Single-molecule measurements of DNA topology and topoisomerases

National Heart, Lung, And Blood Institute

Investigators

Linked publications & trials

Abstract

Research in Progress The first project is focused on elucidating mechanistic details of the interaction between type II topoisomerases and DNA. One aspect of this interaction concerns the ability of type II topos to relax the topology of DNA to below equilibrium values. In vivo, these topoisomerases are responsible for unlinking replicated chromosomes prior to cell division. Since a single link between sister chromosomes can prevent division and induce cell death, it is important that these enzymes preferentially unlink rather than link DNA molecules. This has been demonstrated In vitro, but the mechanism remains a mystery. In a new project in collaboration with Professor Siddhartha Das in the Department of Mechanical Engineering at the University of Maryland College Park, we are using a combination of single-molecule DNA relaxation measurements and molecular dynamics simulations to test the hooked-juxtaposition model of type II topoisomerase unlinking activity. This model suggests that the non-equilibrium topology simplification by type IIA topoisomerases arises from preferential passage of DNA segments that are juxtaposed in a hooked configuration in which the two strands are sharply bent towards each other. We can directly control the degree of this hooked bending and measure how this influences the rate of strand passage in single-molecule experiments combined with MD simulations, which will provide the first experimental test of this hypothesis. In related work in collaboration with Sean Collums at the University of Glasgow, we are examining how the bacterial type II topoisomerase, topo IV, distinguishes between DNA crossings in linked versus supercoiled DNA substrates. This work further contributes to our understanding of how type II topoisomerases select substrates on which to act. This work dovetails with the single-molecule measurements of unlinking by topo IV and will provide a means of testing the predictions of substrate selection obtained from the single-molecule and simulation results. To complement single-molecule approaches, we developed next generation sequencing based approaches to probe topoisomerases-DNA interactions. The in vitro approach provides nucleotide resolution mapping of topoisomerase binding, and cleavage site location and frequency with unprecedented sensitivity. By varying the topology of the DNA plasmids we can quantitatively map the dependence of binding and cleavage site preferences and absolute cleavage levels on DNA topology. Furthermore, we can determine how clinically important topoisomerase poisons alter the cleavage site selection and cleavage levels and how these respond to DNA topology. An ongoing effort is combining the extensive cleavage site data with Artificial Intelligence/ Machine Learning approaches and biophysical modeling to define the mechanisms governing the weak but distinct cleavage site preferences of type II topoisomerases. We have expanded sequence space and directly mapped bacterial Topoisomerase IV and Gyrase to the purified bacterial genome in collaboration with Monica Guo at the University of Washington. We obtained more than an order of magnitude more cleavage sites than have previously been identified on the bacterial genome with the antibiotic ciprofloxacin and have for the first time resolved cleavage in the absence of type II poisons. In parallel work done in collaboration with Neil Osheroff at Vanderbilt University we have completed a comparative study mapping the cleavage of topoisomerase IV and gyrase from several bacterial species, revealing similarities and differences in sequence specificity among bacterial type II topoisomerases with exquisite sensitivity. In collaboration with John Nitiss at the University of Illinois at Chicago, we are using in vitro next generation sequencing approaches to define the cleavage sites and levels of mutant type II topoisomerases that result in increased mutational frequencies in vivo. We have further expanded this collaboration to measure the distribution of yeast type II topoisomerase cleavage over the entire yeast genome in the absence and presence of the eukaryotic type II topoisomerase inhibitor, Etoposide. Similar to the bacterial genome cleavage data, we expect that machine learning or AI approaches will provide a predictive model relating cleavage levels to sequence for this eukaryotic type II topoisomerase. In similar work, we are mapping the cleavage specificity of additional bacterial type II topoisomerases in collaboration with Dagmar Klostermeier at the University of Meunster, Germany. This work contributes to the goal of identifying the protein determinants of cleavage specificity of type II topoisomerases, in addition to understanding species specific differences that could be related to differences in the genomic DNA among different bacterial species. We have developed an approach to leverage the in vitro next generation sequences approaches to obtain sequence-resolved structural information of supercoiled DNA substrates. Unwinding or negative supercoiling of circular plasmid DNA results in the formation of locally melted structures. By selectively cleaving the DNA at melted or distorted regions and sequencing the DNA fragments, we can identify the location and extend of melting or DNA distortion. In collaboration with Alice Pyne at the University of Sheffield and Fenfei Leng at Florida International University, we are determining the locations and frequences of DNA melting and duplex distortion in highly strained mini-circles of DNA, and correlating this information with AFM imaging of the same minicircles to develop a general approach to obtain sequence level information from AFM images of mini-circle DNA. Combining these results with molecular dynamics simulations and computational approaches, we further aim to develop predictive models for the location and frequency of DNA melting and duplex distortion in topologically and mechanically stressed DNA. In another ongoing project in collaboration with Neil Osheroff at Vanderbilt University, we are directly monitoring the poisoning of type II topoisomerases by fluroquinolone antibiotics at the single-molecule level. We can directly measure the transient poisoning of the topoisomerase during ATP driven strand-passage, which allows us to determine the on-rate and off-rate of the poison interacting with the active topoisomerase, and how these rates are influenced by the topology, torque, and force on the DNA. In other collaborative work with the Osheroff lab, we have contributed to the biochemical characterization of the new dual targeting antibiotic Gepotidacin and its interactions with its targets in E. coli: Topoisomerase IV and Gyrase. The third project involves the molecular mechanism of topoisomerase IA activity. We previously directly observed the opening and closing of type IA enzymes as they reversibly cleave and religate a single DNA strand during their catalytic cycle. We are currently investigating the human topoisomerase IIIα along its essential accessory proteins RMI1 and RMI2 that have been predicted to alter gate dynamics. In collaboration with Yuk-Ching Tse-Dinh at Florida International University, we are conducting structure function measurements of the gate dynamics of the bacterial type IA enzymes (Topoisomerase I and Topoisomerase III) to elucidate the critical structural features that govern gate dynamics and performing molecular dynamics simulations to relate the motions we observe experimentally to the molecular scale motions of the proteins. As part of this collaboration, we have identified a novel magnesium-dependent reorganization of a salt-bridge network that regulates the conformational dynamics in type IA topoisomerases. We have evidence that this magnesium-dependent salt-bridge network switching is conserved among type IA topoisomerases, and that it may represent an underappreciated general mechanism through which enzyme activity is regulated by magnesium. The fourth project involves the role of DNA topology in the identification and repair of DNA damage. We previously established that a single mismatched base in 6 kb of DNA will preferentially localize at the tip of a plectoneme in supercoiled DNA. We have recently extended these results to include negatively supercoiled DNA via multiscale simulations of DNA containing mismatches in collaboration with Siddhartha Das in the Mechanical Engineering department at the University of Maryland. Experimental and computational results indicate that supercoiling of DNA can contribute to the localization and identification of mismatches or other DNA damage by repair enzymes that recognize sharply bent DNA with a flipped-out base, both of which are favored when the damaged site is localized at the tip of a plectoneme in supercoiled DNA. These projects have been enabled by the continued development of magnetic tweezers instruments that afford high spatial and temporal resolution measurements of the topology of individual DNA molecules. The ongoing development and improvement of this magnetic tweezers instrument represents a sustained research endeavor. We have recently added a total internal reflection fluorescence (TIRF) modality, and a separate total internal reflection dark-field scattering modality to the magnetic tweezers instruments that permit single-molecule fluorescence measurements and high temporal-spatial resolution tracking of nm scale gold particles acting as local probes of displacement and torque, in conjunction with single-molecule manipulation.