Mouse papillomavirus infection persists in mucosal tissues of an immunocompetent mouse strain and progresses to cancer

Mouse papillomavirus has shown broad tissue tropism in nude mice. Previous studies have tested cutaneous infections in different immunocompromised and immunocompetent mouse strains. In the

current study, we examined mucosal infection in several immunocompetent and immunocompromised mouse strains. Viral DNA was monitored periodically by Q-PCR of lavage samples.

Immunohistochemistry and in situ hybridization were used to determine viral capsid protein and viral DNA respectively. All athymic nude mouse strains showed active infections at both

cutaneous and mucosal sites. Interestingly, NOD/SCID mice, which have a deficiency in T, B, and NK cells, showed minimal disease at cutaneous sites but developed persistent infection at the

mucosal sites including those of the anogenital region and the oral cavity. Three strains of immunocompetent mice supported mucosal infections. Infections of the lower genital tract in

heterozygous (immunocompetent) mice of the NU/J strain progressed to high grade dysplasia and to carcinoma in situ. Anti-MmuPV1 neutralizing antibodies were detected in the sera of all

immunocompetent animals. Our findings demonstrate that the mucosae may be the preferred sites for this virus in mice. The mouse model is expected to be a valuable model for the study of

mucosal papillomavirus disease, progression, and host immune control.

Human papillomaviruses (HPVs) are obligate factors for the development of cervical cancer, which is responsible for the deaths of 250,000 women worldwide each year1. In addition, these

viruses are increasingly implicated in head and neck cancers, anal cancers, and some skin cancers2,3. The three existing vaccines against HPVs are all prophylactic in nature and, while

effective in preventing new infections of important subsets of papillomaviruses, offer little help to the many people already infected with the virus4. In addition, the uptake of the

vaccines has been disappointingly low meaning that those people unable to clear the disease will continue to be at risk for the development of cancer over time5.

Papillomaviruses are species-specific and therefore it is not possible to study a HPV in an animal model. For many years, the cottontail rabbit papillomavirus (CRPV) model was the system of

choice for several laboratories including our own, in part because CRPV lesions progress to cancer over time6,7,8,9,10. Much about the immunology, molecular biology and malignant potential

of papillomaviruses has been learned using this system and it is anticipated that the model will continue to be a valuable resource in the years to come11. However, there are limitations to

the model. For one thing, most HPV-associated cancers are of mucosal origin and the CRPV lesions are cutaneous12. In addition, reagents for the rabbit are quite limited relative to those for

the most common laboratory animal, the mouse. Unfortunately, until 2011 when Ingle et al. reported finding Mouse papillomavirus 1(MmuPV1) in a colony of nude mice in India13, no mouse virus

had been identified that could infect a common laboratory strain14,15.

The discovery of MmuPV1 intrigued the papillomavirus research community although enthusiasm was tempered by the early report that the virus was strictly cutaneous in nature13. A number of

laboratories began to study the virus with most work assuming cutaneous tropism16,17,18,19,20. Sundberg et al. looked at strain and site differences and noted the formation of

trichoblastomas on the dorsal skin21. Handisurya et al. also looked at strain differences and T cell involvement in clearance17. They showed that T cell depletion via anti-CD3 antibody

rendered immunocompetent animals permissive for cutaneous infections. Wang et al. studied immunologic control of the virus and noted cutaneous viral persistence in one strain of

immunocompetent mice, the hairless SKH-118. This work was followed up by that of Jiang et al. in a paper in which the utility of this animal was demonstrated for the study of clearance of PV

disease19. Uberoi et al. reported that systemic immunosuppression induced by a high dose of UVB promoted cancer development in MmuPV1 infected ear skin of FVB/NJ immunocompetent mice20.

In our laboratory, we found and reported on the first mucosal infections with the MmuPV1 virus and have definitively shown that oral, vaginal, anal and penile tissues are all highly

susceptible to the virus, putting to rest the idea that the virus is restricted to cutaneous sites22,23,24,25. These observations were further confirmed by studies in another group21,26. In

addition to the active anogenital infections and dysplasia in these animals, we have also observed that the single circumvallate papilla of the mouse tongue is uniquely susceptible to

infection by the virus. This site is comparable to back of the tongue sites so commonly found in oral papillomavirus-associated cancers in humans, for which an increasing incidence is

reported in younger male Caucasians27. We anticipate that this new mouse model will be of use in studying progression of oral papillomavirus disease.

