Introduction

Top Summary Introduction Materials and Methods Results Discussion References

-aminobutyric acid (GABA) is the major inhibitory neurotransmitter in the vertebrate brain. Fast inhibitory postsynaptic potentials are mediated by GABA_A receptors (GABARs), which contain binding sites for many modulators including benzodiazepines, barbiturates, zinc and general anesthetics including neurosteroids. GABARs have also been identified on peripheral neurons and nonneuronal cell types. The role of GABARs outside the central nervous system (CNS) is not well documented, but it has been proposed that GABARs may play a role in the motility of uterine contractions and the secretion of hormones from endocrine cells.

GABARs are composed of five subunits that together form a transmembrane chloride ion channel. Four different mammalian subunit families (, , , ) and their subtypes (1-6, 1-3, 1-3) have been studied extensively (Macdonald and Olsen, 1994). Two new subunit families,  (Davies et al., 1997) and  (Hedblom and Kirkness, 1997), have recently been identified. In addition, , , , and  subunit subtypes have been shown to be differentially expressed throughout development (Laurie et al., 1992) and in different regions of the rat brain (Wisden et al., 1992). The developmental expression of  and  subunits has not been reported. Pharmacological studies of recombinant receptors have shown that individual subunits and their subtypes confer different sensitivities to such GABAR modulators as benzodiazepines (Pritchett et al., 1989; Wieland et al., 1992) and zinc (Draguhn et al., 1990). The subunit composition of the pentamer seems to be highly regulated and all of the potential subtype combinations do not assemble to create functional GABARs (Angelotti et al., 1993; Burgard et al., 1996). In addition, receptors composed of subunits may be further restricted by a 2:2:1 stoichiometry (Chang et al., 1996; Tretter et al., 1997).

The recently cloned  subunit is most closely related to GABAR  (37%) and  (35%) subunits and to the GABA_C receptor  subunit (33%) and is less similar to other GABAR or glycine receptor subunits. The  subunit was amplified from the cDNA libraries of a number of reproductive tissue (uterus, ovaries), digestive tissue (gall bladder, small intestine), and two brain regions (hippocampus and temporal cortex) with reverse transcription-polymerase chain reaction, but no transcripts were hybridized from whole brain samples with Northern analysis (Hedblom and Kirkness, 1997). In addition, the  subunit is expressed by the teratocarcinoma NT2 neuronal precursor cells (Hedblom and Kirkness, 1997) and the terminally differentiated NT2-N cells (Neelands et al., 1998). We have previously shown that the NT2 neuronal precursor cells also express high levels of mRNA for the 5 and 3 GABAR subunit subtypes and low levels of the 3 subtype. GABAR currents in these cells are highly sensitive to inhibition by zinc, enhanced by loreclezole, and unaffected by application of the benzodiazepine diazepam (Neelands et al., 1997).

There has been some question of whether or not the  subunit coassembles with other GABAR subunits to form functional GABARs. Human embryonic kidney 293 fibroblasts transfected with the  subunit alone or in combination with either an 1 or a 1 subunit did not bind the GABAR ligands, muscimol, or t-butylbicyclophosphorothionate and no current was evoked by application of GABA (Hedblom and Kirkness, 1997). Muscimol and t-butylbicyclophosphorothionate did, however, bind to GABARs composed of the 11 combination but was indistinguishable from binding to 11 GABARs (Hedblom and Kirkness, 1997). Transfection of cells with higher concentrations of  subunit cDNA than 2 subtype cDNA produced GABARs with reduced binding of the benzodiazepine site ligand flumazenil, which suggests that the  subunit was interfering with the ability of the  subunit to incorporate into functional GABARs (Hedblom and Kirkness, 1997). However, there is as yet no electrophysiological evidence that the  subunit is incorporated into functional GABARs. The aim of the present study was to determine whether coexpression of the  subunit with  and  or , , and  subunits produced GABARs with properties similar to or different from those of or receptors, consistent with incorporation of the  subunit into GABARs. In addition we wanted to determine whether incorporation of the  subunit altered the pharmacological and biophysical properties of GABARs.

