Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The interplay of neurotransmitters and their corresponding ligand-gated ion channels plays a pivotal role in inhibition or excitation of synaptic transmission. In the central nervous system (CNS), inhibitory transmission is mediated predominantly through interactions of the neurotransmitter -aminobutyric acid (GABA) with two classes of receptor-chloride channel complexes: -aminobutyric acid_A receptor channels (GABA_ARs) and -aminobutyric acid_C receptor channels (GABA_CRs; Macdonald and Olsen, 1994). These receptor channels are differentially distributed within the CNS, with GABA_ARs ubiquitous throughout the CNS, whereas GABA_CRs are found primarily within the retina (Enz et al., 1995; Lukasiewicz, 1996). The main criteria for distinguishing between these two classes of receptor channels are their differential responses to drugs. For instance, the barbiturates and the benzodiazepines can modulate GABA_ARs by increasing the magnitude of the GABA-induced current (Macdonald and Olsen, 1994), whereas GABA_CRs are insensitive to these two classes of drugs (for reviews, see Johnston, 1996; Lukasiewicz, 1996; Feigenspan and Bormann, 1998).

Metabolites of the stress hormone corticosterone and the female sex hormone progesterone compose another class of GABA_AR modulators: the neuroactive steroids (Harrison and Simmonds, 1984; Callachan et al., 1986, 1987; Majewska et al., 1986; Barker et al., 1987; Harrison et al., 1987; Morrow et al., 1987; Gee et al., 1988; Peters et al., 1988; Turner et al., 1989; Paul and Purdy, 1992; Twyman and Macdonald, 1991; Lambert et al., 1995; Le Foll et al., 1997). The concentrations of these metabolites can increase markedly within the CNS after stress and can vary during menstrual cycles (Purdy et al., 1990, 1991; Paul and Purdy, 1992; Negri-Cesi et al., 1996). Two metabolites of corticosterone, allotetrahydrodeoxycorticosterone (5-THDOC) and tetrahydrodeoxycorticosterone (5-THDOC), are positive modulators of GABA_ARs, with the 5 compound being the more efficacious of the two (Harrison et al., 1987; Peters et al., 1988; Im et al., 1990; Kokate et al., 1994; Xue et al., 1997). The progesterone metabolites pregnanolone, 5-dihydroprogesterone (5-DHP), and allopregnanolone also enhance GABA-induced current from GABA_AR (Harrison et al., 1987; Poisbeau et al., 1997; Reith and Sillar, 1997; Maitra and Reynolds, 1999). In nanomolar concentrations, these neuroactive steroids can potentiate the GABA-elicited currents, and at higher concentrations, they can act as partial agonists on the GABA_AR (Harrison and Simmonds, 1984; Barker et al., 1987; Morrow et al., 1990; Puia et al., 1990; Wittmer et al., 1996; Le Foll et al., 1997).

Thus far, no potentiators have been reported for GABA_CRs (₁). Here, the two-electrode voltage-clamp technique of an oocyte expression system is used to study the effects of neuroactive steroids on ₁ receptor channels. Several metabolites of corticosterone and progesterone, as well as a synthetic steroid alphaxalone (Harrison and Simmonds, 1984; Cottrell et al., 1987), are demonstrated to modulate the activity of homo-oligomeric ₁ receptor channels in a positive or negative fashion. In view of these findings, correlation between the structure of these neuroactive steroids and their differential effect on the ₁ receptor channel is presented.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

The plasmid vector containing the human ₁ cDNA was linearized using restriction enzyme SspI. This restriction site is located a few hundred bases downstream from the stop codon, which during synthesis of the cRNA results in incorporation of additional sequences at the 3' end of the cRNA. These auxiliary sequences may enhance the stability of the synthesized cRNA within the oocyte. The resulting DNA template was in vitro transcribed into cRNA using the T7 Megascript in vitro transcription kit (Ambion, Austin, TX). The quality of the cRNA was determined using electrophoresis of set dilutions of the products on a 1% agarose gel containing formaldehyde.

Xenopus laevis (Xenopus I, Ann Arbor, MI) were anesthetized via hypothermia, and oocytes were surgically removed and placed into oocyte Ringer's solution (OR2; 82.5 mM NaCl, 2.5 mM KCl, 1 mM CaCl₂, 1 mM NaPO₄, 1 mM MgCl₂, 5 mM HEPES, 2.5 mM sodium pyruvate, 50 U/ml penicillin, and 50 µg/ml streptomycin, pH 7.5). The oocytes were then dissociated in 82.5 mM NaCl, 2.5 mM KCl, 1 mM NaPO₄, 1 mM MgCl₂, and 5 mM HEPES, pH 7.5, plus 0.3% collagenase A (Boehringer-Mannheim Biochemicals, Indianapolis, IN) for ~2 h. After separation, the oocytes were washed thoroughly with OR2. Finally, stage VI oocytes were isolated and maintained in OR2 containing 2% horse serum at 18°C.

