(B-77) Characterization of functional diversity of Pseudomonas aeruginosa clinical isolates with broadly representative phenotypes

Introduction::
Pseudomonas aeruginosa is a leading cause of infections in immunocompromised individuals in healthcare settings. The treatment of these infections is complicated by the presence of a variety of metabolic and virulence mechanisms among clinical strains. However, these differences in metabolic functions are poorly characterized. We hypothesize that the metabolic differences in P. aeruginosa are dependent on a complex combination of host and pathogen-specific factors, which can be delineated using a combination of genomic and transcriptomic analyses coupled with genome-scale metabolic modeling to identify the core and unique metabolic functions in different clinical isolates.




Materials and Methods::
To better understand the metabolic functional landscape of the clinical isolates and to inform innovative treatment strategies against this versatile pathogen, 971 clinical isolates of P. aeruginosa from 590 patients at the UVA Health System Clinical Microbiology Laboratory, with corresponding patient metadata, bacterial morphological phenotypes, and antimicrobial susceptibility profiles, were utilized. We selected 25 phenotypically representative isolates from this collection through stratified random sampling while preserving the distribution of original phenotypic characteristics during repeat cultures. These 25 isolates were then cultured in LB medium with 5% Fetal Bovine Serum for whole genome sequencing and parallel Biolog substrate utilization assays. The genome sequence data was used for comparative genomic analysis using the PA14 strain as the reference genome. A dissimilarity matrix was enumerated from the output of multiple local alignment searches to develop a phylogenetic cluster of the isolates. The genotypic clustering was compared to the phenotypic clustering generated from a multi-parametric analysis to assess the genotype-phenotype correlation. Each of the complete genomes of the isolates was annotated based on the KEGG biochemical database and a genome-scale metabolic network reconstruction was developed for each isolate through extensive amendment to an existing PA14 reconstruction, iPau21. In order to assess the impact of mucin (a key component of the physiological niches which P. aeruginosa infects), five clinical isolates along with the lab strain PA14 were selected for transcriptomic sequencing using static cultures in Synthetic Cystic Fibrosis growth medium (SCFM) ± 0.5% MUC5AC under aerobic conditions. Purified RNA was subjected to high-throughput sequencing.









Results, Conclusions, and Discussions::
Whole genome sequencing and annotation of the selected clinical isolates portrayed a complete picture of their functional diversity and uniqueness. It revealed not only how the isolates are similar or unique compared to each other as well as the reference P. aeruginosa PA14 strain, but also how the clinical isolates expand the entire metabolic landscape of the P. aeruginosa pan-genome, sharing metabolic functions from a diverse range of species. We reconstructed the metabolic networks of each of the isolates from the genome annotation and built on top of a highly curated, genome-scale metabolic model of the PA14 strain, iPau21. We enumerated 500 flux samples for each of the isolate models to evaluate their relative distance based on a non-metric multidimensional scaling method. We observed that while most of the strains share much of their metabolic profiles, all of them show some degree of unique metabolic functions. Correlation of the flux sampling results with phenotypic traits identified patient sex and strain isolation site as the most important factors in shaping the pathogen metabolism. These network reconstructions show diverse metabolic functionalities and substrate dependencies, as well as enhance the collective P. aeruginosapangenome metabolic repertoire.









Specific metabolic functions and biological processes modulated by the presence of mucin were identified using differential gene expression analysis. Furthermore, the transcriptomic datasets are being analyzed and incorporated into the genome-scale metabolic models of the clinical isolates; this integration allows us to identify the mucin-driven metabolic changes in the clinical isolates and their differences across the isolates.









Characterizing this rich set of clinical P. aeruginosa isolates allows for a deeper understanding of genotypic and metabolic diversity of the pathogen in a clinical setting and lays a foundation for further investigation of the metabolic landscape of this pathogen and mucin-induced metabolic modulations during infection. Overall, this work paves the foundation for further integrative studies to understand the mechanisms and diversity of metabolic modulations in P. aeruginosa associated with the host environment.














Acknowledgements (Optional): :

References (Optional): :