	Materials and Methods

Top Summary Introduction Materials and Methods Results Discussion References

Transfections. Full-length cDNAs for rat GABAR 5 (obtained from A. Tobin, University of California, Los Angeles), 3 (obtained from D. Pritchett, University of Pennsylvania, Philadelphia, PA), and 3 (obtained from P. Seeburg, Max-Planck Institute for Medical Research, Heidelberg, Germany) subtypes were subcloned into the pCMVNeo expression vector and human  (obtained from E. Kirkness, The Institute for Genomic Research, Rockville, MD) was subcloned into the pCDM8 expression vector. For selection of transfected cells, the plasmid pHook-1 (Invitrogen, San Diego, CA) containing cDNA encoding the surface antibody sFv was also transfected into the cells. L929 cells were maintained in Dulbecco's modified Eagle's medium plus 10% heat-inactivated horse serum, 100 U/ml penicillin and 100 µg/ml streptomycin (Greenfield et al., 1997). Cells were passaged by a 5-min incubation with 0.5% trypsin/0.2% EDTA solution in PBS (10 mM Na₂HPO₄, 0.15 mM NaCl, pH 7.3).

Recording Solutions and Techniques. For both whole-cell and outside-out patch recording, the external solution consisted of 142 mM NaCl, 8.1 mM KCl, 6 mM MgCl₂, 1 mM CaCl₂, 10 mM glucose, and 10 mM HEPES, pH 7.4, and osmolarity adjusted to 295 to 305 mOsM. Recording electrodes were filled with an internal solution of 153 mM KCl, 1 mM MgCl₂, 5 mM K-EGTA, 10 mM HEPES, and 2 mM MgATP, pH 7.4, and osmolarity adjusted to 295-305 mOsM. These solutions provided equilibrium potential for Cl near 0 mV. Patch pipettes for whole-cell recordings were pulled from either borosilicate glass (Fisher Scientific, Pittsburgh, PA) or Labcraft microhematocrit capillary tubes (Curtin Matheson Scientific, Inc., Houston TX) on a P-87 Flaming Brown puller (Sutter Instrument Co., San Rafael, CA) to a resistance of 8-12 M. For single-channel recording, patch pipettes were pulled from thick-walled borosilicate glass with an internal filament (World Precision Instruments, Sarasota, FL), fire polished to a resistance of 5-10 M, and coated with Q-dope (GC Electronics, Rockford, IL) to reduce capacitance.

Data Analysis. Whole-cell currents were analyzed off-line with the programs Axotape and Prism (Graphpad, San Diego, CA). All whole-cell current amplitudes were obtained by measuring the peak current evoked during the application of GABA or GABA plus drug. The magnitude of the enhancement or inhibition of GABAR current by a drug was measured by dividing the peak amplitude of GABAR currents elicited in the presence of a given concentration of the drug (with GABA) by the peak amplitude of control current elicited by GABA alone and multiplying the fraction by 100 to express it as percent of control. Thus, the control response was 100%. Peak GABAR currents at various drug concentrations were fitted to a sigmoidal function with a four-parameter logistic equation (sigmoidal concentration-response) with a variable slope. The equation used to fit the concentration-response relationship was:
where I is the GABAR current at a given GABA concentration, I_max is the maximal GABAR current and n_H is the Hill coefficient. The curve-fitting algorithm minimized the sum of the squares of the actual distance of points from the curve. Convergence was reached when two consecutive iterations changed the sum of the squares by < 0.01%. Data were presented as mean ± S.E.M. All four parameters were "floating", and therefore, the maximum effect observed was not necessarily the upper limit of the fit (e.g., see Figs. 2A and 7B). The only cases in which one of the parameters was not allowed to float were when inhibitory concentration-response curves did not plateau at the low end. In these instances, the lower limit was set at zero to prevent the fitting algorithm from extending into the undefined range below 0%. Figures include the values for the four "floating" parameters: maximal and minimal values are given on the y-axes, and the EC₅₀ value and Hill coefficient are given in the insets. It is important to note that the Hill coefficient does not represent the number of molecules that bind to the receptor but is only an indication of the slope of the concentration-response relationship and is provided throughout the text only to describe the fit. For compounds that had smaller effects at higher concentrations, the fit did not include these higher points and were indicated by dashed lines in the figures. All comparisons of the effect of a compound or voltage on different subtype combinations were done with GABA concentrations that produced equivalent responses (extrapolated from Fig. 1B). Application of GABA was repeated until the peak currents had stabilized and functioned as controls for each cell. Coapplication of the same concentration of GABA with increasing concentrations of individual modulators was performed to determine the maximal effect and potency of each modulator on GABARs. All fits were made to average, normalized data with the current expressed as a percentage of the maximum current elicited by control GABA concentrations for each cell.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Comparison of GABA concentration-response curves. A, representative current traces of 0.1, 3, and 100 µM GABA recorded from L929 cells expressing 53, 533, 53, and 533. Horizontal bars, drug application. B, normalized averaged concentration-response curves for GABA-evoked currents from the same four subunit combinations. Data are mean ± S.E.M. C, comparison of EC50 values calculated from concentration-response curves of individual cells. (*p < .05; **p < .01).