Micropipettes for cRNA injection were made on a Sutter P87 horizontal puller (Sutter Instrument Co., Novato, CA), and the tips were cut off using microscissors. cRNA for injection was drawn up into the micropipette using negative pressure and injected into the oocytes by the application of positive pressure using a PICOSPRITZER II (General Valve Corporation, Fairfield, NJ).

At 2 to 3 days postinjection, the oocytes were placed on a nylon mesh suspended in a recording chamber (volume, ~50 µl). This recording chamber has an inlet in the top and an outlet in the bottom that allowed continuous perfusion of control or drug solution with ~2 ml/min. Twenty separate reservoirs (100-ml glass containers) were connected to four six-way valves, and the outlet of each of these six-way valves (the sixth position was connected to the reservoir containing control solution) was connected to one four-way valve. The outlet of the four-way valve lead to the chamber. In this way, up to 20 different solutions could be introduced to an individual oocyte. The exchange time (dead time plus equilibration time in the chamber), which is ~7 s, is accounted for in the T_1/2 of deactivation measurements. Switching between the different solutions was controlled manually. The oocytes were constantly perfused with recording OR2 (OR2 lacking Na₂HPO₄, Na pyruvate, and antibiotics) and switched to test solutions containing GABA or GABA plus steroid.

The neuroactive steroids were purchased from Sigma Chemical Co. (St. Louis, MO), Research Biochemicals Inc. (Natick, MA), and Steraloids (Newport, RI) to compare batch-to-batch variation. No significant variability in the outcome of the results was noted when samples of the same steroid obtained from different sources were compared. The stock solutions of 10 mM steroids were made in dimethyl sulfoxide. Test solutions containing drugs were made by adding the steroid stock solutions to rapidly stirring recording OR2. Given that these neuroactive steroids are highly hydrophobic, the maximum feasible concentration of these compounds within the recording OR2 appeared to be ~20 µM. The presence of the vehicle solution dimethyl sulfoxide at the maximum tested concentration (0.2%) did not alter the GABA-induced current from ₁ receptor channel.

Recording electrodes were fabricated on a Narishige PP-83 (Narishige Scientific Instrument Lab., Tokyo, Japan). Electrodes were then filled with a solution of 3 M KCl. The oocytes were voltage-clamped at 70 mV using a TURBO TEC-05 npi (Adams and List, Westbury, NY) amplifier, and output was recorded on tape and chart paper by a Gould TA240 chart recorder.

Percent potentiation (PP) and percent inhibition (PI) were calculated as follows:
where I_GABA is the current elicited by a control application of GABA, and I_Steroid is the current with the application of both GABA and steroid. Thus, a 100% potentiation represents a current twice the amplitude of the control GABA-elicited current.

The IC₅₀ value and Hill coefficient were determined by fitting the concentration-response relationship to the following logistic equation using Sigma Plot 4.0. 
where I is the peak current at a given concentration of steroid [S], I_max is the maximal inhibited current, IC₅₀ is the concentration of steroid giving half-maximal inhibition, and n is the Hill coefficient.

The one-tailed Student's t test was used to calculate confidence levels for T_1/2 of deactivation and concentration dependence of 5-THDOC inhibition. Nonpaired analysis was used for the T_1/2 calculation, and paired analysis was used for the inhibition by 5-THDOC.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

