4077

Functional and Developmental Effects of Zika Virus Infection and Hydroxychloroquine Treatment in Mouse Placenta
Kelsey V Meinerz1, Brooke Liang2, James D Quirk3, Indira U Mysorekar2,4, and Joel R Garbow3,5

1Physics, Washington University in St Louis, Saint Louis, MO, United States, 2Obstetrics and Gynecology, Washington University in St Louis, Saint Louis, MO, United States, 3Mallinckrodt Institute of Radiology, Washington University in St Louis, Saint Louis, MO, United States, 4Center for Reproductive Health Sciences, Washington University in St Louis, Saint Louis, MO, United States, 5Siteman Cancer Center, Washington University in St Louis, Saint Louis, MO, United States

Synopsis

Functional and developmental effects of Zika virus (ZIKV) infection and hydroxychloroquine (HCQ) treatment on the murine placenta were measured using MRI and oxygen enhancement to assess placental volume and oxygenation. HCQ treatment helps to prevent adverse placental volume outcomes earlier in gestation, but late-gestation results are similar to those of ZIKV-infected dams. R1 and R2* measurements show that HCQ has a significant effect on placental oxygenation throughout gestation, whereas ZIKV does not significantly affect these metrics.

Introduction

In 2015, Zika virus (ZIKV) emerged as a global health threat after infections in pregnant females in Brazil were causally linked to fetal microcephaly, intrauterine growth restriction (IUGR), and spontaneous abortion. ZIKV can severely impair fetal development, requiring rapid and efficient methods of intervention to prevent vertical transmission from mother to fetus. Mouse models of Zika infection have shown that maternal-fetal transmission of ZIKV is trans-placental1, and that hydroxychloroquine (HCQ) can attenuate infection and improve fetal outcomes2. This pilot study used established MRI methods to non-invasively probe placental structure and function to assess the effects of ZIKV infection and HCQ treatment.

Methods

Pregnant, wild-type mice lacking an intact interferon response were infected subcutaneously with 103 focus-forming units (FFU) of ZIKV (n=17) or saline (n=9) at embryonic day 9.5 (E9.5). Half of each cohort was treated with 40 mg/kg of HCQ intraperitoneally daily from E10.5 to E14.5. To measure placental volume and function, MRI experiments were performed at 4.7-T on E15.5 and E18.5. Placental volumes (5-10/mouse) were calculated by manual segmentation of respiratory-gated, T2-weighted, multi-slice fast-spin-echo images. 3D gradient-echo variable-flip-angle (VFA) and 3D multi-echo, gradient-echo images were acquired while breathing both medical air and carbogen (95% O2/5% CO2) for R1 and R2* mapping, respectively. Optimal relaxation rate constants were determined via Bayesian probability theory based software3,4. Following imaging on E18.5, the mice were euthanized, fetal sizes measured, and maternal and fetal tissues harvested for ZIKV titering and histopathological analysis.

Results

Murine placentas consist of both maternal and fetal structures that can be visualized via MRI and histology (Fig. 1). Placental volumes of ZIKV-infected animals were smaller than those of healthy controls at both time points (Fig. 2). At E15.5, mean placental volumes were 0.060 cm3 (ZIKV-infected) vs. 0.063 cm3 (healthy). At E18.5, the means were 0.059 cm3 vs. 0.066 cm3, respectively. In healthy/HCQ-treated animals, mean volumes were 0.065 cm3 (E15.5) vs. 0.071 cm3 (E18.5); in the ZIKV-infected/HCQ-treated animals, 0.063 cm3 (E15.5) vs. 0.058 cm3 (E18.5).

Figure 3 shows that the average R1 values at both time points, and under both breathing-gas conditions, are similar for control and ZIKV-infected animals in the absence of HCQ treatment, with average R1 values of 0.55-0.59 s-1. In HCQ-treated animals, both ZIKV-infected and uninfected, R1 increased significantly, with an average R1 of > 0.7 s-1 for all cases.

Figure 4 shows average R2* values across time points and under both breathing-gas conditions for each cohort. On room air, both cohorts showed increased R2* values from E15.5 to E18.5 (84 s-1 vs. 119 s-1 or 43% increase for healthy dams; 87 s-1 vs. 110 s-1 or 27% increase for ZIKV-infected). For HCQ-treated animals, there was no significant change in baseline R2* across gestation.

Figure 5 shows the change in R2* (air- carbogen) at both gestational ages. The mean change in R2* for healthy and ZIKV-infected dams was 15-20 s-1 on both days. Similarly, the R2* change in ZIKV-infected/HCQ-treated dams decreased from 23 s-1 (E15.5) to 9 s-1 (E18.5), while for healthy/HCQ-treated dams the change was ~3 s-1 on both days.

Discussion

The smaller placental volumes in infected dams (Fig. 2) suggest that ZIKV inhibits proper development of the murine placenta across gestation. Additionally, ZIKV-infected placentas show no change in mean volume from E15.5 to E18.5, compared with a 3% increase observed in healthy dams. At E15.5, mean placental volumes in ZIKV-infected/HCQ-treated animals are equivalent to those of healthy dams at E15.5, but by E18.5 is comparable to ZIKV without treatment. This suggests that while HCQ may initially preserve placental development in ZIKV-infected animals, without sustained treatment the late-gestation adverse outcomes are unaffected.

