Two γ2L subunit domains confer low Zn2+ sensitivity to ternary GABAA receptors
Article first published online: 5 AUG 2004
The Journal of Physiology
Volume 532, Issue 1, pages 17–30, April 2001
How to Cite
Nagaya, N. and Macdonald, R. L. (2001), Two γ2L subunit domains confer low Zn2+ sensitivity to ternary GABAA receptors. The Journal of Physiology, 532: 17–30. doi: 10.1111/j.1469-7793.2001.0017g.x
- Issue published online: 5 AUG 2004
- Article first published online: 5 AUG 2004
- (Received 24 August 2000; accepted after revision 2 December 2000
- 1The sensitivity of GABAA receptors (GABARs) to Zn2+ inhibition depends on subunit composition. The predominant neuronal forms of mammalian GABARs, αβγ and αβδ, are differentially sensitive to Zn2+ inhibition; αβγ receptors are substantially less sensitive than αβδ receptors. Recently, functional domains involved in Zn2+ sensitivity have been identified in α and β subunits. Our aim in the present study was to localize functional domains of low Zn2+ sensitivity within γ2L subunits.
- 2Chimeric subunits were constructed by progressively replacing the rat γ2L subunit sequence with that of the rat δ subunit sequence. Whole-cell currents were recorded from mouse L929 fibroblasts coexpressing wild-type rat α1 and β3 subunits with a chimeric δ-γ2L subunit.
- 3Unlike α and β subunits, the γ2L subunit was found to contain a determinant of low Zn2+ sensitivity in the N-terminal extracellular region. In addition, we identified determinants in the M2 segment and the M2-M3 loop of the γ2L subunit that are homologous to those found in β and α subunits.
- 4We postulate that the interface between the latter two domains, which may form the outer vestibule of the channel, represents a single functional domain modulating Zn2+ sensitivity. Thus, the Zn2+ sensitivity of ternary GABARs appears to be determined by two functional domains, one in the N-terminal extracellular region and one near the outer mouth of the channel.
The γ-aminobutyric acid type A receptor (GABAR) is the major inhibitory neurotransmitter receptor in the mammalian central nervous system. Functional mammalian GABARs are thought to be pentameric combinations of homologous subunits drawn from seven different families with multiple subtypes: α(1-6), β(1-3), γ(1-3), δ, ε, π and θ (Macdonald & Olsen, 1994; Davies et al. 1997; Hedblom & Kirkness, 1997; Bonnert et al. 1999). Like other members of the ligand-gated ion channel gene superfamily, GABAR subunits have a putative membrane topology consisting of a large N-terminus, four membrane-spanning segments (M1-M4), one extracellular and two intracellular loops connecting the membrane-spanning segments (M1-M2, M2-M3, M3-M4), and an extracellular C-terminus. It has been suggested that neuronal GABARs are predominantly ternary αβγ and αβδ combinations (Angelotti et al. 1993; McKernan & Whiting, 1996).
The divalent cation Zn2+ has been proposed to be an endogenous modulator of synaptic transmission (Smart et al. 1994; Harrison & Gibbons, 1994). Its ability to inhibit native GABARs, however, varies with neuronal type, age and activity (Smart et al. 1994; Kapur & Macdonald, 1997). Studies of recombinant receptors indicate that the inhibitory activity of Zn2+ on GABAR currents depends on subunit composition (Draguhn et al. 1990; Smart et al. 1991). Binary αβ GABARs have a high sensitivity to Zn2+ with IC50 values ranging from 0.1 to 1 μm (Draguhn et al. 1990; Smart et al. 1991; Wooltorton et al. 1997; Horenstein & Akabas, 1998). In ternary αβX GABARs (where X is δ, γ, ε or π), Zn2+ sensitivity is decreased. Receptors containing π subunits have a Zn2+ IC50 of ≈2 μm (Neelands & Macdonald, 1999), receptors containing δ subunits have Zn2+ IC50 values ranging from 5 to 16 μm (Saxena & Macdonald, 1996; Krishek et al. 1998), and receptors containing ε subunits have a Zn2+ IC50 of ≈40 μm (Whiting et al. 1997; Neelands et al. 1999). Incorporation of a γ subunit, however, has an even greater effect on GABAR Zn2+ sensitivity, resulting in IC50 values that range from 20 to 600 μm (Saxena & Macdonald, 1996; Burgard et al. 1996; Krishek et al. 1998).
Functional domains involved in Zn2+ sensitivity have been identified in two GABAR subunit families, α and β. The extracellular end of the M2 segment in the β subunit has been shown to be a major determinant of Zn2+ sensitivity in β homomers and αβ receptors (Wooltorton et al. 1997; Horenstein & Akabas, 1998). The M2-M3 extracellular loop of the α6 subtype has been shown to confer higher Zn2+ sensitivity to α6β3γ2L GABAR isoforms as compared to α1β3γ2L receptors (Fisher & Macdonald, 1998). In addition, in GABAC receptors, which are structurally homologous to GABARs, the N-terminus has an essential role in conferring Zn2+ sensitivity to ρ1 homomers (Wang et al. 1995). Such functional domains, however, have yet to be identified for other GABAR subunit families including δ and γ.
To determine which domains of the γ2L subunit subtype are involved in conferring low Zn2+ sensitivity to GABARs, we constructed GABAR chimeras by replacing varying lengths of the rat γ2L subunit with the homologous sequence of the rat δ subunit. These chimeric subunits were coexpressed in mouse L929 cells with wild-type α1 and β3 subunits, and whole-cell currents were elicited by application of GABA. The relative Zn2+ sensitivities of GABARs containing the chimeric subunits revealed that two different and non-contiguous γ2L subunit domains, one in the N-terminal extracellular region and one near the outer mouth of the channel, contributed to the regulation of Zn2+ sensitivity.
