The Journal of Neuroscience, March 1992, f2(3): 1040-l 062
Brain. I. Telencephalon,
of 13 GABA, Receptor Subunit mRNAs in the Rat
Laboratory of Molecular Neuroendocrinology, Zentrum fijr Molekulare Biologie, University of Heidelberg, D-6900
D. J. Laurie, H. Monyer, and P. H. Seeburg
mRNAs, and these predominated
VI of neocortex
of the forebrain,
midal cells, the ar, and 6 transcripts
In thalamic nuclei, the only abundant
were those of a,, a4, &, and 6. In the medial
lamic nucleus, a,, a,, &, 6, and yJ mRNAs were the principal
GABA, receptor transcripts. The
colocalized and may encode
forms of the GABA, receptor.
tions support the hypothesis
sponsible for benzodiazepine
ceptors containing a*, a3, and a, contribute
the BZ II site. Based on significant
and y mRNAs, we suggest that in vivo, the CY~ subunit
tributes to GABA, receptors
by in situ hybridization.
confined (a6 mRNA was present
cells). Some neuronal
of subunit mRNAs,
patterns, and these areas probably
of GABA, receptors.
with -r2 mRNA. The a,&
in olfactory bulb, globus
nigra pars reticulata,
of 13 GABA,
&-@,, y,--r3, 6) were determined
ranging from ubiquitous
receptor subunit en-
genes in adult
only in cerebellar
In many areas,
for a, and b2 mRNAs,
apparent for the (Y* and fi3
in areas such as amygdala
The a3 mRNA occurred
and in the reticular thalamic
with the exception of hippocampal
only a few
a5 and 8, mRNAs generally
I (BZ I) binding,
that lack BZ modulation.
GABA is the principal inhibitory transmitter in vertebrate brain.
GABA produces its inhibitory effect by interacting with two
and Jutta Rami for efficient secretarial
W.W. held an EMBO
7291, the Deutsche
Im Neuenheimer Feld 282, D-6900
Copyright 0 1992 Society for Neuroscience
Aug. 8, 1991; revised
Oct. 21, 1991; accepted Oct. 23, 1991.
Ulla Keller for expert
skills. D.J.L. was a recipient
fellowship awarded by the Royal
long-term fellowship. This work
Blr Forschung und Technologie,
should be addressed to W. Wisden,
Zentrum fur Molekulare
and dedicated technical
of a European
(SFB 3 17, B9), and the Fonds
364 AZ 231/
of Heidelberg, Biologie,
classes of molecules on the target cell: (1) GABA, receptors,
which are ligand-gated anion channels that exhibit a diverse and
clinically important pharmacology, being the locus of action for
barbiturates, benzodiazepines (BZs), and steroids, which all act
alosterically to modify the efficacy of GABA (Haefely and Pole,
1986; Lambert et al., 1987; Puia et al., 1990); ethanol also
appears to mediate some of its effects through this receptor
(Wafford et al., 1990, 199 1); and (2) GABA, receptors, which
are coupled to G-protein-mediated cellular responses (Bowery,
1989). These two GABA receptor classes act on different time
scales, often in the same synapse (Dutar and Nicoll, 1988).
Our knowledge regarding the molecular composition of the
GABA, receptor has increased considerably over recent years.
Once thought to be a single molecular species (Haring et al.,
1985) this receptor now presents a staggering molecular diver-
sity revealed by cDNA cloning of GABA, receptor subunits. In
keeping with other members of the ligand-gated ion channel
superfamily (Unwin, 1989; Cooper et al., 1991), the GABA,
receptor is probably assembled as a pentameric structure from
a number of possible subunit classes. The subunit stoichiometry
in any given GABA, receptor complex is unknown. In the ro-
dent there are currently six a-subunits ((Y,*~), three @-subunits
(p&3,), three y-subunits (7,-y,), and a &subunit (Olsen and
Tobin, 1990; Seeburg et al., 1990; Herb et al., 1992; Ltiddens
and Wisden, 1991; Wilson-Shaw et al., 1991). Additionally a
p-subunit cDNA has recently been isolated from human retina
(Cutting et al., 199 1). Molecular diversity rather than uniformity
probably accounts for the pharmacological heterogeneity of
GABA, receptors seen by numerous laboratories (Unnerstall et
al., 198 1; Young et al., 1981; Sieghart, 1989; Olsen et al., 1990).
Different subunit combinations confer disparate pharmacol-
ogies on GABA, receptors expressed from combinations of
cDNAs. For example, the y-subunit class is required to confer
a generally robust BZ responsiveness on any cul/3 subunit com-
bination (Pritchett et al., 1989b), such that a minimum require-
ment for conventional GABA, receptor pharmacology would
be an LYJ~~T~ combination (where x is any variant). However,
studies on alxpXyz receptors reveal that it is the members of the
e-subunit class that dictate which type of BZ ligand binds to,
and allosterically modulates, the receptor complex (Pritchett et
al., 1989a,b; Ltiddens et al., 1990; Pritchett and Seeburg, 1990;
Seeburg et al., 1990; Ltiddens and Wisden, 1991). The a,-con-
taining complexes exhibit high affinity for CL 218-872 and
/3-carbolines, whereas the complexes containing cu,, LYE, and LYE
show lower affinity for these compounds (Pritchett et al., 1989a;
Pritchett and Seeburg, 1990). However, all display high affinity
for the BZ antagonist Ro 15-1788 and, particularly, for the
The Journal of Neuroscience, March 1992, f2(3) 1041
combination with a &y2 subunit pair bind Ro 15-45 13 at a site
insensitive to diazepam (Ltiddens et al., 1990; Wisden et al.,
1991 b), and hence CY,&, and a&y, receptors do not bind BZ
agonists. The p-subunits have been reported to modulate current
amplitudes (Sigel et al., 1990) and appear to differ in binding
of GABA analogs and pentobarbital (Bureau and Olsen, 1990).
The role of the &subunit remains unclear, but this subunit forms
homomeric channel complexes gated by GABA (Shivers et al.,
1989). The p-subunit appears to be quantitatively
in rodent CNS, with its mRNA
Clearly, in order to proceed with
pharmacological analyses, it is important
combinations occur in vivo. Two approaches
ping the site of protein expression
sion using in situ hybridization.
approach, although providing useful information
of GABA, receptors, is severely hindered by
the difficulty of raising unique antibodies.
tibodies, for example, monoclonal
mAb 62-361 (Richards et al., 1986, 1987; de Blas et al., 1988),
react with both & and & subunits
al., 1990, 199 1). However, one recent report
peptide-derived antibodies has demonstrated
calization of immunoreactivity
units in rat brain (Zimprich et al., 199 1). The yz and &subunits
have also been detected immunohistochemically
specific antisera (Benke et al., 199 1 a,b).
The sites of gene expression
been partially mapped by in situ hybridization.
these are the mRNAs for (Y, (Sequier et al., 1988; Khrestchatisky
et al., 1989; Lolait et al., 1989; Hironaka
et al., 1990b; Seeburg et al., 1990; MacLennan
(Seeburg et al., 1990; MacLennan
199 1; Wisden et al., 199 la), (Y) (Seeburg et al., 1990; Persohn
et al., 1991; Wisden et al., 1991a), LYE (Wisden et al., 1991b), (Ye
(Khrestchatisky et al., 1989; MacLennan
a14 by these authors), LYE (Kato, 1990; Lilddens
(SCquier et al., 1988; Malherbe
1990; Zhang etal., 1990) p2 and & (Lolait et al., 1989; Seeburg
et al., 1990; Zhang et al., 1990), y, (Ymer et al., 1990), yz (Shivers
et al., 1989; Malherbe et al., 1990b; Persohn et al., 199 1; Wisden
et al., 1991a), and 6 (Shivers et al., 1989). In bovine brain, the
ffl, a2, and (Ye mRNAs
mRNAs (Siegel, 1988), and in chicken
(Bateson et al., 199 la), have also been studied.
However, no systematic comparison
studies have selectively focused on different
or different species, and second, the mRNA
assessed by oligonucleotides, cRNA,
ment probes. The combined results make a systematic
parison between data difficult. In addition,
extended cRNA or DNA probes
sults, since such probes may cross-hybridize
gene family members.
In this and the accompanying article (Laurie
have undertaken a systematic comparison
bution ofthe 13 currently known rat GABA,
by in situ hybridization using unique
specific for each subunit mRNA.
Ro 15-45 13. The cyq and (Ye subunits in
restricted to retina (Cutting et
to know which subunit
or tracing sites of gene expres-
about the cel-
(mAb) bd- 17 and
(Fuchs et al., 1988; Ewert et
specific for (Y,, o(~, and LYE sub-
of some of the subunits have
In rodent brain
et al., 1990; Malherbe
et al., 1991), (Y*
et al., 1991; Persohn et al.,
et al., 1991, termed
et al., 1990), p,
et al., 1990b; Seeburg et al.,
(Wisden et al., 1988, 1989a,b)
the (Y, mRNA
is currently possible. First,
brain regions and/
or DNA restriction
to closely related
may produce re-
et al., 1992), we
of the brain
For detection of GABA,
from the region encoding the divergent intracellular
tative transmembrane domains M3 and M4. The oligonucleotides
constructed complementary to rat cDNA encoding subunit residues as
follows: 01~, 342-356 (Khrestchatisky
stchatisky et al., 1991); (Ye, 361-375 (Malherbe et al., 1990a); (Ye, 15-30
ofthe signal peptide (Ymer et al., 1989a; Wisden et al., 1991b); (Ye, 355-
369 (Khrestchatisky et al., 1989; Malherbe
termed 01~ by us is termed LYE by Khrestchatisky et al., 1989); (Ye, 342-
356 (Lilddens et al., 1990); @,, 382-396 (Ymer et al., 1989b); &, 382-
396 (Ymer et al., 1989b); &, 380-394
354 (Ymer et al., 1990); yz, 338-352 (Shivers et al., 1989); y3, 343-358
(Herb et al., 1992; see also Wilson-Shaw
sequence); 6, 335-349 (Shivers et al., 1989).
