Abstract
Purpose :
Pseudomonas aeruginosa (PA) is a vision-threatening pathogen that causes corneal keratitis. Interactions between PA and the injured cornea are well studied but the host response that protects a healthy cornea from PA infection is not fully understood. Prior work shows that MyD88 and IL-1R limit PA adherence and traversal of the intact corneal epithelium; both factors are involved in NF-κB activation. Specific stimuli that activate receptors upstream of NF-κB are unknown. We aim to show how NF-κB is activated by PA to understand how downstream responses protect the ocular surface.
Methods :
Corneal epithelial cells (hTCEpi) were infected with wt PA (PAO1), a toxin null mutant (PAO1ΔexoSTY), or a type III secretion system (T3SS) mutant (PAO1ΔexsA). Cell lysates were collected at 4 hours post infection and RNAseq was performed. Cells were also exposed to live PA, killed PA by heat, paraformaldehyde (PFA), or amikacin, and TLR 1/2 (Pam3CSK4) or 2/6 (FSL-1) agonists. Cells were fixed and stained for p65. Cells positive for nuclear p65 were scored using an ImageJ macro measuring the ratio of mean intensity in the nucleus per mean intensity in an expanded perinuclear region.
Results :
Cells infected with PAO1, PAO1ΔexoSTY, or PAO1ΔexsA exhibited similar gene expression patterns including upregulation of NF-κB genes. In all cases, transcripts were detected for TLR 2, 3, & 6, NOD-1 & 2, RIG1, MDA5, NLRP1 & 3 but not TLR 1, 4, 5, 7, 8, 9, or 10. NF-κB activation occurred regardless of whether bacteria had a T3SS (39% - 72% positive). Heat and PFA-killed PA and Pam3CSK4 failed to activate NF-κB at several concentrations, whereas amikacin-killed PA and FSL-1 activated NF-κB to similar levels as live PA (47% - 66% positive).
Conclusions :
Corneal epithelial cells respond to PA infection independent of the T3SS and its effectors. This response includes activation of the NF-κB pathway. TLR4 and 5 were not transcribed and Pam3CSK4 failed to activate NF-κB, thus activation may be TLR 4, 5, and 1-independent. However, NF-κB activation may occur via TLR 2/6. Heat and PFA-killed PA failed to activate NF-κB, but amikacin-killed PA did, suggesting that intact bacteria or a molecule sensitive to PFA fixation may be required for NF-κB activation. Future studies will focus on roles of other pattern recognition receptors and the adaptor protein MyD88 in PA infection detection and response.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.