In Brazil, a yellow fever (YF) outbreak was reported in areas considered YF-free for decades. The low vaccination coverage and the increasing forest fragmentation, with the wide distribution of vector mosquitoes, have been related to yellow fever virus (YFV) transmission beyond endemic areas since 2016. Aiming to elucidate the molecular and phylogenetic aspects of YFV spread on a local scale, we generated 43 new YFV genomes sampled from humans, non-human primates (NHP), and primarily, mosquitoes from highly heterogenic areas in 15 localities from Rio de Janeiro (RJ) state during the YFV 2016-2019 outbreak in southeast Brazil. Our analysis revealed that the genetic diversity and spatial distribution of the sylvatic transmission of YFV in RJ originated from at least two introductions and followed two chains of dissemination, here named the YFV RJ-I and YFV RJ-II clades. They moved with similar dispersal speeds from the north to the south of the RJ state in parallel directions, separated by the Serra do Mar Mountain chain, with YFV RJ-I invading the north coast of São Paulo state. The YFV RJ-I clade showed a more significant heterogeneity across the entire polyprotein. The YFV RJ-II clade, with only two amino acid polymorphisms, mapped at NS1 (I1086V), present only in mosquitoes at the same locality and NS4A (I2176V), shared by all YFV clade RJ-II, suggests a recent clustering of YFV isolates collected from different hosts. Our analyses strengthen the role of surveillance, genomic analyses of YVF isolated from other hosts, and environmental studies into the strategies to forecast, control, and prevent yellow fever outbreaks.
Keywords: Brazilian outbreak; amino acid polymorphisms; ecology; epidemiology; genomic surveillance; yellow fever virus.