The procedures used (Monyer et al., 199 1; Wisden et al., 199 1 c) were
a modification of those by Young et al. (1986). Probes were 3’ end
labeled using a 30: 1 molar ratio of &S-dATP
sham) to oligonucleotide, and terminal
(Boehringer Mannheim). Unincorporated
Bio-Span 6 chromatography columns (Bio-Rad).
were removed and frozen on dry ice. Sections (14 wrn) were cut on a
cryostat, mounted onto poly-L-lysine-coated
temperature. Sections were fixed in 4% paraformaldehyde,
phosphate-buffered saline, and dehydrated into 95% ethanol for storage
at 4°C until required. Prior to hybridization,
from ethanol and allowed to air dry. Labeled probe dissolved in hy-
bridization buffer (0.06 fmol, 1000 dpm/pl) was then applied to sections.
Hvbridization buffer contained 50% formamide/
~‘Nacl, 0.0 15 M Na-citrate)/lO% dextran sulfate. Hybridization
42°C overnight. Sections were washed to a final stringency of 1 x SSC
at 60°C before alcohol dehydration and exposure to Kodak XARS film.
Anatomy of sections and autoradiographs
atlas of Paxinos and Watson (1986), and for thalamus the monograph
of Jones (1985) was consulted. Signal specificity was assessed by use of
competition experiments in which radiolabeled
to sections in the presence of an excess (50-fold) unlabeled probe. This
resulted in blank autoradiographs.
reference to previous reports of the distribution
in rat performed by other laboratories or other methods (e.g., Northern
blot analysis): the distribution of (Y, (Northern blot and in situ hybrid-
ization<RNA probes, Khrestchatisky et al., 1989) 01~ (Northern
Khrestchatisky et al., 199 1; in situ hybridization-cRNA
Lennan et al., 199 l), (Ye (Northern blot and in situ hybridization-cRNA
probes, Khrestchatisky et al., 1989; MacLennan
em blot and in situ hybridizationeDNA
em blot, Garrett et al., 1990), & and /3, (Northern
hybridization-oligonucleotides, Lolait et al., 1989; Ymer et al., 19896
Zhang et al., 1990), y2 (in situ hybridization-cRNA
al., 1989; Malherbe et al., 1990b), and y (in situ hybridization-cDNA
probe, Shivers et al., 1989). Our results were in general agreement with
receptor subunit transcripts, 45base antisense
were synthesized, each of a unique sequence often taken
area between pu-
et al., 1989); 01~, 340-344 (Khre-
et al., 1990a; the subunit
(Ymer et al., 1989b); y,, 341-
et al., 1991, for the mouse
(1200 Ci/mmol; Amer-
nucleotides were removed by
slides, and dried at room
sections were removed
x SSC (1 x SSC:O. 15
was determined using the
probes were hybridized
Specificity was also confirmed by
of GABA, transcripts
et al., 199 l), 01~ (North-
probe, Kato, 1990) p, (North-
blot and in situ
probes, Shivers et
domains M3 and M4, all GABA, receptor subunits carry di-
vergent amino acid sequences (Seeburg et al., 1990). Therefore,
this segment is ideal for designing
probes that will not cross-hybridize to transcripts of related
genes. However, several subunit genes (yz and avian 6,) have
been recently shown to be differentially spliced in this region
(Whiting et al., 1990; Bateson et al., 199 1 b; Kofuji et al., 199 1;
Wafford et al., 199 1). These new findings mean that we have to
be cautious in evaluating whether our probes would be truly
selective (or nonselective) for any, as yet undiscovered,
products. For example, our y2 probe detects both known ver-
sions of the yz mRNA.
In situ hybridization was performed with subunit mRNA-
specific %-labeled oligonucleotide probes on various horizontal
and coronal sections through the rat brain in order to cover a
the predicted cytoplasmic loop between transmembrane
subunit-specific nucleic acid
Table 1. Distribution of q-ah, &-&, y,--y3, and 6 mRNAs of GABA, receptors in the CNS
a2 a3 a4
+++ 0 +
CA1 str. pyramidalis
CA3 str. pywnidak
DG granule cells
+ ++ 0
central amygdaloid I
med. amygdaloid n.
lateral amygdaloid n
bed nucleus s. t
Ventr. posterior n.
In situ hybridization
or not detectable,
YGlabeled oligonucleotide probes on serial sections were assessed as intense, + + f; strongly positive, + +; positive, +;
The Journal of Neuroscience, March 1992, C’(3) 1043
broad range of structures. The results compiled from the figures
and also unpublished data are summarized in Table 1. Results
for the olfactory bulb and cerebellum are described and dis-
cussed in detail in the accompanying article (Laurie et al., 1992).
In the neocortex and hippocampus,
present (Figs. 1-14). The (Ye mRNA
the cerebellum (Fig. l), and its expression will not be further
described here (see accompanying article, Laurie et al., 1992).
Considering first the a-subunit
cortex in a laminated pattern, with layers II/III
pressing higher levels than layer IV (Figs. 1, 5, 11). In contrast,
the c+ mRNA is most predominant
present in deeper layers (Figs. 1, 3, 5, 7, 9). The (Ye mRNA
occurs in a gradient reciprocal to that of (Ye, with layer VI ex-
pressing most of this transcript (Figs. 1, 5, 7, 9). The q mRNA
appears to be highest in layers II and III, although significant
levels are present in deeper layers (Figs. 1,3,7,9). The (Ye mRNA
is rare in cortex, but the expression pattern appears weakly
delineated in layer VI (Figs. 1, 3, 5, 7, 9).
For the p-subunit mRNAs in cortex (Figs. 2,4, 6,8, lo), that
of p, resembles (Ye mRNA in that it is present at overall uni-
formly low levels, but layer VI has slightly higher levels. The
& and 6, transcripts are present in similar amounts in the same
pattern of lamination, with layers II/III, V, and VI having higher
levels than layer IV (Figs. 2, 4, 6, 8, 10).
All three y-subunit genes are expressed in cortex (Figs. 2, 4,
6, 8, lo), although only the pattern of yZ expression appears
strongly laminated, similar to that of the LY,, &, and & mRNAs.
The y, mRNA is present at uniformly low levels throughout all
layers of the cortex. Interestingly, parts of the corpus callosum
appear to contain targets hybridizing with the y, probe, as wit-
nessed by the lack of a signal boundary between cortex and
caudate putamen for the y, autoradiographs
In addition, white matter tracts in the hippocampus also appear
to be weakly labeled. This effect seems specific because (1) two
independent y, oligonucleotides
the mRNA (Ymer et al., 1990) give the same result, (2) the
signal can be competed by competition
and (3) y, probes do not label white matter tracts in the cere-
bellum or olfactory bulb (Fig. 2; Laurie et al., 1992). No other
subunit mRNAs could be detected in corpus callosum, and the
demarcation between cortex and caudate-putamen
pronounced for these subunit mRNAs.
The d-subunit mRNA pattern resembles that of the (Ye and CX~
mRNAs, with layer II showing moderate levels (Figs. 3, 5, 7,
(cortex, hippocampal formation, septum,
12 of the 13 subunits are
is restricted completely to
class, (Y, mRNA is present in
and V/VI ex-
in layer II, although it is
(e.g., Figs. 2, 4, 6).
recognizing different parts of
with unlabeled probe,
The piriform cortex expresses every subunit mRNA, to a greater
or lesser extent, except (Ye (data not shown). The most abundant
transcripts are (Y,-(Y~, & &, y2, and 6 (Figs. 3-6). The autora-
diographic images obtained over this area are often extremely
Examining the hippocampal expression of the a-subunit genes,
the CQ mRNA is consistently the most abundant product and is
expressed at high levels in the CA 1, CA3, and dentate gyrus cell
layers (Figs. 1, 7, 9, 12). The 01, and q mRNAs
in all sectors of the hippocampus (Figs. 1, 7, 9, 11, 12). The LYE
mRNA occurs mainly in the dentate granule cells, with some
pyramidal cell expression also. The (Ye mRNA appears as abun-
dant as the (Y* transcript in the CA1 and CA3 areas, but is less
prominent in the dentate gyms (Figs. 1, 7, 9, 12). In fact, the
LYE mRNA appears to encode a predominantly
subunit, since it is virtually absent from most other areas of the
brain (see also Khrestchatisky et al., 1989).
