Abstract
Lyso-PS (lyso-phosphatidylserine) has been shown to activate the Gi/o-protein-coupled receptor GPR34. Since in vitro and in vivo studies provided controversial results in assigning lyso-PS as the endogenous agonist for GPR34, we investigated the evolutionary conservation of agonist specificity in more detail. Except for some fish GPR34 subtypes, lyso-PS has no or very weak agonistic activity at most vertebrate GPR34 orthologues investigated. Using chimaeras we identified single positions in the second extracellular loop and the transmembrane helix 5 of carp subtype 2a that, if transferred to the human orthologue, enabled lyso-PS to activate the human GPR34. Significant improvement of agonist efficacy by changing only a few positions strongly argues against the hypothesis that nature optimized GPR34 as the receptor for lyso-PS. Phylogenetic analysis revealed several positions in some fish GPR34 orthologues which are under positive selection. These structural changes may indicate functional specification of these orthologues which can explain the species- and subtype-specific pharmacology of lyso-PS. Furthermore, we identified aminoethyl-carbamoyl ATP as an antagonist of carp GPR34, indicating ligand promiscuity with non-lipid compounds. The results of the present study suggest that lyso-PS has only a random agonistic activity at some GPR34 orthologues and the search for the endogenous agonist should consider additional chemical entities.
- chimaeric receptor
- evolution
- mutagenesis
- orphan G-protein-coupled receptor (GPCR)
- structure–function relationship
INTRODUCTION
GPR34 is an orphan GPCR (G-protein-coupled receptor) and was first discovered by mining GenBank® for novel GPCR sequences and homology cloning. It has been assigned to the human X chromosome [1,2].
GPR34 belongs to the P2Y12-like receptor group within the family of rhodopsin-like receptors. The P2Y12-like receptor group comprises the ADP receptors P2Y12 and P2Y13, the UDP-glucose receptor P2Y14 and the orphan receptors GPR87, GPR82 and GPR34 [3]. The ADP receptor P2Y12 has a central role in platelet aggregation and is the therapeutic target of clopidogrel [4,5]. However, very little is known about the function of the other members of this group.
Several members of the P2Y12-like receptor group have been assigned to agonists, including nucleotide derivates and lipids [6–8]. Specifically, human GPR34 was shown to be activated by lyso-PS (lyso-phosphatidylserine) in vitro. Lyso-PS is generated by hydrolysis of membrane lipids through phospholipases A1 and A2, e.g. when apoptotic cells expose PS on their surface to these phospholipases [9,10]. Lyso-PS is a potent activator of histamine release from mast cells [11]. Furthermore, lyso-PS has been described as a growth inhibitor of T-cells and as a chemotactic substance for fibroblasts and tumour cells [10–13]. Some parasite-derived lyso-PS species can directly activate Toll-like 2 receptors [14].
Phylogenetic studies have shown that GPR34 has been conserved over the past 450 million years of vertebrate evolution. In contrast with tetrapods, several fish genomes contain two or more GPR34 subtypes [15]. GPR34 appears to be highly relevant for vertebrate physiology, because no GPR34-deficient vertebrate has been identified yet. GPR34-deficient mice are viable and do not display any obvious differences in morphology, histology, laboratory chemistry or behaviour in a standard laboratory environment compared with their wild-type littermates [16]. Detailed expression analyses showed GPR34 expression in immune cells (macrophages and microglia) [16,17], immune system-derived cell lines (HL-60, K562, WEHI-3B, RAW 264.1 and P815) [18] and lymphoma cells [19,20]. Examinations under different immune system-challenging conditions revealed that GPR34 deficiency interferes with a proper immune response and results in a reduced resistance to systemic infections [16]. However, two independent studies showed that mast cells derived from GPR34-deficient mice still degranulate upon lyso-PS stimulation [16,21]. Furthermore, in vitro studies performed with human and mouse GPR34 constructs using different expression and functional systems provided no evidence for P-lyso-PS (1-palmitoyl-lyso-PS)-mediated receptor activation [16,22]. Differences in the functional assay setup, cellular expression systems as well as the different purity and chemical composition of P-lyso-PS may account for this discrepancy with previous findings [8,23]. However, further studies are needed to differentiate between the surrogate agonistic activity of lyso-PS and the clear evidence that lyso-PS/GPR34 is the naturally selected matching pair of agonist and receptor.
