Molecular Biology and Evolution

MBE Advance Access originally published online on November 29, 2006
Molecular Biology and Evolution 2007 24(2):599-610; doi:10.1093/molbev/msl188

This Article Abstract FREE Full Text (PDF) Supplementary Material All Versions of this Article: 24/2/599    most recent msl188v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Tsang, S.-Y. Articles by Xue, H. Search for Related Content PubMed PubMed Citation Articles by Tsang, S.-Y. Articles by Xue, H. Social Bookmarking          What's this?

© The Author 2006. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org

Research Articles

The Evolution of GABA_A Receptor–Like Genes

Shui-Ying Tsang^*, Siu-Kin Ng^*,, Zhiwen Xu^* and Hong Xue^*

^* Department of Biochemistry and Applied Genomics Laboratory, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong
Graduate Program of Bioengineering, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong

E-mail: hxue{at}ust.hk.

	Abstract

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
The inhibitory ligand-gated ion channel family of receptors, including the type A -aminobutryic acid (GABA_A) and glycine receptors, mediates inhibitory neurotransmissions in the central nervous system. In this study, GABA receptor (GABR) evolution was explored through comparative genomics using genomes that span divergent lineages. GABA_A/Gly receptor–like (GRL) gene sequences were retrieved from the genomes of various species ranging from mammal to fish to worm and subjected to cross-species comparison. All vertebrate GRL gene sets in the study but no invertebrate ones exhibit the extensive and conserved pattern of gene clustering that is characteristic of human GABR genes, indicating that the gene clusters were established early in vertebrate evolution, after divergence from the invertebrates. Moreover, the vertebrate gene structure is highly conserved with a basic 9–coding exon structure, whereas, as well as being diverse in copy numbers and chromosomal loci, the invertebrate GRL genes display a variety of gene structures. Remarkably, the invertebrates each possess a unique GRL gene pair that lies in neighboring loci within their respective genomes: zc482.5 and zc482.1 in roundworm, CG8916 and CG17336 in fruitfly, Ci4249 and Ci4254 in Ciona, and these were revealed by phylogenetic analysis to be homologous to human GABR and ß subunits, respectively. The phylogenetic classification of these genes is also corroborated by experimental ligand-binding measurements using recombinant gene products. Furthermore, the 3 invertebrate gene pairs harbor characteristic key residues and exhibit similarities in intron positions to their vertebrate counterparts. The results strongly indicate that such a gene pair originally existed in the bilaterian ancestor from which all 3 phyla evolved and suggest that the extant GABR clusters arose from an ancestral –ß subunit gene pair gave rise to the extant GABR clusters.

Key Words: comparative genomics • GABA_A receptors • evolution • gene structure • bilaterian ancestral genes

	Introduction

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
Neurotransmitter receptors convert chemical into electrical signals at the postsynaptic membranes of neuronal synapses. The most prevalent forms of neurotransmitter receptors are Cys-loop ligand-gated ion channels (LGIC), both cationic excitatory Cys-loop LGIC gated by acetylcholine and serotonin and anionic inhibitory Cys-loop LGIC (ILGIC) comprising the type A -aminobutryic acid (GABA_A) receptors and glycine receptors (GlyR) (Betz 1990; Xue 1998). There are high amino acid sequence similarities within the 2 divisions of Cys-loop LGIC receptors (Grenningloh et al. 1987), which have been linked to a variety of human pathologies (Baulac et al. 2001; Lo et al. 2004), and constitute the targets of much pharmaceutical usage and design (Whiting 2003). The vertebrate ionotropic glutamate receptors are also excitatory LGIC, but their sequence and structure differ significantly from those of the Cys-loop LGIC and they are not included in the superfamily.

Cys-loop LGIC receptors are assembled from a combination of homologous subunits encoded by separate genes (Schofield et al. 1987). In the human genome, there are as many as 24 genes encoding ILGIC subunits, consisting of 19 GABA_A receptor (GABR) and 5 GlyR genes. The mapping of GABR genes began with the assignment of GABRA1 to chromosome 5q34, GABRA3 to Xq28, and the GABRA2–GABRB1 pair to 4p12 (Bell et al. 1989; Buckle et al. 1989). The location and distribution of all 24 genes are now fully assigned, and 14 of them are arranged in gene clusters (two 4-gene clusters and two 3-gene clusters). Because the GABRQ-encoded and GABRE-encoded polypeptides are generally considered to be "ß-like" and "-like", respectively (Sinkkonen et al. 2000), the four clusters thus contain 1 ß like, 1 like, and 1 or 2 subunit genes. In addition, the clusters are similar in terms of both gene order and transcriptional orientation of the member genes, the gene order being ß, , (), with the ß subunit gene in a transcriptional orientation opposite to that of the other genes. This unusual cluster arrangement, also found in other mammalian genomes, could have arisen from an ancestral ß–– subunit cluster through 2 rounds of whole-genome duplication (WGD) early in vertebrate evolution (Bailey et al. 1999; Russek 1999; Simon et al. 2004; Darlison et al. 2005).

Genes with sequence similarity to vertebrate ILGIC genes are also present in invertebrate genomes. In both the nematode Caenorhabditis elegans (roundworm) and the arthropod Drosophila melanogaster (fruitfly), a number of such ILGIC-like genes have been identified (Lee et al. 2003; Schuske et al. 2004). The unc-49 gene in the roundworm encodes 3 distinct alternative variants, and the homomers formed by combinations of these variants were found to display anion channel functionality (Bamber et al. 1999, 2003). Similarly, functional studies on Rdl and Lcch3 genes in the fruitfly have revealed their anion channel properties. Alternative splicing of Rdl yields 4 variants, and the functional homomers formed by these variants yielded differing agonist potencies (Hosie et al. 2001). Heteromers formed by RDL and LCCH3 subunits were also functional chloride channels (Zhang et al. 1995). However, those formed by GRD (encoded by the fruitfly Grd gene) and LCCH3 exhibited GABA-gated cation channel properties (Gisselmann et al. 2004), and the roundworm gene exp-1 also encodes for a novel GABA-gated cation channel (Beg and Jorgensen 2003). The major difference between this EXP-1 protein and conventional GABRs resides in the pore-forming domain (Schuske et al. 2004).