Regarding P-subunit mRNAs,
CA3, and dentate gyrus cell layers (Figs. 2, 8, 10, 13), with p,
and ps being more abundant than & mRNA.
the 01~ mRNA, the p, mRNA is expressed at its highest levels
in the hippocampus but is rare elsewhere. All three y-subunit
mRNAs are detectable in the hippocampus (Figs. 2, 8, 10, 13)
with the yZ mRNA being most abundant. The ys mRNA is rather
rare, cortical levels being higher than those in the hippocampus.
The 6 mRNA is restricted to dentate granule cells at the reso-
lution of x-ray film autoradiographs
are also found
all three are present in CAl,
(Figs. 7, 9, 11, 12).
The tenia tecta, a structure
campus, expresses a number of subunit genes at very high levels
(Figs. 3, 4). These abundant mRNAs
No y-subunit or a-subunit mRNAs are detected in this structure.
embryologically related to hippo-
are CQ, q, q, /I,, and &.
The lateral septum contains a variety of subunit mRNAs,
most abundant of which are LY*, OIL, &, and y, transcripts
3,4). Some subunit mRNAs such as (Ye, y2, and ys are completely
absent from the lateral septum. The medial septal nucleus (data
not shown; Wisden et al., 199 1 b) and the nucleus of the diagonal
band contain very high levels of cy,, &, and y2 mRNAs
1; Figs. 5, 6). This is consistent with these two nuclei being
anatomically linked and containing the same cell types (Bleier
and Byne, 1985). In the ventral septum, a different profile of
subunit transcripts is found. The bed nucleus of the stria ter-
minalis contains very high levels of (Ye and y, transcripts
urating autoradiographic signals), moderate levels of a3, p,, and
ps mRNAs and low levels of LY,, q, &, and y2 mRNAs
1; Figs. 5, 6).
Basal ganglia: caudate-putamen,
In the caudate nucleus the most prevalent a-subunit mRNAs
are CQ and q (Figs. 1, 3, 5). However,
also reveal the presence of (Y, and C+ mRNAs.
was undetectable. The predominant P-subunit in the caudate is
/I, (Figs. 2, 4, 6) followed in order of abundance by & and p,
mRNAs. All three y-subunit mRNAs
in caudate, with the y3 mRNA
others (Figs. 2, 4, 6). The &subunit mRNA
pressed in the caudate nucleus (Figs. 3, 5, 11).
Levels of subunit transcripts
allel their respective levels in caudate. Subunit mRNAs
dant in the caudate are also abundant in the nucleus accumbens,
while those mRNAs rare in nucleus accumbens are also rare in
caudate. The most abundant accumbens transcripts
of Ly2, q, and p,. However, the y3 mRNA appears to be elevated
in the medial parts of the nucleus accumbens (Fig. 4).
In the globus pallidus, the major a-subunit mRNA is that of
(Y, (Fig. 5). There are also smaller but significant amounts of LYE
nucleus accumbens, and
longer exposure times
The LYE mRNA
are present at low levels
slightly elevated relative to the
is moderately ex-
in the nucleus accumbens par-
1044 Wisden et al. - GABA, Receptor mRNA Distribution
Distribution of GABA, receptor a-subunit mRNAs (q-0~~) in horizontal rat brain sections. See Appendix for abbreviations. Scale bar,
and a3 mRNAs in this area (Fig. 5). The & mRNA is the only
@ mRNA found in the globus pallidus (Fig. 6). Regarding the
y-subunit class, y, mRNA is marginally the more abundant
y-transcript in this region (Fig. 6). The 6 and y3 mRNAs are not
expressed in the globus pallidus (Figs. 5, 6).
In the subthalamic nucleus, the (Y, and & mRNAs are the most
significant GABA, receptor transcripts present (Figs. 9, 10). y2
mRNA is also in this area, but at lower levels (Fig. 10). Longer
exposure times also indicate the presence of ys mRNA (not
In the amygdaloid complex, CQ mRNA is found at very high
levels and is the predominant a-subunit mRNA in the medial
amygdaloid nucleus (Fig. 9), lateral amygdaloid nucleus (Fig.
7), and posterior medial cortical nucleus (not shown). All other
The Journal of Neuroscience, March 1992, Q(3) 1045
Figure 2. Distribution
for abbreviations. Scale bar, 4.4 mm.
of GABA, receptor B-subunit mRNAs (j3,-&) and r-subunit mRNAs (n-/J in horizontal rat brain sections. See Appendix
a-subunit mRNAs, except (Ye, are also present in these nuclei
but at differing degrees. The rarest is cr5 mRNA. Similarly, all
P-genes are expressed in these nuclei, with & generally being
the best expressed p-subunit mRNA (Figs. 8, 10).
The differential expression of the y-subunit genes in the amyg-
daloid nuclei is noteworthy. For example, y, mRNA is expressed
at striking levels in the medial amygdaloid nuclei (Fig. lo), an
area where there is relatively little of the y2 and yp mRNAs.
Indeed, this nucleus contains some of the highest amounts of
y, mRNA in the brain. However, yZ mRNA is the main rep-
resentative of the y-class in the general amygdaloid area (Figs.
8, 10). The y3 mRNA is present in diffuse, low levels throughout
the complex. The &transcript is absent from the amygdala (Figs.
Diencephalon (epithalamus, thalamus, and hypothalamus)
The medial habenula expresses significant quantities of CQ and
& mRNAs and some y, and y2 mRNAs (Figs. 7, 8). All other
subunit mRNAs are rare or undetectable.
1048 Wisden et al. l GABA, Receptor mRNA Distribution
Figure 3. Distribution
of (Y,-cY~ and 6 GABA, receptor subunit mRNAs in coronal sections at the level of caudate nucleus and nucleus accumbens.
for abbreviations. Scale bar, 2.8 mm.
The IX, and LY., transcripts are the most prominent a-subunit
mRNAs throughout the whole thalamus. On sections through
caudal portions of the thalamus, the dorsal lateral geniculate,
ventral lateral geniculate, and the ventral posterior nuclei are
clearly positive with the CY, and CQ probes (Figs. 7, 9, 11, 14) as
is the lateral dorsal complex (Fig. 1). The q and a3 mRNAs are
also present in the thalamus, but in more restricted subpopu-
lations (Figs. 7, 9, 14). The q mRNA was absent from all
thalamic nuclei examined (Figs. 7, 9).
Examining the thalamic distribution ofthe a-subunit mRNAs
in more detail, several features become apparent. The q mRNA
is very abundant in most nuclei of the thalamus, with some
exceptions. For example in the zona incerta, a part of the ventral
thalamus, (Y, mRNA is the only a-variant present (Fig. 9), while
in the reticular thalamus q mRNA is absent but q mRNA is
present (Figs. 7, 9). The CQ and (Ye mRNAs are absent from the
ventral posterior nucleus where the (Y, and aa mRNAs appear
(Figs. 7, 9). In very caudal parts of the thalamus, such as the
medial geniculate nucleus, the (Y, and CQ mRNAs predominate,
with a4 mRNA being the most abundant (Fig. 12). The CY* and
The Journal of Neuroscience, March 1992, 72(3)
Figure 4. Distribution of &+3, and y,-y3 GABA, receptor subunit mRNAs in coronal sections at the level of caudate nucleus and nucleus
aecumhens. See Appendix for abbreviations. Scale bar, 2.8 mm.
(Y, mRNAs appear to be restricted to predominantly midline
nuclei (e.g., paraventricular nuclei, rhomboid nuclei) and also
the central lateral nucleus, and seem to colocalize in these struc-
tures. Both the (Ye and (Y) mRNAs are absent from the medio-
dorsal nucleus (Fig. 7).
Certain areas of the thalamus show a surprising degree of
microheterogeneity with regard to a-subunit expression. This is
illustrated by the parafascicular thalamic nucleus (Fig. 14, PF),
which encircles the fasciculus retroflexus (fr) fiber tract (Jones,
1985; Paxinos and Watson, 1986). The LYE transcript is the high-
1048 W&den et al. - GABA, Receptor mRNA Distribution
Figure 5. Distribution
hypothalamic area. See Appendix for abbreviations. Scale bar, 2.8 mm.
of LY,-C+ and 6 GABA, receptor subunit mRNAs in coronal sections at the level of globus pallidus and medial preoptic
est overall in this nucleus and is more abundant in the lateral
and ventral parts than in the most medial part (Fig. 14). In
contrast, (Y, mRNA is largely restricted to the lateral portion
(Fig. 14 (Y,, arrowheads), and the CQ mRNA is present in the
most medial midline portion (Fig. 14, CY~, arrowhead). The (Ye
mRNA is present at diffuse low levels throughout the parafas-
cicular nucleus (Fig. 9).
A pronounced feature of @-subunit mRNA expression in thal-
amus is the diffuse low-level expression of B, and & mRNAs.
Levels for these mRNAs increase in midline nuclei, for example,
paraventricular, rhomboid, and central lateral nuclei. By con-
trast, p2 mRNA is ubiquitous in thalamus (Figs. 2, 8, 10, 13,
14), and its pattern resembles in detail that of (Y, mRNA, both
in spatial distribution and relative abundance.
