During transcription RNAP translocates along the major helical groove of DNA, unwinding DNA in front and rewinding DNA behind itself. According to the “twin-supercoiling domain” model, the torque that RNAP generates may significantly change DNA topology, positively supercoiling the downstream DNA and negatively supercoiling the upstream DNA.

To understand the processes that take place over supercoiled DNA, we first characterized the torsional properties of DNA. We have conducted extensive studies of DNA behavior under biologically relevant torques. Extending previous magnetic tweezers work on DNA buckling (formation of intertwined loops called plectonemes), we made the first direct measurement of the torque required to buckle DNA (Forth et al., PRL, 2008). The high spatiotemporal resolution of the angular optical trap also allowed us to detect the dynamics of the plectoneme formation. We then investigated DNA melting under torque and characterized the resulting DNA structure: left-handed helix with unpaired bases (Sheinin et al., PRL, 2011). In a tour-de-force demonstration of angular trapping capabilities we measured the minuscule (about 1 pN•nm) torques associated with the migration of a Holiday Junction – a recombination intermediate (Forth et al., BJ, 2011). These studies enable us to better understand the mechanical properties of DNA under torsion, and pave the way for more complicated experiments involving DNA-protein complexes.

Transcription under DNA supercoiling. (A) (left) Cartoon of RNAP working against downstream (+) supercoiling. (right) The distribution of the measured downstream stall torques. The smooth blue curve is a fit with a Gaussian function, yielding a mean of 11.0 ± 3.7 pN•nm (mean ± SD). (B) (left) Cartoon of RNAP working against upstream (-) supercoiling. (right) The distribution of measured upstream stall torques. The upstream stall torque distribution shows a singular peak immediately before a sharp cutoff near the DNA melting torque. The smooth curve is a fit with a Gaussian function assuming that the peaked fraction generated torques of at least 10 pN•nm, yielding a mean of 10.6 ± 4.1 pN•nm (mean ± SD).

By using the angular optical trap, which allows us to simultaneously apply and measure torque, our lab has produced the first measurements of the torque that an E. coli RNAP can generate as it works against (+) supercoiling in front, or (–) supercoiling behind (Ma et al., Science, 2013). We systematically investigated how torque can regulate transcription rate (the torque-velocity relation), and we found that a resisting torque slowed RNAP and increased its pause frequency and duration. We also determined that RNAP was able to generate 11 ± 4 pN•nm (mean ± SD) of torque before stalling. This is sufficient to melt DNA of arbitrary sequence, and establishes RNAP as a more potent torsional motor than previously known. The ability to melt DNA may facilitate the initiation of transcription from adjacent promoters, binding of regulatory proteins, and even initiation of replication. We also discovered that a stalled RNAP was able to resume transcription upon torque relaxation, and that transcribing RNAP was resilient to transient torque fluctuations, such as encountered in vivo. These results provide a quantitative framework for understanding how dynamic modification of DNA supercoiling regulates transcription. In addition, the techniques we established may be used to examine the broader impact of torque and DNA supercoiling on other DNA-based translocases.

Previously, the measured force RNAP can generate set the force scale for processes that take place over DNA (Yin et al., Science,1995; Wang et al., Science, 1998). Now these measurements of the torque that RNAP can generate set the relevant torque scale for these processes (Ma et al., Science, 2013).