(Received for publication, August 29, 1996, and in revised form, October 9, 1996)
From the Institut Alfred Fessard, UPR2212, CNRS, 91198 Gif-sur-Yvette cedex, France, the Departments of Psychiatry and
§ Pharmacology, University of Toronto, and the
¶ Laboratory of Molecular Neurobiology, Clarke Institute of
Psychiatry, Toronto, Ontario M5T 1R8, Canada
The existence of dopamine D1C and D1D receptors in Xenopus and chicken, respectively, challenged the established duality (D1A and D1B) of the dopamine D1 receptor class in vertebrates. To ascertain the molecular diversity of this gene family in early diverging vertebrates, we isolated four receptor-encoding sequences from the European eel Anguilla anguilla. Molecular phylogeny assigned two receptor sequences (D1A1 and D1A2) to the D1A subtype, and a third receptor to the D1B subtype. Additional sequence was orthologous to the Xenopus D1C receptor and to several other previously unclassified fish D1-like receptors. When expressed in COS-7 cells, eel D1A and D1B receptors display affinity profiles for dopaminergic ligands similar to those of other known vertebrate homologues. The D1C receptor exhibits pharmacological characteristics virtually identical to its Xenopus homologue. Functionally, while all eel D1 receptors stimulate adenylate cyclase, the eel D1B receptor exhibits greater constitutive activity than either D1A or D1C receptors. Semiquantitative reverse transcription-polymerase chain reaction reveals the differential distribution of D1A1, D1A2, D1B, and D1C receptor mRNA within the hypothalamic-pituitary axis of the eel brain. Taken together, these data suggest that the D1A, D1B, and D1C receptors arose prior to the evolutionary divergence of fish and tetrapods and exhibit molecular, pharmacological, and functional attributes that unambiguously allow for their classification as distinct D1 receptor subtypes in the vertebrate phylum.
Dopamine is a widespread modulatory neurotransmitter in the central and peripheral nervous system of vertebrates. The physiological roles of dopamine range from the sensorimotor control, thermoregulation, and modulation of appetite to the regulation of reproductive and maternal behavior. At the cellular level, the pleiotropic effects of dopamine are mediated by two specific classes of dopamine receptors, termed D1 and D2, distinguishable by their ability to activate (D1 class) or to inhibit (D2 class) the enzyme adenylyl cyclase (1). More recently, molecular studies have revealed that D1 and D2 receptors are composed of several membrane proteins, each belonging to the G protein-coupled superfamily of receptors defined by their shared overall topology and common signal transduction mechanism, which triggers GDP/GTP exchange on heterotrimeric G proteins (2). Two D1-like receptors (D1 and D5, more precisely named D1A and D1B, respectively) and three D2-like receptor subtypes (D2, D3, and D4), each encoded by distinct genes, have been isolated in mammals (3). Based on sequence analysis and gene organization, D1 and D2 dopamine receptor classes appear not to be more closely related to each other than to other catecholamine receptor families (4, 5). As with other members of the monoamine receptor family, D1-like and D2-like receptors are probably of independent origin and have acquired separately and convergently the ability to bind their endogenous ligand (5).
In contrast with dopamine D2-like receptor genes, which have only been extensively characterized in mammals, D1-like receptor gene diversity has been recently examined in a small set of other vertebrate species. Besides D1A and D1B receptors, additional D1-like receptor genes have been isolated from amphibians (Xenopus D1C receptor (6)) and birds (chicken D1D receptor (7)). Similarly, D1-like receptor sequences not classified as either D1A or D1B have been described in fish (Ref. 8 and GenBankTM sequence X81969[GenBank]). Although Xenopus D1C and chicken D1D receptors significantly differ from both vertebrate and mammalian D1A and D1B receptors on the basis of their amino acid sequence and distinct pharmacological profiles, it is unclear whether they are truly reflective of distinct D1 receptor subtypes or if their presence is merely associated with and restricted to these particular species.
As with other members of the catecholamine receptor gene family, the molecular diversity of D1-like receptors appears to be the product of gene duplication events occurring during the evolutionary history of a particular species (9). Since the multiplicity of D1-like receptors is a common characteristic of amphibian, avian, and mammalian genomes, the gene duplication events that underlie the origin of these receptor subtypes must have occurred significantly before or close to the emergence of tetrapods. In order to gain insights into the nature and temporal occurrence of these important genetic events, it is necessary to ascertain the genetic diversity of the D1 receptor family in a species belonging to a phylum that diverged before the emergence of tetrapods. Ray-finned fish (actinopterygians) diverged ~420 million years ago from flesh-finned fish (sarcopterygians) from which the tetrapod ancestor descended. Therefore, we searched for the full complement of D1-like receptor diversity in a modern representative of actinopterygians, the European eel Anguilla anguilla (a teleost). We report here that (a) eel dopamine D1 receptors are comprised of three distinct D1 receptor subtypes, which express molecular, pharmacological, and functional signatures that define unequivocally the characteristics of vertebrate D1A, D1B, and D1C receptors; (b) the diversity of the vertebrate D1 receptor gene family occurred and was clearly established before the divergence of actinopterygians from sarcopterygians.
Freshwater silver eels weighing 250-300 g were caught in Northern France and kept in tanks in running freshwater. Animals were killed by decapitation, the brain and cervical spinal cord were dissected out, and blood samples were taken. RNAs were extracted from total or dissected brain regions by the guanidinium isothiocyanate/acid phenol method (10), and poly(A)+ RNA was obtained by batch chromatography on oligo(dT)-cellulose (Boehringer Mannheim). Genomic DNA was extracted from the nucleated erythrocytes (11).
