Transmission in host population is the main motor of viral evolution 
. Herpesviruses have co-evolved with their host for millions of years and have therefore developed sophisticated mechanisms to persist and transmit in presence of protective immune response 
. This is particularly the case for gammaherpesviruses 
. Until now, most of the immune evasion strategies of gammaherpesviruses have been studied in vitro
or in animal models 
. However, none has been investigated in the light of transmission mainly due to the lack of experimental transmission model. In this study, using in vivo
imaging, we observed that MHV-68 is genitally excreted after latency establishment in intranasally infected female mice (–, S1
). This allowed us to observe, for the first time, experimental transmission to naïve males after sexual contact (–).
The observation of vaginal shedding of MHV-68 is somewhat surprising as numerous people have been working on this model around the world for a long time without reporting such observations. However, several points can be mentioned. First, we have used a method of in vivo
imaging that was recently developed and which is particularly sensitive, allowing the detection of low levels of replicative virus 
. Secondly, we have followed the infections daily and during a long period, generally between 14 and 32 days post infection. To our knowledge, such following of the infectious process has never been reported. Interestingly, the two previous studies using in vivo
luciferase imaging of MHV-68 cycle suggested potential genital infection. Milho et al.
showed that the female genital tract is a site of virus replication after intraperitoneal infection 
and one of the mice used by Hwang et al.
displayed light emission in the genital region after intranasal infection and latency establishment (Hwang et al., , day 18 p.i. 
). The high frequency of genital signal observation in our study (~80% of the infected mice) could reflect particular experimental conditions. For example, a potential co-infection with another pathogen could favour MHV-68 genital excretion. Such synergic relation has been demonstrated for others herpesviruses, notably HSV-2 and human cytomegalovirus, with HIV-1 
. If such a pre-existent infection exists, the causal agent remains to be identified. Another explanation could be related to our housing facility which homes both females and males mice. This can be an important element as male pheromones can modulate estrous cycle in mice 
The observation of genital signal in females was dependent on the estrous cycle as ovariectomy nearly abolished the phenomenon and as estrogens supplementation restored it (). Steroid hormones influence susceptibility, replication as well as transmission of many viruses, including herpesviruses 
. Numerous studies have illustrated the influence of female sex hormones on both susceptibility and immune responses to sexually transmitted pathogens 
. Thus, estrogen and progesterone influence the susceptibility to genital herpes infection 
. However, in those cases the presence of progesterone increased susceptibility to HSV-2 
, whereas the presence of estrogen prevented or decreased the risk of HSV infection in the female genital tract 
. While these hormones can directly influence the sensitivity of the cells of the genital tract, they could also attenuate or modulate the innate 
and/or the adaptive 
immune response against the virus. For example, the abundance of antigen presenting cells, T cells and B cells has been shown to vary in uterus and vagina with the estrous cycle 
. Replication and shedding of MHV-68 could therefore be a consequence of impaired immune surveillance. On another hand, as some of these immune cells harbour MHV-68 latent infection 
, the transient observation of genital signal in females could reflect the variation of abundance of some particular cell types over time. In the future, these point will have to be tackled. Thus, the origin of vaginally excreted virions could be addressed by using cell-type specific Cre/Lox genetic labelling of MHV-68 to track the route of viral excretion in vivo
as it has recently been done to explore the host colonization pathway 
. Interestingly, the fact that the MHV-68-associated luciferase genital signal lasted for at most 3–4 days () could be linked to the cyclic remodelling of the epithelium observed during the estrous cycle. Indeed, infected cells are located in the superior layers of the vaginal epithelium () and could therefore be removed at each cycle.
Besides these indirect roles, steroid hormones have also been shown to directly induce herpesvirus reactivation from latency. Thus, 17β-estradiol promotes HSV-1 reactivation in latently infected neurons 
. Similarly, several studies have shown that dexamethasone, a synthetic corticosteroid, induces Bovine Herpesvirus-1 reactivation from latency either in vitro
or in vivo
in calves 
and rabbits 
. This has been associated with the induction of cellular transcription factors and/or signalling pathways that stimulate viral lytic genes expression and subsequent reactivation 
. In the present study, we did not observe any direct effect of estrogens on latently infected B-cells either in vitro
or ex vivo
(). However, we have no evidence that these cells mimic what happens in vivo
in the infected female genital tract. Further experiments are therefore required to identify the mechanism involved in estrogen-induced MHV-68 vaginal shedding.
