Cloning and characterization of a papillomavirus associated with papillomas and carcinomas in the European harvest mouse (Micromys minutus)

Rodent Papillomaviruses

Preclinical infection model systems are extremely valuable tools to aid in our understanding of Human Papillomavirus (HPV) biology, disease progression, prevention, and treatments. In this context, rodent papillomaviruses and their respective infection models are useful tools but remain underutilized resources in the field of papillomavirus biology. Two rodent papillomaviruses, MnPV1, which infects the Mastomys species of multimammate rats, and MmuPV1, which infects laboratory mice, are currently the most studied rodent PVs. Both of these viruses cause malignancy in the skin and can provide attractive infection models to study the lesser understood cutaneous papillomaviruses that have been frequently associated with HPV-related skin cancers. Of these, MmuPV1 is the first reported rodent papillomavirus that can naturally infect the laboratory strain of mice. MmuPV1 is an attractive model virus to study papillomavirus pathogenesis because of the ubiquitous availability of lab mice and the fact that this mouse species is genetically modifiable. In this review, we have summarized the knowledge we have gained about PV biology from the study of rodent papillomaviruses and point out the remaining gaps that can provide new research opportunities.

Assessing the Diversity of Rodent-Borne Viruses: Exploring of High-Throughput Sequencing and Classical Amplification/Sequencing Approaches

Rodents are distributed throughout the world and interact with humans in many ways. They provide vital ecosystem services, some species are useful models in biomedical research and some are held as pet animals. However, many rodent species can have adverse effects such as damage to crops and stored produce, and they are of health concern because of the transmission of pathogens to humans and livestock. The first rodent viruses were discovered by isolation approaches and resulted in break-through knowledge in immunology, molecular and cell biology, and cancer research. In addition to rodent-specific viruses, rodent-borne viruses are causing a large number of zoonotic diseases. Most prominent examples are reemerging outbreaks of human hemorrhagic fever disease cases caused by arena- and hantaviruses. In addition, rodents are reservoirs for vector-borne pathogens, such as tick-borne encephalitis virus and Borrelia spp., and may carry human pathogenic agents, but likely are not involved in their transmission to human. In our days, next-generation sequencing or high-throughput sequencing (HTS) is revolutionizing the speed of the discovery of novel viruses, but other molecular approaches, such as generic RT-PCR/PCR and rolling circle amplification techniques, contribute significantly to the rapidly ongoing process. However, the current knowledge still represents only the tip of the iceberg, when comparing the known human viruses to those known for rodents, the mammalian taxon with the largest species number. The diagnostic potential of HTS-based metagenomic approaches is illustrated by their use in the discovery and complete genome determination of novel borna- and adenoviruses as causative disease agents in squirrels. In conclusion, HTS, in combination with conventional RT-PCR/PCR-based approaches, resulted in a drastically increased knowledge of the diversity of rodent viruses. Future improvements of the used workflows, including bioinformatics analysis, will further enhance our knowledge and preparedness in case of the emergence of novel viruses. Classical virological and additional molecular approaches are needed for genome annotation and functional characterization of novel viruses, discovered by these technologies, and evaluation of their zoonotic potential.

Recent advances in preclinical model systems for papillomaviruses

Preclinical model systems to study multiple features of the papillomavirus life cycle have greatly aided our understanding of Human Papillomavirus (HPV) biology, disease progression and treatments. The challenge to studying HPV in hosts is that HPV along with most PVs are both species and tissue restricted. Thus, fundamental properties of HPV viral proteins can be assessed in specialized cell culture systems but host responses that involve innate immunity and host restriction factors requires preclinical surrogate models. Fortunately, there are several well-characterized and new animal models of papillomavirus infections that are available to the PV research community. Old models that continue to have value include canine, bovine and rabbit PV models and new rodent models are in place to better assess host-virus interactions. Questions arise as to the strengths and weaknesses of animal PV models for HPV disease and how accurately these preclinical models predict malignant progression, vaccine efficacy and therapeutic control of HPV-associated disease. In this review, we examine current preclinical models and highlight the strengths and weaknesses of the various models as well as provide an update on new opportunities to study the numerous unknowns that persist in the HPV research field.

