Mus musculus papillomavirus 1 is a key driver of skin cancer development upon immunosuppression

Epidemiological and experimental data implicate cutaneous human papillomavirus infection as co‐factor in the development of cutaneous squamous cell carcinomas (cSCCs), particularly in immunocompromised organ transplant recipients (OTRs). Herein, we established and characterized a skin cancer model, in which Mus musculus papillomavirus 1 (MmuPV1) infection caused cSCCs in cyclosporine A (CsA)‐treated mice, even in the absence of UV light. Development of cSCCs and their precursors were observed in 70% of MmuPV1‐infected, CsA‐treated mice on back as well as on tail skin. Immunosuppression by systemic CsA, but not UV‐B irradiation, was a prerequisite, as immunocompetent or UV‐B–irradiated mice did not develop skin malignancies after infection. In the virus‐driven cSCCs the MmuPV1‐E6/E7 oncogenes were abundantly expressed, and transcriptional activity and productive infection demonstrated. MmuPV1 infection induced the expression of phosphorylated H2AX, but not degradation of proapoptotic BAK in the cSCCs. Transfer of primary cells, established from a MmuPV1‐induced cSCC from back skin, into athymic nude mice gave rise to secondary cSCCs, which lacked viral DNA, demonstrating that maintenance of the malignant phenotype was virus independent. This papillomavirus‐induced skin cancer model opens future investigations into viral involvement, pathogenesis, and cancer surveillance, aiming at understanding and controlling the high incidence of skin cancer in OTRs.


| INTRODUC TI ON
Cutaneous squamous cell carcinoma (cSCC) represents the second most common type of skin cancer worldwide. A meta-analysis estimated the numbers of new cases in the United States White population in 2012 to be between 186 000 and 419 000, with increasing incidence. 1 As cSCCs are not specifically disclosed in national cancer registries, the exact numbers are unknown. 1,2 Organ transplant recipients (OTRs) are particularly vulnerable with up to 250-fold higher cSCC rates compared to the general population, and 40% of the afflicted develop skin malignancies within 15 years after transplantation. 3 Recently, an 812 per 100 000 person-years incidence was estimated in this population in the United States. 4 Aside from common risk factors, such as high cumulative ultraviolet (UV) exposure, genetic predisposition, and chronic inflammation, the cSCC rates in OTRs are influenced by intensity, duration, and type of immunosuppression. 5 Commonly used immunosuppressants include calcineurin inhibitors (cyclosporine A [CsA], tacrolimus), mammalian target of rapamycin inhibitors (rapamycin/sirolimus, everolimus), and antimetabolites (azathioprine, mycophenolate mofetil).
Oncogenic mucosal alpha-HPVs can persistently infect mucosa of the anogenital and upper aerodigestive tract to induce cancer at these sites. For keratinocyte cancers, the role of cutaneous HPVs, in particular of genus beta (beta-HPVs), has been a matter of debate for decades. [6][7][8][9] In the rare genodermatosis Epidermodysplasia verruciformis (EV), certain genus beta-HPVs are responsible for the development of cSCCs on sun-exposed areas, which contain high viral loads and actively transcribed E6/E7 oncogenes. 10 OTRs have a higher risk for human papillomavirus (HPV) infections due to suppressed cell-mediated immunity. Numerous epidemiologic studies have implicated beta-HPV skin infection as risk factor for cSCCs in this population, based on viral positivity and on seropositivity at the time of transplantation, [11][12][13] although others argue against a causal role, given that viral DNA is not present in all cancer cells and can also be found on skin surfaces in the general population. 8,9 The carcinogenic potential of beta-HPVs was further demonstrated in several animal models. For instance, transgenic (tg) beta-HPV8 mice, which expressed the viral genome in the skin, developed cSCCs without exposure to UV light or chemical carcinogens. 14 The carcinogenic synergism between the viral oncogenes and the major causative factor for skin cancer development, UV light, was shown in tg beta-HPV mice and in Mastomys natalensis or Mus musculus papillomavirus (MmuPV1)-infected rodents, which developed cSCCs after (long-term) UV irradiation. [15][16][17][18][19][20] In contrast to alpha-HPVs, which can immortalize cells and induce genomic instability after integration, beta-HPV oncogenes were shown to compromise DNA damage repair and inhibit apoptosis, 3 allowing accumulation of mutations in keratinocytes. Hence, beta-HPVs are thought to contribute as initial tumor promoter and progression factors to carcinogenesis rather than to maintenance of malignancy ("hit-and-run-mechanism"). 6,10 Herein, we addressed the questions, whether MmuPV1 infection per se can induce cSCC development in the absence of UV light and which role an intact immune system plays in preventing tumor development. Hence, we aimed at establishing and characterizing a laboratory mouse model, in which experimental MmuPV1 infection caused cSCCs in CsA-immunocompromised mice.