	Results

Top Summary Introduction Materials and Methods Results Discussion References

GABA Sensitivity of Cells Coexpressing 53, 533, 53, and 533 Subtype Combinations

GABA Concentration-Response Curves. To determine which combinations of 5, 3, 3, and  subtypes could assemble in L929 cells to form functional GABARs, we coexpressed different subtype combinations in L929 cells and obtained whole-cell recordings and GABA concentration-response relationships. In parallel transfections, no GABAR currents were recorded from cells coexpressing 53 (n = 5), 33 (n = 5), 5 (n = 5), or 3 (n = 4) subtypes or the  subunit alone (n = 4). Previous work in our laboratory has shown that L929 cells do not form functional channels when cells are cotransfected with or subunits or transfected with an , , or  subunit alone (Angelotti et al., 1993; Saxena and Macdonald, 1994; Burgard et al., 1996; Neelands et al., 1999). In subsequent experiments, concentration-dependent GABAR currents were obtained from cells cotransfected with 53, 53, 533, and 533 subtypes (Fig. 1A). The amplitudes of currents varied considerably among transfections and among individual cells, presumably because of differences in transfection efficiency or cell viability. Given this variability, there were no apparent differences in the maximum amplitudes of GABAR currents among the four subunit combinations tested (p = .374). 53 currents were larger than previous reports from our laboratory on heterodimers. Whether this was caused by the subunit combination, the L929 cells used in this study, or some other factor was not investigated. Complete concentration-response curves were obtained for the four subunit combinations that were GABA-sensitive (Fig. 1B) to determine whether the  subunit altered GABA EC₅₀ values. Peak currents were normalized to the maximum current recorded for each cell. The average maximal currents evoked by GABA were 1002 ± 450, 502 ± 177, 375 ± 103, and 274 ± 96 pA for 53, 53, 533, and 533, respectively. Average amplitudes were normalized to the maximal current evoked for each cell and then plotted as a function of GABA concentration and fit with a four-parameter logistic equation (see Materials and Methods). Cells expressing 53 subtypes had an EC₅₀ value of 0.7 µM (n_H = 1.4; n = 5-13; Fig. 1B). GABA concentration-response curves obtained from cells expressing 53 subtypes with either  and/or 3 subtypes were shifted to the right (Fig. 1B). The EC₅₀ values for the 53, 533, and 533 subtype combinations were 1.3 µM (n_H, 1.4; n = 4-8), 1.5 µM (n_H, 1.6; n = 6-7), and 1.8 µM (n_H, 1.8; n = 6-7), respectively. GABA concentration-response curves for individual cells were also fit with logistic equations and compared across cells (Fig. 1C). The log EC₅₀ values from the different subunit combinations were significantly different (one-way ANOVA, p = .030). Post hoc tests (see Materials and Methods) indicated that the 53 subtype GABA EC₅₀ values were significantly different from each of the other subtype combination GABA EC₅₀ values (p < .05 for 53 and 533 and p < .01 for 533 subtype combinations). There were no significant differences among the GABA EC₅₀ values of 53, 533, and 533 subtype combinations. In addition, the normalized current at 1 µM GABA was significantly larger in cells expressing the 53 subunit combination than the 53, 533, and 533 subtype combinations (p < .01, p < .01, p < .001, respectively) consistent with the shift in apparent EC₅₀ value. The degree of the rightward shift in the GABA concentration-response curve (>half a log unit) for  subunit-containing receptors was similar to that previous reported after coexpression of a  subunit (Angelotti et al., 1993).