Recent advances in genomic sequencing have opened the way to novel insights into gene evolution through cross-species genomic comparisons (Boffelli et al. 2004). In addition to the completely sequenced human and mouse genomes, progress is made in the genomes of a number of other species including vertebrates, such as chimpanzee and chicken, and invertebrates, such as roundworm and fruitfly. The genomes of 2 species of Tetraodontiformes were also sequenced (Aparicio et al. 2002; Jaillon et al. 2004): Takifugu rubripes (FUGU) and Tetraodon nigroviridis (pufferfish) are particularly attractive for comparative genomics in that their genomes are about one-eighth the length of the human genome but carry a similar complement of protein-coding genes (Mulley and Holland 2004). A draft of the protein-coding portions of the genome of the most studied ascidian, Ciona intestinalis (Ciona), has also been generated (Dehal et al. 2002). This invertebrate chordate, endowed with single-copy genes in place of most vertebrate gene families, represents a plausible approximation to ancestral chordates, bridging between vertebrates and invertebrates. Previously, the major phylogenetic branches of ILGIC have been analyzed (Xue 1998). In the present study, advantage is taken of this wealth of genomic data, spanning 500 Myr of evolution of bilaterian animals, protostomes and deuterostomes, to gain insights into ancestral Cys-loop LGIC genes present in the bilaterian ancestor.

	Materials and Methods

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
Database Retrieval of Gene Sequences
Genomic data of 3 vertebrates and 3 invertebrates were obtained from the Ensembl genome browser (Birney et al. 2006). Sequences designated in the InterPro domain IPR006028 (gamma-butyric acid receptor A) were retrieved. Because genes with sequence similarity with GABR or GlyR genes do not necessarily encode inhibitory anion channels, the term GABA_A/Gly receptor like (GRL) hereafter is adopted to refer to the genes and their GABA/Gly-gated functions. By excluding those sequences with definitive annotations for proteins other than GABR or GlyR in the vertebrate genomes, 24 GRL gene sequences were identified in the most recent human assembly (National Center for Biotechnology Information build 35), consisting of 19 GABR genes and 5 GlyR genes. GRL gene sequences were similarly retrieved from the roundworm (WS140), fruitfly (BGDP 4), Ciona (JGI1.95), pufferfish (TETRAODON7), and chicken (Gallus gallus; WASHUC1) genomes. The pufferfish sequences retrieved, unannotated, were assigned based on database-predicted reciprocal hits against the human genome. The invertebrate sequences were likewise largely unannotated. Because their similarity with human sequences was low, no definitive reciprocal hits were obtained against the human genome. They were therefore described only by their database gene names.

Phylogenetic Analysis of GRL Subunits
GRL peptide sequences of human (8 sequences), Ciona (5), roundworm (34), and fruitfly (8) origin were aligned using ClustalX (Chenna et al. 2003). Human acetylcholine (CHRNA1 and CHRNB3) and serotonin (HTR3A and HTR3C) receptor sequences were included as outgroup. Two hundred and seven informative sites were selected (with nonalignable and noninformative sites removed manually) for phylogenetic analysis using the PHYLIP package (Retief 2000). 1 hundred replications were made to the data set by SEQBOOT resampling, and the protein distance matrices were estimated using PROTDIST with the JTT model of amino acid substitution. Phylogenetic trees were generated using FITCH, and the final rooted consensus tree was built using CONSENSE with the majority-rule option.

Construction of Expression Plasmids
Subcloning and deletional mutagenesis were performed with the polymerase chain reaction (PCR)–based Mutagenesis Kit from Stratagene (La Jolla, CA) except that Pfu DNA polymerase was used and DpnI treatment was replaced by gel purification of the linear PCR products. Two 1.5-kb sequences containing the cDNAs encoding the mature peptides of roundworm GABR subunits (Ala16–Lys500 and Val62–Val524 of zc482.1 and f07b10.5, respectively) were amplified from the roundworm cDNA library (Resgen, Invitrogen, Carlsbad, CA) using the oligonucleotides shown in supplementary table 1 (Supplementary Material online). The sequences were cloned into pTrcHis (Invitrogen) after removal of the upstream vector sequence except for the minicistron as previously described (Hang et al. 2000). Based on the initial subclones, the deletional mutants encoding only the extracellular domains were made using the PCR primers shown in supplementary table 1 (Supplementary Material online) with appropriate termination codons. The fidelity of the final constructs was verified by DNA sequencing.

Protein Expression
Roundworm GABR recombinant proteins were expressed and purified as described (Shi et al. 2003; Xu et al. 2005). Protein concentrations were determined by OD₂₈₀ in the presence of 6 M GdCl, with the extinction coefficient ₂₈₀ calculated from the amino acid composition (Pace et al. 1995).

Fluorescence Anisotropy Titrations and Data Analysis
Fluorescence measurements were carried out at room temperature using a PerkinElmer model LS50B luminescence spectrometer. In the saturation experiments, 0.02 µM Bodipy-FL Ro1986 (Molecular Probes, Inc., Eugene, OR) in 10 mM glycine, pH 9.6, was titrated with protein. The fluorescence anisotropy (FA) was measured at excitation wavelength 490 nm and emission wavelength 511 nm, both with a 5-nm slit. The fraction of bound ligand was estimated from the increase in FA with increment of protein (van den Elsen et al. 1997). The dissociation constant k_D was estimated from FA saturation curves by nonlinear least-squares fit to a single-site binding model and linear fit of the Scatchard transformation.

	Results

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
GRL Genes in Vertebrate Genomes
Although the majority of vertebrate GRL genes display relatively conserved transcript lengths of 415–500 amino acids (supplementary table 2, Supplementary Material online), their actual gene sizes vary greatly, ranging from 11 to 652 kb in human and 3 to 27 kb in pufferfish. These different gene sizes parallel the different genome sizes (human, 3272 Mb; pufferfish, 402 Mb) and reflect among other factors large differences in intron lengths. Gene clusters homologous to the 4 human GRL gene clusters (Simon et al. 2004) are present in the other 2 vertebrate genomes, and the GRL gene sets as a whole are grouped on a small fraction of chromosomes (fig. 1A). The 24 human GRL genes are spread over only 7 of the 24 chromosomes, those in the chicken are spread over only 6 of the 34 chromosomes, and of the 28 pufferfish GRL genes identified, 20 have been assembled onto 7 of the 21 chromosomes.

