Gene regulatory network inference and analysis of multidrug-resistant Pseudomonas aeruginosa

Background

Pseudomonas aeruginosa is a human pathogen that is a major cause of hospital- acquired infections (HAIs), primarily in the skin, lower respiratory tract, urinary tract and eyes, infections that can all lead to more serious manifestations, including bacteremia and endocarditis (1). Of particular concern is the prevalence of multi-drug resistance (MDR) strains, lending to the inclusion of P. aeruginosa as one of the ESKAPE pathogens, a set of MDR bacterial pathogens that exert great economic and social burdens. One such strain is the CCBH4851 which has caused multiple outbreaks in Brazil in recent years, where 35% of strains tested were resistant to carbapenems (a common treatment for P. aeruginosa infection) (2). A plethora of encoded resistance mechanisms confer resistance in P. aeruginosa, including efflux pumps, β-lactamases and biofilm formation, which can all be acquired either endogenously due to mutation or via horizontal gene transfer (2) .

Together, these resistance determinants, alongside less specific regulatory architectures that control their activity, provide elements of redundancy and synergy; however, the mechanisms which underpin them remain poorly understood. This is primarily due to the limitations of individual gene studies and reductionist approaches, which fail to capture the full scale of the networks regulating these processes and other cellular behaviors. In recent years, more holistic analysis methods have been applied to better grasp these networks, and one such set of methodologies is Graph theory (3). It is a set of mathematical doctrines which focuses on the construction of a network model, or a graphical representation of the regulatory network. Such network models aim to simulate the relationships and interactions between various genes, in order to infer underlying regulatory circuits and interactions that can be used to further understand the bacterium or to propose possible drug targets. When applying graph theory, genes take the form of nodes, connected by edges which represent the interaction between genes. Network topology, connectivity and clustering of nodes are all metrics that can then be quantified and used to further describe or infer biological processes and regulatory mechanisms (4). As such, the authors of this paper used existing databases and adapted annotations of other P. aeruginosa strains to CCBH4851 to generate a network model, through which the authors were able to gain insights into the bacterium’s pathology and resistance mechanisms (5).

Key findings

• They identified a new sub-network not previously described that regulates the polysaccharide synthesis locus. This locus has roles relevant to pathology and resistance, including drug efflux and biofilm formation.

• They were able to improve on previous network models that focus on the more benign, well-characterised PAO1 strain and were able to identify novel network properties (6). For example, they determined that ~50% of genes had an activatory effect on other genes, which they suggest is important for processes that, once started, need to go to completion. Many of the processes they regulate fit this category, including pathogenic traits, adhesion, virulence, resistance factors and biofilm formation.

• Many genes undergo auto-regulation, where gene’s can influence their own expression via either up-regulation or down-regulation. However they identified that the majority of genes that exhibit auto-regulation have a negative effect, meaning that they repress their own expression, which provides robustness for the system to external fluctuations. Most of these genes were associated with factors regulating biosynthesis and they suggest that negative autoregulation enables P. aeruginosa to prevent wastage of resources.

• Finally, they identified various “hub” genes, or genes that interact with many other genes, including lasR, fur and np20. These are ideal targets for drugs as their disruption will have a greater effect on cellular behaviour than less connected genes.

Why I chose this paper

Given the looming threat of prevalent anti-microbial resistance combined with the low rate of novel antibiotic discovery, we are in dire need of new methodologies to identify novel drug targets. This paper represents one such method that is able to i) identify and understand key regulatory mechanisms that are relevant to pathogenesis and ii) propose a list of candidate hub genes that are possible targets for drug development. Moreover, they were able to perform this analysis entirely computationally, only utilizing previously published data meaning that such studies are low-cost and relatively easy to scale. Such holistic analyses have proved effective in understanding the behavior of viral infections and cancers and providing guidance for drug development. As such, I think that these studies will become much more abundant in future years, given the abundance of publicly-available data and the wealth of insight that such analyses provide (3,4,7).

Future perspectives

• I think further investigation into the “druggability” of some of these hub genes would be beneficial, alongside application of the understanding acquired to better prevent and control infections by P. aeruginosa, particularly in cystic fibrosis patients.

• I think it would be beneficial to apply this analysis to other bacterial pathogens of the ESKAPE classification, given that these too are huge burdens on the human population.

• Deletion and mutation of various hub genes and fitness studies could be performed to further determine ideal candidates.

• Given that hub genes are most likely to be evolutionary ancient and thus more likely to share homology with human orthologs, phylogenetic analysis and sequence comparison could be performed to further filter candidates that are inappropriate due to similarity with human genes.

Bibliography

1. Lyczak JB, Cannon CL, Pier GB. Establishment of Pseudomonas aeruginosa infection: lessons from a versatile opportunist. Microbes Infect [Internet]. 2000 Jul 1 [cited 2019 Jul 9];2(9):1051–60. Available from: https://www.sciencedirect.com/science/article/pii/S1286457900012594

2. Maraolo AE, Cascella M, Corcione S, Cuomo A, Nappa S, Borgia G, et al. Management of multidrug- resistant Pseudomonas aeruginosa in the intensive care unit: state of the art. Expert Rev Anti Infect Ther [Internet]. 2017 Sep 2 [cited 2019 Jul 9];15(9):861–71. Available from: https://www.tandfonline.com/doi/full/10.1080/14787210.2017.1367666

3. Pavlopoulos GA, Secrier M, Moschopoulos CN, Soldatos TG, Kossida S, Aerts J, et al. Using graph theory to analyze biological networks. BioData Min [Internet]. 2011 Dec 28 [cited 2019 Jul 9];4(1):10. Available from: https://biodatamining.biomedcentral.com/articles/10.1186/1756-0381-4-10

4. Arrell DK, Terzic A. Network Systems Biology for Drug Discovery. Clin Pharmacol Ther [Internet]. 2010 Jul 2 [cited 2019 Jul 9];88(1):120–5. Available from: http://doi.wiley.com/10.1038/clpt.2010.91

5. Filho FM, Nascimento APB do, Santos MT dos, Carvalho-Assef APD, Silva FAB da. Gene regulatory network inference and analysis of multidrug-resistant Pseudomonas aeruginosa. bioRxiv [Internet]. 2019 Jun 28 [cited 2019 Jul 9];610493. Available from: https://www.biorxiv.org/content/10.1101/610493v2.full

6. Stover CK, Pham XQ, Erwin AL, Mizoguchi SD, Warrener P, Hickey MJ, et al. Complete genome sequence of Pseudomonas aeruginosa PAO1, an opportunistic pathogen. Nature [Internet]. 2000 Aug [cited 2019 Jul 9];406(6799):959–64. Available from: http://www.nature.com/articles/35023079

7. Mason O, Verwoerd M. Graph theory and networks in Biology. IET Syst Biol [Internet]. 2007 Mar 1 [cited 2019 Jul 9];1(2):89–119. Available from: https://digital- library.theiet.org/content/journals/10.1049/iet-syb_20060038

 

 

doi: https://doi.org/10.1242/prelights.12282

Read preprint (No Ratings Yet)