I didn't read the paper thoroughly yet, but from a brief skimming this is the gist: The technique they used was very clever. The first step was to take what's

I didn't read the paper thoroughly yet, but from a brief skimming this is the gist: The technique they used was very clever. The first step was to take what's called a zinc finger. This is a type of DNA binding protein that can specifically recognize a sequence of DNA and bind to it. In this case, they created two zinc fingers that bound to a region of the first transmembrane domain (CCR5 has 7 transmembrane domains), just upstream of the deletion in the delta32 mutation. Then, attached to the zinc fingers is a type IIS restriction enzyme. What this does is make a double stranded cut in the DNA. In other words, it cuts both strands in half. The cell, sensing the damage, repairs it with nonhomologous end joining (http://en.wikipedia.org/wiki/Non-homologous_end_joining http://en.wikipedia.org/wiki/Non-homologous_end_joining). Due to the nature of the cut, mutations are introduced in the repair process. It seems in many cases, the mutations introduced are sufficient to disrupt the CCR5 protein and make it non-functional. While it's not as much a "sure thing" as the delta 32 mutation, which introduces a frameshift, it seems relatively effective. They used an adenovirus to introduce their zinc fingers + restriction enzyme, so this has the risks associated with using a virus as a delivery system (ie cancer). Link to paper: http://www.scribd.com/doc/3746359/Establishment-of-HIV1-resistance-in-CD4-T-cells-by-genome-editing-using-zincfinger-nucleases- http://www.scribd.com/doc/3746359/Establishment-of-HIV1-resi...