart with fresh Tyrode solution for 10 min, after whichrecirculation was initiated again with buffer containing fluo-4 AM (15�M). Two successive additions of 12 and 10 �M fluo-4 AM to theperfusate were made after 20 and 40 min. The fluo-4 AM solution wasthen washed out to allow deesterification of the dye in the cytoplasm.Recirculation was reinitiated and the paralytic agents were againintroduced into the solution.

Ca2� imaging was accomplished as previously described (1, 12,33, 34). Briefly, the scan line was placed either along the long axis ofindividual myocytes to obtain high resolution imaging of Ca2� cy-cling within a single cell or the line was placed across the short axisof multiple cells (up to �14) to study Ca2� cycling in larger cellpopulations. Multiple sites were imaged in the middle of the LV freewall to get a random sampling of Ca2� cycling behavior among manymyocytes in each heart. Note that kinetics of intracellular Ca2�

changes but not amplitude can be reliably measured in hearts previ-ously exposed to di-4-ANEPPS, because although interference bydi-4-ANEPPS affects the absolute levels of fluo-4 fluorescence inten-sity, it does not alter the relative changes.

In some of the control-fed SHRs, we identified myocytes withdefective Ca2� cycling (slow rise, low release, and slow decay). RAN(10 �M) was then added to the superfusate, and these multicellularsites were imaged after 10 min of exposure to RAN.

Because we were interested in both the release and recovery phasesof sarcoplasmic reticulum (SR) Ca2� cycling, we analyzed rise time(i.e., from 10 to 90% of the peak) and time to peak Ca2� (i.e., frombaseline to 100% of the peak) transient and transient durations at 50,80, and 90% (TD50, TD80, and TD90, respectively) of recovery tobaseline. A customized MATLAB (Mathworks, Natick, MA) routinewas used for measurements of Ca2� transients during pacing at a basiccycle length (BCL) of 700 ms and during rapid pacing (BCL � 300ms) to allow comparisons of Ca2� cycling at both slow and fast heartrates. In addition, restitution of SR Ca2� release was measuredfollowing a 30-s train at the basal pacing cycle length of 700 ms byinterpolating a single beat at an interval ranging from 150 to 600 ms.Restitution as a function of increasing recovery interval was calcu-lated by measuring Ca2� transient magnitude as percentage of thatobserved at the basal rate. The magnitude of Ca2� alternans (cycle-to-cycle “large-small-large-small” alternations in Ca2� transient mag-nitude) was measured at the end of a 10-s period of rapid pacing atprogressively shorter BCL from 400 to 180 ms, following 30 s ofpacing at the basal rate. Alternans ratio (AR) was calculated as ameasure of the variation of the beat-to-beat Ca2� transient amplitude

[AR � 1 � (small/large)] (39). Note that AR 1 signifies a nearlyundetectable Ca2� transient during the small alternation. Finally, thevariability in Ca2� cycling properties within each myocyte wasmeasured as the SD of measured values of each property.

Measurements of t-tubule organization using a fast Fouriertransform. The two-dimensional (2-D) image from each z-stack thatrepresented the middle of each myocyte excluding the external sar-colemma was identified and analyzed using customized MATLABsoftware and then transformed to the spectral domain using a fastFourier transform. Power spectrum analysis was used to determine anindex of t-tubule organization (OI) within a myocyte (9, 28). Toincrease the sensitivity of our calculation of t-tubule organization, wemeasured the area under the power spectrum between frequencies0.43 and 0.53 �m�1 where normally organized t-tubules would befound (band 1, which peaks at a power of �0.46 �m�1). We thenmeasured the area under the power spectrum between frequencies0.33 and 0.43 �m�1 where poorly organized t-tubules would beevident (band 2). All power below 0.33 �m�1 was ignored becausethis component contains only system noise. OI was calculated as thepower for organized t-tubules as a fraction of the total power of bothpoorly organized and normal t-tubules [band 1/(band 1 � band 2)].This approach provides a sensitive measure of both the decrease inwell-organized t-tubules as well as the increase in poorly organizedt-tubules, giving a usable range of �0.45 to 0.95. OIs were measuredfor all cells in each heart to compare the effects of RAN versus CON.In these analyses, OI can decrease but cannot increase beyond amaximum, resulting in a naturally skewed rather than a normaldistribution of OI values. Therefore, 95% confidence intervals for OIwere used as a measure of variability in OI. Note, however, that meanand SD are also included because it will be the OI mean and thevariability around the mean that will ultimately determine the physi-ological effect of OI variability in the whole heart.

Echocardiography. Echocardiography (2-D and M-mode) was per-formed on each animal at baseline (7 mo), and after 3 mo of treatment.animals were restrained and tail blood pressure was measured with aCODA System (Kent Scientific, Torrington, CT). All animals wereimaged by a single, experienced echocardiographer using an i13Lprobe and a Vivid 7 echocardiography machine (General Electric).Short-axis, long-axis, and apical four-chamber views of the heart wereobtained. LV mass, dimensions including LV end-diastolic diameter,fractional shortening, and ejection fraction were quantified, along withmitral inflow parameters. All measurements were made offline usingan EchoPAC workstation (General Electric, Kenosha, Wisconsin).

Measurement of INa,L in isolated LV myocytes from 7- and 10-mo-old SHR. Briefly, 7- and 10-mo-old male SHRs were anesthetized andanticoagulated with heparin sodium. Hearts were removed and imme-diately mounted on a Langendorff apparatus to be perfused at 8 to 9ml/min and 37°C with a solution of the following composition: (inmM) 140 NaCl, 4.4 KCl, 1.5 MgCl2, 120 NaH2PO4, 5 HEPES, 7.5glucose, 16 taurine, and 5 Na pyruvate, adjusted with NaOH to pH7.3. Hearts were then perfused with this solution supplemented withcollagenase type I (0.8 mg/ml; Sigma-Aldrich, St. Louis, MO) and0.2% bovine serum albumin. The LV free wall was dissected andfinely chopped and gently triturated. Isolated myocytes were collectedby filtration and stored in Kraftbrühe solution for 1 h at roomtemperature and then transferred to Eagle’s minimal essential mediumsolution (M0518; Sigma-Aldrich), supplemented with (in mM) 1Ca2�, 5 HEPES, and 5 glucose at pH 7.3 (NaOH). Isolated cardio-myocytes were used within 6–8 h after isolation.