Current-Voltage Relationships. Current-voltage relations were generated for the four GABA-sensitive subtype combinations with EC₃₅ GABA concentrations. Peak currents were measured at 25-mV increments at holding potentials ranging from 100 to +75 mV (Fig. 2A). Because of the large variation in maximal current among subunit combinations and differences in expression levels among cells, currents from individual cells were normalized to the peak current recorded at 75 mV. The average normalized current for each subtype combination was plotted as a function of the holding potential (Fig. 2B). Currents from the 53 subtype combination were outwardly rectifying (Fig. 2A). Average peak currents evoked at positive potentials were 232 ± 14% (+50 mV) and 396 ± 33% (+75 mV) of the peak current recorded at 75 mV (Fig. 2B, inset). 533 (107 ± 10%, n = 7 at +75 mV) and 533 (n = 2, data not shown) subtype currents were both linear over the membrane potentials tested (Fig. 2B). The normalized current amplitudes for these two subunit combinations superimposed. 53 currents, however, still displayed outward rectification (209 ± 46% at +75 mV). An ANOVA performed on the degree of rectification among the 53, 53, and 533 subunit combinations was significant (p < .001). Post hoc tests showed the degree of rectification was significantly greater for currents from the 53 subtype combination compared with currents from 53 or 533 subtype combinations (p < .01 at +75 mV and p < .001 at +50 mV for both combinations). In addition, the 53 currents had significantly greater rectification than 533 currents (p < .05).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Comparison of the current-voltage relations of GABA-evoked currents. A, representative current traces evoked by EC35 GABA concentrations at holding potentials of 75 and + 75 mV for 53, 53, and 533 subunit combinations. The 75 mV current is inverted and superimposed on the + 75 mV trace. Horizontal bars, drug application. B, peak currents at each potential are normalized to the peak current at 75 mV. Average normalized currents are plotted on the y-axis against the membrane holding potential (HP) on the x-axis. Inset, amplitude ratio of peak currents evoked at +75 and 75 mV. Data are mean ± S.E.M. Pharmacology of GABAR-evoked currents from cells coexpressing 53, 533, 53, and 533 subtypes.

Pharmacology of GABAR Currents from Cells Coexpressing 53, 533, 53, and 533 Subtypes

Benzodiazepines. The effects of benzodiazepine site ligands, such as diazepam, have been shown to require a  subunit in the GABAR to exert their effects. To determine whether a  subunit could replace a  subunit in the formation of the benzodiazepine site we tested the ability of diazepam (1 µM) to enhance GABAR currents in cells coexpressing the  subunit with other subtypes (Fig. 3A). Control currents evoked from cells coexpressing the 53 subtypes by 1 µM GABA were not enhanced by diazepam (n = 4). In parallel transfections, diazepam enhanced 533 currents to 159 ± 13% of control (n = 15) but had no effect (105 ± 3%) on 533 currents (n = 5; Fig. 3B, C). The differences in current enhancement by diazepam were significantly different among the subtype combinations tested (p = .012). The magnitude of enhancement of 533 currents was significantly different (p < .05) from that of currents from the other two subtype combinations, but there was no difference between the effect of diazepam on 533 and 53 currents (p > .05).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Comparison of the effect of 1 µM diazepam. A, superimposed current traces of successive applications of 1 µM GABA (black trace) and 1 µM GABA plus 1 µM diazepam (gray trace) for 53, 533, and 533. Horizontal bars, drug application. B, average percentage of modulation of coapplication of 1 µM diazepam. Data are mean ± S.E.M. (*p < .05). C, effects of coapplication of 1 µM diazepam on individual cells expressing 53, 533, and 533.

Loreclezole. The novel anticonvulsant drug loreclezole has been shown to potentiate GABAR currents when the isoforms contained a 2 or 3 subunit subtype but not a 1 subtype (Wafford et al., 1994). Although the degree of potentiation by 1 µM loreclezole varied depending on the isoform, all currents from cells containing subunit combinations that included the 3 subtype were loreclezole-sensitive (Wingrove et al., 1994). To determine whether coexpression of the  subunit altered loreclezole sensitivity, we determined the concentration-dependence of loreclezole enhancement of currents evoked by EC₆₀ GABA concentrations from the 53, 53, 533, and 533 subtype combinations (Fig. 4A). Coapplication of up to 10 µM loreclezole caused a concentration-dependent increase of all subtype combination currents except the 533 combination (Fig. 4B). The percentage of enhancement started to decrease with application of higher loreclezole concentrations, as in previous reports (Donnelly and Macdonald, 1996). Interestingly, 533 and 53 currents were enhanced to a similar extent by loreclezole over the whole range of concentrations tested. The enhancement of loreclezole (3 µM) among subtype combinations was significantly different (p = .221). Enhancement of current from all three of these subunit combinations was statistically different from the loreclezole effect on the insensitive 533 currents at 3 µM loreclezole (p < .05).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Modulation of GABA-evoked currents by loreclezole. A, superimposed current traces of successive applications of EC60 GABA concentrations (upper trace) and coapplication of GABA plus 3 µM loreclezole (LOR; bottom trace) are shown for 53, 53, 533, and 533. Horizontal bars, drug application. B, concentration-response relationship of averaged normalized data for loreclezole modulation of currents evoked by EC60 GABA for the same subunit combinations. Data are mean ± S.E.M.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Dependence on GABA concentration for the effects of loreclezole on GABA-evoked currents from the 533 subunit combination. A, representative current traces of the concentration dependent effects of loreclezole (1-30 µM) on currents evoked by 0.2 µM GABA from the 533 combination. Horizontal bars, drug application. B, average sequential concentration-response relationships for GABA alone and GABA plus 1 µM loreclezole normalized to the peak current evoked by 30 µM GABA alone for the 533 combination. Arrows indicate the GABA concentration used in Figs. 4A and 5A, respectively. Data are mean ± S.E.M.

