Skip Navigation

Cardiovascular Research 1998 38(2):395-404; doi:10.1016/S0008-6363(98)00011-X
© 1998 by European Society of Cardiology
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Bou-Abboud, E.
Right arrow Articles by Nattel, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bou-Abboud, E.
Right arrow Articles by Nattel, S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © 1998, European Society of Cardiology

Molecular mechanisms of the reversal of imipramine-induced sodium channel blockade by alkalinization in human cardiac myocytes

Elias Bou-Abboudb and Stanley Nattela,b,c,*

aDepartment of Medicine, Institut de Cardiologie de Montréal and Université de Montréal, McGill University, Montreal, Quebec, Canada
bDepartment of Pharmacology, Université de Montréal, McGill University, Montreal, Quebec, Canada
cDepartment of Pharmacology and Therapeutics, McGill University, Montreal, Quebec, Canada

* Corresponding author. Research Center, Montreal Heart Institute, 5000 Bélanger Street East, Montreal, Quebec, Canada H1T 1C8. Tel.: +1 (514) 376 3330; Fax: +1 (514) 376 1355; E-mail: nattel@icm.umontreal.ca

Received 4 June 1997; accepted 29 December 1997


   Abstract
Top
 Abstract
1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions
 Appendix A
 References
 
Background: Alkalinizing agents reverse cardiotoxicity of a variety of sodium channel blockers, including tricyclic antidepressants, but their mechanisms of action are poorly understood. Purpose: To establish the mechanisms by which alkalinization diminishes the sodium channel blocking action of imipramine. Methods: The whole-cell voltage-clamp technique was used to measure INa during a variety of depolarizing pulse protocols in isolated human atrial myocytes, in the presence and absence of imipramine. A three-state model was used to analyze state-dependent INa block. Results: Imipramine (1 and 5 µM) strongly inhibited INa. Experimental data and piecewise exponential analysis suggested significant binding to both activated and inactivated states. Alkalosis antagonized imipramine-induced INa blockade by increasing the unbinding rate, with intracellular alkalosis being more effective than extracellular alkalosis. The dissociation constant (Kd) for the inactivated state was increased from 0.55 to 1.40 µM by extracellular alkalosis and to 2.51 µM by intracellular alkalosis. Along with the reversal of drug-induced shifts in the inactivation curve, these data indicate that alkalosis on either side of the membrane antagonized drug interactions with the inactivated state. On the other hand, only intracellular alkalosis antagonized activated state block, increasing the Kd from 0.67 µM to 2.18 µM, while extracellular alkalosis left the activated state Kd unaltered at 0.67 µM. Conclusions: Alkalinization antagonizes the INa-blocking action of imipramine by promoting unbinding from the receptor. Intracellular alkalosis has a particularly important effect related to the activated-state interaction. The lipid-soluble, uncharged moiety appears to be a critical determinant of imipramine's ability to dissociate from the Na+ channel receptor.

KEYWORDS Sodium channel; Cardiac arrhythmias; Local anesthetics; pH; Antiarrhythmic drugs


   1 Introduction
Top
 Abstract
 1 Introduction
2 Methods
 3 Results
 4 Discussion
 5 Conclusions
 Appendix A
 References
 
Imipramine (Tofranil), a widely-used antidepressant drug [1], also has class 1a antiarrhythmic drug properties [2]. Overdose with tricyclic antidepressants causes cardiotoxic manifestations largely due to sodium channel blockade [3–5], which can be lethal. For severely poisoned patients, the most widely advocated therapy is the use of alkalinizing sodium salts [6–8], but the mechanism of action of sodium salt therapy is still poorly understood.

In a recent study, we found that the depressant effects of imipramine (5 µM) on maximum phase 0 upstroke velocity (Vmax) of canine cardiac Purkinje fibers were antagonized to a similar extent by increases in bicarbonate and sodium bicarbonate concentration, but not by increases in sodium concentration alone [9, 10], suggesting that alkalinization is the principal mechanism by which sodium bicarbonate antagonizes imipramine effects. The molecular mechanisms by which pH change modulates the effect of imipramine on INa have not been assessed. The present study was conducted to evaluate the mechanisms of the attenuation of imipramine-induced INa blockade by increased extracellular or intracellular pH, and to compare the interactions of the drug with intracellular vs. extracellular alkalosis.


   2 Methods
Top
 Abstract
 1 Introduction
 2 Methods
3 Results
 4 Discussion
 5 Conclusions
 Appendix A
 References
 
2.1 Cell isolation
Experiments were performed on isolated human atrial myocytes. Specimens were obtained from the apex of the right atrial appendage of the hearts of patients undergoing aortocoronary bypass surgery, with the approval of the Ethics Committee of the Montreal Heart Institute. None of the patients had a history of supraventricular arrhythmias. All atrial specimens were grossly normal at the time of excision. Upon excision, samples were immersed in nominally Ca2+-free Tyrode solution (100% O2, 37°C) containing (in mM): NaCl 126, KCl 5.4, MgCl2 1.0, NaH2PO4 0.33, dextrose 10, and HEPES 10; pH adjusted to 7.4 with NaOH.

