From Data to Decisions: Leveraging Artificial Intelligence and Machine Learning in Combating Antimicrobial Resistance - a Comprehensive Review.

The emergence of drug-resistant bacteria poses a significant challenge to modern medicine. In response, Artificial Intelligence (AI) and Machine Learning (ML) algorithms have emerged as powerful tools for combating antimicrobial resistance (AMR). This review aims to explore the role of AI/ML in AMR management, with a focus on identifying pathogens, understanding resistance patterns, predicting treatment outcomes, and discovering new antibiotic agents. Recent advancements in AI/ML have enabled the efficient analysis of large datasets, facilitating the reliable prediction of AMR trends and treatment responses with minimal human intervention. ML algorithms can analyze genomic data to identify genetic markers associated with antibiotic resistance, enabling the development of targeted treatment strategies. Additionally, AI/ML techniques show promise in optimizing drug administration and developing alternatives to traditional antibiotics. By analyzing patient data and clinical outcomes, these technologies can assist healthcare providers in diagnosing infections, evaluating their severity, and selecting appropriate antimicrobial therapies. While integration of AI/ML in clinical settings is still in its infancy, advancements in data quality and algorithm development suggest that widespread clinical adoption is forthcoming. In conclusion, AI/ML holds significant promise for improving AMR management and treatment outcome.

A large-scale whole-genome comparison shows that experimental evolution in response to antibiotics predicts changes in naturally evolved clinical Pseudomonas aeruginosa.

Pseudomonas aeruginosa is an opportunistic pathogen that causes a wide range of acute and chronic infections. An increasing number of isolates have mutations that make them antibiotic resistant, making treatment difficult. To identify resistance-associated mutations we experimentally evolved the antibiotic sensitive strain P. aeruginosa PAO1 to become resistant to three widely used anti-pseudomonal antibiotics, ciprofloxacin, meropenem and tobramycin. Mutants could tolerate up to 2048-fold higher concentrations of antibiotic than strain PAO1. Genome sequences were determined for thirteen mutants for each antibiotic. Each mutant had between 2 and 8 mutations. For each antibiotic at least 8 genes were mutated in multiple mutants, demonstrating the genetic complexity of resistance. For all three antibiotics mutations arose in genes known to be associated with resistance, but also in genes not previously associated with resistance. To determine the clinical relevance of mutations uncovered in this study we analysed the corresponding genes in 558 isolates of P. aeruginosa from patients with chronic lung disease and in 172 isolates from the general environment. Many genes identified through experimental evolution had predicted function-altering changes in clinical isolates but not in environmental isolates, showing that mutated genes in experimentally evolved bacteria can predict those that undergo mutation during infection. Additionally, large deletions of up to 479kb arose in experimentally evolved meropenem resistant mutants and large deletions were present in 87 of the clinical isolates. These findings significantly advance understanding of antibiotic resistance in P. aeruginosa and demonstrate the validity of experimental evolution in identifying clinically-relevant resistance-associated mutations.

Targeting the Pseudomonas aeruginosa Virulence Factor Phospholipase C With Engineered Liposomes.

Engineered liposomes composed of the naturally occurring lipids sphingomyelin (Sm) and cholesterol (Ch) have been demonstrated to efficiently neutralize toxins secreted by Gram-positive bacteria such as Streptococcus pneumoniae and Staphylococcus aureus. Here, we hypothesized that liposomes are capable of neutralizing cytolytic virulence factors secreted by the Gram-negative pathogen Pseudomonas aeruginosa. We used the highly virulent cystic fibrosis P. aeruginosa Liverpool Epidemic Strain LESB58 and showed that sphingomyelin (Sm) and a combination of sphingomyelin with cholesterol (Ch:Sm; 66 mol/% Ch and 34 mol/% Sm) liposomes reduced lysis of human bronchial and red blood cells upon challenge with the Pseudomonas secretome. Mass spectrometry of liposome-sequestered Pseudomonas proteins identified the virulence-promoting hemolytic phospholipase C (PlcH) as having been neutralized. Pseudomonas aeruginosa supernatants incubated with liposomes demonstrated reduced PlcH activity as assessed by the p-nitrophenylphosphorylcholine (NPPC) assay. Testing the in vivo efficacy of the liposomes in a murine cutaneous abscess model revealed that Sm and Ch:Sm, as single dose treatments, attenuated abscesses by >30%, demonstrating a similar effect to that of a mutant lacking plcH in this infection model. Thus, sphingomyelin-containing liposome therapy offers an interesting approach to treat and reduce virulence of complex infections caused by P. aeruginosa and potentially other Gram-negative pathogens expressing PlcH.

The Effects of Sub-inhibitory Antibiotic Concentrations on Pseudomonas aeruginosa: Reduced Susceptibility Due to Mutations.

