(Received for publication, April 24, 1997)
From The Institute for Genomic Research, Rockville, Maryland 20850
A novel subunit of the -aminobutyrate, type A
(GABAA) receptor family has been identified in human
and rat tissues. The subunit displays 30-40% amino acid identity with
known family members and represents a distinct subunit class (termed
). Transcripts of the
subunit were detected in several human
tissues and were particularly abundant in the uterus. The
subunit
protein can assemble with known GABAA receptor subunits and
confer unique ligand binding properties to the recombinant receptors in
which it combines. Most notably, the presence of the
subunit alters the sensitivity of recombinant receptors to the endogenous steroid, pregnanolone. Identification of the
subunit indicates a new target
for pharmacological manipulation of GABAA receptors that are located outside of the central nervous system.
In mammalian brain, synaptic inhibition of neuronal activity is
mediated mainly by -aminobutyric acid
(GABA)1 at GABAA receptors. In
mammals, the 14 known subtypes of GABAA receptor subunits
have been categorized within five structural classes (
1-6,
1-3,
1-3,
,
). These subunits are thought to assemble in different
pentameric complexes, with most functional receptors containing
/
/(
,
, or
) subunit combinations (1-3). A sixth class
of subunit (
) form homomeric GABA receptors that do not appear to
coexist with GABAA receptor subunits and correspond pharmacologically to a related (GABAC) receptor subtype (4, 5).
In addition to their location on central neurons and astroglia, functional GABAA receptors have been detected on peripheral neurons and non-neuronal cells. The non-neuronal cells include endocrine cells of the pituitary pars intermedia, adrenal medulla, islets of Langerhans, and placenta (6-9). The receptors have also been located on smooth muscle cells of the urinary bladder and uterus (10, 11).
The precise function of GABAA receptors in non-neuronal
cells is presently ill-defined. Clearly, their location on endocrine cells suggests a role in the regulation of hormone secretion. However,
their function in the uterus appears to be related directly to tissue
contractility. Within the uterus, GABA, and its metabolic enzymes are
found at high concentrations (12). The GABAA receptors of
this tissue have been proposed to regulate uterine motility by
inhibiting contractions (13). They may also mediate the relaxing effects of 5,3
-pregnanolone, an endogenous steroid that is
capable of activating GABAA receptors directly (13,
14).
Here, we describe a novel class of GABAA receptor subunit that is expressed at relatively high levels in several peripheral tissues, including the uterus. This subunit can combine with known GABAA receptor subunits and alter the sensitivity of recombinant receptors to modulatory agents such as pregnanolone.
[35S]tert-Butyl
bicyclophosphorothionate (TBPS), [3H] muscimol,
[3H]Ro15-1788, and [3H]flunitrazepam were
purchased from DuPont. Pentobarbital and 5-pregnan-3
-ol-20-one
were from Sigma. The HEK-293 cell line (ATCC CRL 1573) was obtained
from American Type Culture Collection. The PANC cell line was provided
by Dr. E. Jaffe (Johns Hopkins University). Human placenta and uterus
were obtained from National Disease Research Interchange. Samples of
poly(A)+ RNA from other tissues were purchased from
Clontech.
A consensus
sequence of amino acid residues that are conserved between all known
vertebrate GABAA receptor subunits (3) was searched against
the human cDNA data base (15), using the TBLASTN algorithm (16).
The search uncovered an expressed sequence tag (EST) that encodes part
of a subunit homologue that has not been described previously. This EST
was derived from a partial cDNA clone of a pancreatic carcinoma
mRNA (nucleotides 859-3300; Fig. 1). Flanking 5 sequence
(nucleotides 1-984; Fig. 1) was obtained by anchored PCR (17), using a
pancreatic carcinoma cDNA library as template. A fragment of the
amplified cDNA (nucleotides 2-230; Fig. 1) was then used to screen
approximately 5 × 105 plaques of the cDNA
library, using standard procedures (17). Two independent clones that
contained full-length inserts of 3.3 kb were isolated. A fragment of
one insert (nucleotides 76-1503; Fig. 1) was transferred to the pCDM8
vector (Invitrogen) and re-sequenced over its entire length prior to
expression studies.