View larger version (35K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 1.— Chromosomal arrangement of GRL genes. The chromosomal arrangement and transcriptional orientation of GRL genes in the vertebrate genomes of human, chicken, and pufferfish (A) and the invertebrate genomes of roundworm and fruitfly (B) are shown. Only those pufferfish genes with known chromosomal positions are included. The 4 clusters in the human genome and corresponding clusters in the chicken and pufferfish genomes are in shaded boxes. The 2 unique gene pairs in the fruitfly and roundworm genomes are also in shaded boxes. Where only gene fragments are given in the Ensembl database, the genes are shown in gray italics. The forward (+) and reverse (–) transcriptional orientations of the genes are depicted.

 
In the chicken genome, a complete set of homologs to the human genes and possibly an extra copy of GABRP were found. However, instead of the and subunit genes, the chicken genome harbors the ß₄ and ₄ subunit genes, respectively, which are orthologous by virtue of sequence identity as well as relative position in the gene clusters (fig. 1A). The ortholog to human GABRB1 is truncated in chicken, possibly indicating that the ß₁ subunit gene has evolved into a pseudogene (Darlison et al. 2005). The orthologs of the clusters on human chromosomes 4 and X are both located on chicken chromosome 4 (fig. 1A), pointing to regional synteny between chicken chromosome 4 and human chromosomes 4 and X. Moreover, the orthologs of the GlyR genes GLRA3 and GLRB on human chromosome 4 and GLRA4 on human chromosome X are again found on chicken chromosome 4, further supporting the possible synteny. Similarly, the GRL genes on human chromosome 5 are orthologous to those on chicken chromosome 13, and the GRL gene cluster on human chromosome 15 and the GRL gene pair on human chromosome 6 have orthologs on chicken chromosomes 1 and 3, respectively, indicating additional regions of synteny. The relative order and orientation within the gene clusters are much the same in the 2 species: the genes follow the order ß, , (), , with the direction of the ß subunit gene always being opposite to that of the other members in the cluster.

In the pufferfish, some homologs of human GRL genes have as yet undetermined chromosomal positions, and others are missing (supplementary table 2, Supplementary Material online). Some identified homologs are mere fragments that could belong to pseudogenes, and there are some duplicate copies of homologs (fig. 1A). Because only 92% of the pufferfish genome has been sequenced and only 64% of the sequenced DNA mapped to chromosomes, the missing homologs to human GABRE, GABRB1, and GABRG1 genes may be present but are as yet unidentified. Because the missing genes would belong to gene clusters, they are expected to occur near other cluster members. However, a Blast search in such neighborhoods has not revealed their presence. The same genes were also found to be missing from the genome of FUGU, another teleost, suggesting that these genes are absent from these teleosts as well as their common ancestor.

Altogether, there are 2 clusters of pufferfish GRL genes, that homologous to the human GABRB2–GABRA6–GABRA1–GABRG2 cluster on chromosome 7 and the GABRG3–GABRA5–GABRB3 cluster on chromosome 3 (fig. 1A). A GABRB2–GABRA6 homolog and a second GABRG3–GABRA5–GABRB3 homolog cluster are located on pufferfish chromosome 1 and 1_random (consisting of nonassembled contigs of chromosome 1), respectively. Notably, the gene order and orientation in the pufferfish clusters are the same as those in the human clusters, except that the ß subunit gene does not oppose its subunit neighbor in the "half-cluster" on pufferfish chromosome 1. In addition, the pufferfish contains homologs to human GABRA2 and GABRA4 (unknown chromosomal position), and GABRQ–GABRA3, but not full-fledged equivalents to the clusters on human chromosomes 4 and X.

The vertebrate GRL genes, apart from the pufferfish GlyR homologs, exhibit gene structures that are highly conserved both in terms of exon number and intron position. Only their intron lengths are widely divergent, for example, 300 kb in human GABRG3 intron 3, but 0.1 kb in GABRD (fig. 2). There are 9 coding exons in the majority of vertebrate GRL genes. Within the 9-exon structure, the lengths of exons 3–7 are the most conserved (fig. 2). These exons encode for most of the extracellular region, the first transmembrane region, and through to part of the second transmembrane region. Exon 3 (68 bps) and exon 5 (83 bps) are invariant in the vertebrate homologs (except for pufferfish GlyRs). The conserved nature of exon 5, harboring the signature Cys-loop, has previously been employed as one of the criteria for identifying novel members of the human GABR family (Simon et al. 2004). In contrast, the positions of exon–intron boundaries at both the 5' and 3' ends of the genes are more variable.

View larger version (12K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 2.— The conserved gene structure of GRL genes. The regions corresponding to exons 3–7 of the conserved 9-exon structure of human GABR and GlyR are shown. The exons are labeled E3– E7 with the corresponding lengths given (in bp) underneath the boxed exons. In cases where exons of certain subunits have length other than that common to most subunits, these exceptions are shown in italics with the relevant subunits given in brackets. These subunits include GABRP (), GABRR1–3 (), GABRD (), GABRA1–6 (), GLRB (ß), and GLRA4 (4), where indicated. The numbers above the sequence line in the intron positions represent the lengths of introns 3–6 taken from the shortest and longest human subunits GABRD and GABRG3, respectively. The corresponding region for 2 Ciona GRL genes (Ci2641 and Ci4254) is also shown to illustrate an intron gain within E4 and the resultant split exon.

 
GRL Genes in Invertebrate Genomes
The numbers of GRL genes varied considerably among the invertebrate genomes from different phyla. Thirty-nine GRL genes were found in the roundworm genome (gene lengths ranged from 2.3 to 12.9 kb and transcript lengths from 400 to 679 amino acids) and 12 in the fruitfly genome (gene lengths ranged from 1.9 to 44.5 kb and transcript lengths from 422 to 686 amino acids; supplementary table 3, Supplementary Material online). In addition, 5 GRL genes were retrieved from the invertebrate chordate Ciona (gene lengths ranged from 4.3 to 14.0 kb and transcript lengths from 422 to 430 amino acids). The gene and transcript lengths of these invertebrate GRL genes all fell within the same range, with notably less transcript length variations in the Ciona genes. A cluster of 6 genes on roundworm chromosome X (t01h10.1–t01h10.7, not included for further study) was classified as ambiguous because, although they bear some sequence similarity to Cys-loop LGIC genes, they could not be distinguished as GRL or acetylcholine receptor–like channels. These genes might be ancient, undifferentiated forms of Cys-loop LGIC genes, or they might have lost their original character through evolution.

