To evaluate the effects of liothyronine (3,5,3′-triiodo-l-thyronine, T3) on Na+ channel current (I Na) properties, I Na was recorded in adult guinea pig ventricular myocytes. T3 (1 nM) acutely increased whole cell I Na and shifted the steady-stateI Na inactivation curve dose dependently. When the pipette solution contained 100 μM GTP or GTPγS, the effect of T3 on the whole cell I Na was increased two- to threefold. This effect was almost completely abolished by pertussis toxin preincubation. In the cell-attached patch, T3 increased the open probability of singleI Na by reducing the null probability. In the inside-out patch, T3 effect was 10 times faster than that in whole cell and cell-attached patches while GTPγS was present and could be completely washed out. T3 alone slightly increased the channel open probability by increasing the closed state to open state rate constant (k CO) and reducing the null probability. GTPγS exposure only increased the number of functional channels. T3 and GTPγS synergistically enhanced the channel open probability 5.8 ± 0.5-fold by increasingk CO, decreasing the open state to absorbing inactivated state rate constant, and greatly reducing the null probability. These results demonstrate that T3 acts on the cytosolic side of the membrane and acutely activatesI Na. Pertussis toxin-sensitive G protein modulation greatly magnifies the T3 effects on the channel kinetics and null probability, thereby increasing the channel open probability.
- pertussis toxin-sensitive G protein
- whole cell sodium channel current
- single channel current
- cardiac myocytes
thyroid hormones play an important role on cardiac electrophysiological properties (3, 6, 42). The effects of thyroid hormones have been considered to exert their actions through nuclear and/or extranuclear mechanisms. Nuclear effects are mediated by the binding of thyroid hormones to specific nuclear receptors, resulting in increased transcription of thyroid hormone-responsive cardiac genes (16,34). Extranuclear effects, which are independent of nuclear effects, primarily influence the transport of amino acids and sugars through the cell membrane (5, 39). Recently, specific binding sites for thyroid hormones on the plasma membrane have been reported (15, 40). However, few studies (4, 7,43) have demonstrated the effects of thyroid hormones on cardiac ionic channels. We (37) have previously reported that liothyronine (3,3′,5-triiodo-l-thyronine, T3) enhanced single inward rectifier K+ channel currents by increasing the channel open probability (P o) in cell-attached patches. Rubinstein and Binah (36) reported that Ca2+ channel currents and delayed rectifier K+ channel currents are increased in hyperthyroidism, and Ca2+ channel currents are decreased in hypothyroidism in guinea pig myocytes. Subsequently, T3-induced increase in cardiac L-type Ca2+ current was observed in in vitro studies (20, 33). It has also been found that acutely administered T3 promoted slow inactivation kinetics of the Na+ channel current (I Na) in neonatal rat myocytes (19). Dudley et al. (7) reported that T3 quickly induced I Nabursting in cell-attached patches in rabbit ventricular myocytes. However, the molecular mechanism of the effect of thyroid hormone on the ion channels in cardiac myocytes is not well understood. In this study, we explored the detailed mechanisms of G protein-modulated acute T3 action on I Na in ventricular myocytes.
