Early Embryonic Loss Following Intravaginal Zika Virus Challenge in Rhesus Macaques

Abstract

Zika virus (ZIKV) is an arthropod-borne virus (arbovirus) and is primarily transmitted by Aedes species mosquitoes; however, ZIKV can also be sexually transmitted. During the initial epidemic and in places where ZIKV is now considered endemic, it is difficult to disentangle the risks and contributions of sexual versus vector-borne transmission to adverse pregnancy outcomes. To examine the potential impact of sexual transmission of ZIKV on pregnancy outcome, we challenged three rhesus macaques (Macaca mulatta) three times intravaginally with 1 x 10⁷ PFU of a low passage, African lineage ZIKV isolate (ZIKV-DAK) in the first trimester (~30 days gestational age). Samples were collected from all animals initially on days 3 through 10 post challenge, followed by twice, and then once weekly sample collection; ultrasound examinations were performed every 3-4 days then weekly as pregnancies progressed. All three dams had ZIKV RNA detectable in plasma on day 3 post-ZIKV challenge. At approximately 45 days gestation (17-18 days post-challenge), two of the three dams were found with nonviable embryos by ultrasound. Viral RNA was detected in recovered tissues and at the maternal-fetal interface (MFI) in both cases. The remaining viable pregnancy proceeded to near term (~155 days gestational age) and ZIKV RNA was detected at the MFI but not in fetal tissues. These results suggest that sexual transmission of ZIKV may represent an underappreciated risk of pregnancy loss during early gestation.

Keywords: Zika virus; embryonic loss; intravaginal infection; macaques; pregnancy.

Figure 1

Study design. Three female rhesus macaques were time-mated, confirmed pregnant by ultrasound, and challenged intravaginally at 30 days (+/-2) gestational age with 1x10⁷ PFU ZIKV-DAK three times at two-hour intervals. Blood collection* denotes plasma and PBMC isolation at every sampling time point while serum collection was planned only on days 0, 14, 27, and 122 post-ZIKV challenge. Ultrasound^# denotes ultrasound imaging was performed every 3-4 days during early gestation, then weekly thereafter. Hysterotomies were originally planned for each animal at approximately 122 days post-ZIKV challenge.

Figure 2

Intravaginal ZIKV challenge resulted in detection of vRNA in plasma for all three dams. The x-axis shows days post-ZIKV challenge. The y-axis starts at the estimated limit of quantification of the qRT-PCR assay (1x10² copies/ml) and is shown as copies/ml plasma on the log scale. Plasma vRNA loads are displayed for dam 049-101 as orange triangles, for dam 049-102 as blue circles, and for dam 049-103 as magenta squares. For comparison, ZIKV plasma vRNA loads are also shown for three pregnant macaques subcutaneously (SC) inoculated with 1x10⁴ PFU ZIKV-DAK and are displayed as gray dashed lines and noted as 10⁴ PFU SC in the legend (47).

Figure 3

Pregnancy outcomes and maternal-fetal interface (MFI) and fetal tissue vRNA loads. (A) Pregnancy outcomes for three dams intravaginally inoculated 3x with ZIKV at approximately 30 days gestation. Two dams (049-102 and 049-103) were determined by ultrasound to have nonviable embryos at approximately 17-18 days post-ZIKV challenge. Hysterotomies and embryo and MFI tissue collections were performed 3 days after detection. Dam 049-101’s pregnancy continued until scheduled hysterotomy and extensive tissue collection at approximately 155 days gestational age. (B) Average ZIKV vRNA loads for positive embryo and MFI tissues collected at hysterotomy from dams 049-102 (blue) and 049-103 (magenta) following embryonic death at approximately 17-18 days post-ZIKV challenge. The dashed line represents the average of the estimated limit of detection (50-100 copies/mg, average: 75 copies/mg tissue) for the qRT-PCR assay. (C) Average ZIKV vRNA loads for positive MFI tissues collected at hysterotomy from dam 049-101 (orange) at approximately 122 days post-ZIKV infection. Fetal tissues were negative for ZIKV RNA by qRT-PCR. The dashed line represents the average of the estimated limit of detection (50-100 copies/mg, average: 75 copies/mg tissue) for the qRT-PCR assay.

Figure 4

ZIKV genomic RNA was detected by in situ hybridization (ISH) in placental tissues collected from dams 049-102 and 049-103, but not from dam 049-101. For all images, red coloration is indicative of positive staining for ZIKV genomic RNA. Overall, positive staining is focal but visible in multiple areas. Insets show high magnification of the areas denoted by the black arrows in each larger panel. Representative images are shown of (A) primary placental disk from 049-103, (B) secondary placental disk from 049-103, (C) primary placental disk from 049-102, and (D) secondary placental disk from 049-102. Scale bar = 1000 micrometers.

Figure 5

All three dams developed neutralizing antibodies (nAbs) against ZIKV as detected by PRNT₉₀ following intravaginal ZIKV infection. The x-axis is the log10 reciprocal serum dilution and the y-axis is the percent plaque reduction for ZIKV-DAK. Day 0 for all animals is shown as symbols with 049-101 represented by orange triangles, 049-102 represented by blue circles, and 049-103 represented by magenta squares. Dashed gray horizontal lines indicate the PRNT₉₀ and PRNT₅₀ cut-offs, respectively. Neutralization curves were generated using non-linear regression to estimate the dilution of serum required to inhibit 90% of Vero cell culture infection. Neutralization curves are shown for days 14 (dashed lines) and 24 (dotted lines) for dams 049-102 (blue) and 049-103 (magenta), and for days 14, 27, and 122 (dashed and dotted line) for dam 049-101 (orange).