Whole cell INa was recorded at 22 � 1°C using the patch-clamptechnique and a Multiclamp 700B amplifier (Molecular Devices,Sunnyvale, CA). Cardiomyocytes were superfused with bath solutioncontaining (in mM) 135 NaCl, 4.6 CsCl, 1.8 CaCl2, 1.1 MgSO4, 10HEPES, and 10 glucose and 0.01 nitrendipine at pH 7.4. Patchpipettes were made from borosilicate glass (World Precision Instru-ments, Sarasota, FL) using a DMZ Universal puller (Dagan, Minne-apolis, MN). Pipette resistance was 1–2 M when filled with a pipette

H1069INa,L INHIBITION SLOWS THE PROGRESSION TO HEART FAILURE

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00401.2013 • www.ajpheart.org

(internal) solution containing (in mM): 120 aspartic acid, 20 CsCl, 1MgSO4, 4 ATPNa2, 0.1 GTPNa3 and 10 HEPES; pH adjusted to 7.3with CsOH. After establishing a whole cell patch configuration, aperiod of 5–10 min was given for cells to stabilize before conductingexperiments. INa,L was measured as current amplitude between 210and 220 ms after application (at a rate of 0.1 Hz) of a 220-msdepolarizing step to �20 mV from a holding potential of �120 mV.Data were acquired using pClamp 10.2 software (Molecular Devices)and analyzed using Clampfit 10, Microcal Origin (OriginLab,Northampton, MA), and GraphPad Prism (GraphPad Software, LaJolla, CA) software programs. Values of INa,L in the absence andpresence of RAN were corrected by subtraction of tetrodotoxin (10�M)-insensitive current. To determine the half maximal inhibitoryconcentration (IC50) values for inhibition by ranolazine of INa,L,concentration-response relationships were fit using the Hill equation:Idrug/Icontrol � 1/[1 � (D/IC50)nH], where Idrug/Icontrol is fractionalblock, D is drug concentration, IC50 is the drug concentration thatcauses 50% block, and nH is the Hill coefficient.

Protein lysate preparation and Western blot analysis. Proteinlysates were prepared by resuspending the frozen tissue in radioim-munoprecipitation assay buffer (pH 7.4) containing 1% 3-[(3-chol-amidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) and1� protease inhibitors (Complete Mini, Roche) on ice. The tissue washomogenized on ice using six brief pulses of a polytron homogenizer.Homogenates were centrifuged at 14,000 g for 20 min at 4°C toremove pellet. The supernatant was used for Western blot analysis.Cell lysates were heated with Laemmle sample buffer, consisting of0.5 mM Tris·HCl (pH 6.8), 10% SDS, 10% glycerol, 4% �-mercap-toethanol, and 0.05% bromophenol blue at 55°C for 15 min. Equalamounts of protein (40 �g/lane) were separated by SDS-PAGE(4–20% gradient gel, Bio-Rad) and electroblotted on polyvinylidenedifluoride membranes. Membranes were blocked using either 5%BSA or 5% nonfat skim milk in Tris-buffered saline with 0.1% Tween20. Antibodies against voltage-dependent Ca2� channel 1.2 (Cav1.2;1:1,000, Abcam), NCX (1:500, Swant Swiss antibodies), sarco(endo)plasmic reticulum Ca2�-ATPase 2a (SERCA2a; 1:200, Thermo Sci-entific), Nav1.5 (1:1,000, Alomone) and ryanodine receptor type 2(RyR2; 1:500, Thermo Scientific) were incubated with the membranesovernight at 4°C. The membranes were then incubated with horse-radish peroxidase-conjugated anti-mouse IgG or anti-rabbit IgG sec-ondary antibodies for 2 h at 37°C. GAPDH was used as a loadingcontrol. Protein bands were visualized with ECL-Plus reagent (Pierce;Thermo Fisher Scientific, Rockford, IL). Band intensity was quan-tified by densitometric scanning with ImageJ software (NIH,Bethesda, MD).

Statistics. Data are presented as means � SE unless otherwiseindicated. Statistical analyses were performed using SigmaPlot 11.0.Sample means were compared between groups using either paired orunpaired t-tests, and ANOVA was used for comparisons for multiplegroups with Student-Newman-Keul’s tests for post hoc comparisons.Differences between sample means were considered significant whenP � 0.05.

RESULTS

Chronic INa,L block (3 mo) decreases the number of myo-cytes with poor t-tubule organization. The OI was measured asa fluorescence signal from di-4-ANEPPs staining of the t-tu-bule membrane, excluding the sarcolemma. Results of OImeasurements are shown in Fig. 1. Figure 1A shows a myocytefrom a CON heart with poor t-tubule organization, whereasFig. 1B shows a cell from a RAN heart with highly organizedt-tubules. The graphs below each 2-D image illustrate theresults of the fast Fourier transform where the CON cell has asmall peak and a high baseline to the left of the peak, indicatingboth an increase in poorly organized t-tubules and a decrease in

well-organized t-tubules. In contrast, the RAN cell has a verysharp, high-peak and a very low baseline, indicating thatvirtually all fluorescence is found only in the highly organizedt-tubule system.

OI distributions from representative CON and RAN heartsare illustrated in Fig. 1, C and D, respectively. For the CON ratheart (Fig. 1C), the mean OI for all myocytes was 85.1 �10.0% (OI � 0.851) with a 95% confidence interval rangingfrom 65.5 to 104.7%. Note the distinct skewing of the distri-bution to the left, indicating a significant number of myocyteswith poor t-tubule organization. In contrast, the RAN heart(Fig. 1D) has a slightly higher mean OI (86.2 � 10.9%) witha 95% confidence interval ranging from 66.6 to 105.8%.

Fig. 1, E and F, shows the cumulative OI histograms for allCON and RAN rats (n � 8 for each group), respectively. Themean OI for all CON myocytes was 83.8 � 8.2% (OI � 0.838,n � 1,003 cells in 8 hearts) with a 95% confidence intervalranging from 83.3 to 84.3%. Note the distinct skewing of thedistribution to the left. In contrast, myocytes from all RANhearts had a higher mean OI (85.6 � 1.9%, n � 1,230 in 8hearts, P � 0.001 compared with CON), with a 95% confi-dence interval ranging from 85.2 to 85.9%. In addition, thepercentage of myocytes with OI � 80% was significantlylower in hearts from RAN-treated compared with CON rats(14 � 3.0% vs. 37 � 10%, P � 0.05). These data demonstratethat not only is there a significantly greater OI in the myocytesfrom RAN hearts but there are also fewer myocytes withpoorly organized t-tubules in the RAN hearts compared withCON hearts.

Chronic INa,L block prevents the progression of LV hy-pertrophy. Echocardiographic measurements before and after 3mo of CON and RAN feeding are summarized in Table 1.There were no significant differences in cardiac structure andfunction (e.g., LV end-diastolic diameter, fractional shorten-ing, ejection fraction) between CON and RAN-fed rats, withthe exception of LV mass. Hearts from CON rats had theexpected increase in LV mass from 7 to 10 mo (�0.13 � 0.12g), whereas LV mass in RAN rats declined (�0.06 � 0.11 g,P � 0.005 vs. CON).

