Grant: Project Grant

One of the largest aclasses of human regulator factors is the "C2H2 zinc finger proteins". These proteins have a number of unique properties, including the potential to bind to very specific, long DNA sequences. They also evolve quickly, which makes them difficult to study - most do not have counterparts in other species, such as mouse. Until now, they have also been difficult to study biochemically, possible due to the length of their binding sites. We have developed a new technique, GHT-SELEX, which makes it much easier to measure their DNA binding. Here, we will determine what sequences the human C2H2 zinc finger proteins bind, what are the consequences of variation in their binding sites across the genome, and what happens if the regulatory proteins themselves vary between individuals, or are mutated in disease. We will also study how to computationally represent the sequence specificity of these proteins. Computer models are important because they are used to determine the effect of sequence variation among individuals. For C2H2 zinc fingers, it is a particularly difficult problem. The C2H2 zinc fingers are usually found in a row, with a media number of 11 fingers. Each finger binds three DNA bases, so that two fingers bind six bases, three bind to nine, etc. In addition, different segments of the C2H2 "array" can engage the DNA at different binding sites. Thus, a single protein with 11 fingers could potentially recognize thousands of different sites, We propose to develop new computational techniques to complement the GHT-SELEX data, which can dissect which fingers are used at different binding sites, and we plan to develop sequence scanning models that use the resulting information.