In the present study, we assessed the evolutionary conservation of lyso-PS action at several GPR34 orthologues. Only carp GPR34 subtypes displayed significant activity upon lyso-PS stimulation in a heterologous yeast and mammalian cell expression system. Systematically exchanged portions between human and carp orthologues revealed multiple determinants involved in this species-dependent agonist specificity, mainly located in EL2 (second extracellular loop) and TM5 (transmembrane region 5). Our results show that lyso-PS has some residual activity at GPR34 but is not the natural agonist at mammalian GPR34 orthologues.
EXPERIMENTAL
Materials
If not stated otherwise, all standard substances were purchased from Sigma–Aldrich, Merck and Carl Roth. Cell culture material was obtained from Sarstedt and primers were purchased from Invitrogen. For expression of GPR34 orthologues in yeast, the p416GPD vector [24] was used, whereas mammalian expression was performed using the pcDps vector [25] and the pcDNA5/FRT (Invitrogen). Restriction enzymes were purchased from New England Biolabs. ‘Brain’-lyso-PS, whose main component is S-lyso-PS (1-stearyl-lyso-PS), and all other phospholipids were obtained from Avanti Polar Lipids. The adenine nucleotide library was from JenaBioscience (for compound details see http://www.jenabioscience.com/images/7c63e6fc71/LIB-101.pdf).
Preparation and purification of P-lyso-PS
As the P-lyso-PS used by Sugo et al. [8] was no longer commercially available from Sigma–Aldrich or any other company, P-lyso-PS was synthesized by hydrolysis of 2-dipalmitoyl-sn-glycero-3-phospho-L-serine with phospholipase A2. P-lyso-PS was purified from the reaction mixture by extraction and TLC as described previously [16]. S-lyso-PS was used for most studies, since P-lyso-PS and S-lyso-PS always gave identical results with the wild-type orthologues.
Identification of GPR34 orthologues
Mining of NCBI (National Center for Biotechnology Information) trace archives
GPR34 sequences of various vertebrate species were obtained using the respective mouse orthologue nucleotide sequence as a query entry in the sequence analysis service Mega BLAST. The discontiguous version was used to search through all available vertebrate trace archives. Trace files of sequences producing significant alignments were downloaded, followed by assembly, analysis (using SeqManPro of DNAStar Lasergene Software Suite for Sequence Analysis 7.1.) and manual proof-reading. Orthologues analysed in the present study are listed in Supplementary Table S1 (at http://www.BiochemJ.org/bj/443/bj4430841add.htm).
Amplification, sequencing and cloning of GPR34 orthologues
To analyse additional GPR34 orthologue sequences, genomic DNA samples were prepared from tissues of various species (given in Supplementary Table S1). Tissue samples were digested in lysis buffer (50 mM Tris/HCl, pH 7.5, 100 mM EDTA, 100 mM NaCl, 1% SDS and 0.5 mg/ml proteinase K) and incubated at 55°C for 18 h. DNA was purified by phenol/chloroform extraction and ethanol precipitation. Degenerated primer pairs [15] were applied to amplify GPR34-specific sequences. PCR reactions were performed with Taq polymerase under variable annealing and elongation conditions. Specific PCR products were directly sequenced and/or beforehand subcloned into the pCR2.1-TOPO vector (Invitrogen). In case of heterozygosity, allelic separation was performed by subcloning and subsequent sequencing. Sequencing reactions were performed with a dye-terminator cycle sequencing kit and applied on a MegaBACE™ 1000 (GE Healthcare).
Generation of mutant GPR34
All GPR34 constructs were introduced into the mammalian expression vector pcDps and double-tagged with an N-terminal HA (haemagglutinin) tag and a C-terminal FLAG tag to quantify total cellular and plasma membrane expression using ELISA [26]. GPR34 chimaeras and mutants were generated by PCR-based mutagenesis and fragment-replacement strategies. All PCR-derived constructs were verified by sequencing.
Cell culture, transfection and functional assays
The haploid Saccharomyces cerevisiae yeast strain MPY578t5 (provided by Dr Mark Pausch, Wyeth Research, Discovery Neuroscience, CN 8000, Princeton, NJ, U.S.A.) was used for the expression of GPR34 orthologues. Cells were transformed with plasmid DNA using electroporation as described previously [27].
The adenine nucleotide compound library was screened for agonists and antagonists/inverse agonists at the carp GPR34 2b in the yeast expression system. The respective yeast cell suspension (100 μl; D600=0.1) was pipetted into each well of a 96-well plate and, to these samples, 100 μl of a 2× ligand solution or medium was added. Attenuance measurements were performed 24 and 48 h later. Compounds identified to be antagonists were further characterized in concentration–response setups.
