Date Published: April 6, 2017
Publisher: Public Library of Science
Author(s): Alison A. McBride, Alix Warburton, Katherine R. Spindler.
Persistent infection with a subset of “high oncogenic risk” human papillomaviruses (HPVs) can promote the development of cancer. In these cancers, the extrachromosomal viral genome has often become integrated into the host genome. The integration event is thought to drive oncogenesis by dysregulating expression of the E6 and E7 viral oncogenes, leading to inactivation of critical cell cycle checkpoints and increased genetic instability in the host. This Pearl reviews the evidence that gave rise to the current textbook paradigm of HPV integration events and their consequences and incorporates new findings that demonstrate that stochastic integration events can promote oncogenesis in many ways.
Papillomaviruses have a resourceful life cycle that takes advantage of the tissue renewal process of stratified epithelia. Only the lower, basal cells in the epithelium proliferate, but they can divide either symmetrically (to produce more basal cells) or asymmetrically (one of the daughter cells leaves the basal layer and begins the differentiation process). Differentiating daughter cells move up through the epithelium, acquiring specialized properties until they are released from the surface as part of the process of tissue renewal. Papillomaviruses exploit this process; they access and infect the basal cells through a micro-abrasion and establish a long-term infection in these dividing cells. The viral E1 and E2 proteins support replication of the viral genome at a low copy number in the basal cells as a small, dsDNA nuclear plasmid of about 7–8 kbp. It is only when these infected cells differentiate and move towards the surface of the epithelium that high levels of viral DNA are synthesized, packaged in virions, and sloughed from the surface of the epithelium in viral-laden squames . HPVs are often found integrated in premalignant lesions and a range of anogenital and oropharyngeal cancers [2–4], but this is not part of the viral life cycle. In fact, integration is a dead end for the virus, as it is no longer able to form a small, circular genome that can be packaged and transmitted to a new host.
Almost all HPV integration events that have been studied in detail to date are related to HPV oncogenesis. HPV integration events can be detected in premalignant lesions, but the percentage of cells containing integrated HPV increases as cells progress to invasive cancer . Integration usually results in dysregulation of expression of the viral E6 and E7 oncogenes, which promotes cellular proliferation, abrogates cell cycle checkpoints, and causes progressive genetic instability. This gives cells a selective growth advantage and promotes oncogenic progression . Clonal outgrowth of cells with integrated HPV highlights the importance of dysregulated oncogene expression. In fact, HPV-associated cancers are dependent on the expression of the viral E6 and E7 oncogenes for continued proliferation and survival .
Although many HPV-associated cancers contain integrated viral DNA, it is not universal. HPV-associated cancers can contain either integrated HPV DNA, extrachromosomal viral DNA, or a mix of both . However, in tumors with exclusively extrachromosomal viral DNA, the viral genome has usually acquired genetic or epigenetic changes that result in dysregulated E6/E7 gene expression [12, 21, 22]. For example, methylation of the E2 binding sites in the URR can alleviate E2-mediated repression of the viral oncogenes [22–24]. Therefore, while integration is very frequently detected in HPV-associated cancers, it is not absolutely required.
Genome-wide efforts to elucidate a genomic signature of HPV integration events have identified only a handful of recurrent loci. These so-called genomic “hotspots” are highly correlated with common fragile sites  and transcriptionally active regions of the genome [19, 32]. Regions of microhomology (1–10 bp) between viral and human genomic sequences are sometimes found at integration breakpoints , as have AT-rich regions of the genome that have the potential to form stem—loop structures that promote the formation of stalled replication forks during replication stress . There are also examples of HPV integration resulting in insertional mutagenesis and/or potential regulatory effects on neighbouring genes . Increased integration within cancer-associated genes or pathways, including the MYC gene locus, have also been reported [2, 34]. However, this is not a universal phenomenon associated with HPV integration .
There are several stages in the development of an HPV integration site that eventually drives oncogenic progression. The initial integration event likely takes place in regions where the viral and host DNA are in close proximity in the proliferating basal cells of a lesion. HPV hijacks the host DNA damage response to replicate its own DNA in certain phases of the life cycle [26, 27], and this occurs adjacent to regions of the host DNA susceptible to replication stress (e.g., common fragile sites) . This could explain the preferential integration in these regions . Secondly, the target region must be in a transcriptionally competent region of host chromatin that can support viral oncogene expression. In most cases, E6/E7 mRNA is expressed as a viral—host fusion transcript, and this necessitates the presence of a nearby cellular splice acceptor and polyadenylation site (often cryptic). Finally, it is likely that there is epigenetic modulation of the integration site (DNA methylation and chromatin modifications) that further determine whether the integration site is active or silenced [12, 39]. Therefore, many events and processes contribute to the development of an HPV integration event that is a strong driver of oncogenesis; there are likely many dead-end integration events that fail to produce sufficient E6/E7 oncoproteins to drive clonal expansion of the host cell.