The methods for cell isolation have previously been described in detail [11]. Tissue samples were chopped into cubic chunks, placed in Ca2+-free Tyrode solution bubbled with 100% O2 and stirred with a magnetic bar. After 5 min, they were reincubated in a similar solution containing collagenase (150–300 units/ml, type II, Worthington Biochem., Freehold, NJ) and protease (1.6 units/ml, type XXIV, Sigma Chemical, St. Louis, MO). The first supernatant was discarded after 45 min. The chunks were reincubated in a protease-free fresh enzyme-containing solution. When cell yield was maximal, the chunks were suspended in a high K+-solution containing (mM): KCl 25.5, KH2PO4 10, glucose 10, K glutamate 70, β-hydroxybutyric acid 10, taurine 10, EGTA 10, and 1% albumin; pH adjusted to 7.4 with KOH and gently pipetted. Only quiescent rod-shaped cells with a well-delineated cell membrane and clear cross-striations were used. Small cells were selected to optimize spacial voltage control.

2.2 Recording techniques
An aliquot of the cell-containing solution was placed in a 1-ml bath on the stage of an inverted microscope. After cell adhesion to the bath, cells were superfused at 2 ml/min with a solution containing (mM): NaCl 10, MgCl2 1.2, CaCl2 1, CsCl 127.3, HEPES 20, and glucose 11; pH adjusted to 7.3 or 7.6 with CsOH. CdCl2 (100 µM) was added to block calcium current. The bath temperature was maintained at 17°C with a temperature controlling device (model TC-202; Medical Systems, Greenvale, NY). The pipette (intracellular) solution consisted of (mM): NaCl 10, CsF 135, EGTA 10, HEPES 5, MgATP 5; pH adjusted to 7.2 or 7.5 with CsOH. Adequate voltage control was assessed in the I–V curve by (a) the absence of threshold phenomena or abnormal notches; (b) a negative limb spanning 30 mV or more; and (c) the absence of any crossover of INa traces as INa decreased with inactivation.

Borosilicate glass pipettes (1.0 mm o.d.) were pulled with a horizontal puller (model P-87; Sutter Instruments, Novato, CA) and heat-polished on a microforge. Pipettes (resistances of 1–2 M{Omega} when filled) were connected to a patch-clamp amplifier (Axopatch 1-D, Axon Instruments, Foster City, CA). Liquid junction potentials were zeroed before formation of the membrane-pipette seal, and the whole-cell mode achieved by light suction.

Command pulses were generated with an IBM compatible computer interfaced with a 12-bit digital-to-analogue converter. Recordings were low-pass filtered (5 kHz) and digitized (50–100 kHz). Series resistance (Rs) was compensated to minimize the duration of the capacitive transient. The capacitive time constant was 149±71 µs (mean±s.d.) before and 89±21 µs after compensation (cell capacitance 48±21 pF, n=21). Rs averaged 3.1±1.2 M{Omega} before and 1.9±1.1 M{Omega} after compensation. Experiments with a series voltage error >5 mV were discarded.

2.3 Experimental protocols
A holding potential (HP) of –140 mV was used for all experiments. The current-voltage relationship was determined with 40-ms pulses (5 mV steps) to voltages from –80 to –5 mV. The steady-state INa availability curve was constructed by applying 1-s conditioning prepulses to various test potentials followed immediately by a 30-ms test pulse to –30 mV. The recovery of INa was studied with a conditioning train (40 pulses to –30 mV of 50-ms duration, 200-ms interstimulus interval) followed by a 30-ms test pulse to –30 mV. Protocols with varying interpulse interval and pulse duration were used to study the dependence of block on rested and inactivated time respectively.

Results were obtained with extracellular pH (pHe) of 7.3 and intracellular pH (pHi) of 7.2 respectively (NORM condition), with extracellular alkalosis (ALKe; pHe 7.6, pHi 7.2), or with intracellular alkalosis (ALKi; pHe 7.3, pHi 7.5). All protocols were applied under drug-free NORM, ALKe, or ALKi conditions and then in the presence of drug (1 or 5 µM imipramine) at the same pH values. The pH values for study were selected to mimic the pH range observed in clinical imipramine intoxication before and after alkalinizing therapy [7, 8]. The drug concentrations studied were selected to mimic the high end of the therapeutic range (1 µM) and the concentration range at which cardiovascular toxicity becomes prominent (5 µM) [5].

2.4 Data analysis
A detailed treatment of the theoretical analysis is given in the Appendix. A three-state binding model was used to assess interactions with activated, inactivated, and rested channels. Mathematical expressions for the derivation of binding and unbinding rates were based on a piecewise exponential model [12, 13]. Linear regression analysis was used to calculate the slope and the Y-intercept for each dependence as predicted by equation 4 in the Appendix. The precision of the slope and the Y-intercept was characterized by their standard errors that delineate 95% confidence intervals. These errors were used to determine standard deviations for the calculated rate constants.

2.5 Statistical analysis and curve fitting
Data are expressed as the mean±s.d. unless otherwise stated. Analysis of variance with a Scheffé range test was used for multiple comparisons. A value of P<0.05 was taken to indicate statistical significance. Marquardt's procedure was used for non-linear curve fitting of experimental data.