All members of the y-subunit mRNA class are poorly ex-
pressed in the thalamus (Figs. 2, 8, 10, 13). Of these, the yz
mRNA is the most abundant, with the y, and y, probes giving
weaker, diffuse signals. The yS mRNA is present at moderate
The Journal of Neuroscience, March 1992, 12(3) 1049
Figure 6. Distribution of @,+3, and 7,~~ GABA, receptor subunit mRNAs in coronal sections at the level of globus pallidus and medial preoptic
hypothalamic area. See Appendix for abbreviations. Scale bar, 2.8 mm.
levels in the medial geniculate nucleus (Fig. 13), being more
abundant than y, and y2 mRNAs in this nucleus.
The b-subunit mRNA is observed in a broad range ofthalamic
nuclei and colocalizes with the a., mRNA in the medial genic-
ulate, ventral posterior, ventral lateral geniculate, and dorso-
lateral geniculate nuclei (Figs. 5, 7, 9, 11, 12). However, unlike
the a4 mRNA, 6 mRNA is absent from the parafascicular tha-
lamic nucleus (Fig. 9). The 6 mRNA also appears to be absent
from midline nuclei such as the paraventricular nucleus and
rhomboid nucleus, and is also absent from the reticular nucleus
(Fig. 7). However, it is found in the mediodorsal nucleus (Fig.
7). In this respect, b-gene expression is reciprocal to the thalamic
nuclei expressing the a2 and a, subunit genes. For example, the
autoradiographic signals for a2 and aj mRNAs result in the
formation of a trident-like pattern (comprising the centrolateral,
paraventricular, and rhomboid nuclei) in the middle of the thal-
1050 W&den et al. l GABA, Receptor mRNA Distribution
Figure 7. Distribution
Appendix for abbreviations.
of GABA, receptor or-subunit mRNAs (q-q
Scale bar, 3 mm.
and 6) in coronal sections of rat brain at the level of medial habenula. See
The Journal of Neuroscience, March 199% 733)
Figure 8. Distribution of GABA, receptor &subunit mRNAs (8,~&) and r-subunit mRNAs (y,-y3) in coronal sections at the level of medial
habenula. gee Appendix for abbreviations. Scale bar, 3 mm.
amus (Fig. 7), and the 6 mRNA distribution traces the negative
image of this pattern (Fig. 7).
In the hypothalamus, the most prominent mRNA is that of cx2
(Figs. 5, 7, 9, 14). This mRNA is abundantly present in the
medial preoptic, dorsomedial, ventromedial, and arcuate nuclei
and in the dorsal hypothalamic area. The CY,, oJ, and (Ye subunit
mRNAs are also found in these nuclei, but in lower amounts.
The q mRNA appears to be absent from the hypothalamus
Regarding B-subunits, the P3 mRNA predominates in the me-
1052 Wisden et al. - GABA, Receptor mRNA Distribution
Figure 9. Distribution
for abbreviations. Scale bar, 3 mm.
of q-a, and 6 GABA, receptor subunit mRNAs in coronal sections at the level of the parafascicular nucleus. See Appendix
The Journal of Neuroscience, March 1992, 7273) 1053
for abbreviations. Scale bar, 3 mm.
Distribution of/3,-& and 7,-y, GABA, receptor subunit mRNAs in coronal sections at level of the parafascicular nucleus. See Appendix
(Figs. 6,8, 10). A low amount of& mRNA is seen in the arcuate
and the ventromedial nuclei, whereas & mRNA is rare in all
hypothalamic nuclei examined. Considering the y-subunit class,
dorsomedial, ventromedial, and arcuate nuclei y, and y2 mRNAs are found in dorsomedial, ventromedial, and
arcuate nuclei (Figs. 8, 10). The y, mRNA conspicuously pre-
dominates in the medial preoptic area (Fig. 6). The &subunit
mRNA is undetectable in hypothalamus.
1054 Wisden et al. l GABA, Receptor mRNA Distribution
Comparison of the distribution of q, LX.,, &, and d-subunit mRNAs in horizontal sections. See Appendix for abbreviations. scale
The Journal of Neuroscience, March 1992, 12(3) 1055
Fimr Fe 12. Coronal sections at the level of the medial geniculate nucleus and substantia nigra illustrating differential patterns of q-a,
mkNAs. See Appendix for abbreviations. Scale bar, 1.6 mm.
Midbrain (colliculi, substantia nigra, red nucleus)
Throughout the general midbrain area, the most noticeable
mRNAs are (Y,, q, & Pa, and y2 (Figs. 12, 13). However, the
autoradiographic signals obtained with the (Y,, &, and to some
extent the yz probes are very punctate, suggesting expression in
large cells. The patterns obtained with the LYE and p3 probes are
more uniform and diffuse (Figs. 12, 13). Some subunit mRNAs
(LYE, j3,, and 6) are entirely absent from any midbrain structures
examined, whereas others are generally absent but very prom-
inent in certain nuclei, for example, (Ye in substantia nigra com-
pacta (Fig. 12).
The substantia nigra pars reticulata contains high levels of (Y,
(see also Hironaka et al., 1990) and p2 mRNAs with y1 and yZ
transcripts also present (Figs. 12, 13). The substantia nigra pars
compacta contains oj, (Ye, &, and y2 mRNAs (Figs. 12, 13).
The red nucleus (see also Hironaka et al., 1990, for LY,) expresses
the same GABA, receptor subunit genes as that of the substantia
n&a reticula@ that is, a,, &, and y2 mRNAs, with only bor-
derline to zero degrees of expression for the others (Figs. 12,
All levels of the superior colliculus contain cq, p2, and “/z mRNAs,
although deeper layers contain larger amounts (Figs. 12, 13). In
contrast, the q., q, and a)5 transcripts are largely found in the
superficial layer. The (Ye mRNA occurs mainly in the optic nerve
layer. The p, mRNA is restricted to the optic layer, and the &
1056 Wisden et al. l GABA, Receptor mRNA Distribution
Figure 13. Coronal sections at the level of the medial geniculate nucleuskubstantia
See Appendix for abbreviations. Scale bar, 1.6 mm.
nigra illustrating patterns of @,-j3, mRNAs and y,yJ mRNAs.
mRNA is more highly expressed throughout the superior col-
In the inferior colliculus (central nucleus), the main GABA,
receptor transcripts are (Y,, &, and yZ (Figs. 1, 2, 11).
patterns can be analyzed to deduce plausible in vivo subunit
combinations that may constitute molecularly and functionally
distinct receptor subtypes. In the following discussion, we list
such combinations and endeavor to match these suggested sub-
In this study we have documented the regional brain distribution
of 13 rat GABA, receptor subunit mRNAs. Their expression
types with previously published pharmacological characteristics
in both brain membranes and engineered expression systems.
To derive plausible combinations, brain regions in which a
limited subset of subunit genes was expressed were naturally
more amenable than regions with highly complex expression
patterns. For example, dentate granule cells pose an extreme
problem, since they seem to contain every subunit mRNA with
the exception of that encoding (Ye. Thus, either a large complexity
of GABA, receptors exists on one cell type or there are sub-
populations of granule cells, each expressing particular subsets
of receptors. Sophisticated multiple-labeling experiments with
antibodies will be required to address this problem. However,
there is evidence in favor of more than one GABA, receptor
on different parts of hippocampal neurons. GABA, receptors
on pyramidal cell soma differ from those on dendrites in terms
The Journal of Neuroscience, March 1992, L!?(3) 1057
Fi.gz rre 14. Differential distribution of CX,, LY?, ad, and B, mRNAs in the thalamus (enlargements from Figs. 9, 10). Arrowheads defined in text.
Appendix for abbreviations. Scale bar, 1.8 mm.-. -
of agonist preference (Alger and Nicoll, 1982; Nicoll and Dutar,
Similarly, the neocortex seems refractory
to the large repertoire of different cell populations.
there are certain shared regional patterns of subunit mRNAs
with regard to different laminae. For example, the cr,, Q, and 6
mRNAs appear higher in layers II and III, relative to other
laminae (Table 1). The (Ye, c+, and @, transcripts
layer VI relative to their abundance in other layers. Although
the a,, &, &, and yz mRNAs are in every layer, they appear
highest in layers II/III and V/VI.
significant. For the (Y,, (Ye, and (Y) subunits, the cortical poly-
peptide distribution correlates with the mRNA
prich et al., 199 l), and the same mRNA
bovine cortex (Wisden et al., 1988). Moreover,
groupings of mRNAs are consistent
brain regions (thalamus, colliculi, caudate nucleus). For example
(Ye and 6 mRNAs often codistribute,
terns look very similar, and the a,&y,
frequently (see below). This is suggestive of at least three cortical
receptor subtypes containing LYJ, Q3,, and (~,&y~. Other sub-
units, most obviously -r-variants,
part of these “cores.”
to analysis owing
are highest in
These correlations may be
pattern is observed in
with groupings in other
the (Ye and @, mRNA
would presumably also be a
Deduced GABA, receptor subunit combinations
c~,/3JyJ. The (Y, and & mRNAs
in the brain (Table 1). In addition, the -yz mRNA
calizes with this pair. In areas such as the central nucleus of the
are most widely codistributed
inferior colliculi and the red nucleus, only (Y,, &, and yz mRNAs
are found. This core combination
of the olfactory bulb and in cerebellar Purkinje cells (accom-
panying article, Laurie et al., 1992). In coexpression
ments, the cy,&y, receptor embodies the complete profile of
“classical” GABA, electrophysiological
1990; Verdoom et al., 1990). This subunit combination is also
supported by immunoprecipitation
However, it appears that other y-variants
ample, in the globus pallidus, both (~,&y, and cu,&y, combi-
nations may occur. Some regions, such as the zona incerta,
islands of Calleja, and the subthalamic
high levels of cy, and p2 mRNAs,
mRNAs present except for moderate levels of yz mRNA.
