INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There is already persuasive evidence that neurosteroids work at specific allosteric modulatory sites on GABA_A receptor/Cl channel complexes. In particular, the anesthetic alphaxalone (5-pregnane-3-ol-11,20-dione), a synthetic steroid whose structure is related closely to some naturally occurring pregnane steroids, modulates and directly activates GABA_A receptors in a number of neuronal preparations (Lambert et al. 1995; Majewska 1992; Olsen and Sapp 1995; Paul and Purdy 1992). The present study stands in sharp contrast: we have found that in cultured embryonic human dorsal root ganglion (DRG) neurons, alphaxalone activates a Cl conductance independent of the GABA_A receptor.

In previous studies, low concentrations of both alphaxalone and pregnane steroids potently enhanced GABA_A-receptor-mediated membrane responses (Barker et al. 1987; Callachan et al. 1987; Cottrell et al. 1987; Gee et al. 1988; Harrison and Simmonds 1984) by a process involving prolongation of the mean open time of the associated GABA-activated Cl channels (Barker et al. 1987; Mistry and Cottrell 1990; Twyman and Macdonald 1992). In the absence of GABA, higher concentrations of alphaxalone and pregnane steroids are reported to directly open the Cl channels of GABA_A receptors (Barker et al. 1987; Callachan et al. 1987; Cottrell et al. 1987; Peters et al. 1988). The basis of this neurosteroid interaction with GABA_A receptors is unknown, although it is believed that both modulating and activating effects of neurosteroids are mediated by sites that are distinct from the binding sites on the GABA_A receptor/Cl channel complex for GABA, barbiturates, benzodiazepines, and picrotoxin (Cottrell et al. 1987; Gee et al. 1988).

The present studies focus on how alphaxalone affects cultured, embryonic human DRG neurons, the same neurons described in the preceding paper (Valeyev et al. 1999). In contrast to findings at other, nonhuman GABA_A receptors, we found that micromolar concentrations of alphaxalone did not enhance GABA_A-receptor-mediated membrane responses but did directly activate a Cl current. Moreover the alphaxalone-induced Cl current had biophysical and pharmacological properties that differed from those of GABA-activated Cl currents in the same neurons. The data suggest that in cultured, embryonic human DRG neurons, alphaxalone does not bind to the previously described allosteric modulatory site for neurosteroids located on the GABA_A-receptor Cl channel complex. A preliminary account of this work has appeared (Valeyev et al. 1996).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The procedures and techniques used are identical to those described in the preceding paper (Valeyev et al. 1999).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In contrast to effects at GABA_A receptors on other types of neurons, alphaxalone did not potentiate the inward currents produced by GABA under whole cell, voltage-clamp conditions (not illustrated). However, when voltage-clamped at negative membrane potentials, embryonic human DRG neurons responded with inward currents to micromolar applications of alphaxalone (concentrations ranging from 1.0 to 5.0 µM; Fig. 1A) in the absence of exogenous GABA in the medium. This occurred in essentially all neurons. As seen in Fig. 1A, the alphaxalone-induced currents were dependent on the concentration of alphaxalone applied. The Hill coefficient was 1.3 ± 0.15 (mean ± SE; Fig. 1B). When the intrapipette [Cl] was 134 mM, the mean peak current activated by alphaxalone (1.0 µM) measured 28.0 ± 5.5 pA (n = 12) (holding potential, V_h, of 60 mV). The currents required <1.0 s to reach peak amplitudes after exposure to alphaxalone, did not attenuate during prolonged applications (Fig. 4A), and returned quickly to baseline after removal. Activation of the Cl current by alphaxalone (10 µM) was not prevented by internal perfusion of the neurons with EGTA concentrations of 10.0 mM (n = 6).

View larger version (13K): [in this window] [in a new window]   Fig. 1. Alphaxalone concentration-responses in cultured embryonic human dorsal root ganglion neurons. A: whole cell current traces produced by applications of increasing concentrations of alphaxalone in the absence of GABA in the medium. B: alphaxalone concentration-response relationship. Alphaxalone responses are plotted as means ± SE of 3-5 independent determinations of the peak current normalized to the peak current induced by 1.0 µM alphaxalone.