Barbiturates. The barbiturate pentobarbital has been shown to potentiate GABAR currents, directly activate a chloride current, and act as an open channel blocker at high concentrations (Schulz and Macdonald, 1981; Schwartz et al., 1986; Peters et al., 1988; Robertson, 1989; Thompson et al., 1996). We investigated the effect of coexpression of the  subunit with 53 or 533 subtypes on the modulatory actions of pentobarbital. Coapplication of pentobarbital (0.3-100 µM) with equally effective concentrations of GABA were used to determine the EC₅₀ value of pentobarbital for enhancement of 53, 533, and 53 currents. Higher concentrations of pentobarbital were not used because of the direct effects of pentobarbital on GABARs (Fig. 6A). All three subtype combination currents were enhanced in a concentration-dependent manner by pentobarbital. Averaged normalized data were fit with a logistic equation with apparent EC₅₀ values of 25.9 µM (n_H = 1.8, n = 3-5), 39.0 µM (n_H = 1.2, n = 4), and 59.5 µM (n_H = 1.1, n = 4) for 53, 533, and 53 currents, respectively (Fig. 6B). There were no statistical differences between the three subunit combinations when individual EC₅₀ values were compared (p > .05). The concentration range tested did not produce a plateau in the effect of pentobarbital at high concentrations. Therefore, fits were allowed to extend beyond the enhancement of 100 µM pentobarbital (see Materials and Methods) that resulted in maximal "effects" of 501% (53), 373% (53), and 356% (533). The effect of the highest concentration of pentobarbital tested (100 µM) ranged from 467 ± 130% of control for the 53 currents to 307 ± 59% for 53 and 265 ± 61% for 533 currents but were not statistically different (p > .05; Fig. 6B).

View larger version (25K): [in this window] [in a new window]   Fig. 6.   Enhancement of GABA-evoked currents by pentobarbital. A, superimposed current traces of successive applications of EC60 concentrations of GABA (upper trace), GABA plus 30 µM pentobarbital (PB; middle trace), and GABA plus 100 µM pentobarbital (lower trace) for 53, 53, and 533 subunit combinations. Horizontal bars, drug application. B, concentration-response relationship of averaged normalized data for pentobarbital enhancement of currents evoked by EC60 GABA for the same combinations. Data are mean ± S.E.M.

Neurosteroids. Neurosteroids have been shown to have both positive (alphaxalone) and negative [pregnenolone sulfate (PS)] allosteric effects on recombinant GABARs. To determine whether the  subunit altered the effects of either of these compounds, concentration-response curves were generated with EC equivalent concentrations of GABA in cells expressing either the 53, 53, or 533 subtype combinations.

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Enhancement of GABA-evoked currents by alphaxalone. A, superimposed current traces of successive applications of EC60 concentrations of GABA (upper trace) and GABA plus 10 µM alphaxalone (ALX; lower trace) for 53, 53, and 533 subunit combinations. Horizontal bars, drug application. B, concentration-response relationship of averaged normalized data for pentobarbital enhancement of currents evoked by EC60 GABA for the same combinations. EC50 value for alphaxalone enhancement (1 nM-3 µM alphaxalone) was derived independent of the apparent inhibition seen at higher concentrations (hatched line). Data are mean ± S.E.M.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Inhibition of GABA-evoked currents by PS. A, superimposed current traces of successive applications of EC80 concentrations of GABA (lower trace), GABA plus 3 µM PS (middle trace), and GABA plus 30 µM PS (upper trace) for 53, 53, and 533 subunit combinations. Horizontal bars, drug application. B, concentration-response relationship of averaged normalized data for PS inhibition of currents evoked by EC80 GABA for the same combinations. Data are mean ± S.E.M.