For expression in mammalian cells, COS-7 and CHO (Chinese hamster ovary)-K1 cells were grown in DMEM (Dulbecco's modified Eagle's medium), supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin and 100 μg/ml streptomycin, at 37°C in a humidified 5% CO2 incubator. Transient cotransfection experiments of COS-7 cells with the respective GPR34 constructs and the chimaeric G-protein Gαqi4 and inositol phosphate accumulation assays [28] were performed as described previously [15]. To generate stably transfected CHO cell lines, GPR34 constructs were subcloned into pcDNA5/FRT (Invitrogen) and transfected. To select stably transfected cells, CHO cells were kept under hygromycin (500 μg/ml). For cAMP assays, transfected cells were labelled with [3H]adenine (2 μCi/ml; PerkinElmer) for 12 h and washed once in serum-free DMEM containing 1 mM 3-isobutyl-1-methylxanthine (Sigma–Aldrich), followed by incubation in the presence of the indicated compounds and forskolin (10 μM) for 1 h at 37°C. Reactions were terminated by aspiration of the medium and addition of 1 ml of 5% (w/v) trichloroacetic acid. The cAMP content of cell extracts was determined by anion-exchange chromatography as described previously [29]. To measure label-free receptor activation, a dynamic mass redistribution assay (Corning Epic® Biosensor Measurements) with stably transfected CHO cells was performed as described previously [30].
To estimate cell surface expression of receptors carrying an N-terminal HA tag, we used an indirect cellular ELISA [31]. To assess the amounts of full-length HA/FLAG double-tagged GPR34 constructs and to demonstrate that the reduction of cell surface expression levels is not due to a decrease in receptor expression in general, a sandwich ELISA was used and performed as described previously [32].
Sequence alignments and PAML (phylogenetic analysis by maximum likelihood) analyses
A GPR34 nucleotide alignment (corresponding to amino acid positions 1.23–7.70; relative numbering system of GPCR based on [33]) of all functionally tested GPR34 orthologues was generated with the ClustalW algorithm (Bioedit Sequence Alignment Editor 7.0.9; http://www.mbio.ncsu.edu/BioEdit/bioedit.html [34]) followed by manual trimming. Phylogenetic tree inference was conducted in MEGA4 [35] using the neighbour-joining method [36], whereas the evolutionary distances were computed applying the maximum composite likelihood method [37]. Branch support was estimated with 1000 bootstrap replicates [38], resulting in a fully resolved tree representing phylogenetic relationships (see Figure 4A and Supplementary Table S4 at http://www.BiochemJ.org/bj/443/bj4430841add.htm). Tests of selection (ω=dN/dS) were accomplished by maximum likelihood using a codon-based substitution model implemented in PAML version 4.2 [39]. Branch models [40] were applied to determine the mode of evolution of functionally tested GPR34 orthologues. Branch-site models were used to detect positive selection affecting only a few sites on pre-specified foreground branches (test 2: modified model A against modified model A with ω2=1 fixed) [41,42]. All analyses were run twice using different initial ω values to check for convergence. LRTs (likelihood ratio tests) were performed to test nested competing hypotheses.
Vertebrate GPR34 nucleotide alignments (corresponding to amino acid positions 3.52–7.48; relative numbering system of GPCR based on [33]), were generated with ClustalW algorithm (Bioedit Sequence Alignment Editor 7.0.9), followed by manual trimming. Phylogenetic trees were inferred as described above. Trees that were not fully resolved were manually constructed for analyses on the basis of respective mammal [43,44] and fish phylogenies [45–54]. All trees used in PAML are stated in Supplementary Table S4 in Newick format.