LYJ~&J, aJ,(rJ. Another frequently occurring combination
is that of CQ and & mRNAs. For example, in the nucleus ac-
cumbens, caudate nucleus, medial habenula, numerous amyg-
daloid nuclei, and in many hypothalamic
occurs in combination with various y-variant
spinal cord, the &&-pair rule also seems to hold, with probable
a&y, complexes occurring on motor neurons (Persohn et al.,
1991; Wisden et al., 1991a). The (Y* and & mRNAs
calize in the olfactory bulb granule cell layer (Laurie et al., 1992).
It is interesting to note that the (1~~ and & together with CQ and
B, are the most abundant hippocampal mRNAs.
bution and intensity of the LYE and p, probe hybridization
appear to be identical throughout
exception of the olfactory bulb), suggesting that they could well
is also found in mitral cells
responses (Sigel et al.,
studies (Benke et al., 199 la).
can be used. For ex-
nucleus, express very
with no other GABA, receptor
nuclei, the CQ& pair
mRNAs. In the
most of the brain (with the
1058 Widen et al. - GABA, Receptor mRNA Distribution
be partners. Such an q/I,-containing
ly in the hippocampus. The properties of both a,P,y, and cu,&y,
recombinant receptors have been studied in Xenopus oocytes
and transfected kidney cells (Sigel et al., 1990; Puia et al., 199 l),
with receptors containing the 8, subunit being less sensitive to
diazepam potentiation of GABA responses (Sigel et al., 1990).
cr,~u,/3+thalamic receptors. This set of mRNAs
pressed in overlapping areas of the thalamus. With the exception
of ys mRNA in the medial geniculate nucleus, y mRNAs
exiguous in thalamus. This data may be suggestive of an in vivo
receptor complex containing ayz, LYE, &, and &subunits
known stoichiometry. Such a receptor in the absence of a y-sub-
unit might be expected to bind GABA,
(Pritchett et al., 1989a,b; Shivers et al., 1989), and indeed, the
diencephalic distribution of high-affinity
‘HGABA, sites (Palacios et al., 198 1; Bowery et al., 1987; Olsen
et al., 1990) strikingly resembles the distribution of the CX~ mRNA.
For example, high-affinity 3H-muscimol
thalamus and rare in the hypothalamus.
assessed by 3H-flunitrazepam binding are fivefold less abundant
than 3H-muscimol sites (Olsen et al., 1990), suggesting that the
majority of thalamic GABA, receptors do not bind BZs (Un-
nerstall et al., 1981). The a-subunit is probably present in a
subset of the high-affinity muscimol binding receptors since it
is present in a more limited number of thalamic nuclei than the
cy,, q, and & mRNAs. Thus, it is possible that many thalamic
receptors may be a,cw.,& receptors. In certain parts of the thal-
amus, such as the reticular nucleus, cr3 mRNA
place” that of ad, a result consistent with immunocytochemical
studies using qspecific antibodies (Zimprich
With regard to cu,cu,&G receptors, the evidence for the occur-
rence of two different a-variants
tain. An antibody specific for the CX, subunit coprecipitates under
nondenaturing conditions other photolabeled subunits in ad-
dition to that of a, (tagged by either 3H-flunitrazepam
15-45 13), whereas only the labeled (Y, subunit is precipitated
under denaturing conditions (Ltiddens et al., 199 1). On the other
hand, other immunoprecipitation
gest that the (Y, and (Y* subunits are in largely distinct complexes
(Duggan and Stephenson, 1990), in agreement with our in situ
hybridization data. In Xenopus oocytes, addition of cr, or (Y~ to
cr,p,y, combinations made little difference to the receptor prop-
erties (Sigel et al., 1990).
receptor would occur main-
is highly ex-
ligands but not BZs
sites are prevalent in
Thalamic BZ sites as
appears to “re-
et al., 199 1).
in the same complex is uncer-
studies on bovine brain sug-
Correlation of mRNA levels with protein levels
Our inferences of subunit combinations from differential mRNA
distributions is critically dependent on the hypothesis that mRNA
levels reflect protein levels. Unfortunately,
proof that this assumption holds. Additionally, the protein might
be located in processes (dendrites, axon terminals) far from the
soma where the mRNA resides. Nevertheless,
unit mRNA abundances correlate well with the immunocyto-
chemical results obtained with the small number of specific
antibodies tested so far (Benke et al., 199 1 a,b; Zimprich
1991). For example, using (Y,, q, and a3 subunit-specific
bodies on globus pallidus, an anti-,-subunit
very strong signal, (Ye is present in a small number of cells, and
(Ye is absent (Zimprich et al., 1991). Of the three a-subunit
antibodies, that of (Ye produces the most intense reaction in
dentate gyrus (Zimprich et al., 1991), in line with our in situ
hybridization results. The protein levels of the d-subunit appear
to follow mRNA levels and distribution
al., 1989; Benke et al., 1991b). There are, however, some dis-
there is currently no
the relative sub-
antibody gives a
faithfully (Shivers et
crepancies with the yz results. Some of the strongest -y2-im-
munoreactive labeling appeared in the islands of Calleja and in
the substantia nigra (Benke et al., 199 la), areas that, although
positive for yz mRNA, are not the most marked areas of yz
mRNA abundance (Shivers et al., 1989; Malherbe et al., 1990b;
present results). The hippocampus,
of yz mRNA (considerably higher than substantia nigra), con-
tains only moderate levels of yz immunoreactivity.
arguments could be made regarding the relative specificites of
immune sera, it is possible that some degree of distortion
be present in our inferences resulting from differences in mRNA
turnover rates in different cell types.
while containing high levels
GABA, receptor mRNA distribution and pharmacology
The a-subunits. Expression
show that it is the a-subunit class that confers major pharma-
cological differences with respect to BZs on the receptor com-
plexes cyIp;uz. The (Y, &yz complexes display BZ I-type binding,
whereas (~,p,y, and a3@,yz combinations
able BZ II binding (Pritchett et al., 1989a). GABA,
containing the CX~ subunit also display a BZ II-like pharmacology
(Pritchett and Seeburg, 1990) whereas those containing aq do
not appear to bind BZ agonists (Wisden et al., 199 1 b). The brain
areas expressing the highest level of a, mRNA, that is, olfactory
bulb, medial septum (Wisden
zona incerta, central nucleus of the inferior colliculi, red nucleus,
substantia nigra pars reticulata, and cerebellum are precisely
those regions that are mainly of the BZ I type (Young et al.,
198 1; Niddam et al., 1987; Sieghart, 1989). These areas contain
relatively low levels of mRNA for the other BZ agonist-binding
a-subunit variants (+, (Ye, (YJ. One potential difficulty with the
assignment of the cu,&y, combination
BZ I binding is enriched in cortical layer IV (Young et al., 198 1;
Niddam et al., 1987; Olsen et al., 1990), while the mRNAs
not. However, this could be due to spatial mismatches between
receptor protein present in dendrites and mRNA
Conversely, the spinal cord (Persohn et al., 199 1; Wisden et
al., 199 1 a), the nucleus accumbens, caudate nucleus, and parts
of the amygdala have relatively little (Y, mRNA
highly the (Y* and/or a3 mRNAs.
inantly BZ II sites (Young et al., 198 1; Niddam et al., 1987).
These data are concordant with in vitro expression binding data
on the a-subunits. Additionally,
lations of BZ I and BZ II binding sites (e.g., cortex, hippocam-
pus) have mixed populations of mRNAs
A puzzling observation is that, in receptor autoradiography,
the GABA agonists 3H-muscimol
hypothalamic and amygdaloid areas (Bowery et al., 1987; Olsen
et al., 1990), even though in both of these regions CQ and p3
mRNAs are well expressed. It is possible that a2/@, subunit-
containing receptors have a high affinity for certain GABA an-
tagonists, since the distribution
SR-9553 1 matches the distribution
Martin, 1988; Olsen et al., 1990); its binding in cortex decreases
from superficial to deep layers, and the highest density ofbinding
is in the hippocampus and nucleus accumbens, with interme-
diate levels present in caudate nucleus. Low densities of sites
are observed in thalamic nuclei, substantia nigra, and both layers
of the cerebellum.
It appears that forebrain also contains an unusual type of
GABA, receptor, constructed in part from the (Ye subunit. The
recombinant (a&,rJ is characterized by 3H-Ro 15-45 13 binding
studies on recombinant receptors
et al., 1991b), globus pallidus,
as a BZ I subtype is that
in the soma.
These areas contain predom-
areas that have mixed popu-
encoding BZ agonist-
and 3H-GABA fail to decorate
of the GABA,
of a2 mRNA
The Journal of Neuroscience, March 1992, 12(3) 1059
not displaceable by diazepam (Wisden et al., 199 1 b). The prop-
erties of this cu,/3,yz subtype are reminiscent
subunit-containing GABA, receptors (Liiddens et al., 1990).
