Comparative genomics of schistosomes: identifying parasites specific genes, drug and vaccine targets

Abstract

Schistosomiasis is a chronic disease, caused by Schistosoma species, affecting 200 million people worldwide and causing at least 300,000 deaths annually. Currently, no vaccines are available and Praziquantel is the standard anti-schistosomiasis drug. Praziquantel disrupts the tegument of adult worms, but not juvenile parasites and it does not prevent reinfection. Praziquantel resistance is rare, but repeated treatment in the field and laboratory manipulation has increased parasitic resistance. Therefore, it is necessary to develop a vaccine that induces long-term immunity to schistosomiasis with the final goal of complete elimination. Driven by the need to improve disease treatment and prevention, the genomes of three human Schistosoma species have recently become publicly available (S. mansoni, S. japonicum Chinese strain and S. haematobium). The principal goal of the PhD research project is to employ machine learning and Bioinformatics methods to identify novel vaccine and drug targets against the human-infecting Schistosoma parasites from genome sequence information.In the first study, schistosome specific machine learning classifiers were developed for surface proteins and secretory peptides. Schistosome surface proteins, especially those expressed in tegument, represents the interface between host and parasite and its molecules are responsible for essential functions to parasite survival. Also, large number of proteins secreted by schistosomes are important for their survival in their hosts and infection. Knowledge of schistosome surface and secreted proteins is essential for understanding parasite host interaction and finding new candidate targets for vaccines and drugs or developing novel diagnostic methods. The web application SchistoProt has been developed, a schistosome specific classifier, for identifying schistosome specific surface proteins and secretory peptides that might be potential drug and vaccine targets.In the second study, a machine learning prediction tool is developed to predict schistosome specific immunoreactive peptides. The sequence properties of immunoreactive Schistosoma proteins have been determined and compared the significant sequence features of immunoreactive proteins and non-immunoreactive proteins of Schistosoma species. The SchistoTarget web application, for the in silico identification of Schistosoma immunoreactive proteins has been developed. SchistoTarget uses supervised machine learning methods and significant differential features distribution between immunoreactive and non-immunoreactive peptides.In the third study, a comparative analysis of the publicly available Schistosoma genomes S. mansoni, S. Japonicum, S. haematobium, the newly sequenced Schistosoma bovis genome and the non-parasitic, free-living flatworm Schmidtea mediterranea reveals the interesting candidate genes for vaccine targets. Selected genes from this study have been annotated as surface or secretory proteins using the developed web applications from previous two studies. Further, using Gene Ontology and Swiss-Prot annotations, 20 putative vaccine and drug targets have been identified to be biologically validated by wet laboratory experiments in animals and then clinically.The in silico comparative genomics analysis approach for identifying new drug and vaccine candidates represents a valuable resource for the Schistosoma research community. The protocol developed in this PhD research project can be used as a blueprint for other important parasitic diseases including malaria.