Contrasting anesthetic sensitivities of T-type Ca2+ channels of reticular thalamic neurons and recombinant Cav3.3 channels


  • Pavle M Joksovic,

    1. Department of Anesthesiology, University of Virginia Health System, Mail Box 800710, Charlottesville, VA 22908-0710, U.S.A.
    Search for more papers by this author
    • 3

      These authors contributed equally to this work.

  • Barbara C Brimelow,

    1. Department of Anesthesiology, University of Virginia Health System, Mail Box 800710, Charlottesville, VA 22908-0710, U.S.A.
    Search for more papers by this author
    • 3

      These authors contributed equally to this work.

  • Janet Murbartián,

    1. Department of Pharmacology, University of Virginia Health System, Charlottesville, VA 22908, U.S.A.
    Search for more papers by this author
  • Edward Perez-Reyes,

    1. Department of Pharmacology, University of Virginia Health System, Charlottesville, VA 22908, U.S.A.
    Search for more papers by this author
  • Slobodan M Todorovic

    Corresponding author
    1. Department of Anesthesiology, University of Virginia Health System, Mail Box 800710, Charlottesville, VA 22908-0710, U.S.A.
    Search for more papers by this author

Department of Anesthesiology, University of Virginia Health System, Mail Box 800710, Charlottesville, VA 22908-0710, U.S.A. E-mail:


  • 1Reticular thalamocortical neurons express a slowly inactivating T-type Ca2+ current that is quite similar to that recorded from recombinant Cav3.3b (α1Ib) channels. These neurons also express abundant Cav3.3 mRNA, suggesting that it underlies the native current.
  • 2Here, we test this hypothesis by comparing the anesthetic sensitivities of recombinant Cav3.3b channels stably expressed in HEK 293 cells to native T channels in reticular thalamic neurons (nRT) from brain slices of young rats.
  • 3Barbiturates completely blocked both Cav3.3 and nRT currents, with pentobarbital being about twice more potent in blocking Cav3.3 currents. Isoflurane had about the same potency in blocking Cav3.3 and nRT currents, but enflurane, etomidate, propofol, and ethanol exhibited 2–4 fold higher potency in blocking nRT vs Cav3.3 currents.
  • 4Nitrous oxide (N2O; laughing gas) blocked completely nRT currents with IC50 of 20%, but did not significantly affect Cav3.3 currents at four-fold higher concentrations. In addition, we observed that in lower concentration, N2O reversibly increased nRT but not Cav3.3 currents.
  • 5In conclusion, contrasting anesthetic sensitivities of Cav3.3 and nRT T-type Ca2+ channels strongly suggest that different molecular structures of Ca2+ channels give rise to slowly inactivating T-type Ca2+ currents. Furthermore, effects of volatile anesthetics and ethanol on slowly inactivating T-type Ca2+ channel variants may contribute to the clinical effects of these agents.

British Journal of Pharmacology (2005) 144, 59–70. doi:10.1038/sj.bjp.0706020


dorsal root ganglion


low-voltage activated


nitrous oxide


nucleus reticularis thalami


The mechanisms by which general anesthetics produce clinical effects are not well understood, but it is likely that they involve specific effects on multiple classes of ion channels that control excitability of nerve cells. Neuronal voltage-gated Ca2+ channels play a central role in control of cellular excitability and transmitter release. While subtypes of high voltage-activated (HVA) Ca2+ channels play important roles in synaptic transmission in CNS neurons (Miller, 1998), the main proposed function of low voltage-activated (LVA) or transient (T-type) Ca2+ channels include promotion of Ca2+-dependent burst firing, generation of low-amplitude intrinsic neuronal oscillations, elevation of Ca2+ entry, and boosting of dendritic signals, which contribute to the pacemaker activity of CNS neurons, wakefulness, and seizure susceptibility (reviewed in Huguenard, 1996; Perez-Reyes, 2003).

Whereas it has been known that HVA Ca2+ currents arise from multiple forms of Ca2+ channels with distinct pharmacological properties, the extent to which T-type Ca2+ current arises from multiple Ca2+ channel subtypes became clear more recently with the cloning of three genes encoding α1 subunits of T channels, named α1G (Cav3.1; Perez-Reyes et al., 1998), α1 H (Cav3.2; Cribbs et al., 1998), and α1I (Cav3.3; Lee et al., 1999). These distinct genes encode channels with unique biophysical and pharmacological properties, and their region-specific expression contributes to the heterogeneity of T-type Ca2+ currents observed in native cells (Herrington & Lingle, 1992; Todorovic & Lingle, 1998; Perez-Reyes, 2003)

