Vertical transmission of African-lineage Zika virus through the fetal membranes in a rhesus macaque (Macaca mulatta) model

Abstract

Zika virus (ZIKV) can be transmitted vertically from mother to fetus during pregnancy, resulting in a range of outcomes including severe birth defects and fetal/infant death. Potential pathways of vertical transmission in utero have been proposed but remain undefined. Identifying the timing and routes of vertical transmission of ZIKV may help us identify when interventions would be most effective. Furthermore, understanding what barriers ZIKV overcomes to effect vertical transmission may help improve models for evaluating infection by other pathogens during pregnancy. To determine the pathways of vertical transmission, we inoculated 12 pregnant rhesus macaques with an African-lineage ZIKV at gestational day 30 (term is 165 days). Eight pregnancies were surgically terminated at either seven or 14 days post-maternal infection. Maternal-fetal interface and fetal tissues and fluids were collected and evaluated for ZIKV using RT-qPCR, in situ hybridization, immunohistochemistry, and plaque assays. Four additional pregnant macaques were inoculated and terminally perfused with 4% paraformaldehyde at three, six, nine, or ten days post-maternal inoculation. For these four cases, the entire fixed pregnant uterus was evaluated with in situ hybridization for ZIKV RNA. We determined that ZIKV can reach the MFI by six days after infection and infect the fetus by ten days. Infection of the chorionic membrane and the extraembryonic coelomic fluid preceded infection of the fetus and the mesenchymal tissue of the placental villi. We did not find evidence to support a transplacental route of ZIKV vertical transmission via infection of syncytiotrophoblasts or villous cytotrophoblasts. The pattern of infection observed in the maternal-fetal interface provides evidence of paraplacental vertical ZIKV transmission through the chorionic membrane, the outer layer of the fetal membranes.

Fig 1. Experimental design.

The pathway to vertical transmission was evaluated using two cohorts of pregnant rhesus macaques. The fetectomy cohort consisted of a total of eight dams that were inoculated with ZIKV-DAK at gestational day (gd) 30. These pregnancies were surgically terminated (four at 7 days post infection (dpi) and four at 14 dpi), the conceptus was removed and a fetectomy and maternal-fetal interface (MFI) dissection was performed to collect specimens that were evaluated for ZIKV burden using multiple methods. The gross image of the fetus shows the fetus, the placenta (with the chorionic plate facing up) and the relatively thick decidua which is seen curled around at the edges of the placental discs. The perfusion cohort consisted of four dams that were inoculated with ZIKV-DAK at gd 30. At the time of pregnancy termination (either 3, 6, 9, or 10 dpi), the dams underwent terminal perfusion with 4% paraformaldehyde (PFA), and the gravid uterus was collected. The post-fixed uterus was sliced using a slicing box, creating consistent coronal sections of the entire uterus. Dashed line depicts a coronal section. These coronal sections were routinely processed and paraffin embedded using extra-large histology cassettes and viral burden in the coronal sections was evaluated using in situ hybridization. Gray boxes indicate the day of inoculation and black boxes indicate pregnancy termination.

Fig 2. Summary of ZIKV viral burden at 7 dpi.

(A) Artistic representation of ZIKV infection in the maternal-fetal interface at 7 days post infection (dpi). Pink tissue/structure names denote that ZIKV RNA was detected via ISH or RT-qPCR. The number of cases where the tissue had ZIKV RNA detected out of the number of cases in which tissue was evaluated is indicated below each tissue name. An artistic representation of the infection of each tissue is illustrated in pink in the figure. (B) Summary of the ZIKV RNA detected via RT-qPCR. The level of ZIKV burden is summarized as high (>1000 ZIKV RNA copies/mg of tissue), medium (>100 ZIKV RNA copies/mg), low (<100 copies/mg of tissue), no virus detected (below the limit of detection), or NA, sample not available. The theoretical limit of detection for all the tissues is 3 copies/mg. C) Summary of ZIKV RNA detected via ISH. ZIKV RNA was noted as present (+) or absent (-) in the indicated tissues/structures. In some instances, adequate histological samples were not obtained, and could not be properly evaluated (/). (B) (C) Individual animal IDs are listed on the bottom. Please see subsequent figures for precise cellular localization of vRNA.

