Abstract

METHODS

RESULTS

To further elucidate testicular shedding dynamics of ZIKV, we used TEM to visualize virions within the testes (developing sperm) and in epididymal homogenate (mature sperm). We found virus particles surrounding developing sperm at the luminal edge of the seminiferous epithelium of the testes as they are being released by the Sertoli cells during the process of spermination (Figure 1A). Transmission electron microscopy of sperm within epididymal homogenate demonstrated ZIKV particles to be attached to mature sperm (Figure 1B). We further found virions to be present within cytoplasmic droplets attached to mature sperm (Figure 1B) and associated with epdidymal residual bodies (Figure 1B). These observations are consistent with high virus titer in testes and epididymis at 5 days after infection (Figure 1C). We clearly observed a trend where ZIKV-infected mice had a higher proportion of sperm with cytoplasmic droplets compared with epididymal homogenate from mock-infected mice.

Open in a separate window

Figure 1.

Virions were detected in testes and epididymis after Zika virus (ZIKV) infection. Transmission electron microscopy depicts ZIKV-infected cells in the testes (A) and epididymis (B) of Ifnar1^−/− mice at 5 days after infection. Arrows indicate ZIKV virions within the infected cells. Scale bars = 500 nm. C, The ZIKV titers in testes and epididymis at 5 days after infection were examined by plaque assays using Vero cells. The results are expressed as the mean titers. The detection limit of plaque assays is 1.8 log₁₀ plaque-forming units per gram.

Because we and others have shown that Ifnar1^−/− mice are susceptible to ZIKV through intravaginal exposure [3, 9], we aimed to develop a mouse model of sexual transmission by mating an infected Ifnar1^−/− male mouse with a naive Ifnar1^−/− female mouse. To this end, male Ifnar1^−/− mice were infected subcutaneously with 10⁵ PFUs of ZIKV^MEX. As we have previous shown a high level of viremia to be detected 3–7 days after infection [6], we cohoused naive Ifnar1^−/− female mice with infected male mice at 3 days after infection. Female mice were monitored for the presence of a copulation plug and dissected 18 days after plug detection (Figure 2A). Fetal weight was significantly reduced in pups from female mice cohoused with infected male mice compared with pups from the control female mice cohoused with naive male mice (Figure 2B). Of note, some fetuses displayed clear evidence of intrauterine growth restriction (n = 4/38 fetuses from 5 pregnant mice) as compared with those from the control Ifnar1^−/− mice at E18.5 (Figure 2B). Strikingly, we also observed fetuses with ocular malformations in 2 (from different litters) of 38 fetuses from mating with the infected sire. We also detected virions in fetus brain by TEM (Figure 2C). These results strongly demonstrate that ZIKV can transmit from an infected male to a naive female and causes fetal birth abnormalities.

Open in a separate window

Figure 2.

Zika virus (ZIKV) transmits from infected males to females and causes fetal growth restriction. A, Schematic of the experimental model for sexual transmission is described. B, Total body weight in pups at embryonic day 18.5 (E18.5) from infected animals (n = 30 pups from 4 separate litters cohousing with uninfected male mice; n = 38 pups from 5 separate litters cohousing with infected male mice; Student’s t test, *P < .05, average ± SEM). Pups with red or blue symbols show the ocular malformation or growth restriction, respectively. Pictures represent the typical Ifnar1^−/− pups at E18.5. Blue arrow shows ocular malformations in the pups. Scale bar = 1 cm. C, Zika virus–infected cells of fetal brain from the female cohousing with the infected male killed on E18.5 were observed by transmission electron microscopy. Arrows indicate the ZIKV virions within the infected cells. Scale bar = 500 nm. D, Zika virus envelope protein–specific antibodies in serum were measured using an enzyme-linked immunosorbent assay. The detection limit of reciprocal titer is 1.7. * P < .05 and ** P < .01 by 1-way analysis of variance followed by Tukey’s multiple comparison test. Abbreviations: inf, infected; IVAG, intervaginally; PFU, plaque-forming unit; SC, subcutaneously; ZIKV, Zika virus.