PCR1 CloningFor the D1A receptor gene, two degenerate oligonucleotides (U2, ATCYTSAACCTGTGTGCCATCAGC; and L2, TGAYGGGRTTSACGGCRCTGTTSAC) were chosen on the basis of the conserved transmembrane domains III and VII of the rat and human D1A and D1B receptors. 35 cycles of PCR (94 °C, 45 s; 60 °C, 45 s; 72 °C, 1 min) were performed with 50 ng of eel genomic DNA using Taq polymerase (Promega). PCR products were directly subcloned into the pCRII vector (Invitrogen).
For the D1C receptor gene, we used two degenerate oligonucleotides (TGGGTDGCHTTYGAYATHATGTG and CARMANACRAAVACNCCCAT), which encode sequences based on the conserved transmembrane domains III and VI of all the vertebrate D1-like receptors cloned by January 1995 (rat, human, opossum, goldfish, Xenopus, and chicken D1A receptors; rat, human, Xenopus, and chicken D1B receptors; and Tilapia D1C, Xenopus D1C, and chicken D1D receptors). 50 ng of eel genomic DNA served as template in 40 cycles of PCR (94 °C, 45 s; 50 °C, 45 s; 72 °C, 1 min). After subcloning, recombinant clones were screened at medium stringency (60 °C in 5 × SSPE, 0.5% SDS, 5 × Denhardt's solution, 100 µg/ml yeast tRNA) with a Tilapia D1C receptor radiolabeled probe (a gift of Dr. A. Lamers, University of Nijmegen, The Netherlands) and discriminated on the basis of hybridization signals.
Full-length D1A1 and D1C coding sequences were
obtained by a modification of the 5- and 3
-RACE (12): Briefly, 500 ng
of genomic DNA were digested in four separate reactions with
AluI, RsaI, PvuII, and SspI
(New England Biolabs), phenol/chloroform-extracted, and
ethanol-precipitated. The digested DNAs were ligated to 500 pmol of a
partially double-stranded adapter. The four reaction mixes were heated
at 70 °C for 5 min to inactivate the ligase and treated with
terminal deoxynucleotidyltransferase (Promega) in the presence of 0.1 mM dideoxy-ATP to avoid nonspecific 3
extension during
subsequent PCR experiments. DNAs were finally desalted on SH-400
Sephacryl columns (Pharmacia Biotech Inc.). 5 ng of adaptor-ligated
DNAs were amplified with adapter-specific primer and gene-specific
internal primers (corresponding to positions 704-719 for the 5
-RACE
and to position 687-710 for the 3
-RACE in the sequence of the
D1A1 receptor; for the D1C receptor, positions 618-656 for the 5
-RACE and 826-856 for the 3
-RACE). 0.5 µl of PCR
mix were reamplified with nested gene-specific internal primers (corresponding to positions 620-646 for the 5
-RACE and to positions 752-777 for the 3
-RACE in the sequence of the D1A1
receptor; for the D1C receptor, to positions 468-495 for
the 5
-RACE and to positions 976-1007 for the 3
-RACE) for 30 cycles.
The longest PCR products were subcloned into the pCRII vector and
sequenced.
The full-length D1A1 and D1C receptor coding sequences were constructed by PCR performed on eel genomic DNA using Pfu DNA polymerase (Stratagene) with gene-specific primers allowing us to amplify the DNA sequences from upstream of the putative Kozak sequence and downstream of the putative stop codon. The PCR products were cloned into the eukaryotic expression vector pcDNA3 (Invitrogen) and sequenced.
cDNA Library Construction and ScreeningAn unidirectional cDNA library was constructed in part with the cDNA synthesis kit and Uni-ZAP-XR cloning kit (Stratagene). Briefly, 10 µg of poly(A)+ from eel brain and pituitary were reverse-transcribed with avian myeloblastosis virus reverse transcriptase using a hybrid oligo(dT) linker-primer, which contains an XhoI site, in the presence of 5-methyl-dCTP. Second strand synthesis was performed with RNase H and DNA polymerase I, and the cDNAs were blunt-ended with Klenow DNA polymerase, ligated to an EcoRI adapter, and finally digested with XhoI and EcoRI. Following size fractionation by gel filtration on Sephacryl S400, cDNAs over 1 kilobase pair were ligated into the Uni-ZAP XR vector arms and in vitro packaged with Gigapack II Gold extract. The library contained ~6 × 106 independent clones.
Approximately 1 × 106 recombinant phages were plated
and transferred in duplicate onto Hybond N+ filters
(Amersham Corp.). The DNA fragment encoding the eel dopamine D1A receptor gene was labeled with
[-32P]dCTP by random priming to a specific activity
higher than 1.109 cpm/µg. Filters were hybridized at
60 °C in the hybridization medium as described above, containing
1.106 cpm/ml of 32P-labeled D1A
fragment. The filters were washed twice for 30 min in 0.5 × SSPE,
0.1% SDS at 55 °C and autoradiographed. Positive clones were
isolated by three rounds of purification and excised in vivo
according to the manufacturer's protocol.
All DNA clones were sequenced on both strands with internal sequence-specific or universal M13 primers using Sequenase 2.0 (Amersham), either by hand or on a ABI373 sequencer (Genome Express, Grenoble, France). The deduced amino acid sequences of the cloned eel D1-like receptors were aligned with all D1-like sequences available by February 1996: human D1A (X58987), rat D1A (M35077), opossum D1A (S67258), goldfish D1A (L08602), Xenopus D1A (U07863), chicken D1A (L36877), Fugu D1-like sequence (X80174), human D1B (X58454), rat D1B (M69118), Xenopus D1B (U07864), chicken D1B (L36878), Xenopus D1C (U07865), chicken D1D (L36879), Tilapia D1C (X81969), Fugu D5-like sequence (X80177), and Drosophila D1 (X77234). The alignments of sequences, deletions of the invariant or noninformative positions, distance calculation, tree constructions, and bootstrap analysis were carried out on a PC with the MUST software package (13).
