Secondary Infections, Expanded Tissue Tropism, and Evidence for Malignant Potential in Immunocompromised Mice Infected with Mus musculus Papillomavirus 1 DNA and Virus

Abstract

Papillomavirus disease poses a special challenge to people with compromised immune systems. Appropriate models to study infections in these individuals are lacking. We report here the development of a model that will help to address these deficiencies. The MmuPV1 genome was synthesized and used successfully to produce virus from DNA infections in immunocompromised mice. In these early studies, we have demonstrated both primary and secondary infections, expanded tissue tropism, and extensive dysplasia.

TEXT

Papillomaviruses are double-stranded DNA tumor viruses of about 8 kb. These viruses are ubiquitous in nature, and 241 types have been identified to date in both humans and many other animal species (http://pave.niaid.nih.gov). A subset of human papillomaviruses is linked to cancers, most notably cervical cancer, but also cancers of the head and neck, skin, and other anogenital sites. Papillomaviruses require fully differentiating cells for the completion of the viral life cycle and thus are not easily studied in cell culture systems. For a review of these viruses, see reference 1. The cottontail rabbit papillomavirus (CRPV) in vivo model has proven to be very useful in our hands for the study of cutaneous papillomavirus disease (2–8). An in vivo system to study both cutaneous and mucosal disease, papillomavirus pathology, and disease progression in the immunocompromised patient would be a valuable addition to current preclinical models.

In 2011, Ingle et al. reported the isolation of a new mouse papillomavirus (9). This virus was found in lesions on the face of animals in a colony of NMR1-Foxn1^nu/Foxn1^nu mice in India and was subsequently isolated and passaged. The virus was initially labeled MusPV but is now known as MmuPV1. It was reported to be strictly cutaneous in its tropism. In a subsequent report, the sequence of the virus was reported (10), and GenBank accession number GI808564 was assigned. A variant found in a house mouse in Germany was reported in 2012 (HQ625439) (11). We investigated MmuPV1 in the Foxn1^nu/Foxn1^nu immunocompromised mice and in SCID SHO mice to determine its potential as an animal model for papilloma disease. Our goals were to generate infections via DNA, to isolate and passage virus that might result from these infections, and to confirm the tissue tropism of the virus. Sundberg et al. have shown that tissue tropism may be relaxed in immunocompromised animals (12), and we hypothesized that mucosal tissues might also be infected in immunocompromised mice.

We used the published sequence of the MmuPV1 clone (10) to generate two EcoRV-BamHI subclones (Blue Heron). These clones were combined to create the full-length genome cloned at BamHI in pUC19. This genome was used to infect SHO SCID mice (strain 474; Charles River) and athymic nude Fox mice (Harlan) in a manner similar to that used in our well-established rabbit model (13). We found that we were able to develop a robust DNA infection system for the mice by incorporating modest modifications to the protocol. Specifically, mice were anesthetized with a ketamine-xylazine mixture (100 mg/kg and 20 mg/kg body weight). Back skin was stretched to create a taut site. The skin was gently scarified with a scalpel blade to create lesions about 1 cm in diameter that resembled brush burns. Typically, two sites on the right back and two on the left back were prepared.

On day 1 or 24 h postscarification, animals were again anesthetized, the scarified sites were gently scratched, and 5 μg plasmid DNA, either digested with BamHI or undigested, was applied in a 50-μl volume using a pipette tip. The DNA was gently scraped into the skin with a tuberculin needle. Animals were placed in their cages and allowed to recover. They were examined weekly for lesions. Photos were taken each week. Lesions appeared in 2 to 3 weeks; these became less apparent over time. Infections using virus isolated from the lesions were carried out in the same manner.

The animal experiments were reviewed and approved by the Penn State University, College of Medicine, IACUC.