Although (Ye mRNA is abundant in many forebrain areas, Ro
15-45 13 binding that is resistant to BZ agonist displacement
does not seem very common except in the cerebellum (Sieghart
et al., 1987; Turner et al., 1991). However,
diazepam-insensitive Ro 15-45 13 binding has been detected in
cortex, hippocampus, and striatum, although not in the thala-
mus (Turner et al., 1991). This may suggest that a fraction of
the (Ye subunits combines with y-subunits
(cortex, hippocampus, striatum),
example, the thalamus is the principal region for aq gene ex-
pression, but because ofthe relative scarcity ofy-subunit
thalamic (Ye containing receptors may exhibit binding of mus-
cimol but not BZs (see below).
The P-subunits. In recombinant
(Y-, p-, and y-subunits, the three P-subunits appear to be func-
tionally interchangeable isoforms (Pritchett et al., 1989a; Ymer
et al., 1989b), yet clearly their marked differential distributions
of mRNAs (see also Lolait et al., 1989; Zhang et al., 1990; Laurie
et al., 1992) would suggest some functional
p-variants. Recently, a fourth p cDNA variant has been isolated
from avian brain cDNA libraries (Bateson et al., 199 1 b), but it
is not at present clear if a rat fi4 homolog exists or whether it
represents an avian idiosyncrasy.
reported for recombinant p-subunits
current amplitudes in receptors expressed in Xenopus oocytes
(Sigel et al., 1990). This could simply be due to different effi-
ciencies of protein expression in the oocyte system. However,
other subtle differences may emerge if the cloned P-subunits are
tested with their most likely in vivo partners. For example, it
has been suggested that natural receptors containing & and &
subunits differ in their affinity to GABA analogs and pentobar-
bital (Bureau and Olsen, 1990).
The y-subunits. The y-subunits
tentiation by BZs of (Y- and @subunit-containing
(Pritchett et al., 1989b; Ymer et al., 1990; Herb et al., 1992).
Of the three known members, the yz mRNA
versal and abundant. It is also the most studied in terms of
function (Pritchett et al., 1989a,b; Sigel et al., 1990; Verdoom
et al., 1990). Additional complexity has recently arisen in that
the y2 mRNA exists in two splice versions, resulting from an
exonic insertion into the cytoplasmic loop between transmem-
brane segments M3 and M4 (Whiting et al., 1990; Kofuji et al.,
1991; Wafford et al., 1991). This splicing event is predicted to
generate a yz polypeptide with the addition of a target sequence
for protein kinase C. As assessed by PCR analysis, the relative
abundances of the yz splice variants depend on the brain region
(Whiting et al., 1990). However,
to both of these mRNA versions.
In certain limbic areas of the brain such as amygdala, hy-
pothalamus, and septum, yz mRNA
considerable amounts of y, mRNA.
selective for y, subunit-containing
expected selectively to modulate neurons of affective circuits.
The y, subunit would appear to contribute to only a minority
of CNS GABA, receptors, as assessed by its low mRNA
dance. Yet other regions, for example, the tenia tecta and the
thalamus, contain none or very little, respectively, of any of the
known y-subunit mRNAs, even though high levels of non-y-
subunit transcripts are detected in these regions. How the dif-
ferent y-subunits affect the functional properties of the (Y- and
of cerebellar 01~
a low amount of
in some brain regions
but not in other areas. For
receptors assembled from
So far, the only differences
(/3, vs. p,) are those of
are required for allosteric po-
is the most uni-
our y2 probe would hybridize
appears to be replaced by
Thus, future compounds
GABA, receptors might be
P-subunits is a largely unexplored area, and many of the com-
binations we have suggested here have not yet been subjected
to binding or electrophysiological
y-subunits can differentially modulate responsiveness
units to BZs and related compounds (Ymer et al., 1990; Herb
et al., 1992). Specifically, the y, subunit confers positive mod-
ulation by /3-carbolines and DMCM
carboline-3-carbolic acid methyl ester) on LVJ~, complexes (Puia
et al., 199 l), and the replacement of y2 with either y, or ys leads
to a general lowering of the response to BZs. Consequently, the
y2 subunit may not always be the appropriate y-subunit to use
when testing a-subunit pharmacology in vitro.
The&subunit. This subunit’s role in GABA, receptor function
has remained an enigma. In a large number of regions (princi-
pally thalamic nuclei) it colocalizes with the CX,, ad, and pz mRNAs,
although its distribution is more restricted (see Fig. 11). These
results suggest that in vitro experiments should be designed with
recombinant receptors containing both CY~ and &subunits.
originally suggested for the d-subunit alone (Shivers et al., 1989),
ol,a,@-containing receptors may have high affinity to muscimol
but lack BZ binding sites. In the cerebellum, a,a,@-containing
receptors would correspond to high-affinity
the granule cell layer (Laurie et al., 1992). Thus, &subunits may
preferentially associate with a-subunits
analysis. However, different
muscimol sites in
that do not bind BZ
A number of plausible GABA, receptor combinations have been
inferred based on mRNA distribution.
and immunocytochemical studies in addition to modem patch-
clamping methodology on brain slices (Edwards
could be used to test the validity of these predictions,
recombinant receptors as a reference. Although the original clas-
sification of BZ I and BZ II receptors could be criticized for
being too reductionistic, it seems that a systematic comparison
of GABA, receptor subunit distributions
ables this distinction to be maintained with qualifications. Thus,
based on their patterns of gene expression
expression studies, we would propose that the a,&y, receptors
correspond to the BZ I subtype, and receptors containing c&y,,
(~$~y~, and c@,y, are three subtypes of BZ II receptor. Addi-
tional as yet unclassified subtypes of receptor would emerge if
these subunits occur with the y, or y3 subunits. The receptors
containing (Ye and cyg would diverge from the main family of BZ
receptors because of their very restricted
profile in vitro. Indeed, in some brain regions the (Ye subunit
may contribute to BZ-insensitive
mRNA fails to colocalize with that of any known y-subunit.
The physiological significance of such GABA,
plexity remains to be determined.
presynaptic terminals could serve as autoreceptors
negative feedback of transmitter
GABA, receptor complexes could differ in mean channel open
time or desensitization rate, with constraints
ometry dictating the type of desensitization
ductance that is required to produce a given hyperpolarization
per unit time. Alternatively, neurons receiving a strong excit-
atory input may require GABA,
tance and slower densensitization
such input is less. Analogous situations for receptor diversity
also extend to other ligand-gated ion channels in brain, most
notably the neuronal nicotinic receptors (Wada et al., 1989;
Morris et al., 1990), glycine receptors (Betz, 1991; Malosio et
et al., 1989)
in the brain still en-
BZ ligand binding
GABA, receptors because its
of neuronal ge-
rate or channel con-
receptors of greater conduc-
rates than neurons in which
1060 Wisden et al.
l GABA, Receptor mRNA Distribution
al., 199 l), and glutamate receptors (Bettler et al., 1990; Sommer
et al., 1990; Monyer et al., 199 1; Werner et al., 1991). Future
experimental directions could involve homologous
nation experiments to see if it is possible to dissect out the
function of the different receptor subtypes.
List of anatomical abbreviations
central amygdaloid nucleus
central gray, dorsal
centrolateral thalamic nucleus
neocortex, layer 2
dorsal lateral geniculate thalamic nucleus
islands of Calleja
lateral amygdaloid nucleus
laterodorsal thalamic nucleus
lateral posterior thalamic nucleus, mediocaudal
mediodorsal thalamic nucleus
medial amygdaloid nucleus
medial geniculate nucleus
medial habenular nucleus
medial preoptic nucleus
optic nerve layer, superior colliculus
parafascicular thalamic nucleus
pamventricular thalamic nucleus
rhomboid thalamic nucleus
reticular thalamic nucleus
substantia nigra pars compacta
substantia nigra pars reticulata
superficial gray layer, superior colliculus
triangular septal nucleus
ventral lateral geniculate nucleus
ventral posterior thalamic nucleus
anterior commissure, anterior
anteroventral thalamic nucleus
bed nucleus, stria terminalis
fields l-4 of Ammon’s horn
Alger BE, Nicoll RA (1982)
of GABA receptor on rat hippocampal
vitro. J Physiol (Lond) 328: 125-14 1.
Pharmacological evidence for two kinds
pyramidal cells studied in
Bateson AN, Harvey RJ, Wisden W, Glencorse TA, Hicks AA, Hunt
SP, Barnard EA, Darlison MG ( 199 1 a) The chicken GABA,
tor (Y, subunit cDNA sequence and localization
mRNA. Mol Brain Res 9:333-339.
Bateson AN, Lasham A, Darlison MG
receptor heterogeneity is increased by alternative splicing of a novel
@-subunit gene transcript. J Neurochem
Benke D, Mertens S, Trzeciak A, Gillessen D, Mijhler H (199 la) GA-
BA, receptors display association of yZ subunit with (Y,- and j32,3-
subunits. J Biol Chem 266:4478-1483.