   3 Results
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
4 Discussion
 5 Conclusions
 Appendix A
 References
 
3.1 Properties of human atrial INa
Fig. 1A shows representative sodium current tracings from one cell. The time to peak current was the same under all conditions (Table 1). Fig. 1B shows INa density as a function of test potential in nine cells studied at NORM, eight cells at ALKe and 10 cells at ALKi. No statistically significant difference was detected between mean currents at any voltage. The threshold for INa activation was –65 mV, the maximal peak current was between –30 and –35 mV, and the reversal potential estimated by extrapolating the ascending limb of the average current voltage-relationship was close to 0 mV as expected for symmetrical sodium concentrations.


Figure 1
View larger version (26K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 Properties of INa in human atrial myocytes. A. Recordings from one cell obtained with voltage protocol shown in inset (40-ms pulses, 1 Hz). B. Mean (±S.E.) current-voltage relations obtained in nine cells at pHe/pHi 7.3/7.2 (NORM), eight cells at pHe/pHi 7.6/7.2 (ALKe) and 10 cells at pHe/pHi 7.2/7.5 (ALKi). There were no significant pH-dependent differences between currents at corresponding voltages. C. Currents recorded with pulse protocol shown in inset, consisting of 1-s prepulses to voltages between –140 and –65 mV, followed by a 30-ms test pulse to –30 mV, with an interpulse interval of 1.2 s. D. Mean (±S.E.) voltage-dependent inactivation (h{infty}) curve obtained from nine cells at NORM, eight cells at ALKe and 10 cells at ALKi with voltage protocol shown in C. Data were fitted by the Boltzmann equation I/Imax=1/{1+[eV–V1/2/s]}, where I is the current amplitude for prepulse potential V, V1/2 is the potential giving half-maximal inactivation, Imax is the current for a prepulse potential of –140 mV and ‘s’ is the slope factor (curves shown).

 

View this table:
[in this window]
[in a new window]

  		Table 1 Properties of INa under various pH conditions in the absence of imipramine

 
Recordings used to study the voltage dependence of inactivation are illustrated in Fig. 1C, and corresponding mean data are shown in Fig. 1D. Under control conditions, no significant differences in the inactivation curves at NORM, ALKe and ALKi were noted (Table 1). There was a time-dependent inactivation shift with a similar rate for all three conditions, which was greatest initially, and slowed substantially 15 min after membrane rupture. At least 20 min were allowed for stabilization prior to obtaining data. To minimize the influence of time-dependent voltage shifts, we performed protocols in a constant order and as soon as possible after stable drug effects were obtained (maximum of 7 min). The response to a standard pulse to –30 mV was monitored before and after each experimental protocol, and data discarded if INa changed by ≥10%.

3.2 Modulation by alkalosis of steady-state INa blockade by imipramine
Fig. 2 shows the effect of intracellular and extracellular alkalosis on imipramine block of INa. Steady-state block is expressed as the fractional reduction in current relative to control at the same pH and frequency. At both concentrations, ALKi was more effective than ALKe in reducing drug-induced blockade. Only ALKi significantly reduced drug effect at the lower drug concentration, while both ALKe and ALKi reversed drug-induced blockade at the higher concentration. In the presence of 1 µM imipramine, tonic block on the first pulse after 90 s at the HP was small and equivalent for all conditions (4.3±3.3% for NORM, n=9; 3.5±3.4% for ALKe, n=7; 3.1±3.2% for ALKi, n=8). At 5 µM imipramine, tonic block was significantly reduced by both ALKe and ALKi (Table 2). Thus, alkalosis attenuates imipramine-induced INa block, with the effects of ALKi being greater than those of ALKe.


Figure 2
View larger version (20K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 Modulation by alkalosis of imipramine-induced INa blockade. Values shown are the fraction of current blocked by imipramine under a given condition relative to control at the same condition. *P<0.05, **P<0.01, ***P<0.001 vs. NORM; {dagger}P<0.05, {ddagger}P<0.01 for ALKe vs. ALKi.

 

View this table:
[in this window]
[in a new window]

  		Table 2 Selected effects of imipramine (5 µM) under various pH conditions

 
3.3 Modulation of drug effects on the voltage dependence of activation and availability
Imipramine caused a hyperpolarizing shift of both the activation and availability curves, indicating a probable interaction with the open and the inactivated state respectively. The shift of the steady-state availability curve following exposure to 5 µM imipramine was significantly attenuated by both ALKe and ALKi (Table 2). The imipramine-induced shift in the activation curve was affected differently by intracellular compared to extracellular alkalosis: the hyperpolarizing shift induced by imipramine was increased by ALKe and reduced by ALKi. Slope factors for availability and activation curves were unaffected by pH change.

3.4 Effects of pH change on sodium current recovery in the absence and presence of imipramine
Recovery of INa was studied with the use of the protocol shown in the inset of Fig. 3. In the presence of 5 µM imipramine recovery was slowed, to a greater extent under NORM (Fig. 3D) than at ALKe (Fig. 3E) and ALKi (Fig. 3F). Recovery data were well-fitted by monoexponential relations (Fig. 3, bottom). In the absence of drug, recovery was not affected by changes in pH over the range studied (Table 1). In the presence of imipramine, both extracellular and intracellular alkalosis significantly accelerated recovery (Table 2).