COS-7 Cell Expression and Ligand BindingDNAs were
subcloned into the eukaryotic expression plasmid pcDNA3
(Invitrogen). COS-7 cells grown in 150-mm plates in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at
37 °C, 5% CO2 were transfected with recombinant plasmid
(10 µg/106 cells) by either the DEAE-dextran procedure or
by electroporation as described previously (6, 14). 48-72 h after
transfection, cells were collected, and membranes were prepared in
buffer (50 mM Tris, pH 7.4, 5 mM EDTA, 1.5 mM CaCl2, 5 mM KCl, 5 mM MgCl2, 120 mM NaCl). For
saturation experiments, membranes (30-50 µg/ml) were assayed for
D1 receptor activity using increasing concentrations (0.02-2.0 nM) of [3H]SCH-23390 (DuPont NEN;
81.4-85 Ci/mmol; 1 Ci = 37 GBq) in a final volume of 1.5 ml and
incubated for 90 min at 37 °C or room temperature. For competition
binding experiments, [3H]SCH-23390 (150 pM
final concentration) was incubated with increasing concentrations of
dopaminergic compounds (1011 to 10
4
M) and assayed as above. Incubations were terminated by
rapid filtration over Skatron filter mats and monitored for tritium in
a Pharmacia or Beckman liquid scintillation counter. For all experiments, nonspecific binding was defined in the presence of 1 µM (+)-butaclamol. Ki values were
calculated from the estimated IC50 as described by Munson
and Rodbard (38) by using the nonlinear least-square curve fitting
program, Kaleidagraph (Abelbeck Software). All experiments were
conducted in triplicate with all clones tested concurrently.
In cAMP accumulation experiments, COS-7 cells transiently transfected with the receptor clones were grown for 48-72 h in 6- or 24-well dishes and assayed for cAMP accumulation in the presence or absence of various dopaminergic compounds as indicated in Dulbecco's modified Eagle's medium containing 0.5 mM 3-isobutylmethylxanthine and 1 µM propranolol as described (6). For experiments in which the constitutive activity of multiple D1 receptors were assessed, cells were assayed simultaneously for D1 receptor activity using ~1.5-2.0 nM [3H]SCH-23390 to ensure the equivalence of D1A, D1B, and D1C receptor expression. cAMP content was determined by immunodetection as described by the manufacturer (Amersham).
Reverse Transcription-PCR Analysis of the Multiple D1 Receptor mRNA Distribution10 µg of total RNA extracted
from different sections of eel brain were treated with 5 units of RQ1
DNase (Promega) for 30 min at 37 °C, extracted twice with
phenol/chloroform, and precipitated with ethanol. 5 µg of RNA were
then primed with oligo(dT) and reverse-transcribed with 200 units of
Superscript II (Life Technologies, Inc.) at 45 °C for 60 min.
1/50 of the reaction was used as substrate in subsequent PCR
experiments. We firstly checked that all four D1 receptor
transcripts were linearly and similarly enough amplified (never more
than a 4-fold difference) by the receptor-specific primers after 15-30
cycles of PCR amplifications. Then, for each receptor, we performed 25 cycles of PCR with receptor-specific primers (see legend to Fig. 6)
with cDNAs prepared from different parts of the brain. The
quantities of cDNA used as substrate were estimated by
co-amplification of the eel cytochrome b cDNA
(GenBankTM sequence D84302[GenBank]).
PCR products were separated by agarose gel electrophoresis and transferred onto Hybond N+ filters. Filters were hybridized with 32P-labeled receptor-specific probe (1 × 106 cpm/ml) in 5 × SSPE, 0.5% SDS, 5 × Denhardt's solution, 100 µg/ml yeast tRNA for 8 h at 68 °C and washed twice in 0.1 × SSPE, 0.1% SDS at 65 °C for 30 min. After autoradiography, filters were stripped by boiling 0.5% SDS and subsequently hybridized under the same conditions with 1 × 106 cpm/ml of 32P-labeled cytochrome b probe.
Amplification of eel
genomic DNA by PCR with D1 specific primers generated a
fragment that upon sequence analysis displayed 74% identity with
transmembrane segments III-VII of the rat D1A receptor
gene. The full-length sequence of this gene was then obtained using a
modification of the 5- and 3
-RACE technique (see "Experimental
Procedures"). Three clones were obtained from independent PCR
experiments with the same sequence. This D1 receptor fragment was secondarily used as a probe to screen at medium stringency a cDNA library prepared from eel brain and pituitary mRNAs. Six independent clones were isolated. Further hybridization analysis and
DNA digestion with restriction enzymes revealed that they corresponded
to two distinct cDNA sequences. Sequence determination of the
longest cDNAs of each of the two classes (clones Eel 441 and 446)
revealed that they most likely represented D1A and
D1B receptor orthologues. Surprisingly, the D1A
cDNA sequence isolated from the library differed slightly from that
of the PCR products previously obtained from genomic DNA. Therefore, we
named these two D1A sequences D1A1 and
D1A2 in accord with the standard nomenclature of recently
duplicated loci (15).
