Chemogenomic approaches to antibiotic drug discovery
This research has been funded by several different agencies, CIHR-Regional Priority (2011-2013), Cystic Fibrosis Canada (CFC) (2013-2017), Research Manitoba (2017-2019) a University of Manitoba Collaborative Research Grant (UCRP) (2017-2019) and the Cystic Fibrosis Foundation (2018-2020).
The research strategy of this program is two-pronged: the genomic exploration of essential processes as antibiotic targets in B. cenocepacia and the identification of the mechanism of action of novel molecules with antibacterial activity. For the first approach, we have used high-density transposon mutagenesis in B. cenocepacia to determine the essential genome of this bacterium (Gislason, Turner, et al., 2017).
The genetic tool used for transposon mutagenesis delivers a rhamnose-inducible promoter randomly into the chromosome (Bloodworth et al., 2013). Screening of transposon mutants in the presence/absence of rhamnose allowed selection of knockdown mutants, in which the rhamnose-inducible promoter controls an essential gene. This system allowed us to construct a bacterial library that holds hundreds of knockdown mutants in essential genes. An angle of this research program is the exploration of some of the essential proteins identified, as putative targets for antibiotic therapy (Bloodworth et al., 2015; Stietz et al., 2017; Cardona et al., 2018).
Essential gene libraries are great tools to determine the mechanism of action of novel antimicrobials as they observe chemical genetic interactions with antibiotics. The underlying principle is that knockdown mutants that are directly or indirectly related to the mechanism of action have differential fitness in the presence of the drug, in comparison with unrelated knockdowns.
A bottleneck, however, is the laborious process of screening hundreds of mutants against several antimicrobials. To increase the throughput of our method, we have recently tracked the relative fitness of a pooled knockdown mutant library in the presence of antibiotics by Illumina sequencing. This work demonstrated that the specific mutant depletion in response to antibiotics is enhanced when the mutants are grown competitively as a pool, compared with individual growth (Gislason, Choy, et al., 2017). Using Tn-seq to track the knockdown mutants in the presence of antibiotics, we have identified the cell division protein FtsZ as the target of C109, a novel antimicrobial molecule with broad-spectrum activity (Hogan et al., 2018).
Catabolic control of pathogenicity
This research has been continuously funded by the NSERC Discovery Grant Program (2006-2010, 2011-2015, 2016-2020).
When bacteria cause infection, they utilize host nutrients for growth. The type and availability of nutrients found by pathogenic bacteria can influence their virulence, but the mechanisms are not always understood. Most species of the Bcc colonize the lungs of people with cystic fibrosis producing a spectrum of clinical manifestations, ranging from asymptomatic carriage to the lethal “cepacia syndrome.” Despite intense clinical research, the reasons why some cystic fibrosis patients succumb to Bcc infections while others do not are unknown. A major source of carbon and nitrogen source supporting the cystic fibrosis microbiota are the amino acids present in the cystic fibrosis mucus. Using a species of Bcc, B. cenocepacia, we are elucidating how this bacterium senses changes in amino acid availability eliciting different pathogenic responses.
We have identified that a protein that forms the flagellum, flagellin is one of the most upregulated proteins of a species of Bcc, B. cenocepacia in conditions that mimic the nutrients of the cystic fibrosis mucus. We further confirmed that the cystic fibrosis nutritional conditions induced swimming motility and flagellin expression (Kumar and Cardona, 2016). Notably, B. cenocepacia K56-2 exhibited an increase in the number of flagella per cell. On the contrary, when grown in minimal medium with glucose, cells presented only one polar flagellum. Intriguingly, arginine and glutamate-induced swimming motility and protease activity and this response was mediated by a decrease of a global messenger molecule in bacteria, c-di-GMP (Kumar et al., 2018).
We realized that amino acid metabolism has other effects on the Bcc bacteria pathogenicity when we found a link between phenylalanine metabolism and virulence of B. cenocepacia (Law et al., 2008). In B. cenocepacia, phenylalanine funnels to the tricarboxylic acid (TCA) cycle through the phenylacetic acid degradation pathway. One of the steps of this pathway, the epoxidation of phenylacetyl-CoA by the action of the PaaABCDE complex can occur only in aerobic conditions. Therefore, we expect that the anoxic microenvironments of the CF lung will cause intracellular accumulation of phenylacetyl-CoA in B. cenocepacia if phenylalanine is available as a carbon source. Remarkably, deletion of the paaABCDE gene cluster attenuated virulence in B. cenocepacia in a quorum sensing-dependent manner (Pribytkova et al., 2014). The paaABCDE mutant accumulates phenylacetyl-CoA, suggesting that this molecule may be responsible for direct or indirect inhibition of quorum sensing.
Taken together, we have found three overlapping virulence mechanisms that are triggered by amino acids: i) The number of flagella increases in media containing amino acids; ii) Swimming motility and protease activity is upregulated in response to arginine and glutamate by a decrease of c-di-GMP signaling; and iii) pathogenicity is dependent on phenylalanine and phenylacetic acid degradation because accumulation of phenylacetyl-CoA reduces quorum sensing-related virulence. The intertwining of mechanisms with different effects on pathogenicity in response to amino acid content suggests that host and microbiome-derived variations in amino acid content may be factors that play a role in the variable clinical outcomes observed in Bcc infections.
Bcc bacteria hold extraordinary metabolic capacities that can be exploited in biotechnological applications. My laboratory has developed several genetic tools that can be applied in gene and genome editing to manipulate desired features in Burkholderia strains while reducing their pathogenicity. The latest technology we are developing is CRISPR-based interference of gene expression. This novel technique, which has been applied to a limited number of bacteria but needs to be customized to specific characteristics of Bcc bacteria.
In collaboration with Dr. David Levin, From Biosystems Engineering, U. of Manitoba, we are looking into the ability of Bcc bacteria to degrade polyesters used in the production of biodegradable plastics. Once the most active strains are identified, we will study the specificity of the bioplastic degrading enzymes, and the genetic elements involved. The next step will be to edit the strain’s genome to produce a strain active regarding biodegradation, but defective in terms of pathogenicity.