MATERIALS AND METHODS
Single ventricular myocytes were isolated from adult male guinea pig (250–350 g) hearts using a modified procedure, as previously described (34). In brief, the guinea pigs were anesthetized with ether. Hearts were rapidly excised, mounted on a Langendorff-type apparatus, and perfused retrogradely through the aorta at 37°C. Hearts were perfused for 5 min with nominally Ca2+-free Tyrode solution containing (in mM) 136 NaCl, 5.4 KCl, 0.3 NaH2PO4, 1.0 MgCl2, 10 HEPES, 4d-mannitol, 0.6 thiamine HCl, and 5.5 glucose (pH 7.4 with NaOH). Hearts were then perfused for 4–6 min with Ca2+-free solution containing 1 mg/ml type I collagenase (Sigma; St. Louis, MO), 1 mg/ml bovine albumin (Sigma), and protease (7.6 mg/50 ml, Sigma). Thereafter, the collagenase was washed out with a high-K+ and low-Cl− solution, which was composed of (in mM) 115.9 KOH, 80 glutamic acid, 10 taurine, 14 oxalic acid, 10 KH2PO4, 10 HEPES, 25 KCl, 0.5 EGTA, and 11 glucose (pH 7.4 with KOH). The perfusion rate was ∼5–8 ml/min. Ventricular cells were gently dispersed and immersed in 0.1 mM Ca2+-Tyrode solution with albumin (1 mg/ml, bovine, Sigma), which was made by the addition of 0.1 mM CaCl2 to Ca2+-free Tyrode solution. After 10–20 min, 1.7 mM CaCl2 was added to the above solution, and cells were stored in this solution (1.8 mM Ca2+). All solutions were aerated with 95% O2-5% CO2. Only the cells that were Ca2+ tolerant, rod shaped, and having clear striations without any blebs on the surface were selected for the experiments.
Whole cell patch-clamp recording.
Na+ current was recorded from single cardiac myocytes by the whole cell patch- clamp technique described by Hamill et al. (18). An axopatch-1D amplifier (Axon Instruments; Foster City, CA) was used for the voltage-clamp experiments. Protocol generation and data acquisition were controlled with an IBM computer and 12-bit analog-to-digital and digital-to-analog converter by using pCLAMP software. The internal solution contained (in mM) 130 CsCl, 10 NaCl, 2 Mg2Cl, 5 Cs2-EGTA, 10 HEPES, and 4 ATP (pH 7.4). The external solution contained (in mM) 40 NaCl, 95 tetraethylammonium (TEA), 5.4 CsCl, 1 CaCl2, 1 MgCl2, 10 HEPES, and 10 glucose (pH 7.4).
Single channel recording.
Single I Na in ventricular myocytes were recorded by using cell-attached patches, according to the patch-clamp method described by Hamill et al. (18). Pipettes were fabricated from thin-wall glass and were coated near the tip with silicon (Sylgard, Dow Corning) to reduce the capacitance to ground and were then heat polished. When the pipettes were filled with the solution, the resistance of pipettes was 8–15 MΩ. The sealing resistance after negative pressure was applied to the inside of the pipette (∼30 cmH2O) was >30 GΩ. Signals for single channel currents were recorded with the patch-clamp amplifier (Axopatch-200B, Axon Instruments) and stored on a hard drive of an IBM computer and optic disk. Currents were sampled every 50 μs and recorded signals were filtered with a cutoff frequency of 2 kHz (−3 dB, four-pole low-pass Bessel filter). Patches were held at −140 mV and depolarized once per second to test potentials (V T) (−60, −50, −40, and −30 mV) for 50 ms. At least 300 sweeps (5 min) were recorded at each V T. Patches with four channels or less were used for this study. Patches that showed obvious rundown were omitted. Control data were recorded at least 5 min after the gigaohm seal was obtained, as time-dependent changes in kinetics of singleI Na were remarkable during the first 5 min after seal formation. To estimate time-dependent changes in kinetics of single I Na, cells were perfused with an external solution without T3 and data were recorded 0–5, 5–10, 15–20, and 25–30 min, respectively, after the formation of gigaohm seals at a V T of −40 mV. Data that could be recorded for 20 min after the formation of gigaohm seals without deterioration or loss of gigaohm seals were used for analysis. All recordings were obtained at room temperature (22–24°C). For cell-attached recordings, isolated cells were superfused with a high-K+ external solution of the following composition (in mM): 115.9 KOH, 80 glutamic acid, 14 oxalic acid, 10 KH2PO4, 10 HEPES, 25 KCl, 0.5 EGTA, and 11 glucose (pH 7.4 with KOH). The resting potential for these cells was assumed to be 0 mV and no correction was applied for the holding potential and V T. The high-K+external solution with 1 μM T3 was adjusted pH (7.4) after the addition of T3 with KOH. Pipette solution contained (in mM) 140 NaCl, 1 CaCl2, 5 MgCl2, 10 TEA chloride hydrate, and 10 HEPES (pH 7.4 with NaOH). Single channel inside-out patch bath solutions contained (in mM) 140 K+ aspartate, 1.0 EGTA, 0.55 CaCl2, 2.0 Mg2+ ATP, and 10 HEPES (pH 7.25).