Taking into account these new sequences and those from
Xenopus D1C and chicken D1D (6, 7),
we designed other degenerate primers suitable for PCR amplification in
order to detect a larger set of D1-like receptors from
genomic DNA. Using this approach, an amplified DNA fragment sharing
73% sequence identity with the Xenopus D1C gene
was isolated. The full-length gene sequence was subsequently obtained
using a combination of 5- and 3
-RACE procedures as described under
"Experimental Procedures." Finally, PCR analysis revealed that none
of the four eel D1-like receptor genomic sequences contain
introns interrupting their coding regions, as is characteristic for all
cloned vertebrate/mammalian D1 receptor sequences to date (data not shown).
Fig. 1 depicts the deduced amino acid sequence of all
four cloned eel D1-like receptors. Sequence comparisons
with other cloned members of the D1 receptor family clearly
indicate that these four sequences encode fish homologues of mammalian
D1 receptors, which can be subdivided into D1A,
D1B, and D1C receptor subtypes. This contention
is further supported by phylogenetic analysis of multiple receptor
sequences (see below). A putative initiation methionine with the
predicted Kozak sequence (16) was followed by long open reading frames
for all four clones. Two eel D1A-like receptors were
encoded by 1335 nucleotides (D1A1 and D1A2),
the D1B of 1374 nucleotides (eel D1B) and eel
D1C was composed of 1344 nucleotides encoding proteins with
estimated molecular masses of 49,441, 49,310, 51,940, and 50,084 Da,
respectively. The highest observed number of identities shared by all
of the known members of the D1 receptor class are found in
putative transmembrane regions of the deduced amino acid sequence.
Within these regions the two D1A receptors
(D1A1 and D1A2), the D1B receptor,
and the D1C receptor exhibit homologies of ~95, 93, and
97%, respectively, with their corresponding Xenopus,
chicken, and mammalian receptors. As with all D1 receptors
cloned to date, regions of significant sequence divergence between
D1 receptor subtypes are particularly evident in the amino
termini, the third intracellular and extracellular loops, the fourth
extracellular domain, and the carboxyl-terminal tails.
Consensus sequences for putative post-translational modifications as well as amino acid residues known to be critical for dopamine binding have been remarkably conserved in the eel D1 receptor family and are found at the expected positions in the sequence. For instance, the aspartic acid residue in TM2 thought to mediate the sodium ion effect on ligand binding (17) as well as the aspartic acid in TM3 and the three sequential serine residues in TM5, which are believed to be the key determinant of dopamine binding in the rat D1A receptor (18), are conserved in the eel sequences. Similarly, six cysteine residues are found at homologous positions in all D1 receptor sequences. Two cysteines (positions 94 and 185 of the eel D1A1 receptor) are believed to form a disulfide bridge linking the first and second putative extracellular loops. Interestingly, two other cysteines (positions 292 and 304) are also conserved in the third putative extracellular loop and could possibly serve to form a second disulfide bridge constraining receptor structure. Two additional conserved cysteines are found in the cytoplasmic tail of these receptors (positions 344 and 348). The residues homologous to cysteine 337 or 347 of the rat and human D1A sequences, respectively, appear to be myristoylated so as to anchor the C terminus to the membrane and to allow agonist-selective conformational changes to be transmitted at the internal side of their plasma membrane (19, 20). Several sites that are putative substrates for protein kinase A and protein kinase C are conserved in the third intracellular loop of all the vertebrate D1 receptors. The serine residues homologous to the serine 380 of the rat D1A sequence shown in vitro to be a substrate for phosphorylation by protein kinase A (21) is conserved in all of the D1A sequences, being replaced, interestingly enough, by a conservative substitution (threonine) in the eel D1A1 receptor. Conserved consensus sequences for N-linked glycosylation are found within the amino terminus and the third extracellular loop for all four cloned eel D1 receptors. The eel D1C receptor, however, contains one additional consensus site in the amino terminus, as does the Xenopus D1C receptor. Unlike the Xenopus D1C receptor, which contains two additional protein kinase C consensus sites within the carboxyl tail, the eel D1C homologue does not, the functional significance of which, if any, is presently unknown. Finally, as highlighted in Fig. 1, the Xenopus and eel D1C receptor sequences exhibit several specific residues shared with the avian D1D sequence but not with either D1A or D1B receptor sequences. While clearly not definitive, these specific amino acid residues may constitute the "molecular signature" of D1C receptors and, as such, possibly suggest that the chicken D1D receptor is indeed more closely related to the D1C than to the other D1 receptor subtypes.
Molecular Phylogeny Analysis of the D1-like Receptor DiversityThe amino acid sequences deduced from the four eel
D1-like receptor DNA clones were aligned with all cloned
D1-like receptor sequences available so far. As depicted in
Fig. 2, we calculated a matrix of pairwise distances
(i.e. the number of amino acid substitutions) separating the
receptor sequences and constructed a phylogenetic tree with the
neighbor-joining method (22), from which evolutionary relationships
could be hypothesized. Maximum parsimony analysis of the data gave
essentially the same branching order (not shown).
