| Summary: | <p>Abstract</p> <p>Background</p> <p>The Human Papillomavirus (HPV) genome is divided into early and late coding sequences, including 8 open reading frames (ORFs) and a regulatory region (LCR). Viral gene expression may be regulated through epigenetic mechanisms, including cytosine methylation at CpG dinucleotides. We have analyzed the distribution of CpG sites and CpG islands/clusters (CGI) among 92 different HPV genomes grouped in function of their preferential tropism: cutaneous or mucosal. We calculated the proportion of CpG sites (PCS) for each ORF and calculated the expected CpG values for each viral type.</p> <p>Results</p> <p>CpGs are underrepresented in viral genomes. We found a positive correlation between CpG observed and expected values, with mucosal high-risk (HR) virus types showing the smallest O/E ratios. The ranges of the PCS were similar for most genomic regions except <it>E4</it>, where the majority of CpGs are found within islands/clusters. At least one CGI belongs to each <it>E2/E4 </it>region. We found positive correlations between PCS for each viral ORF when compared with the others, except for the LCR against four ORFs and <it>E6 </it>against three other ORFs. The distribution of CpG islands/clusters among HPV groups is heterogeneous and mucosal HR-HPV types exhibit both lower number and shorter island sizes compared to cutaneous and mucosal Low-risk (LR) HPVs (all of them significantly different).</p> <p>Conclusions</p> <p>There is a difference between viral and cellular CpG underrepresentation. There are significant correlations between complete genome PCS and a lack of correlations between several genomic region pairs, especially those involving LCR and <it>E6</it>. <it>L2 </it>and <it>L1 </it>ORF behavior is opposite to that of oncogenes <it>E6 </it>and <it>E7</it>. The first pair possesses relatively low numbers of CpG sites clustered in CGIs while the oncogenes possess a relatively high number of CpG sites not associated to CGIs. In all HPVs, <it>E2/E4 </it>is the only region with at least one CGI and shows a higher content of CpG sites in every HPV type with an identified <it>E4</it>. The mucosal HR-HPVs show either the shortest CGI size, followed by the mucosal LR-HPVs and lastly by the cutaneous viral subgroup, and a trend to the lowest CGI number, followed by the cutaneous viral subgroup and lastly by the mucosal LR-HPVs.</p>
|
|---|