5-THDOC Potentiates GABA-Evoked Currents of ₁ Receptor Channels. Figure 1A depicts representative current traces from the application of GABA and GABA plus different concentrations of 5-THDOC to an oocyte expressing ₁ receptor channels. Serial bath application of GABA (0.4 µM; GABA EC₅₀ = 1.03 ± 0.26) plus 1, 5, 10, and 20 µM 5-THDOC resulted in currents that were greater in magnitude than the GABA current alone. With successively higher concentrations, the amplitude of the GABA currents increased without reaching a plateau even in the presence of the highest feasible concentration of 5-THDOC (20 µM, see Materials and Methods). These currents did not exhibit desensitization even in the presence of the highest concentration of 5-THDOC and GABA. The bar graph representing the mean percent potentiation of GABA currents (±S.E.M) in the presence of different concentrations of 5-THDOC is shown in Fig. 1B (n = 4). The percentage potentiation of the GABA-evoked currents was 24 ± 7% and 87 ± 14% in the presence of 1 and 5 µM 5-THDOC, respectively (n = 4). This value was further increased to 129 ± 10% and 198 ± 9% (n = 4) with the coapplication of 10 and 20 µM 5-THDOC, respectively. Fitting of the Hill equation to the concentration-response data did not yield an EC₅₀ value or a Hill coefficient for 5-THDOC because an extrapolated maximum could not be calculated. The agonistic action of 5-THDOC alone was also examined. This steroid at a concentration as high as 20 µM did not activate ₁ receptor channels (data not shown).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   5-THDOC potentiation of GABA-evoked currents from 1 receptor channels. A, representative current traces from application of GABA or coapplication of GABA with different concentrations of 5-THDOC. Thick lines above current trace represent the duration of application of either GABA or GABA plus steroid. B, the bar graph represents percent potentiation of GABA-evoked currents in the presence of increasing concentrations of 5-THDOC (n = 4). The error bars are the S.E.M. C, average percent increase in T1/2 of deactivation with increasing concentration of 5-THDOC (n = 4). Note the increase in the decay time of the GABA-induced current with coapplication of 5-THDOC. The asterisk denotes a T1/2 of deactivation value significantly different from that of GABA alone (P < .05, one-tailed Student's nonpaired t test). D, residual effect of 5-THDOC on subsequent GABA applications. The return of the GABA-induced current to the control level occurred over several applications of GABA. The times between the drug applications are ~4 min.

Alphaxalone and Allopregnanolone Potentiation of ₁ Receptor Channels. Additional neuroactive steroids (alphaxalone and allopregnanolone) were also tested with the above protocol. Figure 2 shows the current traces as well as the mean percent potentiation of GABA (0.4 µM) responses for ₁ receptor channels in the presence of 1, 5, 10, and 20 µM alphaxalone and allopregnanolone. Both compounds were found to be positive modulators of ₁ receptor channels. The GABA-evoked currents in the presence of alphaxalone or allopregnanolone display characteristics similar to 5-THDOC in that they did not exhibit desensitization even in the presence of the highest concentration of the neuroactive steroids. The times to peak for these currents were also prolonged with increasing concentrations of these compounds.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Alphaxalone and allopregnanolone modulation of 1 receptor channels. A, current traces from application of GABA or coapplication of GABA with either alphaxalone or allopregnanolone. Control currents were scaled to the same peak height to demonstrate the difference in efficacy between these compounds. Thick lines above current trace represent duration of application of either GABA or GABA plus neuroactive steroid. B, average percent potentiation (±S.E.M.) for coapplication of GABA with different concentrations of alphaxalone (n = 3) and allopregnanolone (n = 3).

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Coapplication of either alphaxalone or allopregnanolone with GABA lengthens the T1/2 of deactivation and confers long-lasting effects. A, average percent increase in T1/2 of deactivation over control obtained from coapplication of either alphaxalone (n = 3) or allopregnanolone (n = 3) with GABA. Alphaxalone caused the greatest increase in the T1/2 of deactivation. Significant difference between T1/2 of deactivation in the presence and absence of steroid (*P < .05, **P < .01). B, current traces showing the prolonged effect of alphaxalone or allopregnanolone on 1 receptor channels. The time between the drug applications is ~4 min.