| MmuPV1-induced skin cancer model
Immunocompetent, female FVB/NCrl mice aged 4-5 weeks were obtained from Charles River Laboratories (Sulzfeld, Germany), immunodeficient NMRI-Foxn1 nu/nu mice bred in-house. A schematic representation of the experimental setup is shown in Figure S1. The back tumor area in mm 2 and the tail tumor length in mm were determined employing ImageJ software.

| Hematoxylin/eosin (HE) staining and immunohistochemistry (IHC)
Skin sections taken from the inoculation sites and the draining inguinal lymph nodes were stained with HE for subsequent evaluation by a pathologist blinded to the experimental conditions or used for IHC (Table S1). The MmuPV1-L1/L2-specific polyclonal immune serum was generated by immunization of a New Zealand White rabbit with MmuPV1-L1/L2 pseudovirions (PsVs) in a four-dose regimen at week 0-2-4-8 (Eurogentec, Seraing, Belgium). Images were digitalized using an Aperio slide scanner (Leica Biosystems, Nussloch, Germany). The numbers of immunopositive cells (per mm 2 ) and the immunopositive area (in %) present in the entire skin or lymph node tissue specimens were scored independently by three authors (S.D., K.S., and G.S.) employing ImageJ.

| RNA in situ hybridization
Single molecule RNA in situ hybridization for the MmuPV1-E6/ E7 mRNA was performed employing the RNAscope MusPV-E6-E7 probe (Advanced Cell Diagnostics, Newark, CA). RNA integrity was verified with the endogenous control probe Mm-PPIB, background staining evaluated using the negative control probe specific for the bacterial DapB gene. Brightfield images were acquired using the Aperio slide scanner (Leica Biosystems).

| Quantification of MmuPV1 genomic DNA and E1^E4 splice transcripts
Total RNA and genomic DNA were purified from the same tissue specimens using TRI reagent (Sigma-Aldrich). MmuPV1 DNA copy numbers were normalized to the housekeeping gene gamma-actin and quantified according to standard curves with known amounts of the religated viral genome. 21,22 MmuPV1-E1^E4 spliced transcripts, a marker for infectivity and viral transcription, were determined as described previously 21

| Particle-ELISA and PsV-neutralization assay (NA)
MmuPV1-specific antibodies in mouse sera (diluted 1:100) were assessed by particle-ELISA employing native virions as the antigen. 24 Neutralizing antibodies were determined by PsV-NA employing MmuPV1-SEAP PsVs composed of the MmuPV1 capsid proteins L1/ L2 and the reporter pYSEAP, encoding for secreted alkaline phosphatase. 25 The IC 50 titer is defined as the highest serum dilution showing at least 50% reduction in SEAP signal compared to the preimmune sera or in the absence of immune sera.

| Statistical analyses
Analyses were performed using GraphPad Prism 8. Data represent mean±standard deviation (SD). Differences between groups were analyzed using the nonparametric Kruskal-Wallis test with the Dunn's posttest. A P-value of <.05 was considered statistically significant.  Figure 1D).  Immunopositivity was restricted to tumorous tissues, whereas nontumorous skin areas were negative ( Figure S4A). UV-B irradiation generally induced low-level expression of γH2AX; however, in virusinfected mice, concomitant CsA administration allowed for 2.6fold elevated γH2AX expression compared to uninfected controls.