We also measured cell size characteristics that are associatedwith hypertrophy. Representative images in Fig. 2, A and B,show that cell length but not cell width was greater in the CONheart than following RAN treatment. The summary data showthat cell length (measured in cells at rest) but not cell width wassignificantly greater in CON- than in RAN-treated rats at 10mo (Fig. 2, C and E). The variability in cell length but not incell width was also significantly greater in hearts from RAN-treated compared with CON-treated rats (Fig. 2, D and F).

Chronic RAN treatment prevents the development of Ca2�

cycling defects during the progression of hypertensive heartdisease. An essential characteristic of altered Ca2� cycling inHF is a slowing in Ca2� release and reuptake (14, 34). Theformer occurs presumably because t-tubule disruption reducesCa2� influx via L-type Ca2� channels and thus reduces thetrigger for Ca2�-induced Ca2� release from the SR. This isprimarily responsible for systolic dysfunction as slow, nonuni-form, and defective Ca2� release throughout the myocyte leadsto slow and inefficient myofibril activation and resulting cellcontraction. Prolongation of the Ca2� transient occurs in partbecause of reduced expression of SERCA2a, thereby slowingCa2� reuptake. However, even if Ca2� removal mechanisms

H1070 INa,L INHIBITION SLOWS THE PROGRESSION TO HEART FAILURE


were normal, the delayed Ca2� release in itself results inextended reuptake. This slowing in relaxation is likely tocontribute to diastolic dysfunction.

Figure 3 shows transverse line scan recordings across manycells in intact hearts of both 10-mo-old CON (A)- and RAN(B)-fed animals. The BCL for pacing was 700 ms. The trans-verse scan mode facilitates simultaneous measurement of Ca2�

transients in multiple myocytes. Note that the black horizontallines indicate cell edges; thus Fig. 3A shows recordings from

seven myocytes, and Fig. 3B shows recordings from six cells.Average fluorescence intensity is shown in the intensity profileabove each line scan. On the right are intensity profiles fromfour selected myocytes from each image. The cells in the CONheart generally show a slower rise and delayed time to peak aswell as slower decay to diastolic Ca2� levels than the myocytesin the RAN heart. The summary data in Fig. 3, C–F, show thatthere is a faster rate of Ca2� release (shorter rise time) inmyocytes from RAN-fed animals and shorter overall time to

Fig. 1. Measurements of t-tubule organizationindex (OI) in myocytes from hearts of con-trol- and ranolazine-fed rats. A and B: exam-ple 2-dimensional images of a cardiac myo-cyte from a control (A) and a ranolazine-treated (B) rat. The graphs below each panelshow the fast Fourier transform (frequency onthe x-axis; power on the y-axis) for each cellindicated with a red arrow in the images. Thefrequency range for bands 1 and 2 (B1 andB2, respectively) is indicated for calculationof OI, as described in METHODS. C andD: frequency (y-axis) histograms showing ex-amples of measurements of OI (x-axis) incardiomyocytes from individual control- andranolazine-fed rats. E and F: summary datafor all myocytes (n � 1,003 and 1,230 forcontrol and ranolazine, respectively) from allhearts (n � 8 rats in each group).



peak release. The durations of Ca2� transients at both 50 and90% of recovery (TD50, TD90) were also significantly shorterin the RAN than in the CON heart (Fig. 3, E and F).

Chronic block of INa,L increases synchrony of Ca2� cyclingwithin the myocyte. In addition to improving overall Ca2�

cycling properties (rates of Ca2� release and removal), RANalso prevented the loss of synchronization in Ca2� cycling

within the myocyte. Figure 4, A and B, shows line scan imagesof Ca2� transients recorded along the length of individualmyocytes in intact hearts of CON and RAN SHRs during basalpacing at BCL � 700 ms. Note that the Ca2� transient is morevariable along the cell length (rise time, magnitude, and dura-tion) in the myocyte from CON than from the RAN-fed rat.Ca2� cycling was then analyzed in 2-�m increments along the

Table 1. Summary of values of structural and functional parameters measured using echocardiography in control- andranolazine-treated rats

Echo Trait

Control Ranolazine Difference, 10 vs. 7 mo

7 mo 10 mo P value 7 mo 10 mo P value Control Ranolazine P value

LV mass, g 1.05 � 0.07 1.18 � 0.08 0.018 1.14 � 0.09 1.08 � 0.10 0.17 0.13 � 0.12 �0.06 � 0.11 0.005LVEDD, mm 7.80 � 0.53 7.50 � 0.21 0.15 7.84 � 0.79 7.40 � 0.40 0.20 �0.31 � 0.54 �0.44 � 0.87 0.73EF, % 83 � 3 84 � 5 0.76 84 � 5 84 � 6 0.84 0.8 � 6.7 0.6 � 8.3 0.97FS, % 47 � 3 49 � 7 0.56 48 � 6 48 � 7 0.97 1.8 � 8.0 �0.13 � 10.2 0.69E/A ratio 1.99 � 0.28 1.89 � 0.30 0.49 1.64 � 0.25 2.01 � 0.36 0.077 �0.11 � 0.40 0.38 � 0.42 0.054

Values (means � SE) were collected before (age of 7 mo) and after (10 mo) feeding of spontaneously hypertensive rats with either normal chow (control)or chow containing ranolazine for 3 mo (n � 8 rats in each group). LV, left ventricular; LVEDD, LV end-diastolic diameter; EF, ejection fraction; FS, fractionalshortening; E/A, ratio of the early (E) inflow of blood to the left ventricle through the mitral valve upon its opening, to the subsequent peak of the atrial (A)contraction-induced inflow of blood.

Fig. 2. Summary of data for mean cardiomyo-cyte cell length and width in control- andranolazine-fed spontaneously hypertensiverats (SHRs). A: 2-dimensional image ofmyocytes from a control heart. B: image of aranolazine heart. C: summary of mean celllength. D: summary of variability (SD) ofcell length. E: mean cell width. F: variabilityof cell width in hearts from control- (blackbars) and ranolazine-fed (light gray bars)rats. N � 1,024 and 1,234 myocytes from 8hearts in each group. *P � 0.05 vs. control;**P � 0.01 compared with control.



entire cell length, approximating the length of individual sar-comeres, and the SDs of individual parameters were calculatedto measure variability in Ca2� release and reuptake along thecell length during both basal (BCL � 700 ms) and rapid pacing(BCL � 300 ms). Values of SD for Ca2� release (rise time,Fig. 4, C and G) and for duration of the Ca2� transient at 50,80, and 90 of recovery (TD50, TD80, TD90; Fig. 4, D–F andH–J) were significantly decreased at both heart rates in theRAN myocytes compared with CON. These results indicatethat the increased variability in Ca2� cycling in CON isreduced by RAN and that this effect was independent of heartrate.