Each of the D1 eel receptor sequences is clearly assigned to a well defined receptor subtype. The bootstrap resampling method applied to the distance tree estimated as very robust the branching delineating the D1A (100%), D1B (99%), and D1C (98%) sequences (Fig. 2). Although the D1A sequences appear to have diverged more slowly than D1B or D1C sequences in the same group of species, these divergences remain modest and do not hinder the significance of the analysis. The European eel possesses, therefore, two distinct D1A (D1A1 and D1A2), one D1B, and one D1C receptor genes. Other D1-like receptors have been recently isolated from other fish, such as a D1A receptor from goldfish retina (23), and several other D1-like receptor sequences have been isolated from Tilapia (24) and Fugu (8), allowing for the opportunity to assess their evolutionary relationships. Our phylogenetic analysis strongly suggests that eel D1C, Tilapia D1C, Fugu D5-like, and Xenopus D1C are orthologues (bootstrap value = 98) and that they constitute a new subtype of vertebrate D1 receptor. As such, the Fugu sequence should be renamed as D1C. The situation for the chicken D1D is less clear, since molecular distance analysis could not unambiguously determine whether this receptor belonged to the D1C subtype, but with a high rate of sequence divergence, or if it represented the first member of another subtype of D1-like receptors specifically duplicated in the avian lineage. Parsimony methods as well as the presence of some specific residues shared by D1C and D1D sequences (see Fig. 1) appear to support the existence of a D1C/D1D clade (data not shown). This issue could be resolved with the availability of D1D receptor sequences from nonavian species. In any event, the existence of the same paralogous D1-like receptor subtypes in teleost fish and tetrapods demonstrates that the D1C receptor subtype is common to most vertebrates and that the gene duplication events at the origin of the D1-like receptor diversity arose prior to the separation of actinopterygian fishes from the other vertebrates, ~420 million years ago.
Eel D1A1 and D1A2 Are Encoded by Two Distinct GenesThe existence of two D1A receptors
with an overall amino acid homology of 94% (within the transmembrane
segments) suggests that the genetic diversity of D1
receptors in eel is in fact greater than that observed in other
vertebrates. Alternatively, the existence of two highly homologous
receptor sequences suggested the possibility that D1A1 and
D1A2 receptors could represent two alleles of a single
gene, even if sequence differences between the two receptors were much
higher than usually expected for allelic variants. This possibility was
initially strengthened by the fact that the genomic DNA from which the
D1A1 receptor was amplified and the mRNAs used to
construct the library from which D1A2 had been cloned did
not originate from the same animals and were obtained from wild
populations, in which large genetic diversity is expected. To
distinguish between these two hypothesis, the presence of
D1A1 and D1A2 sequences was detected by PCR
experiments in 12 individual genomes from eels obtained from different
suppliers and very distant geographical areas. Fig. 3
depicts the results obtained following PCR amplification of eel genomic
DNA with D1A1- and D1A2-specific primers. As
can be seen in Fig. 3, the two receptor sequences, corresponding to D1A1 and D1A2, coexist in a single eel genome.
Identical results were obtained with 11 other eel samples (data not
shown). The extremely low probability of 12 animals coming from
different parts of France to be heterozygous for the same alleles
prompted us to conclude that D1A1 and D1A2
sequences indeed corresponded to two different genes (also see
below).
Pharmacological Characterization of the Eel D1-like Receptors
In order to further justify our proposed receptor classification scheme based on molecular phylogentic analysis, we characterized D1-like receptors found in the European eel in terms of their pharmacological profiles following transient expression in COS-7 cells. In particular, a comparative analysis of the pharmacological properties obtained for each of the proposed D1 receptor subtypes in the eel with those observed for well defined mammalian or other vertebrate D1-like receptor subtypes (7, 25) would strongly support the differentiation of multiple D1-like receptors into three distinct subclasses. Following expression in COS-7 cells, all four receptors bound the D1 receptor antagonist, [3H]SCH-23390, in a saturable manner to a single class of binding site with high affinity and with estimated dissociation constants (Kd) of 0.140, 0.90, 0.130, and 0.065 pM for eel D1A1, D1A2, D1B, and D1C receptors, respectively. Saturation analysis revealed receptor densities (Bmax) that were on average 0.6, 0.8, 1.0, and 0.8 pmol/mg protein for D1A1, D1A2, D1B, and D1C, respectively. The expression levels of the four eel receptors were essentially similar, facilitating the comparison of their respective pharmacological characteristics. [3H]SCH-23390 binding to membranes of COS-7 cells expressing D1A2, D1B, or D1C receptors was inhibited by various dopaminergic agonists and antagonists, in a stereoselective, concentration-dependent, and uniphasic manner (as indexed by Hill coefficients close to unity) with a pharmacological profile clearly indicative of a D1 receptor. Estimated Ki values for these agents are listed in Table I.
|
One unique distinguishing pharmacological feature between the mammalian
dopamine D1-like receptors is the inherent ability of the
D1B receptor to display higher affinity for the endogenous neurotransmitter dopamine than D1A (26). As listed in Table I, and consistent with our proposed classification of these receptors based on molecular phylogeny, the eel D1A2 receptor
displayed an affinity for dopamine (~3 µM) ~3-4-fold
less than the eel D1B receptor (~880 nM),
paralleling their vertebrate and mammalian, particularly rat,
counterparts. Moreover, similar to the vertebrate/mammalian D1B receptor, 6,7-ADTN exhibited higher affinity for the
Xen D1B receptor than for the D1A
receptor, while most antagonists exhibited lower affinities for the eel
D1B receptor similar to that seen with the human
D1B/D5 receptor. Estimated
Ki values for the inhibition of
[3H]SCH-23390 binding by a series of compounds to eel
D1A2 or D1B receptors are highly correlated to
Ki values obtained on the vertebrate or human
D1A or D1B receptors, respectively, with
essentially 1:1 correspondence in drug affinities (see correlation with
Xenopus receptors in Fig. 4, B and
C). As such, the major differences in drug affinities
discriminating between the D1A and D1B receptor
subtypes in either vertebrate or mammalian species appear to be
conserved and also found for eel sequences. Although not as extensively
characterized, the D1A1 receptor appears in this respect to
behave as a bona fide D1A receptor, displaying affinities
for dopaminergic agonists and antagonists similar to that observed for
D1A2.