Cav3.3 is selectively expressed in brain regions such as the thalamus (only in reticular nucleus but not relay nuclei), cerebellum, and olfactory cortex (Talley et al., 1999). Unlike Cav3.1 and Cav3.2, Cav3.3 T-type Ca2+ channels have unusually slow inactivation kinetics and activate over a broader range of membrane potentials. Native and recombinant channels arising from Cav3.1 and Cav3.2 genes give rise to T currents that inactivate faster, with inactivation time constant between 20 and 30 ms, while Cav3.3 currents have up to three-fold slower inactivation (Perez-Reyes, 2003). These distinct kinetic features enable Cav3.3 channels to mediate slow and sustained influx of Ca2+ ions into neurons, which in turn increases their excitability by generating high frequency bursts (Kozlov et al., 1999; Chemin et al., 2002). It has been shown that volatile general anesthetics block native T currents in hippocampal (Study, 1994), dorsal root ganglion (Todorovic & Lingle, 1998), thalamic relay (Ries & Puil, 1999) neurons, as well as recombinant Cav3.1 and Cav3.2 (Todorovic et al., 2000) T-type Ca2+ currents, at clinically relevant concentrations. This suggests that anesthetic depression of neuronal T-type Ca2+ currents may contribute to the inhibitory effects of anesthetics in neuronal pathways. Therefore, we wished to investigate, under identical recording conditions, whether the anesthetic sensitivities of the kinetically distinct, slow Cav3.3 T-type channel variant are similar to the sensitivity of the faster inactivating T currents encoded by Cav3.1 and Cav3.2 genes. Furthermore, we describe contrasting pharmacological sensitivities to anesthetics and ethanol of slowly inactivating native T currents in thalamus and recombinant Cav3.3 currents using the most abundantly expressed ‘b’ splice variant (Murbartián et al., 2002).


Cell preparation

HEK-293 cells (A293, ATCC, Manassas, VA, U.S.A.) were transfected with the rat Cav3.3b cDNA (α1Ib; AY128644) contained in the vector pcDNA3. Stable colonies were isolated in DMEM media supplemented with 1 mg ml−1 G418 (Invitrogen, Carlsbad, CA, U.S.A.). The Ca2+ currents from three stable lines were characterized, and found to be similar to that observed in the transiently transfected cells (Murbartián et al., 2002), so line rat Ib-17 was chosen for this study. Cells were typically used 1 to 2 days after plating. Average cell capacitance (Cm) was 12.9±5.7 pF (mean±s.d.), and the average series resistance (Rs) was 5.29±2.08 (N=101).

Electrophysiological methods

Recordings were made with standard whole-cell voltage-clamp techniques. Electrodes were fabricated from microcapillary tubes (Drummond Scientific Company, Broomall, PA, U.S.A.), and fire-polished. Pipette resistances were 2–5 MΩ. Voltage commands and digitization of membrane currents were carried out with Clampex 8.2 of the pClamp software package (Axon Instruments, Foster City, CA, U.S.A.) running on an IBM-compatible computer. Membrane currents were recorded with an Axopatch 200A patch-clamp amplifier (Axon Instruments). Typically, cells were held at −90 mV and depolarized to −30 mV every 20 s to evoke inward currents. Data were analyzed using Clampfit (Axon Instruments) and Origin 7.0 (Origin Lab Corporation, Northhampton, MA, U.S.A.). Currents were filtered at 5 kHz. Reported series resistance values were taken from the reading of the amplifier. All experiments were carried out at room temperature (20–23°C). In most experiments, leak subtraction was performed online with a P/5 protocol for on-line leakage subtraction.

Analysis of current blockade

The percent reduction in peak T current at a given anesthetic concentration was used to generate concentration–response curves. For each concentration–response curve, all points are averages of multiple determinations obtained from at least five different cells. In most of the cells, we tested at least two different concentrations of the same anesthetic. On all plots, vertical bars indicate standard errors (s.e.). Mean values on concentration–response curves were fit to the following function:


where PB is the percent block of the peak T current, PBmax is the maximal block, IC50 is the concentration that produces 50% of maximal inhibition, and n is the apparent Hill coefficient for blockade. Fitted values are typically reported with 95% linear confidence limits.

Solution exchange procedures

The solution application system consisted of multiple, independently controlled glass capillary tubes driven by gravity. During an experiment, solution was removed from the end of the chamber opposite the glass capillary tubes with the use of constant suction. Switching between solutions was accomplished by manually controlled valves. Test solutions were maintained in closed, weighted all-glass syringes (to avoid saline evaporation and loss of volatile agents). Changes in Ca2+ current amplitude in response to rapidly acting drugs or ionic changes were typically complete in 10–20 s. Switching between separate perfusion syringes, each containing control saline, resulted in no changes in Ca2+ current.

Solutions and current isolation procedures

The standard extracellular saline for recording of Cav3.3. Ca2+ current contained (in mM): 160 TEA-Cl, 10 HEPES, 2 CaCl2, adjusted to pH 7.4 with TEA-OH, (316 mOsm). Internal solution consisted of (in mM): 110 Cs-methane sulfonate, 14 phosphocreatine, 10 HEPES, 9 EGTA, 5 Mg-ATP, and 0.3 Tris-GTP, pH adjusted to 7.20 with CsOH (300 mOsm).

Drugs and chemicals

Etomidate powder and isoflurane were obtained from Abbott (Abbott Park, IL, U.S.A.). All other chemicals were obtained from Sigma or Aldrich Chemicals (Milwaukee, WI, U.S.A.). Stock solutions of propofol (100 mM) and etomidate (300 mM) were prepared in dimethyl sulfoxide (DMSO) and kept at 4°C until use. DMSO (0.3%) had no effect when tested alone on Cav3.3 Ca2+ currents (n=3). All barbiturates were prepared in stock solutions in 0.1 N TEA-OH; the pH of the final extracellular solution was adjusted with HCl to 7.4. Isoflurane and enflurane solutions were prepared from saturated saline solutions (Todorovic & Lingle, 1998; Todorovic et al., 2000), and the final concentration in the bath was determined by gas chromatography for each experiment (Evers et al., 1986).

