Dissection of functional residues in receptor activity-modifying proteins through phylogenetic and statistical analyses

Abstract

Type I and type-II functional divergences have been stated to highlight specific residues carrying out differential functions in evolutionary-divergent protein clusters from a single common ancestor. Briefly, type I analysis is based on residue constraints reflecting a gain of function just in one cluster of an entire family of proteins; while the type-II approach is based on residue constraints showing a different chemical nature in every cluster of a protein family. This last evidence is understood as differential functionality among clusters. The Receptor Activity-Modifying Proteins constitute a family characterized by its paralogous distribution in vertebrates. They are known as G-Protein Coupled Receptor modulators. Although several studies have determined their involvement in ligand binding, specificity, and enhancement of signal transduction, the responsible residues supporting those functions are unclear. Using different bioinformatic approaches, we predicted residues involved in different RAMP functional tasks. Many residues localized in an extracellular coil of RAMP proteins were predicted to be under functional divergence suggesting a gain of function in their respective proteins. Interestingly, the transmembrane region also showed important results for residues playing relevant roles where most of them showed a biased distribution on the structure. A relevant role was conferred by the enrichment of type-II residues observed in their sequences. We show a collection of residues explaining possible gain of function and differential functionality in RAMP proteins. These residues are still experimentally unexplored with regards to functionality. Finally, an evolutionary history could be discerned. Mainly, the RAMP2 cluster has evolved in a higher manner than other RAMP clusters. However, a deacceleration in the aminoacid substitution rate of RAMP2 was observed in mammals. Such effect could be caused by the co-evolution of ligands and receptors interacting with RAMP2 through evolution and/or the specialization of this cluster in GPCR modulation.

Keywords: evolutionary history; functional divergence; receptor activity-modifying proteins.

Figure 1. Phylogenetic analyses of separately RAMP clusters

Phylogenetic analyses of different RAMP clusters were firstly submitted through the Prottest application (Abascal et al. 2005) to confirm the best model for each cluster. A—The bestfit for Cluster RAMP1/4 was the Jones Thorton Taylor model plus Gamma distribution plus amino acids frequencies [JTT+G+F]. B—Best model explaining phylogeny of cluster RAMP2/5 was determined by JTT+G one. C—Phylogeny in cluster RAMP3 was resolved by JTT+I [Invarianted frequencies]+ G model. Branches reflect phylogenetic distances and bootstraps over 1000 replicates.

Figure 1. Phylogenetic analyses of separately RAMP clusters

Phylogenetic analyses of different RAMP clusters were firstly submitted through the Prottest application (Abascal et al. 2005) to confirm the best model for each cluster. A—The bestfit for Cluster RAMP1/4 was the Jones Thorton Taylor model plus Gamma distribution plus amino acids frequencies [JTT+G+F]. B—Best model explaining phylogeny of cluster RAMP2/5 was determined by JTT+G one. C—Phylogeny in cluster RAMP3 was resolved by JTT+I [Invarianted frequencies]+ G model. Branches reflect phylogenetic distances and bootstraps over 1000 replicates.

Figure 1. Phylogenetic analyses of separately RAMP clusters

Phylogenetic analyses of different RAMP clusters were firstly submitted through the Prottest application (Abascal et al. 2005) to confirm the best model for each cluster. A—The bestfit for Cluster RAMP1/4 was the Jones Thorton Taylor model plus Gamma distribution plus amino acids frequencies [JTT+G+F]. B—Best model explaining phylogeny of cluster RAMP2/5 was determined by JTT+G one. C—Phylogeny in cluster RAMP3 was resolved by JTT+I [Invarianted frequencies]+ G model. Branches reflect phylogenetic distances and bootstraps over 1000 replicates.

Figure 2. Phylogenetic tree of RAMP proteins

Full phylogenetic analysis was done through protest software for all seventy-one RAMP proteins evaluated in this study. The phylogeny model consisted of JTT+G one. Branches reflect only phylogenetic distances based on this model.

Figure 3. Multiple alignments of RAMP groups

RAMP proteins were clustered in two different groups according their vertebrate class origin being mammals or fishes. A—Thirty-two mammal RAMP sequences were aligned by T-Coffee software (Notredame et al. 1998). The sequences from Orictolagus cuniculus and Loxodonta africana were excluded because these partial sequences perturbed analyses. B—Twenty-six fish RAMP sequences were aligned as above except the RAMP protein from Tetraodon nigroviridis, which was a partial sequence and also disturbed the analyses. Amino acid numeration and secondary structure are based on previous RAMP analyses (Benitez-Paez, 2006).

Figure 4. Site-specific profile of Type-I functional divergence of RAMP proteins

Multiple alignments of each RAMP group (mammal and fish) and respective phylogenies were submitted to DIVERGE 2.0 software (Wang and Gu, 2001) and Type-I functional divergence analysis was performed. The posterior probabilities were plotted separately for each group. The dashed line shows the threshold value for significant residues under type-I functional divergence. mRAMP = RAMP proteins coming from mammals (specific sequences of some organisms belong to Boreoeutheria subclade of Placentalia Infraclass and Order Didelphimorphia of Marsupialia Infraclass). fRAMP = RAMP proteins proceeded from fishes. White bars show scores for the RAMP1/RAMP2 pairwise comparison; Grey bars show scores for the RAMP1/RAMP3 pairwise comparison; Black bars show scores for the RAMP2/RAMP3 comparison. Secondary structures are defined for each group; non homogeneous distribution was caused by no type-I information for some residues.

Figure 5. Type-0/Type-I/Type-II residue distributions on transmembrane helix

Residues under Type-0, Type-I, and Type-II functional divergences were drawn in a helical wheel plot of the human RAMP3 transmembrane helix. A—Helical wheel plot showing Type-0 [blue shaded symbols], Type-I residues [red shaded symbols], and Type-II residues [yellow shaded symbols]; orange shaded symbols represents residues with Type-I and Type-II prediction at same time. B—Molecular surface representation of transmembrane helix showing a biased distribution of type-0, type-I, and Type-II residues; color of residues is maintained according to Figure 5A.

Figure 6. Site-specific profile of Type-II functional divergence of RAMP proteins

According to procedure used in type-I analysis, the multiple alignments of each RAMP group were processed by DIVERGE 2.0 (Wang and Gu, 2001) to collect type-II functional divergence data. The posterior ratios retrieved from this analysis were plotted separately for each group. The dashed line shows the threshold value for significant residues under type-II functional divergence. mRAMP = RAMP proteins from mammals. fRAMP = RAMP proteins from fishes. White, grey, and black bars show the scores for the RAMP1/RAMP2, RAMP1/RAMP3, and RAMP2/RAMP3 pairwise comparisons respectively. Secondary structures are defined for each group; non homogeneous distribution was caused by loss of type-II information for some residues.