Zika virus infection in pregnant rhesus macaques causes placental dysfunction and immunopathology

Abstract

Zika virus (ZIKV) infection during pregnancy leads to an increased risk of fetal growth restriction and fetal central nervous system malformations, which are outcomes broadly referred to as the Congenital Zika Syndrome (CZS). Here we infect pregnant rhesus macaques and investigate the impact of persistent ZIKV infection on uteroplacental pathology, blood flow, and fetal growth and development. Despite seemingly normal fetal growth and persistent fetal-placenta-maternal infection, advanced non-invasive in vivo imaging studies reveal dramatic effects on placental oxygen reserve accompanied by significantly decreased oxygen permeability of the placental villi. The observation of abnormal oxygen transport within the placenta appears to be a consequence of uterine vasculitis and placental villous damage in ZIKV cases. In addition, we demonstrate a robust maternal-placental-fetal inflammatory response following ZIKV infection. This animal model reveals a potential relationship between ZIKV infection and uteroplacental pathology that appears to affect oxygen delivery to the fetus during development.

Fig. 1

Experimental design and ZIKV infection relative to fetal brain development. The pregnant dams were infected with ZIKV at 31, 51, 114, and 115 dGA. For all animals, C-section delivery was performed at 135dGA and maternal and fetal tissues were collected. Maternal peripheral blood, urine, and saliva samples (red asterisks) were obtained at the times indicated. Ultrasound analyses of the fetus and placenta were performed at the times indicated with gray asterisks

Fig. 2

ZIKV infection alters decidua and villous placenta monocyte and CD4+ T-cell populations and activation status. Decidua and villous placental lymphocytes were isolated from ZIKV-infected and control RM and analyzed by flow cytometry for the presence of the cellular activation marker CD169 on a dendritic cells; b monocytes; c monocyte subsets; and d NK cells, or the marker for Ki67 proliferation marker on e CD4+ central memory T cells and f CD4+ effector memory T cells. Cells were stained with a panel of antibodies recognizing HLA-DR, CD14, CD11c, CD123, CD20, CD3, CD16, and CD169 or a secondary panel containing CD3, CD4, CD8β, CD95, CD28, and Ki67. One-way ANOVA was performed with Tukey’s multiple comparison test, bars represent mean and SEM, ***p < 0.001, **p < 0.01, *p < 0.05

Fig. 3

Cytokine and chemokine profiling. A 29-plex-cytokine magnetic bead assay was performed on a plasma isolated from ZIKV-infected and control fetal and cord blood, b amniotic fluid; and c fetal and maternal cerebral spinal fluid at 135dGA. Plasma analysis revealed changes in the cytokines IL-RA, IL-1β, IL-2, IL-12, IFNγ, and MIF; and chemokines MCP-1 (CCL-2), MDC (CCL-22), IP-10 (CXCL-10), and ITAC (CXCL-11). Compared to controls increased levels of IL-6, Eotaxin (CCL-11), MCP-1 (CCL-2), MIF and FGF-basic were detected in amniotic fluid from infected animals. Levels of IL-1RA, Eotaxin (CCL-11), MCP-1 (CCL-2), MIF, and IP-10 (CXCL-10) were higher in fetal vs. maternal CSF. One-way ANOVA was performed with Tukey’s multiple comparison test, bars represent mean and SEM, **p < 0.01, *p < 0.05 UV, umbilical vein; UC, umbilical cord