Currents directly activated by alphaxalone are carried by Cl ions

Current-voltage (I-V) curves show that when Cl was the main intracellular anion ([Cl]_o/[Cl]_i = 151 mM/140 mM), the reversal potential for the alphaxalone-induced response was close to 0 mV (4.4 ± 2.1 mV, n = 6). Moreover, as can be seen in Fig. 2, alphaxalone-induced currents displayed strong inward rectification at positive holding potentials (i.e., outward currents were smaller than inward currents at equivalent potentials). In addition, reduction of [Cl]_i (to 90 mM by isotonic replacement of CsCl by CsF) shifted the reversal potential to 33.3 ± 5.6 mV (n = 6), a value that conformed to the expected change in the theoretical Nernstian Cl equilibrium potential for a Cl-current (theoretical value: 30.0 mV).

View larger version (19K): [in this window] [in a new window]   Fig. 2. Whole cell current-voltage (I-V) relationship for alphaxalone-activated Cl current. Peak amplitudes of currents induced by alphaxalone (2.5 µM) applied by 1- to 5-s pressure pulses plotted against membrane potential for 2 levels of [Cl]i (n = 6 neurons for each Cl level). Inward rectification was seen at positive membrane potentials. Brackets indicate SE of mean.

Alphaxalone-induced currents are not blocked by GABA_A receptor antagonists

If the Cl currents activated by alphaxalone are mediated by GABA_A receptors, these currents should be blocked by antagonist compounds that bind to various sites on the GABA_A receptor/channel complex (Barker et al. 1987; Cottrell et al. 1987; Ong et al. 1988). However, as seen in Fig. 3, alphaxalone-induced whole cell currents were found to be insensitive both to the specific, competitive GABA_A receptor antagonist bicuculline (100 µM, n = 6), and to the channel blockers picrotoxin (100 µM, n = 6) and t-butylbicyclophosphorothionate (TBPS, 100 µM, n = 6).

View larger version (15K): [in this window] [in a new window]   Fig. 3. Alphaxalone-activated currents are unaffected by GABAA antagonists. Downward deflections in whole cell patch-clamp recordings represent inward Cl currents. Left: currents activated by alphaxalone (2.5 µM) applied by pressure injection. Right: currents activated by alphaxalone in the presence of bicuculline methochloride (BCC, 100 µM), picrotoxin (PTX, 100 µM), and t-butylbicyclophosphorothionate (TBPS, 100 µM). All neurons were voltage clamped at 60 mV. Applications of alphaxalone represented by bars above current traces.

Fluctuation analysis of alphaxalone-evoked current noise

Current fluctuation analysis was carried out on neurons voltage-clamped at 60 mV. Alphaxalone applied in a concentration of 1.0 µM evoked nondesensitizing responses lasting 60-90 s (Fig. 4A). Between three and seven spectra were obtained from each neuron studied. Figure 4B shows an example of the power density spectrum of alphaxalone-induced Cl current fluctuations. The spectrum in all neurons was well described by the sum of two Lorentzian components with slow (₁) and fast (₂) time constants. The mean open times ₁ and ₂ were 67.2 ± 11.7 and 3.2 ± 2.1 ms (n = 12), and the elementary conductance determined by the relationship between mean amplitude of whole cell current and variance was 30 ± 3.4 pS (n = 12).

View larger version (21K): [in this window] [in a new window]   Fig. 4. Fluctuation analysis of alphaxalone-activated Cl current under whole cell clamp. A, top: low-gain DC record; bottom: high-gain AC record of currents evoked by alphaxalone (1.0 µM) in a neuron voltage-clamped at 60 mV. Application of alphaxalone represented by bar above current traces. B: spectral density plot of current fluctuations during alphaxalone application. Difference spectrum fit by a double Lorentzian function. , corner frequencies ( fc) of 20.2 and 34.6 Hz corresponding to estimated apparent mean open times for alphaxalone-activated Cl channels of 78.8 and 4.6 ms.