For experiments with nitrous oxide (N2O), the extracellular solution was bubbled with air or N2O/O2 mixtures using a bubbling stone (Todorovic et al., 2001). The bubbling container was sealed with Parafilm and was punctured with a small escape perforation. The solution was equilibrated with gas for at least 30 min, during which time gas-equilibrated solution was drawn into a closed glass syringe. The syringe served as a solution reservoir for the gravity-fed local perfusion system. For most experiments, 80% N2O/20% O2 was used so that bottled air 80% N2/20% O2 could be used as a control. All test solutions were prepared on the day of the experiment by diluting the stock solutions with the appropriate amount of extracellular saline.

In vitro slice tissue preparation

Most of experiments were performed with 200–250 μm thick transverse rat brain slices taken through the middle portion of the nucleus reticularis thalami (nRT) (stereotaxic levels 4.6–5.82 mm below bregma; (Paxinos & Watson, 1982). Gravid Sprague–Dawley rats were housed in a local animal facility, and used in accordance with guidelines (NIH Guide for the Care and Use of Laboratory Animals) and protocols approved by the University of Virginia Animal Use and Care Committee. Young rats, aged 7–14 days, were anesthetized with halothane and decapitated, and the brains rapidly removed and placed in chilled (4oC) cutting solution consisting of (in mM): 2 CaCl2, 260 sucrose, 26 NaHCO3, 10 glucose, 3 KCl, 1.25 NaH2PO4, 2 MgCl2, equilibrated with a mixture of 95% O2 and 5% CO2. A block of the tissue containing the thalamus was glued to the chuck of a vibrotome (TPI, St Louis, MO, U.S.A.) and 200–250 μM slices were obtained in a transverse plane. The slices were incubated in oxygenated saline at 36°C for 1 h prior to placing in a recording chamber that had been modified for superfusion at a rate of 1.5 cc/min. Incubating saline consisted of (in mM): 124 NaCl, 4 KCl, 26 NaHCO3, 1.25 NaH2PO4, 2 MgCl2, 10 glucose, 2 CaCl2, equilibrated with a mixture of 95% O2 and 5% CO2. Slices were maintained at room temperature in the recording chamber and remained viable for at least up to 1 h under these conditions.

Recording procedures

We recorded Ca2+ currents with whole-cell patch clamp technique in coronal thalamic slices from young rats from a total of 122 visually identified neurons (Rs<20 MΩ) of nRT at room temperature. Whole-cell recordings were obtained from nRT visualized with an IR DIC camera (Hammamatsu, C2400) on the Zeiss 2 FS Axioscope (Carl Zeiss, Jena). Electrodes were fabricated from microcapillary tubes with thin walls, and were not fire-polished, with final resistances of 3–6 MΩ. The standard extracellular saline for recording of T-type Ca2+ currents consisted of (in mM): 2 CaCl2, 130 NaCl, 2.5 MgCl2, 10 glucose, 26 NaHCO3, 1.25 NaH2PO4, 2 CsCl, 10 TEACl, 1 4-AP and 1 μM TTX to block voltage-gated Na+ currents. To isolate T-type from HVA Ca2+ currents, an internal solution was used that contained (in mM): 135–140 TMAOH, 10 EGTA, 40 HEPES, and 2 MgCl2, titrated to pH 7.15–7.25 with hydrofluoric acid (Todorovic & Lingle, 1998). In some experiments, we used the same internal solution as in experiments with HEK cells and this did not influence the effects of pentobarbital, ethanol, or N2O (n=3 for each experiment, data not shown). For recordings of T currents from nRT, cells were held at −100 mV and depolarized to −50 mV to avoid small contaminating component of HVA current that was resistant to intracellular F. For the data presented in this study, correction of 10 mV in voltage-clamp experiments was made for liquid junction potentials.

Since voltage control is compromised in slices due to the presence of extensive cell processes, we included in our results only cells where voltage-dependent current activation was smooth; no excessive delay in the onset of current was observed and the onset and offset kinetics depended on voltage but not on the amplitude of current. All measurements of amplitudes from holding, peak, and steady-state currents are made at time points sufficient to ensure reasonably well-clamped current conditions.

Similarly, anesthetic delivery in vitro could be compromised in slices due to decline in drug concentration along the length of the bath and diffusion through the slice tissue. Highly lipid-soluble anesthetic agents pose challenges to many standard solution application methods. We have found that propofol and etomidate are readily absorbed into silicon and some other types of tubing within minutes. To minimize this problem, glass or metal tubing is used. We have found that a system consisting of multiple pieces of glass tubing connected with small PVC joints allows reliable and consistent delivery of drugs to the bath. Furthermore, most of the recordings are carried out from neurons in the superficial 50 μm of slices to facilitate equilibration of anesthetics and slice tissue. We demonstrate on Figure 4 that prolonged applications of etomidate, propofol, and pentobarbital do not suppress more currents after first 3 min of application. This argues that our recording conditions are adequate to perform proposed experiments and that an apparent equilibrium is reached in all experiments when measurements were taken.

Figure 4.

Pentobarbital, etomidate, and propofol block slow thalamic T currents. (a–e). Time course of block of peak T current in different nRT cells in slices perfused with different concentrations of pentobarbital, etomidate, and propofol. Bars indicate time of application. Insets depicts traces of T current evoked by depolarizing to −50 from −100 mV for the given experiments where time course of block is presented. An asterisk (*) marks currents obtained while drugs were applied. Two control traces indicate currents before and after drug application. Calibration for T current in top left panel applies for all T currents. (f) Average concentration–response curves for propofol (open squares, n=17 cells), etomidate (filled triangles, n=18 cells), and pentobarbital (filled circles, n=19 cells). Solid lines indicate best fit for average data giving IC50 of 25±2, 25±4 and 264±76 μM for etomidate, propofol, and pentobarbital, respectively. Slope of the curve indicating steepness of concentration–response relationship was 1.7±0.5, 1.6±0.7, and 1.3±0.2 for propofol, pentobarbital, and etomidate, respectively.