Experimental protocols and data analysis.
The voltage dependence of the steady-state inactivation relationship was examined with a standard two-pulse protocol. AV T of −20 mV was preceded by a 500-ms preconditioning pulse ranging from −150 to +20 mV with a 0.1-Hz pulse interval. Under these conditions, the slow inactivation-dependent shift in the inactivation curves was minimized. The normalized curves were fitted using a Boltzmann distribution equation.
Single channel currents from patch-clamp recordings were analyzed using the software program pCLAMP (version 6.01, Axon Instruments). Recordings were corrected off-line for leak and capacitive currents by analog circuitry and by subtracting the average of recordings without channel openings (null sweeps) at each given potential. Openings detected by the computer were confirmed by visual observation and used for further analysis. The threshold to judge the open state was set at half of the unit amplitude of the single channel current. The number of channels in each patch was estimated from channel overlap at the test potentials of −20 to −10 mV over 1,000 sweeps. Open time histograms were constructed from the data in which opening events with overlapping openings of two or more channels were excluded. The first bin was omitted to fit open time histograms because dead time for data filtered at 2 kHz was 90 μs. Ensemble currents were calculated by the following equation: 1,000 (sum of sweeps)/[(no. of channels)(no. of sweeps)] as the currents of 1,000 channels. Null probability (P N) was calculated by taking the nth root of the fraction of observed null sweeps (where nrepresents the estimated number of channels in the patch).P o plots were constructed by averaging the probability of openings at each time per one channel with the number of sweeps. Assuming that each channel opening shows independent identical distribution, the number of openings per one sweep with openings was calculated by the following equation: (averaged no. of openings per one sweep)/[(1 − P N)(no. of channels)].
All data were expressed as means ± SD unless otherwise indicated. Statistical significance was evaluated by Student's paired and unpaired t-test, where appropriate. Differences with a value of P < 0.05 were considered to be significant.
T3 increases whole cell INa.
T3 (1 nM) significantly increased whole cell Na+ current density by 10 ± 2% compared with that in control conditions at a holding potential of −100 mV (n = 18, P < 0.05). Figure1 A illustrates the acute effect of 10 nM T3 on the whole cellI Na. At a V T of −10 mV and a holding potential of −100 mV at 24°C, the peakI Na increased 47%. As shown in Fig.1 B, T3 increased the peak of the current-voltage (I-V) relationship curve, but did not change the shape of the curve or the reversal potential. After exposure to 10 nM T3, the steady-state availability curve for whole cellI Na was slightly but significantly shifted to the right (3.4 ± 0.9 mV, n = 18,P < 0.05, Fig. 1 C). As shown in Fig.1 D, the effect of T3 on whole cellI Na was dose dependent. The EC50 of T3 was 8.7 nM. The activation curve was slightly shifted to the left by T3 (0.9 ± 0.4 mV), but this change was not statistically significant (n = 18,P = 0.058, data not shown).
Reverse 3′,5′,3-l-triiodothyronine (rT3) (10 nM), which has been reported to partially compete with T3and tetraiodothyronine on the membrane receptors, did not increase whole cell I Na (Fig.2 A). In cells preincubated with 10 nM rT3 (30 min), the T3 (10 nM) effect on the whole cell I Na at −40 mV was blocked by 79% (P < 0.05) 25 min after superfusion with T3 was started. If cells preincubated with rT3were then washed for 2 h, the effect of T3 could be partially restored (37%) (n = 8, P < 0.01).