Inhibition of ₁ Receptor Channels by Pregnanolone, 5-THDOC, and 5-DHP. In contrast to the above neuroactive steroids, pregnanolone, 5-THDOC, and 5-DHP were found to inhibit ₁ receptor channels. Figure 4A shows the current trace from bath application of either GABA (0.4 µM) alone or GABA (0.4 µM) and 1 to 20 µM concentrations of pregnanolone to an oocyte expressing ₁ receptor channels. The GABA (0.4 µM)-evoked currents were inhibited by pregnanolone in a concentration-dependent manner. Pregnanolone at concentrations of 0.5 and 1 µM decreased GABA-evoked currents by 18 ± 5 and 29 ± 5% (n = 4), respectively (Fig. 4B). The currents were further reduced by 45 ± 5 and 46 ± 5% with coapplication of 5 and 10 µM concentrations of this steroid. The inhibition appeared to reach its maximum around 5 µM pregnanolone. At a concentration of 20 µM, pregnanolone caused a partial reversal of the inhibition (32 ± 5% reduction in 0.4 µM GABA-evoked current). The concentration-response relationship for pregnanolone is shown in Fig. 4C. Fitting these data points to the Hill equation yielded an IC₅₀ value of 0.55 µM and a Hill coefficient value of 1.8 for pregnanolone.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Pregnanolone, 5-THDOC, and 5-DHP inhibition of 1 receptor channels. A, representative current traces from application of GABA or coapplication of GABA and steroid. Control currents were scaled to the same peak height to demonstrate the relative efficacies of pregnanolone, 5-THDOC, and 5-DHP. Thick lines above current trace represent duration of application of either GABA or GABA plus steroid. B, concentration-response curves for the three inhibitory neuroactive steroids. Imax is the current amplitude in the absence of the neuroactive steroids, and I is the current in the presence of steroid. As shown, pregnanolone is the most potent of these compounds.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Long-lasting effects of pregnanolone, 5-DHP, and 5-THDOC. Thick lines above current trace represent application time of either GABA or GABA plus neuroactive steroid. Concentrations of pregnanolone (10 µM) and 5-THDOC (10 µM) that elicited the maximum inhibition were used for this experiment. Pregnanolone, the most potent inhibitor, also required the greatest period of time for recovery. The time between each drug applications is ~4 min.

Modulation by 5-THDOC and 5-THDOC Is GABA Concentration Dependent. Figure 6 shows the degree of potentiation by 5-THDOC (10 µM) in the presence of 0.2, 0.4, 0.8, 1, 2, 4, and 10 µM GABA. These GABA concentrations range from fractions of to nearly 10 times the GABA EC₅₀ value for ₁ receptor channels (1.03 ± 0.26 µM). There was a significant potentiation in the presence of 0.2 µM GABA. In the presence of low concentrations of GABA (0.2 and 0.4 µM), 5-THDOC (10 µM) caused a significant potentiation (216 ± 10% and 95 ± 7%, respectively) in the magnitude of the GABA-evoked current (Fig. 6B, n = 4). The effects of 5-THDOC on GABA-induced currents, however, decreased with increasing concentrations of GABA. Furthermore, there was a slight inhibition when concentrations of GABA greater than the EC₅₀ value were used. For example, coapplication of 10 µM 5-THDOC and 4 or 10 µM GABA reduced the GABA-induced currents by 8 ± 4% and 10 ± 4% of the control value, respectively.

View larger version (26K): [in this window] [in a new window]   Fig. 6.   GABA concentration-dependent modulation of 1 receptor channels by 5-THDOC and 5-THDOC. A, representative current traces from application of increasing concentrations of GABA in the presence of 5-THDOC (10 µM) or 5-THDOC (10 µM). Control currents for all concentrations of GABA were scaled to same peak height. Note that 5-THDOC potentiation occurs at concentrations of GABA equivalent to a fraction of the EC50 value, whereas at higher concentrations of GABA, 5-THDOC causes a slight inhibition. Thick lines above current trace represent application time of either GABA or GABA plus steroid. B, average percent potentiation (or inhibition; ±S.E.M.) for 10 µM concentration of either 5-THDOC or 5-THDOC in the presence of different concentrations of GABA. The decrease in inhibition at higher concentrations of GABA is statistically significant (*P < .05).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, the effects are shown of several neuroactive steroids on the ₁ receptor channel. Allopregnanolone, alphaxalone, and 5-THDOC all potentiated the GABA-induced currents and prolonged the decay time. In contrast, the coapplication of GABA with 5-THDOC, pregnanolone, or 5-DHP inhibited the ₁ GABA-evoked current. Collectively, the degree of potentiation and, to a lesser extent, inhibition of ₁ GABA-elicited currents by these neuroactive steroids were dependent on the GABA concentration. These effects were most prominent in the presence of low concentrations of GABA, equivalent to a fraction of the EC₅₀ value. Finally, the effects of the neuroactive steroids on ₁ receptor channels were shown to be long lasting because the applications of GABA alone did not return to the control level for several minutes subsequent to neuroactive steroid treatment.

