Commonly used ZIKV strains.
ZIKV was first identified in the 1940s as a mosquito-borne virus; however, sexual transmission, which is uncommon for arboviruses, was demonstrated more than 60 years later. Tissue culture and animal models have allowed scientists to study how this transmission is possible. Immunocompromised mice infected with ZIKV had high viral loads in their testes, and infection of immunocompetent female mice was achieved following intravaginal inoculation or inoculation via mating with an infected male. These mouse studies lead researchers to investigate the individual components of the male reproductive system. In cell culture and mouse models, ZIKV can persist in Sertoli and germ cells of the testes and epithelial cells in the epididymis, which may lead to sexual transmission even after ZIKV has been cleared from other tissues. ZIKV has also been studied in nonhuman primates (NHPs), which appears to mimic the limited human epidemiological data, with low rates of symptomatic individuals and similar clinical signs. Although refinement is needed, these animal models have proven to be key in ZIKV research and continue to help uncovering the mechanisms of sexual transmission. This review will focus on the animal models used to elucidate the mechanisms of sexual transmission and persistence of flaviviruses.
- Zika virus
- sexual transmission
- animal models
- Sertoli cell
Zika virus (ZIKV) is a single-stranded, positive-sense virus of the genus
|MR766||1947||African||Ziika forest, Uganda||Contemporaneously available strains are heavily passaged in mice, altering their virulence|
|PRVABC59||2015||Asian (American)||Puerto Rico||Isolated from human serum during the American outbreak|
|FSS13025||2010||Asian||French Polynesia||Common isolate used collected prior to the American outbreak. Human isolate|
|IBH30656||1968||African||Nigeria||Isolated from human blood sample|
|H/PF/2013||2013||Asian||French Polynesia||Human isolate|
|BeH815744||2015||Asian (American)||Brazil||Human isolate|
Prior to the major outbreaks in the 2000s, ZIKV had not been detected in the Americas and reported human cases of ZIKV-caused disease were scarce worldwide. Seroprevalence studies in Asia and Africa revealed human populations were exposed to the virus but disease was rarely reported, likely due to a high percentage of asymptomatic infection or because the signs and symptoms were indistinguishable from other known diseases . Data collected during recent outbreaks in Yap Island and across the Americas revealed ZIKV is usually asymptomatic, with symptomatic cases being characterized by flu-like symptoms including self-limiting fever, headache, rash, and myalgia .
Worldwide outbreaks in Gabon, Micronesia and French Polynesia expanded the virus’ known range and susceptible population, allowing prospective epidemiological studies. ZIKV likely arrived in the Americas in 2014, and in 2015 it quickly spread, starting its largest outbreaks to date [4, 5]. The large number of infected patients and heightened medical and scientific response allowed rare outcomes and transmission routes to be noticed. It was possible to identify that ZIKV causes congenital malformations . Prior to this, cases of miscarriage, microcephaly and other congenital malformations, such as microcephaly or damage to brain and eye tissues, had been identified and were retrospectively observed to be correlated with infection [7, 8]. Infected travelers returning home initiated small traveler-associated transmission cycles, some of which were between sexual partners. Sexual transmission was proposed by Foy and colleagues, who described a probable case of ZIKV sexual transmission of a scientist that visited Senegal in 2008 . Many more cases of sexual transmission were reported after the outbreaks in the Americas, confirming that sexual transmission played a role in the spread of ZIKV, although the full impact of sexual transmission may be underappreciated since it occurs concomitantly with the more efficient transmission by mosquito vectors [10, 11, 12, 13].
