Structure and Dynamics of Sigma Subunit of RNA Polymerase from Bacillus Subtilis

M. Zachrdla1,2, L. Žídek1,2, A. Rabatinová3 , H. Šanderová3, L. Krásný3

1NCBR, Faculty of Science, Masaryk University, Kamenice 5, 62500 Brno, Czech Republic

2CEITEC, Masaryk University, Kamenice 5, 62500 Brno, Czech Republic

3Department of Molecular Genetics of Bacteria, Institute of Microbiology, Academy of Sciences of the Czech Republic, Prague 14220, Czech Republic

RNA polymerase of gram-positive bacteria contains several unique subunits in comparison to RNA polymerase of gram-negative bacteria. Sigma subunits play a critical role in recognition of DNA promotor sequence. We focused on sigma A factor (SigA) from Bacillus subtilis. SigA belongs to group 1 transcriptional factors. SigA is composed of four domains, σ1.1, σ2, σ3, and σ4 that are connected by flexible linkers. SigA is capable of DNA recognition without interaction with other partner, therefore a regulatory mechanism to prevent DNA binding in inappropriate times has to exist. σ1.1 domain is responsible for auto-regulation of SigA.

Solution state nuclear magnetic resonance was used to solve the structure of σ1.1. Although σ1.1 is relatively small in size, 9.3 kDa, the resonance frequency assignment of σ1.1 is not a trivial task because it contains 23 glutamine or glutamic acid residues. As a result, one third of amino acids have very similar chemical shifts and therefore the standard set of NMR experiments 2D 1H-15N HSQC, 3D HNCA, 3D HN(CO)CA, 3D HNCACB, 3D HN(CO)CACB for sequential resonance frequency assignment can lead to ambiguous assignment. This problem was solved using 3D HCCCONH experiments. Side-chain assignment was done using additional 3D HCCH-TOCSY for aromatic and aliphatic spectral regions and 3D HSQC-TOCSY. We obtained backbone assignment for all but two N-terminal residues. Side-chain assignmnent of 97% nuclei was obtained not including His-tag. All experiments were performed on 700 MHz and 850 MHz spectrometers.

Set of 15N-edited 3D HSQC-NOESY experiment and 13C-edited 3D HSQC-NOESY experiments  for aromatic and aliphatic spectral regions was acquired to obtain proton-proton distances from experiments based on nuclear Overhauser effect (NOE). The NOESY cross-peak assignment was done using program CANDID. 3-bond scalar couplings, residual dipolar couplings, and NOE restrains were utilized by program CNS to calculate the structure. There is structure of one homologue protein from Thermotoga maritima available in the PDB database. However, even though the secondary structure prediction reflects a very similar pattern, our structure of σ1.1 exhibites significant differences in comparison.

In order to describe the internal motions of σ1.1 ,  we analysed auto-relaxation rates R1 and R2, longitudinal and transverse cross-correlated relaxation rates, and steady-state Nuclear Overhauser enhancement. Data was obtained on 600MHz, 850MHz, and 950MHz spectrometers. Obtained relaxation rates were used for spectral density mapping and model-free analysis.

This work was supported by Czech Science Foundation, grant number GA 13-16842S.