Codon usage indicates that amphibians, reptiles and birds are major hosts for Zika and other arboviruses: implications for epidemiology and surveillance

Codon usage studies estimate the efficiency of viral replication in putative hosts and vectors for which there has been a supposed history of adaptation. This can help reconstructing the global epidemiology of arthropod-borne viruses (arboviruses), clarifying aspects regarding their ecology and evolution. Current studies concentrate on primate hosts, promoting a lack of information about alternative reservoirs. We studied if and how the codon usage of a set of 13 arboviruses comprising Zika, chikungunya, dengue, and yellow fever is adapted to a range of 30 different putative animal hosts. Our results indicate that all considered arboviruses are more adapted to amphibians, reptiles and birds than to humans and other mammals. This outcome is consistent over different indexes and correlates with the same analyses in West Nile and Japanese encephalitis viruses, two arboviruses known for their bird-to-bird transmission cycle. This is an unexpected finding indicating that non-mammals are the likely primary hosts of most arboviruses. Thus, we advocate that wild environments may have a stronger influence than previously thought in the reservoiring and spreading of Zika and other arboviruses. This has profound implications for the actual surveillance and management of outbreaks.