In males, initial infection was localized in the superior layers of the penis epithelium and of the corpus cavernosum (). Infection then spreads to draining lymph nodes and spleen ( and ). Again, cell-type specific Cre/Lox genetic labelling of MHV-68 
will be helpful to track the route of viral infection after sexual transmission. As infectious virions were rarely detected in vaginal lavages although MHV-68 induced luciferase signal was frequent, we hypothesize that close contacts between genital organs of males and females are necessary to transmit infection. Indeed, the penis of the male mice is recovered of spines called filiform papilla. These structures could therefore induce abrasion of the vaginal epithelium and promote virus transmission. Interestingly, cells that were initially infected on penis were located around these filiform papilla (). Infection persisted at this site for at least three weeks (). The importance of this observation for MHV-68 epidemiology will therefore have to be tested. For example, it has recently been shown that male circumcision significantly reduces the incidence of HSV-2 and HIV-1 infection and the prevalence of HPV infection 
. Our results suggest that it could also be the case for some gammaherpesviruses.
Until now, we did not manage to establish experimental conditions to repeatedly transmit the virus from genitally infected males to naïve females. Human herpesvirus transmission generally occurs at a low rate even between close contacts 
. However, our results (Figure S7
) can only be considered suggestive of male to female MuHV-4 transmission. Important aspects of mouse behaviour such as scent marking may not be properly reproduced with conventional housing. We conclude that under the experimental conditions used, male to female transmission is possible but inefficient, certainly much less so that female to male transmission. The normal mode of male to female MuHV-4 transmission remains to be determined.
Sexual transmission constitutes an easy way of spread for a virus in natural populations of wild animals. This is particularly the case for rodents. Indeed, rodents live generally in small groups spread on a relatively large territory. Sexual contact could therefore be a relatively efficient route of transmission. Interestingly, Telfer et al.
showed that gammaherpesvirus (identified serologically as MuHV-4, though likely Wood Mouse Herpesvirus) infection in wood mice was more prevalent in heaviest, sexually active, males than in any other category of animal 
. The fact that the viral shedding in the female genital tract is linked to sexual cycle and more precisely to the period of estrus (high rates of estrogens) would be very beneficial for transmission as re-excretion would occur during the periods of female receptivity for mating. Sexual transmission has also been proposed for EBV and KSHV 
but is mainly important for HSV-1 and -2 
. The observation of MHV-68 sexual transmission from infected females to naïve males could therefore be particularly interesting in the general context of herpesvirus transmission.
Shedding of MHV-68 in the female genital tract could also have an effect on progeny. However, in contrast to what was reported by Stiglincova et al.
, we did not observe premature termination of pregnancy, reduced number of newborns, vertical transmission or transmission through milk of MHV-68 in mice (). We have no explanation for this discrepancy. However, mother to child transmission of human gammaherpesviruses, both transplacental or perinatal, seems also to be very limited 
The identification of a route of transmission for MHV-68 in mice opens new fundamental research perspectives. Thus, it will allow testing the importance of various immune evasion strategies, such as those based on the gp150 glycoprotein 
in the light of transmission. It will also be interesting to test if transmission requires latency establishment and reactivation of the virus or, conversely, if it is enhanced by immunosuppression through the use of drugs like Cyclosporine A 
or of depletion of specific cell types such as CD8 
. Finally, it will be possible to test antiviral and/or vaccinal strategies in the context of infection epidemiology.
Altogether, in this study we identified for the first time a genital excretion site of MHV-68 after latency establishment in intranasally infected female mice. This has allowed us to observe sexual transmission of the virus from infected females to naïve males. These results open new perspectives for the study of gammaherpesvirus in particular but also for the study of sexually transmitted infections in general.