Chronic RAN treatment prevents the increased vulnerabilityto rate-dependent Ca2� alternans found in HF. We recentlyreported a shift in the rate dependence of Ca2� alternans tomuch slower heart rates during the progression of HF in theSHR model; this shift parallels disease development and aging(11, 34). The result is that Ca2� alternans develops at progres-sively slower heart rates during HF development, making thesehearts more vulnerable to reentrant arrhythmias. Alterations ofCa2� release from the SR feedback on Ca2�-sensitive sar-colemmal ionic currents and exchangers such as the L-typeCa2� channel current and the NCX current to increase thetemporal variability of action potential duration (APD) (25,26), which in turn establishes ventricular repolarization gradi-ents and therefore reentrant arrhythmias (36).

Figure 5 shows transverse line scan recordings of multiplemyocytes in two intact hearts during four paced beats at a BCL �300 ms. There was nearly complete Ca2� alternans in themyocytes from the CON heart (note average intensity profileabove the image; Fig. 5A). In contrast, none of the myocytes in

the RAN heart in Fig. 5B demonstrates Ca2� alternans. Figure5C shows the rate dependence of alternans development formyocytes from a RAN heart (black circles) and a CON heart(white circles). Note that Ca2� alternans magnitude is calcu-lated as alternans ratio so that alternans magnitude increaseswith decreasing BCL (increasing heart rate). The estimatedcycle length at 50% alternans (ECL50) was 341 ms in the CONheart compared with 274 ms in the RAN heart. The datasummarized in Fig. 5D for all myocytes of each type depict asignificant increase in vulnerability to Ca2� alternans (shift ofECL50 to longer cycle lengths) found in hearts of CON-relative to RAN-treated rats.

We have proposed that the Ca2� cycling instabilities respon-sible for Ca2� alternans development may be related to aslowing in the rate of recovery (restitution) of SR Ca2� releaseas heart rate increases (32). The rate of restitution of Ca2�

release in a CON myocyte is shown in Fig. 5E; the 50%recovery interval (R50) was 347 ms. This rather long recoveryinterval is consistent with findings that we have previouslyreported for the SHR (11, 34). The recovery interval was muchshorter in a myocyte from a RAN heart (R50 � 271 ms),demonstrating that the slowing in the recovery of SR Ca2�

release during HF development in the SHR was prevented byRAN treatment. The summary histogram in Fig. 5F shows thatSR Ca2� release restitution in RAN-treated SHR was fasterthan in CON-treated SHR.

Acute INa,L block increases Ca2� release rate and Ca2� re-uptake rate in HF myocytes. In three CON hearts (n � 9myocytes), we added RAN to the superfusate to investigate itsacute effects on Ca2� cycling in 10-mo-old SHRs. Figure 6, A and B,shows transverse line scan images from a cell in an intact CON

Fig. 3. Calcium transients in myocytes ofhearts from control- and ranolazine-fed rats.A: line scan image across multiple cells in anintact heart from a control rat. Horizontalblack lines indicate cell boundaries. Averageintensity profile for the entire site is shownabove the image. Intensity profiles for se-lected myocytes are shown to the right of theimage. B: line scan image from a ranolazineheart. F/F0, fluorescence intensity ratio.C–F: summary data for Ca2� transient risetime, time to peak, and durations at 50 and90% recovery (TD50 and TD90, respectively)to baseline in myocytes from control- (redbars) and ranolazine-fed (green bars) rats.n � 72–146 myocytes in 3 hearts each. *P �0.05 and **P � 0.01 compared with control.



heart at BCLs of 700 and 300 ms, respectively. Note that therise time of the Ca2� transient is slow at BCL � 700 ms,giving a rounded appearance to the Ca2� transient; this wasexaggerated during faster pacing (and consistent with resultssummarized in Fig. 3). RAN (10 �M) was added to thesuperfusate for 10 min and recordings were repeated at bothcycle lengths. Note the faster Ca2� release at both BCL � 700and 300 ms (Fig. 6, C–H) in the presence of RAN. The verticalline drawn between the recordings at the faster rate indicatesthe time at which peak magnitude of the Ca2� transient isachieved during RAN treatment. The time at which peakfluorescence occurs after RAN exposure is earlier than peakrelease before RAN treatment. The summary data in Fig. 6,E–L, show that the rate of Ca2� release (rise time, time to peak)is increased by RAN during rapid pacing and that the slowCa2� reuptake in HF is accelerated by RAN, again with agreater effect during rapid pacing. These data demonstrate that,in addition to chronic effects to slow the onset of Ca2� cyclingdefects, acute exposure to RAN also increases the rates ofCa2� release and reuptake, which are underlying determinantsof systolic and diastolic function.

Effects of chronic INa,L block on expression of Ca2� cyclingproteins. We further investigated whether ranolazine has anypotential effect on the expression level of different Ca2�

handling proteins essential to excitation-contraction coupling.Protein lysates were obtained from LV tissue from both CON(n � 4) and RAN rats (n � 5). We quantified Cav1.2, NCX,SERCA2a, Nav1.5, and RyR2 protein levels. Figure 7 showsthat expression of these proteins was not significantly changedin hearts from rats treated with RAN compared with CONanimals. Thus any differences between the experimentalgroups must occur as the result of RAN treatment that does notinvolve changes in expression levels of these proteins.

Ranolazine block of INa,L in SHRs at both 7 and 10 mo ofage. Finally, we made direct measurements of INa,L in SHRmyocytes to determine the magnitude of the current andwhether this current is blocked by relevant concentrations ofranolazine. Figure 8A shows a representative recording of INa,L

and the effects of RAN to reduce this current followed bycomplete block with tetrodotoxin. The concentration-responserelationships for ranolazine (Fig. 8B) show that the IC50 valuesfor block of INa,L by RAN were 3.7 � 0.01 and 4.7 � 0.25 �Min myocytes from 7- and 10-mo-old rat hearts, respectively(P � not significant). These data demonstrate that the 5.4 �Mconcentration of RAN found in the plasma of RAN-treated ratswas likely to have caused significant block of INa,L in myocytesfrom SHRs used in this study. Finally, the magnitudes of INa,L

were significantly different between LV myocytes from heartsof 3- to 4-mo-old Wistar and 10-mo-old SHRs (Fig. 8C, **P �0.01).