As listed in Table I, the D1C receptor, however, displayed pharmacological characteristics consistent with both types of receptors with an observed affinity for dopamine (~1380 nM), somewhat intermediate to that of D1A2 and D1B receptors. Fig. 4A illustrates a similar pattern for NPA, displaying an affinity for the D1C receptor intermediate to that of D1A2 or D1B. Similar results were obtained for 6,7-ADTN. All other agonists displayed somewhat higher affinity for the eel D1C receptor than either the D1A or D1B receptors, similar to that seen with the Xenopus D1C receptor. Most D1 receptor antagonists exhibited affinities at the D1C receptor that were either intermediate to those of D1A2 and D1B receptors or identical to that of D1A2 (see Table I). As illustrated in Fig. 4D, the estimated Ki values of various agonists and antagonists at the eel D1C receptor are highly correlated with those obtained on the cloned Xenopus D1C with a virtual 1:1 correspondence in drug affinities as indexed by the line of equimolarity. Since none of the receptors displayed guanine nucleotide sensitivity (data not shown), consistent with previous observations (6, 7), it is possible to directly compare affinities of these receptors with their vertebrate/mammalian counterparts expressed in the same cells and assayed under similar conditions. While the eel D1C receptor subtype can be pharmacologically differentiated from either D1A or D1B, none of the compounds tested selectively identified or exhibited preferential affinity for the D1C receptor subtype. Future work will need to identify ligands, possibly not of the benzazepine class, that can pharmacologically differentiate the D1C receptor.
Despite the rather unique pharmacological profiles exhibited by eel
D1 receptor subtypes, all four receptors were found to couple to the same second messenger system when expressed in COS-7 cells. As illustrated in Fig. 5A, dopamine
(10 µM) stimulated the eel D1A2
receptor-mediated production of cAMP ~10-fold over basal levels, an
effect that is consistently blocked by pretreatment with the
D1 receptor antagonist SCH-23390 (1 µM). The
D1A1 receptor exhibits properties identical to those of the
D1A2 receptor, and corresponding data are therefore not
presented. Cells transfected by the nonrecombinant vector remained
insensitive to dopamine. D1B receptor activation stimulated
adenylate cyclase activity ~6-fold over basal levels, an effect
blocked by SCH-23390 and similar to that seen with the D1C
receptor. Given that receptor expression levels in any given experiment
were similar (0.7-0.9 pmol/mg protein), it appears that the eel
D1C receptor displays a somewhat increased efficacy in its
ability to maximally stimulate adenylate cyclase activity. The
concentration-dependent cAMP accumulation induced by
dopamine at either D1A2, D1B, and
D1C receptors, however, did not reveal any significant
differences in estimated affinity with EC50 values of 1.2, 1.4, and 0.9 µM, respectively (data not shown). Moreover,
as seen in Fig. 5A, while SKF-38393 partially stimulated
cAMP accumulation at D1A2, D1B, and
D1C receptors, consistent with its partial agonist status,
the eel D1C receptor appeared to be somewhat more
responsive to this compound, consistently yielding maximal cAMP
responses (55% relative to dopamine) greater than either
D1A2 or D1B receptors.
As depicted in Fig. 5B, the basal levels of cAMP accumulation measured in D1B-transfected cells were significantly higher (~2-3-fold) than in cells expressing D1A receptors. As illustrated in the figure, D1B constitutive activity could be antagonized by both butaclamol and flupentixol but not by SCH-23390. This is virtually identical to the profile exhibited by these ligands at mammalian D5/D1B receptor subtypes. Moreover, these data confirm the notion that SCH-23390 is a pure D1 antagonist, while butaclamol and flupentixol are inverse agonists at the D1B receptor and that the sequence specific motifs regulating this event are absolutely conserved within the D1B/D5 receptor family. Interestingly, the D1C receptor also exhibited a significant intrinsic activity promoting cAMP production (~1.5-2-fold) in transfected COS cells, albeit less pronounced than that of D1B and consistent with its binding profile, which is somewhat intermediate to those of the D1A and D1B receptors. As with the D1B receptor, the eel D1C receptor appears to recognize flupentixol and butaclamol as inverse agonists, inhibiting constitutive D1C adenylate cyclase activity to below basal levels.
In summary, pharmacological and functional data obtained from the four eel D1 receptors transiently expressed in a single cell type clearly support the conclusion drawn from the sequence analysis that they belong to three different D1 receptor subtypes, namely D1A, D1B, and D1C, which have undisputable homologues in the other vertebrate species.
Tissue Distribution of the Four D1-like Receptor Transcripts in the Eel BrainThe low abundance of most of the D1-like receptor transcripts in the eel brain prevented the use of Northern blot analysis to study their regional distribution. The D1A2 receptor mRNA was a notable exception that seems to be the most abundant in the eel brain (data not shown). We therefore used reverse transcription-PCR analysis to obtained a relative estimate of the amount of the four mRNA transcripts in the dissected telencephalon, midbrain, pituitary, cerebellum, and brain stem. After 25 cycles of PCR amplifications with gene-specific primers, the abundance of each of the D1-like transcript was measured by quantifying the corresponding hybridization signals relative to cytochrome b transcripts, used as an internal standard. As illustrated in Fig. 6 each of the D1-like receptors exhibited a differential distribution pattern. D1A1 mRNAs were found in all the brain regions, being the most abundant in the brainstem. In contrast, D1A2 was mainly expressed in the midbrain, as was also observed for D1B and D1C receptor mRNAs. D1C receptor transcripts were almost absent from the cerebellum. Surprisingly, the pituitary, where D1 receptor-mediated effects are best documented in teleost fish (27), is not a prominent area for D1-like transcription. In the the telencephalon, where dopamine projections are highly variable from one species to another (28), the four D1 mRNA subtypes were present at a high level. Quantification of the hybridization signals obtained for the D1A1 and D1A2 transcripts highlighted further their differential distribution and is documented in Table II. In summary, the data obtained from this crude distribution of the D1-like receptor mRNA transcripts provide evidence for the tissue-specific expression of the four corresponding genes in eel, as is also observed for the D1A and D1B receptors in mammals (29) or for D1A, D1B, and D1C/D1D in Xenopus and chicken (6, 7).