Construction of expression vectors and mutagenesis
Full-length cDNAs for the rat GABAR α1 (Dr A. Tobin, University of California, Los Angeles, CA, USA), β3 (Dr D. Pritchett, University of Pennsylvania, Philadelphia, PA, USA), δ (Dr K. Angelides, Baylor College of Medicine, Houston, TX, USA) and γ2L subunits (F. Tan, University of Michigan) were subcloned into the expression vector pCMVneo (Huggenvik et al. 1991). Chimeras were constructed using the splice overhang extension method. Point mutations were generated using the QuikChange mutagenesis kit (Stratagene, La Jolla, CA, USA). Oligonucleotide primers were synthesized by the University of Michigan DNA synthesis core. Sequences of chimeras and point mutants were verified by fluorescent DNA sequencing (University of Michigan DNA sequencing core).
Cell culture and transient transfection
The mouse fibroblast cell line L929 (American Tissue Type Collection, Rockville, MD, USA) was grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10 % horse serum, 100 i.u. ml−1 penicillin and 100 μg ml−1 streptomycin (all from Gibco-BRL, Grand Island, NY, USA). Cells were maintained in a 37 °C incubator with 95 % air-5 % CO2 in 10 cm culture dishes. Cells were passaged every 3-4 days using 0.5 % trypsin-0.2 % EDTA (Boehringer-Mannheim, Indianapolis, IN, USA) in divalent-free, phosphate-buffered saline (PBS; 10 mm Na2HPO4, 0.15 mm NaCl, pH 7.3). Twenty-four hours prior to transfection, cells were seeded at a density of 300 000 in 60 mm culture dishes. Cells were transfected using a modified calcium phosphate precipitation method. Plasmids encoding GABAR subunits were mixed in a 1:1:1 ratio using 3-4 μg of each along with 2 μg of pHook-1 (Invitrogen, San Diego, CA, USA), which encodes the cell surface antibody sFv. Following addition of DNA, cells were incubated for 4-5 h in 3 % CO2 and shocked for 30 s with 15 % glycerol in BBS buffer (50 mm Bes, 280 mm NaCl, 1.5 mm Na2HPO4). The following day, transfected cells were selected and concentrated using Capture-Tec beads (magnetic beads coated with hapten; Invitrogen). Selected cells were replated on 35 mm culture dishes and used for recording about 24 h later.
Electrophysiological recording solutions and techniques
For whole-cell recording, the external bath solution consisted of (mm): 142 NaCl, 8.1 KCl, 6 MgCl2, 1 CaCl2, 10 glucose and 10 Hepes at pH 7.4 and osmolality between 311 and 325 mosmol kg−1. The concentration of Mg2+ in the bath solution did not change the GABAR IC50 values for Zn2+. The IC50 values for the wild-type receptors α1β3δ and α1β3γ2L were similar to those reported previously in the literature (Gingrich & Burkat, 1998; Krishek et al. 1998). Recording electrodes were filled with an internal solution of (mm): 153 KCl, 1 MgCl2, 5 K-EGTA, 10 Hepes and 2 Mg-ATP at pH 7.3 and osmolality adjusted to 295-300 mosmol kg−1. These solutions provided a chloride equilibrium potential near 0 mV. Patch pipettes were pulled from microhaematocrit tubes made of soda-lime glass (i.d. = 1.1-1.2 mm, o.d. = 1.3-1.4 mm; Fisher Scientific, Pittsburgh, PA, USA) on a P-87 Flaming Brown puller (Sutter Instrument Co., San Rafael, CA, USA). Pipettes had resistances of 5-10 MΩ and were coated with polystyrene Q-Dope (GC Electronics, Rockfield, IL, USA) before use. Currents were recorded with an Axoclamp 200A patch clamp amplifier (Axon Instruments, Foster City, CA, USA), DigiData 1200 interface (Axon Instruments) and Zenith Pentium computer as well as on Beta videotape. Series resistance was compensated by 90 %. GABA and ZnCl2 were prepared as stock solutions of 100 mm in water. All working solutions were prepared on the day of the experiment by diluting stock solutions in external solution. Drugs were applied to cells using a modified U-tube delivery system with a 10-90 % rise time of 70-150 ms (Greenfield & Macdonald, 1996). GABA-induced currents were recorded at a holding potential of -75 mV. All experiments were performed at room temperature.
Analysis of whole-cell currents
Whole-cell currents were analysed off-line using Axoscope (Axon Instruments) and Prism software (GraphPad, San Diego, CA, USA). Normalized concentration-response data for the different isoforms were fitted with a four-parameter logistic equation:
where X is the concentration of drug, nH represents the Hill coefficient and I represents current expressed as a percentage of the maximum current elicited by saturating concentrations of GABA for each cell or in the case of Zn2+, as a percentage of the current elicited by GABA alone (Imax). Data in the figures were derived from a single fit of averaged responses from all cells. Data in Tables 1 and 2 were derived from fits of individual cells and are reported as means ±s.e.m. Statistical tests were performed by one-way ANOVA and Newman-Keuls multiple comparisons tests where P < 0.05. Results of significance tests are presented in the relevant table legends.