RESULTS AND DISCUSSION
Lyso-PS is an agonist at carp GPR34 subtypes but not at the mouse and human GPR34
Lyso-PS has been assigned as an agonist at GPR34 [8,23]. It has been shown that mast cells express GPR34 at high levels [8]. Indeed, lyso-PS is a potent activator of histamine release from mast cells [11]. However, lyso-PS-induced histamine release was not altered in mast cells from GPR34-deficient mice [16,21]. To address this obvious discrepancy, we and others tested lyso-PS at GPR34 in several functional in vitro assays. In the present and previous independent studies [16,22], all attempts, including cAMP inhibition assays (Figure 1A), fura-2 calcium measurement, and studies with chimaeric G-proteins linking signal transduction of a Gi-protein-coupled receptor to the phospholipase C/inositol phosphate pathway [15] with transiently and stably transfected mammalian cells failed to clearly verify that lyso-PS is a robust agonist for the mouse and human GPR34. A small reduction of the forskolin-induced cAMP levels was found for mouse GPR34 and human GPR34 (Figure 1A). This may explain the agonistic effect of lyso-PS seen by other groups [8,23]. This effect was not observed in a dynamic mass redistribution assay [30] with stably transfected CHO-K1 cells (Figure 1B). However, we found that P-lyso-PS and S-lyso-PS activate the carp GPR34 subtypes 2a and 2b heterologously expressed in yeast and mammalian cells (Figure 1 and Supplementary Table S2 at http://www.BiochemJ.org/bj/443/bj4430841add.htm). As shown in Figure 1(C), all three carp GPR34 subtypes responded to S-lyso-PS in a concentration-dependent manner. Carp GPR34 subtypes showed significant differences in basal activity and EC50 values. GPR34 orthologues from human (see Figure 1C), mouse, chicken, Xenopus and zebra fish showed no significant response in the yeast cell expression system (Table 1). Furthermore, different phospholipids were tested at human, mouse and carp GPR34 orthologues. Among many other lyso-lipids tested, only S-lyso-P-choline was a weak agonist at the carp GPR34 type 2, but not at mouse and human orthologues (Supplementary Table S2). To prove that all constructs are expressed in the cell and at the cell surface, ELISA studies were performed. As shown in Table 1, all orthologues were expressed and properly delivered to the cell surface.
GPR34 orthologues that were analysed are shown in this Table. Yeast cells expressing these receptors were grown as described in the Experimental section. D600 readings were performed after 24 h to measure basal activity and growth under stimulation with S-lyso-PS. Additionally, COS-7 cells were transfected with receptor constructs and cell surface expression and total expression levels were analysed. Cell surface and total expression levels of constructs were measured by cell surface and total cellular ELISA (see the Experimental section). Specific attenuance readings (D492/620 value of HA/FLAG-tagged construct minus D492/620 value of GFP-transfected cells) are given as a percentage of human GPR34. Results are the means ± S.D. of three independent experiments, each performed in triplicate. *P<0.05 and ***P<0.001, significantly different from basal D600 (Student's t test).
(A) To test functional properties of GPR34 orthologues in a mammalian cell expression system, CHO cells were stably transfected with the indicated constructs and cAMP inhibition assays were performed as described in the Experimental section. Cells were incubated with forskolin (10 μM) alone or together with S-lyso-PS (10 μM) and cAMP levels were determined. Results are given as means ± S.D. of three experiments. (B) For label-free measurements of receptor activation, a dynamic mass redistribution assay (Corning Epic® Biosensor Measurements) with stably transfected CHO cells was performed essentially as described previously [30]. The response 6 min after application of S-lyso-PS is shown. ATP served as a positive control to verify cell vitality and responsiveness. Results are given as means ± S.D. of three independent measurements. (C) Yeast cells expressing the indicated GPR34 orthologues were incubated with increasing concentrations of S-lyso-PS. Receptor activation-dependent growth was measured as D600 after 24 h. EC50, Emax, basal values of the concentration–response curve and cell surface/total cellular expression levels in COS-7 cells are given in Table 1. Results are given as means ± S.E.M. of two representative assays, each performed in triplicate. **P<0.01, ***P<0.001.
The fact that lyso-PS displayed very species-specific agonistic activity at GPR34 argues strongly against the assignment of GPR34 as the specific receptor for lyso-PS. Usually, there is very high consistency in the signalling properties of non-peptide GPCR orthologues and their designated agonists during vertebrate evolution. Although receptor subtypes (paralogues) can differ in their signal transduction and expression regulation, the potency of agonists at different subtypes remains very constant (e.g. aminergic GPCR). GPR34 subtypes only exist in fishes [15]. The carp GPR34 subtypes showed very different EC50 values for lyso-PS (compare type 1 with type 2, Figure 1C). This would suggest a very unlikely scenario that the GPR34 subtypes monitor a lyso-PS gradient, but only in some fishes.