ZIKV or viral RNA was found to persist in human semen, vaginal secretions and blood [14, 15]. Acquiring samples to study viral persistence in these tissues is difficult because ZIKV infection is frequently asymptomatic and identification of ZIKV-positive samples has to be done by serology or molecular biology, since other viruses that cause similar signs and symptoms are present in the same regions as ZIKV. In Brazil, for example, there are several other flaviviruses and alphaviruses in circulation which produce clinical pictures that cannot be readily differentiated from the signs and symptoms associated with ZIKV. Therefore, molecular and serological tools are critical for the identification and classification of such samples . In many locations, this requires a concerted collaborative effort between hospitals and laboratories, and many samples may need to be shipped before determining which patients came in contact with ZIKV. Additionally, sample collection is made difficult because once symptoms have rescinded, sample collections can no longer be made by hospitals and people may be opposed to having samples collected from them when they are no longer in need of medical care. In spite of these obstacles, several studies were able to assess ZIKV persistence in human samples. In semen, ZIKV RNA was detected for up to 6 months and ZIKV infectious particles could be isolated up to 1 month after primary infection . Vaginal secretions were also found to be positive for RNA for up to 6 months after onset of symptoms . These observations suggest sexual transmission could potentially happen even long after resolution of primary infection. Caution must be taken when trying to directly interpret the consequences of persistent viral RNA since this does not directly imply infectious viral particles. Regardless, transmission is possible, as demonstrated by a woman infecting her partner 44 days after onset of symptoms . Although sexual transmission has most often been reported from males to females, there have been reports of suspected male-to-female , male-to-male  and female-to-male  transmissions. The actual rate of sexual transmission remains unknown.
Our current knowledge of the potential for sexual transmission among members of the
Examples of sexual transmission may be more common in the family
The fact that ZIKV was shown to be sexually transmitted raised many questions, as many aspects of this mode of transmission are poorly understood for these viruses. Namely, the importance of sexual transmission for ZIKV in different regions of the world, the role of cells of the reproductive tract involved in ZIKV persistence and transmission, mutations in the virus which favor successful sexual transmission, interactions of ZIKV with the immune system in the reproductive tract and molecular interactions between ZIKV and host cells allowing for sexual transmission to take place, are all critical points of study. Although abundant data on human disease was generated during the recent outbreaks, animal experiments are an important tool to answer these questions because they allow controlled and rationally designed experiments which could ultimately lead to the development of new vaccines, therapeutics and prophylactic measures. This review will focus on what is currently known for ZIKV infection in the male reproductive tract.
2. Animal models of sexual transmission
Upon discovery of ZIKV, animal experiments were first conducted by Dick . They found that white mice younger than 2 weeks were susceptible to ZIKV infections via intraperitoneal route, whereas mice older than 2 weeks were only susceptible to intracerebral inoculations . Non-human primates inoculated subcutaneously developed neutralizing antibodies against ZIKV and only one animal presented a slight elevation in temperature [39, 40]. Several decades after this initial study, several other model species were used as animal models for ZIKV, but efforts were mostly focused on mice [41, 42, 43] and non-human primates (NHPs) [44, 45, 46, 47], with most studies being conducted after the outbreak in the Americas in 2015–2016. The experimental design and analyses should carefully consider the anatomical and physiological differences from the species used to humans as well as confounding factors such as different viral strains and inoculation titer and route.
2.1 Mouse models
Mouse models of sexual transmission for ZIKV initially utilized either interferon receptor deficient mice lacking type I (A129)  and or type I and II (AG129) interferon receptors . Other models included interferon responsive mouse strains for which a transient knock-down of interferon response was induced by exogenous treatment with antibody against murine type I interferon receptor of wild-type mice or Rag1−/− mice (lacking both B and T lymphocytes) . Several of these models have demonstrated sexual transmission from needle-inoculated male mice to naive female mice following mating. These studies, coupled with the use of surrogate breeding females from which uteri were gavaged, allowed direct assessment of virus and viral RNA shedding efficiency over time. AG129 males infected with ZIKV strain PRVABC59 were shown to shed infectious virus from 7 to 21 days post infection (dpi). Vasectomized males also were shown to shed virus; however, the magnitude was shown to be significantly lower compared to non-vasectomized males . This finding was consistent with a previous study reporting ZIKV RNA shedding in symptomatic men infected with ZIKV in which vasectomized men shed significantly lower levels of ZIKV RNA .