Isolation of currents

Figure 1(a and b) depicts average traces of T currents from 13 nRT cells and 11 HEK cells stably transfected with Cav3.3 constructs. It is evident that both currents in similar recording conditions slowly inactivate with a single exponential inactivation τ of about 56±5 ms for nRT (Figure 1c) and about 78±5 ms for Cav3.3 currents (Figure 1d) (P<0.01).

Figure 1.

Both recombinant Cav3.3 and native nRT T currents have slow inactivation kinetics. (a) Traces evoked from Vh −100 mV and Vt −50 mV from 13 nRT cells are averaged. (b) Similar currents averaged from 11 HEK cells transfected with Cav3.3 constructs are evoked from Vh −90 mV to Vt −30 mV. Bars indicate calibration. (c) A single exponential fit (dark line) of inactivating portion of the nRT current from panel (a) gave an average inactivation τ of 56±5 ms. (d) A single exponential fit (dark line) of inactivating portion of the Cav3.3 current from panel (b) gave an average inactivation τ of 78±5 ms.

Effects of barbiturates on Cav3.3 current

Barbiturates have been used clinically for induction of general anesthesia, as anticonvulsant drugs, and also as neuroprotective agents. Here, we examined the effects on Cav3.3 currents of three barbiturates that we used previously in our studies with recombinant channels. Figure 2 depicts the concentration-dependent effects of pentobarbital, thiopental, and phenobarbital on recombinant Cav3.3 currents. All barbiturates blocked currents completely and reversibly, with pentobarbital and thiopental being about 4–5 times more potent than phenobarbital (Figure 2d). Figure 2b also illustrates the stability of current when recorded over a prolonged period of time. Pentobarbital depressed peak inward Ca2+ currents with an IC50 of 180±42 μM and thiopental blocked currents with IC50 of 260±52 μM, and phenobarbital blocked with IC50 of 1000±200 μM. All barbiturates depressed the peak of the current with very little effect on current activation and inactivation kinetics. The time course of current activation was assessed by measuring 10–90% rise time, and inactivation time constant (τ) by fitting the decaying portion of the current with a single exponential function before and during application anesthetic at a concentration that blocks ∼50% of the peak current. Overall, the three barbiturates did not significantly change (P>0.05) the 10–90% rise time or inactivation rate (τ) in these cells (n=5–8 cells for each agent, data not shown).

Figure 2.

Barbiturates block Cav3.3 currents. (a) T currents are reversibly blocked by three different concentrations of thiopental. Inactivation τ changed from 80 ms in control to 62 ms in the presence of 0.3 mM thiopental. (b) Time course of an experiment showing the block of peak currents by thiopental (same cells as in panel (a). Peak inward current is plotted as function of time and horizontal solid bars indicate times of application. Note fast onset and offset, as well as near complete recovery from thiopental blocking action. (c) Representative traces from cells where 1 mM of thiopental, pentobarbital, and phenobarbital are used. Note that this concentration of thiopental and pentobarbital blocked almost completely, while phenobarbital blocked only 42% of peak inward current. (d) Concentration–response curves for three barbiturates are shown, with each point being the average of 6–9 different cells. Symbols indicate different anesthetic as indicated on this figure. Vertical lines are s.e. the solid lines are best fits with the Hill equation where the steepness of the slope is described by the coefficient h. IC50 values for block of Cav3.3 current were 180±42 μM (h=1.0±0.2, n=17) for pentobarbital, 260±52 μM (h=1.64±0.5, n=9) for thiopental, and 1000±200 μM (h=1.0±0.2, n=12) for phenobarbital.

Effects of propofol and etomidate on Cav3.3 currents

Propofol (2,6 diisopropylphenol) is a new intravenous anesthetic agent that has gained widespread use. Figure 3a, c and e indicate that propofol blocked Cav3.3 currents in a concentration-dependent manner with an IC50 of 60±12 μM. Propofol caused a significant increase in current inactivation rate that was evidenced by up to three-fold decrease in inactivation τ (n=5, P<0.001).

Figure 3.

Effects of propofol and etomidate on Cav3.3 currents. (a) Experiments from an HEK cell show effects of three different concentrations of propofol on Cav3.3 currents. Propofol increased the apparent inactivation of current by decreasing inactivation τ (from 65 ms in control saline, to 27 ms with application of 30 μM propofol). Note that propofol blocked current nearly completely when higher concentrations are used. (b) This panel depicts representative traces showing reversible block of Cav3.3 currents with 0.1 and 0.3 mM etomidate. Similar to propofol, etomidate increased inactivation of the current by decreasing inactivation τ by about 50% (from 78 ms in control saline to 40 ms with anesthetic application). (c) The time course of block by randomly applied multiple concentrations of propofol is plotted for the same cell as in (a). Bars indicate time of application. (d) Time course of etomidate-induced Cav3.3 current blockade is depicted (same cell shown in panel b). Horizontal bars indicate time of application of 0.1 and 0.3 mM etomidate. (e) The concentration dependence for block of Cav3.3 currents by propofol (open circles, n=13 cells, IC50 60 ±12 μM, h=1.5±0.4). (f) Concentration–response curve for etomidate block of Cav3.3 currents from total of 13 cells. Each symbol represents the average from at least five different cells. The solid line represents the fit to the average data: IC50 of 90±14 μM and h=1.54±0.34.