DISCUSSION

Recently, it has been suggested that the development of HFoccurs as the result of the accumulation of myocytes withdisrupted t-tubules (4, 5, 34, 35). Disorganization of the t-tu-bule network is at least partially responsible for the defects inintracellular Ca2� cycling and contractile function character-istic of myocytes in HF. When the number of myocytes withpoorly organized t-tubules is low, there is very little change incardiac performance. However, as the number of myocyteswith disrupted t-tubules increases, paralleled by a decrease inthe number of myocytes with normal t-tubule organization andcell function, there will be a progression from a compensatedhypertrophic stage to decompensated HF (35). This processprogresses with time until there are too few normal myocytes,

Fig. 4. Variability in Ca2� cycling along thecell length in myocytes in intact hearts fromcontrol- and ranolazine-fed rats. A andB: line scan images of Ca2� transients inindividual myocytes recorded longitudinallyfrom intact hearts of control and ranolazinetreated rats, respectively, during pacing at abasic cycle length (BCL) of 700 ms. Aver-age fluorescence intensity profiles are shownabove each image. C–J: heterogeneity in-dexes (HIs; i.e., values of SD) of meanvalues of rise time, TD50, TD80, and TD90 ofCa2� transients recorded in cardiomyocytesfrom control- (red bars) and ranolazine-fed(green bars) rats during pacing at BCL �700 (left) and 300 ms (right). * P � 0.05,** P � 0.01 compared with control.



too many myocytes with defective t-tubules (both ultrastruc-turally and functionally), and eventually replacement of deadmyocytes with fibrotic tissue.

The goal of the experiments in the present study was todetermine if block of INa,L might slow the progression to HF ina hypertension-induced model of cardiac hypertrophy. Themajor finding was that treatment of 7-mo-old SHR for 3 mowith ranolazine, an inhibitor of INa,L, reduced the percentage ofcardiac myocytes with poor t-tubule organization (i.e., thosewith an OI of 0.8 or lower; Fig. 1), shortened the duration andthe longitudinal variability of the Ca2� transient in individualmyocytes (Figs. 3, 4), reduced rate-dependent Ca2� alternans(Fig. 5), and improved the restitution of SR Ca2� release (Fig.5), compared with untreated control rats. Thus ranolazineimproved the stability and synchrony of cellular Ca2� han-dling, especially at a high rate (BCL � 300 ms) of pacing.These findings are consistent with previous reports that rano-lazine reduces Na�/Ca2� overloading and occurrences of de-layed afterdepolarizations in myocytes from failing hearts (30).The findings suggest that inhibition of late INa during thedevelopment of cardiac hypertrophy and failure may be bene-ficial to reduce both the progressive loss of t-tubule organiza-tion and the increase of contractile and rhythm disturbances.

There is growing evidence that INa,L is increased and con-tributes to electrical instability (arrhythmias) and LV diastolic

dysfunction in HF and during ischemia (6, 17, 30). We alsofound that the amplitude of INa,L in this study was greater inLV myocytes from SHRs than in control rats. It has beensuggested that CaMKII is activated (possibly by increaseddiastolic [Ca2�] or reactive oxygen species) and then phos-phorylates Na� channels, resulting in an increase in INa,L (16).The reverse may also be true, namely, that an increase in INa,L

raises [Na�]i and therefore [Ca2�]i, which in turn activatesCaMKII (40). Regardless of how INa,L is enhanced, the resultwould be the intracellular accumulation of Na�, which hasbeen identified in nearly all forms of HF in both animal modelsand in patients (22–24, 31, 32). An increase in Na� would inturn lead to a further increase in intracellular Ca2� via reverse-mode NCX, increasing Ca2� uptake into the SR, but because ofincreased SR Ca2� leak in HF (13), failing to increase SR Ca2�

load while maintaining an increased diastolic Ca2�. This ideais supported by the fact that inhibition of INa,L improved Ca2�

cycling defects and diastolic relaxation of contractile force inthis and previous (30, 31) studies.

From a more global perspective, it is interesting that therewere no significant changes in myocardial mechanics in controlSHRs as measured using advanced echocardiographic imagingover the 3-mo period of study. Systolic and diastolic functionswere unchanged despite clear evidence of t-tubule disruptionand accompanying alterations in Ca2� transients. The most

Fig. 5. Calcium alternans and restitution of sarcoplasmic reticulum Ca2� release in myocytes in hearts of control- and ranolazine-fed rats. A and B: line scanimages across multiple cells in hearts of control- (A) and ranolazine-fed (B) rats during pacing at BCL � 300 ms. Cells are separated by horizontal black lines.C: relationship between the magnitude of the alternans ratio [AR � 1 � (small/large)] and BCL in hearts from control- (Œ) and ranolazine-fed (�) rats. ECL50

is the estimated cycle length at which AR � 0.5. D: summary of ECL50 data for all myocytes [n � 11 and 16 in 3 hearts each from control- (red bar) andranolazine-fed (green bar) rats]. E: recovery of sarcoplasmic reticulum Ca2� release as a function of recovery interval. R50 is the cycle length at 50% recoveryof Ca2� transient magnitude (relative to magnitude recorded at a BCL of 700 ms). F: summary of R50 data for all myocytes [n � 7 and 16 in 3 hearts each fromcontrol- (red bar) and ranolazine-fed (green bar) rats]. *P � 0.05 and **P � 0.01 compared with control.



likely explanation for this apparent discrepancy is the fact thatthis study was performed during the earliest phases of diseaseprogression. It is known that hypertrophy is well developedand hypercontractile behavior occurs at about the initiationtime point of this study (27), so the heart is still in a fullycompensated state. It is not surprising that after 3 mo there arenot enough myocytes with poor t-tubule organization to inducean overall decline in cardiac function detectable by echocar-diographic imaging. Also, since our measurements were madeonly on the epicardial surface of the middle LV, we do notknow the extent of t-tubule remodeling throughout other re-gions of the ventricle (endocardium, base vs. apex) duringdisease progression. Our work and that of others (35) havedemonstrated that the number of affected cells increases withdisease severity in general, but it is likely that the fact thatwe do not see any overt changes in cardiac function suggeststhat the number of affected cells is just too small to producesignificant changes in myocardial mechanics. Our ongoingwork in older animals (J. A. Wasserstrom, G. L. Aistrup, L.Beussink, and S. Shah, unpublished observations) suggests thatthe decline in myocardial function occurs with a delay related toa requirement for a critical number of affected cells rather thanshowing a sudden onset when enough cells are affected (a thresh-old effect). Later, as disease progresses farther, there is a closercorrelation between t-tubule remodeling and disease progressionin the form of growing systolic and diastolic dysfunction (35)(J. A. Wasserstrom, G. L. Aistrup, L. Beussink, and S. Shah,unpublished observations).