The necessity for the use of mice lacking the ability to respond to interferon is due to the inherent resistance of murine STAT2 from being inhibited by ZIKV NS5 . In contrast, human STAT2 has been demonstrated to be highly susceptible to antagonism by ZIKV NS5 . As such, the subsequent development of humanized STAT2 mice have proven useful as a fully interferon responsive model for which ZIKV can replicate and induce pathogenic responses .
2.1.1 Sexual transmission murine models
A number of different models have been developed in order to directly assess sexual transmission Sexual transmission has been modeled via direct intravaginal inoculation. In one model, AG129 male mice euthanized and caudal epididymal lumen (containing sperm) was collected. This suspension was used to inoculate female AG129 mice. In this study, the antioxidant ebselen was used to alleviate testicular pathology caused by ZIKV. Although intravaginally inoculated female AG129 mice with sperm from male mice treated with ebselen demonstrated reduced mortality, sexual transmission was not prevented, as female organs (ovary/fallopian tubes, spleen and brain) were infected . Other studies have used homogenized accessory gland fluid and epididymal lumen fluid from Ifnar1−/− male mice subcutaneously inoculated with PRVABC59 to intravaginally inoculated female AG129 mice. In this study, females became viremic and succumbed to infection. Furthermore, progesterone pre-treatment of female mice before intravaginal inoculation was shown to increase mortality of females . IFNAR1−/− male mice inoculated subcutaneously with PRVABC59 and then at 14 dpi or 35 dpi, prostatic and seminal vesicular homogenates and epididymal flushes were collected. Female AG129 mice inoculated intravaginally with this insemination fluid failed to become viremic .
Intravaginal inoculation of di-estrus timed AG129 mice or LysMCre+IFNARfl/fl mice (lacking IFNAR in myeloid cells) was shown to result in viremia and virus replication in peripheral organs and in the vaginal tissues measured by viral assay by RT-PCR through 10 dpi . In an alternative study, immunocompetent C57BL/6 N female mice intravaginally inoculated with PRVABC59 showed a slight increase in viral RNA in the lower female reproductive tract (LFRT) from dpi 1 to 2 and mRNA expression of type I and III IFNs, IRF3/7, RIG-I, and MDA-5 was comparable to uninfected controls, suggesting that a dampened antiviral immune response occurs in the LFRT in response to infection with ZIKV. When mice were treated with an enhancer of RIG-1 signaling, the increase in ZIKV RNA in LFRT was not observed . After intravaginal inoculation during the diestrus phase, ZIKV was shown to replicate in vaginal mucosa of wild type (WT) C57BL/6 and Ifnar1−/− mice. Fetuses from pregnant WT C57BL/6 mice developed intrauterine growth restriction and fetal brains were infected following intravaginal ZIKV inoculation. Fetuses from pregnant IFNAR1−/− mice developed severe intrauterine growth restriction and fetal death was observed. The pregnant females were found to develop viremia. These data are suggestive of an ascending infection from the vaginal tissues to the uterus for fetal involvement .
