REGULATION OF PAPILLOMAVIRUS GENE EXPRESSION

乳头瘤病毒是嗜上皮病毒，可诱导良性和恶性肿瘤

The papillomaviruses are epitheliotropic viruses which induce benign and malignant lesions in a variety of squamous epithelia such as skin and the cervix. The papillomavirus life cycle is intimately linked with the differentiation state of the squamous epithelium which it infects. One goal of this project is to establish both tissue culture and animal systems that can be used to study the full viral life cycle. Alison McBride, in collaboration with this lab, has successfully obtained productive infection with cloned bovine papillomavirus type 1 (BPV-1) DNA by transfection into bovine keratinocytes in culture and differentiation of these cells with a combination of organotypic culture and nude mouse xenografts. We have also cloned HPV-16 genomic DNA from a cell line (W12) which harbors extrachromosomal HPV-16. Preliminary experiments indicate that we are able to transfect primary human foreskin keratinocytes (HFK) with this DNA and establish cell lines that maintain extrachromosomal HPV-16 DNA. These systems should allow us to use mutant papillomaviruses to study the roles of viral cis elements and trans factors in the viral life cycle. We have also developed a nude mouse xenograft system to study the differentiation-dependent processing of papillomavirus pre-mRNAs. This system uses integrated expression vectors and allows the assay of cis-processing elements that overlap viral genes essential for viral replication. Surprisingly, mice grafted with HaCat cells expressing a BPV-1 late pre-mRNA (containing the E4 and E5 ORFs but not the E6 or E7 ORFs) developed large tumors with characteristics of papillomavirus-induced papillomas. This system will also be used to investigate the signaling pathways through which E5 functions and the roles of E4 and E5 in the development of papillomas. BPV-1 late pre-mRNAs are alternatively spliced in a differentiation dependent manner. This alternative splicing is an essential component of the early to late switch in viral gene expression. Several cis- elements have been identified which regulate splice site choice in vitro. Immediately downstream of the first of two alternative 3' splice sites is a bipartite splicing regulatory element consisting of a purine- rich exonic splicing enhancer (ESE) and a pyrimidine-rich exonic splicing suppressor (ESS). A second ESE-like element is located a short distance upstream of the second alternative 3' splice site. This element could act either as an enhancer on the upstream 3' splice site or as a repressor on the downstream 3' splice site. We have now shown that mutation of either ESE switches splice site usage in vivo, highlighting the importance of these two elements. Both ESE elements bind the same set of SR splicing factors. This arrangement potentially allows the coordinated regulation of two alternative splice sites by the same transacting factors. We have now carried out a detailed characterization of the ESS. In vitro splicing studies using several heterologous pre- mRNAs indicate that suppression of splicing requires a suboptimal 3' splice site but not an ESE, suggesting that the ESS works directly on the 3' splice site. Native gel electrophoresis has been used to demonstrate that the ESS inhibits early steps in spliceosomal assembly. Mutational analysis of the ESS showed that although the entire ESS is necessary for maximal suppression of splicing in vitro, the central C- rich region is the most important. UV cross-linking and immunoprecipitation studies showed that U2AF65 binds the U/C-rich 5' half, the C-rich central region binds 30 and 55 kDa SR proteins, and the AG-rich 3' end binds the SR protein ASF/SF2.