![]() | ![]() |
Formats:
|
||||||||||||||||
Copyright © 2006, Cold Spring Harbor Laboratory Press The nicotinic acetylcholine receptor gene family of the honey bee, Apis mellifera
1MRC Functional Genetics Unit, Department of Physiology Anatomy and Genetics, University of Oxford, Oxford OX1 3QX, United Kingdom; 2Centre de Recherches sur la Cognition Animale, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 5169, Université Paul Sabatier, 31062 Toulouse, France 3
Corresponding author.
E-mail david.sattelle/at/anat.ox.ac.uk; fax 44-1865-282-651. Received August 10, 2005; Accepted November 14, 2005. Freely available online through the Genome Research Open Access option. This article has been cited by other articles in PMC.Abstract Nicotinic acetylcholine receptors (nAChRs) mediate fast cholinergic synaptic transmission and play roles in many cognitive processes. They are under intense research as potential targets of drugs used to treat neurodegenerative diseases and neurological disorders such as Alzheimer's disease and schizophrenia. Invertebrate nAChRs are targets of anthelmintics as well as a major group of insecticides, the neonicotinoids. The honey bee, Apis mellifera, is one of the most beneficial insects worldwide, playing an important role in crop pollination, and is also a valuable model system for studies on social interaction, sensory processing, learning, and memory. We have used the A. mellifera genome information to characterize the complete honey bee nAChR gene family. Comparison with the fruit fly Drosophila melanogaster and the malaria mosquito Anopheles gambiae shows that the honey bee possesses the largest family of insect nAChR subunits to date (11 members). As with Drosophila and Anopheles, alternative splicing of conserved exons increases receptor diversity. Also, we show that in one honey bee nAChR subunit, six adenosine residues are targeted for RNA A-to-I editing, two of which are evolutionarily conserved in Drosophila melanogaster and Heliothis virescens orthologs, and that the extent of editing increases as the honey bee lifecycle progresses, serving to maximize receptor diversity at the adult stage. These findings on Apis mellifera enhance our understanding of nAChR functional genomics and provide a useful basis for the development of improved insecticides that spare a major beneficial insect species. The honey bee, Apis mellifera, is an important beneficial insect in agriculture. In addition to producing honey and beeswax, the contribution of A. mellifera to crop pollination is valued at more than $14 billion dollars per year in the U.S. alone (United States Department of Agriculture http://www.ars.usda.gov/main/main.htm). Honey bees live in societies of considerable complexity and thus are studied as models for social behavior (Robinson et al. 1997). The neonicotinoids are the newest major group of insecticides, which includes acetamiprid, clothianidin, dinotefuran, imidacloprid, nitenpyram, thiacloprid, and thiamethoxam (Tomizawa and Casida 2005). The worldwide annual sales of neonicotinoids amounts to ~1 billion dollars, and they are used against piercing-sucking pests (aphids, leafhoppers, and white-flies) of major crops. In France, the use of imidacloprid has been suspended over concerns that it may be having a drastic effect on bee populations (http://www.pan-uk.org/press/pr140604.htm), highlighting the importance that effective insecticides should also show selectivity within insects so that pollinators such as A. mellifera are spared. While the link between imidacloprid use and bee population decline has yet to be proven, studies have shown that imidacloprid is highly toxic to A. mellifera (Suchail et al. 2004) and at sublethal doses can alter honey bee foraging and learning (Guez et al. 2001; Lambin et al. 2001; Decourtye et al. 2004). Neonicotinoids act as agonists on their molecular targets, nicotinic acetylcholine receptors (nAChRs) (Matsuda et al. 2001), which are prototypical members of the cys-loop ligand-gated ion channel (LGIC) superfamily (Karlin 2002). The fast actions of acetylcholine (ACh) at synapses are mediated by nAChRs, which consist of five homologous subunits arranged around a central ion channel (Corringer et al. 2000; Unwin 2005). Analyses of completed genomes have revealed diverse nAChR gene families with mammals possessing 16 subunit genes, chicken, 17 (Millar 2003), Fugu rubripes, 28 (Jones et al. 2003), and Caenorhabditis elegans, at least 27 (Jones and Sattelle 2004). In contrast, Drosophila melanogaster and Anopheles gambiae have notably smaller nAChR gene families, each consisting of 10 