3,5,3′-Triiodothyroacetic acid (TAA) (10 nM), an analog of T3, significantly increased whole cellI Na currents in a manner similar to T3 (Fig. 2 A) (37). The increment inI Na caused by TAA was the same as the increment caused by a comparable concentration of T3 (62% ± 8% at −40 mV and 45 ± 7% at V T = −100 mV, n = 5).
G protein modulates effect of T3 on whole cell INa.
As shown in Fig. 2 B, the effect of T3 could be observed 8 min after cells were exposed to 10 nM T3 and reached a steady state in ∼25–35 min, but these effects could not be washed out during the subsequent 30 min of observation. Therefore, the detailed time-dependent change using a control solution without T3 was evaluated. There was no significant change in whole cell I Na current observed during 35 min of superfusion, but significant time-dependent rundown was observed at 45 min after superfusion was started (data not shown). These tendencies were observed at either temperature (24 and 37°C). As shown in Fig.2 B, when the pipette solution contained 100 μM GTP, the onset of T3 (10−8 M) effect onI Na was significantly earlier. The time to one-half peak response was 12.7 ± 2.2 min with GTP compared with 21.8 ± 3.7 min without GTP (n = 15,P < 0.01). The maximum effect was reached within 17–21 min after T3 exposure.
Figure 3 A shows the current-voltage relation for the tetrodotoxin-sensitiveI Na effected by T3 recorded with either 100 μM GTP or 100 μM GTPγS in the pipette. When GTP or GTPγS was present in the pipette solution, the peakI-V curves were increased 73 ± 11% or 75 ± 14% by T3 compared with control conditions. These increments were twofold greater than that recorded in the normal pipette solution. Our preliminary study has shown that incubation of myocytes with 200 ng/ml pertussis toxin at 33°C for 3 h resulted in virtually complete ADP ribosylation of sensitive G protein by endogenous NAD+ (41). Therefore, the same amount of pertussis toxin was used for preincubating myocytes. Pertussis toxin abolished most of the effect of T3 on theI Na when either GTP or GTPγS was present in the pipette solution. The voltage dependence of the T3 (10 μM) effect on I Na inactivation was recorded when the pipette solution with or without GTP or GTP-γS. When GTP or GTP-γS was present, T3 significantly shifted the inactivation curve toward a more positive potential (7 ± 2 mV,n = 15, P < 0.05). This effect was significantly greater than that seen without GTP or GTP-γS (P < 0.05) (Fig. 1 C).
T3 effects on single INa in cell-attached patch experiments.
In considering the slow onset of T3 effect, the time dependence of the single I Na activity was examined under control conditions. It seemed that traces recorded at 0–5 min have less null sweeps and more openings. The peak of ensemble currents recorded at 0–5 min was significantly greater than that recorded at 5–10 (P < 0.005), 15–20 (P < 0.01), and 25–45 min (P < 0.05). There was no significant change in the peak of ensemble currents during 5–45 min. Within 5 min after the formation of gigaohm seals, the changes varied; 7 of 10 patches showed a decrease, 1 of 10 patches showed an increase, and 2 of 10 patches showed no change. The unit amplitude did not change while recording. The number of openings decreased by 8% in 5 min and thereafter remained constant. The mean open time (t o) gradually increased by 12%, and P N gradually increased by 17%, reaching a plateau in 5 min, but these changes were not statistically significant. These results suggest that singleI Na was not stable within the first 5 min and relatively stable over the 5–45 min after the formation of gigaohm seals under our experimental conditions. Therefore, we recorded all data for the estimation of T3 effects on singleI Na during a period of 5–30 min after the formation of gigaohm seals.