Figure 3
View larger version (24K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 Recovery of INa in representative cells under control conditions (top) and in the presence of imipramine (middle) under NORM (left), ALKe (middle) and ALKi (right) conditions. Recordings from A and D, B and E, and C and F were from one cell under each condition before and after drug exposure, at the recovery intervals indicated (for protocol see inset). The recovery intervals for all control recordings are given in panel C, those for imipramine and NORM condition are shown in D, while those for imipramine and ALKe or ALKi are given in panel F. Current at each recovery interval (RI) was normalized to current after a 5-s pause (control) or a 90-s pause (drug) and plotted as a function of recovery interval under control (G) and drug (H) conditions. The best-fit monoexponential functions (solid lines shown) to the data from the cells shown provided a {tau}rec of 12.4 ms at NORM, 7.4 ms at ALKe and 3.9 ms at ALKi under control conditions; and 18 s under NORM, 7.4 s under ALKe and 7.3 s under ALKi conditions in the presence of imipramine.

 
3.5 Effect of pH alteration on kinetics of use-dependent block and calculation of binding and unbinding rates
Fig. 4 shows INa recordings during a pulse train. Results in A and D were obtained from one cell under NORM conditions before and after imipramine, those in B and E were obtained from another cell under ALKe while as those in C and F from a different cell under ALKi. Under control conditions, there was very little use-dependent reduction of INa. In the presence of imipramine, however, INa showed a progressive reduction with successive pulses. The onset of use-dependent block at various interpulse intervals is illustrated in panels G–I. Block onset becomes progressively more rapid and steady-state block decreases as the interpulse interval (rested time) increases.


Figure 4
View larger version (28K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 4 Use and frequency-dependent block of INa by imipramine under NORM (left), ALKe (middle) and ALKi (right) conditions. Thirty 50-ms pulses from –140 to –30 mV were applied at interpulse intervals (IPI) of 0.2, 0.5, 1, 1.5, or 2 s. Trains were separated by 10 s (control) or 90 s (drug). Top: Recordings during voltage pulses 1, 3, 5, 10, 20 and 30 with interpulse interval 200 ms are shown for control under NORM (A), ALKe (B) and ALKi (C) conditions. Middle: Recordings from same cells during same pulses as on top in presence of imipramine (5 µM) under NORM (D), ALKe (E) and ALKi (F) conditions. G, H and I show the onset of block in the presence of drug from the representative experiments shown in A–F, with peak INa during each pulse normalized to INa of the first pulse of the train. The best monoexponential curve fits are shown by the solid lines through the points.

 
To vary inactivated time, the interpulse interval was fixed at 200 ms (long enough to assure recovery of unblocked channels), and pulse duration for each train varied from 2 to 200 ms. Fig. 5 (top panels) shows recordings of INa for pulses 1, 3, 5, 10, 20, 30 and 40 of trains with the pulse durations indicated. INa was constant for trains of different pulse width under control conditions, but in the presence of imipramine, use-dependent block developed at different rates and to different levels of steady-state inhibition depending on pulse width. Block onset was an exponential function of pulse number at each pulse width, with pulse onset faster and steady-state block greater as pulse width increased (bottom).


Figure 5
View larger version (31K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 5 Pulse-width dependence of block of INa by imipramine under NORM (left), ALKe (middle) and ALKi (right) conditions. Thirty to 40 pulses (from –140 to –30 mV) of 10-, 50- and 200-ms durations were applied at a fixed interpulse interval of 200 ms. Trains were separated by 10 s (control) or 90 s (drug). For each pulse width tested, recordings of pulses 1, 3, 5, 10, 20, 30, and 40 (except for PW 200 ms, for which pulse 40 superimposed on pulse 30) are shown for the same cell before and after exposure to 5 µM imipramine. The lower graphs show the onset of block in the presence of drug from the same experiments as illustrated above. The best monoexponential curve fits are shown by the solid lines through the points.

 
As described in the Appendix, kinetic analysis of the type of data obtained in Figs. 4 and 5Go allows for the calculation of binding kinetics for Na+ channel interaction with the drug. To calculate forward (ks) and reverse (ls) rate constants for any channel state (‘s’), we estimated the uptake rate ({lambda}s) with the data at 1 and 5 µM imipramine. This provided two equations of the form {lambda}s=ksC+ls, which were used to calculate ks and ls (Equation 1 in Appendix). Single-channel theory and observations suggest that the mean open time corresponds to the inactivation time constant ({tau}i) of the whole-cell current [14]. We therefore used mean {tau}i under each condition as an estimate of ta, as have previous investigators [15]. The overall uptake rate {lambda}* was found (as predicted) to be a linear function of the rested time given by the interpulse interval Tr (when interpulse interval is changed) and inactivated time given by the pulse duration Ti (when pulse width is varied). From these linear relations, the uptake rates {lambda}r and {lambda}i can be obtained for each series of experiments and the rate constants ka, la, ki, li, kr and lr can be estimated, providing the results shown in Table 3. Whereas both ALKe and ALKi reduced drug affinity for inactivated channels, only ALKi increased the Kd for activated channels, consistent with the greater efficacy of ALKi in antagonizing imipramine-induced INa block.