subunits (Jones et al. 2005; Sattelle et al. 2005). To date, four A. mellifera nAChR subunits (Apisα2, Apisα3, Apisα7-1, and Apisα7-2) have been identified (Thany et al. 2003, 2005), which are expressed in brain structures that play roles in learning and memory, olfactory signal processing, mechanosensory antennal input, and visual processing. These findings are consistent with ACh being a major excitatory neurotransmitter in the insect nervous system (Breer and Sattelle 1987; Lee and O'Dowd 1999). Patch clamp studies have demonstrated the existence of a distinct nAChR subtype in the honey bee nervous system that is blocked by the nAChR antagonists α-bungarotoxin (α-Btx), dihydroxy-β-erythroidine and methyllycaconitine, while nicotine and imidacloprid acted as partial agonists on this receptor (Goldberg et al. 1999; Déglise et al. 2002; Wustenberg and Grunewald 2004). Another study has shown the presence of two nAChR populations that differ in their responses to imidacloprid but not ACh (Nauen et al. 2001). The involvement of nAChRs in honey bee behavior has also been investigated. Injection of the nAChR agonist, nicotine, showed that potentiation of the cholinergic system improves short term memory (Thany and Gauthier 2005) and injection of the nAChR antagonist, mecamylamine, inhibited olfactory learning or memory recall depending upon the site of injection (Lozano et al. 1996, 2001). Recently it has been demonstrated that one distinct nAChR sub-type, which is α-Btx sensitive, is involved in long-term memory, whereas a second subtype, which is α-Btx insensitive, but is affected by mecamylamine, plays a role in retrieval processes (Dacher et al. 2005). Interestingly, this mirrors to a certain extent the mammalian central nervous system, where there are two predominant nAChR subtypes, the α7 and α4β2 receptors, that are α-Btx sensitive and insensitive, respectively, and both receptor subtypes play a role in memory (for review, see Hogg et al. 2003). Since individual nAChR subunits can confer distinct pharmacological properties on a receptor (Romanelli and Gualtieri 2003), the multiple nAChR subtypes present in the honey bee nervous system are likely to be determined by their subunit composition. Identifying the full complement of honey bee nAChR subunits represents a critical step in understanding the variety of roles played by nAChRs in the honey bee nervous system and the exquisite repertoire of bee behavior, as well as in identifying particular targets of chemical compounds. Here we have used the A. mellifera genome to describe the complete honey bee nAChR gene family. Results Existence of II candidate nAChR subunit genes in the A. mellifera genome Using TBLASTN, we identified 11 candidate nAChR subunits in the A. mellifera genome. The complete open-reading frames of each subunit were confirmed and completed RT-PCR and RACE–PCR. An alignment of their protein sequences shows that the honey bee nAChR candidate subunits possess features common to members of the cys-loop LGIC superfamily (Fig. 1
The honey bee nAChR subunits show substantial sequence similarity with known nAChR subunits, in particular, those of other insects. As shown in Table 1, Apis and Drosophila nAChR subunits can share up to 84% amino-acid identity. With regard to vertebrate nAChR subunits, they show 25%–38% identity. A phylogenetic tree demonstrating the relationship between Apis nAChR subunits and those of Drosophila and Anopheles indicates orthologous relationships between the honey bee and fruit fly/ mosquito subunits (Fig. 2
Features particular to certain Drosophila and Anopheles nAChR subunits are also evident in their Apis counterparts (Fig. 1 Analysis of Drosophila and Anopheles nAChRs shows that each insect possesses a distantly related subunit sharing relatively low-sequence identity with other nAChR subunits. In the case of Drosophila, the subunit is of the non-α type (Dβ3) (Lansdell and Millar 2002), whereas in Anopheles it is an α subunit (Agamα9) (Jones et al. 2005). Interestingly, the honey bee has two distantly related subunits, one α (Amelα9) and the other non-α (Amelβ2), which are designated members of the “Dβ3 Group” (Fig. 2 A comparison of Apis and Drosophila nAChR gene structures shows that only one ortholog pair (Dα6 and Amelα6) shares an identical set of exon–intron boundaries (Fig. 3
Splice variants increase Apis nicotinic receptor diversity Two Apis nAChR subunits, Amelα4 and Amelα6, have alternatively spliced exons most likely arising from tandem exon duplication (Kondrashov and Koonin 2001). As with Dα4 and Agamα4 (Lansdell and Millar 2000; Jones et al. 2005), Amelα4 possesses two alternatives for exon 4 (denoted exon4 and exon4′) (Fig. 4A