Several infected animals were followed over time, and secondary infections were found to develop in all animals. These lesions were detectable initially at about 3 to 4 months and by 6 months were generally widespread. Interestingly, the secondary infections displayed a broadened tissue tropism. To date, we have identified lesions in numerous anatomical regions that did not receive the initial inoculum such as the inner cheek, tail, face, lips, back, urethra, anus, and vagina. All lesions were positive for MmuPV1 DNA by in situ hybridization and most for capsid protein. Figures 1 to 6 demonstrating staining of vaginal/vulvar, back, lip, tail, anus, snout, face, inner cheek, and urethral tissues are representative. We hypothesize that these infections were generated by transfer from the primary sites to the secondary sites as a result of scratching, abrasions, and injuries inflicted by cage mates. Hyperplastic epidermis was noted in most sections with expansion of the basal and spinous layers and with numerous atypical cells. These cells contained abundant finely granular amphophilic cytoplasm and large vesicular nuclei. Both in situ hybridization and group-specific antigen colocalized with the atypical amphophilic cells. The vaginal-vulvar lesion (Fig. 1) showed definitive mucosal staining, as did the anal lesion (Fig. 4). The neck lesion (Fig. 2) displayed marked dysplasia, with basaloid cells comprising nearly the entire thickness of the epithelium and loss of progressive differentiation from basement membrane to lumen. There were numerous suprabasilar mitoses. The pathology of this lesion was reminiscent of cervical intraepithelial neoplasia (CIN). Marked dysplasia was also noted in the facial lesion (Fig. 5D and E), lending additional support for the idea of a possible malignant potential of the virus. Figure 6 shows inner cheek (panel A) and urethral (panel B) in situ positivities. This new rodent animal model has significant potential for the study of both mucosal and cutaneous papillomavirus diseases. It will be especially useful for the investigation of papillomavirus transmission in immunocompromised animals, with special relevance to patients with HIV/AIDS, and for interventions to minimize disease in these individuals. The mouse model will complement the in vivo rabbit model also in use in our laboratory and should accelerate research on many of the fundamental aspects of papillomavirus disease at both mucosal and cutaneous sites.

Fig 1.

Vaginal/vulvar lesion from a nude mouse. (A) Hematoxylin and eosin (H&E) staining (magnification, ×4). (B) In situ hybridization (ISH) (magnification, ×4). (C) H&E (magnification, ×20). Hyperplastic epithelium tissue with expansion of the basal and spinous layers and numerous atypical cells with abundant finely granular amphophilic cytoplasm and large vesicular nuclei is shown. (D) ISH of the area shown in panel C. Hybridization colocalizes with atypical amphophilic cells. (E) Immunohistochemistry (IHC) of the area in panels C and D showing capsid antigen staining (Griffonia simplicifolia agglutinin [GSA]). Nuclear staining is prominent in atypical amphophilic cells.

Fig 6.

(A) Inner cheek; in situ positivity. (B) Urethra; in situ positivity.

Fig 4.

Oblique section of haired skin and anocutaneous junction. (A) H&E. (B) In situ hybridization of area shown in panel A. Scattered positive nuclei are seen in mucosa.

Fig 2.

Neck lesion from a SCID SHO mouse. (A) H&E (magnification, ×4). A pedunculated papilloma arises from the haired skin. (B) ISH of same lesion (magnification, ×4). (C) Within the papilloma, there are multiple areas with loss of progressive differentiation from basement membrane to lumen. Here, basaloid cells comprise nearly the entire thickness of the epithelium, resembling CIN; there are frequent suprabasilar mitoses (magnification, ×20). (D) ISH of the same area (magnification, ×20). (E) Image of same area (magnification, ×40). Note frequent mitotic figures well apical of the basement membrane. Only the apical quarter of the epithelium shows any normal squamous differentiation.

Fig 5.

(A) Snout, H&E. (B) Snout, GSA. (C) Snout, in situ positivity. (D) Facial lesion of a nude mouse; in situ positivity. (E) The same facial lesion. Multiple mitotic figures are seen per high power field. (F) The same facial lesion. Triradiate chromosomes in this slide indicate genomic instability.

Fig 3.

(A) Haired skin (lip) lesion from a nude mouse. H&E staining shows hyperplastic epidermis tissue with tortuous rete ridges, expansion of the basal and spinous layers, and numerous atypical cells with scattered finely granular amphophilic cytoplasm and large vesicular nuclei. (B) ISH of the area shown in panel A. Hybridization is prominent at the rostral margin. (C) IHC of the area shown in panels A and B. Nuclear staining is prominent in atypical amphophilic cells. (D) Tail lesion, H&E. (E) Strong in situ positivity in tail lesion. (F) Strong GSA positivity in amphophilic cells of tail lesion is shown.

Additional studies are focused on the generation of primary infections at mucosal sites, the continued investigation of the malignant potential of the virus, and the investigation of the ability of the virus to produce productive infections in animals with intact immune systems. The observation that it is possible to generate virus from DNA provides a unique opportunity to study viral mutants. Altered genomes with targeted mutations will help to elucidate basic mechanisms of MmuPV1 biology as has been done by us and others for CRPV biology using mutant genomes of the cutaneous CRPV (5, 14, 15).