Somewhat surprisingly, HCQ treatment had a larger effect on placental relaxation parameters, R1 and R2*, than did ZIKV infection (Figs. 3 – 5). The increase of R1 in both healthy and ZIKV-infected animals following HCQ treatment could suggest increased oxygenation of the placental tissue. Both healthy and ZIKV mice show significant R2* increases with aging, an effect eliminated by HCQ treatment. At both gestational ages, carbogen breathing produced a significant decrease in R2* in untreated healthy and ZIKV-infected mice, but not healthy/HCQ-treated animals. At E15.5, a similar decrease in R2* in response to carbogen is observed for ZIKV-infected/HCQ-treated mice, an effect that is lost at E18.5. In untreated mice, changes in R1 and R2* from E15.5 and E18.5 reflect expected changes in O2 content and utilization associated with placental aging. The mechanism by which HCQ affects R1 and R2* values in healthy and ZIKV-infected animals, and its differential effects on aging, are the subject of ongoing research.

Acknowledgements

Research supported, in part, by Mallinckrodt Institute of Radiology Pilot Fund 18-016 (WUSM), the Howard Hughes Medical Institute, and National Institutes of Health grants R01 HD086323, 5T32 EB014855, and R01 HD091218. Valuable discussions with Professor Joseph JH Ackerman, Washington University in St Louis, and Professor Bin Cao, Xiameng University, are gratefully acknowledged. Special thanks to the Center for Reproductive Health Sciences at Washington University School of Medicine for assistance with histopathological processing of tissue samples.

References

  1. Miner, J.J., et al., Zika Virus Infection during Pregnancy in Mice Causes Placental Damage and Fetal Demise. Cell, 2016.165: p. 1081-91.
  2. Cao, B., et al., Inhibition of autophagy limits vertical transmission of Zika virus in pregnant mice. J. Exp. Med., 2017. 214(8): p. 2303-13.
  3. Bretthorst G.L. (1990) An Introduction to Parameter Estimation Using Bayesian Probability Theory. In: Fougère P.F. (eds) Maximum Entropy and Bayesian Methods. Fundamental Theories of Physics, vol 39. Springer, Dordrecht.
  4. G. Larry Bretthorst, Bayesian Analysis of Common NMR Problems, a Bayesian signal analysis toolbox available for free download at: http://bayesiananalysis.wustl.edu/index.html

Figures

A) Representative coronal T2w fast spin echo MR image of fetoplacental units within a pregnant mouse at gestational age E15.5. A single placenta is marked on this image, with green outlining the labyrinth zone and yellow the junctional zone. B) Histological sample of murine placenta annotated with black lines to mark the edges of the labyrinth and junctional zones. The blue box indicates the field of view shown in C) magnified 10x. C) Zoomed-in region showing: junctional zone (i) and labyrinth zone (ii).

No pattern represents E15.5 and diagonal lines represent E18.5. No fill represents untreated and gray fill represents HCQ-treated. Each point is the average placental volume of all fully segmented placentas within a single dam, arranged by cohort. Average placental volume in infected dams was smaller than those of healthy dams. Additionally, infected dams showed no change in average volume from E15.5 to E18.5, whereas healthy dams show a slight increase. ZIKV-infected/HCQ-treated showed increased average placental volume at E15.5 compared to the ZIKV-infected group, but this effect disappeared by E18.5.

No pattern represents E15.5 and diagonal lines represent E18.5. Blue represents animals breathing air and red represents animals breathing carbogen. No fill represents untreated and gray fill represents HCQ-treated. Boxes with no fill represent E15.5 animals and boxes with diagonal lines represent E18.5 animals. Each point is the average R1 for all placentas within a single dam, arranged by cohort. HCQ treatment significantly increased the average R1 for both infected and uninfected animals.

No pattern represents E15.5 and diagonal lines represent E18.5. Blue represents animals breathing air and red represents animals breathing carbogen. No fill represents untreated and gray fill represents HCQ-treated. Boxes with no fill represent E15.5 animals and boxes with diagonal lines represent E18.5 animals. Each point is the average R2* for all placentas within a single dam, arranged by cohort. Both the healthy and ZIKV-infected animals showed a decrease in R2* upon switching from air to carbogen and an increase in R2* as gestational age increases. These patterns appear to be disrupted by HCQ treatment.

No pattern represents E15.5 and diagonal lines represent E18.5. No fill represents untreated and gray fill represents HCQ-treated. Each point is the average change in R2* (upon switching from air to carbogen) for all placentas within a single dam, arranged by cohort. The R2* of both healthy and Zika-infected dams increased when breathing carbogen at both gestational ages. Zika-infected/HCQ-treated mice also showed increased R2* with carbogen at E15.5. Healthy/HCQ-treated showed almost no change in R2* at both gestational ages, as did Zika-infected/HCQ-treated dams at E18.5.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
4077