Sixteen consecutive sweeps at V T = −50 mV in control and 25 min after superfusion with 10 nM T3 are shown in Fig. 4 A. This patch had three channels. After superfusion with T3, the number of null sweeps was decreased and the number of openings was increased. The unit amplitude was unchanged after T3 superfusion. Figure 4 B shows the change in the ensemble currents by T3. The peak of the ensemble currents with T3was significantly greater than the peak without T3 atV T = −60 to −30 mV. AtV T = −30 mV, the percent change in the peak of the ensemble currents induced by T3 was 45% (n = 8, P < 0.005) and was similar to that seen in the whole cell recordings (Fig. 4 C). As shown in Fig. 5 A,P N was 0.4–0.7 in controls and the relation between P N and the membrane potential was V shaped with a turning point at V T = −40 mV. The value of P N was the same as that previously reported (38). T3 reducedP N significantly atV T = −60 to −30 mV by 16–43%. No increase in the number of available channels was observed while superfusing with T3. In controls, t owas 0.2–0.8 ms at V T = −60 to −20 mV and longer at more positive membrane potentials but showed minimal voltage dependence over a range of V T = −30 mV (Fig. 5 B). This tendency is consistent with previous reports, but the values of t o were shorter than that previously reported (0.3–3.0 ms) (24, 26, 38). Shorter t o is considered to be due to a little higher experimental temperature (22–24°C) and/or shorter sampling interval (50 μs). The values and the tendency against the membrane potential were not altered by T3. The number of openings was 1.3–1.7 and increased at more negative membrane potentials in controls (35). T3 did not changet o or the number of openings atV T = −60 to −30 mV (Fig. 6, Aand B). The relation between the unit amplitude and the membrane potential was shown in Fig.6 C. The slope conductance was 5.3 pS in control and 5.2 pS 25 min after superfusion with T3. The effects of T3 on P o were also evaluated. Figure7 A demonstrated theP o in a test pulse duration of 50 ms in control solution and 25 after superfusion with 10 nM T3 atV T = −30 mV. T3 increased the peak P o from 8 to 16%. The effects of T3 on the peak P o were plotted against the membrane potential in Fig. 7 B. The peakP o was significantly increased atV T = −60 to −30 mV (57% atV T = −30 mV, n = 8,P < 0.005).
T3 effects on single INa in inside-out patch experiments.
Figure 8 A demonstrates 16 consecutive sweeps and the ensemble currents before and 25 min after superfusion with 1 nM T3 and 100 μM GTPγS atV T = −40 mV in an inside-out patch configuration containing two channels. Similar to our results in the cell-attached patch experiments, T3 decreased the number of null sweeps and increased the peak of the ensemble currents. As shown in Table 1, the peak of the ensemble currents increased 64% by 1 nM T3 (P < 0.01), P N was reduced 35% (P < 0.01), and t o was slightly increased (P = 0.052) at V T = −40 mV. The effect of 1 nM T3 with 100 μM GTPγS on the peak of the ensemble currents was 39.7 ± 8.7% greater compared with cell-attached patch experiments (P < 0.01). However, the effects on P N were quantitatively the same as that in cell-attached patch experiments. To elucidate the detailed mechanism, we assumed a five-state model, which is given by the scheme
where R1, R2, and C are closed, rested states, O is an open state, I is a closed, absorbing inactivated state, and the rate constants k CO,k OC, k CI, andk OI represent closed state to open state, open state to closed state, closed state to absorbing inactivated state, and open state to absorbing inactivated state, respectively (38). As shown in Table 1, 1 nM T3 with 100 μM GTPγS significantly increased k CO for the transition from the C to O state. k OI, for the transition from the O to I state, was significantly decreased by T3. However, the other rate constants were not significantly changed at any membrane potential tested.