Etomidate is an intravenous anesthetic that blocks native T channels in DRG neurons (Todorovic & Lingle, 1998) and recombinant Cav3.1 and Cav3.2 currents in HEK cells (Todorovic et al., 2000) without affecting their apparent rate of current activation and inactivation. In contrast, etomidate blocked Cav3.3 currents at lower concentrations, and caused a significant increase in the rate of current decay (Figure 3b). Etomidate (100 μM) decreased the apparent inactivation τ by 56±7% (n=7 cells, P<0.05), with change in the 10–90% rise time of only 12±5% (P>0.05). Block of the peak current was reversible and concentration-dependent with an IC50 of 90±11 μM (Figure 3b, d and f).

Effects of etomidate, propofol, and pentobarbital on slowly inactivating nRT currents

We tested the hypothesis that the slowly inactivating T-type currents from nRT are carried by Cav3.3 channels by comparing their sensitivity to intravenous anesthetics. Figure 4 summarizes our findings from these experiments. At lower concentration, all anesthetics in most of the cells partially blocked peak current evoked from −100 to −50 mV (Figure 4a, c and e). However, at higher concentrations of 0.1–1 mM, all agents induced a complete and reversible block of nRT currents (Figures 4b and d). Pentobarbital as a representative barbiturate was about two times less potent in blocking nRT than Cav3.3 currents with an IC50 of about 264 μM, while etomidate and propofol were 2–4 times more potent in blocking nRT than Cav3.3 currents with IC50s of about 25 μM. (Figure 4f). In contrast to their effects on Cav3.3 currents, neither of the tested anesthetics significantly affected inactivation kinetics. For example, average control inactivation τ's in these experiments were 53±5 ms (n=17) and 54±4 ms (n=6, P>0.05), 63±14 ms (n=6, P>0.05), and 68±8 ms (n=5, P>0.05) for pentobarbital, etomidate, and propofol, respectively. Overall, our data indicate that different classes of intravenous anesthetics block native nRT currents differently than recombinant channels.

Differential sensitivities of ethanol on Cav3.3 and nRT T-type Ca2+ currents

Similar to many general anesthetics, ethanol depress neuronal activity by enhancing inhibitory GABAA currents or blocking excitatory NMDA currents (Catlin et al., 1999). However, recent reports indicate that ethanol modulates HVA voltage-gated Ca2+ channels in the same concentration range found to affect ligand-gated currents (Walter & Messing, 1999), Since effects of ethanol on T-type Ca2+ channels are not well characterized, we examined effects of ethanol on slowly inactivating T-type Ca2+ channel variants. We found that nRT currents were completely blocked with 50–200 mM ethanol (Figure 5a and b) and Cav3.3 currents were relatively resistant to these concentrations of ethanol (Figure 5c and d). At 200 mM, ethanol blocked only about 26% of Cav3.3 currents, while this concentration blocked slow nRT T currents reversibly and almost completely (84%) (Figure 5a–d). Concentration–response relationship for effects of ethanol upon nRT T currents gave an IC50 of about 100 mM and Cav3.3 currents were about four times less sensitive (IC50 365 mM) (Figure 5e). Ethanol blocked completely and reversibly both recombinant and native T currents but did not change inactivation τ of either Cav3.3 or nRT currents.

Figure 5.

Ethanol blocks Cav3.3 and nRT T currents. (a) Traces show nRT currents activated at −50 mV in the presence and absence of 2 concentrations of ethanol. (b) Time course from the same experiment showing reversibility of ethanol's effects on nRT currents. (c) Traces showing very little block of Cav3.3 currents by 100 and 200 mM ethanol. Bars indicate the calibration. (d) Percent block of peak nRT current is plotted as a function of time. Bars indicate the time of application. Note that 200 mM ethanol blocked only about 37% of Cav3.3 current while almost completely blocked nRT current (panel b). (e) Comparison of sensitivity to ethanol of Cav3.3 (open circles) and nRT currents (solid squares). All symbols are averaged from multiple determinations (n=9 cells for Cav3.3 and n=25 cells for nRT). Solid lines represent fit to the average data: Cav3.3, IC50 of 365±32 mM, h=2.2±0.4; and for nRT, IC50 of 100±1 mM and h=2.4±0.1.

Effects of volatile anesthetics isoflurane, enflurane, and N2O on Cav3.3 currents

The frequently used potent volatile anesthetic isoflurane blocks faster-inactivating (Cav3.1 and Cav3.2) T-type Ca2+ channel variants at subanesthetic concentrations (Todorovic et al., 2000), but the sensitivity of slowly inactivating variants like Cav3.3 is not known. Therefore, we examined effects of isoflurane and its isomer enflurane on Cav3.3 currents in HEK cells. We found that both volatile anesthetics blocked Cav3.3 currents completely and reversibly in a concentration-dependent manner (Figure 6a–d). Both anesthetics had a similar potency with IC50s of 260±23 and 360±21 μM, for isoflurane and enflurane, respectively. Similar to etomidate and propofol, the two volatile anesthetics had very little effect on the time course of current activation, but significantly increased current inactivation rate more than two-fold. Isoflurane decreased inactivation τ by 66±6% (n=6, P<0.01), and enflurane decreased inactivation τ by 62±8% (n=5, P<0.001).