| MmuPV1 infection induces the expression of phosphorylated H2AX (γH2AX) in cSCCs
Tumorous and nontumorous skin tissues stained positive for γH2AX, the latter presumably attributable to the effects of UV-B ( Figure   S4A). Expression of γH2AX was not detected in the tissues of uninfected, non-CsA/non-UV-B-treated and uninfected, CsA-treated animals ( Figure 3A

| DISCUSS ION
The role of cutaneous genus beta-HPVs in the pathogenesis of keratinocyte carcinomas of the skin in the normal immunocompetent population has long been a matter of debate. In OTRs, whose viral infections are not adequately controlled due to iatrogenic immunosuppression, the highly elevated risk to develop cSCCs raises the possibility of a viral contribution to pathogenesis. The mouse model presented could contribute to site-specific susceptibility to tumorigenesis (skin of ear versus back and tail skin). 18,29 Intriguingly, cSCCs on back skin were 1.7 times more frequent in CsA-treated compared to CsA-/ UV-B-treated mice. Possible explanations are that in the latter group efficiency of infection was hampered by UV-B irradiation, which had already induced a certain degree of skin hardening prior to inoculation. We observed an increased thickening of the epidermis, which was most pronounced in mice of the CsA-/UV-B-followed by the CsA treatment group (data not shown), presumably due to different factors, such as the inoculation process, UV-B irradiation and as side effect of CsA, which could interfere with lesional outgrowth.  40 In contrast to a recently published MmuPV1 mouse model, 18 herein mice were exposed to cumulative rather than a single dose of UV-B, recapitulating the situation in humans, where lifelong sun exposure is the major causal factor in the development of keratinocyte skin malignancies. However, we showed that while the applied UV-B dose induced DNA damage in the irradiated tissues of mice, as demonstrated by the expression of γH2AX, a surrogate marker for DNA double-strand breaks and chromatin instability, interestingly, MmuPV1 per se seems to be capable to induce genomic instability, even without concomitant irradiation.
Inhibition of CPD repair by MmuPV1 might occur in the presence of CsA, which allows for elevated viral replication in the skin tissues, possibly favoring carcinogenesis. Another important mechanism in skin carcinogenesis is the inhibition of keratinocyte apoptosis in response to UV damage, mediated by the viral E6 oncoproteins. In this line, the E6 of several beta-HPVs were shown to inactivate BAK 41,42 and the survival of DNA-damaged cells promotes the progression of skin malignancies. Conversely, to beta-HPVs, we did not observe BAK degradation in skin tissues of infected mice. This could be due to differences in the susceptibility to UV-B light between mouse and human keratinocytes because of their evolutional history. It could be possible that neither the intrinsic antiapoptotic mechanisms in the keratinocytes nor those provided by MmuPV1 infection are enough to fully overcome the UV-B-induced kill of the mouse keratinocytes, whether normal or transformed. Since mice are evolutionarily not a subject to UV damage as humans, it is possible that MmuPV1 may not have evolved ways of counteracting it to the extent that skintropic HPVs have.
In summary, our data provide evidence that MmuPV1 has carcinogenic potential and skin infection causes cSCCs in mice upon systemic immunosuppression even in the absence of UV-B light.
Strikingly, the "hit-and-run" mechanism proposed for beta-HPVs ap- This allows for enhanced migratory capacity and altered responsiveness to apoptotic stimuli, facilitating cancer invasion and progression with metastatic expansion, even when the virus gets lost.
Hence, this study may instigate investigations regarding different options to prevent development or restrict (metastatic) invasion of cSCCs in OTRs, such as antiviral strategies (e.g., vaccination prior to organ transplantation, timely antiviral treatment), replacement of CsA by other immunosuppressants with less cancer-and/or virus-promoting properties, and strategies that inhibit and/or reverse epithelial-to-mesenchymal transition.

ACK N OWLED G M ENTS
We thank Felicia Spitzer, BSc., for assistance with the in vivo experiments and Saeed Shafti-Keramat for assistance with the PsV-NA.

D I SCLOS U R E
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.