Richardson, William
(2022)
Structural investigation of Pseudomonas aeruginosa quorum sensing receptor PqsR to aid virulence inhibition.
PhD thesis, University of Nottingham.
Abstract
Pseudomonas aeruginosa represents a significant threat to public health. P. aeruginosa is a ubiquitous organism and opportunistic pathogen able to infect immunologically compromised individuals and burn patients. The organism coordinates collective action, such as the release of virulence factors, through quorum sensing (QS). QS in P. aeruginosa is facilitate by diffusible signalling molecules from three systems: las, rhl and pqs. The latter utilises alkyl quinolones (AQ) produced by the pqsABCDE biosynthetic operon and PqsH. PqsR, a LysR-type transcriptional regulator (LTTRs), binds to 2-heptyl-3-hydroxy-4-quinolone (PQS) and its precursor 2-heptyl-4-quinolone (HHQ) to activate the pqs operon and virulence factors including pyocyanin and elastase. Therefore, disruption of this system represents a potential strategy to attenuate P. aeruginosa virulence. Prior to this thesis a large high throughput screen – as part of the SENBIOTAR research project to sensitise P. aeruginosa to antimicrobial therapeutics - was conducted on bioreporter strain to identify high potency antagonists.
In this thesis the binding of SENBIOTAR antagonists and derivatives were characterised by x- ray crystallography of the ligand binding domain (LBD) which has revealed relationships required for effective binding. The position of halogens within sub pocket B are considerably restricted and movement of this halogen appears to have a deleterious effect on potency observed in a bioreporter assay. It was also observed that π stacking interactions to Tyr 258, located in pocket A, could be exploited through several distinct geometries with concomitant effects of potency. Additionally, a fragment-based drug discovery campaign was launched using thermal shift (TSA) to identify compounds binders that were validated by isothermal titration calorimetry (ITC). Thus far literature studies have utilised more high cost SPR screening, but our efforts have shown an in-silico pre-screen and thermal shift to be effective. A benzothiazole fragment, with ligand efficiency of 0.38 was identified as a suitable starting point that could be rationally modified to increase binding affinity. However, lead-like compounds produced by growing and linking methods gave weak potency. Attempts to elucidate binding of fragments was hampered by weak resolution and difficulties outcompeting bound solvent.
To compensate for moderate-low resolution observed during crystal soaking of antagonists a crystal engineering campaign was undertaken to identify a new crystal form that diffracted to a high resolution. A total of 7 new constructs were trialled based upon 4 design different rationales: elimination of high entropy residues; disulphide locking monomer units together; improving thermostability of partially disordered loops and changing the N- terminal boundaries. Two new crystal forms were identified that elucidated previously unknown conformations of the PqsR LBD. One crystal form revealed that PqsR may have some redox sensing function via a surface exposed C108 which was confirmed by a complementation analysis to modulate protein activity. The other form revealed a canonical dimer arrangement that is adopted in most full length LTTRs structures that suggests the protein’s mode of activation. Additionally, binding pocket changes were observed that will aid further medicinal chemistry development.
To gain a better understanding of PqsR functionality attempts were made to express and purify the full length receptor, reported as insoluble in the literature. Expression of the full length using standard overexpression vectors did not appear to yield soluble protein and native expression in P. aeruginosa failed to produce the protein that could be captured by affinity chromatography. Larger tags such as Maltose binding protein (MBP) produced aggregated material, but it was found that smaller crystal carriers developed in-house were able to effectively solubilise PqsR. Subsequent crystallisation trials identified a lithium sulphate condition that gave rise to a plate crystal habit. Despite extensive optimisation the diffraction of these crystals could not be substantially enhanced >4 Å and a suitable phase solution could not be attained by molecular replacement. Furthermore, through gel filtration, ITC and complementation experiments it was concluded the protein lacked biological activity. However, examination of truncations at positions 310 and 296 found that the disordered C-terminal was partially dispensable for function.
Finally, through a screening campaign we examined the druggability of PqsA, an anthranilate ligase required to catalyse the first step in AQ synthesis. Screening of 496 compounds yielded a single inhibitor, found to function non-competitively in the micromolar range. In P. aeruginosa the inhibitor reduced pqsA promoter activation by ~50% with an IC50 of 0.68 μM. Subsequent testing of pyocyanin, eDNA and AQ production
revealed a lack of significant inhibition. The attachment phenotype appeared to be altered with treated cells more prone to aggregate to plastic surfaces. Molecular docking was used to analyse molecular interactions which elucidated a probable binding event to the active site. The highest scoring docking poses showed the inhibitor to occupy the same position as the anthraniloyl – adenosine monophosphate intermediate product.
In conclusion, this thesis has broadened the structural characterisation of PqsR antagonists to drive further medicinal chemistry optimisation. The discovery that similar chemical scaffolds, such as the quinazolinone, can demonstrate subtle structural differences can aid in defining chemical space for antagonist growth. Additionally, this thesis has identified and validated a low-cost thermal shift assay for PqsR fragment discovery and elucidate new fragment starting points with high ligand efficiency. To generate superior diffracting crystals for ligand determination 2 new crystal forms have also been generated that shed light on LTTR activation and dimer interface rearrangement. Purification of the full length PqsR showed it adopted a tetrameric arrangement canonical of LTTRs but the protein failed to produce highly ordered crystals. Further optimisation of purification and crystallisation strategies will be required for structural determination.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[thumbnail of WRichardson2021v3Final.pdf]](https://eprints.nottingham.ac.uk/style/images/fileicons/application_pdf.png "WRichardson2021v3Final.pdf")