Fig 3. ZIKV infection in the decidua, trophoblastic shell, and chorionic membrane at 7 dpi.

ZIKV infection was seen in (A)–(C) a decidual vessel, (D)–(F) decidual macrophages, (G)(H) in the trophoblastic shell, and (I–K) chorionic membrane. (A) H&E staining of the section shown in (B) Pink staining shows ZIKV RNA detected via ISH (C) IHC staining for cytokeratin. (D) H&E staining of the corresponding section shown in (E). (E) ISH (pink) staining showing ZIKV RNA in the decidua. (F) IF staining for CD163 (red) to identify macrophages, ZIKV (green), and DAPI nuclear staining (blue). Colocalization of CD163, ZIKV, and DAPI is outlined by the dashed circles. The scale bar represents 200 μm. (G) Pink staining shows ZIKV RNA detected via ISH (H) IHC staining for cytokeratin, highlights the trophoblastic shell. (I) H&E staining of the section shown in (J). (J) Pink staining showing ZIKV RNA in the chorionic membrane. (K) IHC staining for cytokeratin identifies the trophoblast layer of the chorionic membrane. (A)–(E), (G)–(K) Scale bars represent 500 μm.

Fig 4. Infectious ZIKV in extraembryonic coelomic fluid at 7 dpi.

Infectious virus detected by plaque assay. The LOD of 50 plaque-forming units (PFUs) is represented by a dashed line.

Fig 5. Summary of ZIKV viral burden at 14 dpi.

(A) Artistic representation of ZIKV infection in the maternal-fetal interface at 14 days post infection (dpi). Pink tissue/structure names denote that ZIKV RNA was detected via ISH or RT-qPCR. The number of cases where the tissue had ZIKV RNA detected out of the number of cases in which tissue was evaluated is indicated below each tissue name. An artistic representation of the infection of each tissue is illustrated in pink in the figure. (B) Summary of the ZIKV RNA detected via RT-qPCR. The level of ZIKV burden is summarized as high (>1000 ZIKV RNA copies/mg of tissue, medium (>100 ZIKV RNA copies/mg or mL), low (<100 copies/mg of tissue), below the limit of detection, or NA, sample not available. The theoretical limit of detection for all the tissues is 3 copies/mg. (D) Summary of ZIKV RNA detected via ISH. ZIKV RNA was noted as present (+) or absent (-) in the indicated tissues/structures. In some instances, adequate histological samples were not obtained, and could not be properly evaluated (/). (B) (C) Individual animal IDs are listed on the bottom.

Fig 6. ZIKV infection in the chorionic membrane and extraembryonic coelomic and amniotic fluid at 14 dpi.

(A)(C)(E)(G) Images from 14–3 chorionic membrane. (B)(D)(F)(H) Images from 14–4 chorionic membrane. (A)(B) Pink staining showing ZIKV detected via ISH predominantly within the mesenchymal layer of the chorionic membrane. (C) Corresponding H&E stain of (A). (D) Corresponding H&E stain of (B). The mesenchymal layer is indicated by m and the trophoblast layer is indicated by t. (E)(F) IF staining for cytokeratin (red) and ZIKV (green), and DAPI nuclear stain (blue). (F) The dashed circle indicates the area of colocalization showing ZIKV infection in the trophoblasts of the chorionic membrane. (G)(H) IF staining with CD163 (red), ZIKV (green), and DAPI nuclear staining (blue). The dashed circles indicate areas of colocalization showing ZIKV infection in macrophages. (A)–(D)The scale bar represents 100 μm, (E)–(H) 200 μm. (I) Infectious virus detected by plaque assay. The LOD of 50 plaque-forming units (PFUs) is represented by the dashed line.