The GRL genes in the roundworm and fruitfly are dispersed over many chromosomal loci, with no more than 2 GRL genes in close proximity to each other (supplementary table 3, Supplementary Material online; fig. 1B). The gene pair zc482.1 and zc482.5 display opposing transcriptional orientations on roundworm chromosome III, whereas the CG8916 and CG17336 pair display the same orientation on fruitfly chromosome X. Because the sequence data of the Ciona genome have only been assembled on to scaffolds, the exact chromosomal positions of its 5 GRL genes are unknown. Nevertheless, Ci4249 and Ci4254 are located on the same scaffold, close to each other and in opposing transcriptional orientations. Therefore, a unique pairing of GRL genes occurs in all 3 invertebrate genomes.

The number of exons varies from 5 to 16 in the roundworm GRL genes and between 2 and 9 in the fruitfly GRL genes (supplementary table 3, Supplementary Material online). Intron positions also vary extensively. In contrast, Ciona GRL genes possess either 9 or 10 exons (except for Ci2641, which has 12 exons), and there is also more conservation in intron positions in the region encoding the putative extracellular domain. The intron positions that show significant conservation are similar to those of vertebrate introns 2–6 (numbered according to the vertebrate 9-exon structure; fig. 3, positions a–e), except for Ci4249 at position b and Ci9266 at position c. There is also an extra intron between positions b and c in all 5 Ciona genes (positions underlined in fig. 3). Interestingly, when this region of the unique gene pair from both roundworm and fruitfly were aligned (fig. 3), the corresponding intron positions are also similarly conserved despite some intron loss.

View larger version (67K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 3.— Multiple sequence alignment of GRL proteins showing conserved intron positions. Partial sequences of the 5 Ciona GRL protein subunits and the transcription products of the unique fruitfly and roundworm gene pairs were aligned along with a representative human GRL subunit protein sequence GABRB1 using the ClustalX program (Chenna et al. 2003). The exon–intron boundaries as given in the Ensembl database are in boldface and highlighted. The positions of conserved boundaries in the invertebrate sequences are boxed and labeled a–e and correspond to the positions of introns 2–6 of human GABRB1. Positions of "extra" introns (no equivalent in the vertebrate gene) are underlined. The numbering given is according to the ClustalX alignment.

 
Remarkably similar exon arrangements are also found in a few other roundworm and fruitfly genes. For example, fruitfly CG10537 has 9 exons, and its 3–4–5 exons have lengths of 68–227–83 nucleotides, respectively. The roundworm c09g5.1 has 8 exons, and its 2–3–4 exons have lengths of 68–221–83 nucleotides, respectively. However, despite this similarity in exon number and exon lengths, these fruitfly and roundworm genes display low identity with the vertebrate GRL genes, showing only 30% sequence identity.

Phylogenetic Analysis
A phylogenetic tree, obtained earlier using a selection of vertebrate and invertebrate GRL sequences (Xue 1998), provided a means to classify GRL genes: and subunit–like sequences are assigned to class I; ß, , , and subunit–like sequences to class II; GlyR-like sequences and Glu receptor–like sequences that form anion channels to class III; and invertebrate sequences outside of the other 3 classes to class IV. The phylogenetic tree generated in the present study using amino acid sequences of GRL subunits is classified accordingly (fig. 4). Each of the Ciona subunits is homologous to human counterparts and thus can be readily classified as follows: Ci4249 homologous to GABRA1 and GABRG1 (class I); Ci2641 to GABRR1, Ci5403 to GABRP and GABRD, and Ci4254 to GABRB1 (all class II); and Ci9266 to GLRA3 and GLRA4 (class III).

View larger version (24K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 4.— Phylogenetic relationship of invertebrate GRL sequences with representative human homologs. The bootstrap majority-rule consensus tree was obtained from 100 replicates of 47 invertebrate and 8 human GRL protein sequences using the human acetycholine and serotonin receptor sequences as outgroup and 207 informative sites. The programs SEQBOOT, PROTDIST, FITCH, and CONSENSE from the PHYLIP package (Retief 2000) were used in the construction as described in Materials and Methods. The numbers at the branch nodes give the bootstrap values of 100 replications. The human and Ciona sequences are prefixed with hs and Ci, respectively, and the fruitfly and roundworm sequences are labeled with the gene names used in the Ensembl database. The sequences are divided into clades A–H, and the 4 major classes I–IV (Xue et al. 1998) are highlighted. Class III contains clades D and E, class IV contains clades F–H, and clade C is intermediate of classes I–III.

 
From the topology of the tree, a number of roundworm and fruitfly subunits can also be assigned to classes I–III: roundworm subunits zc482.5 and f07b10.5 and fruitfly subunits CG7446 and CG8916 belong to class I, whereas roundworm zc482.1 and fruitfly CG17336 belong to class II. Clade E that includes the 7 genes from c27h5.8 and CG7535–f11a5.10 contains proteins that are annotated as Glu-gated chloride channels in the Ensembl database and so belong to class III similar to the vertebrate Gly-gated chloride channels. No invertebrate GlyR has been identified, and invertebrate Glu receptors may be regarded as the invertebrate equivalent of vertebrate GlyR (Vassilatis et al. 1997). On the other hand, clade C with subunits t21c12.1, f11h8.2, and CG10537 cannot be definitively assigned to the classes I–III. However, the demonstrated GABR properties of the t21c12.1 and CG10537 subunits (Zhang et al. 1995; Bamber et al. 2003) and the depicted phylogeny support a close relationship between this clade and the vertebrate GRL subunits, placing them as intermediate of the 3 classes.

The other roundworm (clades F and G) and fruitfly (clade H) subunits in the consensus tree are assigned to class IV, representing more ancient and poorly differentiated genes compared with those in the other 3 classes. The separation of the roundworm genes into 2 clades suggests that there are 2 distinct classes of receptors. From the tree topology, the genes in the deepest branching G and H clades appear to be more ancient than those in F, but the functional significance of most of these genes remains unknown. Of the few that have been studied, k06c4.6 (mod-1) is a serotonin-gated chloride channel (Ranganathan et al. 2000; Menard et al. 2005), whereas c53d6.3, f55d10.5, f58g6.4, and t27e9.9 are reported to be acetylcholine-gated chloride channels (Putrenko et al. 2005).