T-tubule remodeling and INa,L. The mechanism(s) by whichranolazine decreased the percentage of myocytes with poort-tubule organization (OI�0.8) is unclear. The fact that INa,L

inhibition with ranolazine reduced t-tubule remodeling sug-gests that INa,L-mediated Na� influx and the subsequent NCX-mediated Ca2� accumulation are contributors to t-tubule dis-ruption during disease progression in the SHR. An increase inthe diastolic Ca2� concentration may ultimately cause mito-chondrial Ca2� overload and apoptosis (8) and/or necrosis(18). Expression of cardiac sodium channels with the long-QTsyndrome 3 mutation N1325S, which enhances INa,L, wasassociated with increased cardiomyocyte apoptosis and con-tractile dysfunction in mice (41). Ranolazine was reported toreduce cytosolic Ca2�, mitochondrial permeability transitionpore opening, and cell death after global ischemia and reper-fusion (which increase INa,L) in the guinea pig isolated heart(2). In sum, these findings and ours suggest that enhancers andinhibitors of INa,L may increase or reduce, respectively, injuryand cell death associated with Ca2� overload, including dis-ruption of t-tubule organization. However, because ranolazineis reported to stabilize cardiac ryanodine receptors (20) and toinhibit the rapid delayed rectifier K� current (IKr) (3), it ispossible that other actions of the drug, in addition to inhibitionof INa,L, may contribute to its beneficial effect in this study.However, since there is no IKr present in rat ventricle, it isunlikely that drug actions on this current could contribute to theeffects of ranolazine described here.

Fig. 6. Effects of acute ranolazine treatment on Ca2� transient characteristics during slow (BCL � 700 ms) and rapid (BCL � 300 ms) pacing. A and B: linescan images showing Ca2� transients in a control rat heart at BCL � 700 and 300 ms. Intensity profile is shown below. C and D: recordings from the same heartas in A and B after 10 min of exposure to ranolazine (10 �M). E–L: summary data for Ca2� transient characteristics before (red bars) and after 10 min exposureto ranolazine (green bars). n � 9 myocytes in 3 hearts. *P � 0.05 and **P � 0.01 compared with control.



We found that chronic INa,L block with ranolazine preventedthe desynchronization of Ca2� cycling along the cell lengthobserved in CON myocytes. This effect would be expected ift-tubule organization was better preserved, because Ca2� cy-cling properties should remain more synchronized and uniformwhen trigger Ca2� can uniformly activate sarcoplasmic retic-ular Ca2� release to enable rapid, synchronous, and efficientcontraction. It is clear that t-tubule disruption occurs as part ofthe pathophysiology that is ultimately responsible for HF.Early intervention that prevents t-tubule disruption at the cel-lular level may slow or even reverse the progression fromcompensated, asymptomatic to decompensated HF.

INa,L block decreases rate sensitivity to Ca2� alternans de-velopment. We found that chronic INa,L inhibition had impor-tant effects on the rate vulnerability of Ca2� alternans. Thisform of Ca2� instability is probably responsible for APDalternans and the resulting T-wave alternans that are thought tobe highly arrhythmogenic. APD alternans may underlie therepolarization gradients that are a substrate for reentrant ar-rhythmias (21, 25, 37). Thus the development of Ca2� altern-ans at slow heart rates increases the vulnerability to reentry.We have demonstrated that there is a shift in Ca2� alternanssusceptibility to progressively slower heart rates during theprogression of HF in the SHR model (11, 34). The underlyingmechanism responsible for this increasing vulnerability toCa2� alternans is likely to be related to the slowing in therecovery of SR Ca2� release as disease develops, making Ca2�

cycling unstable at increasing heart rates (11, 34). In this studywe found that chronic INa,L inhibition prevented both theincrease in Ca2� alternans vulnerability to slow rates and theslowing in SR Ca2� release restitution. This suggests 1) thatINa,L in SHRs contributes to the slowing in Ca2� restitution andresulting increased rate-dependent vulnerability to Ca2� alter-nans and 2) that chronic block of INa,L in the SHR decreasesvulnerability to Ca2� alternans. These results demonstrate for

the first time that chronic inhibition of INa,L not only preventsultrastructural and hypertrophic remodeling but also reducesthe substrate for reentrant arrhythmias that arise during devel-opment of HF. Inhibition of INa,L by ranolazine of electricalalternans may also contribute to its beneficial effects on cardiacfunction. Ranolazine reduces the transmural and beat-to-beatvariability of action potential (AP) repolarization (3, 38).Effects of ranolazine to decrease inward INa,L and formation ofearly afterdepolarizations and to reduce the [Na�]i and NCX-mediated Ca2� overloading may both be important to decreasethe susceptibility of cardiac tissue to Ca2� and electricalalternans. Again, it is interesting that ranolazine is effective inreducing alternans sensitivity in SHRs where IKr is absent andso cannot play a role in alternans formation, suggesting that theefficacy of the drug occurs through mechanisms that do notinvolve this current component.

Acute INa,L inhibition corrects Ca2 cycling defect. We alsofound that acute block of INa,L improved both Ca2� release andreuptake in CON myocytes with slow Ca2� release. The factthat acute RAN treatment improved Ca2� cycling properties inmyocytes showing distinct defects in excitation-contractioncoupling suggests that acute administration of this agent mightalso improve cellular and cardiac function in HF. Indeed, thismay represent a unique cellular target and novel approach forpharmacotherapy in decompensated HF.

The Western blot results demonstrated that RAN did notalter expression levels of a number of key proteins involved inexcitation-contraction coupling. Consequently, we concludethat the beneficial actions of RAN (5.4 �M) on HF develop-ment are likely to arise from its selective action to block INa,L.However, it is worth noting that it is also possible that RANcould indirectly reduce CaMKII activity (40), which couldhave important effects on Ca2� cycling and could contribute tothe beneficial effects of RAN treatment.

Fig. 7. Expression of Ca2� cycling proteins in left ventric-ular tissue from rats-fed control (C) or ranolazine-contain-ing (R) chow for 3 mo. A: Western blot showing expressionof voltage-dependent Ca2� 1.2 (Cav1.2), Na�/Ca2� ex-changer (NCX), sarco(endo)plasmic reticulum Ca2�-ATPase 2a(SERCA2a), voltage-gated Na� channel 1.5 (Nav1.5), andryanodine receptor type 2 (RyR2), with GAPDH as control.B: relative protein expression normalized to GAPDH inhearts from control- (black bars) and ranolazine-fed (whitebars) rats. Values indicate means � SE; n � 4 for C andn � 5 for R.



Clinical implications of chronic block of INa,L during HFdevelopment. Our results demonstrate that it is possible tointervene during the transition from hypertension to decom-pensated HF in a manner that 1) slows the progression of bothcellular and whole organ hypertrophy to decompensation, 2)significantly reduces t-tubule remodeling and the magnitude ofCa2� cycling defects that contribute to both systolic anddiastolic HF, and 3) reduces the vulnerability to Ca2� alternansand its potential contributions to arrhythmias. A putative rolefor INa,L in HF development raises the possibility that anyunderlying pathophysiological mechanisms responsible for in-creased INa,L might not only contribute to disease developmentand the progression to HF but may also provide potentialtherapeutic targets for the prevention of HF.