In vitro and Animal Models for Antiviral Therapy in Papillomavirus Infections

The need for antiviral therapies for papillomavirus infections is well recognized but the difficulties of reproducing the infectious cycle of papillomaviruses in vitro has hindered our understanding of virus-cell interactions and the regulation of viral gene expression during permissive growth. Recent advances in understanding the temporal expression and function of papillomavirus proteins has enabled consideration of a targeted approach to papillomavirus chemotherapy and in particular the inhibition of viral replication by targeting the E1 and E2 proteins. There are in vitro culture systems available for the screening of new chemotherapeutic agents, since significant advances have been made with culture systems which promote epithelial differentiation in vitro. However, to date, there are no published data which show that virions generated in vitro can infect keratinocytes and initiate another round of replication in vitro. In vivo animal models are therefore necessary to assess the efficacy of antivirals in preventing and treating viral infection, particularly for the low-risk genital viruses which are on the whole refractory to culture in vitro. Although papillomaviruses affect a wide variety of hosts in a species-specific manner, the animals most useful for modelling papillomavirus infections include the rabbit, ox, mouse, dog, horse, primate and sheep. The ideal animal model should be widely available, easy to house and handle, be large enough to allow for adequate tissue sampling, develop lesions on anatomical sites comparable with those in human diseases and these lesions should be readily accessible for monitoring and ideally should yield large amounts of infectious virus particles for use in both in vivo and in vitro studies. The relative merits of the various papillomavirus animal models available in relation to these criteria are discussed.

Intranuclear Inclusions in Renal Tubular Epithelium in Immunodeficient Mice Stain with Antibodies for Bovine Papillomavirus Type 1 L1 Protein

The kidneys from six immunodeficient mice examined by Cerberus Sciences and the Animal Resources Centre, displayed karyomegaly with pale eosinophilic, intranuclear inclusions upon histopathological examination. Electron microscopy performed on kidney tissue from 5/6 mice demonstrated margination of the chromatin in large nuclei. Laboratory tests were used to detect nucleic acid of papillomaviruses, polyomaviruses, circoviruses and anelloviruses (4/6 mice), a specific PCR was used to detect murine polyomavirus (1/6), and a panel of serological tests was used to detect seroconversion to major murine pathogens (1/6). All molecular and serological tests were negative. Immunohistochemistry using polyclonal anti-bovine papillomavirus type 1 (BPV-1) L1 antibody, Camvir monoclonal anti-papillomavirus antibody (directed against the seven amino acids GFGAMDF found in human papillomavirus (HPV) 16 L1 protein), a commercially available mixture of two monoclonal antibodies, anti-BPV-1 L1/1H8 + Camvir antibodies, and a monoclonal anti-Hsc70 antibody revealed specific, positive staining of murine renal tubular epithelial intranuclear inclusions in 6/6 mice using the anti-BPV-1 L1 containing antibodies only. Methyl pyronin green, PAS and Feulgen histochemical reactions revealed that the intranuclear inclusions did not consist of RNA, DNA or carbohydrate. An immunohistochemical method now exists that can be used to confirm and evaluate suspected cases of murine inclusion body nephropathy.