Cloning of the Rat
Initially, two
fragments of the rat subunit gene were amplified by PCR from total
genomic DNA using degenerate primers that were derived from the human
subunit protein sequence. For fragment 1, the primers were
5
-GGWAATGATGTKGARTTYACYTGG-3
and 5
-CGATCTKGTKACTAAKGTRAARTA-3
(nucleotides 709-732 and 799-822; Fig. 1). For fragment 2, they were
5
-CACTAGRTTRGTYTTRCARTTTGA-3
and 5
-GCAGGTTCTTGCAGGGACTGAATC-3
(nucleotides 843-866 and 961-984; Fig. 1). Amplification at
95 °C for 45 s, 55 °C for 60 s, 72 °C for 2 min was
performed for 30 cycles using the GenAmp system (Perkin-Elmer).
Sequences of the cloned fragments were then used to design primers for
amplification of two overlapping fragments of the rat
subunit
cDNA. Total RNA (5 µg) was isolated from rat uterus (18), and
reverse-transcribed using 200 units of Moloney murine leukemia virus
reverse transcriptase (Life Technologies, Inc.) at 42 °C for 60 min.
The primer was a 15-mer poly(dT) oligonucleotide. Samples (5%) of the
reaction products were then subjected to 35 cycles of PCR as described above. For the 5
-region of the cDNA, the primers were
5
-gcggaattcCAGAGCCTCAACAACTACCTG-3
(nucleotides 94-114; Fig. 1) and
5
-gcggaattcCCAAAAGGAGACCCAGGATAA-3
(nucleotides 820-840 of GenBankTM
accession number U95368[GenBank]). For the 3
-region, the primers were
5
-gcggaattcGATTCGGTACGTGGACTCGAA-3
(nucleotides 635-654 of U95368[GenBank])
and 5
-gcggaattcGTTGAAGACCTATGGCATGCA-3
(nucleotides 1497-1516; Fig.
1). The two amplified cDNA fragments (858 and 772 base pairs) were
purified and sequenced directly. The protein-coding region of the
cDNA sequence was then amplified from total uterus cDNA using
the XL-PCR system (Perkin-Elmer). The primers were
5
-ccggccactaGTGAATCAGCTCCTTCAATATGAGCTACA-3
(nucleotides 27-55 of
GenBankTM number U95368[GenBank]) and
5
-ccggccactagTCAAAAATACATGTAGTATGCCCAGTAA-3
(nucleotides
1341-1368 of GenBankTM number U95368[GenBank]). The reaction products were
ligated into the pCDM8 vector, and individual clones were sequenced
over their entire lengths to ensure that no mutations had been
introduced.
Poly(A)+ RNA was
prepared from placenta, uterus, and PANC cells using the Oligotex
system (Qiagen). Approximately 2-µg samples of poly(A)+
RNA were electrophoresed on 1.2% formaldehyde-agarose gels,
transferred to nylon membranes, and hybridized with a
32P-labeled fragment of the 3-untranslated
subunit
cDNA (nucleotides 1501-2340; Fig. 1). The blots were washed at
60 °C in 0.1 × SSC, 0.1% SDS prior to exposure. The blots
were stripped of probe by boiling in 0.5% SDS, and re-hybridized with
a 32P-labeled fragment of the cDNA that encodes human
glyceraldehyde phosphate dehydrogenase (nucleotides 789-1140; Ref.
19).
Human embryonic
kidney cells (HEK-293) were maintained in Dulbecco's modified Eagle's
medium, supplemented with calf serum (10%), penicillin (50 units/ml),
and streptomycin (50 µg/ml). Cells were plated at ~20% confluence,
24 h prior to transfection. Transfection of subunit cDNAs was
performed with a calcium phosphate-DNA precipitate in HEPES buffer
(17). Following incubation with the precipitate for 24 h, the
cells were washed and cultured for a further 48 h before
harvesting. All subunit cDNAs were derived from rat and were cloned
in pCDM8. Transfection efficiencies were standardized by cotransfecting
a -galactosidase cDNA, as described previously (20).