As previously observed for Drosophila nAChRs, alternative splicing introduces amino-acid changes in functionally significant regions (Lansdell and Millar 2000; Grauso et al. 2002). For the two versions of Amelα6 exon 8, residues in the region linking TM2 with TM3 are altered (Fig. 4A Both Dα3 and Agamα3 possess extraordinarily long intra-cellular domains between TM3 and TM4 (Schulz et al. 1998; Jones et al. 2005). However, the Apis ortholog, Amelα3, does not have such an extended region (Fig. 1 Truncated transcripts for several Drosophila nAChR subunits have also been described. For instance, Dα4 cDNAs lacking exon 2 (Dα4Δexon2) have been identified (Lansdell and Millar 2000), while in other cases, omission of exon 4 from Dα4 (Dα4Δexon4) and exon 5 from Dα5 (Dα5Δexon5) result in frameshifts and the introduction of premature stop codons (Lansdell and Millar 2000; Grauso et al. 2002). For Dα7, a premature stop codon is introduced by lack of splicing intron 5 (Grauso et al. 2002). RT– PCR was performed to determine whether similar truncated transcripts could be detected for the corresponding Apis nAChR subunits. As with Anopheles (Jones et al. 2005), truncated honey bee cDNAs analogous to Dα4Δexon2 and Dα5Δexon5 were not detected, whereas Amel4Δexon4 and truncated Amelα7 transcripts were identified, both having premature stop codons (Fig. 4B The two Amelα4 splice variants are differentially expressed RT–PCR was performed to determine which of the 11 Apis nAChR subunits as well as all splice variants are transcribed at different stages of honey bee development, including four larval stages (L0–L3), three pupal stages (P1, P3, and P4), and the following tissues from adults: mushroom bodies, optic lobes, brains, head, and whole bodies. All 11 subunits, as well as all splice variants, are transcribed in each developmental stage and tissue tested (see Supplemental material) with one exception. Amelα4 exon4 transcripts were detected in all stages and tissues, whereas transcripts of Amelα4 exon4′ splice variants were not observed in larvae and were particularly more abundant in the mushroom bodies, optic lobes, and brain (see Supplemental material). Since alternative splicing of Amelα4 exon4′ substitute residues in the vicinity of the cys-loop, which has been shown to be important for complete receptor assembly (Green and Wanamaker 1997) and radio-ligand-binding assays, indicate that Dα4 with exon 4′ assembles less efficiently than with exon4 (Lansdell and Millar 2000), Amelα4 exon 4′ subunits may serve to modulate receptor assembly during the later stages of honey bee development and in tissues rich in neural activity such as the mushroom bodies and optic lobes. Amelα6 undergoes A-to-I pre-mRNA editing Pre-mRNA A-to-I editing modifies select adenosine (A) residues to inosine (I) in transcripts, which is interpreted as guanosine (G), thereby generating mRNA with a nucleotide composition that differs from the corresponding genomic DNA (Seeburg 2002). RNA editing has been observed in several Drosophila nAChR subunits, including two sites in loop D of Dβ1, one site in TM2 of Dβ2, one site in TM3, three sites in TM4 of Dα5, and seven sites in loops E to F in Dα6 (Grauso et al. 2002; Hoopengardner et al. 2003; Sattelle et al. 2005). To determine whether orthologous Apis nAChR subunits are also RNA edited, the equivalent regions of Amelβ1, Amelα8, Amelα5, and Amelα6 were amplified with high-fidelity proofreading DNA polymerase. For Amelβ1, Amelα8, and Amelα5, the sequences of the resulting amplification products were identical to those of genomic DNA with no indication of A-to-G changes (data not shown), showing that these regions of the three subunits are not RNA edited. For Amelα6, however, six RNA-editing sites were observed, two of which are conserved in the Drosophila and Heliothis virescens orthologs, Dα6 and Hvα7-2, respectively (Grauso et al. 2002) (Fig. 5