GRANTS

This project was funded by a grant from Gilead Sciences (to J. A. Wasserstrom).

DISCLOSURES

N. El-Bizri, S. Rajamani, J. C. Shryock, and L. Belardinelli are employ-ees of Gilead Sciences. Ranolazine (Ranexa) is owned by GileadSciences.

AUTHOR CONTRIBUTIONS

G.L.A., D.K.G., J.E.K., N.S., J.N., S.R., J.C.S., L. Belardinelli, S.J.S.,and J.A.W. conception and design of research; G.L.A., J.E.K., M.J.O.,A.F.N., N.C., S.M., L. Beussink, N.S., N.E.-B., S.R., and J.A.W. performedexperiments; G.L.A., J.E.K., M.J.O., A.F.N., N.C., S.M., L. Beussink, N.S.,J.N., M.R., T.M., N.E.-B., S.R., S.J.S., and J.A.W. analyzed data; G.L.A.,D.K.G., N.S., J.C.S., L. Belardinelli, S.J.S., and J.A.W. interpreted resultsof experiments; G.L.A. and J.A.W. drafted manuscript; G.L.A., S.R.,J.C.S., L. Belardinelli, S.J.S., and J.A.W. edited and revised manuscript;G.L.A., D.K.G., J.E.K., M.J.O., A.F.N., N.C., S.M., L. Beussink, N.S.,J.N., M.R., T.M., N.E.-B., S.R., J.C.S., L. Belardinelli, S.J.S., and J.A.W.approved final version of manuscript; M.J.O., N.E.-B., S.R., and J.A.W.prepared figures.

Fig. 8. Measurements of late sodium current (INa,L) in cardiac myocytes isolated from 7- and 10-mo-old SHRs. A: voltage-clamp protocol (top) along withrepresentative recordings of total and late (INa,L) sodium current from a single myocyte (10-mo-old SHR) in the absence (control) and presence of sodium channelblockers ranolazine (RAN, 5 �M) and tetrodotoxin (TTX, 10 �M). Inset: expanded tracings of INa,L. Test potential was �20 mV. B: concentration-responserelationships for inhibition by ranolazine of INa,L in 7- (�) and 10-mo-old (Œ) SHRs. nH, Hill coefficient. C: summary of normalized INa,L magnitude in 3-mo-oldWistar rats (WRs, n � 9), 7-mo-old SHRs (n � 4) and 10-mo-old SHRs (n � 7). nH **P � 0.01.



REFERENCES

1. Aistrup GL, Kelly JE, Kapur S, Kowalczyk M, Sysman-Wolpin I,Kadish AH, Wasserstrom JA. Pacing-induced heterogeneities in intra-cellular Ca2� signaling, cardiac alternans, and ventricular arrhythmias inintact rat heart. Circ Res 99: e65–e73, 2006.

2. Aldakkak M, Camara AK, Heisner JS, Yang M, Stowe DF. Ranolazinereduced Ca2� overload and oxidative stress and improved mitochondrialintegrity to protect against ischemia reperfusion injury in isolated hearts.Pharmacol Res 64: 381–392, 2011.

3. Antzelevitch C, Belardinelli L, Wu L, Fraser H, Zygmunt AC,Burashnikov A, Di Diego JM, Fish JM, Cordiero JM, Goodrow RJ Jr,Scornik F, Perez G. Electrophysiologic properties and antiarrhythmicactions of a novel antianginal agent. J Cardiovasc Pharmacol Ther 9,Suppl 1: S65–S83, 2004.

4. Balijepalli RC, Kamp TJ. Cardiomyocyte transverse tubule loss leads theway to heart failure. Future Cardiol 7: 39–42, 2011.

5. Balijepalli RC, Lokuta AJ, Maertz NA, Buck JM, Haworth RA,Valdivia HH, Kamp TJ. Depletion of t-tubules and specific subcellularchanges in sarcolemmal proteins in tachycardia-induced heart failure.Cardiovasc Res 59: 67–77, 2003.

6. Belardinelli L, Antzelevitch C, Fraser H. Inhibition of late (sustained/persistent) sodium current: a potential drug target to reduce intracellularsodium-dependent calcium overload and its detrimental effects on cardi-omyocyte function. Eur Heart J 6, Suppl 1: 13–17, 2004.

7. Brette F, Orchard C. T-tubule function in mammalian cardiac myocytes.Circ Res 92: 1182–1192, 2003.

8. Chen X, Zhang X, Kubo H, Harris DM, Mills GD, Moyer J, BerrettaR, Potts ST, Marsh JD, Houser SR. Ca2� influx-induced sarcoplasmicreticulum Ca2� overload causes mitochondrial-dependent apoptosis inventricular myocytes. Circ Res 97: 1009–1017, 2005.

9. Everett TH, Kok LC, Vaughn RH, Moorman R, Haines DE. Frequencydomain algorithm for quantifying atrial fibrillation organization to increasedefibrillation efficacy. IEEE Trans Biomed Eng 48: 969–978, 2001.

10. Ju YK, Saint DA, Gage PW. Hypoxia increases persistent sodium currentin rat ventricular myocytes. J Physiol 497: 337–347, 1996.

11. Kapur S, Aistrup GL, Sharma R, Kelly JE, Arora R, Zheng J,Veramasuneni M, Kadish AH, Balke CW, Wasserstrom JA. Earlydevelopment of intracellular calcium cycling defects in intact hearts ofspontaneously hypertensive rats. Am J Physiol Heart Circ Physiol 299:H1843–H1853, 2010.

12. Kapur S, Wasserstrom JA, Kelly JE, Kadish AH, Aistrup GL. Aci-dosis and ischemia increase cellular Ca2� transient alternans and repolar-ization alternans susceptibility in the intact rat heart. Am J Physiol HeartCirc Physiol 296: H1491–H1512, 2009.

13. Lehnart SE, Wehrens XH, Laitinen PJ, Reiken SR, Deng SX, ChengZ, Landry DW, Kontula K, Swan H, Marks AR. Sudden death infamilial polymorphic ventricular tachycardia associated with calciumrelease channel (ryanodine receptor) leak. Circulation 109: 3208–3214,2004.

14. Louch WE, Mork HK, Sexton J, Stromme TA, Laake P, Sjaastad I,Sejersted OM. T-tubule disorganization and reduced synchrony of Ca2�

release in murine cardiomyocytes following myocardial infarction. JPhysiol 574: 519–533, 2006.

15. Lyon AR, MacLeod KT, Zhang Y, Garcia E, Kanda GK, Lab MJ,Korchev YE, Harding SE, Gorelik J. Loss of t-tubules and otherchanges to surface topography in ventricular myocytes from failing humanand rat heart. Proc Natl Acad Sci USA 106: 6854–6859, 2009.

