Response to Klyuyev and Vassylyev: on the mechanism of tagetitoxin inhibition of transcription

Abstract

In their commentary, Klyuyev and Vassylyev dispute a model of transcription inhibition by tagetitoxin (Tgt) proposed by us based on biochemical analysis and computational docking. We maintain that, although an alternative explanation can be provided for any single observation reported by us, taken together our results support a model in which Tgt acts by trapping the trigger loop (TL) in an inactive state (Artsimovitch et al.). This model is consistent with all the data collected with a physiological target for the inhibitor, the transcription elongation complex (EC). The Tgt-binding pose in our model is indeed different from that observed in the structure of the Thermus thermophilus RNA polymerase (RNAP) holoenzyme in the absence of nucleic acids (Vassylyev et al. Nat Struct Mol Biol 2005; 12:1086). The latter can hardly be considered a dogma because RNAP undergoes conformational changes in the course of the transcription cycle and during catalysis and small molecules containing phosphates likely bind to several sites on RNAP, with the crystallographic site/pose not necessarily being the one most relevant mechanistically. Furthermore, the model proposed based on the Tgt/holoenzyme structure does not explain the inhibitor's effects on transcript elongation and RNAP translocation. These arguments necessitate further inquiry into the mechanism of inhibition by Tgt by techniques orthogonal to X-ray crystallography. In our opinion, elucidation of a molecular mechanism of any RNAP inhibitor and the follow-up design of more potent derivatives requires a combination of approaches, including genetics, biochemistry, biophysics, X-ray crystallography and computational analysis.

Figure 1. Crystallographic models of Tgt bound to the T. thermophilus RNAP. (A) The Tgt pose reported in ²; the omit Fo-Fc map contoured at 3σ is shown as a green mesh. (B) An alternative chemically plausible orientation of Tgt that agrees well with the omit electron density. Both of these poses appear plausible at the current resolution based on bonding considerations as the negatively charged carboxylate and the phosphate swap locations in the two orientations of Tgt. Other plausible sites can be found that involve translations, in addition to this rotation of the Tgt, especially since RNAP has evolved to accommodate multiple phosphates in its active site or along its nucleic acid binding surfaces. Therefore, Tgt may not have a unique binding site on RNAP, even as a static enzyme.

Figure 2. A clash between Tgt and the folded TL. The docking model of the Tgt-EC complex was superimposed upon the T. thermophilus EC structure (2o5j). From the docking model only Tgt (purple sticks) is shown. From 2o5j, the bridge helix and the TL are shown as semi-transparent surfaces and cartoons, teal and orange respectively. The TL residue Arg1239, which exhibits the most severe clash with Tgt, is shown as orange sticks.

Figure 3. Tgt polar contacts in (A) 2be5 (pdb.org) and (B) the refined 2be5 final (cmbi.ru.nl/pdb_redo/) structures. Tgt is shown as purple sticks, amino acids forming polar contacts (red dashed lines) as green sticks, Mg²⁺ ions as magenta spheres, and water molecules as dark blue spheres.