Expression of αβδ-γ2L chimeric GABARs
Although the δ and γ2L subunit families share only 35 % homology with each other (Shivers et al. 1989), the four chimeras we constructed produced functional channels when coexpressed with wild-type α and β subunits. The GABA EC50 values for all of the chimeric isoforms were distinct from those observed for binary αβ receptors (≈1 μm; Angelotti et al. 1993), indicating that the chimeric subunits were incorporated and influenced channel properties.
Design of δ-γ2L chimeric subunits
To determine which domains of the γ2L subunit were involved in conferring low Zn2+ sensitivity to ternary GABARs, four chimeric subunits were created by progressively replacing the wild-type γ2L subunit sequence from the N-terminus with wild-type δ sequence. Figure 1 depicts the putative membrane topology for each of the δ-γ2L chimeras including the large N-terminus, the four transmembrane segments (M1-M4) and their interconnecting loops, and the extracellular C-terminus. These chimeras divided the GABAR subunit sequence into five sections: (1) the N-terminus, (2) the M1 segment, (3) the M1-M2 loop and the M2 segment, (4) the M2-M3 loop and (5) the M3 and M4 segments, the M3-M4 loop and the C-terminus.
GABA sensitivity of GABARs containing wild-type and chimeric subunits
Wild-type and chimeric forms of the δ and γ2L subunits produced functional GABAR channels when coexpressed with wild-type α1 and β3 subunits in L929 fibroblasts. Cells were voltage clamped at -75 mV, and whole-cell currents were elicited by application of GABA. Comparison of concentration-response curves for GABARs containing wild-type subunits indicated that α1β3γ2L receptors had a higher EC50 for GABA than α1β3δ receptors (Fisher & Macdonald, 1997; P < 0.001; Table 1).
|Receptor isoform||EC50 (μM)||Hill slope||Peak current (pA)||n|
|αβδ||4.4 ± 1.0||1.4 ± 0.1||136.0 ± 36.7||5|
|αβγ||15.5 ± 3.6||1.5 ± 0.1||472.9 ± 145.9||5|
|αβδ-γ(M1e)||2.8 ± 0.9||1.5 ± 0.2||506.4 ± 142.9||5|
|αβδ-γ(M1i)||25.4 ± 4.8||1.1 ± 0.07||362.0 ± 250.9||3|
|αβδ-γ(M2e)||23.9 ± 3.4||0.9 ± 0.02||219.3 ± 50.8||5|
|αβδ-γ(M3e)||3.5 ± 0.4||1.3 ± 0.1||158.7 ± 62.8||6|
|αβδ(MVS STI)||5.1 ± 1.1||1.2 ± 0.1||191.7 ± 44.0||3|
|αβγ(STI MVS)||19.1 ± 5.0||1.4 ± 0.3||2164 ± 1203||3|
|αβδ(M278S)||4.3 ± 1.7||1.2 ± 0.1||366.6 ± 129.2||3|
|αβδ(V279T)||3.0 ± 0.8||1.6 ± 0.2||137.4 ± 110.5||3|
|αβδ(S280I)||2.5 ± 0.3||1.0 ± 0.1||328.5 ± 134.0||3|
|αβγ(I282S)||13.8 ± 1.4||1.7 ± 0.1||408.0 ± 115.6||4|
|αβδ(S283K)||3.0 ± 0.5||1.5 ± 0.1||197.2 ± 50.8||4|
|αβγ(K285S)||13.2 ± 1.4||1.6 ± 0.2||636.6 ± 100.5||3|
Currents from GABARs containing chimeric subunits were differentially sensitive to GABA (Fig. 2A). Two of the isoforms containing chimeric δ-γ2L subunits, α1β3δ-γ2L(M3e) and α1β3δ-γ2L(M1e), had GABA EC50 values similar to that of GABARs containing the wild-type δ subunit (Fig. 2B, Table 1). The other two chimeric receptors, α1β3δ-γ2L(M2e) and α1β3δ-γ2L(M1i), had higher GABA EC50 values that were similar to that of GABARs containing the wild-type γ2L subunit (Fig. 2B, Table 1). These results suggest that the juxtaposition of the γ subunit transmembrane and/or loop sequences with the δ subunit N-terminal extracellular sequence could influence GABA binding and/or channel gating such that a higher GABA EC50 value was observed.
Zn2+ sensitivity of GABARs containing wild-type and chimeric subunits
Currents evoked from wild-type α1β3δ and α1β3γ2L receptors by EC50 concentrations of GABA were differentially inhibited by coapplication of 10 μm Zn2+ (Fig. 3A). The Zn2+ IC50 of α1β3γ2L receptors was significantly higher than that of α1β3δ receptors (Fig. 3B, P < 0.001; Table 2).