Figure 6.

Isoflurane and enflurane block Cav3.3 currents in subanesthetic concentrations. (a) Traces show Cav3.3 currents activated at −30 mV in the presence and absence of isoflurane. Similar to propofol and etomidate, isoflurane increased the apparent inactivation rate (inactivation τ was 87 ms in control conditions, and 32 ms after isoflurane was applied). (b) Traces from another experiment where isoflurane's isomer, enflurane was applied. Enflurane (520 μM) blocked 75% of the peak current and decreased the inactivation τ about three-fold (from 72 to 23 ms). (c) Time course of the effects of four different concentrations of isoflurane on peak Cav3.3 current is plotted. Horizontal bars indicate times of application. Note fast onset and offset of isoflurane's effects on the peak inward current. (d) Percent block of peak Cav3.3 current is plotted as a function of isoflurane (open circles) and enflurane (solid circles) concentrations (n=8 cells for enflurane and n=9 cells for isoflurane). Solid lines represent fit to the average data: isoflurane, IC50 of 260±23 μM, h=1.7±0.3; and for enflurane, IC50 of 360±21 μM and h=1.7±0.2. (e) Traces show minimal effect (about 7% block) of 80% N2O on peak Cav3.3 current. In contrast, 600 μM isoflurane blocked about 70% of the peak current. (f) Time course of the experiment depicted on panel (e) of this figure is presented. Horizontal bars indicate times of application of 80% N2O and 600 μM isoflurane.

N2O (laughing gas) has been used for almost two centuries in medical and dental practice because of effective analgesic properties that are achieved at subanesthetic concentrations. While cellular targets for N2O were not known for a long time due to difficulties in working with this agent in vitro, it was recently reported that, unlike most of the other general anesthetics that augment GABA currents, N2O blocks NMDA currents in CNS neurons (Jevtovic-Todorovic et al., 1998). N2O blocks potently native DRG and Cav3.2 T-type Ca2+ currents but not Cav3.1 currents in HEK cells (Todorovic et al., 2001). Here, we found that 80% N2O (maximal concentration that could be used in in vitro experiments) had small blocking effect on peak Cav3.3 currents (8±2%, n=6). However, this was not significantly different from effects of air (6±2% block, n=4, data not shown) which was used as a control (P>0.05). Cells that were not sensitive to N2O had the usual sensitivity to isoflurane (Figure 6e and f).

Effects of fluorinated volatile anesthetics and N2O on nRT T currents

We reported that volatile anesthetics isoflurane and enflurane block nRT currents completely with IC50 of 281 and 148 μM, respectively (Joksovic et al., 2004).

We found that at low concentrations of 8%, N2O consistently and reversibly increased peak T current by about 30±9% in six out of eight tested cells (Figure 7a and b). The increase in T current by N2O was not accompanied in changes in apparent current kinetics or holding current. In three cells, 8% N2O increased current while 80% N2O blocked peak T current in the same cells (data not shown). This indicates that an increase in peak T current by low concentrations of N2O does not reflect action on a distinct subpopulation of thalamic cells but strongly suggests that it may be a dual mechanism of T channel modulation by N2O in nRT neurons.

Figure 7.

Effects of N2O on T currents in nRT cells. (a) Representative traces showing effects 8% N2O on an nRT cell. (b) Time course from the same experiments depicted on panel (a) indicate reversible increase of about 50% of peak current. (c) At concentrations of 80% N2O almost completely blocked nRT current as indicated on this cell. (d) Time course from the same experiment depicted on panel (c) indicate fast onset and offset of block by N2O. (e) Filled circles indicate average block of nRT currents by 20, 40, and 80% N2O (n=25 cells). Solid line is best fit of the data giving IC50 of 20.0±0.1%, fitted maximal block of 85±1% and h of 2.4±0.1. Open symbol indicates average block of Cav3.3 currents by 80% N2O.

N2O at higher concentrations (20–80%) reversibly inhibited nRT currents with a maximal current block of 82±4% at 80% N2O and an apparent IC50 of about 20% N2O (Figure 7e, n=9). Unlike isoflurane and enflurane (Joksovic et al., 2004), this blocking effect was not accompanied by significant change in inactivation kinetics. Average inactivation τ was 57±7 ms in control conditions and 83±11 ms in the presence of 40% N2O (P>0.05, n=9). This suggests that N2O and fluorinated general anesthetics block slow T current by different mechanisms. In control experiments, air did not significantly affect peak T current (n=6 cells, data not shown). Therefore, these data indicate that T currents in nRT are more sensitive to N2O than any other T-current variant previously tested.

Comparison of anesthetic sensitivities of different T-type Ca2+ channel variants

Table 1 describes the similarity and difference of the blocking effects of tested anesthetics on Cav3.3 as well as Cav3.1 and Cav3.2 currents in HEK cells, under identical recording conditions (Todorovic et al., 2000). It appears that all three cloned T channels, despite their different kinetic properties, have similar sensitivities to the most tested anesthetics. The exception is that Cav3.3 currents are resistant to N2O like Cav3.1 currents, more sensitive to etomidate and pentobarbital, and somewhat less sensitive to propofol. In contrast, nRT T current were less sensitive to pentobarbital, but 2–4-fold more sensitive to etomidate, propofol, and N2O. When compared to Cav2.3 currents in HEK cells (Nakashima et al., 1998) or oocytes (Kamatchi et al., 1999), it is evident that T-type Ca2+ channels variants are more sensitive to anesthetics, particularly volatile anesthetics like isoflurane (e.g. four-fold difference in sensitivity). It must be noted that in our studies, for most anesthetics, we did not test for voltage- or use-dependence of block. Substantial voltage- or use-dependence of block can affect estimated IC50 values. However, there was no indication of slow block by any of the anesthetics we examined.

