Computational prediction of a novel transcriptional regulation for the PqsH encoding gene in Pseudomonas aeurginosa
Author(s):
Scott Christley, Olga Zaborina, John Alverdy, Gary An, University of Chicago
Background: Pseudomonas aeruginosa is one of the most clinically significant microbes in healthcare associated infections, and is well known to transform into virulent phenotypes through host-pathogen interactions in conditions of host-stress. Quorum sensing is known to be involved in virulence transformation, however the ubiquitous challenge remains: the task of integrating vast amounts of data to generate new knowledge that can be tested in an efficient manner. Herein we present a novel approach using advanced computational modeling and bioinformatics to identify a novel putative regulatory interaction in P. aeruginosa quorum sensing.
Hypothesis: Our computational analysis suggests the hypothesis that under low oxygen conditions, Anr is produced and suppresses transcription of PqsH.
Methods: We developed a computational model from a literature survey that integrates the three primary P. aeruginosa quorum-sensing systems, LasRI, RhlRI, and MvfR-PQS. We performed large-scale simulations of the model (>25 million in 2 days using a desktop GPU machine) using novel computational algorithms developed by our group to extract useful predictions without requiring (typically unknown) kinetic rates. Bioinformatic search of DNA binding sites was performed using the motif search tool at http://www.pseudomonas.com.
Results: The computational simulations results suggest that the only currently known transcriptional regulation of the pqsH gene by the LasRI quorum system is insufficient to explain observed data under high phosphate (25mM) conditions. Specifically, an inhibitory regulation of PqsH is required. PqsH encodes an enzyme for producing PQS from HHQ, and this metabolic reaction is limited by oxygen availability, which suggests that PqsH may be regulated by the oxygen-sensing system via its transcription factor Anr. A bioinformatic search of the Anr binding motif, TTGN{3,8}TCAA, reveals a potential binding site within the transcriptional unit of the pqsH gene.
Conclusions: This research has shown that the compilation of biological knowledge from existing literature into a computational model, and then the use of dynamic simulations to analyze that model, can lead to novel and specific hypotheses. While the hypothesis that Anr suppresses transcription of PqsH is yet to be validated, the significant result is that computational analysis was able to pinpoint a gap in our current knowledge.