View this table:
[in this window]
[in a new window]

  		Table 3 Binding (k) and unbinding (l) rates for imipramine interaction with sodium channels

 

   4 Discussion
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
5 Conclusions
 Appendix A
 References
 
In the present study, we evaluated the alterations in imipramine effects on INa caused by alkalinization at clinically-relevant pH values and drug concentrations. The results show that intracellular alkalosis is more effective in reducing INa blockade than extracellular alkalosis. Whereas both intracellular and extracellular alkalosis antagonize the interaction between imipramine and inactivated channels, only intracellular alkalosis reduces drug interactions with activated channels.

4.1 Comparison with previous studies of imipramine's sodium channel blocking action
There are no previously published studies regarding the effect of ph change on imipramine-induced ina blockade. Several groups of investigators have demonstrated the ability of imipramine to depress Vmax in cardiac preparations [2, 16–21]. Delpon et al. found that 1 µM imipramine produced 3.9% tonic block of Vmax in the guinea pig ventricle, comparable to the value of 4.3±3.3% we obtained for the same concentration [21]. Habuchi et al. [22] estimated an imipramine Kd of 0.66 µM for the inactivated state, similar to the value (0.55 µM) that we obtained. Ogata and Narahashi [23] reported an 18±2 mV shift in the inactivation curve with 5 µM imipramine, compared to our observed value of 9.4±2.4 mV. Barber et al. [24] found that extracellular (but not intracellular) alkalosis accelerates the time constant of amitryptiline-induced INa block from about 14 to 4 s. The details of pH modulation of amitryptiline's state-dependent effects were not explored in that study.

4.2 State-dependent interactions of imipramine with cardiac sodium channels
The piecewise exponential formulation allows for the determination of kinetic constants characterizing the drug-channel interaction [12, 13, 25]. This approach has generally been applied to two-state models, which require that the blocking drug have negligible interactions with one of the three classical sodium channel states. A three-state analysis has only been applied in two previous studies [26, 27], and has not previously been used to analyze pH-dependent drug-channel interactions. Unlike previous more limited studies of imipramine [22, 23], our results indicate that the open-state interaction is significant, and that in fact the Kd for the activated state is in the same range as for the inactivated state (Table 3).

4.3 Comparison with previous studies of pH modulation of the effects of cardiac sodium channel blockers
We did not observe significant changes in INa recovery or voltage dependence when extracellular pH was increased from 7.3 to 7.6 under control conditions. Previous studies have suggested small and variable changes in INa properties over the pH range we examined [28–30]. Decreases in pH slow the recovery of INa from block by lidocaine [30, 31], quinidine [32], nicardipine [29] and W6211 [33]. Recent work has suggested that increases in pH can augment a slowly developing component of block in the presence of disopyramide [34], lidocaine [34] and mexiletine [35]. We did not observe a biexponential onset of imipramine block during pulse trains in the present experiments. Little information is available regarding differences between extracellular and intracellular pH effects on drug-INa interactions. Nettleton and Wang found that increased extracellular pH decreases, while increased intracellular pH increases, the Kd for cocaine block of Na+ channels in lipid bilayers [36]. These results resemble ours, in that we found intracellular alkalosis to be much more effective in antagonizing the actions of imipramine.

4.4 Mechanisms of pH effect on imipramine binding
We found that ALKe and ALKi produced qualitatively similar changes in the inactivated state interaction, as indicated by shifts in the availability curve (Table 2) and alterations in kinetic constants (Table 3). On the other hand, intracellular alkalosis reduced activated-state block more effectively than extracellular alkalosis. Extracellular alkalosis increased the binding and unbinding rates to the activated state to a comparable extent (Table 3), leaving the overall Kd unchanged, while intracellular alkalosis greatly increased the unbinding rate of activated channels while slightly decreasing the binding rate. Since alkalosis increases the non-ionized fraction of the drug, these results are compatible with the concept that imipramine binding to the activated Na+ channel is determined by the access of uncharged drug from the extracellular side, while the unbinding rate is determined by a hydrophobic pathway on the intracellular side [37]. The similar effects of intracellular and extracellular alkalosis on interactions with the inactivated state suggest that unbinding from this state can occur via a hydrophobic pathway at the intracellular or extracellular side of the channel.

4.5 Potential significance
We found that intracellular alkalosis is more effective than extracellular alkalosis in reversing imipramine effects on INa. Interventions which alter pHe have varying effects on pHi, with changes in bicarbonate concentration being less effective in altering pHi than changes in extracellular CO2 content [38]. Both sodium bicarbonate administration [7, 8] and hyperventilation [39] have been used to treat cardiotoxicity caused by tricyclic antidepressants. Our findings raise the possibility that the efficacy of these interventions may depend more on their ability to alter pHi than on changes in pHe as reflected by standard arterial blood gas measurements. Furthermore, acute myocardial ischemia causes greater changes in pHi than pHe, because intracellular acidosis is the direct result of anaerobic metabolism caused by myocardial ischemia [38, 40]. Consequently, our finding that intracellular pH change can modulate drug-induced INa blockade differently from extracellular pH change may have important potential implications for antiarrhythmic drug action on the acutely ischemic myocardium.

