Single Amino Acid Mutations Affect Zika Virus Replication In Vitro and Virulence In Vivo

Abstract

The 2014-2016 Zika virus (ZIKV) epidemic in the Americas resulted in large deposits of next-generation sequencing data from clinical samples. This resource was mined to identify emerging mutations and trends in mutations as the outbreak progressed over time. Information on transmission dynamics, prevalence, and persistence of intra-host mutants, and the position of a mutation on a protein were then used to prioritize 544 reported mutations based on their ability to impact ZIKV phenotype. Using this criteria, six mutants (representing naturally occurring mutations) were generated as synthetic infectious clones using a 2015 Puerto Rican epidemic strain PRVABC59 as the parental backbone. The phenotypes of these naturally occurring variants were examined using both cell culture and murine model systems. Mutants had distinct phenotypes, including changes in replication rate, embryo death, and decreased head size. In particular, a NS2B mutant previously detected during in vivo studies in rhesus macaques was found to cause lethal infections in adult mice, abortions in pregnant females, and increased viral genome copies in both brain tissue and blood of female mice. Additionally, mutants with changes in the region of NS3 that interfaces with NS5 during replication displayed reduced replication in the blood of adult mice. This analytical pathway, integrating both bioinformatic and wet lab experiments, provides a foundation for understanding how naturally occurring single mutations affect disease outcome and can be used to predict the of severity of future ZIKV outbreaks. To determine if naturally occurring individual mutations in the Zika virus epidemic genotype affect viral virulence or replication rate in vitro or in vivo, we generated an infectious clone representing the epidemic genotype of stain Puerto Rico, 2015. Using this clone, six mutants were created by changing nucleotides in the genome to cause one to two amino acid substitutions in the encoded proteins. The six mutants we generated represent mutations that differentiated the early epidemic genotype from genotypes that were either ancestral or that occurred later in the epidemic. We assayed each mutant for changes in growth rate, and for virulence in adult mice and pregnant mice. Three of the mutants caused catastrophic embryo effects including increased embryonic death or significant decrease in head diameter. Three other mutants that had mutations in a genome region associated with replication resulted in changes in in vitro and in vivo replication rates. These results illustrate the potential impact of individual mutations in viral phenotype.

Keywords: A129 mouse model; Zika virus; infectious clone; microcephaly; mutation; quasispecies; variant.

Figure 1

Schematic representation of the Zika virus (ZIKV) genome. This representation shows viral proteins and locations of amino acid mutations chosen for in-depth examination in this study. Numbers indicate the polyprotein amino acid residue that was manipulated to generate synthetic mutants. Orange = structural proteins; blue = non-structural proteins; green = mutation locations.

Figure 2

Mutants display varying growth characteristics in Vero and C6/36 cell lines. Vero and C6/36 cell lines were infected with wild-type (WT) or indicated mutant viruses at a multiplicity of infection (MOI) of 0.01 (Vero) or 0.05 (C6/36) and harvested for FACS analysis at the indicated time points. Representative time course plots are shown for the WT virus in Vero cells (A) and C6/36 cells (B). The fold change in infected cells observed with each mutant relative to WT are shown at the 48 h time point for Vero cells (C) and at the 90 h time point for C6/36 cells (D). ** p < 0.01, **** p < 0.0001.

Figure 3

Weight loss and recovery after infection with synthetic ZIKV strains in adult mice. (A,B) Daily weight measurements in males (A) and females (B) measured as a percentage of starting weight. Mice of both sexes generally begin to regain weigh by 10 days post-infection. (C,D) Significant weight loss measured at day 7, compared to WT in males (C) and females (D). Values significantly different from WT are indicated by a bracket. p-values are indicated (Student’s t-test). Mutant 2074/86 = 2074/2086.

Figure 4

Embryo survival rates and embryo size measurements after infection with synthetic ZIKV strains. (A) total number of embryos recovered per pregnancy group. Gray—non-viable embryos, pink—viable embryos. A significant increase in number/percentage of non-viable embryos was found with mutants 123, 894, and 2074 compared to WT. (*** = p < 0.001, **** = p < 0.0001). (B) Head diameter measurements compared to WT. Significantly smaller head diameter was noted in viable embryos of mutant 894 compared to WT (*** = p < 0.001). (C) Ratio of head diameter to crown–rump length; smaller ratios indicate smaller head diameters in relation to body size. (* = p < 0.05; *** = p < 0.001).

Figure 5

Viral copies of synthetic ZIKV strains in tissues of adult nonpregnant female mice at 14 days post-infection. Differences are compared to WT. Strain difference for mutant 1404 was compared to WT at day 7 since no animals remained at day 14 when the other samples were collected. (A) ZIKV genome copies identified per µg total RNA in whole blood. (* = p < 0.05; ** = p < 0.01). (B) ZIKV genome copies identified per µg total RNA from brain tissue. (C) ZIKV genome copies identified per µg total RNA in spleen. (* = p < 0.05). (D) ZIKV genome copies identified per µg total RNA in ovary tissue. No significant differences were identified at day 14. Mutant 2074/86 = 2074/2086.

Figure 6

Viral copies of synthetic ZIKV strains in tissues of adult male mice. Differences are compared to WT. Strain difference for mutant 1404 was compared to WT at day 7 since no animals remained at day 14 when the other samples were collected. (A) ZIKV genome copies identified per µg total RNA in whole blood. (* = p < 0.05). (B) ZIKV genome copies identified per µg total RNA from brain tissue. (C) ZIKV genome copies identified per µg total RNA in spleen. (D) ZIKV genome copies identified per µg total RNA in testis. No significant differences were identified at day 14. Mutant 2074/86 = 2074/2086.

Figure 7

Tissues from pregnant mice (brain PF and spleen PF), and placentas were assayed via RT-qPCR for viral RNA. (A) ZIKV genome copies identified per μg total RNA in spleen of pregnant females. (B) ZIKV genome copies identified per μg total RNA in brain tissue of pregnant females. (C) ZIKV genome copies identified per μg total RNA in E14.5 placenta associated with embryos infected via mother at E4.5 of development. No significant difference was detected between any of the mutants and the WT virus.