EVALUATING IN VITRO AND IN VIVO LOOPING EFFICIENCIES OF ARTIFICIAL DNA-BINDING PROTEINS

Abstract

DNA looping plays an important role in gene regulation by increasing the local concentration of a transcriptional activator or repressor at its primary binding site. Several <italic>in vitro</italic> and <italic>in vivo</italic> studies on DNA looping showed that the stability of protein-mediated DNA loops depends on the flexibility of both the looping protein and the DNA that contains the binding sites. We designed two types of short and rigid DNA looping proteins, based on a coiled-coil motif, in order to probe DNA flexibility on the thermodynamics of protein-mediated DNA loop formation. <italic>In vitro</italic> characterization of the putatively tetrameric DNA binding protein lzee by electrophoretic mobility shift assays (EMSA) did not show evidence of a sandwich complex, which is a necessary precursor to DNA looping. A quantitative <italic>in vivo</italic> looping assay, adapted from the reporter gene assay developed by Becker, Kahn, and Maher (2005), showed relatively weak enhancement of repression on GCN4 operators spaced >300 bp apart by lzee and the dimeric looping proteins LZD73, LZD80 and LZD87. However, lzee and LZD87 expression triggers cell toxicity orhighly decreased reporter protein expression on reporter strains containing GCN4 sites <240 bp apart. We proposed recombination events to explain the unexpected behavior in this distance range. Results from <italic>in vitro</italic> Plasmid Conformation Capture (PCC) revealed a very weak increase in crosslinking efficiency on 450- and 900-bp DNA loops. The apparent failure in capturing DNA loops by <italic>in vivo</italic> and <italic>in vitro</italic> PCC was attributed to the LZD proteins not being able to crosslink to DNA.

Lastly, we introduced two kinds of modifications to our DNA binding proteins. The first modification sought to improve the linker sequence in lzee in order to select for better tetrameric looping proteins. The other modification introduced lysine residues at the DNA binding domains in the dimeric GCN4 looping protein LZD87 to enhance their ability to crosslink DNA. The <italic>in vivo</italic> repression assay failed to select for lzee mutants that are better repressors than lzee, while the crosslinking assays on the LZD single mutants did not show clear evidence that the new proteins can crosslink DNA. Taking all of these results together, we have concluded that the inability of the LZD proteins to stabilize DNA loops <italic>in vivo<italic> support the model that DNA plays a more passive role in the thermodynamics of DNA looping. However, the unique ability of lzee to trigger recombination in our repression assays can be utilized to design assays that detect recombination as a consequence of DNA

looping, although further studies are required to understand this behavior. The molecular tools presented in this work would serve not only in providing a deeper understanding the thermodynamics of DNA looping, but could also be used as a starting point to develop better systems for modulating gene expression.