In addition to sexual transmission models that have utilized intravaginal exposure to model sexual transmission potential, several studies have addressed transmission dynamics through direct coitus models. AG129 males were found to sexually transmit to naïve AG129 females in 50% of all matings as measured by subsequent viremias in mated females. This initial study demonstrated
2.1.2 Potential sources of sexually transmitted virus/ZIKV infection of the murine testes and epididymides
The majority of ZIKV detected in the seminal fluid of infected AG129 mice during the peak timing of sexual transmission (10–12 dpi) was from the supernatant fraction, suggesting cell-free ZIKV may be largely responsible for sexual transmission. In this study, the testes and epididymides were determined to be infected concurrently and epididymal epithelial cells were identified as the predominant cell population infected in epididymides and shown to contain replicating ZIKV by
Tissue restricted ZIKVs generated through the incorporation of microRNA target sequences within recombinant ZIKVs were utilized to assess the importance of different cell populations for sexual transmission potential. Testes-restricted ZIKVs could still infect the epididymides, demonstrating a hematogenous/lymphogenous route of infection. Epididymides-restricted ZIKV had high titers in epididymides by plaque assay, but immunohistochemical analysis confirmed epididymides-restricted ZIKV did not replicate in the epididymal epithelium, suggesting that ZIKV is transported from testes to epididymides via excurrent ducts in a cell-free form or transported in sloughed spermatids/infected luminal leukocytes, and ZIKV can infect the epididymides via hematogenous/lymphogenous route of infection .
2.1.3 Persistence and tropism of ZIKV on the male reproductive tract
Tropism and persistence of ZIKV in the male reproductive tract may be the key factor responsible for the presence of the virus in semen even long after initial infection. A model of the male mouse reproductive tract is shown in Figure 1A. Based on
Intracellular viral persistence is likely an important component for long-term sexual transmission. Although the cells that act as reservoirs of ZIKV in testes and the reproductive tract are unknown, possible reservoirs in the host are SC [67, 68] germ cells , Leydig cells  and epididymal epithelial cells [59, 62] which have been shown to support persistent infections of ZIKV (Figure 1). When primary SCs persistently infected with two strains of ZIKV (PRVABC59 or MR766) were monitored for a period of 6 weeks, it was found that 15% of the cells were still positive for both strains of ZIKV . In this same study, Leydig cells were not observed to support persistence . The mechanisms underlying persistence of ZIKV in testes are likely multifactorial and represent a complex phenomenon involving interactions between viral and host factors that needs to be studied in depth. The interactions that occur between ZIKV and host factors that are required for long-term infection of the testes/epididymides are poorly understood and viral persistence likely requires a balance between efficient viral replication and damage caused to host cells. The AXL receptor tyrosine kinase, which was previously shown to be required for entry of ZIKV and other flaviviruses into certain cell types, has been shown to be required for ZIKV entry in SCs , but may also be involved in negatively regulating SCs innate immune response . ZIKV infection of SCs results in gene expression changes, with upregulation of antiviral pathways, dysregulation of junction and growth pathways [67, 73]. Although the immune system of the reproductive tract in females is able to eventually clear ZIKV, largely due to interferon signaling [74, 75], such as the type III interferon lambda , the microenvironment of the male testes is an immune-privileged site that lacks adequate response for clearance .
2.1.4 Viral genetic changes associated with persistence
Mosquito-vectored flaviviruses that can result in persistent infections are usually associated with an accumulation of adaptive mutations, suggesting that viral genetics may also play an important role for establishing persistence. Examples of mutations in viruses causing persistence were shown with ZIKV, WNV [78, 79, 80], JEV , YFV [24, 82], and DENV . These mutations happen in different viral proteins and appear to affect various viral functions, but in general attenuate the virus to cause less cytopathic effects, which is likely important for persistence. Identifying the mutations and cellular components that allow flaviviruses to cause long-term infections may elucidate mechanisms by which ZIKV is able to remain stable in seminal fluid long after infection. Although functional mutations in ZIKV may have been found in the male reproductive tract and ejaculates of mice , these were do not appear to be specific to these tissues as they were also found after passaging in Vero cells , and no mutation has been identified to be required for persistence and/or sexual transmission. Continuing research efforts using animal models are paramount to understand the mechanisms of ZIKV persistence, sexual transmission, reservoirs and interactions with cellular targets.