Characterization of Mus musculus papillomavirus 1 infection in situ reveals an unusual pattern of late gene expression and capsid protein localization

Full-length genomic DNA of the recently identified laboratory mouse papillomavirus 1 (MusPV1) was synthesized in vitro and was used to establish and characterize a mouse model of papillomavirus pathobiology. MusPV1 DNA, whether naked or encapsidated by MusPV1 or human papillomavirus 16 (HPV 16) capsids, efficiently induced the outgrowth of papillomas as early as 3 weeks after application to abraded skin on the muzzles and tails of athymic NCr nude mice. High concentrations of virions were extracted from homogenized papillomatous tissues and were serially passaged for >10 generations. Neutralization by L1 antisera confirmed that infectious transmission was capsid mediated. Unexpectedly, the skin of the murine back was much less susceptible to virion-induced papillomas than the muzzle or tail. Although reporter pseudovirions readily transduced the skin of the back, infection with native MusPV1 resulted in less viral genome amplification and gene expression on the back, including reduced expression of the L1 protein and very low expression of the L2 protein, results that imply skin region-specific control of postentry aspects of the viral life cycle. Unexpectedly, L1 protein on the back was predominantly cytoplasmic, while on the tail the abundant L1 was cytoplasmic in the lower epithelial layers and nuclear in the upper layers. Nuclear localization of L1 occurred only in cells that coexpressed the minor capsid protein, L2. The pattern of L1 protein staining in the infected epithelium suggests that L1 expression occurs earlier in the MusPV1 life cycle than in the life cycle of high-risk HPV and that virion assembly is regulated by a previously undescribed mechanism.

The biology of papillomavirus latency

The presence of viral DNA in the absence of disease has suggested that papillomaviruses, like many other viruses, can exist as latent infections in the skin or other epithelial sites. In animal models, where detailed investigation has been carried out, papillomavirus DNA can be found at sites of previous infection following immune regression, with the site of latent infection being the epithelial basal layer. Such studies suggest that immune surveillance can restrict viral gene expression in the basal and parabasal layers without efficiently suppressing viral genome replication, most probably through the action of memory T-cells in the skin or dermis. Although gradual papillomavirus genome loss appears to occur over time at latent sites, immunosuppression can arrest this, and can lead to an elevation in viral genome copy number in experimental systems. In addition to immune-mediated latency, it appears that a similar situation can be achieved following infection at low virus titres and/or infection at epithelial sites where the virus life cycle is not properly supported. Such silent of asymptomatic infections do not necessarily involve the host immune system and may be controlled by different mechanisms. It appears that virus reactivation can be triggered by mechanical irritation, wounding or by UV irradiation which changes the local environment. Although the duration of papillomavirus latency in humans is not yet known, it is likely that some of the basic principles will resemble those elucidated in these model systems, and that persistence in the absence of disease may be the default outcome for at least some period of time following regression.

Isolation of three novel rat and mouse papillomaviruses and their genomic characterization

Despite a growing knowledge about the biological diversity of papillomaviruses (PV), only little is known about non-human PV in general and about PV mice models in particular. We cloned and sequenced the complete genomes of two novel PV types from the Norway rat (Rattus norvegicus; RnPV2) and the wood mouse (Apodemus sylvaticus; AsPV1) as well as a novel variant of the recently described MmuPV1 (originally designated as MusPV) from a house mouse (Mus musculus; MmuPV1 variant). In addition, we conducted phylogenetic analyses using a systematically representative set of 79 PV types, including the novel sequences. As inferred from concatenated amino acid sequences of six proteins, MmuPV1 variant and AsPV1 nested within the Beta+Xi-PV super taxon as members of the Pi-PV. RnPV2 is a member of the Iota-PV that has a distant phylogenetic position from Pi-PV. The phylogenetic results support a complex scenario of PV diversification driven by different evolutionary forces including co-divergence with hosts and adaptive radiations to new environments. PV types particularly isolated from mice and rats are the basis for new animal models, which are valuable to study PV induced tumors and new treatment options.