Fig 7. Infection in the placental villi at 14 dpi.

(A) Pink staining shows ZIKV detected via ISH. (C) corresponding H&E section. The red square in (A) identifies the area magnified in (B) as does the black square in (C) for (D). (A)(C) The scale bars represent 500 μm. (B)(D) The scale bars represent 100 μm. (E) IF staining for CD163 (red) ZIKV (green) and DAPI (blue). The scale bar represents 400 μm. The orange square shows the portion that is magnified in (F). (G) IF staining for cytokeratin (red) ZIKV (green) and DAPI (blue). (E)(G)The scale bars represent 400 μm. (H) Multiplex fluorescence in situ hybridization (mFISH) to detect genomic, positive sense ZIKV RNA (green) and replicative intermediate, negative sense RNA (red) with nuclear DAPI staining (blue). The scale bar represents 50 μm. The replicative intermediate negative sense RNA (red) is shown alone in (I).

Fig 8. ZIKV burden in the fetus at 14 dpi.

(A) ZIKV RNA detected in fetal tissues and organs. The limit of detection of 3 copies/mg is depicted by the dashed line. (B) ZIKV RNA detected in fetal fluids. The limit of detection of 150 copies/mL is depicted by the dashed line. CSF = cerebrospinal fluid. (C)–(E) ZIKV RNA detected via ISH (pink). (C)(D) images from 14–3 and (E)(F) from 14–4. (C)–(F) Scale bars represent 1000 μm. (C) Infection detected in the eye of the fetus with ZIKV (+) in the material (vitreous/aqueous humor) surrounding the lens (L) and within the sclera of the eye indicated by the arrow. ZIKV was also detected subjacent to the nasal mucosa of the fetal nasal cavity (nc). (D) ZIKV RNA detected in the kidneys (k), adrenal glands (a), and intestinal muscular layers (m) with minimal scattered positivity in the spinal cord (sc). (E) Abundant ZIKV was detected throughout the myocardium of the right (rt) and left (l) ventricles of the heart and (F) in the pulmonary interstitium of the lungs with sparing of the respiratory epithelium of the bronchi and bronchioles. Scale bar represents 100 μm.

Fig 9. ZIKV infection in large decidual vessels at 6 dpi.

(A) Photomicrograph of a representative coronal section of the gravid uterus evaluated via ISH for ZIKV RNA (pink). Pink tissue/structure names indicate the detection of ZIKV RNA via ISH. (B) Magnification of ZIKV RNA detected in a decidual vessel. (C) H&E staining of a serial section. (D) IF staining of the same decidual vessel for cytokeratin (red), ZIKV (green), and DAPI for nuclear staining (blue). (E) Multiplex fluorescence in situ hybridization (mFISH) to detect genomic, positive sense ZIKV RNA (green) and replicative intermediate negative sense RNA (red) with nuclear DAPI staining (blue). Colocalization of genomic RNA and replicating RNA (yellow) showing that the ZIKV RNA detected in the EVTs represents replicating virus. (B)–(E) The scale bars represent 200 μm.

Fig 10. ZIKV infection of chorionic membrane and peripheral margin of the placenta at 9 dpi.

(A) Photomicrograph of a representative coronal section of gravid uterus evaluated via ISH for ZIKV RNA (pink). Pink tissue/structure names indicate the detection of ZIKV RNA via ISH. (B) Magnification of ISH staining showing ZIKV RNA in the trophoblasts of the chorionic membrane and in the peripheral margin of the placenta. * indicates the placenta. (C) Sparse infection of the decidua. (B)(C) The scale bars represent 100 μm.

Fig 11. ZIKV infection at 10 dpi shows extensive infection of the chorionic membrane.

