Abstract:
Multi-drug resistant (MDR) Pseudomonas aeruginosa strains are increasingly becoming a cause for global concern, with the World Health Organisation (WHO) listing P. aeruginosa as one of the critical pathogens in urgent need of new antibiotics. Colistin is a last resort drug for Gram negative infections; however the nephrotoxicity of colistin and emerging resistance to it are a deterrent for treatment. Resistance to colistin by P. aeruginosa can be mediated via chromosomal mutations: including modification of lipid A and loss of lipopolysaccharides (LPS). One method to prevent the spread of resistance is to employ reliable and practical antimicrobial susceptibility testing (AST) methods for colistin. Currently, the broth microdilution method (BMD) is considered the gold standard for AST, however this method is relatively labour intensive, slow and impractical. Several commercial products are now available, which work to the same principle as the BMD, but in a more convenient and user-friendly format. There are a number of difficulties encountered specifically in colistin susceptibility testing: such as the use of different susceptibility breakpoints applied in different studies, the cationic nature of colistin, its large molecular size, and heteroresistance.
Another method of slowing the development of antimicrobial resistance (AMR) is to consider compartments of infection in clinical situations where the bacteria are receiving sub-inhibitory levels of colistin. The effects of such exposure on virulence and related traits such as biofilm formation and serum resistance after exposure are underexplored. Biofilm formation is one of the antibiotic-resistance mechanisms in P. aeruginosa, enabling persistence and survival, and low penetration of antibiotics into the bacterial community. Evolution experiments in combination with next generation sequencing (NGS) can uncover molecular mechanisms behind development of colistin resistance, and novel devices such as morbidostats can considerably speed up the process of generating resistant isolates.
The aims of this study were two fold. In order to accurately measure the level of resistance to colistin developing within the morbidostat, it was necessary to select an AST method, which can account for difficult isolates straddling the MIC breakpoint, and results in higher number of errors. Thus, we aimed to generate a comprehensive collection of colistin resistant P. aeruginosa strains with a range of MICs using the morbidostat, and to compare and evaluate five colistin susceptibility testing methods: three commercial BMD products, the gradient Etest, and the colorimetric reaction based Rapid Polymyxin Pseudomonas test (Rapid PP) against the reference BMD method, using these morbidostat generated isolates.
The second aim of the study was to generate P. aeruginosa isolates cultivated with exposure to colistin, metronidazole and a combination of the two antibiotics for 21 days, and complete RNA-Seq to uncover the transcriptional changes over time. The changes in the virulence related phenotypes were measured with isolates at four key time points.
For the first aim, a total of 131 P. aeruginosa isolates were used for colistin susceptibility test evaluation (100 colistin susceptible, 31 colistin resistant). The 31 colistin resistant isolates are derived from 18 colistin susceptible P. aeruginosa strains that evolved resistance in the morbidostat at different MIC ranges. The categorical agreement (CA) rates for MICRONAUT-S, SensiTest and Rapid Polymxyin Pseudomonas were 94.7%, 93.9%, 92.4%, respectively. The Sensititre achieved the highest CA score (96.9%), while the Etests had the lowest CA score (84%). The very major discrepancies (VMD) rates for all tests were between 3.2-67.4%. We concluded that the commercial BMD methods are a suitable alternative to the gold standard BMD method, with the Micronaut MIC strip tests providing a wide range of MICs for testing.
To investigate effects of antibiotic pressure, we cultivated P. aeruginosa in a semi-automated morbidostat device with colistin, metronidazole and a combination of the two antibiotics for 21 days, and completed RNA-Seq to compare differential gene expression over time. The susceptibility of isolates to colistin was measured using the commercial BMD method Micronaut MIC strip. Within 21 days, strains developed resistance to colistin.
After seven days of colistin exposure, strains developed an ability to grow in serum, indicating that colistin is a driver of bacterial modifications which protect against serum complement factors. At Day 21, we saw the biggest increase in serum resistance, with no significant changes in susceptibility to serum found in isolates exposed to only metronidazole or in the control condition, with LB medium. Colistin-resistant strains also show significantly increased biofilm formation: the cell density in biofilm increases under exposure to colistin, while metronidazole modulates this effect. Notably, strains exposed to colistin showed a decrease in virulence, when measured using the Galleria mellonella infection model. After 21 days there was a significant attenuation of virulence potential compared to the baseline strain. This effect was not seen in the combination drug condition, with metronidazole seemingly having a modulatory effect on the impact of colistin exposure on virulence.
These phenotypic changes were underlined by a series of differential gene expression changes, in particular LPS modifications, spermidine synthesis (via speH and speE) and the major stress response regulator rpoS.
Our results suggest a clinically important bacterial evolution under sub-lethal antibiotic concentration leading to potential for significant changes in the clinical course of infection.
antibiotic resistance
Print-on-Demand