Benke D, Mertens S, Trzeciak A, Gillessen D, Mohler H ( 199 1 b) Iden-
tification and immunohistochemical
subtypes containing the b subunit in rat brain. FEBS Lett 283:145-
Bettler B, Boulter J, Hermans-Borgmeyer
neris ES, Moll C, Borgmeyer U, Hollmann
Cloning of a novel glutamate receptor subunit GluR 5: expression in
the nervous system during development.
Betz H (199 1) Glycine receptors: heterogeneous and widespread ex-
pression in the mammalian brain. Trends Neurosci 14:458-46 1.
Bleier R, Byne W (1985) Septum and hypothalamus. In: The rat ner-
vous system, Vol 1 (Paxinos G, ed), pp 87-l 18. Sydney: Academic.
Bowery NG ( 1989) GABA, receptors and their sianificance in mam-
malian pharmacology. Trends Pharmacol Sci lo:20 l-407.
Bowery NG, Hudson AL, Price GW
receptor site distribution in the rat central nervous system. Neurosci-
Bristow DR, Martin IL (1988) Light microscopic
localization in rat brain of the bindina sites for the GABA,
antagonist [‘HI SR 9553 1: comparison-with
Eur J Pharmacol 148:283-288.
Bureau M, Olsen RW (1990) Multiple
r-aminobutyric acid-A receptor protein show different ligand-binding
affinities. Mol Pharmacol 37:497-502.
Cooper E, Couturier S, Ballivet M (1991)
subunit stoichiometry of a neuronal nicotinic acetylcholine receptor.
Cutting GR, Luo L, O’Hara BF, Kasch LM, Montrose-Rafizadeh
Donovan DM, Shimada S, Antonarakis
Kazazian HH (199 1) Cloning of the GABA p, cDNA: a novel GA-
BA, receptor subunit highly expressed in retina. Proc Nat1 Acad Sci
de Blas AL, Vitorica J, Friedrich P (1988) Localizaton
receptor in rat brain with a monoclonal
peptide of the GABA, receptor/benzodiazepine
complex. J Neurosci 8:602-6 14.
Duggan MJ, Stephenson FA (1990)
istence of y-aminobutyrate, receptor isooligomers. J Biol Chem 265:
Dutar P, Nicoll RA (1988) A physiological role for GABA,
in the central nervous system. Nature 322: 156-l 58.
Edwards FA, Konnerth A, Sakmann B, Takahashi T (1989)
slice preparation for patch clamp recordinn from neurones in the
mammalian central nervous system. Pfluegers Arch 4 14:600-6 12.
Ewert M, Shivers BD, Liiddens H, Miihler
Subunit selectivity and epitope characterization
against the GABA,/benzodiazepine
Ewert M, de Blas AL, Mohler H, Seeburg PH (1991)
epitope on GABA, receptors is recognized by two different mono-
clonal antibodies. Brain Res, in press.
Fuchs K, Mohler H, Sieghart W (1988)
brain, specifically and irreversibly labelled by [3H]flunitrazepam,
distinct (Y subunits of the GABA-benzodiazepine
Neurosci Lett 90~3 14-3 19.
Garrett KM, Saito N, Duman RS, Abel MS, Ashton RA, Fujimori
Beer B, Tallmann JF, Vitek MP, Blume AJ (1990)
pression of y-aminobutyric acid,, receptor subunits. Mol Pharmacol
Haefely W, Pole P (1986) Physiology of GABA enhancement by ben-
zodiazepines and barbiturates. In: Benzodiazepine/GABA,
and chloride channels: structural and functional properties (Olsen
RW, Venter JC, eds), pp 97-133. New York: Liss.
HIring P, Stahli C, Schoch P, Takacs B, Staehelin T, Mijhler H (1985)
Monoclonal antibodies reveal structural homogeneity
of the corresponding
(199 1 b) y-Aminobutyric acid,
mapping of GABA, receptor
M, Heinemann S (1990)
Neuron 5: 583-595.
(1987) GABA, and GABA,
[‘HI GABA, distribution.
distinct subunits of the
Pentameric structure and
SE, Guggino WB, Uhl GR,
of the GABA,
antibody to the 57,000 M,
Biochemical evidence for the ex-
H, Seeburg PH (1990)
of mAbs directed
receptor. J Cell Biol 110:2043-
Various proteins from rat
The Journal of Neuroscience, March 1992, 12(3) 1061
Nat1 Acad Sci USA 82:48374841.
Herb A, Wisden W, Lilddens H, Puia G, Vicini S, Seeburg PH (1992)
A third y subunit of the GABA,
USA, in press.
Hironaka T, Morita Y, Hagihira S, Tateno E, Kita H, Tohyama
(1990) Localization of GABA, receptor a, subunit mRNA-contain-
ing neurons in the lower brainstem of the rat. Mol Brain Res 7:335-
Jones EG (1985) The thalamus. New York: Plenum.
Kato K (1990) Novel GABA, receptor LY subunit is expressed only in
cerebellar granule cells. J Mol Biol 2 14:6 19-624.
Khrestchatisky M, MacLennan AJ, Chiang M-Y, Xu W, Jackson MB,
Brecha N, Stemini C, Olsen RW, Tobin AJ (1989) A novel (Y subunit
in rat brain GABA, receptors. Neuron 3:745-753.
Khrestchatisky M, MacLennan AJ, Tillakaratne
Tobin AJ (199 1) Sequence and regional distribution
encoding the a2 polypeptide of rat y-aminobutyric
Kofuji P, Wang JB, Moss SJ, Huganir RL, Burt DR (199 1) Generation
of two forms of the y-aminobutyric
by alternative splicing. J Neurochem 56:7 13-7 15.
Lambert JJ, Peters JA,Cottrell GA (1987) Actions of synthetic and
endogenous steroids on the GABA,
Laurie DJ, Seeburg PH, Wisden W (1992)
GABA, receotor subunit mRNAs in the rat brain. II. Olfactorv bulb
and cerebellum. J Neurosci 12: 1063-1076.
Lolait SJ, O’Carroll AM, Kusano K, Mahan LC (1989)
logical characterization and region specific expression in brain of the
& and &-subunit of the rat GABA,
Liiddens H, Wisden W (199 1) Function and pharmacology of multiple
GABA, receptor subunits. Trends Pharmacol Sci 12:49-5 1.
Liiddens H, Pritchett DB, Kiihler M, Killisch I, Keinlnen
H, Sprengel R, Seeburg PH (1990)
lective for a behavioural alcohol antagonist. Nature 346648-65
Lilddens H. Killisch I. Seebura PH (1991)
may exist in a GABA,/benzidiazepine
MacLennan AJ, Brecha N, Khrestchatisky M, Stemini C, Tillakaratne
NJK, Chiang M-Y, Anderson K, Lai M, Tobin AJ (199 1) Indepen-
dent cellular and ontogenetic expression of mRNAs encoding three
(Y polypeptides ofthe rat GABA, receptor. Neuroscience 43:369-380.
Malherbe P, Sigel E, Baur R, Pershon E, Richards JG, MGhler H ( 199Oa)
Functional expression sites of gene transcription
of the GABA, receptor in rat brain. FEBS Lett 260:261-265.
Malherbe P, Sigel E, Baur R, Pershon E, Richards JG, Mbhler
(1990b) Functional characteristics and sites of gene expression of the
o;;~$,, y,-isoform of the rat GABA,
receptors in different brain areas. Proc
receptor family. Proc Nat1 Acad Sci
NJK, Chiang M-Y,
of the mRNA
acid, receptors. J
acid, receptor y2 subunit in mice
receptor. Trends Pharmacol Sci
The distribution of 13
receptor. FEBS Lett 258: 17-2 1.
receptor se- Cerebellar GABA,
More than one (Y variant
receptor complex. J Receptor
of a novel LY subunit
receptor. J Neurosci 10:2330-
Malosio M-L, Marqueze-Pouey
spread expression of glycine receptor subunit mRNAs in the adult
and develooina rat brain. EMBO J 10:2401-2409.
Monyer H, .Seeb&g PH, Wisden W (199 1) Glutamate-operated
nels: developmentally early and mature forms arise by alternative
splicing. Neuron 6:799-8 10.
Morris BJ, Hicks AA, Wisden W, Darlison MG, Hunt SP, Barnard EA
(1990) Distinct regional expression of nicotinic acetylcholine recep-
tor genes in chick brain. Mol Brain Res 7:305-3 15.
Nicoll RA, Dutar P (1989) Physiological roles of GAB%
receptors in synaptic transmission in the hippocampus. In: Allosteric
modulation of amino acid receptors: therapeutic implications
nard EA, Costa E, eds), pp 195-205. New York: Raven.
Niddam R, Dubois A, Scatton B, Arbilla
radiographic localization of [3H]zolpidem
Comparison with the distribution
J Neurochem 49:890-899.
Olsen RW, Tobin AJ (1990) Molecular biology of GABA,
FASEB J 4: 1469-l 480.
Olsen RW, McCabe RT, Wamsley JK (1990) GABA,
types: Autoradiographic comparison of GABA, benzodiazepine,
convulsant binding sites in the rat central nervous system. J Chem
Palacios JM, Wamsley JK, Kuhar MJ
B, Kuhse J, Betz H (1991) Wide-
S, Langer SZ (1987) Auto-
binding in the rat CNS.
of [3H]flunitrazepam binding sites.
Brain Res 222:285-307.