|
Four D1 dopamine receptors have been isolated by molecular cloning in the teleost fish A. anguilla (European eel). All these receptors are expressed in the brain and have been pharmacologicaly and functionally characterized.
The assignment of each of the eel sequences to a particular subtype of D1 receptor was important in order to analyze, through an evolutionary perspective, the functional significance of the dopamine D1 receptor genetic diversity observed in modern vertebrates. The major interest of a comparative approach is to point out the parameters that are really conserved and specific of each of the receptor subtypes and therefore relevant to the physiological role of each of these subtypes. In principle, several different criteria may be used to classify the various molecular forms of G protein-coupled receptors, which include the relative affinity and activity of ligands, the modulation of particular intracellular signaling pathways, the specific tissue distribution, and finally the sequence similarities analyzed by molecular phylogeny methods (5, 30). The convergence of these criteria is probably the best way to provide a robust definition and classification of each of the receptor subtypes isolated from different animal species.
One major criterion of receptor classification into distinct receptor subtypes depends on sequence identities. In this respect, we have recently emphasized that molecular phylogeny methods are very useful to unravel the relationships of the various monoamine receptors and, in particular, to identify orthologous versus paralogous receptors (5). Indeed, the "characters" represented by nucleotides or amino acids at each position in the sequence alignments constitute objective information that can be used for the classification of these molecules on the basis of their shared similarities and differences. Despite the limitations represented by high divergence rates, saturation of sequence similarities, or gene homogenization, the phylogenetic trees constructed from sequence alignments also provide valuable information regarding evolutionary relationships among corresponding molecules and, therefore, on gene duplication events giving rise to the molecular diversity currently observed. We document here one such example for members of the dopamine D1-like receptor family, which is so far the receptor class most largely analyzed in a wide range of vertebrate species.
As illustrated in Fig. 2, molecular phylogeny unambiguously assigns the four eel D1-like receptors to three D1 receptor subtypes named D1A, D1B, and D1C. As such, it provides strong support for the view that the D1C receptor reflects a true receptor subtype, common to most of the vertebrate species. Indeed, Macrae and Brenner (8) recently isolated two D1-like receptors from another teleost fish, the puffer fish Fugu rubripes. The phylogenetic tree clearly shows that these two receptor sequences are, respectively, orthologous to the D1A and D1C subtypes. We can now conclude that D1A and D1B orthologous receptors are found in all the vertebrate species analyzed to date, whereas D1C orthologues are up to now found only in fish and amphibians, challenging the presence of this subtype in mammals.
The presence of two distinct D1A receptors in the eel is also an intriguing feature brought to light by this study. We have accumulated evidence to show that D1A1 and D1A2 receptors are not simply different polymorphic alleles; their sequences are too different to correspond to polymorphic variations, since they are found in 12 different individuals fished from different parts of France and their mRNAs are differentially transcribed in the eel brain. These data imply that an ancestor D1A gene duplicated either recently in the eel lineage or more precociously in an ancestor of modern teleosts, although this assumption awaits the demonstration of the presence of two D1A receptors in other fishes. The duplication of the ancestor of the eel D1A receptor genes is probably not the consequence of genome tetraploidization, as is the case for the toad Xenopus laevis or salmons, since European eels as well as their Japanese relatives are not tetraploids (31).
Finally, the chicken D1D sequence (7), although too divergent from the other vertebrate D1-like sequences to be unambiguously assigned to one of the three D1 subtypes, shares impressive synapomorphies with the D1C receptor subtype. As such, it is still unclear whether the chicken D1D receptor represents a fast diverging D1C sequence or if it corresponds to a new paralogous subtype of the D1 receptor appearing late in the vertebrate phylum, perhaps specifically in the Thecodontia (birds and crocodiles) lineage.
The most commonly used criterion of receptor classification is based on the rank order of potency and relative binding affinity of various dopaminergic agonists and antagonists in vitro. Since this pharmacological parameter basically results from the ligand interactions with a small number of amino acid residues in the binding pocket of the receptors, the classification obtained by this criterion should closely resemble that obtained on the basis of sequence comparison. It is indeed the case for the eel D1 receptors that display pharmacological profiles that closely conform to those previously observed for each of the three D1 receptor subtypes (6, 7). In particular, the two eel D1A receptors (D1A1 and D1A2) have very similar binding abilities.