|Receptor isoform||IC50 (μM)||Hill slope||n|
|αβδ||8.6 ± 1.8||−0.7 ± 0.1||5|
|αβγ||307.6 ± 68.7||−0.9 ± 0.2||3|
|αβδ-γ(M1e)||23.9 ± 2.2||−0.9 ± 0.1||4|
|αβδ-γ(M1i)||19.8 ± 4.6||−0.9 ± 0.06||3|
|αβδ-γ(M2e)||11.3 ± 1.0||−0.8 ± 0.05||4|
|αβδ-γ(M3e)||5.4 ± 1.2||−0.9 ± 0.1||5|
|αβδ(MVS STI)||15.8 ± 2.9||−0.9 ± 0.03||4|
|αβγ(STI MVS)||55.1 ± 6.0||−0.6 ± 0.03||3|
|αβδ(M278S)||7.2 ± 3.4||−0.9 ± 0.1||3|
|αβδ(V279T)||7.0 ± 1.9||−1.1 ± 0.3||2|
|αβδ(S280I)||5.3 ± 1.2||−1.1 ± 0.7||3|
|αβγ(I282S)||345.7 ± 154||−0.9 ± 0.1||3|
|αβγ(S283K)||9.8 ± 2.3||−0.9 ± 0.1||5|
|αβγ(K285S)||178.2 ± 26.7||−0.8 ± 0.05||4|
Currents evoked from chimeric GABARs by EC50 concentrations of GABA were also differentially inhibited by coapplication of 10 μm Zn2+ (Fig. 4A). Progressive replacement of the γ2L subunit from the N-terminus with the δ subunit sequence produced a progressive decrease in Zn2+ IC50 (Fig. 4B). The α1β3δ-γ2L(M1e) isoform had an average Zn2+ IC50 that was substantially lower than that of α1β3γ2L receptors (P < 0.001; Table 2). Replacement of the wild-type γ2L subunit with this chimeric subunit accounted for 63 % of the log difference between Zn2+ IC50 values of α1β3γ2L and α1β3δ-γ2L(M3e) receptors, suggesting that the N-terminal extracellular domain contained a critical determinant. Replacement of additional γ2L subunit sequence (the M1 segment) with the δ subunit sequence produced little change in Zn2+ sensitivity (Fig. 4B); the α1β3δ-γ2L(M1i) isoform had an average Zn2+ IC50 that was unchanged from that of the α1β3δ-γ2L(M1e) isoform (Table 2), suggesting that the M1 segment was not involved in conferring low Zn2+ sensitivity. Extending the δ subunit sequence through the γ2L subunit M1-M2 loop and the M2 segment, however, again decreased Zn2+ IC50 (Fig. 4B). The α1β3δ-γ2L(M2e) isoform had an average Zn2+ IC50 that was substantially lower than that of the α1β3δ-γ2L(M1e) isoform (Table 2). This chimeric subunit accounted for an additional 19 % of the log difference between Zn2+ IC50 values of α1β3γ2L and α1β3δ-γ2L(M3e) receptors, suggesting that the M1-M2 loop and M2 segment contained another determinant of low Zn2+ sensitivity. Extending the δ subunit sequence through the γ2L subunit M2-M3 loop also decreased Zn2+ IC50 (Fig. 4B). The α1β3δ-γ2L(M3e) isoform had an average Zn2+ IC50 that was lower than that of the α1β3δ-γ2L(M2e) isoform (Table 2). This chimeric subunit accounted for an additional 18 % of the log difference between Zn2+ IC50 values of the α1β3γ2L and α1β3δ-γ2L(M3e) receptors, suggesting that the M2-M3 loop contained a third determinant of low Zn2+ sensitivity.
Point mutations in the outer vestibule
Of the three structural determinants identified by the δ-γ2L chimeras, we proceeded to further delineate two adjacent domains: (1) the M1-M2 loop and the M2 segment and (2) the M2-M3 loop. We chose to focus on these adjacent regions since previous studies of GABARs had identified functional domains for Zn2+ in the M2 segment of β subunits and in the M2-M3 loop of α subunits, suggesting the presence of homologous domains in γ2L and δ subunits (Wooltorton et al. 1997; Fisher & Macdonald, 1998; Horenstein & Akabas, 1998). We targeted the extracellular end of the M2 segment and the proximal end of the M2-M3 extracellular loop, subunit regions putatively associated with the outer vestibule of the channel. Specific amino acid residues in the wild-type δ and γ2L subunits were targeted for site-directed mutagenesis based on studies of Zn2+ sensitivity involving other GABAR subunits.
The M2 segment.
The putative channel-lining M2 segment is highly conserved across all GABAR subunits. There are, however, a few sequence differences among subunit families. Between the δ and γ2L subunits, there are differences at four positions in the M2 sequence (Fig. 5A). Three differences occur in a triplet of amino acids at the extracellular end of the M2 segment. Residues in the α1 subunit that are homologous to the second and third position of the triplet, I270 and S271, have been shown to be water accessible in αβγ receptors (Xu & Akabas, 1996). Residues in the β1 and β3 subunits homologous to the third position of the triplet, H292 and H267, respectively, were shown to be major determinants of Zn2+ inhibition in β homomers and binary αβ receptors and also appeared to be water accessible (Wooltorton et al. 1997; Horenstein & Akabas, 1998). Therefore, we focused on the extracellular triplet of residues and made the mutations M278S, V279T, S280I (MVS STI) and S280M, T281V, I282S (STI MVS) in the δ and γ2L subunits, respectively.
Coexpression of δ or γ2L M2 mutant subunits with wild-type α1 and β3 subunits resulted in functional GABARs. The GABA EC50 values of the α1β3δ(MVS STI) and α1β3γ2L(STI MVS) isoforms were similar to those of their respective wild-type receptors (Table 1), suggesting that the triple mutations in the M2 segment did not affect GABA sensitivity.
The Zn2+ IC50 values of GABARs containing either of the M2 mutants differed from those of GABARs containing wild-type subunits (Fig. 5B and C). The α1β3δ(MVS STI) isoform had an average Zn2+ IC50 that was substantially higher than that of the α1β3δ isoform (Table 2). This mutant subunit accounted for 17 % of the log difference between Zn2+ IC50 values of α1β3γ2L and α1β3δ receptors. The α1β3γ2L(STI MVS) isoform had an average Zn2+ IC50 that was substantially lower than that of the α1β3γ2L isoform (P < 0.001; Table 2). This mutant subunit accounted for 48 % of the log difference between Zn2+ IC50 values of α1β3γ2L and α1β3δ receptors. These results indicated that the extracellular end of the M2 segment plays an important role in determining the Zn2+ sensitivity of ternary GABARs. Introduction of the γ2L triplet into the δ subunit background decreased Zn2+ sensitivity whereas removal of the triplet from the γ2L subunit background increased Zn2+ sensitivity. The effects of these mutations, however, were not equivalent for the γ2L and δ subunits (see Discussion).
