Identification of G protein-coupled receptors in Schistosoma haematobium and S. mansoni by comparative genomics

Abstract

Background: Schistosomiasis is a parasitic disease affecting ~200 million people worldwide. Schistosoma haematobium and S. mansoni are two relatively closely related schistosomes (blood flukes), and the causative agents of urogenital and hepatointestinal schistosomiasis, respectively. The availability of genomic, transcriptomic and proteomic data sets for these two schistosomes now provides unprecedented opportunities to explore their biology, host interactions and schistosomiasis at the molecular level. A particularly important group of molecules involved in a range of biological and developmental processes in schistosomes and other parasites are the G protein-coupled receptors (GPCRs). Although GPCRs have been studied in schistosomes, there has been no detailed comparison of these receptors between closely related species. Here, using a genomic-bioinformatic approach, we identified and characterised key GPCRs in S. haematobium and S. mansoni (two closely related species of schistosome).

Methods: Using a Hidden Markov Model (HMM) and Support Vector Machine (SVM)-based pipeline, we classified and sub-classified GPCRs of S. haematobium and S. mansoni, combined with phylogenetic and transcription analyses.

Results: We identified and classified classes A, B, C and F as well as an unclassified group of GPCRs encoded in the genomes of S. haematobium and S. mansoni. In addition, we characterised ligand-specific subclasses (i.e. amine, peptide, opsin and orphan) within class A (rhodopsin-like).

Conclusions: Most GPCRs shared a high degree of similarity and conservation, except for members of a particular clade (designated SmGPR), which appear to have diverged between S. haematobium and S. mansoni and might explain, to some extent, some of the underlying biological differences between these two schistosomes. The present set of annotated GPCRs provides a basis for future functional genomic studies of cellular GPCR-mediated signal transduction and a resource for future drug discovery efforts in schistosomes.

Figure 1

Summary of results for the identification and classification of GPCRs in Schistosoma haematobium and S. mansoni. Top to bottom: First, from the inferred proteomes [47,49], the numbers of sequences with transmembrane (TM) domains, the numbers of GPCRs of each class predicted by Hidden Markov Models (HMMs) and the numbers of significant matches to known* GPCRs (from databases such as SwissProt, TrEMBL and KEGG) are presented. Second, the preliminary sets of GPCRs categorised to the class level are shown, after filtering sequences that did not contain 5–8 transmembrane domains. Third, orthologs not detected in published gene sets were identified in the phylogenetic trees generated (paired one-to-one orthologs were expected for the two closely related schistosome species). Fourth, the final sets of GPCRs for each class, including the numbers of sequences found by homology but not predicted by HMMs, are shown. Finally, the numbers following sub-classification by SVM1 (class A subclass) and SVM2 (class A amine sub-classification – ligand affinity) are given.

Figure 2

Phylogenetic trees displaying the relationships of GPCRs representing subclasses of class A, and also classes B and F identified in Schistosoma haematobium and S. mansoni. In each tree, the amino acid changes per site are shown. The posterior probability (pp) of each node is indicated by small circles (pp = 0.7-0.8) or dots (pp > 0.8) at nodes. Medium to high transcription (cf. Methods section) in the adult stages of the schistosomes is identified by asterisks. The class A amine tree shows the SVM sub-classification, coloured according to ligand affinity. Asterisks indicate experimentally validated GPCRs. Relationships of the sequences representing the SmGPR clade [23] (inset a shows an enlargement). Shae and Smp are codes for sequences from S. haematobium and S. mansoni, respectively.

Figure 3

Alignment of sequences representing the SmGPR clade (cf. Figure2). Conserved, transmembrane (TM) domains are outlined in red; the most conserved amino acid residues in other sequence regions are in blue, with the least conserved residues in palest. Most sequences within this clade were predicted, using a Support Vector Machine (SVM), to bind to dopamine, with the exception of those with extended N-termini.