Comparison of the binding properties of the eel D1 receptors with those of the other vertebrate D1 receptors delineates salient conserved characteristics for each of the D1 receptor subtypes. Most importantly, a higher affinity of the D1B subtype for the endogenous ligand dopamine, as compared with D1A and D1C, is a distinguishing property of the D1B receptor subtype in all of the vertebrate receptors analyzed so far. In this respect, the D1C receptor displays an intermediate affinity for dopamine in both Xenopus and eel. Most antagonists bind the eel D1C receptor with estimated affinity values very close to that of D1A2. One interesting conserved discriminating feature, however, is the rank order of affinity for NPA (D1A > D1C > D1B), 6,7-ADTN (D1B > D1C > D1A), haloperidol (D1A > D1C > D1B), and spiperone (D1A > D1C > D1B). In contrast, the benzazepine SCH-23390, the canonical D1 receptor ligand, does not discriminate between the various receptor subtypes in all of the vertebrate species tested and should be considered indeed as a generic marker of the the D1 receptor family in vertebrates. Whether the newly described D1C receptor subtype can be further discriminated from D1A or D1B receptors to display exquisite sensitivity for other dopaminergic compounds or "second generation" D1-like agonists and antagonists is currently under investigation.
It is worth mentioning that drug-based discrimination of the various subtypes of vertebrate dopamine D1 receptors is better described by a "profile" of binding affinities for several drugs that can be readily compared in different species than by the particular properties of single "specific" ligands. Indeed, confusion in receptor classification may occur when minor sequence differences between orthologous receptors (species homologues) are recognized by specific ligands or, on the contrary, when paralogous receptors (true subtypes) are not distinguished by different ligands (32). Examples of such heterodox behavior are also observed in the D1 receptor family, where flupentixol displayed ~7-fold higher affinity for the chicken than for the Xenopus D1B receptor (7).
The third classification criterion is provided by the differential
coupling of the receptors to intracellular signaling molecules via
direct interactions with and
subunits of heterotrimeric G
proteins. This criterion is difficult to fully apply in the case of
cloned receptors, since their activity is evaluated by introducing them
in cell lines where the full range of "natural" intracellular
pathway activation may not always be obtained (33). Nevertheless, the
four eel D1-like receptors are able to significantly activate adenylyl cyclase in COS-7 cells. This property, which historically led to the definition of the D1 receptor
class, remains the key parameter of D1 receptor
characterization. From an evolutionary point of view, this
plesiomorphic property was acquired by the common ancestor of all of
the D1 receptors and should be found even in early
diverging species. This statement is supported by the fact that the
Drosophila dopamine D1 receptor mediates
dopamine activation of adenylyl cyclase, although this molecule retains little of the pharmacological profile that defines vertebrate D1 receptors (25).
Interestingly, cells transfected with the eel D1B receptors consistently exhibit a higher basal cAMP level than control, D1A-transfected, or D1C-transfected cells, as found in other species. It suggests that one functional differentiating characteristic of the D1B receptor is its constitutive activity, a property inherent in the mammalian D5/D1B receptor (34, 35) and which appears to have been absolutely conserved throughout the evolutionary course of the D1B receptor subfamily. As such, constitutive activation of adenylate cyclase by the D1B receptor system appears fundamental to this receptor subtype and functionally relevant to the physiology of dopamine in the vertebrate nervous system. At present, it is difficult to ascertain whether all vertebrate D1C receptors share, at a somewhat attenuated level, the ability to display constitutive activity (see Fig. 5) and whether this property is common throughout the evolutionary history of the D1C receptor subtype. In this regard, the Xenopus D1C receptor does not appear to be constitutively active (6), although direct comparison with the eel D1C receptor is difficult due to the widely disparate levels of receptor expression. The precise determination of D1C subtype characteristics regarding adenylyl cyclase modulation will require analysis in a larger set of animal species in order to determine whether its relatively high intrinsic activity is indeed conserved.
The fourth criterion that may identify receptor subtypes expressed in a single animal species is its tissue specific distribution profile. Although, as of yet, we were not able to provide a precise and complete description of the tissue distribution of the four eel D1 receptor mRNAs, the semiquantitative PCR analysis suggested that the various eel D1 receptor subtypes are differentially transcribed in various segments of the eel brain. Differential distribution of receptor mRNAs also characterizes mammalian D1A and D1B receptors. D1A receptors are present at very high levels in the striatal and olfactory regions, whereas D1B receptors are expressed mainly in the hippocampus and cortex in the human and also in the parafascicular nucleus of the thalamus in rat (29, 36, 37). The localization of the D1A and D1B receptors seems to be essentially nonoverlapping in mammals. This characteristic could be extended to the other vertebrate species and may render the multiplicity of D1 receptors essentially nonredundant. The physiological consequence for the expression of four dopamine D1 receptors in a teleost fish can be only hypothetical at present, but it probably relates to dopamine functions (not only cellular effects) selected for eel adaptation to a changing milieu. European eels, like their American and Asian relatives, have a very complex physiology in which dopamine is thought to contribute significantly to the sensorimotor, feeding, and reproductive behavior during its whole life cycle. The role of dopamine in such a life cycle could have driven the conservation of more dopamine receptor subtypes than found in other vertebrates living in more constant environments. Be that as it may, the acquisition of differential expression territories by the various duplicated genes during evolution would be a mechanism of utmost importance for the conservation of paralogous genes.
In summary, the comparison of the the eel D1 receptor sequences as well as of some of their pharmacological and functional characteristics with those of the other D1 vertebrate subtypes, although still incomplete, pinpoints defining characteristics and features specific for each of the D1A, D1B, and D1C receptor subtypes found in vertebrates. Understanding the physiological relevance of these defined functional homologies will now require an appreciation of the relationship between the differential localization of the various D1 receptor subtypes within the brains of the main groups of vertebrates and the synaptic organization of dopaminergic pathways in these species.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U62918[GenBank], U62919[GenBank], U62920[GenBank], U62921[GenBank].
We thank Delphine Valente (Commissariat à l'Energie Atomique, Saclay, France) for help with cAMP assays and Sylvie Dufour (Museum National d'Histoire Naturelle, Paris, France) for having introduced us to the eel world.