4.6 Limitations of the model
The conditions required for accurate voltage change of INa differ from conditions prevailing in vivo. In particular, low extracellular [Na+] and temperature are necessary to control the amplitude and activation rate of INa. Extrapolation to the clinical condition must therefore be cautious. A decrease in [Na+]o can affect the drug-Na+ channel interaction [41–43]. We used low concentrations of Cd2+ (100 µM) to prevent contamination of sodium currents by ICa. While Cd2+ can alter INa as well [44, 45], Cd2+ was present under all conditions and should not have significantly altered the drug-channel interaction. Previous investigators studying the actions of blocking drugs on INa have used a variety of agents to suppress ICa, including Cd2+ [30], Co2+ [22, 23] and nicardipine [29].

Results for the rested state interaction were problematic because of the negative forward rate constants (Table 3), which clearly have no physical meaning. Possible explanations include the very low affinity for the rested state, resulting in a forward rate constant so close to zero that experimental error produces negative values, and biological violation of essential assumptions of the model.

Extracellular and intracellular pH were assumed to be equal to pH in the bath and pipette solution respectively. This is a common and reasonable assumption, given the much greater volume of the bath and the pipette compared to the cell. Nevertheless, it is possible that the pH within the channel or in a restricted subsarcolemmal space was not controlled. This possibility is difficult to assess directly and needs to be kept in mind in interpreting our findings.


   5 Conclusions
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions
Appendix A
 References
 
Alkalosis antagonizes imipramine's INa inhibiting effect. Whereas both intracellular and extracellular alkalosis reduce drug affinity for inactivated channels, only intracellular alkalosis reduces drug binding to activated channels, making it more effective in reversing drug-induced INa block.

Time for primary review 19 days.


   Appendix A
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions
 Appendix A
References
 
Sodium channel blockade can be described by a bimolecular interaction of drug D, with unblocked channel, U, to form blocked channel, B:


Formula

where ks and ls are state-dependent binding and unbinding rates. For each channel state, the blocking time course can be described by a monoexponential function: where


Formula

the rate constant {lambda} at each state s, {lambda}s, and drug concentration C is given by


Formula 1

(1)
and equilibrium block


Formula 2

(2)
During periodic application of voltage steps, the sodium channel moves from the resting state (r) through the activated state (a) to the inactivated state (i). Since activated and inactivated states generally have a higher affinity for the drug than the resting state, the peak of the sodium current decreases with each successive pulse in an exponential fashion [12, 13]:


Formula 3

(3)
where Iss is the steady-state current, Io is the current elicited by the first pulse and {lambda}* is the apparent binding rate in pulse–1. According to the piecewise exponential model, {lambda}* has the following relation to individual binding and unbinding rates [13]:


Formula 4

(4)
where {lambda}r, {lambda}a, and {lambda}i are interaction rates with the resting, activated and inactivated states and tr, ta, and ti the intervals that channels spend in each state. The overall uptake rate {lambda}* is predicted to be a linear function of the product of the uptake rate of each state and the time the channel spends in that state. The uptake rate for each state is constant at a given drug concentration. Since sodium channels inactivate rapidly after opening, ta is relatively small and constant, and was estimated on the basis of {tau}i under each condition as described in section 3.5. The inactivated time is well-approximated by pulse duration (Ti), and ti differs from Ti by the time channels are in the activated state prior to inactivation (ta) and by the limited time channels remain inactivated following return to the holding potential. For pulse durations ≥10 ms, the resulting error is small and constant and Ti can therefore be used as a surrogate for true inactivated time. Data for pulse durations <10 ms were not used in kinetic analysis. Similarly Tr is a close reflection of rested state time over the range of pulse protocols used in the present study.


   Acknowledgements
 
The authors thank Guylaine Nichol and Johanne Doucet for technical assistance, and Diane Campeau and Carolyn Gillis for typing the manuscript. They would also like to thank Doctors Conrad Pelletier, Michel Carrier, Raymond Cartier, Michel Pellerin, and Jean-Paul Martineau for providing samples of atrial tissue for cell isolation, and Doctor Vladislav Nesterenko for his very valuable suggestions and critique of the modeling work.


   References
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions
 Appendix A
 References
 