To determine whether sexual transmission of ZIKV resulted in measurable viral replication in the vaginal mucosa, we examined ZIKV RNA levels in the vaginal wash in the dams every other day after sexual contact. However, we could not detect the ZIKV RNA in the vaginal wash, suggesting that ZIKV replicated minimally within the genital tract of the dams. To assess whether naive females seroconverted and produce ZIKV-specific antibodies, we examined antibody responses against ZIKV envelope protein by ELISA. As shown in Figure 2D, 1 of 7 pregnant females generated ZIKV-specific antibodies over the time points examined, suggesting that ZIKV transmitted from infected male Ifnar1^−/− mouse to female Ifnar1^−/− mouse. Because high virus titers were detected in testes and epididymis in infected male Ifnar1^−/− mice at 5 days after infection (Figure 1C), we estimated virus titers in the inoculated semen and infected female mice intravaginally to be approximately 750 PFUs of ZIKV per ejaculate. As previous models of vaginal infection required the use of Depo-Provera, a progestin that drives female mice into the diestrus phase when the vaginal wall is the thinnest and most susceptible to viral infections [3, 9], we examined the virus-specific responses with or without Depo-Provera treatment. Although direct intravaginal infection with 750 PFUs of ZIKV without the use of Depo-Provera did not induce virus-specific antibody in the serum at 21 days after infection, mice treated with Depo-Provera elicited significantly higher virus-specific antibodies (Figure 2D).

DISCUSSION

Sexual transmission of arthropod-borne viruses before ZIKV has not previously been documented. Yet, there have been several incidences where a confirmed ZIKV infection has been found among sexual partners where arthropod involvement in transmission was not suspected [10]. These reports suggest ZIKV is unique among flaviviruses in this transmission ability.

We now demonstrate using a mouse model that ZIKV can be sexually transmitted and cause intrauterine growth restriction in fetus/pups. Similar to Ebola virus, ZIKV is able to persist in semen for a long time [11], and our TEM results show intact virions both on and within sperm, suggesting that seminal fluid, as well as the sperm within the semen, can serve as a vehicle of transmission. This finding suggests the possibility that ZIKV could invade an ovum with a sperm, causing direct infection of the ovum, resulting in deleterious developmental effects. Semen also contains proteins and lipids. A previous study showed that certain seminal factors, such as semen-derived amyloid fibrils, facilitate virion attachment to cells and increase the infectivity of human immunodeficiency virus type 1 in vitro [12]. Although the contribution of seminal fluid on ZIKV infection has not yet been investigated, seminal factors that enhance or suppress ZIKV infection could potentially be examined using our sexual transmission model. Further, we noted that ZIKV-infected mice had a higher proportion of sperm with cytoplasmic droplets compared with mock-infected mice. The retention of cytoplasmic droplets on sperm has been associated with infertility. These observations, therefore, may help to explain why sperm from ZIKV-infected mice exhibit reduced fertility [13]. Although ZIKV RNA and antigens have been previously detected in sperm [5, 8], our TEM data definitively demonstrates that sperm are ZIKV carriers and strongly support sexual transmission.

Our study also demonstrates that female mice can acquire ZIKV through the vaginal route during the peak of estrus phase, a stage required for successful pregnancy. Interestingly we found direct intravaginal infection with 750 PFUs of ZIKV without the use of Depo-Provera did not induce virus- specific antibody production, but the administration of Depo-Provera elicited significantly higher virus-specific antibodies. These results suggest that vaginal ZIKV inoculation from males by semen may be insufficient to induce an antibody response against ZIKV during the peak of estrus phase, a stage required for successful pregnancy. Interestingly, 1 of 2 female mice cohoused with an infected male mouse for 3 weeks, presumably resulting in multiple mating events, showed seroconversion without pregnancy (data not shown), suggesting that viral input from multiple seminal inoculations or viral input during diestrus phase can elicit adaptive immunity. Nevertheless, our results are consistent with the notion that the vaginal route of infection with ZIKV is more efficient during the diestrus-like phase than the estrus-like phase [14].

Recently a ZIKV sexual transmission model was published using female CD-1 or AG129 mice mated with ZIKV-infected male AG129 mice, whereas we used the type I interferon receptor–deficient mouse [15]. Although this group demonstrated sexual transmission, here we add to the literature through finding fetal growth restriction and ocular malformations. In conclusion, we demonstrated that ZIKV is associated with sperm, transmits from male to female sexually, and causes fetal abnormalities. This sexual transmission model recapitulates phenotypes observed in humans and may be useful for drug or vaccine evaluation in vivo.

Notes

Acknowledgements. R. U., K. A. J., A. I., and E. F. conceived the study. R. U., K. A. J., A. I., and E. F. designed the experiments. R. U., K. A. J., and J. H. performed the experiments and analyzed the data. K. S. and T. L. H provided intellectual guidance in the production and interpretation of TEM data. R. U., K. A. J., A. I., and E. F. wrote the manuscript with contributions from all coauthors. We thank Sarah Householder and Laura J. Yockey (Yale University) for discussion.

Financial support. This work was supported by the National Institutes of Health (5T32 AI007019-40 to K. A. J.; 4T32HL007974-15 to J. H.; and R21AI131284 to A. I.). E. F. and A. I. are investigators within the Howard Hughes Medical Institute.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References