High affinity GABA
Paxinos G, Watson C (1986) The rat brain in stereotaxic coordinates,
2d ed. Sydney: Academic.
Persohn E, Malherbe P, Richards JG (1991) In situ hybridization
histochemistry reveals a diversity of GABA, receptor subunit mRNAs
in neurons of the rat spinal cord and dorsal root ganglia. Neuroscience
Pritchett DB, Seeburg PH (1990)
subunit creates novel type II benzodazepine
J Neurochem 54: 1802-1804.
Pritchett DB, Ltlddens H, Seeburg PH (1989a)
GABA, benzodiazepine receptors produced in transfected cells. Sci-
Pritchett DB, Sontheimer H, Shivers BD, Ymer S, Kettenmann
Schofield PR, Seeburg PH (1989b)
receptor subunit for benzodiazepine
Puia G, Santi M, Vicini S, Pritchett DB, Purdy RH, Paul SM, Seeburg
PH, Costa E (1990) Neurosteroids act on recombinant human GA-
BA, receptors. Neuron 4~759-765.
Puia G, Vicini S, Seeburg PH, Costa E ( 199 1) Influence of recombinant
y-aminobutyric acid, receptor subunit composition
allosteric modulators of y-aminobutyric
Richards JG, Mijhler H, Haefely W (1986) Mapping benzodiazepine
receptors in the CNS by radiohistochemistry
istry. In: Neurohistochemistry: modem methods and applications
(Panula P, Pairannta H, Soinita S, eds), pp 629-677. New York: Liss.
Richards JG. Schoch P. Hiring P. Takacs B. Miihler
localization in the CNS with monoclonal
Seeburg PH, Wisden W, Verdoom
A, Ltlddens H, Sprengel R, Sakmann B (1990) The GABA, reeeptor
family: molecular and functional diversity. Cold Spring Harb Symp
@ant Biol 55:29-40.
Siquier JM, Richards JG, Malherbe P, Price GW, Mathews S, Miihler
H (1988) Mapping of brain areas containing RNA homologous to
cDNAs encoding the LY and @ subunits of the rat GABA,
butyrate receptor. Proc Nat1 Acad Sci USA 85:78 15-78 19.
Shivers BD, Killisch I, Sprengel R, Sontheimer H, Kiihler M, Schofield
PR, Seeburg PH (1989) Two novel GABA,
in distinct neuronal s&populations.
Siegel RE (1988) The mRNAs encoding GABA,/benzodiazepine
ceptor subunits are localized in different cell populations ofthe bovine
cerebellum. Neuron 1:579-584.
Sieghart W (1989) Multiplicity of GABA,-benzodiazepine
Trends Pharmacol Sci lo:4074 11.
Sieghart W, Eichinger A, Richards JG, Miihler H (1987) Photoaffinity
labelling of benzodiazepine receptor proteins with the partial agonist
[‘H]Ro 15-45 13: a biochemical and autoradiographic
Sigel E, Baur R, Trube G, Miihler
of subunit composition of rat brain GABA,
function. Neuron 5:703-7 11.
Sommer B, Kainiinen K, Verdoom TA, Wisden W, Bumashev N, Herb
A, Kiihler M, Takagi T, Sakmann B, Seeburg PH (1990)
flop: a cell specific functional switch in glutamate operated channels
of the CNS. Science 249:158&1585.
Turner DM, Sapp DW, Olsen RW (199 1) The benzodiazepine/alcohol
antagonist Ro 15-45 13: binding to a GABA,
is insensitive to diazepam. J Pharmacol Exo Ther 257:1236-1242.
Unnerstall JR, Kuhar %IJ, Niehoff DL, Palacios JM (1981)
diazepine receptors are coupled to a subpopulation
tors: evidence from a quantitative
macol Exp Ther 218:797-804.
Unwin N (1989) The structure of ion channels in membranes of ex-
citable cells. Neuron 3:665-676.
Verdoom TA, Draguhn A, Ymer S, Seeburg PH, Sakmann B (1990)
Functional properties of recombinant GABA, receptors depend upon
subunit composition. Neuron 4:919-928.
Wada E, Wada K, Boulter J, Deneris E, Heinemann S, Patrick J, Swan-
son LW (1989) Distribution of alpha2, alpha3, alpha4, and beta2
neuronal nicotinic receptor subunit mRNAs in the central nervous
system: a hybridization histochemical study in the rat. J Comp Neurol
y-Aminobutyric acid, receptor QL~
Type I and type II
Importance of a novel GABA,
pharmacology. Nature 338:582-
on the action of
acid-gated Cl- currents. Mol
H f 1987) Re-
receptors:’ cellular and‘ subckllular
antibodies. J Neurosci 7:
TA, Pritchett DB, Werner P, Herb
receotor subunits exist
study. J Neu-
H, Malherbe P (1990) The effect
receptors on channel
receptor subtype that
of GABA recep-
study. J Phar- autoradiographic
1062 Wisden et al. - GABA, Receptor mRNA Distribution Download full-text
Wafford KA, Burnett DM, Dunwiddie
differences in the ethanol sensitivity of GABA,
in Xenopus oocytes. Science 249:29 l-293.
Wafford KA, Burnett DM, Leidenheimer NJ, Burt DR, Wang JB, Kofuji
P, Dunwiddie TV, Harris RA, Sikela JM (199 1) Ethanol sensitivity
of the GABA, receptor expressed in Xenopus oocytes requires 8 amino
acids contained in the y2L subunit. Neuron 7~27-33.
Werner P. Voiat M. Keiniinen K. Wisden W. Seebera PH (199 1) Clon-
ing of a p&i& high-affinity kainate receptor ixpresged piedomi-
nantly in hippocampal CA3 cells. Nature 35 1:742-744.
Whiting P, McKeman RM, Iversen LL (1990)
for creating diversity in y-aminobutyric
ing directs expression of two forms of y2 subunit, one of which con-
tains a protein kinase C phosphorylation
Wilson-Shaw D, Robinson M, Gambarana
(199 1) A novel y subunit of the GABA,
the polymerase chain reaction. FEBS Lett 2842 1 l-2 15.
Wisden W, Morris BJ, Darlison MG, Hunt SP, Barnard EA (1988)
Distinct GABA, receptor (Y subunit mRNAs show differentiated pat-
terns of expression in bovine brain. Neuron 1:933-947.
Wisden W. McNauahton LA. Darlison MG. Hunt SP. Barnard EA
(1989a) ‘Differentyal distribbtion
ebellum-localization of az mRNA in Bergmann glia layer. Neurosci
Wisden W, Morris BJ, Darlison MG, Hunt SP, Barnard EA (1989b)
Localization of GABA, receptor a subunit mRNAs in relation to
receptor subtypes. Mol &a&Res
Wisden W, Gundlach AL, Barnard EA, Seeburg PH, Hunt $P (199 la)
Distribution ofGABA, receptor subunit mRNAs in rat lumbar spinal
cord. Mol Brain Res 10:179-183.
TV, Harris RA (1990) Genetic
type A receptors: RNA splic-
site. Proc Nat1 Acad Sci
C, Siegel RE, Sikela JM
receptor identified using
of GABA, receptor in bovine cer-
Wisden W, Herb A, Wieland H, KeinHnen K, Liiddens H, Seeburg PH
(199 1 b) Cloning, pharmacological
tern of the rat GABA, receptor a4 subunit. FEBS Lett 289:227-230.
Wisden W, Morris BJ, Hunt SP (1991~)
synthetic DNA probes. In: Molecular
proach, Vo12 (Chad J, Wheal H, eds), pp 205-225. Oxford: Oxford
Ymer S, Draguhn A, Kiihler M, Schofield PR, Seeburg PH (1989a)
Sequence and expression of GABA,
Ymer S, Schofield PR, Draguhn A, Werner P, Kiihler M, Seeburg PH
(198913) GABA, receptor j3 subunit heterogeneity: functional ex-
pression of cloned cDNAs. EMBO J 8:1665-1670.
Ymer S, Draguhn A, Wisden W, Werner P, Keiniinen K, Schofield PR,
Sprengel R, Pritchett DB, Seeburg PH (1990) Structural and func-
tional characterization of the y, subunit of GABA,ibenzodiazepine
receptors. EMBO J 9:3261-3267.
Young WS III, Niehoff DL, Kuhar MJ, Beer B, Lippa AS (198 1) Mul-
tiple benzodiazepine receptor localization
diohistochemistrv. J Pharmacol
Young WS III, Mezey E, Siegel RE -( 1986) Quantitative
bridization histochemistry reveals increased levels of corticotropin
releasing factor mRNA after adrenalectomy in rats. Neurosci Lett 70:
Zhang JH, Sato M, Noguchi K, Tohyama M (1990) The differential
expression patterns of the mRNAs encoding @ subunits (j3,, & and
&) of GABA, receptors in the olfactory bulb and its related areas in
the rat brain. Neurosci L&t 119:257-260.
Zimprich F, Zezula J, Sieghart W, Lassmann H (1991)
tochemical localization of the (Y,, (Ye and (Y) subunits of the GABA,
receptor in the rat brain. Neurosci Lett 127: 125-128.
characteristics and expression pat-
In situ hybridization
receptor (r4 subunit. FEBS Lett
by light microscopic ra-
EXD Ther 216:425436.
in situ hy-