Monitoring mouse papillomavirus-associated cancer development using longitudinal Pap smear screening

Abstract

A substantial percentage of the population remains at risk for cervical cancer due to pre-existing human papillomavirus (HPV) infections, despite prophylactic vaccines. Early diagnosis and treatment are crucial for better disease outcomes. The development of new treatments heavily relies on suitable preclinical model systems. Recently, we established a mouse papillomavirus (MmuPV1) model that is relevant to HPV genital pathogenesis. In the current study, we validated the use of Papanicolaou (Pap) smears, a valuable early diagnostic tool for detecting HPV cervical cancer, to monitor disease progression in the MmuPV1 mouse model. Biweekly cervicovaginal swabs were collected from the MmuPV1-infected mice for viral DNA quantitation and cytology assessment. The Pap smear slides were evaluated for signs of epithelial cell abnormalities using the 2014 Bethesda system criteria. Tissues from the infected mice were harvested at various times post-viral infection for additional histological and virological assays. Over time, increased viral replication was consistent with higher levels of viral DNA, and it coincided with an uptick in epithelial cell abnormalities with higher severity scores noted as early as 10 weeks after viral infection. The cytological results also correlated with the histological evaluation of tissues harvested simultaneously. Both immunocompromised and immunocompetent mice with squamous cell carcinoma (SCC) cytology also developed vaginal SCCs. Notably, samples from the MmuPV1-infected mice exhibited similar cellular abnormalities compared to the corresponding human samples at similar disease stages. Hence, Pap smear screening proves to be an effective tool for the longitudinal monitoring of disease progression in the MmuPV1 mouse model.

Importance: Papanicolaou (Pap) smear has saved millions of women's lives as a valuable early screening tool for detecting human papillomavirus (HPV) cervical precancers and cancer. However, more than 200,000 women in the United States alone remain at risk for cervical cancer due to pre-existing HPV infection-induced precancers, as there are currently no effective treatments for HPV-associated precancers and cancers other than invasive procedures including a loop electrosurgical excision procedure (LEEP) to remove abnormal tissues. In the current study, we validated the use of Pap smears to monitor disease progression in our recently established mouse papillomavirus model. To the best of our knowledge, this is the first study that provides compelling evidence of applying Pap smears from cervicovaginal swabs to monitor disease progression in mice. This HPV-relevant cytology assay will enable us to develop and test novel antiviral and anti-tumor therapies using this model to eliminate HPV-associated diseases and cancers.

Keywords: 2014 Bethesda system; HSIL; LSIL; Pap smear; RNAscope; cytology; immunohistochemistry (IHC); in situ hybridization; longitudinal; lower genital infection; mouse model; qPCR; squamous cell carcinoma; the mouse papillomavirus (MmuPV1); transmission electron microscope (TEM); viral copy number.

Fig 1

Pap smear slide preparation using a conventional swab. Single-used swabs (A, Puritan PurFlock Ultra Flocked Swabs) were dipped into sterile 0.9% NaCl and inserted into the vaginal tract of infected mice. The cells on the swabs were spread onto a superfrost white slide (B) by rotating the swabs on the slide in a circle manner. The cells will be fixed by spraying (C) the SlideRight cytology fixative (D, fishersci) from 6 inches above. The fixed slides were stored in a slide box at 4°C before staining with the standard Pap smear protocol.

Fig 2

Viral DNA (A) over the time course of papillomavirus infection and anti-MmuPV1 E4 antibodies (B) were detected in the NU/J heterozygous mice. In addition to the previously used high titers of virus for infection, we also assessed a viral titer as low as 1 × 10⁵ genome equivalents. Despite a delayed detection of viral DNA copies in the group with the lower titer, no significant difference was observed among these different groups at week 22 post-infection (the dotted line indicates the threshold of viral detection). Higher levels of viral DNA were detected in Rag1ko mice at several representative time points (indicated by a square). Significantly higher levels of anti-MmuPV1 E4 IgG3 antibody were detected in all infected NU/J heterozygous mice initiated with different doses of viral infection, compared to those in Rag1ko mice (indicated by a square), at dilutions ranging from 1:100 to 1:1,000 (B, P < 0.05, one-way ANOVA).

