Protein crystallography provides the most detailed and compelling account of the structure and function of biological macromolecules and leads to a better understanding of all cellular processes. Building on the experience we have accumulated during previous crystallographic studies covering a wide range of proteins, protein-ligand and protein-nucleic acid complexes from various biological pathways, since 2002 my lab focused on structural analysis of transcription, using crystallography as a tool and RNA polymerase (RNAP) as a major target. Crystallographic studies of transcription are particularly challenging, as most of the proteins from this pathway appear to be very flexible and unstable, thus being refractory to crystallizations and structure determination. Together with the ribosome, multi-subunit RNAPs are among the biggest asymmetric macromolecules studied by X-ray diffraction so far and are therefore at the cutting edge of modern protein crystallography. The 2.6 A resolution crystal structure of a bacterial multi-subunit RNAP holoenzyme (MW ~450kDa) provided key insights into the mechanism of transcription initiation [Vassylyev et al. (2002), Nature]. The high resolution structure of the single-subunit T7 RNAP elongation complex (EC, MW ~120kDa) shed significant light on the basic principles of transcription elongation [Tahirov et al. (2002), Nature]. The atomic structure of the T7 RNAP EC with the incoming substrate analog revealed an intriguing two-step mechanism of substrate selection that might be common for all RNAPs [Temiakov et al. (2004), Cell]. The high resolution structure of a bacterial RNAP holoenzyme in complex with the "magic spot" (ppGpp) yielded a new model of transcriptional regulation during stringent control, an adaptive response of bacteria to amino acid starvation [Artsimovitch et al. (2004), Cell]. Subsequently, the crystal structure of the DksA protein, known to amplify the ppGpp effect, allowed us to propose and verify a detailed molecular mechanism for the DksA/ppGpp synergism [Perederina et al. (2004), Cell]. The structure of the Gfh1 protein that belongs to a family of the transcript cleavage Gre-factors, but possesses distinct effects on transcription revealed an unexpected structural variability of the protein that likely determines its unique functional properties [Symersky, et al. (2005), JBC].
The mechanism of substrate loading in multi-subunit RNA polymerase (RNAP) is crucial for understanding the general principles of transcription yet remains hotly debated. We have determined the 3.0Å resolution structures of the T. thermophilus EC with a non-hydrolyzable substrate analog, AMPcPP, and with AMPcPP plus the inhibitor streptolydigin (Stl). Our structural and biochemical data suggest that the trigger loop (TL) re-folding is vital for catalysis and have three major implications. First, despite differences in the details, the two-step, pre-insertion/insertion mechanism of substrate loading may be universal for all RNAPs. Second, freezing of the pre-insertion state is an attractive target for design of novel antibiotics. Finally, the TL emerges as a prominent target whose re-folding can be modulated by regulatory factors [Vassylev et al. (2007), Nature].
Bacterial RNAP is an attractive target for the design of antibiotics. First, despite the overall homology in structure and function of bacterial and eukaryotic RNAPs, each exhibits many distinct features, in particular, in the regulation of transcription. Second, consisting of over 3,000 amino acids, bacterial RNAP comprises a huge solvent-exposed surface that contains numerous cavities and channels, many of which are used for binding nucleic acids and/or transcription factors. Blocking off these cavities with a steric inhibitor may therefore result in disruption of transcription and eventually cell death. In order to utilize structure-based drug design of a novel antibiotics targeting RNAP one should understand the molecular mechanism of their action for which determination of the atomic structures of RNAP in complexes with the promising inhibitors is of central importance. To this end we have determined the high resolution crystal structures of the RNAP holoenzyme complexed with four bacteria-specific inhibitors: rifapentin and rifabutin (clinically important compounds from the rifamycin line of antibiotics) [Artsimovitch et al., (2005), Cell]., streptolydigin [Temiakov et al., (2005), Mol. Cell]., tagetin [Vassylyev et al., (2005), Nature Struc. Mol. Biol.], myxopyronin [Belogurov et al., (2008), Nature]. These works shed considerable light on the mechanistic aspects of the transcription regulation by these compounds thereby allowing intellectual improvement of their inhibitory properties.