A novel model of HPV infection in meshed human foreskin grafts

Abstract

The present study describes a novel meshing procedure that provided successful low-risk papillomavirus propagation and reproducible wart induction in human foreskin xenografts. The initial HPV6 and/or 11 inocula were collected from clinically excised human wart tissues and confirmed to be free of HPV16, 18 and 31 by PCR analysis. Human foreskin grafts were collected from a circumcision clinic, and pre-inoculated with HPV virions by scarification. Meshing was carried out with a Zimmer Skin Grafter Mesher. Grafts were cut to appropriate size (1 cm × 1 cm or 5 mm × 5 mm) for cutaneous or subcutaneous grafting to NIH-nu-bg-xid mice under halothane anesthesia. Cutaneous xenografts were dressed with antibiotics and protective band-aids for 3 weeks. In the paralleled experiment using the same viral stock containing both HPV6 and 11, and matched grafts, no visible papillomas were observed in non-meshed cutaneous xenografts (n = 4 up to 6 months). In comparison, six of eight cutaneous xenografts treated with the meshing procedure formed visible papillomas within 4 months. This high frequency of distinct papilloma induction over the surface of meshed xenografts were reproduced in subsequent experiments with viral stocks containing both HPV11 and 6 (8 out of 10 grafts), or with a single-type HPV11 inoculum (80–100%). In contrast, an initial viral stock of single-type HPV6 provided lower frequency and more delayed papilloma induction. Serial passage of HPV6 in the meshed xenograft appeared to improve both the induction frequency and growth rate up to the 3rd generation. Histology, in situ hybridization, and immunohistochemical analysis revealed similarity of xenograft warts to those observed in the clinic. The highly reproducible papilloma induction rate and successful viral stock propagation associated with the meshing procedure provide a novel feature in the HPV xenograft model.

Introduction

Human papillomavirus (HPV) infection is widespread in the population and associated with a broad spectrum of human diseases ranging from benign condyloma, common skin warts, and laryngeal papillomas to anogenital cancers (Gross, 1997, Howett et al., 1990, Kreider et al., 1990, Pfister, 1984). While HPV-associated anogenital cancers may be life threatening, there is also an important clinical need for more efficacious therapy for benign HPV diseases due to their morbidity and related economic and social costs (Gross, 1997, Phelps et al., 1998, Phelps and Alexander, 1995, Tewari et al., 2000). To assist in the discovery and development of new antivirals against HPV, several in vitro and in vivo models of HPV infections have been developed (Bonnez et al., 1998, Bonnez et al., 1993, Brandsma et al., 1995, Brown et al., 1998, Bryan et al., 2000, Christensen and Kreider, 1999, Christensen et al., 1997, Culf and Christensen, 2003, Culf and Christensen, 2004, Dollard et al., 1989, Dollard et al., 1992, Howett et al., 2000, Iyer et al., 2002, Laimins, 1993, Majewski and Jablonska, 1997, Tewari et al., 2000, Yiu et al., 1990). However, severe limitations exist for all currently described models (Christensen and Kreider, 1999, Stanley et al., 1997).

In vivo modeling of HPV infection faces numerous challenges. Amongst them are the strict host-specific tropism of HPVs and the lack of conventional methods of viral propagation (Gangemi et al., 1994, Kreider et al., 1990, Pfister, 1984). Kreider et al. first described HPV papilloma induction in human xenografts (Kreider et al., 1987a, Kreider et al., 1987b, Kreider et al., 1986). In this model, human skin tissues were surgically implanted under the renal capsule of the nude mouse. The grafts were allowed to remain in the animal (2–3 months) until papillomatosis was observed in recovered grafts. Viral preparations derived from these xenograft warts were used to induce papillomas in subsequent experiments, verifying successful HPV propagation in this model. The choice of renal capsule as the receiving grafting bed was advantageous due to the enriched blood supply at this anatomical location, which might have been critical for survival, growth, and proliferative papillomatosis. However, the choice of this grafting location also created a limitation for the model, specifically a difficulty in accessing the grafts for treatments and observations. Several other laboratories also tried to extend the xenograft model to cutaneous HPV infections, and limited success in HPV16 virus and DNA inoculation have recently been reported (Christensen and Kreider, 1999, Stanley et al., 1997). Nevertheless, all of the previously published models quantify gross papilloma growth by the measurement of whole graft size rather than direct measurement of distinct visible papillomas similar to that observed in the clinic. Therefore, a highly reproducible and convenient model of cutaneous low-risk HPV infection for virus propagation and antiviral drug evaluations has not been established yet.

Following a number of attempts to infect cutaneously grafted human foreskin grafts in NIH-nu-bg-xid mice with low-risk HPV inocula from clinical wart specimens, we have developed a highly reproducible model of low risk HPV infection that features a novel procedure of graft meshing.