NEW YORK (GenomeWeb) – A team from China, Saudi Arabia, and Denmark has mapped human papillomavirus integration sites in cervical cancer genomes, cell lines, or pre-cancerous cervical neoplasias, garnering hints about integration hotspots that may contribute to cervical cancer development.
As they reported in Nature Genetics today, the researchers used genome sequencing and high-throughput viral integration detection (HIVID) — a method based on sequence data and computational breakpoint detection — to track down HPV integration sites in more than 100 cervical carcinoma tumors, dozens of cervical intraepithelial neoplasias (CINs), and five HPV-positive cell lines.
The resulting map of HPV integration sites provided a peek at sequences that are frequently altered by these integrations, along with the genes associated with them. It also pointed to the possibility of increasing HPV integration during the progression from CIN to cervical cancer.
"The integration hot spots identified in our study could be used as biomarkers for early personalized treatment and prognosis assessment of patients with cervical carcinoma," researchers from Huazhong University of Science and Technology, BGI-Shenzhen, and other institutions wrote.
A series of studies done over the past few decades has described sites in the genome that seem prone to HPV integration, the team noted. While PCR-based approaches formed the basis for many of these integration maps, those involved in the study and others have been exploring methods for cataloging viral integration sites and related mutation patterns using genome or transcriptome sequence data.
For the current study, researchers relied on a viral integration detection method called HIVID, which members of the team first described as a tool for tracking down hepatitis B virus integration sites in the human genome in a 2013 study in the journal Genomics.
After validating their HPV enrichment, whole-genome sequencing, and HIVID methods in HPV-positive cell lines and cervical carcinoma samples, the researchers applied HIVID to 104 cervical carcinoma tumors, 26 CIN samples, and five HPV-positive cell lines.
The approach uncovered some 3,666 HPV-related breakpoints in the genomes of 103 cervical tumors, 14 CINs, and four of the cell lines — integration sites the team verified using Sanger sequencing and RNA sequencing.
"During the progression of cervical lesions, the rate of HPV integration rose markedly from 53.8 [percent] of CINs to 81.7 [percent] of cervical carcinomas," the study's author noted, adding that the "number of integrations increased from a median of [one] integration per case in CINs … to a median of [eight] integrations per case in cervical carcinomas."
"The increase of both integration rate and number from CINs to cancer highlights their potential values as predictors of disease progression," they continued.
The HPV insertions appeared to be peppered across the genome, researchers reported, though some sequences were punctuated by hot spots more frequently affected by the virus. Some 90 percent of the breakpoints cropped up fewer than 500,000 bases away from annotated genes. And almost half of those integration sites were within genes.
The team detected four or more integrations affecting the same set of 33 genes, for instance, as well as nine genes affected by at least five HPV-integration events, prompting speculation that the virus might gain a selective advantage by integrating at certain places in the genome during cancer development.
While the hotspots involved several genes associated with HPV integration in the past — including POU5F1B, FHIT, KLF12, and other genes — the researchers also uncovered recurrent integrations in new genes such as HMGA2, DLG2, and SEMA3D.
The virus seemed prone to integrating in and around such genes in both the pre-cancerous CIN samples and in tumor samples, the team noted. Nevertheless, the analysis hinted at differences in the gene expression consequences of such integrations depending on where in the gene they occur. For instance, the expression of genes such as FHIT was muted when HPV made its way into intron sequences, while integrations in sites flanking other genes, including HMGA2, enhanced their expression.
"Our data support the hypothesis that HPV may efficiently survey the human genome, activating or inactivating genes that favor positive clonal selection," the authors argued, "thereby providing more opportunities for the malignant transformation of host cells."
In other arms of their analyses, the researchers considered both the mechanism of HPV integration into the genome and the breakpoint consequences of such integration on the genome of the virus itself.
From the higher-than-usual levels of microhomology between HPV and human sequences uncovered in integration regions, for example, the group speculated that the virus may be helped along by microhomology-mediated DNA repair pathways as it melds with the human genome.