Single alphaxalone-activated channel currents in cell-attached patches

On-cell recordings from 21 intact neurons were used to evaluate directly the elementary properties of alphaxalone-activated Cl channels. Without alphaxalone in the intrapipette solution, no spontaneous channel activity was recorded. Single-channel currents, seen in all patches when pipette solutions contained alphaxalone (10 µM), were composed of small-amplitude, brief-duration openings that usually occurred as single events or as brief bursts of openings and closings (see Fig. 5 for sample records). By measuring their amplitudes, it was possible to produce histograms for which the distribution of amplitudes was best fitted with one Gaussian function (Fig. 6). These amplitude distributions revealed no indication of subconductance states. Alphaxalone-induced single channel currents demonstrated a mean current amplitude of 0.32 ± 0.04 pA (V_p = 57 mV). Figure 7 illustrates that the amplitude of single-channel currents varied linearly with the patch potential (V_p). The calculated chord conductance derived from single-channel I-V relationships was 29.0 ± 3.1 pS.

View larger version (19K): [in this window] [in a new window]   Fig. 5. Single-channel currents evoked by alphaxalone. Representative single-channel currents recorded in cell-attached configuration at different patch potentials. Alphaxalone (10 µM) was present in the pipette solution. Patch potential (Vp) indicated for each trace. Upward deflections show outward current corresponding to inward flow of Cl ions.

View larger version (12K): [in this window] [in a new window]   Fig. 6. Alphaxalone-activated single-channel current histogram. Channel currents were recorded in a cell-attached patch with alphaxalone (10 µM) in the recording pipette. Distribution of amplitudes was best fit by a single Gaussian function with a peak amplitude of 0.32 ± 0.04 pA. Vp was 57 mV.

View larger version (16K): [in this window] [in a new window]   Fig. 7. I-V relationship of alphaxalone-activated single channel currents. Single-channel currents induced by alphaxalone (10 µM) in cell-attached patches from 4 neurons plotted against Vp. Each point represents the mean of a Gaussian distribution of channel amplitudes.

To evaluate the kinetic properties of the single-channel conductance state activated by 10 µM alphaxalone in on-cell patches, both open and closed dwell-time distributions were determined. Figure 8A shows that the distribution of open-channel events for a patch held at V_p = +40 mV was best fitted with the sum of two exponential components with mean time constants of 2.7 ± 1.9 and 65.3 ± 23.5 ms. In open-time histograms from six patches, a best fit also was obtained with double exponential functions. These data suggest that alphaxalone-activated channels have at least two open states. Similarly, kinetic analysis of the closed-time distributions for a patch held at V_p = +40 showed two exponential components with mean time constants of 17.9 ± 9.5 and 147.2 ± 26.8 ms (Fig. 8B). The best fits of the distributions of closed times from six patches was obtained with two exponential functions with mean time constants of 17.9 ± 9.5 and 147.2 ± 26.8 ms.

View larger version (21K): [in this window] [in a new window]   Fig. 8. Kinetics of alphaxalone-gated channels. Data refer to single-channel on-cell recordings obtained with alphaxalone (10 µM) in the recording pipette. Vp = +40 mV. A: open-time distribution for single channel currents best fit by the sum of 2 exponentials having decay time constants of 2.7 ± 1.9 and 65.3 ± 23.5 ms. B: closed-time distribution best fit by the sum of 2 exponentials having time constants of 17.9 ± 9.5 and 147.2 ± 26.8 ms.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Alphaxalone is reported to activate a Cl conductance via the GABA_A receptor/Cl channel complex in a number of neuron types (Lambert et al. 1995; Majewska 1992; Olsen and Sapp 1995; Paul and Purdy 1992). Of interest in the present results are the unique findings in cultured embryonic human DRG neurons: namely that micromolar concentrations of alphaxalone directly activate a Cl current with electrophysiological and pharmacological properties distinct from the GABA-activated Cl current in the same neurons. In other words, the actions of alphaxalone in these experiments do not appear to be mediated either by allosteric modulation of the GABA_A receptor complex or by direct allosteric activation of the Cl ion conductance associated with GABA_A receptors.

This view is supported by several different lines of evidence. First of all, micromolar concentrations of alphaxalone did not potentiate Cl currents produced by GABA. Nor did the compounds direct actions require the presence of GABA. We found alphaxalone to directly activate Cl channels in cell-attached patches in the absence of GABA in the recording pipette. Alphaxalone also activated whole cell Cl currents when exogenous GABA was not added to the medium.