The most striking finding of this study was the differential modulation of ₁ receptor channels by neuroactive steroids. The 5 derivatives were potentiators, whereas the 5 compounds were inhibitors of the GABA-evoked currents. This is intriguing because all of the 5 and 5 steroid derivatives examined in this study are known to be potentiators of GABA_ARs (Harrison et al., 1987; Kokate et al., 1994; Le Foll et al., 1997). The ability of the neuroactive steroid to potentiate or inhibit ₁ receptor channels depended on the position of the hydrogen atom attached to the fifth carbon (5 versus 5 neuroactive steroids; Fig. 7A). For instance, 5-THDOC significantly potentiated the GABA-induced current of the ₁ receptor channels, whereas 5-THDOC was inhibitory. A comparison of allopregnanolone and pregnanolone further corroborates this hypothesis given that the only difference in the structure of these two compounds is the position of the hydrogen atom on the fifth carbon (Fig. 7B). A comparison of the architectural differences of these compounds reveals that switching of the fifth carbon hydrogen from the  to the  position induces a trans- or cis-configuration at the site of the A and B ring fusion. This structural switch can result in multiple physical changes, including ~10% alteration in length (Fig. 7A), as well as a shift in the dipole moment of the molecule. It is tempting to speculate that these physical differences can influence the relative position of these compounds within their effector site, which can then influence the gating components of the ₁ ion channel complex in an opposing fashion. In comparison, other structural differences among the neuroactive steroids tested appear to affect only the relative potencies and efficacies for these compounds on the ₁ receptor channel. For example, replacement of the hydroxyl bound to the third carbon with a ketone decreased the potency of 5-DHP in comparison to pregnanolone.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   A, structures of the neuroactive steroids 5-THDOC, alphaxalone, allopregnanolone, pregnanolone, 5-THDOC, and 5-DHP. B, superimposed view of pregnanolone and allopregnanolone. Note the architectural alteration caused by alternative placement of the hydrogen bound to the fifth carbon.

The concentration of GABA played an important role in the degree of modulation by these steroids. For example, the modulation of ₁ receptor channels by 5-THDOC was dependent on the GABA concentration, yielding potentiation only at exceedingly low concentrations of GABA (below the EC₅₀ value) yet causing inhibition at higher concentrations. In comparison, potentiation of the GABA_AR by neuroactive steroids occurs over a greater range of GABA concentrations, including concentrations of GABA above the EC₅₀ value (Le Foll et al., 1997). Differences in the activation and deactivation kinetics of hetero-oligomeric GABA_ARs and homo-oligomeric ₁ receptor channels (Amin and Weiss, 1994, 1996) may explain why potentiation of the latter is more dependent on GABA concentration.

Previous studies with expression of retinal mRNA or cloned ₁ subunits within X. laevis oocytes have suggested that the bicuculline-insensitive receptor channels (GABA_C, ₁ receptor channel) do not respond to neuroactive steroids (Woodward et al., 1992). It is important to note that there are no contradictions between results of the previous studies and the data presented here. First, the concentration of GABA used in those studies can dampen the effect of the tested neuroactive steroids. Second, lower concentrations of these steroids were used in those studies, which could in turn result in a less intense response. Finally, the most effective compound used here (5-THDOC) was not tested in the aforementioned study.

In addition to a contrast in modulation by neuroactive steroids, there are other key differences between the responses of the GABA_AR and the ₁ receptor channel to neuroactive steroids. Overall, GABA_ARs display higher sensitivity to neuroactive steroids compared with ₁ receptor channels. For instance, modulation of the GABA_AR is detectable with concentrations of neurosteroids in the nanomolar range (Harrison et al., 1987; Kokate et al., 1994), whereas for 1 receptor channel, micromolar concentrations of neurosteroids are required to exert an effect. Moreover, these neuroactive compounds can directly activate the GABA_ARs at concentration used in this study but show no agonistic properties on ₁ receptor channels. What could account for the difference in the sensitivity between the ₁ receptor channel and the GABA_AR? It has been demonstrated recently that mutation of a single tryptophan (Trp328) within the third transmembrane domain of the ₁ subunit to any hydrophobic residues confers generic pentobarbital sensitivity to ₁ receptor channels (Amin, 1999). The converse mutation of the corresponding residue within the  and  subunits of the GABA_AR to a Trp residue blocks the action of the general anesthetic enflurane (Mihic et al., 1997). It is possible that the action of neurosteroids on the ₁ receptor channel could also be masked, at least in part, by the presence of a large amino acid such as a Trp residue. Other residues within the second transmembrane domain of GABA_AR have also been implicated in the action of anesthetics and ethanol (Belelli et al., 1997; Mihic et al., 1997), which could influence the potency of the neuroactive steroids on ₁ receptor channels.

Further studies using site-directed mutagenesis and kinetic analysis of ₁ and GABA_ARs are needed to explore the mechanisms involved in differential neuroactive steroid action on these two closely related receptor channels.