Classification of Invertebrate Subunits
The classification of various invertebrate subunits into classes I and II in figure 4 is based on their homologies to the GABR and subunits and ß subunits, respectively. Although the Ciona Ci4249 and Ci4254 gene pair encodes for proteins with over 50% sequence identity with human and ß subunits, respectively, sequence identities for the roundworm (30%) and fruitfly (40%) subunits are lower (table 1). Moreover, whereas the identities between the Ciona or fruitfly subunits and the human subunit differ by more than 10% from those with the human ß subunit and are supportive of the given classification, this is not the case for the roundworm subunits, where the difference is marginal.

View this table: [in this window] [in a new window]   Table 1. Comparison of Identity between GRL Protein Sequences

 
Close inspection of the vertebrate protein sequences reveals key residues that distinguish between class I and class II proteins. Figure 5 shows a conserved region where these key residues are highlighted. These defining residues are also found in invertebrate proteins, where they readily classify the proteins into the separate subtypes in total agreement with the topology of the consensus tree. Therefore, although the overall identity between invertebrate and vertebrate or ß subtypes is not high, their classification is supported by the conservation of key residues. In addition, key residues reveal that invertebrate class I sequences are more subunit like than subunit like: invertebrate subunits contain exclusively Gly at positions where vertebrate subunits contain Gly, whereas subunits contain Ser (fig. 5). This suggests that subunit equivalents were originally absent from invertebrates and evolved later only within the vertebrates.

View larger version (35K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 5.— Consensus sequence of GRL proteins. Nineteen human GABR subunit protein sequences and 10 invertebrate homologs, 4 from Ciona, 3 from roundworm, and 3 from fruitfly were aligned with ClustalX. A conserved region of the alignment (without gaps) covering residues 108–199 is shown (numbering based on precursor sequence of human GABR 1 subunit). Only 6 of the 19 human sequences are presented. The underlined sequences (red) are either vertebrate class I members or invertebrate sequences clustered with class I members in the consensus tree shown in figure 4. The sequences named in boldface (blue) are those of vertebrate class II members or invertebrate homologs clustered with them in the consensus tree. Residues characteristic to class I vertebrate genes are underlined (-like; red) and those to class II are in boldface (ß-like; blue). Characteristic residues that do not apply to the entire class but can be used to distinguish between the classes are given in italics in the - and ß-like consensus sequence. Residues common to classes I and II in both vertebrate and invertebrate gene sets are in dark gray (consensus; green).

 
The classification was further ascertained by an examination of the functional properties of the gene products. For this purpose, the benzodiazepine (BZ)-binding fragments of roundworm subunits zc482.1 and f07b10.5 were cloned and expressed (fig. 6). The fragment from f07b10.5 yielded a BZ-binding affinity of k_D = 4.4 µM (cf. k_D = 2.2 µM for the corresponding fragment from human GABRA1) and that from zc482.1 gave a much reduced affinity of k_D = 10.5 µM (cf. k_D = 3.6 µM for the corresponding fragment from human GABRB2). Because class I subunits are more relevant to BZ binding than class II subunits (Xue 1998), these ligand measurements confirmed the phylogenetic sequence classification of f07b110.5 into class I and zc482.1 into class II (fig. 4).

View larger version (26K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 6.— BZ-binding properties of recombinant roundworm protein fragments. (A) Sodium dodecyl sulfate–polyacrylamide gel electrophoresis analysis of extracellular domain of roundworm zc482.1 and f07b10.5 subunits. Lane 1: molecular weight marker; lane 2: zc482.1; and lane 3: f07b10.5. (B) Saturation curves of FA and Scatchard transformations (inset) of Bodipy-FL Ro1986 (BFR) binding to the extracellular domains. FA was measured with excitation and emission wavelengths at 490 and 511 nm, respectively. The BFR–protein dissociation constants kD for the zc482.1 and f07b10.5 fragments are 10.5 ± 0.9 and 4.4 ± 0.4, respectively.

 

	Discussion

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
The utility of comparative genomics rapidly gains in importance as the wealth of extractible information increases with the number of completed genomes. Previously, major phylogenetic branches of ILGIC receptors were identified by comparing amino acid sequences from a selection of vertebrate and invertebrate genomes (Xue 1998). The availability of additional genomic sequences has made possible an expanded analysis of the GRL genes.

In the present study, complete GRL gene sets were extracted from the 3 vertebrate genomes of humans, chickens, and pufferfish and the 3 invertebrate genomes of sea squirt (Ciona), fruitfly, and roundworm. The set of GRL genes in the human genome find homologs in other mammalian genomes such as mouse and rat (Milani et al. 1998; Simon et al. 2004). In the recently published chimpanzee and canine genome sequences, very similar gene sets likewise are present. Because all mammalian GRL gene sets are similar to one another in terms of copy number, gene structure, and coding sequence, the human GRL gene set was selected as the only mammalian representative for further analysis along with the 2 nonmammalian vertebrate representatives chicken (avian) and pufferfish (teleost). The ascidian Ciona is included in addition to the other 2 model representatives of invertebrate species because it occupies a unique evolutionary position as an invertebrate chordate, bridging between the vertebrate and invertebrate lineages, and hence sharing properties with both lineages. Moreover, its gene complement can be regarded as a first approximation to the ancestral chordate complement (Mulley and Holland 2004).

With the vertebrate GRL genes, there is considerable similarity in number, gene structure, and chromosomal arrangement between fish, avian, and mammals. The majority of vertebrate GRL genes have 9 coding exons with the lengths of exons 3–7 being the most conserved. There are whole or partial GRL gene clusters in each genome, with a distinctive cluster arrangement of ß, , (), and the ß subunit gene orientated opposite to the other member genes. These clusters are suggested to originate from 2 WGDs early in vertebrate evolution (Darlison et al. 2005). In the teleosts, the presence of partial clusters may be explained by the loss of the GABRB1, GABRG1, and GABRE genes from the pufferfish and FUGU subsequent to cluster duplication, reducing the size of the original clusters, whereas the duplicate clusters likely originated from the WGD event early in teleost evolution, specific to the ray-finned fishes (Jaillon et al. 2004; Vandepoele et al. 2004). Conservation of the cluster arrangement throughout vertebrate evolution underlines its importance, and the factors contributing to the conservation might be more complex than coordination of expression (Simon et al. 2004). It seems, therefore, quite clear that the vertebrate gene set was formed early in vertebrate evolution and underwent surprisingly little change over some 420 Myr of evolution.