Since loop E contributes to ligand binding and N-glycosylation has also been linked to ligand binding as well as channel desensitization and conductance (Corringer et al. 2000; Nishizaki 2003), editing at this site has considerable potential to alter receptor function. In the remaining cases, editing alters residues near or within the cys-loop, which, like alternative splicing of Amelα4 exon4, may affect receptor assembly. Analysis of Amelα6 editing at different stages of honey bee development shows that in larvae, five of the six sites undergo editing, the extent of which increases throughout development so that in adults, four sites are predominantly edited. In pupae, from P3 onward, editing was observed at the sixth site, which increases considerably the potential diversity of subunit isoforms, as a lysine residue can be converted to either arginine, glutamic acid, or glycine. Interestingly, the elevated editing in the later stages of development is consistent with findings that RNA editing is particularly important in the nervous system function of Drosophila adults (Palladino et al. 2000) and that the highest levels of RNA editing are seen in adult flies (Keegan et al. 2005). Discussion We have used the available A. mellifera genome information to complete the characterization of the honey bee nAChR gene family, thus describing the first complete set of Hymenoptera nAChR subunits and the third insect nAChR gene family following those of the two Diptera, A. gambiae (Jones et al. 2005) and D. melanogaster (Sattelle et al. 2005). The three insect species represent ~280 million years of evolution (Carpenter and Burnham 1985; De Gregorio and Lemaitre 2002) where the nAChR gene family has remained compact with A. mellifera having 11 genes encoding nAChR subunits, whereas both D. melanogaster and A. gambiae possess 10 genes (Jones et al. 2005; Sattelle et al. 2005). The nAChR subunit composition of Apis most closely resembles that of Anopheles in that both possess nine α and one β subunit, while Drosophila has seven α and three β. The extra honey bee subunit is a β subunit (Amelβ2) making A. mellifera only the second insect known to possess more than one non-α type subunit. The characterization of the full complement of honey bee nAChR subunits presents an important basis for associating particular nAChR subtypes with key aspects of behavior, identifying receptor subtypes targeted by neonicotinoids as well as developing insecticides with improved selectivity. Indeed, comparison of complete insect nAChR gene families has identified a highly divergent subunit group (the Dβ3 group) as well as species-specific proteome diversification arising from alternative splicing and RNA editing, all of which represent promising subunit differences to target for future rational insecticide design. While studies using heterologous expression systems such as Xenopus laevis oocytes have proven instructive in characterizing vertebrate nAChRs (Corringer et al. 2000) and low levels of functional expression of an insect α subunit, αL1, have been observed in Xenopus oocytes (Marshall et al. 1990), expression of functional insect nAChRs has so far proven elusive (Sattelle et al. 2005). Nevertheless, Drosophila nAChR α subunits can form robust functional channels when coexpressed with a vertebrate β2 subunit (Bertrand et al. 1994) and studies on such hybrid receptors have provided insights into the selectivity of neonicotinoids for insect nAChRs over those of vertebrates (Matsuda et al. 1998; Ihara et al. 2003), regions of subunit proteins involved in neonicotinoid interactions (Shimomura et al. 2002, 2003, 2004), and the actions of different neonicotinoids (Ihara et al. 2004). Also, computer models of insect nAChRs have been recently generated, which permit docking experiments to assess interactions with compounds of interest (Sattelle et al. 2005). Similar studies combining functional expression with molecular modeling of Apis nAChRs are likely to prove useful in screening for novel compounds that show low selectivity for honey bee receptors and in dissecting the mechanisms of insecticide actions and selectivity on nAChRs. Methods Identification of nAChR subunits in the A. mellifera genome To identify putative nAChR subunits, we screened the A. mellifera genome (database version 34.2b available at http://www.ensembl.org/Apis_mellifera/index.html and assembly version 3.0 available at http://www.ensembl.org/Apis_mellifera/) with each of the 10 D. melanogaster nAChR subunit cDNA sequences using the TBLASTN algorithm (Altschul et al. 1990). Candidate honey bee nAChR subunits were identified based on their considerable sequence homology with previously characterized nAChR subunits (sequences with lowest similarity had E-value 1e-21), particularly at the N-terminal ligand-binding domain and the four transmembrane regions. RT–PCRs were performed to verify the open-reading frame sequences of each subunit. Since BLAST was unable to identify the highly variable N-terminal signal peptides, 5′-RACE, using the Roche 5′/3′ RACE kit, was performed to complete the nAChR subunit sequences. The multiple protein-sequence alignment was constructed with CLUSTALX (Thompson et al. 1997) using the slow-accurate mode with a gap-opening