Figure 8 B shows a reproducible inside-out patch recording from a cell sequentially treated with T3, GTPγS, and then with T3 and GTPγS together. With no GTP or GTPγS present, application of 10 nM T3 onto the cytoplasmic surface of the cell membrane only slightly increased theP o of single I Na (17 ± 4%, P < 0.05) by reducingP N (P < 0.05). The rate constant, k CO, was only slightly increased (8 ± 2%, P = 0.043), andk OI was slightly decreased (4 ± 1%), but this trend did not reach statistical significance (P = 0.052). These effects could be completely washed out. GTPγS (10 μM) alone significantly increased the peak of ensemble current 48.9 ± 12.2% (n = 25, P < 0.01) and slightly increased the P o of the singleI Na. GTPγS increased the number of functional channels in 7 of 15 patches with two channels, but did not alter the single I Na amplitude, t o,P N, or the channel kinetics. ATP was not present in the bath solution, suggesting that the effects of GTPγS were not mediated by activating adenylate cyclase within the patch (23). In the same patch, when 10 μM GTPγS was present, the P o of single I Na was immediately increased by the addition of 10 nM T3 onto the cytoplasmic side of the membrane. This effect was observed within 30 s after T3 exposure and reached a plateau within 3 min. T3 and GTPγS synergistically increased theP o by 5.8 ± 0.5-fold, and this effect was much greater than that seen in whole cell and cell-attached experiments. P N was reduced 91 ± 3% compared with that in control conditions, but singleI Na amplitude was unchanged. The rate constantk CO increased 3.2 ± 0.7-fold, andk OI was significantly decreased (18 ± 4%,P < 0.01). The increase in I Nawas reversible on T3 and GTPγS washout.
T3 activates INa through extranuclear mechanism.
Our results provide direct evidence that T3 produces a dose-dependent increase in I Na in adult ventricular myocytes through an extranuclear mechanism. Although a T3 receptor on the plasma membrane of cardiac myoyctes has not been structurally characterized, numerous studies have suggested its functional existence. The onset of T3 effects onI Na was evident within 5–15 min in whole cell and cell-attached patch experiments. The time course of the T3 effect on I Na is consistent with that on K+ or Ca2+ channel current in adult cardiac myocytes (19, 33, 37). In neonatal myocytes, however, the effect of T3 on I Naoccurred relatively faster, although the changing of current density and the shifting of the inactivation curve were similar to adult ventricular myocytes (20). This difference might result from the different membrane permeability to thyroid hormone in the neonatal myocytes compared with adult myocytes. The time frame of T3 effect on ion channels is considered to be too short to express specific genes within the nucleus of the cell (39). The same amount of enhancement inI Na current was induced by TAA, an analog of T3 that was considered, does not induce DNA transcription in the nucleus (37). However, recent studies (1,29) suggest TAA also has a transcriptional effect. In the present study, we used an inside-out patch technique to examine the effects of T3 on the I Na in the plasma membrane. For the first time, we have been able to provide direct evidence that T3 acts on the cytoplasmic side of the plasma membrane and activates the I Na through an extranuclear mechanism. This explains the relatively slow action of T3 observed in the whole cell and cell-attached patch experiments reported previously by our laboratory and other groups (19, 20, 33, 37). In one exception, Dudley et al. (7) found a significant increase in the singleI Na bursting in the cell-attached patches with T3 in the pipette solution. In that study, the comparison between control and T3 groups was made in different patches. The bursting occurred less than a millisecond after patch excision. Concern has been raised that the recording was made immediately after the patch formed, which is in the unstable period for parameters of single I Na. This data could be affected by time-dependent kinetic changes of singleI Na and their variability (12, 17). Thus the observations may be hardly distinguishable from artifacts of the patch maneuver in both cell-attached patch and inside-out patch reported previously (23, 24). On the other hand, the high tension on the patched membrane might facilitate the T3 in the pipette to permeate into the membrane and promote the fast action of T3 on the channel.
G protein modulates the T3 effect on INa.
