Análisis de los factores implicados en el salto de huésped del virus de la rabia

Abstract

Emerging diseases are the second death cause in the world. That entails a great loss of both human and animal lives as well as economic resources. A key process in the emergence of many diseases is the pathogen jump from a reservoir host to a new host, called host jump. Despite the importance of host jumps to understand disease epidemiology and evolution, very little is known on the factors that affect the success of this process. The aim of this project is to identify such ecological and genetic factors. To achieve this purpose, Rabies virus (RABV) has been used as a model as it can infect a wide range of hosts. The project is divided into three main parts: i) phylogenetic analysis of the RABV population and estimation of virus host jump frequency, ii) construction of a data base containing information about ecological and genetic factors potentially involved in host jumps, and iii) identification of the factors involved in host jumps. Firstly, 3,665 RABV nucleoprotein sequences from 48 hosts were compiled in order to reconstruct the virus phylogeny. A Maximum Likelihood tree was inferred from these nucleoprotein sequences and NST analyses were computed. The results obtained from these analyses showed that the RABV population is genetically structured according to host species. Subsequently, RABV host jumps were modelled as a stochastic diffusion process among a set of discrete states (host species) using the Bayesian Markov Chain Monte Carlo (MCMC) method. The model was divided into internal nodes and external nodes, so we could distinguish between the successful host jumps that have remained through the RABV evolutionary history (internal nodes) and the unsuccessful host jumps, or spill-over jumps, that appear only temporarily in the virus evolutionary history (external nodes). Secondly, information regarding 16 different host ecological factors from the utilized 48 host species was gathered from diverse publications and public databases. To do so, we created a tool that allowed us to efficiently access all that information by using SPARQL queries. Three host and viral genetic variables that were calculated from either the viral nucleoprotein or the hosts’ cytochrome b sequences were included. All 19 factors considered in this database had enough variability to be treated as potentially predictors of the host jump process. Collinearity analyses (Principal Component Analysis) were performed in order to detect self-correlations between factors, which resulted in a final set of 16 potential predictors. Finally, since RABV populations were genetically structured according to the host, statistical analyses (Generalized Lineal Models) were performed using the host jump and spillover frequencies as response variables and the 16 independent factors as predictor variables. These analyses identified three factors as main predictors of successful host jumps: host genetic diversity, virus genetic diversity and number of ecosystems inhabited by each host. Similar analyses yielded four factors as good predictors of spill-over jumps: host population density, overlap between the geographic distribution of the reservoir and the receptor hosts, RABV prevalence and host genetic diversity. In conclusion, the results presented in this project provide evidence that even though ecological factors are important in the host jump process, genetic factors are decisive in the success of this jump. Moreover, this project highlights the importance of studying both successful hosts jump and spill-over processes as two different but complementary processes and empathizes the importance of considering the combined effect of all factors that could be involved in these processes.