  1. Simon G.E., VonKorff M., Heiligenstein J.H., et al. Initial antidepressant choice in primary care. Effectiveness and cost of fluoxetine vs tricyclic antidepressants. JAMA (1996) 275:1897–1902.[Abstract/Free Full Text]
  2. Glassman A.H., Bigger J.T. Jr. Cardiovascular effects of therapeutic doses of tricyclic antidepressants: a review. Arch Gen Psychiatry (1981) 38:815–820.[Abstract/Free Full Text]
  3. Thorstrand C. Clinical features in poisonings by tricyclic antidepressants with special reference to the ECG. Acta Med Scand (1976) 199:337–344.[Web of Science][Medline]
  4. Freeman J.W., Mundy G.R., Beattie R.R., Ryan C. Cardiac abnormalities in poisoning with tricyclic antidepressants. Br Med J (1969) 2:610–611.[Abstract/Free Full Text]
  5. Marshall J.B., Forker A.D. Cardiovascular effects of tricyclic antidepressant drugs: therapeutic usage, overdose, and management of complications. Am Heart J (1982) 103:401–414.[CrossRef][Web of Science][Medline]
  6. Bismuth C., Bodin F., Pebay-Peroula F. Intoxication par l'imipramine avec insuffisance cardiaque aigue. La Presse Med (1968) 48:78–79.
  7. Brown T.C.K., Barber G.A., Dunlop M.E., Loughnan P.M. The use of sodium bicarbonate in the treatment of tricyclic antidepressant-induced arrhythmias. Anaesth Intensive Care (1973) 1:203–210.[Medline]
  8. Hoffman J.R., McElroy C.R. Bicarbonate therapy for dysrhythmia and hypotension in tricyclic antidepressant overdose. West J Med (1981) 134:60–64.[Web of Science][Medline]
  9. Bou-Abboud E., Cardinal R., Nattel S. Drug-specificity of active moiety of sodium salts in reversing local anesthetic cardiotoxicity. Can J Cardiol (1993) 9:32E.
  10. Bou-Abboud E., Nattel S. Relative role of alkalosis and sodium ions in reversal of class I antiarrhythmic drug- induced sodium channel blockade by sodium bicarbonate. Circulation (1996) 94:1954–1961.[Abstract/Free Full Text]
  11. Wang Z., Fermini B., Nattel S. Delayed rectifier outward current and repolarization in human atrial myocytes. Circ Res (1993) 73:276–285.[Abstract/Free Full Text]
  12. Starmer C.F., Grant A.O. Phasic ion channel blockade: a kinetic model and parameter estimation procedure. Mol Pharmacol (1984) 28:348–356.[Web of Science]
  13. Starmer C.F. Theoretical characterization of ion channel blockade: ligand binding to periodically accessible receptors. J Theor Biol (1986) 119:235–249.[CrossRef][Web of Science][Medline]
  14. Aldrich R., Corey D., Stevens C.F. A reinterpretation of mammalian sodium channel gating based on single channel recording. Nature (1983) 306:436–441.[CrossRef][Medline]
  15. Packer D.L., Grant A.O., Strauss H.C., Starmer C.F. Characterization of concentration- and use-dependent effects of quinidine from conduction delay and declining conduction velocity in canine Purkinje fibers. J Clin Invest (1989) 83:2109–2119.[Web of Science][Medline]
  16. Matsuo S. Comparative effects of imipramine and propranolol on the transmembrane potentials of the isolated rabbit atria. Jpn J Pharmacol (1967) 17:279–286.[Medline]
  17. Manzanares J., Tamargo J. Electrophysiological effects of imipramine in non-treated and in imipramine-pretreated rat atrial fibres. Br J Pharmacol (1982) 79:167–175.[Web of Science]
  18. Rodriguez S., Tamargo J. Electrophysiological effects of imipramine on bovine ventricular muscle and Purkinje fibres. Br J Pharmacol (1980) 70:15–23.[Web of Science][Medline]
  19. Rawling D., Fozzard H. Effects of imipramine on cellular electrophysiological properties of cardiac Purkinje fibres. J Pharmacol Exp Ther (1979) 209:371–375.[Abstract/Free Full Text]
  20. Brennan F. Electrophysiological effects of imipramine and doxepine on normal and depressed cardiac Purkinje fibres. Am J Cardiol (1980) 46:599–606.[CrossRef][Web of Science][Medline]
  21. Delpón E., Valenzuela C., Tamargo J. Tonic and frequency-dependent Vmax block induced by imipramine in guinea pig ventricular muscle fibers. J Cardiol Pharmacol (1990) 15:414–420.[Web of Science][Medline]
  22. Habuchi Y., Furukawa T., Tanaka H., Tsujimura Y., Yoshimura M. Block of Na+ channels by imipramine in guinea pig cardiac ventricular cells. J Pharmacol Exp Ther (1991) 256:1072–1081.[Abstract/Free Full Text]
  23. Ogata N., Narahashi T. Block of sodium channels by psychotropic drugs in single guinea pig cardiac myocytes. Br J Pharmacol (1989) 97:905–913.[Web of Science][Medline]
  24. Barber M.J., Starmer C.F., Grant A.O. Blockade of cardiac sodium channels by amitriptyline and diphenylhydantoin. Evidence for two use-dependent binding sites. Circ Res (1991) 669:677–696.
  25. Starmer C.F., Nesterenko V.V., Gilliam F.R., Grant A.O. Use of ionic currents to identify and estimate parameters in models of channel blockade. Am J Physiol (1990) 259:H626–H634.[Web of Science][Medline]