16. Maltsev VA, Reznikov V, Undrovinas NA, Sabbah HN, Undrovinas A.Modulation of late sodium current by Ca2�, calmodulin, and CaMKII innormal and failing dog cardiomyocytes: similarities and differences. Am JPhysiol Heart Circ Physiol 294: H1597–H1608, 2008.

17. Maltsev VA, Silverman N, Sabbah HN, Undrovinas AI. Chronic heartfailure slows late sodium current in human and canine ventricular myo-cytes: implications for repolarization variability. Eur J Heart Fail 9:219–227, 2007.

18. Nakayama H, Chen X, Baines CP, Klevitsky R, Zhang X, Zhang H,Jaleel N, Chua BH, Hewett TE, Robbins J, Houser SR, Molkentin JD.Ca2�- and mitochondrial-dependent cardiomyocyte necrosis as a primarymediator of heart failure. J Clin Invest 117: 2431–2444, 2007.

19. Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, RedfieldMM. Trends in prevalence and outcome of heart failure with preservedejection fraction. N Engl J Med 355: 251–259, 2006.

20. Parikh A, Mantravadi R, Kozhevnikov D, Roche MA, Ye Y, Owen LJ,Puglisi JL, Abramson JJ, Salama G. Ranolazine stabilizes cardiacryanodine receptors: a novel mechanism for the suppression of earlyafterdepolarizations and torsade de pointes in long QT type 2. HeartRhythm 9: 953–960, 2012.

21. Pastore JM, Girouard SD, Laurita KR, Akar FG, Rosenbaum DS.Mechanism linking t-wave alternans to the genesis of cardiac fibrillation.Circulation 99: 1385–1394, 1999.

22. Pieske B, Houser SR. [Na�]i handling in the failing human heart.Cardiovasc Res 57: 874–886, 2003.

23. Pieske B, Maier LS, Piacentino V 3rd, Weisser J, Hasenfuss G, HouserS. Rate dependence of [Na�]i and contractility in nonfailing and failinghuman myocardium. Circulation 106: 447–453, 2002.

24. Pogwizd SM, Sipido KR, Verdonck F, Bers DM. Intracellular Na inanimal models of hypertrophy and heart failure: contractile function andarrhythmogenesis. Cardiovasc Res 57: 887–896, 2003.

25. Restrepo JG, Karma A. Spatiotemporal intracellular calcium dynamicsduring cardiac alternans. Chaos 19: 037115, 2009.

26. Shiferaw Y, Sato D, Karma A. Coupled dynamics of voltage and calciumin paced cardiac cells. Phys Rev E Stat Nonlin Soft Matter Phys 71:021903, 2005.

27. Shorofsky SR, Aggarwal R, Corretti M, Baffa JM, Strum JM, Al-Seikhan BA, Kobayashi YM, Jones LR, Wier WG, Balke CW. Cellularmechanisms of altered contractility in the hypertrophied heart: big hearts,big sparks. Circ Res 84: 424–434, 1999.

28. Song LS, Sobie EA, McCulle S, Lederer WJ, Balke CW, Cheng H.Orphaned ryanodine receptors in the failing heart. Proc Natl Acad Sci USA103: 4305–4310, 2006.

29. Sossalla S, Maurer U, Schotola H, Hartmann N, Didie M, Zimmer-mann WH, Jacobshagen C, Wagner S, Maier LS. Diastolic dysfunctionand arrhythmias caused by overexpression of CaMKII�C can be reversedby inhibition of late Na� current. Basic Res Cardiol 106: 263–272, 2011.

30. Undrovinas A, Maltsev VA. Late sodium current is a new therapeutictarget to improve contractility and rhythm in failing heart. CardiovascHematol Agents Med Chem 6: 348–359, 2008.

31. Verdonck F, Volders PG, Vos MA, Sipido KR. Intracellular Na� andaltered Na� transport mechanisms in cardiac hypertrophy and failure. JMol Cell Cardiol 35: 5–25, 2003.

32. Verdonck F, Volders PG, Vos MA, Sipido KR. Increased Na� concen-tration and altered Na/K pump activity in hypertrophied canine ventricularcells. Cardiovasc Res 57: 1035–1043, 2003.

33. Wasserstrom JA, Kapur S, Jones S, Faruque T, Sharma R, Kelly JE,Pappas A, Ho W, Kadish AH, Aistrup GL. Characteristics of intracel-lular Ca2� cycling in intact rat heart: a comparison of sex differences. AmJ Physiol Heart Circ Physiol 295: H1895–H1904, 2008.

34. Wasserstrom JA, Sharma R, Kapur S, Kelly JE, Kadish AH, BalkeCW, Aistrup GL. Multiple defects in intracellular calcium cycling inwhole failing rat heart. Circ Heart Fail 2: 223–232, 2009.

35. Wei S, Guo A, Chen B, Kutschke W, Xie YP, Zimmerman K, WeissRM, Anderson ME, Cheng H, Song LS. T-tubule remodeling duringtransition from hypertrophy to heart failure. Circ Res 107: 520–531, 2010.

36. Weiss JN, Karma A, Shiferaw Y, Chen PS, Garfinkel A, Qu Z. From pulsusto pulseless: the saga of cardiac alternans. Circ Res 98: 1244–1253, 2006.

37. Weiss JN, Nivala M, Garfinkel A, Qu Z. Alternans and arrhythmias:from cell to heart. Circ Res 108: 98–112, 2011.

38. Wu L, Rajamani S, Li H, January CT, Shryock JC, Belardinelli L.Reduction of repolarization reserve unmasks the proarrhythmic role ofendogenous late Na� current in the heart. Am J Physiol Heart Circ Physiol297: H1048–H1057, 2009.

39. Wu Y, Clusin WT. Calcium transient alternans in blood-perfused isch-emic hearts: Observations with fluorescent indicator fura red. Am J PhysiolHeart Circ Physiol 273: H2161–H2169, 1997.

40. Yao L, Fan P, Jiang Z, Viatchenko-Karpinski S, Wu Y, Kornyeyev D,Hirakawa R, Budas GR, Rajamani S, Shryock JC, Belardinelli L.Nav1.5-dependent persistent Na� influx activates CaMKII in rat ventric-ular myocytes and N1325S mice. Am J Physiol Cell Physiol 301: C577–C586, 2011.

41. Zhang T, Yong S, Drinko JK, Popovic ZB, Shryock JC, Belardinelli L,Wang QK. LQTS mutation N1325S in cardiac sodium channel geneSCN5A causes cardiomyocyte apoptosis, cardiac fibrosis and contractiledysfunction in mice. Int J Cardiol 147: 239–245, 2011.



inhibition of the late sodium current slows t-tubule disruption during the progression of...

Documents