Putting on the brakes: Bacterial impediment of wound healing

Abstract

The epithelium provides a crucial barrier to infection, and its integrity requires efficient wound healing. Bacterial cells and secretomes from a subset of tested species of bacteria inhibited human and porcine corneal epithelial cell migration in vitro and ex vivo. Secretomes from 95% of Serratia marcescens, 71% of Pseudomonas aeruginosa, 29% of Staphylococcus aureus strains, and other bacterial species inhibited epithelial cell migration. Migration of human foreskin fibroblasts was also inhibited by S. marcescens secretomes indicating that the effect is not cornea specific. Transposon mutagenesis implicated lipopolysaccharide (LPS) core biosynthetic genes as being required to inhibit corneal epithelial cell migration. LPS depletion of S. marcescens secretomes with polymyxin B agarose rendered secretomes unable to inhibit epithelial cell migration. Purified LPS from S. marcescens, but not from Escherichia coli or S. marcescens strains with mutations in the waaG and waaC genes, inhibited epithelial cell migration in vitro and wound healing ex vivo. Together these data suggest that S. marcescens LPS is sufficient for inhibition of epithelial wound healing. This study presents a novel host-pathogen interaction with implications for infections where bacteria impact wound healing and provides evidence that secreted LPS is a key factor in the inhibitory mechanism.

Figure 1. Inhibition of cell migration in vitro by some bacterial secretomes.

(a) Images of Calcein AM stained HCLE cells treated with secretomes. *OD₆₀₀ = 1.0 secretomes used in the shown experiment. HT = secretome incubated at 95 °C for 10 minutes. (b) Images of Calcein AM stained human foreskin fibroblast cells treated with LB (mock) and Serratia marcescens PIC3611 secretomes. Initial wound = cells incubated in the presence of a silicone stopper to determine size of original wound. Image taken is half of wound area. Scale bar = 50 μm.

Figure 2. S. marcescens (S.m.) secretomes inhibit corneal wound healing ex vivo.

LB (mock) (n = 6) and secretomes (n = 6) were added onto wounded corneas and incubated for 48 hours. To observe epithelial defects the corneal tissue was stained with Richardson solution (blue stain). Initial wounds (n = 3) are corneas wounded and stained at end of experiment to determine original wound size. HT = secretome incubated at 95 °C for 10 minutes (n = 6). (a) Representative images of porcine corneas treated with secretomes. (b) Measurements of corneal wounds from ex vivo corneal organ culture. Error bars represent one standard deviation. *p < 0.05 by Tukey’s post hoc analysis.

Figure 3. S. marcescens secretomes are not cytotoxic to corneal epithelial cells.

LB (mock) and secretomes were added to HCLEs and incubated overnight. Ethanol treatment was used as a positive control for inviable cell staining. Cells were stained with 0.5 μM Calcein AM and 1 μM propidium iodide (PI). Unstained corneal epithelial cells were imaged to verify there was no background fluorescence. Scale bar = 10 μm.

Figure 4. S. marcescens secretomes alter HCLE actin cytoskeleton.

(a) LB (mock) and secretomes were added to HCLEs and incubated for 4 hours. HCLEs were stained with Alexa-488 phalloidin for actin (green) and Hoechst 33342 for DNA (blue) and imaged. Scale bar = 10 μm. (b) Actin projections per 30 μm cell area were quantified (LB n = 36, S.m. n = 40). Error bars represent standard deviation. *p < 0.05 by Student’s T test. (c) Stratified HCLEs were treated with LB (mock) and secretomes for 3 hours. Cells were fixed and stained as described above. The center of the “wound” was imaged by confocal microscopy. Scale bar = 50 μm.

Figure 5. S. marcescens secretomes from a mutant in the LPS biosynthetic locus transposon mutant (waaG) and LPS depleted secretomes do not inhibit cornea cell migration.

LB (mock) and secretomes were added to HCLEs and incubated for 18–24 hours. (a) HCLE cell migration assays treated with secretomes from S.m. (pMQ131 vector control), LPS transposon mutant (waaG pMQ131 vector control), pMQ491 (waaG alone), and pMQ505 (waaG and orf10). (b) HCLEs treated with mock and secretomes from waaG transposon mutant (waaG), LPS depleted (pmxB), and agarose bead control treated secretomes (S.m. sepharose). Scale bar = 50 μm.

Figure 6. Purified LPS from S. marcescens (S.m.) PIC3611 inhibits corneal wound healing ex vivo.

waaG secretomes and 10000 ng/ml waaG purified LPS (n = 3) do not inhibit corneal wound healing and complementation with pMQ505 (waaG and orf10) restores wound healing inhibition. LB (mock) (n = 4), secretomes (n = 6), and purified LPS were added dropwise onto wounded corneas and incubated for 48 hours. Initial wounds (n = 2) are corneas wounded and stained at end of experiment to determine original wound size. To determine remaining wound size corneas were stained with Richardson solution (blue stain).

Figure 7. S. marcescens (S.m.) PIC3611 LPS, but not LPS derived from E. coli or S.m. waaC and waaG mutants, inhibits corneal cell migration in vitro.

LPS was purified from E. coli K746, S. m. WT, waaC and waaG LPS mutants. Scale bar = 50 μm. (a) E. coli K746 and S. m. LPS cell migration experiments. Loss of staining indicates cell death or removal of corneal cells from surface. (b) waaC and waaG LPS cell migration experiments.