Fig 3

Viral particles were found in the cervicovaginal swabs of infected mice. Cervicovaginal swabs were collected from both NU/J heterozygous (A) and Rag1ko (B) infected mice and tested for viral particle presence by TEM (scale bar 100 nm). Consistent with previous reports, viral particles around 50 nm in diameter were found in both mouse strains. Rag1ko (B) mice produced more viral particles in the cervicovaginal swab samples when compared to those in NU/J heterozygous mice (A). Higher numbers of viral capsid protein L1 were detected in the infected vaginal tissues of Rag1ko (20×, D) mice than those in NU/J heterozygous (C) mice by IHC using in-house monoclonal antibody MPV.B9 (20×, Red dots, arrows). MmuPV1 L1 was detected using an in-house monoclonal antibody MPV.B9 via western blot (E).

Fig 4

Papillomavirus infection induces stages of neoplastic progression identifiable by Papanicolaou-stained Pap smear in the form of cytologically characterized neoplasia in exfoliated cells of mouse and corresponding human Pap Smear samples (inserts at the left corner, 40×). Representative cytology images showing stages of progression of mucosal neoplasia after mouse papillomavirus infection in both immunocompromised (Rag1ko, N = 5) and immunocompetent (NU/J heterozygous, N = 5) mice infected with 1 × 10⁹ MmuPV1 viral DNA equivalent and followed longitudinally. (A) Normal: normal squamous cells with oval-shaped and small nuclei are among a large number of anucleate cornified squamous cells. They are negative for intraepithelial lesion or malignancy (Fig. 4A); (B) ASC-US: Atypical squamous cells of uncertain significance with nucleus size 1.5–2 times larger than a normal squamous cell (Fig. 4B, arrow); (C) LSIL: Low-grade squamous intraepithelial lesion with koilocytes (hyperchromatic raisin-shaped nuclei with nucleus size 2.5–3 times larger than a normal squamous cell and a clear halo around it, arrow; or binucleated cells with a clear halo, arrow); (D) ASC-H: Atypical squamous cells with hyperchromatic nucleus sized 2–2.5 times larger than a normal squamous cell and increased N-C ratio (arrow), cannot exclude HSIL; (E) HSIL: High-grade squamous intraepithelial lesion (HSIL, hyperchromatic cells with high N-C ratio scattered individually or forming sheets, arrow), encompassing cervical intraepithelial neoplasia grades 2 and 3; CIN2 and CIN3; and F) SCC: squamous cell carcinoma (SCC, keratinizing with marked pleomorphism of cell size and shape, arrow; the presence of tadpole cells with dense orangeophilic cytoplasm and hyperchromatic nuclei, arrow; spindle-shaped cells, and inflammation and necrotic debris on the background). These stages of neoplastic progression are based on the 2014 Bethesda system and represent increased grades of intraepithelial neoplasia. Both immunocompromised (G, Rag1ko) and immunocompetent (H, NU/J heterozygous) mice infected with MmuPV1 at the lower genital tract developed LSIL as early as week four post-infection. An increased proportion of Pap smear samples became HSIL over time in both mouse strains. After 6 months, most infected animals developed HSIL and all of them progressed to SCC after 8 months post-infection in the vaginal tissues.

Fig 5

Papillomavirus infection induces identifiable stages from vaginal intraepithelial neoplasia (VAIN) to squamous cell carcinoma (A). Infected mice were sacrificed at different time points post-viral infection for histological validation. We observed the agreement between Pap smear LSIL and VAIN1; HSIL and VAIN2/3; and SCC in our tested animals. Overall, it may take around 5.5 months after mouse papillomavirus infection in the mucosa of infected mice to develop advanced diseases (HSIL to SCC). All mice developed SCC after 8 months post-infection. Corresponding human cervical intraepithelial neoplasia tissues (A, inserts at the right corner, 20×) share similar pathology with the mouse tissues. Histology of squamous cell carcinoma (2× and 10×, B) with a scirrhous response (B, arrows) typical of carcinomas was detected in SCC. Strong viral E4 protein (2× and 10×, C, arrows) and RNA transcripts (2× and 10×, D, arrows) were detected by IHC and RNA-ISH, respectively. Most inflammation was lymphoplasmacytic. There is a low level of immune cells in the stroma of the vagina (red circle), but this cluster was more than anticipated = score 1.

Fig 6

Increased mitotic signal (Ki67) and endothelial cellular adhesion molecule (CD31) were found in representative infected mucosa of mice. Representative images of either Ki67 or CD31 expression in the normal (20×, A, C) or infected vaginal tissues with advanced SCC (20×, B, D) of both tested Rag1ko and NU/J heterozygous mice. Significantly higher Ki67-positive cells (E, P < 0.01, unpaired Student’s t-test) and CD31-positive blood vessels (F, P < 0.01, unpaired Student’s t-test) were found in SCC tissues when compared to those in normal tissues.