2.2 Non-human primate (NHP) models
NHP models are also frequently used to study ZIKV since their biology more closely relates to a human than does mouse biology. Several non-human primates were shown to be susceptible to ZIKV and used to study pathogenesis and infection of the reproductive tract, including cynomolgus macaques (
Because NHPs are evolutionarily close to humans, it is not surprising that these animals and their reproductive tract are anatomically and physiologically more similar to humans than other models such as mice; thus, NHPs are often considered to be one of the most relevant animal models [85, 86]. Unlike mice, immunocompetent NHPs are more suitable to study ZIKV infections, although the extent to which disease is mimicked is difficult to assess and varies with NHP species. Most NHP models were similar to humans in that most infections were asymptomatic, with clinical signs observed in some individuals [87, 88]. Considering that many human infections are asymptomatic, it is currently unclear if the clinical signs displayed by NHPs are a good model for disease rate and severity in humans. In a study using four cynomolgus macaques one individual was shown to present reduced body temperature . Studying ZIKV pathogenesis in NHPs is important to understand disease progression, clinical signs, negative effects to the reproductive tract and outcomes of pregnancy. These data also inform the design of sexual transmission studies as far as titers, timepoints and strain used for infection, tissues that are important for disease and persistence so sexual transmission can take place. It is worth noting that these experimental infections are performed using virus in a needle while most  natural infections occur by a mosquito bite; mosquito saliva has immunomodulatory properties that have been shown to enhance disease in other flaviviral infection [90, 91, 92] and could be another confounding factor in these studies.
2.2.1 Cynomolgus macaque
Cynomolgus macaques were shown to be suitable as models for ZIKV pathogenesis and sexual transmission [87, 88, 93]. Infection with 5 log10 of plaque forming units (PFU) of ZIKV from various geographical origins resulted in viremia which peaked at 2-4 dpi at 4-7 log10 ZIKV genome copies/ml [87, 88]. The outbreak strain PRVABC59 was more virulent than the Asian FSS13025 and the African IBH30656 ZIKV, with the animals infected with the PRVABC59 strain being viremic for extended periods of time. Bodily fluids checked did not include seminal fluid, and shedding of virus in the urine and saliva was not observed with either the FSS13025  or IBH30656 strains. ZIKV FSS13025 was detected in testes , consistent with murine infections with this strain . On the other hand, ZIKV PRVABC59 was detected in urine, saliva and testes . The fact that virus was found in testes suggests this species of NHP could model virus persistence in testes well and may also be a positive feature of this model for sexual transmission in general. The cynomolgus macaque model has also been shown to likely support sexual transmission, as macaques inoculated with 7 log10 PFU of virus intravaginally and intrarectally became viremic 50% and 100% of the time, respectively . To understand the implications of these findings would be important to know the ranges of ZIKV titers in the semen. Although it is not clear what the viral titers are in most human semen samples, studies detected up to 9 log10 RNA copies/ml of virus in semen of patients [17, 89].
2.2.2 Olive baboon
As with cynomolgus macaques, infection of olive baboons with different strains of ZIKV did not result in overt clinical signs. Following subcutaneous inoculation of a French Polynesian ZIKV (H/PF/2013), the baboons presented viremia that peaked at 3 and 4 dpi . Around 40 days post infection, tissues were collected and virus was found in lymph nodes and epididymides, suggesting these are the places where the virus can persist even after viral clearance from the blood . This suggests the testes were likely infected at some point and that the epididymides may also play a role in virus persistence . Olive baboons were also used to model ZIKV infection during pregnancy. Infection of 4 log10 of ZIKV H/PF/2013 resulted in vertical transmission in 3 out of 4 pregnant NHPs . Unlike the non-pregnant animals, all dams presented rash and conjunctivitis . Fetal death and defects in the frontal cortex of the fetus were observed .