Zinc. The divalent cation zinc inhibited and isoform currents with low IC₅₀ values (<5 µM), isoforms with moderate IC₅₀ values (22-42 µM), and isoform currents with high IC₅₀ values (>100 µM; Draguhn et al., 1990; Saxena and Macdonald, 1994; Whiting et al., 1997). To determine the effect the  subunit on the sensitivity of GABAR currents to zinc, we obtained inhibition curves for zinc for the 53, 533, 53, and 533 subtype combinations. Zinc (100 nM-1 mM) inhibited currents evoked by EC₈₀ GABA concentrations for all of the subtype combinations tested (Fig. 9A). Averaged normalized currents were plotted as a function of zinc concentration and fit with logistic equations with IC₅₀ values of 2.1 µM (n_H = 0.8, n = 6), 2.4 µM (n_H = 0.9, n = 6), 43.3 µM (n_H = 0.5, n = 4), and 67.3 µM (n_H = 0.6, n = 5) for 53, 53, 533, and 533 currents, respectively (Fig. 9B). When comparing IC₅₀ values obtained from individual cells, the IC₅₀ values for zinc inhibition were statistically different (p < .0001). However, IC₅₀ values for zinc inhibition of 53 and 53 currents or 533 and 533 currents were not significantly different from one another. The individual IC₅₀ values for zinc inhibition of 53 and 53 currents were statistically different from inhibition of both 533 and 533 (p < .0001). The increase in the zinc IC₅₀ for -containing subunit combinations was consistent with previous reports (Draguhn et al., 1990) but was of smaller magnitude, possibly because of the expression of the 3 subtype rather than the 2 subtype (used in most recombinant studies of GABAR pharmacology).

View larger version (35K): [in this window] [in a new window]   Fig. 9.   Inhibition of GABA-evoked currents by zinc. A, superimposed current traces of successive applications of EC80 concentrations of GABA (lower trace), GABA plus 1 µM zinc (middle trace), and GABA plus 30 µM zinc (upper trace) for 53, 53, 533, and 533 subunit combinations. Horizontal bars, drug application. B, concentration-response relationship of averaged normalized data for zinc inhibition of currents evoked by EC80 GABA for the same combinations. Data are mean ± S.E.M.

Lanthanum. The trivalent cation lanthanum inhibited 6 subtype-containing receptor currents but enhanced 1 subtype-containing receptor currents (Saxena et al., 1997), but the effects of lanthanum on 5 subtype-containing receptor currents have not been reported. To determine the effects of  and 5 subunits on the sensitivity of GABARs to lanthanum we obtained concentration-response curves for lanthanum for the 53, 533, and 53 subtype combinations. Lanthanum inhibited currents evoked by EC₆₀ GABA concentrations for all of the combinations tested (Fig. 10A). Averaged normalized currents were plotted as a function of lanthanum concentration and fit with logistic equations with IC₅₀ values of 297 µM (n_H = 1.1, n = 4), 540 µM (n_H = 1.1, n = 5), and 522 µM (n_H = 0.9, n = 5) for 53, 53, and 533 currents, respectively (Fig. 10B). When comparing IC₅₀ values obtained from individual cells, there was a significant difference between subunit combinations (p = .0051). Lanthanum inhibited 53 currents with a significantly lower IC₅₀ value than either 533 or 53 currents (p < .05, p < .01, respectively). In addition, the percentage inhibition produced by a high concentration of lanthanum (1 mM) was different among subunit combinations (p = .0042). Lanthanum (1 mM) significantly inhibited 53 currents (82.0 ± 3.5%) more than either 53 (70.2 ± 3.2%) or 533 (66.9 ± 3.9%) currents (p < .05, p < .01, respectively). Similar to the inhibition by PS, this difference in "maximal" effect was most likely caused by the shift in the apparent IC₅₀ value and not to a change in efficacy.

View larger version (26K): [in this window] [in a new window]   Fig. 10.   Inhibition of GABA-evoked currents by lanthanum. A, superimposed current traces of successive applications of EC80 concentrations of GABA (lower trace), GABA plus 1 mM lanthanum (La3+; upper trace) for 53, 53, and 533 subunit combinations. Horizontal bars, drug application. B, concentration-response relationship of averaged normalized data for lanthanum inhibition of currents evoked by EC80 GABA for the same combinations. Data are mean ± S.E.M.