To determine whether the influence of the M2 triplet on Zn2+ sensitivity was dependent on only one of the residues, we made a set of point mutations. Mutagenesis at the third position of the triplet in α and β subunits has been shown to influence the Zn2+ sensitivity of binary αβ receptors (Wooltorton et al. 1997; Horenstein & Akabas, 1998). Therefore, point mutations were made at this position in the δ and γ2L subunits at S280 and I282, respectively. In addition, the contributions of the first and second residues of the triplet were tested by making the point mutations M278S and V279T in the δ subunit.
Coexpression of the δ or γ2L M2 triplet mutants with wild-type α1 and β3 subunits resulted in functional GABARs. The GABA EC50 values of the α1β3δ(M278S), α1β3δ(V279T), α1β3δ(S280I) and α1β3γ2L(I282S) isoforms were similar to those of their respective wild-type receptors (Table 1), suggesting that these point mutations in the M2 triplet had little effect on GABA sensitivity.
The Zn2+ IC50 values of GABARs containing any of the M2 triplet mutants were similar to those of GABARs containing wild-type subunits (Fig. 6A and B). The α1β3δ(M278S), α1β3δ(V279T) and α1β3δ(S280I) isoforms had average IC50 values that were similar to that of α1β3δ receptors (Table 2). The α1β3γ2L(I282S) isoform had an average IC50 that was similar to that of α1β3γ2L receptors (Table 2). These results indicated that mutations of the individual residues in the M2 triplet did not replicate the effect of the triple mutation on the Zn2+ sensitivity of ternary GABARs.
The M2-M3 loop.
The M2-M3 loop is involved in modulating the Zn2+ sensitivity of αβγ receptors containing the α1 or α6 subunit subtypes (Fisher & Macdonald, 1998). α6β3γ2L receptors are more sensitive to Zn2+ than α1β3γ2L receptors. Residue H273 of the α6 subunit confers increased Zn2+ sensitivity to α6 subunit-containing receptors whereas the homologous residue in the α1 subunit, N274, confers decreased Zn2+ sensitivity to α1 subunit-containing receptors. In α1β1γ2 receptors, α1(N274) has been shown to be water accessible (Xu & Akabas, 1996), and sequence alignment indicates that the position occupied by α6(H273) differs between the δ and γ subunit families (Fig. 7A). We hypothesized that the positively charged K285 in the γ2L subunit might electrostatically repulse Zn2+ from the outer mouth of the channel whereas the polar, uncharged S283 residue in the δ subunit might interact with Zn2+ (Karlin & Zhu, 1997) and thereby stabilize the cation in the channel. It has been previously reported that replacement of K285 with an alanine residue did not change the Zn2+ insensitivity of ternary receptors (Smart et al. 1994). To assess the role of this M2-M3 residue in determining Zn2+ sensitivity, the point mutations S283K and K285S were made in the δ and γ2L subunits, respectively.
Coexpression of δ or γ2L M2-M3 mutant subunits with wild-type α1 and β3 subunits resulted in functional GABARs. The GABA EC50 values of the two mutant subunit-containing isoforms, α1β3δ(S283K) and α1β3γ2L(K285S), were similar to those of their respective wild-type receptors (Table 1). These results indicated that these point mutations in the M2-M3 loop had little effect on GABA sensitivity.
Mutations in the M2-M3 loop of δ or γ2L subunits had different effects on the Zn2+ sensitivities of ternary GABARs (Fig. 7B and C). The α1β3δ(S283K) isoform had an average IC50 that was slightly higher than that of α1β3δ receptors (Table 2), suggesting that substitution of a lysine residue at S283 in the wild-type δ background had minimal effect on Zn2+ sensitivity. However, the α1β3γ2L(K285S) isoform had an average IC50 that was lower than that of α1β3γ2L receptors (Table 2). This mutant subunit resulted in a reduction in Zn2+ sensitivity that was 15 % of the log difference between Zn2+ IC50 values of α1β3γ2L and α1β3δ receptors, suggesting that substitution of a serine residue at K285 in the wild-type γ2L subunit caused an increase in Zn2+ sensitivity. Similar to the M2 triple mutants, these point mutations in the M2-M3 loop of δ and γ2L subunits did not have equivalent effects on Zn2+ sensitivity, suggesting that the subunit context was an added factor (see Discussion).
For the purposes of this discussion, we defined a functional domain as one or more amino acids that determined the effectiveness of a modulatory compound (e.g. Zn2+). The functional domain could represent the actual binding site for the compound, a transduction device between binding and drug effect, or a structural feature that could indirectly influence binding and/or transduction. In our study, the functional domain was defined by functional differences (e.g. IC50) that were introduced by chimera construction and mutagenesis. Using this approach we identified novel structural determinants of low Zn2+ sensitivity for γ2L subunit-containing ternary GABARs (Fig. 8). One functional domain in the γ2L subunit for low Zn2+ sensitivity was localized to a subunit region forming the outer vestibule of the channel and was composed of residues in the M2 segment and the M2-M3 loop. The other functional domain in the γ2L subunit for low Zn2+ sensitivity was localized to the N-terminal extracellular region. Together, these two functional domains appear to form the basis for differences in Zn2+ sensitivity between αβδ and αβγ receptors.