There is considerable variation in the number and gene structure of GRL genes from the 3 invertebrate genomes. The differences in copy numbers pointed to lineage-specific events, especially in the roundworm where extensive duplication, translocation, and differentiation led to a large diversity of novel receptors whose functions remain unclear. Such bursts of lineage-specific duplications in the roundworm have also been reported for nuclear receptors (Bertrand et al. 2004). Unlike the majority of fruitfly and roundworm GRL genes, a certain degree of conservation in the intron positions is apparent for the Ciona genes (fig. 3). These conserved splicing positions coincide with vertebrate introns 2–6 that define the boundaries of lengthwise most conserved vertebrate exons 3–7. Compared with the vertebrate genes, the Ciona genes have an extra intron in this region. These extra introns might be a result of "intron slide" in the case of Ci4249 and Ci9266, brought about by the loss of an ancestral intron and the subsequent gain of a new intron at a distinct site (Stoltzfus et al. 1997). In Ci5403, Ci4254, and Ci2641, the extra introns could have arisen from intron gain, splitting the equivalent of exon 4 in the vertebrate 9-exon structure into 2 separate exons (fig. 2).

The characteristic cluster arrangement displayed by the vertebrate GRL genes is not found in the invertebrate gene sets. Instead, the invertebrate genomes each possess a unique gene pair (zc482.5 and zc482.1 in roundworm; CG8916 and CG17336 in fruitfly; and Ci4249 and Ci4254 in Ciona). Interestingly, in the Ciona and roundworm, these gene pairs take on opposing transcriptional orientations, as in the case of vertebrate GABR and ß subunit genes within the gene clusters. Although the gene structures of the roundworm and fruitfly GRL gene pairs depart from the conserved vertebrate 9-exon structure in terms of both exon number and exon lengths, they exhibit similarities with respect to intron positions (fig. 3), as noted above for the Ciona genes. The smaller exon numbers in the fruitfly gene pair could be caused by the loss of ancestral introns, as indicated by a lack of introns at such conserved splicing positions as positions a and b in CG8916 (fig. 3). This is in accord with the extensive intron loss observed in nematodes and arthropods, in sharp contrast to the remarkable conservation of ancestral introns in vertebrates and plants (Rogozin et al. 2003). In the roundworm, intron gain events might have countered the loss of ancestral introns, hence the larger exon numbers. The presence of GRL genes in all 3 invertebrate phyla suggests that ancestral GRL genes existed in their common ancestor, and indeed, the 5 Ciona genes could be the ancestral genes that gave rise to the vertebrate GRL genes. Moreover, the conservation of intron positions as far back as nematodes and arthropods suggests that the basic gene structure of GRL genes must be already in place in the early chordates, perhaps even in the bilaterian ancestor.

The phylogenetic tree generated using the protein sequences of GRL subunits is entirely in accord with the earlier study (Xue 1998), despite the absence of Ciona and some roundworm genes and the inclusion of more vertebrate genes in the earlier study (genes were annotated with SwissProt accession numbers in the earlier study, but with gene names in the present study). In the earlier study, the grouping of chloride channels gated by Gly or Glu as class III was based on the close evolutionary distance between the 2 subtypes and their functional distinction from the inhibitory GABA-gated channels (class I and class II) as well as from the less differentiated receptors in class IV. The grouping is reflected in the present study, although separation of the 2 subtypes into distinct clades (clades D and E) was more obvious than in the earlier study. In addition, although the distinct branching within class IV was not so apparent earlier, it is clear that these genes are not close homologs to extant vertebrate GRL genes and do not belong to classes I, II, or III. In a recent phylogenetic study on the Cys-loop LGIC superfamily, it has been hypothesized that these subunit types were present in the bilaterian ancestor, but subsequently lost from the chordate lineage (Dent 2006).

Most interestingly, the subunits encoded by the unique gene pair in each of the invertebrate genomes are respectively homologous to the extant vertebrate and ß subunits. Apart from 2 other -like subunits, no other roundworm or fruitfly GRL subunit was so closely homologous to extant GABR subunits so as to be definitively assigned to class I or II. The phylogenetic classification of these genes is also corroborated by the presence of key characteristic residues as well as experimental ligand-binding measurements using recombinant gene products, where the class I roundworm f07b10.5 protein fragment showed higher BZ ligand–binding affinity than the class II roundworm zc482.1 protein, in agreement with the reported involvement of vertebrate / subunits (class I) in BZ binding. Thus one ancient subunit–like gene and one ancient ß subunit–like gene lie very close to one another in the genomes from distinct phyla. This strongly indicates that such a gene pair originally existed in the bilaterian ancestor from which all three phyla evolved and suggests that the extant GABR clusters arose from an ancestral –ß subunit gene pair.

In addition to the –ß subunit gene pairs, at least one Glu/Gly receptor gene is found in each of the 3 invertebrates, suggesting that a Glu/Gly receptor gene also existed in the bilaterian ancestor. An evolution model for the diversification of GRL genes based on these 3 ancestral genes is proposed in figure 7. It involves 2 initial duplications of the ß subunit–like gene, the copies of which adapted to novel functions, becoming subunit–like and subunit–like. This would correspond to the evolutionary stage of the Ciona genome, close to the chordate ancestor. After divergence from the urochordates, the vertebrate subunit–like gene duplicated, forming the ancient subunit and the ancestral ––ß gene cluster in the vertebrate ancestor. Subsequently, the 4 gene clusters in the land vertebrates and other subunit genes arose through 2 vertebrate-specific WGD events (Dehal and Boore 2005). A third WGD accompanied by extensive subsequent gene loss likely shaped the extant teleost gene set (Vandepoele et al. 2004).

View larger version (23K): [in this window] [in a new window] [Download PowerPoint slide]   FIG. 7.— Proposed evolutionary model for vertebrate GRL genes. The hypothetical gene set in the bilaterian ancestor from which the vertebrate GRL genes evolved consists of 3 genes: an –ß subunit–like gene pair and a Glu/Gly receptor gene. These genes undergo single-gene duplications and WGD with some level of gene loss ( /) to become the extant vertebrate GRL gene sets. Other GRL genes present in the bilaterian ancestor are lost from the chordate lineage and become specific to the fruitfly and roundworm lineages. The given genomic locations of the clusters refer to the 4 present-day human clusters (see fig. 1).

