Flipping states: a few key residues decide the winning conformation of the only universally conserved transcription factor

Abstract

Transcription factors from the NusG family bind to the elongating RNA polymerase to enable synthesis of long RNAs in all domains of life. In bacteria, NusG frequently co-exists with specialized paralogs that regulate expression of a small set of targets, many of which encode virulence factors. Escherichia coli RfaH is the exemplar of this regulatory mechanism. In contrast to NusG, which freely binds to RNA polymerase, RfaH exists in a structurally distinct autoinhibitory state in which the RNA polymerase-binding site is buried at the interface between two RfaH domains. Binding to an ops DNA sequence triggers structural transformation wherein the domains dissociate and RfaH refolds into a NusG-like structure. Formation of the autoinhibitory state, and thus sequence-specific recruitment, represents the decisive step in the evolutionary history of the RfaH subfamily. We used computational and experimental approaches to identify the residues that confer the unique regulatory properties of RfaH. Our analysis highlighted highly conserved Ile and Phe residues at the RfaH interdomain interface. Replacement of these residues with equally conserved Glu and Val counterpart residues in NusG destabilized interactions between the RfaH domains and allowed sequence-independent recruitment to RNA polymerase, suggesting a plausible pathway for diversification of NusG paralogs.

Figure 1.

Regulatory mechanisms of RfaH and NusG.

Figure 2.

The key features of the RfaH and NusG families. (A) Structural alignment of E. coli RfaH and NusG is shown in the middle. The NTD alignment was derived from the superposition of PDBs 2OUG (RfaH) and 2K06 (NusG), and the CTD alignment was derived from the superposition of PDBs 2LCL (RfaH) and 2KVQ (NusG). A profile above the alignment was generated from the sequence alignment of 751 RfaH sequences, while profile underneath was generated from the sequence alignment of 9204 NusG sequences. Red circles in the middle indicate the ΔΔG_bind value; large, >1.5 kcal/mol, small, 1–1.5 kcal/mol. IDI contact areas are shown as blue circles; large, IDI contact areas >50 Ų, small, <50 Ų. The residues with large IDI contact areas and ΔΔG_bind are shaded in magenta and labeled with the residue number in RfaH.

Figure 3.

Effects of RfaH variants on pausing at the ops site. (Top) Transcript generated from the T7A1 promoter on a linear DNA template; transcription start site (a bent arrow), ops element (gray box), and transcript end are indicated on top. (Bottom) Halted A24 TECs were formed as described in Materials and Methods. Elongation was restarted upon addition of NTPs and rifapentin in the presence of the indicated transcription factor. Aliquots were withdrawn at times indicated above each lane (in s) and analyzed on an 8% denaturing gel. Positions of the paused and run-off transcripts are indicated with arrows; the position of the RfaH-induced pause at G39, with a circle. Pausing at ops (U38; fraction of total RNA) and arrival at the C71 position (fraction of final at 180 sec) were quantified to assess the anti-pausing effects of elongation factors; 30-s values are shown below each panel. The experiment was repeated three times; errors were <15%.

Figure 4.

Effects of RfaH variants on pausing in the absence of ops. (A) The experiment was performed as in Figure 3, except that a mutant ops element, with G8 substituted for a C (white oval), was used. (B) Arrival at the C71 position was quantified; the error bars are omitted for clarity. A representative example (30-s) is shown below in panel A, along with the fraction of U38 RNA; errors are standard deviations calculated from three repeats.

Figure 5.

Probing the RfaH domain dissociation by chymotrypsin digestion. (A) Accessibility of aromatic residues in the full length RfaH and the isolated domains. The NTD is shown in gray and the CTD in cyan; both states of the CTD are shown. The aromatic residues are shown as sticks (red in the NTD; blue in the CTD), with their surfaces hidden. This figure was prepared with Pymol 1.8.2.3 (Schrödinger, LLC) using PDB IDs 2OUG and 2LCL. (B) Chymotrypsin cleavage of selected protein variants. The assays were performed as described in Materials and Methods; the samples were analyzed on 17-well 4–12% Bis–Tris gels. The WT, 193E and F130V samples were analyzed on one gel, and the isolated domains (along with the full-length protein, not shown) on another. Chymotrypsin is visible above the uncut proteins.