The reversal potential for the alphaxalone-induced response was close to the theoretical Nernstian Cl equilibrium potential calculated from the given extra- and intracellular Cl concentrations both when Cl was the main intracellular anion and when [Cl]_i was reduced. And because nonspecific cation currents demonstrate negligible anion permeability, it is doubtful that the alphaxalone-activated current represented a nonspecific cation current (Colquhoun et al. 1981; Yellen 1982). Similarly, the finding that alphaxalone-induced currents were not affected by the presence of EGTA in the internal solution presumably indicates that a Ca²⁺-dependent Cl conductance cannot account for the Cl current investigated here (Korn and Weight 1987; Mayer 1985).

Patterns of rectification set the alphaxalone-induced Cl current apart from the GABA-induced Cl current in embryonic human DRG neurons and in nonhuman neurons. In various preparations, both a lack of rectification and the presence of outward rectificationbut never inward rectificationof GABA-generated whole cell Cl currents have been reported (Allen and Albuquerque 1987; Curmi et al. 1993; Fatima-Shad and Barry 1992; Gray and Johnston 1985; Hamill et al. 1983; Smith et al. 1989; Valeyev et al. 1999; Weiss et al. 1988). In sharp contrast, alphaxalone-induced whole cell currents displayed strong inward rectification at positive holding potentials. In our cultured embryonic human DRG neurons, GABA-activated whole cell Cl currents demonstrated outward rectification (Valeyev et al. 1999). It should be noted that the alphaxalone-induced current also differs from the inwardly rectifying Cl conductance described in frog oocytes, Aplysia neurons, and rat hippocampal neurons (Chesnoy-Marchais 1983; Mager et al. 1995; Parker and Miledi 1988; Staley 1994). The latter current is voltage dependent, whereas the amplitude of single-channel currents activated by alphaxalone varies linearly with the patch potential. With regard to their single-channel I-V relationships, however, results from both alphaxalone- and GABA-activated single-channel studies were similar to one another, showing a lack of rectification (Valeyev et al. 1999). Such a linear I-V relationship favors a voltage-dependent change in channel gating as the mechanism underlying the display of rectification in whole cell recordings.

Nearly all reports indicate that subconductance states are generated by GABA at GABA_A receptors (Bormann and Clapham 1985; Hamill et al. 1983; Macdonald et al. 1989; Mistry and Hablitz 1990; Ozawa and Yuzaki 1984; Taleb et al. 1987; Weiss et al. 1988). Likewise, in cultured embryonic human DRG neurons, GABA activates a subconductance state (Valeyev et al. 1999). Significantly, recordings of single-channel activity produced by exposure to alphaxalone failed to reveal a subconductance state. And finally, alphaxalone-induced Cl currents were not blocked by bicuculline, picrotoxin, or TBPS when these compounds were applied in concentrations that block GABA responses both in cultured embryonic human DRG cells (Valeyev et al. 1999) and in noncultured, adult DRG neurons from a number of nonhuman species (Akaike et al. 1985; Dunlap 1984; Gallagher et al. 1978; Inoue and Akaike 1988). Bicuculline, picrotoxin, and TBPS appear to act at different sites of the GABA_A receptor/Cl channel complex. Bicuculline is established firmly as a specific, competitive antagonist of the GABA_A recognition site (Akaike et al. 1985, 1987); picrotoxin and some bicyclic cage compounds such as TBPS behave as noncompetitive Cl channel blockers at GABA_A receptors (Akaike et al. 1985; Newland and Cull-Candy 1992). That none of these compounds blocked alphaxalone-induced currents is further evidence of its independence from the GABA_A receptor. In sum, the Cl currents activated by alphaxalone and by GABA in human embryonic DRG neurons have disparate characteristics.

An association between the effects of general anesthetics and synaptic inhibition has been contemplated for many years (Eccles et al. 1963). It has been suggested that direct GABA_A receptor activation is the mechanism by which anesthetics induce general anesthesia (Schulz and Macdonald 1981). We propose that direct activation of a Cl conductance that does not involve GABA_A receptors in neuronal membranes also may be a factor in the anesthetic actions of alphaxalone. This proposal is plausible because the concentrations used in the present experiments are equivalent to those that have been measured in the plasma during surgical anesthesia with alphaxalone in humans (Sear and Prys-Roberts 1979).