 

	Conclusion

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
By comparing gene sets spanning over 500 Myr of evolution, the presence of unique –ß subunit–like gene pairs, namely, Ci4249 and Ci4254 in Ciona, CG8916 and CG17336 in fruitfly, and zc482.5 and zc482.1 in roundworm, as well as some of the conserved ancestral intron positions, has been identified, providing useful insights into both the likely nature of the bilaterian ancestral genes and the modes of their subsequent evolution in different descendant lineages. In vertebrates, the similarities in number, gene structure, and chromosomal arrangement between fish, avian, and mammals suggest that the GRL genes are core vertebrate genes. Moreover, the analysis points to the evolution of into extant vertebrate GABR clusters an ancestral subunit–like and ß subunit–like gene pair into extant vertebrate GABR gene clusters, with subsequent amplification by gene duplications through single-gene events as well as in WGDs. Overall, the findings suggest that ancestral forms of GRL genes, with structures comparable with extant genes, already existed in the ancestral bilaterians. The longevity and importance of the GRL gene superfamily is therefore clearly underlined.

	Supplementary Material

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
Supplementary tables 1–3 are available at Molecular Biology and Evolution online (http://www.mbe.oxfordjournals.org/).

	Acknowledgements

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 
We thank Professor J. Tze-Fei Wong for helpful discussion. Financial support by the Hong Kong Research Grants Council (Project no. DAG05/06.SC06) is gratefully acknowledged.

	Footnotes

 
David Irwin, Associate Editor

	References

TOP Abstract Introduction Materials and Methods Results Discussion Conclusion Supplementary Material Acknowledgements References

 

Aparicio S, Chapman J, Stupka E, et al. (2002) Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes. Science 297:1301–1310 (41 co-authors).[Abstract/Free Full Text]

Bailey ME, Matthews DA, Riley BP, Albrecht BE, Kostrzewa M, Hicks AA, Harris R, Muller U, Darlison MG, Johnson KJ. (1999) Genomic mapping and evolution of human GABA(A) receptor subunit gene clusters. Mamm Genome 10:839–843.[CrossRef][ISI][Medline]

Bamber BA, Beg AA, Twyman RE, Jorgensen EM. (1999) The Caenorhabditis elegans unc-49 locus encodes multiple subunits of a heteromultimeric GABA receptor. J Neurosci 19:5348–5359.[Abstract/Free Full Text]

Bamber BA, Twyman RE, Jorgensen EM. (2003) Pharmacological characterization of the homomeric and heteromeric UNC-49 GABA receptors in C. elegans. Br J Pharmacol 138:883–893.[CrossRef][ISI][Medline]

Baulac S, Huberfeld G, Gourfinkel-An I, Mitropoulou G, Beranger A, Prud'homme JF, Baulac M, Brice A, Bruzzone R, LeGuern E. (2001) First genetic evidence of GABA(A) receptor dysfunction in epilepsy: a mutation in the gamma2-subunit gene. Nat Genet 28:46–48.[CrossRef][ISI][Medline]

Beg AA and Jorgensen EM. (2003) EXP-1 is an excitatory GABA-gated cation channel. Nat Neurosci 6:1145–1152.[CrossRef][ISI][Medline]

Bell MV, Bloomfield J, McKinley M, Patterson MN, Darlison MG, Barnard EA, Davies KE. (1989) Physical linkage of a GABAA receptor subunit gene to the DXS374 locus in human Xq28. Am J Hum Genet 45:883–888.[ISI][Medline]

Bertrand S, Brunet FG, Escriva H, Parmentier G, Laudet V, Robinson-Rechavi M. (2004) Evolutionary genomics of nuclear receptors: from twenty-five ancestral genes to derived endocrine systems. Mol Biol Evol 21:1923–1937.[Abstract/Free Full Text]

Betz H. (1990) Ligand-gated ion channels in the brain: the amino acid receptor superfamily. Neuron 5:383–392.[CrossRef][ISI][Medline]

Birney E, Andrews D, Caccamo M, et al. (2006) Ensembl 2006. Nucleic Acids Res 34:D556–D561 (51 co-authors).[Abstract/Free Full Text]

Boffelli D, Nobrega MA, Rubin EM. (2004) Comparative genomics at the vertebrate extremes. Nat Rev Genet 5:456–465.[CrossRef][ISI][Medline]

Buckle VJ, Fujita N, Ryder-Cook AS, et al. (1989) Chromosomal localization of GABAA receptor subunit genes: relationship to human genetic disease. Neuron 3:647–654 (11 co-authors).[CrossRef][ISI][Medline]

Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, Thompson JD. (2003) Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res 31:3497–3500.[Abstract/Free Full Text]

Darlison MG, Pahal I, Thode C. (2005) Consequences of the evolution of the GABA(A) receptor gene family. Cell Mol Neurobiol 25:607–624.[CrossRef][ISI][Medline]

Dehal P and Boore JL. (2005) Two rounds of whole genome duplication in the ancestral vertebrate. PLoS Biol 3:e314.[CrossRef][Medline]

Dehal P, Satou Y, Campbell RK, et al. (2002) The draft genome of Ciona intestinalis: insights into chordate and vertebrate origins. Science 298:2157–2167 (87 co-authors).[Abstract/Free Full Text]

Dent JA. (2006) Evidence for a diverse Cys-loop ligand-gated ion channel superfamily in early bilateria. J Mol Evol 62:523–535.[CrossRef][ISI][Medline]

Gisselmann G, Plonka J, Pusch H, Hatt H. (2004) Drosophila melanogaster GRD and LCCH3 subunits form heteromultimeric GABA-gated cation channels. Br J Pharmacol 142:409–413.[CrossRef][ISI][Medline]

Grenningloh G, Gundelfinger E, Schmitt B, Betz H, Darlison MG, Barnard EA, Schofield PR, Seeburg PH. (1987) Glycine vs GABA receptors. Nature 330:25–26.

