Tse, David

Abstract Architectural proteins alter the shape of DNA. Some distort the double helix by introducing sharp kinks. This can serve to relieve strain in tightly-bent DNA structures. Here, we design and test artificial architectural proteins based on a sequence-specific Transcription Activator-like Effector (TALE) protein, either alone or fused to a eukaryotic high mobility group B (HMGB) DNA-bending domain. We hypothesized that TALE protein binding would stiffen DNA to bending and twisting, acting as an architectural protein that antagonizes the formation of small DNA loops. In contrast, fusion to an HMGB domain was hypothesized to generate a targeted DNA-bending architectural protein that facilitates DNA looping. We provide evidence from Escherichia coli Lac repressor gene regulatory loops supporting these hypotheses in living bacteria. Both data fitting to a thermodynamic DNA looping model and sophisticated molecular modeling support the interpretation of these results. We find that TALE protein binding inhibits looping by stiffening DNA to bending and twisting, while the Nhp6A domain enhances looping by bending DNA without introducing twisting flexibility. Our work illustrates artificial approaches to sculpt DNA geometry with functional consequences. Similar approaches may be applicable to tune the stability of small DNA loops in eukaryotes.

Using DNA ligase‐catalyzed cyclization assays, the in vitro persistence length of DNA has been determined to be about 150 base pairs (bp). In living E. coli bacteria, where DNA loops regulate the lac operon, DNA looping appears to be much more probable than expected from in vitro data. Why? A possible explanation for this paradox is that DNA loops in bacteria are bridged by proteins, reducing the amount of required costly DNA bending. Current in vitro methods to measure DNA flexibility are tedious and artificial, in that they require DNA cyclization, not just DNA looping. No high‐throughput method has been described to measure DNA flexibility for loops involving bridging proteins of different sizes. We are developing a high‐throughput method to measure the probability of DNA looping when bridged by a fusion protein that includes PCV2 and micrococcal nuclease domains. A bead‐tethered 1000‐bp DNA molecule of a known sequence is conjugated to this fusion protein through the PCV2 domain, and looping allows self‐cleavage such that the distribution of cleavage sites revealed by next generation sequencing reflects the energetics of DNA looping. We hypothesize that this distribution will be affected as a function of bridging protein size. Measuring how the size of bridging proteins affects tight DNA loop probability enables improved understanding of genetic switches.