Abstract
Purpose :
Pseudomonas aeruginosa is the leading cause of contact lens-associated keratitis. Studying bacterial gene expression relating to keratitis pathogenesis is complicated by the potential for bacterial populations to exhibit spatial/temporal transcriptional heterogeneity and difficulties in modeling in vivo conditions. Here, we utilized new tools for quantifying gene expression in individual bacteria in a murine contact lens model, and demonstrate bacterial transcriptional heterogeneity in vivo, but not in vitro.
Methods :
Fluorescent P. aeruginosa transcriptional reporters were constructed to probe bacterial physiological traits on contact lenses and the murine cornea. These included: interbacterial communication via quorum-sensing molecules (rsaL, lasB, and rhlA promoters), biofilm development (cdrA), and nutrient/oxygen availability (pvdS, crcZ, cbb3). Reporters expressed constitutive ebfp2 (blue), and gfp (green) driven by promoters of interest, and were tested for gfp induction in vitro. Induction of a quorum-sensing reporter (RsaL-GFP) was tested on murine contact lenses incubated in PBS for 24 h in vitro, or worn in vivo for 24 h on one cornea of C57BL/6 mice. Bacteria, contact lenses, and intact eyes ex vivo were imaged by confocal microscopy and analyzed by Imaris.
Results :
In vitro, all reporters exhibited >2-fold increase in GFP-fluorescence in response to control perturbations, showing functionality. Bacteria remaining on lenses worn for 24 h were located in or near large aggregates, or biofilms. Quorum-sensing activation depended on biofilm proximity; bacteria within 5 µm exhibited a >2-fold increase in reporter activity than those 5 µm or further away (p < 0.01, 2-tailed t-test). While 65% of bacteria in vivo showed quorum activation, there were significant spatial differences; 84% within 5 µm activated quorum-sensing versus 24% at 5 µm or further (n = 908 and 409 bacteria respectively). In contrast, lenses incubated in PBS for 24 h in vitro did not harbor biofilms; bacteria were evenly distributed, but ~87% showed quorum activation (n = 1204).
Conclusions :
Bacterial gene expression and community structure on contact lenses are vastly different in vivo versus in vitro. These new reporters could be combined with bulk high-throughput data such as RNA-Seq for more complete analyses of gene expression in the eye, or adapted as probes to monitor factors such as tear exchange or oxygen under contact lenses.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.