Comparison of Single-Channel GABAR Currents from Cells Expressing 53, 53, and 533 Subtypes

Single-channel currents were recorded from outside-out patches pulled from fibroblasts transfected with 53, 53, and 533 subtype combinations to determine whether there were any differences in the conductance levels of single-channel currents when the  subunit was included in the transfection. Previously heterodimeric pentamers were reported to have relatively small conductance levels, ranging from 11 to 13 pS (Moss et al., 1990; Verdoorn et al. 1990; Angelotti and Macdonald, 1993; Fisher and Macdonald, 1997) whereas heterotrimeric pentamers composed of , , and subunits had larger conductance levels ranging from 24 to 27 pS (Fisher and Macdonald, 1997; Neelands et al., 1999). 53 and 533 single-channel current openings were relatively long in duration and were separated into closely grouped bursts of openings in contrast to 53 single-channel current openings which were generally brief, independent openings (Fig. 11A). Detailed kinetic analysis of the open and closed times was not performed on the limited number of openings obtained for this study. The amplitudes of single-channel openings from cells transfected with either 53, 53, and 533 subtypes were measured at holding potentials ranging from 75 to +75 mV and were fit with linear regression analysis to determine single-channel conductance. Not all holding potentials were obtained for each patch and the n values given in the figure represent the range in the number of patches recorded at each holding potential. The average conductance of channels from individual patches varied depending on the subtype combination. It should be noted, however, that linear regression analysis of individual patches gave values similar to the fits of the average data (data not shown). Conductance levels of 15.2, 23.8, and 26.9 pS were calculated for channel openings from the 53 (n = 1-3), 53 (n = 2-4), and 533 (n = 3) subtype combinations, respectively (Fig. 11B). Although their openings had a slightly larger conductance than previously reported heterodimers, the pattern of 53 openings was similar to that of these studies. Further single-channel analysis should be performed to determine whether this small change in conductance was significant and represented a difference of 53 heterodimers compared with other combinations. Because we were more interested in whether adding a  or  subunit changed the properties of the single-channel openings, this difference was not pursued. Incorporation of either a  or 3 subunit into the GABAR complex increased the main conductance level from 53 alone and was in the range previously reported for heterotrimeric GABAR pentamers. 53 and 533 single-channel currents also displayed a bursting type behavior of openings not seen with heterodimers (Fig. 11).

View larger version (14K): [in this window] [in a new window]   Fig. 11.   Single-channel currents recorded from 53, 53, and 533 isoforms. A, representative outside-out single-channel openings recorded from fibroblast cells transfected with either 53, 53, or 533. B, current-voltage relationships are shown for each combination. Data are mean ± S.E.M.

	Discussion

Top Summary Introduction Materials and Methods Results Discussion References

Coexpression of the  Subunit-Altered GABAR Properties. By using electrophysiological techniques, we demonstrated that coexpression of the  subunit with 5, 3, or 3 subtypes alone did not result in expression of functional GABARs. However, coexpression of the  subunit with 53 or 533 subtypes resulted in GABARs with pharmacological and biophysical properties that were different from those of 53 and 533 GABARs. Expression of the 53 subunit combination in L929 cells resulted in GABAR currents that had a higher GABA EC₅₀ value (Table 1), less outward rectification, and larger single-channel conductance than 53 receptors. 53 channels tended to open into bursts of longer duration than single, brief openings typically recorded from patches containing 53 subunit channels. In addition, 53 currents were inhibited by lanthanum with a higher IC₅₀ value and were inhibited by PS with a lower IC₅₀ value than 53 currents (Table 1). In contrast, there were no significant differences in the effects of zinc, loreclezole, pentobarbital, diazepam, or alphaxalone on 53 and 53 GABAR currents (Table 1). No multiple component inhibition curves were obtained for 53 GABAR currents, and single-channel recordings did not reveal low 53 conductance levels, which suggests that it was unlikely that a significant proportion of 53 receptor channels were also expressed with 53 receptor channels.

View this table: [in this window] [in a new window]   TABLE 1 Pharmacological properties of 53, 53, 533, and 533 GABARs

Preferred Assembly of 53 Isoforms in NT2 Neuronal Precursor Cells. The results of this study imply that the  subunit can coassemble with other GABAR subunits to form functional recombinant GABARs. However, as noted above, it is not known if this coassembly actually occurs in neurons. Interestingly, NT2 neuronal precursor cells, a cell line that can be differentiated into a neuronal cell type by retinoic acid treatment, have been shown to express mRNA encoding the  subunit (Hedblom and Kirkness, 1997) along with 5, 3, and 3 GABAR subtype mRNAs (Neelands et al., 1997). Thus, comparisons of the pharmacological and biophysical properties of NT2 neuronal precursor cell and recombinant GABARs were made to determine whether functional evidence of incorporation of the  subunit into a GABAR could be demonstrated in a neuronal precursor cell type (Neelands et al., 1997). The pharmacological and biophysical properties of NT2 neuronal precursor cell GABAR currents, however, were most similar to those of recombinant 53 GABARs, suggesting that although  subunit mRNA was expressed by NT2 neuronal precursor cells, the  subunit protein was not incorporated into functional GABARs in these cells. Thus expression of the  subunit in a population of native neurons has not been clearly demonstrated and whether the  subunit is incorporated into native GABARs remains unclear.