Molecular diagnosis of a laboratory mouse papillomavirus (MusPV)

MusPV, a novel papillomavirus (PV) that naturally infects laboratory mice, was isolated and characterized from a colony of NMRI-Foxn1(nu)/Foxn1(nu) (nude) mice in India. Because MusPV may have been missed during routine pathogen screening of mice in colonies worldwide, a variety of detection methods are described to detect MusPV. The clinical and histologic lesions of productive MusPV infections fit PV-associated features, including papillomas, koilocytes within the stratum granulosum of the hyperplastic/acanthotic papillomatous epithelium, and the presence of intranuclear virus particles in koilocytotic cells visualized by electron microscopy. Antiserum against disrupted PV virions, isolated from another species (canine), identified conserved viral antigens in productively infected cells by immunohistochemistry. A rolling circle technique was used to amplify viral circular DNAs followed by endonuclease restriction enzyme digestion to determine the correct size of PV DNA. Consensus PV degenerative primers, My09/11, commonly used to detect many different types of PVs by polymerase chain reaction (PCR), particularly mucosotropic HPVs, also identified MusPV and all rodent PVs tested. Since there was one nucleotide mismatch between the My09/11 primer set and the MusPV template, a new primer set, MusPV-My09/11, was designed to specifically detect MusPV in latent infections and spontaneous MusPV-induced papillomas. Southern blot analysis verified the presence of full size PV DNA in infected tissues. Virus-like particles (VLPs), generated from MusPV L1 genes, provided a substrate for serological testing of naturally and experimentally infected mice. In summary, a series of diagnostic assays were developed and validated to detect MusPV infection in skin tumors and serological response in laboratory mice.

Genomic analysis of the first laboratory-mouse papillomavirus

A papillomavirus (PV) that naturally infects laboratory mice will provide an extremely valuable tool for PV research. We describe here the isolation, cloning and molecular analysis of the first novel laboratory-mouse PV, designated MusPV. This agent, recently identified in the tissues from florid and asymmetrical papillomas on the face of nude mice (NMRI-Foxn1(nu)/Foxn1(nu)), was demonstrated to be transmissible to immunocompetent mice (Ingle et al., 2010). The MusPV genome is 7510 bp in length, is organized similarly to those of other PVs and has at least seven ORFs (E1, E2, E4, E6, E7, L1 and L2). Phylogenetic analysis indicates that MusPV belongs to the π genus together with four other rodent PVs (McPV2, MaPV1, MmiPV and RnPV1). Of the rodent PVs, MusPV appears most closely related to Mastomys coucha PV (McPV2), with 65 % genomic homogeneity and 80 % L1 amino acid similarity. Rodent PVs, except for MnPV1, do not contain any identifiable retinoblastoma protein (RB) binding sites. MusPV has one putative RB-binding site on the E6 protein but not on the E7 protein. Non-coding regions (NCRs) of PVs maintain multiple binding sites for transcription factors (TFs). The NCR of MusPV has numerous sites for TF binding, of which at least 13 TFs are common to all PVs in the π genus. MusPV provides a potentially valuable, novel mouse model to study mechanisms of infection, oncology and novel preventive and therapeutic approaches in mice that can be translated to diseases caused by human PVs.

Novel laboratory mouse papillomavirus (MusPV) infection

Most papillomaviruses (PVs) are oncogenic. There are at least 100 different human PVs and 65 nonhuman vertebrate hosts, including wild rodents, which have species-specific PV infections. Florid papillomatosis arose in a colony of NMRI-Foxn1(nu)/Foxn1(nu) (nude) mice at the Advanced Centre for Treatment Research and Education in Cancer in India. Lesions appeared at the mucocutaneous junctions of the nose and mouth. Histologically, lesions were classical papillomas with epidermal hyperplasia on thin fibrovascular stalks in a verrucous pattern. Koilocytotic cells were observed in the stratum granulosum of the papillomatous lesions. Immunohistochemically, these abnormal cells were positive for PV group-specific antigens. With transmission electron microscopy, virus particles were observed in crystalline intranuclear inclusions within keratinocytes. The presence of a mouse PV, designated MusPV, was confirmed by amplification of PV DNA with degenerative primers specific for PVs. This report is the first of a PV and its related disease in laboratory mice.