Fig. 4

Ultrasound and MR imaging of utero-feto-placental blood flow. a Uterine artery pulsatility index, b uterine artery blood flow volume (cQUtA) corrected for maternal body weight, c umbilical artery pulsatility index, and d umbilical vein blood flow volume (cQUV) corrected for fetal abdominal circumference measured by Doppler ultrasound in the ZIKV-exposed animals and a cohort of control, non-infected rhesus macaques (n = 6, black closed circles; mean ± SEM) across gestation. e Ultrasound images showing the appearance of the placenta (white outline) in a representative control uninfected dam (102dGA) and f a ZIKV-infected dam at 61dpi (112dGA). Both images were acquired at the same gain setting. g CE-US calculated microvascular flux rate from control (n = 27 spiral artery sources from n = 7 animals) and ZIKV-infected animals. Co-registered axial MRI images from animal D23046 are shown in h (T2w HASTE acquisition) and i (quantitative T2* map). Regions of interest (ROIs) delineating both placental lobes are indicated by the dashed green lines, and a focal region of pronounced hypoperfusion is circled. j Plots normalized histograms of placental relaxation rate (R2* = 1/T2*) for all five ZIKV-infected animals (colored curves), along with the median histogram for the six control animals (black curve). k Plots median values of PSv_i (averaged over all placental lobules) for each of the six control animals (black points) versus values for the five ZIKV-infected animals (colored points), stratified by gestational age at time of infection. l Shows relative blood flow for both the primary and secondary placental lobes in D23046 determined from DCE-MRI measurements for the perfusion domains corresponding to each placental lobule, as determined from DCE-MRI measurements. Darker shades indicated regions of the placenta that are hypoperfused. m Image of the fetal side of the placenta in this animal, demonstrating infarcted regions (white arrows) corresponding to the regions of hypoperfusion seen on DCE-MRI. Of note, ZIKV RNA was detected at five sites across the two placental lobes, indicated by the white asterisk. These ZIKV + areas predominantly correspond to hypoperfused cotyledons. p < 0.0001 one-way ANOVA with Tukey’s multiple comparison test, ***p < 0.001, **p < 0.01

Fig. 5

Uteroplacental histopathology of ZIKV-infected cases compared with gestational-age-matched negative controls. a All five ZIKV-infected cases showed placental infarctions (asterix) with large gross infarctions seen in the cases infected at days 31–51. b ZIKV cases had conspicuous villous stromal calcifications (arrows), which are a consequence of stromal fibroblast cell death (c). ZIKV cases were positive for chronic decidualitis with plasma cells and lymphoplasmacytic leuokocytoclastic vasculitis (d), which was associated damage to the uterine spiral arteries (e) and vascular luminal (astericks) narrowing (f). Scale bar is 100 µm

Fig. 6

Magnetic resonance imaging of fetal brains. A T2W brain template generated from 16 control fetuses at gestation age 135dGA ± 2d (a), a typical control at 135dGA (b), and the 5 ZIKV-exposed fetal brains (c) are shown in sagittal (left) and axial (right) views at the indicated slice locations (insets). Anterior/posterior, superior/inferior, lateral/medial, and left/right are marked on the 3D brain surfaces (insets) and template images (a). Despite the absence of microcephaly, a thinner somatosensory gyrus (c, red arrows), and a missing secondary sulcus (a–c, yellow arrows), was identified for animal F35467, but not other ZIKV-infected animals

Fig. 7

Cortical surface model and apparent diffusion coefficient (ADC) map of fetal brains. For animal F35467, a 3D cortical surface model was generated and curvature of the cortical surface was rendered on the surface (a, top, lateral view; bottom, dorsal view). Compared to an age-matched control (b), a significantly thinner postcentral gyrus, especially at the dorsal portion, can be detected (a, arrows). A missing secondary sulcus (b, asterisks) was also identified on the postcentral gyrus of animal F35467. c An ADC map generated from 8 control fetuses at gestation age 135dGA ± 2d and d one ADC map from ZIKV animal F35797 are shown at 4 axial slices at indicated levels (inset). Lower ADC is observable on the right hemisphere at the regions of motor, somatosensory, as well as auditory cortices compared to the left hemisphere (red dashed circles)

Fig. 8

Histological images of fetal genitourinary tract tissues. At delivery, fetal tissues from testis (a, b), seminal vesicle (c), prostate (d), seminal vesicle (e), and pelvic floor (f) were paraffin-embedded, sectioned and stained with hematoxylin and eosin. Shown are representative images. Arrows indicate the presence of the following: a testicular hemorrhage; b testicular apoptotic cell; c seminal vesicle degeneration and apoptotic bodies in both the epithelial and smooth muscle layers; d vacuolar degeneration and apoptotic cells in the prostate; e seminal vesicle vacuolar degeneration and apoptotic bodies; and f lymphocytic infiltrates in the skeletal muscle of the pelvic floor adjacent to the bladder. Scale bar is 200 µm (a, c), 100 µm (f), and 50 µm (b, d, and e)

Fig. 9

Schematic summary of placental damage following maternal ZIKV-infection. Cross section diagram of placental vasculature demonstrating maternal spiral arteries that perfuse the intervillous space. Notations A–D summarize the placental tissue damage, alterations in perfusion, and immune response to ZIKV infection during pregnancy