Direct Readout of DNA sequences by proteins involves hydrogen bonding and van der Waals contacts between the protein and the base edges of DNA in the major groove, where the location and types of groups are distinguishable among the different DNA bases. Indirect readout involves all forms of recognition excluding direct readout. Since direct readout is visible using structural studies and point mutations, it has been much better characterized. Indirect readout is less obvious, and inferred by the presence of specificity without direct readout contacts. It may involve water mediated hydrogen bonds or distortions in DNA that distiguish sequences energetically. Follow the link here to read more about our studies of indirect readout of DNA sequence.

We are interested in the design of zinc finger proteins for various biomedical applications including designer nucleases for gene therapy. In order to recognize a unique site in the human genome, a DNA binding protein must recognize at least 18 base pairs. Zinc fingers which recognize particular sequences in DNA (3-4 base pairs in length) with high specificity are not available for all possible sequences. Further, the linking of zinc fingers in a single molecule to provide specificity for at least 18 base pairs has been problematic. We are studying the recognition properties of artificial zinc fingers that recognize the rarer DNA sequences, and using structure-function studies to understand the requirements for linking zinc finger domains into single protein molecules recognizing long DNA sequences.

We have unexpectedly discovered the presence of domain swapping (where a region of a protein from molecule swaps places with the same domain in a second copy of the protein) in a sequence specific endonuclease from Streptomyces griseus. The domain swapping stabilizes a tetrameric form of the nuclease, which may be responsible for the unusual properties of this enzyme: after cleavage of its target site in DNA, it will cleave additional different (secondary) sequences. The enzyme is stimulated to cleave these secondary sites at least 10 fold. The secondary sites would not be protected in the organism, hence cleavage of the secondary sites could result in severe DNA damage and consequently cell death. The existence of this property may have evolved as a way to control phage growth in colonies of Streptomyces griseus, where infected phage DNA is cleaved by the endonuclease, and the cleaved DNA results in stimulation of the secondary site cleavage and cell death. Programmed cell death, or apoptosis, is a critically important function in multi-cellular organisms, and the understanding the evolution of this pathway is important for understanding the pathway in humans.

We are engaged in structure-function studies of several DNA repair factors from human, yeast, and archaeal sources. Of interest is how damage is recognized in DNA, and how the pathway coordinates to repair DNA damage. Repair of DNA damage is critical to preventing mutations, and many repair proteins are targets for prevention and treatment of cancer. The archaeal proteins, which often show greater similarity to human proteins than bacterial, shed light on the evolution of repair pathways, and the adaptation to extreme environments.

We are also interested in the mechanism of replication and insertion of viral DNA into the human genome. Viruses such as human Parvo B19 cause outbreaks of infection in humans, and in some cases severe side effects. Alternatively, incorporation of particular coding sequences into the human genome can be of use in reprogramming cells into stem cells or into differentiated cells for the treatment of many human diseases. We study these enzymes in order to provide new opportunities for rational drug design against the viruses, and to engineer them for new uses in biomedical research.