Determinants in EL2 and TM5 are responsible for species-specific receptor activation by S-lyso-PS
Usually, receptor–ligand interaction is optimized during evolution, and, only in rare cases, mutations in GPCR can increase efficacy and potency without alteration of other properties, such as specificity, cell surface expression, basal activity or protein stability. So it should not be possible to significantly improve GPCR functionality by mutations. To prove whether the human GPR34 can gain function upon lyso-PS application by introducing parts of the carp GPR34 2a, we systematically generated (Supplementary Table S3 at http://www.BiochemJ.org/bj/443/bj4430841add.htm) and tested carp/human GPR34 (Table 2). As shown in Figure 2(A), replacing TM3–TM5 (Chim5) with the respective carp subtype 2a region resulted in an S-lyso-PS activation of human GPR34. In contrast, the opposite chimaera, TM3–TM5 of human GPR34 in the carp subtype 2a (Chim2), showed no function. Successive exchange of portions within the TM3–TM5 region (Chim8 and Chim10) revealed that all constructs containing carp EL2 and TM5 responded to S-lyso-PS (Figure 2A, Table 2). Again, proper construct expression was monitored via cell surface and total expression ELISA (Table 2).
(A) Yeast cells expressing human GPR34 (hGPR34), carp GPR34 subtype 2a (cGPR34 2a) and chimaeric constructs (see Supplementary Table S2 at http://www.BiochemJ.org/bj/443/bj4430841add.htm) were incubated with increasing concentrations of S-lyso-PS. Receptor activation-dependent growth was measured as D600 after 24 h. EC50, Emax, basal values of the concentration–response curve and cell surface/total cellular expression levels in COS-7 cells are given in Table 2. Results are given as means ± S.E.M. of two representative assays, each performed in triplicate. (B) To demonstrate functional properties in a mammalian cell expression system, COS-7 cells were transfected with the indicated constructs and a Gqi4 chimaeric G-protein. S-lyso-PS was applied in equal concentrations (10 μM). GFP (green fluorescent protein)-transfected cells were used as a negative control. Inositol phosphate (IP) assays were performed as described in the Experimental section. Results are given as means ± S.E.M. (n=4), with each experiment performed in triplicate.
Structural determinants enabling the carp GPR34 2a to recognize S-lyso-PS as agonist were approached by systematic generation and testing of carp/human GPR34 chimaeras. Ten chimaeric constructs with transmembrane-wide exchanges are shown in this Table. Yeast cells expressing these chimaeric receptors were grown as described in the Experimental section. D600 readings were performed after 24 h to measure basal activity and growth under stimulation with S-lyso-PS. Additionally, COS-7 cells were transfected with receptor constructs and cell surface expression and total expression levels were analysed. Cell surface and total expression levels of constructs were measured by cell surface and total cellular ELISA (see the Experimental section). Specific attenuance readings (D492/620 value of HA/FLAG-tagged construct minus D492/620 value of GFP-transfected cells) are given as a percentage of human GPR34. Results are the means ± S.D. of three independent experiments, each performed in triplicate. ***P<0.001, significantly different from basal D600 (Student's t test).
To test whether receptor pharmacology is retained in a mammalian expression system, COS-7 cells were transfected with human and carp GPR34 orthologues and the constructs Chim2 and Chim5. For measurements in inositol phosphate assays, all constructs were cotransfected with the Gqi4 chimaeric G-protein [15]. As shown in Figure 2(B), carp subtype 2a and Chim5 displayed agonist-induced inositol phosphate production, whereas human GPR34 and Chim2 did not respond upon S-lyso-PS stimulation. We found no elevated basal activity.
With respect to basal activity, Emax and EC50 values, the function of Chim8 was very similar to carp GPR34 2a, indicating that probably all determinants required for successful activation are located within TM4–TM5. Therefore, this receptor portion was systematically subdivided to identify distinct positions involved in agonist specificity (Chim11–Chim20, Table 3). First, comparison of Chim8 against Chim11 and Chim17 against Chim18 revealed that the presence of EL2 mainly increased basal receptor activity, but contributed only partially to agonist specificity. Secondly, the N-terminal third of TM4 did not contribute to functional differences between carp and human orthologues (Chim8 against Chim12) (Table 3). Thirdly, the middle third of TM5 significantly contributed to agonist specificity, because S-lyso-PS lost almost all of its agonistic activity at Chim14, whereas Chim17 showed no loss (Table 3). Similar results were obtained with Chim17 against Chim19 with replacement of both parts.