  26. Starmer C.F., Nesterenko V.V., Undrovinas A.I., Grant A.O., Rosenshtraukh L.V. Lidocaine blockade of continuously and transiently accessible sites in cardiac sodium channels. J Mol Cell Cardiol (1991) 23:I73–I83. (suppl.I).
  27. Furukawa T., Koumi S.I., Sakakibara Y. An analysis of lidocaine block of sodium current in isolated human atrial and ventricular myocytes. J Mol Cel Cardiol (1995) 27:831–846.[CrossRef][Web of Science][Medline]
  28. Courtney K.R. pH and voltage dependence of INa recovery kinetics in atrial cells exposed to lidocaine. Am J Physiol (1988) 255:H1554–H1557.[Web of Science][Medline]
  29. Gilliam F.R. III, Rivas P.A., Wendt D.J., Starmer C.F., Grant A.O. Extracellular pH modulates block of both sodium and calcium channels by nicardipine. Am J Physiol (1990) 259:H1178–H1184.[Web of Science][Medline]
  30. Wendt D.J., Starmer C.F., Grant A.O. pH dependence of kinetic and steady-state block of cardiac sodium channels by lidocaine. Am J Physiol (1993) 264:H1588–H1598.[Web of Science][Medline]
  31. Grant A.O., Strauss L.J., Wallace A.G., Strauss H.C. The influence of pH on the electrophysiological effects of lidocaine in guinea pig ventricular myocardium. Circ Res (1980) 47:542–550.[Free Full Text]
  32. Grant A.O., Trantham J.L., Brown K.K., Strauss H.C. pH-dependent effects of quinidine on the kinetics of dV/dtmax in guinea pig ventricular myocardium. Circ Res (1982) 50:210–217.[Free Full Text]
  33. Moorman J.R., Yee R., Bjornsson T., Starmer C.F., Grant A.O., Strauss H.C. pKa does not predict pH potentiation of sodium channel blockade by lidocaine and W6211 in guinea pig ventricular myocardium. J Pharmacol Exp Ther (1986) 238:159–166.[Abstract/Free Full Text]
  34. Sunami A., Fan Z., Nitta J.I., Hiraoka M. Two components of use-dependent block of Na+ current by disopyramide and lidocaine in guinea pig ventricular myocytes. Circ Res (1991) 68:653–661.[Abstract/Free Full Text]
  35. Ono M., Sunami A., Sawanobori T., Hiraoka M. External pH modifies sodium channel block by mexiletine in guinea pig ventricular myocytes. Cardiovasc Res (1994) 28:973–979.[Abstract/Free Full Text]
  36. Nettleton J., Wang G.K. pH-dependent binding of local anesthetics in single batrachotin-activated sodium channel: cocaine vs. quaternary compounds. Biophys J (1990) 58:95–106.[Web of Science][Medline]
  37. Hille B. Local anesthetics: hydrophillic and hydrophobic pathways for the drug-receptor reaction. J Gen Physiol (1977) 69:497–515.[Abstract/Free Full Text]
  38. Poole-Wilson P.A. Measurement of myocardial intracellular pH in pathological states. J Mol Cell Cardiol (1978) 10:511–526.[CrossRef][Medline]
  39. Kingston M.E. Hyperventilation in tricyclic antidepressant poisoning. Crit Care Med (1979) 7:550–555.[Web of Science][Medline]
  40. Tsien R.W. Possible effects of hydrogen ions in ischemic myocardium. Circulation (1976) 53:I14–I16.[Medline]
  41. Kohlhardt M. A quantitative analysis of the Na+-dependence of Vmax of the fast action potential in mammalian ventricular myocardium. Pflugers Arch (1982) 392:379–387.[CrossRef][Web of Science][Medline]
  42. Ranger S., Sheldon R., Ferminim B., Nattel S. Modulation of flecainide's cardiac sodium channel blocking actions by extracellular sodium: A possible cellular mechanism for the action of sodium salts in flecainide cardiotoxicity. J Pharmacol Exp Ther (1993) 264:1160–1167.[Abstract/Free Full Text]
  43. Johns J.A., Takafumi A., Bennet P.B., Snyders D.J., Hondeghem L.M. Temperature and voltage dependence of sodium channel blocking and unblocking by o-desmethyl encainide in isolated guinea pig myocytes. J Cardiovasc Pharmacol (1989) 13:826–835.[Web of Science][Medline]
  44. Grant A.O. Evolving concepts of cardiac sodium channel function. J Cardiovasc Electrophysiol (1990) 1:53–67.[CrossRef]
  45. Sheets M.F., Hanck D.A., Fozzard H.A. Divalent block of sodium current in single canine cardiac Purkinje cells. Biophysical J (1987) 53:535a.

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Mol. Pharmacol.Home page
N. Lenkey, R. Karoly, J. P. Kiss, B. K. Szasz, E. S. Vizi, and A. Mike
The Mechanism of Activity-Dependent Sodium Channel Inhibition by the Antidepressants Fluoxetine and Desipramine
Mol. Pharmacol., December 1, 2006; 70(6): 2052 - 2063.
[Abstract] [Full Text] [PDF]

Home page
 CirculationHome page
Part 4: Advanced Life Support
Circulation, November 29, 2005; 112(22_suppl): III-25 - III-54.
[Full Text] [PDF]

Home page
 J. Pharmacol. Exp. Ther.Home page
C. Nau, M. Seaver, S.-Y. Wang, and G. K. Wang
Block of Human Heart hH1 Sodium Channels by Amitriptyline
J. Pharmacol. Exp. Ther., March 1, 2000; 292(3): 1015 - 1023.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Bou-Abboud, E.
Right arrow Articles by Nattel, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bou-Abboud, E.
Right arrow Articles by Nattel, S.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?