Our results demonstrate that a pathophysiological concentration of T3 acts directly on the cytosolic side of the membrane and weakly, but significantly activates I Na. The synergism of a pathophysiological concentration of pertussis toxin-sensitive G protein and T3 greatly magnifies the effects of T3. In a whole cell configuration with GTP-free pipette solution, 10 nM T3 increased the peakI Na 47%, and slightly shifted the inactivation curve. With the addition of 100 μM GTP or GTPγS into the pipette solution, the effects of T3 on the peakI-V curve of I Na and the shifting inactivation curve were shifted more than two fold. Pertussis toxin greatly reduced, but did not completely abolish, the effect of T3, suggesting a direct effect of T3 onI Na. Because most G proteins involve channel activation, the effects of T3 on I Nawere also more prominent at hyperpolarized potentials (27,28). The rightward shift of the steady-state inactivation curve seen in this study was similar to that observed in other hormonal effects which involve the second messenger or signal transducing components (24). Nevertheless, previous studies (27,28) have shown that the direct effects of G protein onI Na are not associated with changes in inactivation kinetics. Therefore, T3 was responsible for the change in the inactivation kinetics. In cell-attached configurations, the physiological concentration of GTP was preserved. T3 effects were initiated earlier than that in whole cell recordings with GTP-free medium, but still later than that in the whole cell patches with a saturating concentration of GTP in the medium. In the inside-out configuration, without GTPγS present, T3only slightly increased the single I Na P o, but this effect was initiated 10 times faster than that in whole cell and cell attached configurations. This effect could be washed out completely in the inside-out configuration. These observations provide direct evidence that T3 acts on the cytosolic side of the membrane and directly activatesI Na. In the presence of the GTPγS, T3 greatly increased the P o of the channel within 1 min and reached a maximum within 3 min. This time course was consistent with other G protein-involved channel regulation, such as isoproterenol (28). The strong interaction of the channel with a pathophysiological concentration of T3 in the presence of GTPγS suggests that G proteins might induce a conformational change that has a profound impact on the interaction of the channel with T3. Petit-Jacques et al. (35) reported the synergistic activation of muscarinic K+channel by G proteins and Na+ and Mg2+. Our findings imply that variations in the local levels of G protein may have great impact on the interaction of a pathophysiological concentration of T3 with I Na in ventricular myocytes.
Detailed mechanism of T3 effects on kinetics of single INa.
The present study demonstrates that T3 increased theP o of single I Na by reducing the P N without changingt o or the number of available channels. Horn et al. (21) proposed that I Na can inactivate without opening. Nulls are considered to occur by inactivation of the channel directly from the closed state without entering the open state. However, Scanley et al. (38) mentioned four alternative possibilities that could cause overestimation and at least one that would cause underestimation ofP N: 1) an opening occurred that was too short to be detected (2), 2) an early opening occurred that was obscured by the capacity transient,3) channels were already in the inactivated state at the time of the step, or 4) channels remaining in a closed, noninactivated state throughout the step (31).P N is reported to be higher at more negative potentials (31). However, in this study, the relation between the membrane potential and P N was V shaped. At positive potentials, such as −30 mV, waiting time to the first opening was very short and reopening was rare. Therefore, the capacitive transient could obscure early openings atV T = −30 mV more than those at more negative membrane potentials and P N could be overestimated. In a basic five-state model for singleI Na, P N isk CI/(k CI +k CO) (38). As we reported previously, T3 accelerated the transition from a deep closed state to a shallow closed state in a single inward rectifier K+ channel using a three-state model (C1-C2-O) (31). In the present study, T3 also significantly increasesk CO, accelerates the transition, and decreasesP N. Scanley et al. (38) mentioned that the relation between P N and the membrane potential showed a negative shift at warmer temperatures because thek CO transition rate could be more sensitive to temperature than the k CI rate. This suggests that the sensor to temperature effects in I Namight be same as that to thyroid hormonal effects.
Most prior reports on t o for cardiac singleI Na have indicated that the distribution is well fitted to a single exponential function (12, 32), although some investigators have reported a two exponential distribution (26). Burst type openings have been reported to be different from nonburst type openings (9, 25, 31). The time constant of t o in bursts is longer (3–5 times) than that of nonburst type openings. This suggests that the single I Na have at least two open states. However, the incidence of burst type openings is rare (<1%) (9, 25, 31). In this study, its incidence was <0.3% and was not altered by T3. Therefore, the burst type openings were excluded from analysis.