(A) A Representative coronal section of gravid uterus evaluated via ISH for ZIKV RNA (pink). Pink tissue/structure names indicate the detection of ZIKV RNA via ISH. The fetal head has been added to this image from a different coronal section. The yellow and red squares outline the areas magnified in (B) and (C), respectively. (B)(C) IF staining for cytokeratin (red) and ZIKV (green) and nuclear DAPI (blue). Colocalization of green and red shows that ZIKV has infected the trophoblasts of the chorionic membrane. (D) H&E stained section corresponding to section shown in (E) depicting ZIKV RNA detected via ISH (pink) in the trophoblast layer (t) of the chorionic membrane. The mesenchymal layer is indicated by m. (F) ZIKV RNA detected via ISH in the peripheral margin of the placental disc. (G) H&E stained section. (H) Full slide evaluated via ISH the black square outlines the portion magnified in (E) and the orange square outlines the portion magnified in (F). The scale bar represents 400 μm in (B), 200 μm in (C)–(G), and 2350 μm in (H).

Fig 12. Mesenchymal tissue of the chorionic plate and villi infected with ZIKV at 10 dpi.

(A) full slide evaluated with ISH shown as pink staining. The orange square outlines the area of the slide magnified in portion (B). (B) CP indicates chorionic plate. The black square outlines the portion of the slide magnified in (C–E). (D) H&E of the corresponding section is shown in (C). (C)(D) Arrows indicate endothelial cells, m indicates mesenchymal tissue and t indicates trophoblasts in the placenta. (E) Corresponding section with IHC staining for CD163. (F)(G) IF staining for CD163 (red) ZIKV (green) and DAPI nuclear staining (blue). (F) Photomicrograph of the placental villi with colocalization of green and red showing that ZIKV is within Hofbauer cells in the placenta villi, but not exclusively. (G) Depicts the chorionic plate of the placenta where ZIKV almost exclusively co-localized with CD163 showing ZIKV predominantly within macrophages in the chorionic plate. Scale bars represent 200 μm in (B)–(F) and 400 μm in (G).

Fig 13. ZIKV infection in the yolk sac.

(A) ISH staining for ZIKV shown as pink staining. Red square indicates the portion of the section magnified in (C). (B) Corresponding H&E stained section. Black square indicates the portion of the section magnified in (D). (C)(D) Structures of the yolk sac outlined. All scale bars represent 500 μm.

Fig 14. ZIKV RNA detected at 10 dpi via ISH.

(A) ISH staining for ZIKV (pink) in the endothelial lining of the umbilical vessels (v) and multifocally scattered throughout Wharton’s jelly surrounding the affected vessels of the umbilical cord. (B) Corresponding H&E stained section. (C) Full slide evaluated for ZIKV RNA via ISH. The black square outlines the area magnified in (A) and the red square is magnified in (E). (D) The corresponding H&E stained section shown in (C). The area outlined in blue is magnified in (B) and the area outlined in purple is magnified in (F). (E) ZIKV RNA detected in the body of the fetus via ISH (pink). There is marked diffuse ZIKV RNA (pink) in the majority of tissues with less intense and single cell positivity in the spinal cord (SC) and liver (L). ZIKV RNA was not detected in the cartilage and immature bone of the skeleton of the fetus. (F) The corresponding H&E stained section of the sample shown in (E). (G) Full slide evaluated for ZIKV RNA via ISH. The pink square outlines the portion of the slide magnified in (H). (H) The head of the fetus with ZIKV RNA identified via ISH (pink) in the sclera of the eye (arrow), brain (B), and tissues of head. ZIKV RNA was not found in the lens (L) of the eye. (I) (J) Corresponding H&E stained section of specimens shown in (G) and (I), respectively. (A)(B)(H)(J) Scale bars represent 500 μm. (C)(D)(G)(I) The scale bars represent 5000 μm. (E)(F) The scale bars represent 1000 μm.