Retroviruses are one of the most common vehicles for delivering therapeutic payloads via gene therapy in animal models of disease and human patients.
Viruses integrate into host DNA to replicate, but exactly where they insert themselves has become a topic of increasing importance. This is of special concern when integration is near an oncogene that may lead to uncontrolled, cancerous cell growth.
Now, researchers at the University of Pennsylvania School of Medicine have completed the first whole-genome survey of where three commonly used retroviruses integrate into human DNA. The team, led by Frederic Bushman, PhD, Professor of Microbiology, compared vectors derived from human immunodeficiency virus (HIV), murine leukemia virus (MLV), and avian sarcoma-leukosis virus (ASLV). They found that HIV integrated near active genes; MLV near points on the chromosome where protein translation starts (which confirms earlier work by another lab); and ASLV integrated more randomly throughout the entire genome. That each studied virus preferred a unique integration pattern or site suggests that viruses home in on certain chromosomal features for inserting themselves within the genome. This work appears in the August 17 issue of PLoS Biology, a new open-access journal.
“There’s a picture forming of where different retroviruses integrate in human cells, and it seems to be quite different from virus to virus, which is not something anyone would have ever suspected,” says Bushman. “We can only speculate as to the mechanism at present, but one attractive idea is that retroviral-integration complexes bind to cellular DNA binding proteins attached to specific locations on chromosomes.” For HIV, integrating into active genes may help promote efficient viral gene expression. The reason for the choice of target is less clear in other retroviruses.
These findings are important for devising safer human gene-therapy vehicles. From studies in yeast, the researchers speculate that there is a system of biochemical recognition between proteins bound on human chromosomes and viral proteins, which helps guide integration, and that specific recognition seems to differ from virus to virus. “There’s a prospect of modulating or engineering that kind of system, once we understand it better to direct integration to different locations,” comments Bushman.