Table 1.  Comparison of sensitivities of native nRT T currents and recombinant T-type currents
 1G)a,b1H)a,b1E)c 1I)
  • a

    Data from Todorovic et al. (2000), using 2 mM Ca2+ as charge carrier.

  • b

    Data for N2O are from Todorovic et al. (2001), using 10 mM Ba2+ as a charge carrier.

  • c

    Data from Nakashima et al. (1998) using HEK-293 cells, except isoflurane data which is an estimate of IC50 from Kamatchi et al. (1999) using Xenopus oocytes. Both studies used 10 mM Ba2+ as charge carrier.

  • d

    This study used 2 mM Ca2+, data for isoflurane and enflurane from Joksovic et al. (2004).

AnestheticIC50±s.e.m., maximal block
PentobarbitalIC50 310±40 μMIC50 345±94 μMIC50 600 μMIC50 264±76 μMIC50 180±42 μM
 max. 100%max. 100%max. 100%max. 100%max. 100%
ThiopentalIC50 280±40 μM   IC50 260±52 μM
 max. 100%   max. 100%
PhenobarbitalIC50 1.5±0.2 μM IC50 2.7 μM IC50 1.0±0.2 mM
 max. 100% max. 100% max. 100%
PropofolIC50 20±2.0 μMIC50 27±3 μM IC50 25±4 μMIC50 60±12 μM
 max. 100%max. 100% max. 100%max. 100%
EtomidateIC50 160±46 μM  IC50 25±2 μMIC50 90±14 μM
 max. 100%  max. 100%max. 100%
IsofluraneIC50 277±24 μM IC50 1 mMIC50 281±40 μMIC50 260±23 μM
 max. 100% max. 100%max. 100%max. 100%
Nitrous oxideNo effect up to 80%IC50 58±17 μM IC50 20.0±0.1%No effect up to 80%
  max. 66% max. 85% 
Enflurane   IC50 148±10 μMIC50 360±21 μM
    max. 100%max. 100%


It was recognized as early as two decades ago that volatile anesthetics block certain types of (e.g. L-type) Ca2+ currents in myocardial tissue (Lynch et al., 1981; Terrar & Victory, 1988; Bosnjak et al., 1991). However, multiple classes of Ca2+ channels exist in nerve cells and our more recent work (Todorovic & Lingle, 1998; Todorovic et al., 2000, 2001) and the work of others (Herrington & Lingle, 1991; Takenoshita & Steinbach, 1991; Study, 1994; McDowell et al., 1996) indicates that T-type voltage-gated Ca2+ channels in both central and peripheral neurons are inhibited by volatile general anesthetics at concentrations which occur under clinical conditions. However, previous studies did not investigate anesthetic sensitivities of kinetically distinct slow T-type Ca2+ channel variants. Here, we have explored the sensitivity of the recombinant rat Cav3.3b T-channel isoform and native nRT T currents to various classes of anesthetics and compared with sensitivities of other T-channel variants (Cav3.1 and Cav3.2), as well as one class of channels that contributes to native R-type HVA currents (Cav2.3).

Anesthetic sensitivities of various LVA currents

We find that, in general, sensitivities of most classes of anesthetics among different variants of recombinant T channels do not vary greatly (i.e. less than two-fold). Exceptions are N2O, propofol, and etomidate. Propofol, is about 2–4-fold less potent in blocking Cav3.3 current than Cav3.1 and Cav3.2 currents (Table 1). On the contrary, etomidate is about two-fold more potent in blocking Cav3.3 currents than other recombinant T-channel isoforms. N2O blocks potently recombinant Cav3.2 and related native DRG at subanesthetic concentrations, while higher concentrations are required to block Cav3.1 (Todorovic et al., 2001) and Cav3.3 currents (this study). This indicates that N2O is a valuable tool as a selective blocker for Cav3.2-based currents in recombinant and native systems.

We also compared the sensitivities of slow and fast T-channel variants to the sensitivities of Cav2.3 isoform, which in native cells gives rise to R-type HVA currents (Piedras-Renteria & Tsien, 1998). However, in certain instances, Cav2.3- and R-based currents may resemble T currents because of somewhat similar kinetic features, similar permeation properties, and similar sensitivity to nickel (Randall & Tsien, 1997; Nakashima et al., 1998). We found that Cav3.3, as well as other T-channel variants are in most instances at least two-fold more sensitive to anesthetic blockade than Cav2.3 currents (Table 1). Therefore, our data suggest that anesthetics, particularly volatile anesthetics like isoflurane, may be used to differentiate native currents arising from Cav3.3 and Cav2.3 genes.

An interesting feature of anesthetic-induced blockade of Cav3.3 currents is that two mechanisms of channel blockade may occur. Volatile anesthetics like isoflurane and enflurane share the ability to increase the rate of inactivation of both recombinant and native T currents. On the contrary, neither of the barbiturates, N2O, nor ethanol significantly affected current kinetics of Cav3.3 or nRT T channels. Similarly, we noticed this effect on current inactivation for isoflurane in our experiments with Cav3.1 and Cav3.2 recombinant channels (Todorovic et al., 2000). Halothane produces a similar increase in inactivation rate of T currents in native GH3 cells (Herrington & Lingle, 1991, 1992). This suggests that inhibition of both native and recombinant T currents can occur by either of the two mechanisms, either one of which can produce complete inhibition of T current.