Role of the outer vestibule in regulating Zn2+ sensitivity
One functional domain in the γ2L subunit conferring low Zn2+ sensitivity was localized to the region of the outer vestibule of the channel. It was composed of the extracellular end of the M2 segment and the proximal end of the M2-M3 loop. The γ2L subunit determinant in the M2 segment for reduced Zn2+ sensitivity consisted of three amino acid residues rather than a single one as was found for the β subunits (Wooltorton et al. 1997; Horenstein & Akabas, 1998). Individual substitutions within the triplet based on γ2L and δ sequence differences did not change the Zn2+ sensitivity of ternary receptors. As in the α1 and α6 subunit subtypes (Fisher & Macdonald, 1998), the γ2L subunit M2-M3 loop was also found to make a contribution to Zn2+ sensitivity. Although these two γ2L subunit regions were identified as separate determinants, it is likely that they form a single functional domain that was inadvertently divided by the design of our chimeric subunits.
It is likely that the M2 triplet of γ2L and δ subunits is not directly involved in Zn2+ binding. In the β subunit family, the homologous residue to the first position in the triplet has been shown to influence sensitivity to a variety of modulatory compounds. Replacement of the β1 subunit subtype with either the β2 or β3 subunit subtype confers greater receptor sensitivity to furosemide inhibition as well as to loreclezole, ethanol, enflurane, etomidate and β-carboline potentiation (Wingrove et al. 1994; Stevenson et al. 1995; Belelli et al. 1997; Mihic et al. 1997; Thompson et al. 1999). The differences in drug sensitivities have been attributed to N265 (rat β3 subunit) and the homologous S265 (rat β1 subunit). These observations suggest that the first position in the M2 segment of the γ2L and δ GABAR subunits has a role in transduction of modulator effects rather than in direct binding. Transduction of Zn2+ binding may be a role subserved by the second and third positions of the M2 triplet as well as by γ2L(K285) in the M2-M3 loop. Residues γ2(T281) and γ2(I282) have recently been identified as two of the three transduction elements required to couple benzodiazepine binding to GABA current potentiation (Boileau & Czajkowski, 1999). The third element that was identified is γ2(S291), which is six amino acids C-terminal to γ2(K285) (Boileau & Czajkowski, 1999).
The third position of the M2 triplet has been shown to largely determine the Zn2+ sensitivity of β homomers and binary αβ receptors (Wooltorton et al. 1997; Horenstein & Akabas, 1998). Substitution of a serine (as in the δ subunit) or an isoleucine (as in the γ2L subunit) for the histidine in this β subunit position was found to decrease the Zn2+ sensitivity of α1β1 GABARs (Horenstein & Akabas, 1998). Substitution of a histidine at the homologous position of the α1 subunit (S271) was found to increase the Zn2+ sensitivity of α1β1 receptors (Horenstein & Akabas, 1998). Thus, this position appears to be available for Zn2+ binding in α and β subunits. In γ2L subunits, however, Zn2+ interaction does not seem plausible as substitution of a histidine at γ2(I282) was not found to increase the Zn2+ sensitivity of α1β1γ2 receptors (Horenstein & Akabas, 1998).
Role of the N-terminus in regulating Zn2+ sensitivity
Another determinant of the low Zn2+ sensitivity of γ2L subunits was localized to the N-terminus. This result was somewhat unexpected in the light of recent studies pointing to the extracellular portion of the M2 segment and the M2-M3 loop as important determinants of GABAR Zn2+ sensitivity (Wooltorton et al. 1997; Fisher & Macdonald, 1998; Horenstein & Akabas, 1998). However, it was shown that recombinant GABAC receptor ρ1 subunits possessed a determinant of Zn2+ sensitivity in the homologous extracellular region (Wang et al. 1995). A single residue, H156, was found to be critical for Zn2+ sensitivity. Residue H156 is homologous to a highly conserved asparagine, which is found among all GABAR subunit families except π. Interestingly, this residue is adjacent to H101 (rat sequence) in the α1 subunit, which is required for benzodiazepine sensitivity in ternary αβγ receptors (Smith & Olsen, 1995). The amino acid sequence in the vicinity of this residue is highlighted by two tryptophan residues (W69 and W94 in the rat GABAR α1 subunit) that are conserved in all members of the ligand-gated ion channel superfamily and in GABARs demarcate GABA and benzodiazepine binding regions (Smith & Olsen, 1995). In fact, mutation of the ρ1(H156) has also been found to influence the GABA sensitivity of ρ1 homomers (Kusama et al. 1994).
Functional domains determining the Zn2+ sensitivity of another member of the ligand-gated ion channel superfamily, the glycine receptors (GlyRs), have also been localized to the extracellular N-terminus. Mutagenesis based on chimeric subunit analysis has identified amino acid residues influencing potentiation and inhibition of human α1 homomers by Zn2+ (Lynch et al. 1998; Laube et al. 2000). Replacement of D80 resulted in a loss of potentiation and replacement of T112 resulted in a loss of inhibition. These two residues are located near the conserved tryptophan residues. Mutagenesis based on histidine targeting has also identified residues influencing potentiation and inhibition of GlyR currents by Zn2+ (Harvey et al. 1999). Mutation of H107 abolished Zn2+-mediated inhibition and mutation of H215 abolished Zn2+-mediated potentiation. Mutation of H109, however, abolished both inhibition and potentiation by Zn2+. Histidines 107 and 109 are proximal to the conserved tryptophan residues whereas H215 is not.