Final structural determinants were built into the human GPR34 to switch over this receptor into an S-lyso-PS-recognizing GPCR. Chimaeric constructs and different amino acid-exchange mutants are shown in this Table. Yeast cells expressing these chimaeric receptors were grown as described in the Experimental section. D600 readings were performed after 24 h to measure basal activity and growth under stimulation with S-lyso-PS. Additionally, COS-7 cells were transfected with receptor constructs and cell surface expression and total expression levels were analysed. Cell surface and total expression levels of constructs were measured by cell surface and total cellular ELISAs (see the Experimental section). Specific attenuance readings (D492/620 value of HA/FLAG-tagged construct minus D492/620 value of GFP-transfected cells) are given as a percentage of human GPR34. Results are means ± S.D. of three independent experiments, each performed in triplicate. *P<0.05, **P<0.01 and ***P<0.001, significantly different from basal D600 (Student's t test)
Next, we checked whether the determinants within TM5 (Chim20) and EL2 (Chim21) can mediate agonist specificity individually. Thus the specific portions in human GPR34 were exchanged with the respective parts of carp GPR34 2a. As shown in Table 3, neither carp TM5 nor carp EL2 were able to restore full function in human GPR34 alone, indicating some co-operation of both receptor parts.
In summary, the systematic replacement approach revealed portions in the carp GPR34 which enabled the human GPR34 to recognize lyso-PS as an agonist. We next attempted to identify distinct determinants responsible for this gain of functionality.
Identification of single amino acid positions determining the species-specific receptor activation by S-lyso-PS
Structural comparison of the EL2/TM5 region between GPR34 orthologues responding or not responding to S-lyso-PS revealed several residues co-migrating with the functional phenotype (Supplementary Figure S1 at http://www.BiochemJ.org/bj/443/bj4430841add.htm). Thus, Lys210 and Gly215 and a segment between amino acid position 231–236 of human GPR34 were successively replaced with those residues found in carp GPR34 2a. As shown in Figure 3 and Table 3, double mutant K210L/G215W partially restored agonistic activity of S-lyso-PS in the human GPR34. However, the Emax and the basal activity values were significantly reduced when compared with carp GPR34 2a. Additional introduction of the TM5 segment (Chim22 and Chim23) almost completely installed agonistic activity of S-lyso-PS (see Figure 3). Interestingly, the pharmacology of Chim22 and Chim23 was very similar to carp GPR34 2b (see Figure 1 and Supplementary Table S2). The two carp GPR34 subtypes differ in the TM5 portion only at two positions (see Supplementary Figure S1). For control purposes, the respective mutations in EL2 (L187K/W192G) abolished the function of carp GPR34 2a (Table 3).
Yeast cells expressing human GPR34 (hGPR34), carp GPR34 2a (cGPR34 2a) and chimaeric constructs (see Supplementary Table S2 at http://www.BiochemJ.org/bj/443/bj4430841add.htm) were incubated with increasing concentrations of S-lyso-PS. Receptor activation-dependent growth was measured as D600 after 24 h. EC50, Emax, basal values of the concentration–response curve and cell surface/total cellular expression levels in COS-7 cells are given in Table 3. Results are given as means ± S.E.M. of two representative assays, each performed in triplicate.
Our results clearly show that activation of human GPR34 by S-lyso-PS requires the modification of multiple positions in EL2 and TM5. The modified positions correspond to the ligand-binding pocket rather than to structures involved in G-protein interaction. However, one cannot exclude changes that now enable lyso-PS to activate human GPR34.
Mode of GPR34 evolution
Most vertebrate GPCRs show a purifying mode of evolution (negative selection), reflecting high structural conservation. Species-specific differences in orthologue functionality should result in a significantly reduced structural conservation. In contrast with tetrapods, many fish genomes contain more than one copy of GPR34. One copy or even both parts of the receptor–gene pair may mutate and acquire unique functionality without risking the fitness of the organism, which is ensured by the homologue. Furthermore, gene duplications often retained overlapping expression patterns and preserved partial to complete redundancy consistent with a role in boosting robustness or gene doses. On the other hand, if not advantageous, continuous accumulation of mutations (neutral evolution) will eliminate one of the duplicated genes. As for other genes, disadvantageous mutations in GPCR are removed from a population through purifying selection and kept by positive selection of distinct positions. As several GPR34 subtypes were retained in fish species, this may be indicative of either new functionality or redundancy. The difference in the functionality found between GPR34 orthologues and subtypes may reflect positive selection, but also stochastic ligand promiscuity as found for other GPCRs [55].
