Classification of some of the γ-aminobutyric acidA receptors
GABAAreceptor subtype | Composition4-a | Characteristic properties |
---|---|---|
A1a | α1βn γ2 | High affinities and efficacies for classical BZ agonists,4-b CL 218872 (partial agonist), zolpidem, 2′-oxoquazepam4-c |
A1b | α1 βnγ3 | Same as for A1a, but ∼400-fold less sensitive to zolpidem and affinities are lower for 2′-oxoquazepam (in the same range as for A2, A3, and A5) and for classical BZ agonists4-d |
A1c | α1 βn γ1 | Same as for A1b, but flumazenil and Ro 15-4513 have low affinity and act, like β-carbolines (inverse agonists at A1a,b), as low-potency positive agonists4-e |
A2a | α2 βnγ2 | Similar to A3 for the ligands noted there, but other properties not yet defined |
A2c | α2 βnγ1 | BZ/ω agonists have 2- to 20-fold lower potency than on A2a, with FG8205 the most selective. The affinity of zolpidem is 5-fold greater on A2c but with very low efficacy. Insensitive to antagonists (e.g., flumazenil, CGS-8216, and Ro 15-4513). DMCM is an agonist4-f |
A3a | α3 βnγ2 | High affinities and potencies for classical BZ agonists and β-carbolines, similar to those of A1, but intermediate for zolpidem, for CL 218872 and 2′-oxoquazepam, ∼10-fold lower than on A1a4-g |
A4a | α4 βnγ2 | Insensitive to classical BZ agonists, zolpidem and many other BZ/ω agonists. Notable exceptions are bretazenil, CGS 20625, and some pyrazoloquinolines. Intermediate affinities for most β-carboline inverse agonists (∼10 times higher than at α6 βn γ2), but high affinity for DMCM. Flumazenil and Ro 15-4513 are agonists. The direct activation by propofol or pentobarbital is absent4-h |
A5a1 | α5β1/3 γ2 | A5: High affinity for classical benzodiazepine agonists but insensitive to imidazopyridines. Intermediate affinity for CL218872 and 2′-oxoquazepam. Certain 8-acetylenic imidazobenzodiazepines (inverse agonists) and L-655,708 (BZ/ω agonist) are highly selective |
A5b3 | α5β3 γ3 | Affinities of A5b3 are as for A5a1, but triazolam and β-carbolines are ∼10- to 30-fold weaker and CL 218872 is 10-fold stronger |
A5a2 | α5 β2γ2 | A5a1 differs from A5a2 in its outward rectification and its slower desensitization at depolarized voltages4-i |
A6a1 | α6 β1γ2 | Insensitive to all BZ/ω ligands except bretazenil and some other partial agonists; flumazenil and Ro 15-4513 become partial agonists and DMCM an antagonist (fig. 5) |
A6a2 | α6β2/3 γ2 | Same as for A6a1, but A6a2 is antagonized selectively by furosemide (see notes to table 1) |
A16a2 | α6α1 β2/3γ2 | Combines the binding sites of A1a and A6a24-j |
A0r | A0r: insensitive to all BZ/ω ligands, but also to bicuculline and pentobarbital. Not activated by isoguvacine | |
A0r1 | ρ1 | |
A0r2 | ρ2 | |
A0r12 | ρ1ρ2 | A0r12 (alone) has very low sensitivity to picrotoxin (in the rat) (for references: see Sections I.A. and IV.D.) |
A0r3 | ρ3 | |
A01, A02 | Insensitive to all BZ/ω ligands, not because of ρ, but (e.g.) α + β + δ or α + β + ε. Sensitive to bicuculline | |
A01 | α1 βn δ | Same as above, and highly sensitive to zinc |
A04 | α4βn δ | Generally similar to A01 |
A06 | α6βn δ | Same as for A01; in cerebellar granule cells only4-k |
A01e | α1 βn ε | Generally similar to A01 |
↵4-a This means that, e.g., the GABAA2a receptor has a pharmacology which mimics that of the co-expressed recombinants α2βnγ2, where n = 1–3 (in tests so far), unless distinctions are known because of the β isoform present; β1/3 means β1 or β3. The stoichiometry within the assembly is not implied. Both binding affinities and effects on GABA-evoked currents are considered in the pharmacologies compared in column 3. For ease of comparison here, in the second column the isoform numbers are not written as subscripts (which will be the correct general usage). Species differences in these receptors can occur within the mammals; where possible, data on the human receptors have been used here but, where not available, data on the rat are used. In general, the rule of preference, failing available human receptor pharmacology, will be rat, then other mammals, then birds.
↵4-b Classical BZ agonists are diazepam, flunitrazepam, clonazepam and other BZs of similar activity. Except where noted otherwise, flumazenil is an antagonist and Ro 15-4513 is a partial inverse agonist, both with high affinity.
↵4-c Likewise for most inverse agonists at the BZ site, for example β-carboline, ethylcarboxylate, and Ro 19-4603.
↵4-d Herb et al. (1992); Luddens et al.(1994); Hadingham et al. (1995).
↵4-e Ymer et al. (1990); Puia (1991); Giusti et al. (1993).
↵4-f Wisden and Seeburg (1992); Hadingham et al.(1993); Wafford et al. (1993). Apart from possible locations on some brain neurons, A2c is the subtype on cerebellar Bergmann glia and on α cells of the pancreas.
↵4-g For example, “high affinity” for zolpidem would cover reported Ki values (for the rat receptors), for α1 β(1–3) γ2 of 19 to 30 nM and “intermediate affinity” would cover 688 nM Ki, or 650 nM EC50 in potentiation of the GABA-evoked current (Faure-Halley et al., 1993; Lüddens et al., 1994).
↵4-h Yang et al. (1995); Huh et al. (1996);Knoflach et al. (1996); Scholze et al. (1996);Wafford et al. (1996).
↵4-i A5: Pritchett and Seeburg (1990); Faure-Halley et al. (1993); Hadingham et al. (1993). A5a1 (which, in the γ2L form, resembles a native receptor in CA1 pyramidal neurons): Burgard et al. (1996). A5b3: Luddens et al. (1994); Hadingham et al. (1995). A5a2: Burgardet al. (1996). 8-Acetylenic imidazobenzodiazepines and L-665,708 (tested so far on αn β2 γ2receptors): Liu et al. (1996); Quirk et al.(1996).
↵4-j A6: On A6a1/A6a2, partial agonists bretazenil and CGS-9895 and the antagonist flumazenil show 44 to 270 nM Kivalues (human, rat). A16a2 has high-affinity flumazenil and Ro 15-4513 binding sites, diazepam-sensitive. References: A6a1, A6a2,Lüddens et al. (1990); Korpi et al. (1995);Yang et al. (1995); Hadingham et al. (1996); Huhet al. (1996); Nusser et al. (1996); Waffordet al. (1996). A16a2: Pollard et al. (1995); Khanet al. (1996).
↵4-k A01: Saxena and Macdonald (1994); Ducic et al.(1995). A06: Quirk et al. (1995); Saxena and Macdonald (1996); Jones et al. (1997).