Complete genomic characterization of a murine papillomavirus isolated from papillomatous lesions of a European harvest mouse (Micromys minutus)

The papillomaviruses form a large group of species-specific pathogens that cause epithelial proliferations in a wide spectrum of animal hosts. Previous reports demonstrated a relatively high frequency of a variety of skin lesions in captive European harvest mice. The Micromys minutus papillomavirus (MmPV) was isolated from one of these lesions found on a captive European harvest mouse in a regional zoo in Chicago. In this study we present the entire genomic sequence of MmPV. The MmPV genome is organized into the seven classical papillomaviral open reading frames. Phylogenetic analysis places MmPV together with a papillomavirus (PV) isolated from a Syrian golden Hamster (HaOPV) in the genus Pipapillomavirus. The similar clustering pattern of the MmPV–HaOPV pair and their rodent hosts support the hypothesis of papillomaviral and host co-phylogenetic descent. The availability of the complete genomic sequence of a mouse PV should allow researchers to use MmPV as a model for PV carcinogenesis.

Isolation and cloning of a papillomavirus from a North American porcupine by using multiply primed rolling-circle amplification: the Erethizon dorsatum papillomavirus type 1

The complete genome of a novel papillomavirus was isolated from a cutaneous papillomatous lesion of a North American porcupine (Erethizon dorsatum) using multiply primed rolling-circle amplification (RCA). The nucleotide sequence, genome organization, and phylogenetic position of the Erethizon dorsatum papillomavirus type 1 (EdPV-1) were determined. EdPV-1 is only distantly related to other benign cutaneous papillomavirus sequences and is the first member of the novel Sigma papillomavirus genus.

Flow cytometric quantitation of the protective efficacy of dendritic cell based vaccines in a human papillomavirus type 16 murine challenge model

A murine model for the assessment of protective immunity to human papillomavirus (HPV) type 16, a virus that does not naturally infect mice, is described. In this system, protection was tested following intranasal challenge of mice with a recombinant vaccinia virus expressing both the selected HPV antigen and a beta-galactosidase (beta-gal) reporter. The extent of viral infectivity was determined by measuring beta-gal positive lung cells using flow cytometry. The efficacy of this model to measure protective immunity was evaluated by priming mice with the beta-gal vaccinia virus then challenging the mice with the same virus. Vaccinia primed mice had negligible numbers of beta-gal positive cells in the lung 5 days following viral challenge indicating protection, whereas around 50% of cells were infected in immunologically naive, challenged mice. The protective efficacy of two dendritic cell vaccines for HPV16 was measured in this model. Both vaccines provided some protection to subsequent viral challenge, compared with their controls. Although this protection model was applied to HPV in this study, it may also have broad application to other viruses that do not infect mice naturally.

Therapeutic evaluation of compounds in the SCID-RA papillomavirus model

A previous study by Kreider (Kreider et al., 1979) indicated that rabbit skin, which had been transplanted to immunodeficient nude mice, could be successfully infected with cottontail rabbit papillomavirus (CRPV). We have extended this observation in developing a rodent model for evaluation of compounds for activity against the papillomaviruses. In this model (called the SCID-Ra model), rabbit ear skin is transplanted to the dorsum of SCID mice and allowed to heal for 3 weeks. Infection with CRPV by scarification leads to the growth of warty lesions within 2 3 weeks in >95% of the animals. Topical and/or systemic therapy can be initiated at various times post infection (PI). Weekly lesion scores are recorded and compounds are evaluated for their ability to suppress wart growth when compared to untreated control mice. Ribavirin, which has had a suppressive effect both in the clinic for the treatment of respiratory papillomatosis and on the growth of warts in the rabbit back model, was evaluated and showed significant anti-proliferative activity with oral dosing. Both antiviral and antiproliferative compounds including podophyllin and 5-fluorouracil, which have been used clinically for the treatment of human papillomavirus (HPV) infections, were evaluated in this model. The anti-mitotic compound, Navelbine (vinorelbine tartrate), which is used for the treatment of non-small cell lung carcinoma was evaluated in this system and showed significant inhibition of wart growth with somewhat less topical cytotoxicity when compared to podophyllotoxin.