We noticed that N2O at lower concentrations enhances nRT currents. Similar dual mechanism of modulation of T currents by ethanol was recently reported in thalmocortical neurons in slice preparation (Mu et al., 2003). Even though the mechanism of this dual modulation is not clear, this suggests that thalamic neurons may be uniquely regulated by general anesthetics and ethanol.

Contribution of T channels to clinical effects of volatile anesthetics

The block of recombinant Cav3.3 and nRT currents by volatile anesthetics at concentrations believed to exist in brain tissue during anesthesia states suggests that this block may be clinically relevant. For volatile anesthetics in most mammals, clinically relevant concentrations are assumed to be similar to aqueous concentrations of anesthetic in brain tissue at inspired partial pressures (minimal alveolar concentration – MAC) that prevent purposeful movement to painful stimuli, such as surgical incision in 50% of subjects. For isoflurane, 1 MAC for in vitro experiments at room temperature is 320 μM, and for enflurane it is 620 μM (Franks & Lieb, 1994). In our studies, IC50s for inhibition of Cav3.3 currents are 260 and 360 μM for isoflurane and enflurane, respectively. Furthermore, enflurane blocks nRT T current even more potently with an IC50 of 148 μM (Joksovic et al., 2004). Therefore, it appears that suppression of both Cav3.3 and nRT, as well as other T-channel isoforms (Table 1), occurs at subanesthetic concentrations of fluorinated ether anesthetics.

Native T-channel isoforms with properties similar to recombinant Cav3.3 currents have not been characterized pharmacologically, although slow-inactivating T currents have been recorded in various brain regions such as nRT (Huguenard & Prince, 1992), lateral habenula (Huguenard et al., 1993), and inferior olive (Llinás & Mühlethaler, 1988), where these channels are thought to underlie long-lasting subthreshold burst firing. Furthermore, in situ hybridization studies indicate that these brain regions have high expression of Cav3.3 mRNA (Talley et al., 1999). This suggests that suppression of Cav3.3-based currents by volatile anesthetics in CNS neurons would occur during clinical anesthesia. However, our comparison of anesthetic sensitivities of Cav3.3 and nRT T currents reveals that substantial differences exist in pharmacological sensitivities to anesthetics between these kinetically similar T-current variants. Particularly, paramount sensitivity of nRT T currents to N2O raises the possibility that these currents may arise from Cav3.2 gene. This is supported by findings that mRNA for both Cav3.2 and Cav3.3 is abundant in nRT neurons (Talley et al., 1999). Since expression of native T channels may be dependent on post-translational modifications and tissue-specific factors, more work is necessary to elucidate molecular structure of slowly inactivating T currents in nRT neurons. Since Cav3.3 channel may undergo post-translation modification giving rise to multiple splice variants in the CNS (Mittman et al., 1999; Murbartián et al., 2002), it is important to study in future experiments the effects of anesthetics on native cells where these splice variants are expressed.

Inhibition of Cav3.3 and nRT currents by parenteral anesthetics and ethanol

We tested a variety of parenteral anesthetics for their ability to suppress recombinant Cav3.3 and native nRT currents. It is more difficult to determine clinically relevant concentrations of parenteral anesthetics for in vitro experiments. This is mainly because these anesthetics are used for induction of the anesthetic state due to their rapid redistribution from CNS tissue to body lipids. It has been estimated that clinically relevant anesthetic concentrations for propofol and etomidate are in the low μM range, and 25, 50, and 100 μM for thiopental, pentobarbital, and phenobarbital, respectively (Franks & Lieb, 1994). It appears from our data that inhibition of Cav3.3 currents, as well as nRT currents, may occur during clinical use of volatile anesthetics but is less likely that this would occur during general anesthesia with parenteral anesthetics. These differential sensitivities of various T currents to general anesthetics may contribute to different clinical properties. For example, volatile anesthetics that are more potent in blocking T-type Ca2+ currents are also much better analgesics and more profoundly suppress sensory information processing measured with sensory evoked potentials, than parenteral anesthetics like propofol and etomidate (Barash et al., 1992). At least part of these effects of volatile anesthetics may be mediated by depression of T currents in peripheral sensory and central neuronal pathways.

On the other hand, clinically relevant concentrations of ethanol are 50–100 mM (Catlin et al., 1999; Walter & Messing, 1999). We found that Cav3.3 currents are relatively resistant to up to 100 mM ethanol, but ethanol blocked completely native nRT T currents with an IC50 of 100 mM. This indicates that blockade of nRT but not Cav3.3 currents may occur under clinically relevant blood levels of ethanol and reveals another important pharmacological difference among nRT and Cav3.3 currents.

In conclusion, we have examined the sensitivity of Cav3.3 and native nRT T currents to various classes of general anesthetics and ethanol. Our results indicate that in spite of similar kinetic properties, anesthetic sensitivity of these two T-channel isoforms is greatly different for most of the tested agents. Furthermore, the effects of volatile general anesthetics and ethanol on these currents may contribute to their clinical and behavioral effects.


We thank Jennifer DeLisle for technical assistance. This work was supported by NIDA KO8 award DA 00428 to SMT, and NS38691 to EPR.