The complex nature of functional domains for Zn2+ in the N-terminal extracellular region of GlyRs suggests that such domains may also be complex in the homologous region of the GABAR γ2L and δ subunits. If the tertiary subunit of ternary GABARs participates in Zn2+ binding, then one might expect that the Zn2+ functional domains could be identified by looking for the presence or absence of common Zn2+ co-ordinating residues (H, C, D and E; Karlin & Zhu, 1997). This approach will not work, however, if the functional domains are involved in a role other than Zn2+ binding. Thus, chimeric subunit analysis would prove more fruitful.
Subunit context of mutations and shifts in Zn2+ sensitivity
The shifts in Zn2+ sensitivity conferred by our site-directed mutant subunits were asymmetric. The triple mutations in the M2 segment of the γ2L subunit induced a 48 % shift towards δ subunit-like sensitivity whereas in the δ subunit, the mutations induced only a 17 % shift towards γ2L subunit-like sensitivity. The point mutation in the M2-M3 loop of the γ2L subunit induced a 15 % shift towards δ subunit-like sensitivity whereas in the δ subunit, the mutation induced only a 4 % shift towards γ2L subunit-like sensitivity. Although it is tempting to presume that asymmetry is an indicator of irrelevance of a given residue or residues, we would argue that, in fact, it points to the importance of context (i.e. tertiary structure) for functional domains. In the ‘native’ context of the γ2L subunit, removal of the M2 triplet and the M2-M3 loop residue can be interpreted as a loss of elements necessary for low Zn2+ sensitivity. In the ‘non-native’ context of the δ subunit, introduction of the γ2L subunit M2 triplet and M2-M3 loop residue cannot, however, be interpreted as acquisition of these same elements because the appropriate subunit context is not available. Therefore, these mutations would not be expected to have the symmetrical effect on Zn2+ sensitivity in an otherwise ‘non-native’δ subunit context.
The importance of subunit context for functional domain properties also was demonstrated by the shifts of Zn2+ sensitivity induced by the δ-γ2L chimeric subunits. Removal of the γ2L subunit N-terminus (δ-γ2L(M1e)) resulted in a 63 % shift in sensitivity from γ2L subunit-like to δ subunit-like. The estimated contribution of the N-terminus to the low Zn2+ sensitivity of γ2L subunits based on the site-directed mutant data would indicate a shift of ≈40 % (see above). Sequential removal of the M2 triplet (δ-γ2L(M2e)) and the M2-M3 loop residue (δ-γ2L(M3e)) resulted in shifts of ≈20 % each from γ2L subunit-like to δ subunit-like. The sum of these shifts (≈40 %) is somewhat lower than the sum of those induced by the site-directed mutations involving these subunit segments where the subunit context included the γ2L N-terminus (≈60 %). Thus, the potency of the γ2L subunit outer vestibule in lowering Zn2+ sensitivity appears to be reduced in the context of the δ subunit N-terminus.
Although our study demonstrated that the N-terminus of the γ2L and δ GABAR subunits contributed to Zn2+ sensitivity, it would not be surprising if the homologous regions of α and β subunits were also found to play a role. The alignment of various N-terminal residues implicated in agonist and benzodiazepine binding among the different subunit families suggests that residues involved in Zn2+ inhibition are also aligned (Sigel & Buhr, 1997). Indeed, differences in the mechanism of Zn2+ antagonism (non-competitive, competitive, mixed) among various receptor isoforms (Legendre & Westbrook, 1991; Gingrich & Burkat, 1998; Krishek et al. 1998) might be explained by the availability of Zn2+ functional domains, not only near the channel mouth but also at subunit interfaces whereby agonist binding could be influenced.
- 1993). Assembly of GABAA receptor subunits: analysis of transient single cell expression utilizing a fluorescent substrate/marker gene combination. Journal of Neuroscience 13, 1418–1428. , & (
- 1997). The interaction of the general anesthetic etomidate with the γ-aminobutyric acid type A receptor is influenced by a single amino acid. Proceedings of the National Academy of Sciences of the USA 94, 11031–11036. , , , & (
- 1999). Identification of transduction elements for benzodiazepine modulation of the GABAA receptors: three residues are required for allosteric coupling. Journal of Neuroscience 19, 10213–10220. & (
- 1999). Theta, a novel gamma-aminobutyric acid type A receptor subunit. Proceedings of the National Academy of Sciences of the USA 96, 9891–9896. , , , , , , , , , , & (
- 1996). Properties of recombinant γ-aminobutyric acidA receptor isoforms containing the α5 subunit subtype. Molecular Pharmacology 50, 119–127. , , & (
- 1997). Insensitivity to anaesthetic agents conferred by a class of GABAA receptor subunit. Nature 385, 820–823. , , & (
- 1990). Functional and molecular distinction between recombinant rat GABAA receptor subtypes by Zn2+. Neuron 5, 781–788. , , , & (
- 1997). Single channel properties of recombinant GABAA receptors containing γ2 or δ subtypes expressed with α1 and β3 subtypes in mouse L929 cells. Journal of Physiology 505, 283–297. & (