In a single open state, t o is described by the inverse of the sum of rate constants leaving the open state. Therefore, in a basic five-state model, t o could be calculated as (k OC +k OI)−1. In cell-attached patches, a physiological concentration of G protein remained. The addition of T3 extracellularly did not change t oat V T = −60 to −20 mV, suggesting that T3 does not significantly change the sum ofk OC and k OI. In inside-out patches, the increase of k CO and the decrease of k OI induced by a pathophysiological concentration of T3 were revealed, and these effects were independent of G protein. The presence of a pathophysiological concentration of GTPγS and T3 resulted in a stimulation of channel activity that was much greater than the simple sum of their individual effects. G protein increases the number of functional channels, which could have impact on the effect of T3 on the channel P o (27). However, this does not fully explain the accelerated effects of T3 on channel kinetics in the presence of G protein. It is more likely that T3 and G protein exert their combined effects by synergistic interaction of different channel sites. G protein induces channel conformational changes that may sensitize the channel for T3-induced alterations in channel kinetics, thereby reducing P N and increasing the channelP o. The increased number of functional channels by G protein might further magnify the action of T3 on the channel open probability.
The upstroke velocity of the action potential (V max) is attributed toI Na currents and there is a controversy as to whether thyroid hormones can affect V max or not. A previous investigation has reported that V maxdid not change in hyperthyroid rabbit papillary muscle (42). Acutely administered T3 has been reported to show no effect on V max in guinea pig ventricular myocytes (10). However, Freedberg et al. (14) reported that V max was increased in hyperthyroidism and decreased in hypothyroidism at certain driving frequencies in rabbit atrial cells. Johnson et al. (22) also demonstrated the increase inV max in hyperthyroidism in rabbit atrial cells. Enhanced I Na by T3, which could result in an increased V max, would therefore increase the conduction rate and contractility. However, this effect may be more pronounced in the atrial cells than ventricular cells. The reason for this difference is not clear, it could be due to T3 penetrates the membrane of atrial cells easier than ventricular cells, and/or atrial cells responses to T3better than ventricular cells. On the other hand, a synergistic effect of G protein and T3 would significantly decrease thek OI. Relatively higher inhibitory Gα (Gαi) in the atrial cells may also responsible for the amplified T3 effect in the atrium.
Recent studies (8) have shown that slowingI Na inactivation prolongs the QT interval and aggravates adrenaline-induced arrhythmias. The incidence of ventricular arrhythmias in hyperthyroidism is much lower than supraventricular arrhythmias (44). However, hyperthyroid patients with cardiac hypertrophy and heart failure were generally not included in these studies (44) because ventricular arrhythmias in this patient population are not considered to be related to hyperthyroidism. Substantial increases in Gαi in the setting of hypertrophy, ischemic cardiomyopathy, and heart failure have been found both in animal models and in patients (11). Although pathophysiologically elevated T3 level may have only minor effect on the ion channels. The synergistic interaction between the pathophysiologically increased G protein and T3levels may have a profound impact on Na+ channel kinetics. This synergy significantly slows Na+ channel inactivation, increases fast inward I Na, and consequently increases the membrane excitability that may be the cofactor of ventricular arrhythmias in hyperthyroid patients with hypertrophy, ischemia, and/or heart failure (12).
We thank Dr. Bramah H. Singh for support, Dr. Paul J. Davis for discussions, and Di-Chong Xu for help.
This study was partially supported by grants from the American Heart Association, Greater Los Angeles Affiliate, and the Laubish Fund, University of California Regents.
Address for reprint requests and other correspondence: L. Sen, Div. of Cardiology, UCLA Medical Center, The David Geffen School of Medicine, 47-123 CHS, 10833 Le Conte Ave., Los Angeles, CA 90095-1679 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
July 26, 2002;10.1152/ajpheart.00326.2002
- Copyright © 2002 the American Physiological Society