The technique involves using proteins called zinc finger nucleases to bind to a specific sequence of DNA, which contains a harmful mutation, and then cut both strands. The cells own DNA repair mechanisms are stimulated into action by the broken double-stranded DNA. A piece of DNA containing the correct sequence is also inserted into the cell, which acts as the blueprint for the repair mechanisms to produce the desired sequence.

 

The technique has now been performed on human cells with some success. It is hoped that it may prove useful for people suffering severe genetic diseases, and that it will have fewer side effects than existing methods.

 

From: http://www.newscientist.com/article.ns?id=dn7224&feedId=online-news_rss20

Sounds cool.

 

I hope it does better than that other attempt.

Yeah sounds really cool. Apparently the engeneered proteins involved are modular in the sequence they recognise so they may potentially be assembled in different fasions. It's nice to have something with some specificity so you don't accidently end up inducing cancer causing mutations.

Given it only recognises short sequences (though being modular it could be expanded) I imagine it would bind to more than one site in the genome. That might cause some problems. It looks better than the current method though.

Acutally you can get the exactly opisite effect from having modular proteins if designed right. It increases the specificity alot of the time. Think of it as a key the more notches you have on the key the more specific it is to the lock it fits into. The trick is to "line up the notches" in such a manner so either they all bind or none bind.

So maybe one may bind but because the one adjacent to it can't fit in it get's pushed away.