- 1998). The role of an α subtype M2-M3 His in regulating inhibition of GABAA receptor current by zinc and other divalent cations. Journal of Neuroscience 18, 2944–2953. & (
- 1998). Zn2+ inhibition of recombinant GABAA receptors: an allosteric, state-dependent mechanism determined by the γ-subunit. Journal of Physiology 506, 609–625. & (
- 1996). Whole cell and single channel α1β1γ2S GABAA receptor currents elicited by a ‘multipuffer’ drug application device. Pflügers Archiv 432, 1080–1090. & (
- 1994). Zn2+: an endogenous modulator of ligand- and voltage-gated ion channels. Neuropharmacology 33, 935–952. & (
- 1999). Identification of an inhibitory Zn2+ binding site on the human glycine receptor α1 subunit. Journal of Physiology 520, 53–64. , , , & (
- 1997). A novel class of GABAA receptor subunit in tissues of the reproductive system. Journal of Biological Chemistry 272, 15346–15350. & (
- 1998). Location of a high affinity Zn2+ binding site in the channel of α1β1γ-aminobutyric acidA receptors. Molecular Pharmacology 53, 870–877. & (
- 1991). Regulation of the human enkephalin promoter by two isoforms of the catalytic subunit of cyclic adenosine 3′5′-monophosphate-dependent protein kinase. Molecular Endocrinology 5, 921–930. , , , & (
- 1997). Rapid seizure-induced reduction of benzodiazepine and Zn2+ sensitivity of hippocampal dentate granule cell GABAA receptors. Journal of Neuroscience 17, 7532–7540. & (
- 1997). Classification of mononuclear zinc metal sites in protein structures. Proceedings of the National Academy of Sciences of the USA 94, 14231–14236. & (
- 1998). Interaction of H+ and Zn2+ on recombinant and native rat neuronal GABAA receptors. Journal of Physiology 507, 639–652. , & (
- 1994). Mutagenesis of the GABA ρ1 receptor alters agonist affinity and channel gating. NeuroReport 5, 1209–1212. , , & (
- 2000). Kinetic and mutational analysis of Zn2+ modulation of recombinant human inhibitory glycine receptors. Journal of Physiology 522, 215–230. , & (
- 1991). Noncompetitive inhibition of γ-aminobutyric acidA channels by Zn. Molecular Pharmacology 39, 267–274. & (
- 1998). Zinc potentiation of the glycine receptor chloride channel is mediated by allosteric pathways. Journal of Neurochemistry 71, 2159–2168. , , & (
- 1994). GABAA receptor channels. Annual Review of Neuroscience 17, 569–602. & (
- 1996). Which GABAA-receptor subtypes really occur in the brain? Trends in Neurosciences 19, 139–143. & (
- 1997). Sites of alcohol and volatile anaesthetic action on GABAA and glycine receptors. Nature 389, 385–389. , , , , , , , , , , & (
- 1999). Spontaneous and γ-aminobutyric acid (GABA)-activated GABAA receptor channels formed by ε subunit-containing isoforms. Molecular Pharmacology 55, 168–178. , , & (
- 1999). Incorporation of the π subunit into functional GABAA receptor channels. Molecular Pharmacology 56, 598–610. & (
- 1996). Properties of putative cerebellar γ-aminobutyric acidA receptor isoforms. Molecular Pharmacology 49, 567–579. & (
- 1989). Two novel GABAA receptor subunits exist in distinct neuronal subpopulations. Neuron 3, 327–337. , , , , , & (
- 1997). The benzodiazepine binding site of GABAA receptors. Trends in Pharmacological Sciences 18, 425–429. & (
- 1991). GABAA receptors are differentially sensitive to zinc: dependence on subunit composition. British Journal of Pharmacology 99, 643–654. , , & (
- 1994). Modulation of inhibitory and excitatory amino acid receptor ion channels by zinc. Progress in Neurobiology 42, 393–441. , & (
- 1995). Functional domains of GABAA receptors. Trends in Pharmacological Sciences 16, 162–168. & (
- 1995). β-carboline γ-aminobutyric acidA receptor inverse agonists modulate γ-aminobutyric acid via the loreclezole binding site as well as the benzodiazepine site. Molecular Pharmacology 48, 965–969. , , & (
- 1999). Residues in transmembrane domains I and II determine γ-aminobutyric acid type A receptor subtype-selective antagonism by furosemide. Molecular Pharmacology 55, 993–999. , , , , & (
- 1995). A single histidine residue is essential for zinc inhibition of GABA ρ1 receptors. Journal of Neuroscience 15, 7684–7691. , , & (
- 1997). Neuronally restricted RNA splicing regulates the expression of a novel GABAA receptor subunit conferring atypical functional properties. Journal of Neuroscience 17, 5027–5037. , , , , , , , , , , , & (
- 1994) The modulatory action of loreclezole at the γ-aminobutyric acid type A receptor is determined by a single amino acid in the β2 and β3 subunit). Proceedings of the National Academy of Sciences of the USA 91. 4569–4573. , , & (
- 1997). Identification of a Zn2+ binding site on the murine GABAA receptor complex: dependence on the second transmembrane domain of β subunits. Journal of Physiology 505, 633–640. , , & (
- 1996). Identification of channel-lining residues in the M2 membrane-spanning segment of the GABAA receptor α1 subunit. Journal of General Physiology 107, 195–205. & (
We thank Sharon Baughman, Hyun Chung, Nadia Esmaeil, José S. Santos, Lisa M. Sharkey, Fang Sun and Jie Zhang for technical assistance and Matt T. Bianchi for critical reading of the manuscript. This work was supported by National Institutes of Health grants R01-NS33300 (R.L.M.) and 5T32-NS07222 (N.N.) and an Epilepsy Foundation of America fellowship (N.N.).