Abstract
During infection, the capsid (C) protein of many flaviviruses localizes to the nuclei and nucleoli of several infected cell lines; the underlying basis and significance of C protein nuclear localization remain poorly understood. In this study, double alanine-substitution mutations were introduced into three previously proposed nuclear-localization signals (at positions 6-9, 73-76 and 85-100) of dengue virus C protein, and four viable mutants, c(K6A,K7A), c(K73A,K74A), c(R85A,K86A) and c(R97A,R98A), were generated in a mosquito cell line in which C protein nuclear localization was rarely observed. Indirect immunofluorescence analysis revealed that, whilst C protein was present in the nuclei of PS and Vero cells throughout infection with a dengue serotype 2 parent virus, the substitution mutations in c(K73A,K74A) and c(R85A,K86A) resulted in an elimination of nuclear localization in PS cells and marked reduction in Vero cells. Mutants c(K6A,K7A) and c(R97A,R98A) exhibited reduced nuclear localization at the late period of infection in PS cells only. All four mutants displayed reduced replication in PS, Vero and C6/36 cells, but there was a lack of correlation between nuclear localization and viral growth properties. Distinct dibasic residues within dengue virus C protein, many of which were located on the solvent-exposed side of the C protein homodimer, contribute to its ability to localize to nuclei during virus infection.
