Highly efficient maternal-fetal Zika virus transmission in pregnant rhesus macaques

Abstract

Infection with Zika virus (ZIKV) is associated with human congenital fetal anomalies. To model fetal outcomes in nonhuman primates, we administered Asian-lineage ZIKV subcutaneously to four pregnant rhesus macaques. While non-pregnant animals in a previous study contemporary with the current report clear viremia within 10-12 days, maternal viremia was prolonged in 3 of 4 pregnancies. Fetal head growth velocity in the last month of gestation determined by ultrasound assessment of head circumference was decreased in comparison with biparietal diameter and femur length within each fetus, both within normal range. ZIKV RNA was detected in tissues from all four fetuses at term cesarean section. In all pregnancies, neutrophilic infiltration was present at the maternal-fetal interface (decidua, placenta, fetal membranes), in various fetal tissues, and in fetal retina, choroid, and optic nerve (first trimester infection only). Consistent vertical transmission in this primate model may provide a platform to assess risk factors and test therapeutic interventions for interruption of fetal infection. The results may also suggest that maternal-fetal ZIKV transmission in human pregnancy may be more frequent than currently appreciated.

Fig 1. Study layout and viral RNA burden in pregnant rhesus fluids.

(A) Schematic representation of the timeline of infection, sampling for maternal viral burden, and experimental cesarean section, for all animals in the study. Animals received a ZIKV challenge in the first or late second/early third trimesters of pregnancy, and blood and other fluid samples were collected according to the schedule indicated in detail in S1 Fig. (B) ZIKV viral load in pregnant macaque fluids. Viral RNA loads (vRNA copies/ml) measured in plasma, urine, saliva, and amniotic fluid presented individually for the four pregnant animals. The day post-inoculation is indicated below each graph, and gestational age (days) for each animal is indicated above (term = 165±10 days). Limit of assay quantification is 100 copies/mL. Limit of detection is 33 copies/mL. Colors for individual animals are continued through the rest of the Figures, including the Supplementary Figures.

Fig 2. Maternal viral control and immune responses to ZIKV inoculation.

(A) Peripheral blood plasma viremia in pregnant macaques infected with ZIKV. Results are shown for animals infected at 38 days gestation (animal 827577, dark blue), 31 days gestation (animal 660875, light blue), 103 days gestation (animal 357676, red) or 118 days gestation (animal 598248, yellow). The day of gestation is estimated +/- 2 days. Grey tracings represent viremia in nonpregnant/male rhesus monkeys infected with the identical dose and strain of ZIKV in a previous study [28]. The horizontal line indicates the quantitative limit of detection. (B) Peripheral blood cell response to infection. Absolute numbers of Ki67+ NK cells (left) or CD8+TEM cells (right) are presented as a percentage relative to baseline set at 100% (dashed line), with first trimester and third trimester animals represented in the same colors as presented in Fig 1A. (C) Plasmablast expansion over time from each pregnant animal. The plasmablast expansions of two nonpregnant animals from Dudley et al [28] are shown as grey lines. (D) Neutralization by ZIKV immune sera from pregnant and nonpregnant ZIKV-infected macaques. Immune sera from macaques infected with ZIKV in either the first trimester (dark or light blue), third trimester (red or yellow), or nonpregnant contemporary controls (gray) from Dudley et al [28] were tested for their capacity to neutralize ZIKV-FP. Infection was measured by plaque reduction neutralization test (PRNT) and is expressed relative to the infectivity of ZIKV-FP in the absence of serum. The concentration of sera indicated on the x-axis is expressed as log₁₀ (dilution factor of serum). The EC90 and EC50, estimated by non-linear regression analysis, are also indicated by a dashed line. Neutralization curves for each animal at 28 dpi are shown.

Fig 3. Complete blood counts (CBCs) and serum chemistries for pregnant macaques infected with ZIKV.

Animals were infected with 10⁴ PFU of ZIKV. Animals infected in the first or third trimesters are represented by color coding (A) as presented in Fig 1. All animals had CBC analysis performed on EDTA blood and chemistry analysis performed on serum at -7, -3, 0, 1–10 and additional indicated dpi. B. AST blood chemistries, C. ALT serum chemistries, D. CK serum chemistries, E. WBC counts, F. % lymphocytes, G. red blood cell (RBC) counts.

