Genomic Modification by Zinc Finger Nuclease
The ability to make site-specific genomic modifications in human cells, as well as other commonly studied model organisms, has significant therapeutic and experimental potential. A reliable, accessible technology would allow knock-ins and knockouts to be created for protein coding sequences, microRNA, or cis-regulatory elements, making an almost limitless contribution to the understanding of any biological pathway. Unfortunately current approaches have been limited by the low frequency of homologous recombination in these cell types. However, a double strand break (DSB) created within the target sequence can increase the rate of recombination by 10,000 fold or greater, making gene repair in human and other cells plausible. This has led to extensive research on the development of zinc finger nucleases (ZFNs) that provide a malleable DNA-binding domain – nuclease fusion with the potential to generate DSBs at unique loci in any genome. Engineered Cys2His2 zinc fingers have successfully targeted nucleases in species as diverse as Drosophila, Zea mays, zebrafish and human cell lines to create knockouts and knock-ins.
Previously described ZFNs are limited by the types of sequences they can target and the prevalence of off-target lesions, complicating their use for gene repair. These limitations may be due to bias inherent to the zinc finger domain or failings of current selection systems. Furthermore, any improved efficiency of recombination due to a DSB falls off dramatically with the distance between the cut site and desired genetic conversion. As a result, efficient gene repair in human cells may require targeting of the DSB precisely to the genetic lesion, which would render any specificity limitations problematic. My lab focuses on the development of systems that utilizes a combination of DNA-binding domains (DBDs) and overlapping in vivo and in vitro selection platforms that allow for the selection against off-target cutting. These domains are selected to consider both direct and indirect influences on specificity with the goal of providing engineered domains with optimal specificity for any desired sequence. Ideally, these domains will be able to target a break precisely to a unique locus while eliminating off-target lesions.
Understanding Regulatory Networks of Gene Expression
A great deal of genomic data that is now available for many different species yet the regulatory networks that control gene expression remain relatively undefined. One of the central limitations has been the lack of transcription factor DNA-binding specificity data that would allow for the prediction of functional binding elements in a genome. Certainly methods for characterizing individual transcription factors have been available for decades but they are laborious and time consuming. In addition, because individual factors typically bind short, degenerative sequences, predictions based on single specificities would predict thousands to millions of binding sites throughout a genome. However, a comparison of known regulatory elements indicates that functional binding sites are often found in clusters, containing somewhere on the order of 10 binding sites for 3 or more different factors within a 500bp fragment of DNA. Therefore it is likely that predicting these functional elements will require specificity data for multiple factors that function in a common pathway.
To overcome current specificity limitations high throughput methods such as protein binding microarrays, bacterial and yeast hybrid assays, as well has ChIP assays have been used to characterize many transcription factors. These assays have been very successful characterizing common DNA-binding domains and in some cases the number of transcription factors with known monomeric specificities in a given species is approaching 20% or greater. However, these relatively high throughput methods are just beginning to address the complication of dimeric specificities or transcription factors with potentially long target sites. These higher order, combinatorial interactions may prove to be far more complex and diverse than the simple set of monomeric specificities. My lab uses a bacterial one-hybrid (B1H) system based on the omega subunit of RNA polymerase to characterize the DNA-binding specificity of monomeric transcription factors from multiple species. Modifications to the B1H system as well as the development of new selection platforms are under development for the high throughput characterization of dimeric interactions with DNA as well as the protein-protein interactions that define the set of potential dimeric partners for any given transcription factor.