Modeling cell infection via virus-producing cells rather than free infectious virus significantly improves fits of an in vitro viral kinetic data

Chikungunya and Zika viruses are arthropod-borne viruses that pose significant threat to public health. Experimental data show that during in vitro infection both viruses exhibit qualitatively distinct replication cycle kinetics. Chikungunya viral load rapidly accumulates within the first several hours post infection whereas Zika virus begins to increase at much later times. We sought to characterize these qualitatively distinct in vitro kinetics of chikungunya and Zika viruses by fitting a family of mathematical models to time course viral load datasets. We demonstrate that the standard viral kinetic model, which considers that new infections result only from free virus penetrating susceptible cells, does not fit experimental data as well as a model in which the number of virus-infected cells is the primary determinant of infection rate. We provide biologically meaningful quantifications of the main viral kinetic parameters and show that our results support cell-to-cell or localized transmission as a significant contributor to viral infection with chikungunya and Zika viruses.