The phylogeography of begomoviruses: mapping informative regions in viral genomes

Abstract

Begomoviruses are notorious for their ability to infect a wide range of dicotyledonous plant species. These viruses are transmitted by a complex of cryptic whitefly species, collectively known as Bemisia tabaci. Begomovirus genomes consist of either one (monopartite) or two (bipartite) single-stranded DNA molecules, and their rapid evolution is primarily driven by mechanisms such as mutation, recombination, and pseudorecombination. The dynamic interplay among these mechanisms significantly contributes to their high genetic variability and capacity to swiftly adapt to new host species. Previous studies on microevolution of begomoviruses have consistently revealed the geographical segregation of their populations, suggesting limited gene flow across distinct regions. While numerous investigations have focused into phylogenetic and population genetic analyses of begomovirus genomes from various continents, sub-regions, or countries, there remains a noticeable research gap concerning the evidence of isolation by distance. This study aimed to investigate the evidence of isolation by distance using a multivariate Procrustean approach applied to datasets containing full-length DNA-A (or DNA-A-like) sequences of begomoviruses. Additionally, a sliding window approach was employed to perform a fine-scale mapping of the geographical signal, utilizing 200-nucleotide segments derived from the segmentation of full-length DNA-A sequences. To achieve this objective, a detailed curation of spatial data associated with each DNA-A sequence was conducted, drawing from GenBank records and related scientific publications. Subsequently, an analysis of genetic divergence among begomovirus isolates was carried out by calculating patristic distances derived from maximum likelihood trees. An extensive correlation analysis between distance matrices, encompassing both spatial and genetic information, was performed using the Procrustean Approach to Cophylogeny (PACo). The study yielded robust evidence of isolation by distance in at least three begomovirus species datasets, comprising sequences from isolates of bean golden mosaic virus, cotton leaf curl Gezira virus and tomato yellow leaf curl virus. Furthermore, the results unequivocally underscored the uneven distribution of the geographical signal across genomes. While population segregation across different geographic regions was discernible in various genomic regions, evidence of isolation by distance tended to be more pronounced in localized segments, often interspersed with regions lacking any isolation by distance signal. Additionally, this study shed light on how recombination-induced variation can obscure evidence of isolation by distance, even in datasets containing a limited number of recombinant DNA-A sequences. Finally, we concluded that recent begomovirus incursions into distant regions from their original sites of origin also contributed to the reduced global congruence between spatial and genetic data.