Snapshots of Pseudomonas aeruginosa SOS response reveal structural requisites for LexA autoproteolysis

Abstract

Antimicrobial resistance poses a severe threat to human health and Pseudomonas aeruginosa stands out among the pathogens responsible for this emergency. The SOS response to DNA damage is crucial in bacterial evolution, influencing resistance development and adaptability in challenging environments, especially under antibiotic exposure. Recombinase A (RecA) and the transcriptional repressor LexA are the key players that orchestrate this process, determining either the silencing or the active transcription of the genes under their control. By integrating state-of-the-art structural approaches with in vitro binding and functional assays, we elucidated the molecular events activating the SOS response in P. aeruginosa, focusing on the RecA-LexA interaction. Our findings identify the conserved determinants and strength of the interactions that allow RecA to trigger LexA autocleavage and inactivation. These results provide the groundwork for designing novel antimicrobial strategies and exploring the potential translation of Escherichia coli-derived approaches, to address the implications of P. aeruginosa infections.

Keywords: Biological sciences; Biophysics; Microbiology; Natural sciences.

Graphical abstract

Figure 1

Structural analysis of LexA_Pa^CTD (A) Analytical size exclusion chromatography of LexA_PaS125A (blue) and LexA_Pa^CTDG91D (yellow; chromatograms on the left and standard curve interpolation on the right). (B) SDS-PAGE-based RecA_Pa∗-induced autoproteolysis assay of 4 LexA_Pa variants: full-length LexA_Pa, either wt or S125A inactive mutant, and LexA_Pa^CTD, either wt or G91D uncleavable mutant. One representative gel is shown, see also Figure S1E for band quantification. (C) Overall view of the LexA_Pa^CTDG91D dimer (chains A and B), as revealed by X-ray crystallography. The catalytic dyad (S125/K162) and the mutated self-cleavage site (A90-D91) of each monomer are shown as orange sticks. Boxed regions are zoomed in panels D and E. Superposed (transparent green cartoon) is the closed conformation of LexA_Pa cleavable loop found in LexA_PaS125A bound to RecA_Pa∗. (D) Detailed view of the cleavable loop (chain A) in the “open” (inactive) conformation. Hydrogen bonds engaging the residues of the loop are represented as dashed lines, while residues involved in a hydrophobic cluster are depicted as orange sticks. (E) Detailed views of the homodimerization surface of LexA_Pa^CTD. Dashed lines indicate H-bonds, salt bridges and cation-π interactions, while residues involved in a hydrophobic cluster are depicted as orange sticks. (F) LexA_Pa cleavable loop in the “closed” (active) conformation. Dashed lines indicate H-bonds stabilizing the loop in this state, while orange sticks correspond to the catalytic dyad and to the hydrophobic residues indicated in panel B. The movement of the loop brings the cleavage site inside the catalytic pocket and at the same time opens a hydrophobic cavity (Y117, L119, V159, I85) that hosts I94 in the open conformation.

Figure 2

Cryo-EM structure of RecA_Pa∗ (A) RecA_Pa∗ cryo-EM density map. (B–D) Coloring of density regions corresponding to RecA_Pa∗ protomers and (C and D) zoom on the atomic model (two perpendicular views). (E) Zoom on two adjacent RecA_Pa∗ protomers assembled on ssDNA (RecA_Paⁿ and RecA_Paⁿ⁺¹, moving from 5′ to 3′ on ssDNA). Detailed views of the cryo-EM map around ssDNA (F) and ATPγS (G), and RecA_Pa residues interacting with them.

Figure 3

Cryo-EM structure of RecA_Pa∗-LexA_PaS125A (A) Cryo-EM density map of the RecA_Pa-LexA_PaS125A complex. The displayed map has been locally sharpened using LocScale2. (B) Coloring of density regions corresponding to RecA_Pa∗ protomers (purple tones) and LexA_Pa CTD chains A (yellow) and B (orange). The boxed region represents a low-resolution density, which was not interpreted by the atomic model and that might be due to the LexA_Pa NTD. (C and D) (C) Side and (D) front views of the RecA_Pa∗-LexA_PaS125A atomic model. The dashed line in panel C represents a virtual plane where the model was cut in panel D to allow LexA_Pa clear visualization. (E) Electrostatic surface potential of RecA_Pa∗ and LexA_Pa^CTD, showing complementarity on the interacting surfaces. (F) LexA_Pa^CTD dimer and the main binding determinants on four RecA_Pa protomers (chains G–J), zoomed in panels (G–J). (K) Details of the interfaces buried between LexA_Pa and different RecA_Pa∗ protomers. (L and M) The corresponding interacting surfaces are represented in panels (L) (on RecA_Pa∗ surface) and (M) (on LexA_Pa surface, front and side views). Contour lines are colored as the interacting chain.

Figure 4

Analysis of RecA_Pa interactions with its natural ligands (ATPγS, ssDNA and LexA_Pa) (A–C) FP-based titrations of (A) FAM-32mer ssDNA with RecA_Pa (ATPγS in molar excess), (B) RecA_Pa/FAM-32mer ssDNA with ATPγS and (C) FlAsH-LexA_Pa^CTDS125A with activated RecA_Pa (RecA_Pa∗, RecA_Pa/ssDNA/ATPγS). Points represent the average of three replicates while error bars represent standard errors of the mean (SEM). (D) Relative activity of RecA_Pa variants (wt, M201A, F202A) in terms of oligomerization on ssDNA and induction of FlAsH-LexA_Pa^CTD autoproteolysis. Bars represent averages of three replicates ±SEM. See also Figure S5 for time-course traces. (E) Overview of the model proposed for the molecular process promoted by RecA_Pa∗, that leads to the autocleavage of LexA_Pa. LexA_Pa can bind RecA_Pa∗ if it is free from DNA and with the cleavable loop in the closed conformation. The binding to RecA_Pa∗ allows the self-cleavage of LexA_Pa, that otherwise is mainly prevented.