Zika viruses of African and Asian lineages cause fetal harm in a mouse model of vertical transmission

Abstract

Congenital Zika virus (ZIKV) infection was first linked to birth defects during the American outbreak in 2015/2016. It has been proposed that mutations unique to the Asian/American-genotype explain, at least in part, the ability of Asian/American ZIKV to cause congenital Zika syndrome (CZS). Recent studies identified mutations in ZIKV infecting humans that arose coincident with the outbreak in French Polynesia and were stably maintained during subsequent spread to the Americas. Here we show that African ZIKV can infect and harm fetuses and that the S139N substitution that has been associated with the American outbreak is not essential for fetal harm. Our findings, in a vertical transmission mouse model, suggest that ZIKV will remain a threat to pregnant women for the foreseeable future, including in Africa, Southeast Asia, and the Americas. Additional research is needed to better understand the risks associated with ZIKV infection during pregnancy, both in areas where the virus is newly endemic and where it has been circulating for decades.

Fig 1. In vitro and in vivo characterization of ZIKV strains.

(a) In vitro growth kinetics of ZIKV-PR-IC and mutant ZIKV-PR-N139S on Vero cells. Data points represent means of three replicates at each time point ± standard deviation. Cells were inoculated at an MOI of 0.01 PFU/cell. Titer was measured (PFU/ml) by plaque assay. Growth curves were not significantly different. (b) Survival curves of neonatal BALB/c mice intracranially inoculated with 10 PFU of different strains of ZIKV. PBS: n = 18; ZIKV-CAM: n = 36; ZIKV-PR-IC: n = 24; ZIKV-PR-N139S: n = 30; ZIKV-DAK: n = 20. All strains caused significant mortality by 28 dpi when compared to PBS (Fisher’s exact test). As compared to PBS controls: ****p < 0.001; ** p <0.002; ns, not significant. (c) Weight in grams of surviving intracranially inoculated pups at 28 days post infection. ****p < 0.0001; ns, not significant (student’s t-test). (d) Time-mated Ifnar1-/- dams were inoculated with 10³ PFU of ZIKV on E7.5 and maternal infection was confirmed by plaque assay on day 2 post inoculation. ***p < 0.002; **** p < 0.0001; ns, not significant (one-way ANOVA).

Fig 2. Fetal outcomes after maternal infection with ZIKV strains.

(a) Rate of grossly normal (black) versus abnormal (red) fetuses at E14.5 after maternal infection at E7.5. An abnormal fetus was defined as resorption-prone. Data presented are for individual fetuses from 5–6 litters per treatment group. The n for each group is indicated above each bar. ****p<0.0001; ns, not significant (Fisher’s exact test). (b) Fetus size as assessed by crown-rump length (CRL) in mm using ImageJ software. CRL was only measured for fetuses determined to be grossly normal at E14.5. ****p<0.0001; ns, not significant (unpaired Student’s t-test). (c) Representative images of fetuses on E14.5 from each treatment group. Scale bar, 2 mm. PBS characterized as normal. ZIKV-PR-IC, ZIKV-PR-N139S, ZIKV-DAK characterized as abnormal. (d-f) Viral burdens were measured by qRT-PCR assay from individual homogenized placentas (d), fetuses (e), and concepti (when the fetus and placenta could not be separated due to severe resorption). (f) Symbols represent individual placenta, fetus, or conceptus from 3–5 independent experiments for each treatment group. Bars represent the mean viral burden of each treatment group. *p<0.05; ns, not significant (one-way ANOVA).

Fig 3. Placenta histopathology analysis: Hematoxylin and eosin (H&E) staining of placenta and fetus.

(a-c) Normal histologic features of each placental zone (decidual layer (D), labyrinth layer (L), and junctional zone (JZ)) from concepti from dams inoculated with PBS. BV, normal decidual blood vessels. (d-f) Severe histopathologic injury patterns for each zone from placenta from ZIKV-inoculated dams. (d) Myometrium (M) and decidua (D) from a ZIKV-PR-IC placenta with increased inflammation, multiple thrombi (T) in the decidua, and a necrotic JZ. (e) L from a ZIKV-DAK placenta with focal necrosis (N), lack of blood in most vascular spaces, and numerous degenerating cells. (f) JZ from a ZIKV-DAK placenta with markedly dilated blood vessels, focal thrombi, and a layer of necrosis at the interface with the decidua. (g) D and M from a ZIKV-PR-N139S placenta with inflammation (I). (h) L from a ZIKV-PR-N139S placenta with T and infarction (IF). (i) JZ from a ZIKV-PR-IC with N and I. (j-l) The degree of placental pathology was rated on a scale of 0–4: zero represents normal histologic features and 4 represents the most severe features observed. Each zone of the placenta was scored individually for general overall pathology, amount of inflammation, and amount of vascular injury with a consensus score for each placenta derived from three independent pathologists. Only ‘General’ scores are shown because they were representative of the ‘inflammation’ and ‘vascular injury’ categories and do not differ significantly from ‘general’. Error bars represent 95% confidence interval from the median. Data are representative of 3–5 independent experiments for each treatment group. Scale bar, 50 μm.