A crude membrane fraction was prepared from transfected cells as described previously (3). The [35S]TBPS binding activity of membranes (50-100 ng of protein) was assayed by incubation for 120 min at 25 °C in 100 µl of a solution containing Tris-HCl (20 mM), NaCl (1 M), pH 7.5. Assays included [35S]TBPS at 1-40 nM (for Scatchard analysis) or 5 nM (all other experiments). Nonspecific binding was determined in the presence of picrotoxin (100 µM) and was equal to the binding of mock-transfected cell membranes (<0.03 pmol/mg of protein). Following incubation, 5 ml of ice-cold assay buffer was added, and the membranes filtered through Whatman GF/C filters under vacuum. The filters were washed twice with 5 ml of the same buffer prior to scintillation counting. The binding of [3H]muscimol (20 nM) was assayed by incubation of membranes for 60 min at 0 °C in 100 µl of TEN (10 mM Tris-HCl, 1 mM EDTA, 100 mM NaCl, ph 7.5). Nonspecific binding was determined in the presence of GABA (1 mM). The binding of [3H]Ro15-1788 and [3H]flunitrazepam was assayed by incubation of membranes for 60 min at 0 °C in 100 µl of TEN, NaCl (0.1 M). Nonspecific binding was determined in the presence of flunitrazepam (10 µM). Assays included [3H]Ro15-1788 at 0.5-6 nM (for Scatchard analysis) or 2 nM (all other experiments). Bound ligand was detected by filtration as described above. All binding data are mean values ± S.D. of three experiments, using membranes from at least two independent transfections.
A novel class of GABAA receptor subunit was identified
by searching a data base of ESTs with a peptide consensus sequence of
known family members. The sequence of this EST was used to isolate a
3.3-kb cDNA from a pancreatic carcinoma cDNA library. The
cDNA contains a large open reading frame that is flanked by stop
codons and encodes a polypeptide of 440-amino acid residues (Fig.
1). The polypeptide, termed the subunit, has all of
the hallmarks of a ligand-gated anion channel subunit. The N-terminal half of the protein exhibits five sites for potential
N-linked glycosylation, and two cysteine residues separated
by 13 amino acids (residues 160 and 174). The C-terminal half of the
protein contains four hydrophobic regions that are potential
transmembrane domains. The first of these contains a conserved proline
residue, while the second contains an octamer sequence (residues
280-287) that is similar, but not identical, to that found in most
GABAA and glycine receptor subunits. Within this octamer,
the
subunit contains a serine residue at position 284, where
threonine is found in all other GABAA receptor subunits.
This minor difference in protein sequence may explain why the
subunit cDNA has not been uncovered previously by the
cross-hybridization approaches that have relied almost exclusively on
conservation of this motif.
The amino acid sequence of the subunit is most closely related to
GABAA receptor
subunits (37% amino acid
identity), followed by the
subunit (35%) and
subunits
(33%). It displays less similarity to other GABA and glycine receptor
subunits. The maximum levels of sequence identity are similar to those
found between the different classes of GABAA receptor
subunits (20-40%; Refs. 1 and 2) and justifies its categorization
within a distinct subunit class. Of the 78 amino acid residues that are
conserved between all known GABAA receptor subunits (3),
only six are substituted in the
subunit sequence (Fig. 1). A rat
cDNA that encodes a homologue of the human
subunit was also
isolated. The rat and human
subunits display a high degree of
sequence conservation (93% amino acid identity; Fig. 1). This
conservation includes the six amino acid residues that vary from all
other known GABAA receptor subunits.
Expression of subunit mRNA in human tissues was first assessed
by searching EST data bases for additional examples of
subunit
cDNA fragments. The dbEST data base (21; release 32897) contains
eight entries2 that were derived from six
partial human
subunit cDNA clones. In common with the cDNA
described here, four of these clones were derived from two independent
pancreatic carcinomas. The remainder were isolated from a
neuroepithelial cell line (NT2) and breast tissue.
A more systematic examination of expression patterns employed a
collection cDNA libraries that were derived from most major human
tissues (22). After 35 cycles of PCR, using approximately 1 × 106 phage, a detectable subunit cDNA fragment was
amplified from the cDNA libraries of uterus, prostate, ovaries,
placenta, gall bladder, lung, small intestine, and two brain regions
(hippocampus and temporal cortex; data not shown). When available,
these tissues were examined by Northern analysis to obtain a more
quantitative estimation of subunit mRNA expression levels. A
hybridizing transcript of the predicted size (3.3 kb) could be detected
in several tissues (Fig. 2). Although this transcript
was barely detectable in small intestine, ovaries, and placenta, it was
relatively enriched in lung, thymus, and prostate and was particularly
abundant in the uterus. The
subunit transcript was also enriched in
a cell line (PANC) that was derived from a human pancreatic
adenocarcinoma.3 Notably, no hybridizing
transcripts were detected in samples of whole brain or pancreas. It
appears that the
subunit is expressed in only a minor subpopulation
of cells in these tissues. Future in situ hybridization
studies should help to identify these cell types.