Effects of Loreclezole on 5 and  Subtype-Containing GABARs. Loreclezole enhanced 533 and 53 subunit combination currents with similar EC₅₀ values. In contrast, 533 subunit combination currents were not enhanced by loreclezole but were inhibited by higher loreclezole concentrations. Interestingly, the 533 combination was enhanced less by 10 µM loreclezole than the 53 combination. This could have been caused by expression of two distinct populations of GABAR isoforms. However, the degree of potentiation of the current was not attenuated at lower concentrations, which would have been expected if a proportion of the channels were insensitive. The 2 and 3 subunit subtypes have been shown to have a single amino acid necessary for loreclezole enhancement (Wingrove et al., 1994), and no GABAR isoform containing either of these subtypes has been reported to be loreclezole-insensitive (Wafford et al., 1994). At lower GABA concentrations, however, 533 currents were enhanced. It is possible that the 533 isoform has a higher affinity for loreclezole at the inhibitory site than other isoforms. Therefore, at high GABA concentrations, loreclezole potentiation of 533 GABAR currents was masked by inhibition of the current by loreclezole. Regardless, the coexpression of the  subunit with 533 changed the concentration-dependent effects of loreclezole. These changes in loreclezole sensitivity provide further evidence for assembly of a recombinant 533 isoform.

Effects of Lanthanum on 5 and  Subtype-Containing GABARs. Lanthanum has been shown to potentiate currents from GABARs containing the 1 subtype and to inhibit currents from GABARs containing the 6 subtype (Fisher et al., 1997; Saxena et al., 1997). The effect of lanthanum on GABAR isoforms containing the other  subunits has not been reported. In this study, we showed that GABARs containing an 5 subtype were inhibited by lanthanum but with a much higher IC₅₀ value than for inhibition of GABARs containing an 6 subtype (Table 2). The coexpression of  or 3 subtype with 5 and 3 subtypes slightly increased the IC₅₀ value for lanthanum inhibition compared with the IC₅₀ value for lanthanum inhibition of 53 currents. Although the enhancing effect of lanthanum has been shown to be restricted to the amino-terminal extracellular domain (Fisher et al., 1997), the exact site or sites of action of lanthanum have not been determined. It is possible that the inhibitory and enhancing effects of lanthanum act at completely different sites on the GABAR complex. Different  subunits may have only one site or may have both sites but with a different rank order of potency for lanthanum. Determining the effects of lanthanum on the other  subunits in combination with studies of chimeric or mutant receptors may give insights into the mechanism(s) of action of lanthanum.

View this table: [in this window] [in a new window]   TABLE 2 Pharmacological properties of , , , and GABARs

Potential Biological Roles for the  Subunit. We have demonstrated that coexpression of the  subunit with or subunits in a non-neuronal expression system produced GABAR isoforms with different properties consistent with assembly of or receptors (Table 1). In addition, the pharmacological properties of receptors were different from those of , , and receptors, which suggests that incorporation of the  subunit into GABARs could alter regulation of native GABAR currents by GABAR modulators (Table 2). The other heterotrimeric pentameric receptors confer specific pharmacological properties to recombinant GABARs [such as benzodiazepine sensitivity ()], eliminate the sensitivity to neurosteroids (), or produce spontaneously active channels that are sensitive to allosteric modulation (; Pritchett et al., 1989; Zhu et al., 1996; Neelands et al., 1999). Currents from  subunit-containing GABARs, which have a high sensitivity to inhibition by zinc, insensitivity to diazepam, and differential modulation by neurosteroids, might play a critical role in neuronal development or regulation of neuronal excitability. It remains to be determined, however, if the  subunit is incorporated into native GABARs. The GABAR subunits expressed along with the  subunit in NT2 neuronal precursor cells are predominately expressed in the brains of perinatal rats and not well expressed in the adult rat brain (Laurie et al., 1992). It is possible that the  subunit is incorporated primarily into "immature" GABAR isoforms, similar to the  subunit of the nicotinic acetylcholine receptor (Witzemann et al., 1990). On the other hand, the expression of  subunit mRNA is highest in non-neuronal tissue (Hedblom and Kirkness, 1997), which indicates that it may be more critical in the function of peripheral GABARs than central GABARs. Future work on the  subunit will be required to determine its role in GABA-mediated inhibition in the brain.