To analyse the mode of GPR34 evolution, several tests for positive selection of subtype branches and specific positions were performed. All mammalian, avian, reptilian and amphibian GPR34 orthologues evolved under strong purifying selection (Supplementary Table S5 at http://www.BiochemJ.org/bj/443/bj4430841add.htm). This indicates that the structure/function-determining GPR34 sequence is highly preserved during tetrapod evolution. GPR34 of sharks and rays displayed slightly increased ω values (ω=dN/dS, the number of non-synonymous substitutions per non-synonymous site (non-synonymous substitution rate dN) normalized to the number of synonymous substitutions per synonymous site (synonymous substitution rate dS), most likely owing to some changes in constraint among these orthologues. Comparing the ω of fish GPR34 subtype 1 and fish GPR34 subtype 2 revealed higher values for subtype 2. GPR34 proteins within six species were functionally tested, and only carp orthologues showed responsiveness towards S-lyso-PS. Therefore we generated a phylogenetic tree (Figure 4A) including all these orthologues and applied different branch and branch-site models to test for positive selection in carp GPR34 orthologues. This approach [applying branch-site model A with all carp GPR34 subtypes or only carp GPR34 2 subtypes as foreground lineages in the branch-site test of positive selection (null model ω2=1 fixed)] provided significant support (P=0.003 and P=0.014 respectively) for several positively selected sites, with BEB (Bayes empirical Bayes) probability above 0.8 (see Supplementary Tables S5 and S6 at http://www.BiochemJ.org/bj/443/bj4430841add.htm). The particular amino acids are highlighted in Figure 4(B). Many of these are found in EL2 and TM5, further supporting the hypothesis that determinants in this region gave raise for a change in agonist specificity of these orthologues (see above). Further analyses of the region 3.52–7.48 (relative numbering system of GPCR based on [33]) of especially Otocephala GPR34 orthologues revealed some more positively selected sites in zebrafish, goldfish and carp GPR34 subtype 2 orthologues (P=0.006) and only goldfish and carp GPR34 subtype 2 (P=0.003) (Supplementary Tables S5 and S6 and Supplementary Figure S2 at http://www.BiochemJ.org/bj/443/bj4430841add.htm).
A phylogenetic tree of all functionally tested GPR34 orthologues was constructed. GPR34 nucleotide alignments (corresponding to amino acid positions 1.23–7.70; relative numbering system of GPCR based on [33]) were generated with the ClustalW algorithm (Bioedit Sequence Alignment Editor 7.0.9; http://www.mbio.ncsu.edu/BioEdit/bioedit.html [34]) followed by manual trimming. (A) A ‘free ratio’ model implemented in PAML [39] was used to calculate dN/dS ratios (ω shown in bold below branches) and the number of non-synonymous and synonymous substitutions (shown in parentheses) for each branch. (B) Using branch-site model A with the carp GPR34 branches or the carp GPR34 2a and 2b branches as foreground lineage in the branch-site test of positive selection (null model ω2=1 fixed), LRT provided significant support (P=0.003 and P=0.014 respectively) for positively selected sites with BEB probability above 0.8 as highlighted (Supplementary Tables S4–S6 and Supplementary Figure S2 at http://www.BiochemJ.org/bj/443/bj4430841add.htm).
Altogether, phylogenetic analyses revealed a reduced constraint in carp and other Otocephala GPR34 orthologues. Most of this signal is caused by positive selection of several residues in EL2 and TM5 (see Figure 4). This is consistent with the finding that the ability to recognize lyso-PS as an agonist can be installed in human GPR34 by replacing EL2 and TM5 of carp GPR34 2a. However, there is only a partial overlap of selected positions (see Figure 4) and conservation (see Supplementary Figure S1), with single residues sufficient to induce changes in agonist specificity. This indicates that structural changes enabling lyso-PS to activate carp GPR34 orthologues are rather accidental and make lyso-PS surrogate agonists, as found, for example, in trace-amine-associated receptors [55].