The pharmacological properties that are conferred to GABAA
receptors by the subunit were examined after transient expression in HEK-293 cells. Cells that were transfected with only the
subunit
cDNA did not express binding sites for the GABAA
receptor ligands, [3H]muscimol or [35S]TBPS
(Table I). These cells also failed to elicit chloride currents in response to GABA or glycine (each 100 µM).4 The
subunit is
therefore unlike the
class of GABA receptor subunits (2) or the
class of glycine receptor subunits (23), which can each assemble
homomeric chloride channels that are gated by these ligands. When cells
were cotransfected with cDNAs encoding the
subunit, and either
an
1 subunit or a
1 subunit, there was also a failure to detect
expression of any ligand binding activities. The
subunit therefore
does not behave like a typical GABAA receptor
or
subunit (2). As expected, transfection of cells with a combination of
,
, and
subunits resulted in the expression of both
[3H]muscimol and [35S]TBPS binding sites
(Table I). However, the transfected cell membranes did not bind either
[3H]Ro15-1788 or [3H]flunitrazepam.
Therefore, the
subunit is also unlike the
class of subunits,
which confers a sensitivity to benzodiazepines when expressed with
and
subunits (2).
|
In common with the and
subunit classes, the absence or presence
of binding sites for [35S]TBPS,
[3H]muscimol, and [3H]Ro15-1788 failed to
demonstrate that the
subunit can assemble with known
GABAA receptor subunits. However, such an ability could be
inferred from the observation that inclusion of the
subunit with
/
/
subunit combinations caused a significant reduction of
[3H]Ro15-1788 binding to transfected cell membranes.
This was demonstrated clearly by transfecting different ratios of the
subunit cDNAs (Table II). When the
subunit
cDNA was transfected with an
/
/
subunit combination in an
equimolar ratio, the density of [3H]Ro15-1788 binding
sites was reduced by 45%. A 5-fold increase in the molar ratio of
transfected
subunit cDNA caused a much larger reduction (90%)
of the Bmax value for
[3H]Ro15-1788 binding. In contrast, a 5-fold increase in
the molar ratio of transfected
2 subunit cDNA, blocked the
reduction caused by the presence of the
subunit, and returned the
Bmax value to 84% of that obtained with
/
/
subunits alone. Notably, these variable subunit
combinations had little effect on the binding of
[35S]TBPS and [3H]muscimol (Table II). It
was concluded that expression of the
subunit with the
/
/
subunit combination results in expression of
/
/
/
receptors
that lack [3H]Ro15-1788 binding sites and/or a mixed
population of
/
/
and
/
/
receptors.
|
Clearly, transfection of cells with a combination of ,
,
, and
subunit cDNAs can give rise to multiple receptor types, composed of different subunit combinations and stoichiometries. For
this reason, a less complex system was chosen to perform an initial
characterization of the pharmacological properties that are conferred
to GABAA receptors by the
subunit. This system utilized
the
3 subunit, which assembles homomeric chloride channels that bind
[35S]TBPS with high affinity (24, 25). Cells were
transfected with either the
3 cDNA alone or a combination of
3 and
cDNAs. The presence of the
subunit had no
significant effects on the Kd and
Bmax values for [35S]TBPS binding
to transfected cell membranes. Also, displacement of
[35S]TBPS binding by pentobarbital, yielded similar
Ki values for the two membrane preparations (9 and
10 µM, respectively; Fig. 3). However,
displacement of binding by pregnanolone demonstrated a significant
difference between the recombinant receptors. The concentration of
steroid that displaced binding from the
3 homomeric receptor by 50%
(3.5 µM) had no significant effect on binding to
membranes of
3/
transfected cells (Fig. 3). The
Ki values for the two membrane preparations were 2 and 18 µM, respectively. The same effect was observed
with human
3 and
subunit cDNAs, expressed in either HeLa or
HEK-293 cells (data not shown).
The physiological significance of this reduced sensitivity to
pregnanolone is unknown at present. To address this question, it will
be necessary to identify the subunits with which the subunit
combines in vivo and re-examine the properties of relevant subunit combinations in recombinant systems. However, this effect of
the
subunit provides further evidence that it can assemble with
known GABAA receptor subunits and alter the sensitivity of receptors to modulatory agents. The actions of pregnanolone at GABAA receptors have been proposed to regulate uterine
motility by inhibiting contractions (13, 14). A quiescence of uterine motility is essential for embryonic implantation and maintenance of
pregnancy (26). The identification of a novel subunit class in the
uterus is therefore likely to provide new insights to the role of
GABAA receptors in this important process.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U95367[GenBank] and U95368[GenBank].