Fig 4. Fetal growth following ZIKV infection.

Growth curves of femur length (FL), biparietal diameter (BPD), and head circumference (HC) obtained from fetal ultrasound images throughout gestation are presented as individual lines or symbols with specific colors as in Fig 1. (A) FL, (B) BPD (C) and HC were determined for the fetuses in this study and plotted against data from Tarantal [35], which is presented as the mean (solid black line) and 1, 2, and 3 standard deviations from the mean as grey lines above and below the mean. The data from the last month of pregnancy are also presented as a magnified view of the scatter of individual data points on the right. Representative ultrasound images of FL, BPD, and HC are also shown at the right.

Fig 5. Fetal growth as assessed by predicted gestational ages.

The predicted gestational age (pGA) as described by Tarantal [35] from each of the pregnancies is plotted against the actual day of gestation estimated from breeding activity and animal menstrual records. The pGA was derived from the average of BPD+FL (dashed lines), or the HC (solid lines). A (animal 827577) and B (animal 660875), first trimester infection. C (animal 357676) and D (animal 598248), late second/early third trimester infection.

Fig 6

Charts summarizing (A) ZIKV RNA copy numbers, and (B) histologic evaluation and semiquantitative scoring of all normal and lesioned tissues, presenting all maternal and fetal tissues analyzed. Keys for ZIKV RNA copy number burden per mg of tissue, and description of histopathology scores (“Normal” to “Severe”) appear at the left. Animal numbers are color coded as introduced in Fig 1.

Fig 7. Maternal and fetal histopathology analyses: hematoxylin and eosin (H&E) staining of selected tissues.

Maternal spleen, 660875: increased neutrophils (arrows) throughout splenic sinusoids. Maternal decidua, 827577: multifocal stromal, intravascular, and perivascular inflammation (arrows). Placenta, 660875: moderate multifocal necrosis and loss of trophoblastic epithelium (+) with viable and degenerative neutrophils (arrows) between villi (*) and throughout the intervillous space. Amniotic membrane, 598248: scattered neutrophils (arrows) within the amniotic basement membrane and underlying perivascular stroma. Umbilical cord, 660875: segmental thrombosis (*) with entrapped neutrophils (arrows). Fetal lung, 660875: fetal squamous cells (*) and neutrophils (arrows) admixed with fibrin within alveolar spaces. Scale bar = 50μm.

Fig 8. H&E Staining of fetal tissues of the visual system.

Panels 1A-C, animal 660875. Panel 1A: Mild infiltration of lymphocytes in the bulbar conjunctival substantia propria (arrow). Panel 1B: Moderate neutrophilic infiltration in the ciliary body stroma (arrows). Panel 1C: Moderate gliosis of the laminar and post-laminar optic nerve characterized by overall hypercellularity of the neuropil especially as indicated by asterisks. Panels 2A-C, animal 827577. Panel 2A: Minimal infiltration of lymphocytes in the bulbar conjunctival substantia propria (arrows). Panel 2B: Normal ciliary body stroma (arrows). Panel 2C: Moderate gliosis of the laminar and post-laminar optic nerve characterized by overall hypercellularity of the neuropil especially as indicated by asterisks. Panels 3A-C, animal 357676, and Panels 4A-C, animal 598248. Panels 3A and 4A: Normal bulbar conjunctival substantia propria. Panels 3B and 4B: Normal ciliary body stroma. Panels 3C and 4C: Normal optic nerve. Scale bar = 200μm.

Fig 9. Immunohistochemical localization of ZIKV in fetal [and maternal] tissues.

(A) Immunofluorescent staining for ZIKV NS2B (red) and macrophage marker CD163 (green) in fetal axillary lymph node with a high vRNA burden. The white scale bar = 100 μm. (B) H&E stained near section of the tissue presented in 9A. (C) Nonspecific immunostaining with control isotypes for ZIKV NS2B and CD163.