penalty of 10 and a gap-extension penalty of 0.1 as well as applying the Gonnet 250 protein weight matrix (Benner et al. 1994). The protein alignment was viewed using GeneDoc (http://www.psc.edu/biomed/genedoc). The neighbor-joining method (Saitou and Nei 1987) and bootstrap resampling (Felsenstein 1985), available with the CLUSTALX program, were used to construct a phylogenetic tree, which was then displayed using the TreeView application (Page 1996). Signal peptide cleavage sites were predicted using the SignalP 3.0 server (Dyrlov Bendtsen et al. 2004) and membrane-spanning regions were predicted using the TMpred program (available at http://www.ch.embnet.org/software/TMPRED_form.html). The PROSITE database (Falquet et al. 2002) was used to identify potential cyclic AMP (cAMP), protein kinase C (PKC), CK2, and potential kinase phosphorylation sites. Dissection of A. mellifera tissues Honey bee pupae and larvae were taken from the hive. Their developmental stage was determined using pigmentations of eyes, joints, and legs as described by Winston (1987). Adult honey bees were collected at the entrance of the hive. Bees were anaesthetized on ice and dissection of A. mellifera tissues were performed under a stereomicroscope in sterile 1X PBS. The brain was removed from the capsule head free of cuticle and trachea. When necessary, brain parts were separated manually. The tissues were then frozen in liquid nitrogen before RNA and genomic DNA extraction. Reverse transcription and polymerase chain reaction Genomic DNA was extracted from adult bees using the DNeasy Tissue Kit (Qiagen) and total RNA was extracted from various developmental stages and tissues using the RNeasy Mini Kit (Qiagen). First-strand cDNA was synthesized from 1 μg of total RNA using Superscript III First-Strand Synthesis Super Mix (Invitrogen). Nested RT–PCR reactions were performed to detect transcript of all honey bee nAChR subunits and variants. Primer pairs that recognize different exons were used to allow identification of cDNA-specific products (see Supplemental material for PCR primer sequences). The PCR reactions were performed in a total volume of 50 μL composed of Taq polymerase and 1X PCR buffer (Sigma), 0.2 mM dNTP mix (Roche), 0.4 μM each primer, and 2 μL first-strand cDNA template. The nested PCR approach involved two reactions each with 30 cycles of 95°C for 30 sec, 55°C for 30 sec, and 72°C for 30 sec/500 bp amplified. The first PCR was used at a final dilution of one in 5000 as template for the second nested PCR reaction. For RNA-editing analysis, nested PCR using the proofreading Pfu Turbo DNA polymerase (Stratagene) in 2 × 30-cycle reactions was performed on at least two independently made first-strand cDNAs. PCR products were analyzed by electrophoresis in a TAE gel and then purified using the QIAquick Gel Extraction Kit (Qiagen) before being sequenced by the dye termination method at the Biochemistry Sequencing Facility, University of Oxford. Acknowledgments We are indebted to the A. mellifera Genome Project (Human Genome Sequencing Center), which provided the starting point for this study. We thank Sandrine Paute for technical support. We also thank Ryszard Maleszka and Chris Ponting for encouragement, support, and helpful comments on the manuscript. Footnotes [Supplemental material is available online at www.genome.org. Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under accession nos. DQ026031–DQ026039.] Article published online before print. Article and publication date are at http://www.genome.org/cgi/doi/10.1101/gr.4549206. References
|
PubMed related articles
Your browsing activity is empty. Activity recording is turned off. |
|||||||||||||||
Bioessays. 1997 Dec; 19(12):1099-108.
[Bioessays. 1997]Annu Rev Pharmacol Toxicol. 2005; 45():247-68.
[Annu Rev Pharmacol Toxicol. 2005]Pest Manag Sci. 2004 Mar; 60(3):291-6.
[Pest Manag Sci. 2004]Neurobiol Learn Mem. 2001 Sep; 76(2):183-91.
[Neurobiol Learn Mem. 2001]Arch Insect Biochem Physiol. 2001 Nov; 48(3):129-34.
[Arch Insect Biochem Physiol. 2001]Ecotoxicol Environ Saf. 2004 Mar; 57(3):410-9.
[Ecotoxicol Environ Saf. 2004]Insect Mol Biol. 2003 Jun; 12(3):255-62.
[Insect Mol Biol. 2003]J Neurosci. 1999 Jul 1; 19(13):5311-21.
[J Neurosci. 1999]J Physiol. 1999 Feb 1; 514 ( Pt 3)():759-68.
[J Physiol. 1999]Neurosci Lett. 2002 Mar 15; 321(1-2):13-6.
[Neurosci Lett. 2002]J Comp Physiol A Neuroethol Sens Neural Behav Physiol. 2004 Oct; 190(10):807-21.
[J Comp Physiol A Neuroethol Sens Neural Behav Physiol. 2004]J Biol Chem. 1986 Jun 25; 261(18):8085-8.
[J Biol Chem. 1986]Insect Mol Biol. 2003 Jun; 12(3):255-62.