Identification of a nucleotide derivative as an antagonist at carp GPR34
Our functional and evolutionary results suggest that lyso-PSs are surrogate agonists at GPR34. GPCRs are not always highly specific and, for example, many aminergic and nucleotide receptors display high ligand promiscuity [7,55,56]. GPR34 is phylogenetically related to nucleotide receptors (P2Y12, P2Y13 and P2Y14) [3]. Signalling of carp orthologues provided the unique opportunity to screen for potential nucleotide derivates as ligands for GPR34. Thus a comprehensive nucleotide compound library was screened for ligands in the absence (screen for agonists) and presence (screen for antagonists) of S-lyso-PS. Although there was no agonist among ~80 purine nucleotide compounds, we identified EDA-ATP [2′/3′-O-(2-aminoethyl-carbamoyl)-ATP] as a low-affinity antagonist (Figure 5A) that blocks S-lyso-PS activation of carp GPR34 in yeast and mammalian expression systems in a competitive manner (Figure 5B). In mammalian CHO cells, EDA-ATP blocked S-lyso-PS-mediated carp GPR34 activation and reduced the receptor basal activity, indicating inverse agonistic activity (Figure 5C). This supported our hypothesis further that GPR34 shows ligand promiscuity.
(A) Yeast cells expressing the human ADP receptor P2Y12 (control) and carp GPR34 2b (cGPR34 2b) were incubated with a comprehensive compound library in the presence of MeS-ADP [2-(methylthio)adenosine 5′-diphosphate; an agonist of P2Y12] and S-lyso-PS. Receptor activation-dependent growth was measured as D600 after 18 h and set to 100% (P2Y12 activity at 10 μM MeS-ADP: 0.431±0.007; carp GPR34 2b activity at 10 μM S-lyso-PS: 0.310±0.121). EDA-ATP was identified as an antagonist at carp GPR34. Results are given as means ± S.D. of two independent assays performed in triplicate. (B) S-lyso-PS-induced yeast cell growth showed a right-shifted concentration–response curve when co-incubated with 10 μM EDA-ATP, indicating orthosteric binding of the antagonist at carp GPR34 2b. (C) Forskolin-induced cAMP levels in CHO cells stably expressing carp GPR34 subtypes were determined in the presence of S-lyso-PS and EDA-ATP. Results are given as means ± S.D. (n=3).
Conclusions
Using different expression and functional systems we and other researchers [21,22] cannot provide clear evidence for activation of the human and mouse GPR34 by lyso-PS in vitro and in vivo, in contrast with previous findings [8,23]. Differences in the functional assay setup, cellular expression systems, and differences in the purity and chemical composition of P-lyso-PS used may account for this discrepancy. However, a line of evidence, including phylogenetic and functional data, support lyso-PS species as surrogate agonists: (i) lyso-PS action is not conserved during vertebrate evolution, although the orthologues are under strong negative selection; (ii) the potency of lyso-PS significantly differs between fish GPR34 subtypes; (iii) mutations in human GPR34 can significantly improve signalling; and (iv) GPR34 shows ligand promiscuity to an ATP derivative. Therefore our results suggest that lyso-PS has only random agonistic activity at some GPR34 orthologues and the search for the endogenous agonist should consider additional chemical entities.
AUTHOR CONTRIBUTION
Lars Ritscher constructed the chimaeric receptors and performed all of the functional experiments. Thomas Hermsdorf performed EPIC measurements. Eva Engemaier sequenced the fish GPR34 orthologues. Claudia Stäubert and Holger Römpler performed the phylogenetic studies. Ines Liebscher contributed the preliminary work on carp GPR34. Philipp Schmidt screened a nucleotide library for GPR34 ligands. Angela Schulz and Torsten Schöneberg organised the study and wrote the manuscript in co-operation with all the authors.
FUNDING
This work was supported by the Deutsche Forschungsgemeinschaft [grant numbers Scho 624/7-1 and FOR 748] and the State of Saxony (LIFE).
Acknowledgments
We thank Susann Lautenschläger for excellent technical assistance.
Abbreviations: BEB, Bayes empirical Bayes; CHO, Chinese hamster ovary; DMEM, Dulbecco's modified Eagle's medium; EDA-ATP, 2′/3′-O-(2-aminoethyl-carbamoyl)-ATP; EL, extracellular loop; GPCR, G-protein-coupled receptor; HA, haemagglutinin; LRT, likelihood ratio test; PAML, phylogenetic analysis by maximum likelihood; P-lyso-PS, 1-palmitoyl-lyso-phosphatidylserine; PS, phosphatidylserine; S-lyso-PS, 1-stearyl-lyso-phosphatidylserine; TM, transmembrane region
- © The Authors Journal compilation © 2012 Biochemical Society
References
May 2012
Volume: 443 Issue: 3
Open in Utopia Docs