Active infections can be readily established in immunocompromised animals at mucosal sites22,23,24. To study viral-host interactions, an immunocompetent mouse strain with intact immune

response is desirable. Different immunocompetent mouse strains including C57BL/6, hairless SKH-1, and FVB/NJ have revealed differences in cutaneous site susceptibilities in previous

studies17,18,19,20,21,28. We were interested in following up on these observations and in determining whether we could identify a mucosally susceptible immunocompetent strain as well. The

experiments reported in this manuscript were designed to expand on our observations of mucosal MmuPV1 infections by investigating several different mouse strains. Among the animals selected,

we first tested immunocompetent C57BL/6 mice and SKH-1 hairless elite mice. They showed strong immune responses to MmuPV1 infection and cleared the infection quickly. We then decided to

investigate the heterozygous siblings of the homozygous immunocompromised NU/J, Hsd: NU and B6 animals, which we had previously shown to be permissive for viral infections17,21. These

heterozygotes are immunocompetent. To follow the infections longitudinally, we monitor viral DNA copy numbers via QPCR analysis of DNA isolated from lavage samples, which we collect

regularly over time23. This allows us to use a small number of animals to obtain robust data. It obviates the need for large numbers of animals to be sacrificed over time while still

allowing for extensive data collection. The lavages have proven to be a powerful tool and have been used to study vaginal, penile, anal and oral infections23. We found that the NU/J, Hsd: NU

and C57BL/6 heterozygotes were permissive for infection at the anogenital tissues. Furthermore all NU/J mice proceeded to develop carcinoma in situ in the vaginal infected tissues by 7.5

months post infection. Interestingly, the cutaneous tissues of these same mice showed only subclinical infections and were not permissive for papillomavirus lesion development. These

important findings provide opportunities for the study of mucosal papillomavirus infections and malignancies under the influence of an intact immune system and in a biologically relevant

site, the vaginal canal. They further help to cement the utility of this new MmuPV1 model for the study of papilloma diseases, progression, immunological response and viral-host interaction.

Previous studies have demonstrated that adaptive immunity is sufficient to eliminate MmuPV1 infection in outbred hairless euthymic SKH1-Elite (Crl: SKH1-Hrhr) and C57BL/6J immunocompetent

mice at cutaneous sites19,20. Whether or not mucosal sites, including the lower genital tract, were susceptible to MmuPV1 infection was not tested. Four SKH1-Elite and eight inbred C57BL/6J

mice were infected vaginally with MmuPV1 (Table 1).

Infection was followed by the detection of viral DNA in vaginal lavage, a tool that has proven to be very robust in our hands23. Viral DNA was detected in the lavages at week two

post-infection in both strains but was undetectable at week four post-infection (Supplementary Fig. 1A and B). We detected anti-viral antibodies in serum samples from these infected animals

(Supplementary Fig. 1C and D) indicating that transient mucosal infections probably occurred26. Serum conversion was also reported in cutaneously- infected immunocompetent mice19. These

findings suggest that both immunocompetent mouse strains are susceptible to MmuPV1 infection at vaginal mucosae, and that the infections rapidly clear.

Eight C57LB/6J mice were used to test the ability of these immunocompetent mice to sustain oral infection. Viral DNA was detected in the DNA from oral lavages by QPCR at week three

post-infection in six of the mice and became undetectable after week four (Supplementary Fig. 2A). Two mice with the highest viral DNA copy numbers were sacrificed for histological analysis.

In neither of these animals was viral DNA detected by in situ hybridization. Serum samples from all animals were harvested for antibody detection at week seven post-infection. All eight

orally-infected B6 mice generated detectable antibodies against the mouse papillomavirus (Supplementary Fig. 2B). No viral DNA was detected at the infected sites by in situ hybridization

analysis in any of the animals at week seven post infection, supporting the rapid clearance of disease.

In our previous studies of immunocompromised mice, we have found anal sites to be somewhat less permissive to MmuPV1 infection than vaginal and oral sites23. Previous studies have shown that

CD4 and CD8 T cells are crucial to the elimination of MmuPV1 in immunocompetent mice at cutaneous sites17,18. We, therefore, decided to deplete CD4 and CD8 cells in three C57BL/6J mice by

using anti-mouse CD4 (Clone GK1.5) and anti-mouse CD8 (clone 2.43 against CD8a) for seven weeks following anal viral infections. The CD4 (Supplementary Fig. 3A) and CD8 (Supplementary Fig. 

3B) levels were evaluated after the termination of the experiment by one-color flow cytometry analysis29. Although significantly lower levels of CD8 were found in the depleted animals

(Supplementary Fig. 3C, P  0.05, unpaired Student T-test). Anal infections were detected at week five post-infection (Supplementary Fig. 3D and E, P