[Insect Mol Biol. 2003]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Genetics. 2002 Apr; 160(4):1519-33.
[Genetics. 2002]Gene. 2005 Jan 3; 344():125-32.
[Gene. 2005]Neurosci Lett. 2004 Jun 17; 363(3):195-8.
[Neurosci Lett. 2004]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]J Neurochem. 2002 Jan; 80(1):149-57.
[J Neurochem. 2002]J Neurochem. 2002 Mar; 80(6):1009-18.
[J Neurochem. 2002]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Genomics. 2003 Oct; 82(4):441-51.
[Genomics. 2003]Hum Mol Genet. 2001 Nov 1; 10(23):2661-9.
[Hum Mol Genet. 2001]Neuropharmacology. 2000 Oct; 39(13):2604-14.
[Neuropharmacology. 2000]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Genetics. 2002 Apr; 160(4):1519-33.
[Genetics. 2002]Neuropharmacology. 2000 Oct; 39(13):2604-14.
[Neuropharmacology. 2000]Genetics. 2002 Apr; 160(4):1519-33.
[Genetics. 2002]Proc Natl Acad Sci U S A. 1996 Jun 11; 93(12):6118-23.
[Proc Natl Acad Sci U S A. 1996]J Neurochem. 1998 Aug; 71(2):853-62.
[J Neurochem. 1998]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Nature. 1988 Dec 15; 336(6200):677-80.
[Nature. 1988]J Cell Biol. 2001 Apr 2; 153(1):1-12.
[J Cell Biol. 2001]Neuropharmacology. 2000 Oct; 39(13):2604-14.
[Neuropharmacology. 2000]Genetics. 2002 Apr; 160(4):1519-33.
[Genetics. 2002]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Genome Biol. 2004; 5(2):R8.
[Genome Biol. 2004]Brain Res Mol Brain Res. 2003 Sep 10; 117(1):15-26.
[Brain Res Mol Brain Res. 2003]J Biol Chem. 1997 Aug 15; 272(33):20945-53.
[J Biol Chem. 1997]Neuropharmacology. 2000 Oct; 39(13):2604-14.
[Neuropharmacology. 2000]Neuron. 2002 Jul 3; 35(1):17-20.
[Neuron. 2002]Genetics. 2002 Apr; 160(4):1519-33.
[Genetics. 2002]Science. 2003 Aug 8; 301(5634):832-6.
[Science. 2003]Bioessays. 2005 Apr; 27(4):366-76.
[Bioessays. 2005]Genetics. 2002 Apr; 160(4):1519-33.
[Genetics. 2002]Annu Rev Pharmacol Toxicol. 2000; 40():431-58.
[Annu Rev Pharmacol Toxicol. 2000]Brain Res Mol Brain Res. 2003 Jun 10; 114(2):172-6.
[Brain Res Mol Brain Res. 2003]Cell. 2000 Aug 18; 102(4):437-49.
[Cell. 2000]EMBO J. 2005 Jun 15; 24(12):2183-93.
[EMBO J. 2005]Genomics. 2005 Feb; 85(2):176-87.
[Genomics. 2005]Bioessays. 2005 Apr; 27(4):366-76.
[Bioessays. 2005]Nature. 2002 Oct 3; 419(6906):496-7.
[Nature. 2002]Annu Rev Pharmacol Toxicol. 2000; 40():431-58.
[Annu Rev Pharmacol Toxicol. 2000]EMBO J. 1990 Dec; 9(13):4391-8.
[EMBO J. 1990]Bioessays. 2005 Apr; 27(4):366-76.
[Bioessays. 2005]Eur J Neurosci. 1994 May 1; 6(5):869-75.
[Eur J Neurosci. 1994]Br J Pharmacol. 1998 Feb; 123(3):518-24.
[Br J Pharmacol. 1998]J Mol Biol. 1990 Oct 5; 215(3):403-10.
[J Mol Biol. 1990]Nucleic Acids Res. 1997 Dec 15; 25(24):4876-82.
[Nucleic Acids Res. 1997]Protein Eng. 1994 Nov; 7(11):1323-32.
[Protein Eng. 1994]Mol Biol Evol. 1987 Jul; 4(4):406-25.
[Mol Biol Evol. 1987]Comput Appl Biosci. 1996 Aug; 12(4):357-8.
[Comput Appl Biosci. 1996]J Mol Biol. 2004 Jul 16; 340(4):783-95.
[J Mol Biol. 2004]