Pseudomonas aeruginosa chronically infects in the lungs of people with cystic fibrosis and other forms of lung disease. Infections are treated with antibiotics, but over time, the bacteria acquire mutations that reduce their antibiotic susceptibility. The effects of inhibitory amounts of antibiotics in selecting for antibiotic-resistant mutants have been well studied. However, the concentrations of antibiotics that reach infecting bacteria can be sub-inhibitory and but may nonetheless promote emergence of antibiotic-resistant bacteria. Therefore, the aim of this research was to investigate the effects of sub-inhibitory concentrations of antibiotics on the antibiotic susceptibility of P. aeruginosa. Two P. aeruginosa reference strains, PAO1 and PA14, and six isolates from individuals with cystic fibrosis were studied. The bacteria were passaged in the presence of antibiotics (ceftazidime, ciprofloxacin, meropenem or tobramycin) at sub-inhibitory amounts. Fifteen populations of bacteria (up to five per strain) were exposed to each of the four antibiotics. Antibiotic susceptibility was determined following 10 passages on agar supplemented with antibiotic and compared with susceptibility prior to antibiotic exposure. Antibiotic exposure resulted in susceptibility being significantly (>2-fold) reduced for 13 of the 60 populations. Seven samples had reduced susceptibility to ciprofloxacin, three to tobramycin, two to ceftazidime and one to meropenem. Whole-genome sequencing revealed the mutations arising following antibiotic exposure. Mutants with reduced antibiotic susceptibility had mutations in genes known to affect antibiotic resistance, including regulators of efflux pumps (mexR, mexS, mexZ and nalC) and the fusA1 gene that is associated with aminoglycoside resistance. Genes not previously associated with resistance, including gacS, sigX and crfX and two genes with no known function, were also mutated in some isolates with reduced antibiotic susceptibility. Our results show that exposure to sub-inhibitory amounts of antibiotics can select for mutations that reduce the susceptibility of P. aeruginosa to antibiotics and that the profile of mutations is different from that arising during selection with inhibitory antibiotic concentrations. It is likely that exposure to sub-inhibitory amounts of antibiotics during infection contributes to P. aeruginosa becoming antibiotic-resistant.

Genome evolution drives transcriptomic and phenotypic adaptation in Pseudomonas aeruginosa during 20 years of infection.

The opportunistic pathogen Pseudomonas aeruginosa chronically infects the lungs of patients with cystic fibrosis (CF). During infection the bacteria evolve and adapt to the lung environment. Here we use genomic, transcriptomic and phenotypic approaches to compare multiple isolates of P. aeruginosa collected more than 20 years apart during a chronic infection in a CF patient. Complete genome sequencing of the isolates, using short- and long-read technologies, showed that a genetic bottleneck occurred during infection and was followed by diversification of the bacteria. A 125 kb deletion, an 0.9 Mb inversion and hundreds of smaller mutations occurred during evolution of the bacteria in the lung, with an average rate of 17 mutations per year. Many of the mutated genes are associated with infection or antibiotic resistance. RNA sequencing was used to compare the transcriptomes of an earlier and a later isolate. Substantial reprogramming of the transcriptional network had occurred, affecting multiple genes that contribute to continuing infection. Changes included greatly reduced expression of flagellar machinery and increased expression of genes for nutrient acquisition and biofilm formation, as well as altered expression of a large number of genes of unknown function. Phenotypic studies showed that most later isolates had increased cell adherence and antibiotic resistance, reduced motility, and reduced production of pyoverdine (an iron-scavenging siderophore), consistent with genomic and transcriptomic data. The approach of integrating genomic, transcriptomic and phenotypic analyses reveals, and helps to explain, the plethora of changes that P. aeruginosa undergoes to enable it to adapt to the environment of the CF lung during a chronic infection.

Genomic and phenotypic comparison of environmental and patient-derived isolates of Pseudomonas aeruginosa suggest that antimicrobial resistance is rare within the environment.

Patient-derived isolates of the opportunistic pathogen Pseudomonas aeruginosa are frequently resistant to antibiotics due to the presence of sequence variants in resistance-associated genes. However, the frequency of antibiotic resistance and of resistance-associated sequence variants in environmental isolates of P. aeruginosa has not been well studied. Antimicrobial susceptibility testing (ciprofloxacin, ceftazidime, meropenem, tobramycin) of environmental (n=50) and cystic fibrosis (n=42) P. aeruginosa isolates was carried out. Following whole genome sequencing of all isolates, 25 resistance-associated genes were analysed for the presence of likely function-altering sequence variants. Environmental isolates were susceptible to all antibiotics with one exception, whereas patient-derived isolates had significant frequencies of resistance to each antibiotic and a greater number of likely resistance-associated genetic variants. These findings indicate that the natural environment does not act as a reservoir of antibiotic-resistant P. aeruginosa, supporting a model in which antibiotic susceptible environmental bacteria infect patients and develop resistance during infection.