Presence of human papilloma virus type 16 DNA sequences in human nonmelanoma skin cancers

The presence of human papillomaviruses (HPV) has been shown to be associated with the development and progression of invasive cancers of the genital tract, skin, and head and neck. In this study we analyzed 37 human nonmelanoma skin cancers (21 squamous cell carcinomas and 16 basal cell carcinomas) for the presence of HPV sequences. The polymerase chain reaction (PCR) was employed using primers designed to amplify DNA encoding the E6-E7 region of HPV types 6b/11, 16, and 18. HPV type 6b/11 and 18 sequences were absent from the DNA of all 37 tumors examined. However, HPV type 16 sequences were detected in four of 21 squamous cell carcinomas (SCC) (19%) and three of 16 basal cell carcinomas (BCC) (19%) as indicated on agarose gel electrophoresis by the presence of a single specific DNA band of predicted length. Furthermore, HPV type 16 sequences were absent in the DNA of normal skin from these seven skin cancer patients. The presence of HPV type 16 sequences in the seven skin tumors was confirmed by dot blot hybridization of PCR-amplified material to 32P-labeled HPV type 16, but not to HPV type 6/11 or 18-specific probes. These data indicate that HPV type 16, but not 6b/11 or 18, is associated with development of some human nonmelanoma skin cancers.

Viruses and oral cancer

Oral cancer is a disease with a complex etiology. There is evidence for important roles of smoking, drinking, and genetic susceptibility, as well as strong indications that DNA viruses could be involved. The herpes simplex virus type 1 has been associated with oral cancer by serological studies, and animal models and in vitro systems have demonstrated that it is capable of inducing oral cancer. Papillomaviruses are found in many oral cancers and are also capable of transforming cells to a malignant phenotype. However, both virus groups depend on co-factors for their carcinogenic effects. Future research on viruses and oral cancer is expected to clarify the role of these viruses, and this will lead to improvements in diagnosis and treatment of the disease.

Enhancement of u.v.-induced skin carcinogenesis in the hairless mouse by inoculation with cell-free extracts of skin tumours

The presence of papillomaviral-like DNA has been described in ultraviolet light-induced tumours in the skin of the hairless mouse (14). Here we describe the effects of the inoculation of cell-free extracts of ultraviolet light-induced tumours into the scarified skin of normal hairless mice, prior to exposure of the mice to a cumulative carcinogenic dose of ultraviolet light. Extracts from papillomas or squamous cell carcinomas enhanced the susceptibility of the inoculated mice to ultraviolet light-induced tumorigenesis, if the extracts contained papillomaviral DNA sequences detected by cross-hybridization with Mastomys natalensis papillomaviral DNA. The recipient mice developed a greater tumour incidence, tumour yield, tumour diameter and degree of malignancy.

Papillomas and carcinomas associated with a papillomavirus in European harvest mice (Micromys minutus)

Papillomaviruses, group-specific papillomavirus antigens, or extrachromosomal papillomavirus DNA were detected in cutaneous, mucocutaneous, and pulmonary tumors affecting a colony of European harvest mice (Micromys minutus). Skin lesions were classified as acanthomatous hyperplasia, epidermal inclusion cysts, squamous papillomas, inverted papillomas, trichoepitheliomas, and sebaceous carcinomas. Cutaneous horns (hyperkeratotic papillomas) were on mucocutaneous junctions of one animal. One mouse, with a cutaneous sebaceous carcinoma, had multiple pulmonary keratinaceous cysts. Papillomavirus antigens, detected by the avidin-biotin technique, were in 20 of 31 lesions tested. In contrast, by Southern blot hybridization all 28 lesions tested contained papillomavirus DNA. Papillomavirus DNA was demonstrated in two of ten benign cutaneous lesions by in situ hybridization.