Hang J, Shi H, Li D, Liao Y, Lian D, Xiao Y, Xue H. (2000) Ligand binding and structural properties of segments of GABAA receptor alpha 1 subunit overexpressed in Escherichia coli. J Biol Chem 275:18818–18823.[Abstract/Free Full Text]

Hosie AM, Buckingham SD, Presnail JK, Sattelle DB. (2001) Alternative splicing of a Drosophila GABA receptor subunit gene identifies determinants of agonist potency. Neuroscience 102:709–714.[CrossRef][ISI][Medline]

Jaillon O, Aury JM, Brunet F, et al. (2004) Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype. Nature 431:946–957 (61 co-authors).[CrossRef][Medline]

Lee D, Su H, O'Dowd DK. (2003) GABA receptors containing Rdl subunits mediate fast inhibitory synaptic transmission in Drosophila neurons. J Neurosci 23:4625–4634.[Abstract/Free Full Text]

Lo WS, Lau CF, Xuan Z, Chan CF, Feng GY, He L, Cao ZC, Liu H, Luan QM, Xue H. (2004) Association of SNPs and haplotypes in GABAA receptor beta2 gene with schizophrenia. Mol Psychiatry 9:603–608.[CrossRef][ISI][Medline]

Menard C, Horvitz HR, Cannon S. (2005) Chimeric mutations in the M2 segment of the 5-hydroxytryptamine-gated chloride channel MOD-1 define a minimal determinant of anion/cation permeability. J Biol Chem 280:27502–27507.[Abstract/Free Full Text]

Milani N, Mulhardt C, Weber RG, Lichter P, Kioschis P, Poustka A, Becker CM. (1998) The human glycine receptor beta subunit gene (GLRB): structure, refined chromosomal localization, and population polymorphism. Genomics 50:341–345.[CrossRef][ISI][Medline]

Mulley J and Holland P. (2004) Comparative genomics: small genome, big insights. Nature 431:916–917.[CrossRef][Medline]

Pace CN, Vajdos F, Fee L, Grimsley G, Gray T. (1995) How to measure and predict the molar absorption coefficient of a protein. Protein Sci 4:2411–2423.[Abstract]

Putrenko I, Zakikhani M, Dent JA. (2005) A family of acetylcholine-gated chloride channel subunits in Caenorhabditis elegans. J Biol Chem 280:6392–6398.[Abstract/Free Full Text]

Ranganathan R, Cannon SC, Horvitz HR. (2000) MOD-1 is a serotonin-gated chloride channel that modulates locomotory behaviour in C. elegans. Nature 408:470–475.[CrossRef][Medline]

Retief JD. (2000) Phylogenetic analysis using PHYLIP. Methods Mol Biol 132:243–258.[Medline]

Rogozin IB, Wolf YI, Sorokin AV, Mirkin BG, Koonin EV. (2003) Remarkable interkingdom conservation of intron positions and massive, lineage-specific intron loss and gain in eukaryotic evolution. Curr Biol 13:1512–1517.[CrossRef][ISI][Medline]

Russek SJ. (1999) Evolution of GABA(A) receptor diversity in the human genome. Gene 227:213–222.[CrossRef][ISI][Medline]

Schofield PR, Darlison MG, Fujita N, et al. (1987) Sequence and functional expression of the GABA A receptor shows a ligand-gated receptor super-family. Nature 328:221–227 (12 co-authors).[CrossRef][Medline]

Schuske K, Beg AA, Jorgensen EM. (2004) The GABA nervous system in C. elegans. Trends Neurosci 27:407–414.[CrossRef][ISI][Medline]

Shi H, Tsang SY, Tse MK, Xu Z, Xue H. (2003) Recombinant extracellular domain of the three major subunits of GABAA receptor show comparable secondary structure and benzodiazepine binding properties. Protein Sci 12:2642–2646.[Abstract/Free Full Text]

Simon J, Wakimoto H, Fujita N, Lalande M, Barnard EA. (2004) Analysis of the set of GABA(A) receptor genes in the human genome. J Biol Chem 279:41422–41435.[Abstract/Free Full Text]

Sinkkonen ST, Hanna MC, Kirkness EF, Korpi ER. (2000) GABA(A) receptor epsilon and theta subunits display unusual structural variation between species and are enriched in the rat locus ceruleus. J Neurosci 20:3588–3595.[Abstract/Free Full Text]

Stoltzfus A, Logsdon JM Jr, Palmer JD, Doolittle WF. (1997) Intron "sliding" and the diversity of intron positions. Proc Natl Acad Sci USA 94:10739–10744.[Abstract/Free Full Text]

van den Elsen JM, van Pomeren E, Poolman JT, Wilting J, Herron JN, Crommelin DJ. (1997) On the interaction between a bactericidal antibody and a PorA epitope of Neisseria meningitidis in outer membrane vesicles: a competitive fluorescence polarization immunoassay. Anal Biochem 247:382–388.[CrossRef][ISI][Medline]

Vandepoele K, De Vos W, Taylor JS, Meyer A, Van de Peer Y. (2004) Major events in the genome evolution of vertebrates: paranome age and size differ considerably between ray-finned fishes and land vertebrates. Proc Natl Acad Sci USA 101:1638–1643.[Abstract/Free Full Text]

Vassilatis DK, Elliston KO, Paress PS, Hamelin M, Arena JP, Schaeffer JM, Van der Ploeg LH, Cully DF. (1997) Evolutionary relationship of the ligand-gated ion channels and the avermectin-sensitive, glutamate-gated chloride channels. J Mol Evol 44:501–508.[CrossRef][ISI][Medline]

Whiting PJ. (2003) GABA-A receptor subtypes in the brain: a paradigm for CNS drug discovery? Drug Discov Today 8:445–450.[CrossRef][ISI][Medline]

Xu Z, Fang S, Shi H, et al. (2005) Topology characterization of a benzodiazepine-binding beta-rich domain of the GABAA receptor alpha1 subunit. Protein Sci 14:2622–2637.[Abstract/Free Full Text]

Xue H. (1998) Identification of major phylogenetic branches of inhibitory ligand-gated channel receptors. J Mol Evol 47:323–333.[CrossRef][ISI][Medline]

Zhang HG, Lee HJ, Rocheleau T, ffrench-Constant RH, Jackson MB. (1995) Subunit composition determines picrotoxin and bicuculline sensitivity of Drosophila gamma-aminobutyric acid receptors. Mol Pharmacol 48:835–840.[Abstract]

Accepted for publication November 27, 2006.

CiteULike    Connotea