2.2.3 Rhesus macaque
The rhesus macaque is perhaps the most utilized NHP model to study ZIKV infection. Many studies of ZIKV were done using this species, including pathogenesis analyses in pregnant and non-pregnant animals, immunological and serological studies, and testing of anti-ZIKV vaccines or drugs [44, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106]. A study in which ZIKV was inoculated intravaginally in rhesus macaques to mimic sexual transmission found ZIKV RNA in the reproductive tract of all 6 animals infected, thus raising the question of whether fetal disease could be more pronounced after sexual transmission when compared to vectored transmission . This study lacked subcutaneous inoculation controls and further studies need to be conducted to confirm these findings. Another study comparing the intravaginal and subcutaneous routes using ZIKV PRVABC59 found peak viremia at 5–8 dpi that were variable in titer (3–7.5 log10 PFUs). Although the types of tissue found to be positive differed between subcutaneous and intravaginal inoculated animals, there was no obvious preference for reproductive tract tissues in the intravaginal route . The data showed that intravaginal infection resulted in less CD11Chi myeloid cells, reduced expression of programmed cell death protein 1 (PD-1) in natural killer cells (NK) and more Ki67+ CD8+ central memory cells, indicating the route of infection may play a role in shaping the immune response .
Persistent ZIKV in reproductive tissues may play a role in sexual transmission long after primary infection. After intravenous inoculation of 5 log10 of a Brazilian ZIKV (BeH815744) in female rhesus macaques, ZIKV RNA was detected in multiple tissues of the NHPs, including reproductive tissues 14 dpi . Lymphoid tissues had the highest detectable amount of viral RNA, suggesting these organs, which span many parts of the body, may act as possible viral reservoir .
2.3 Other animal models
Several other animal models have been used to study ZIKV but most are not focused on sexual transmission. These models include guinea pigs [109, 110, 111], hamsters , bats , chick embryos, piglets [114, 115] and boars . Porcine fetuses were shown to present mild to severe neuropathology upon ZIKV infection [114, 115]. Boar semen was inoculated with ZIKV, but it was concluded that ZIKV does not appear to cause cell damage and cannot replicate efficiently or persist in the semen of this species . With respect to the neotropical chiropteran model, viral RNA was found in different tissues of fruit bats 28 dpi, suggesting ZIKV can infect bats which may serve as virus reservoirs. Bats did not show any signs of disease . Stat-2 knockout hamsters infected with ZIKV have shown presence on infected cells with morphology of SCs and spermatogonia, suggesting this could be a suitable model to study persistence of the virus in testes . The guinea pig models report conflicting results. One study with ZIKV challenge mid-gestation showed no evidence of infection ; however, Kumar et al. show guinea pigs inoculated subcutaneously with PRVABC59 had viremia and presented signs such as fever, hunched posture and detectable viral RNA in the blood . Deng et al. showed that intranasally-infected guinea pigs have virus in the sera, saliva and tears .
ZIKV has emerged explosively since 2007, causing an epidemic in the Americas in 2015/16 and become a matter of global health importance. Although data from epidemiologic analyses and case studies helped shed light on the diseases caused by ZIKV, animal models will be important to substantiate and extend these findings under controlled experimental settings. Animal models can be—and have already proven to be—useful as a tool to understand ZIKV transmission and ZIKV-caused illnesses. Animal models do not fully recapitulate diseases as seen in humans; therefore, it is critical to consider the advantages and drawbacks of each model when designing and executing the experiments as well as interpreting the data. Models to study sexual transmission are currently scarce and need further development. Challenges include scarcity of good models to study ZIKV sexual transmission, low number of animals used and the requirement to improve reproducibility of the findings from animal models, which is caused by differences in experimental conditions and the number of animals used. As in certain cases increasing the number of animals used is not possible due to it being prohibitively expensive or posing ethical issues, future experiments assessing sexual transmission of ZIKV should focus on optimizing experimental design and analysis, when possible standardizing experimental conditions so they can be compared between studies.
We thank Drs. Sasha Azar and Nikos Vasilakis for their critical reading of the chapter and helpful discussions.
Conflict of interest
We have no conflicts of interest to declare.