Neutrophil extracellular trap (NET)-mediated killing of Pseudomonas aeruginosa: evidence of acquired resistance within the CF airway, independent of CFTR

Abstract

The inability of neutrophils to eradicate Pseudomonas aeruginosa within the cystic fibrosis (CF) airway eventually results in chronic infection by the bacteria in nearly 80 percent of patients. Phagocytic killing of P. aeruginosa by CF neutrophils is impaired due to decreased cystic fibrosis transmembrane conductance regulator (CFTR) function and virulence factors acquired by the bacteria. Recently, neutrophil extracellular traps (NETs), extracellular structures composed of neutrophil chromatin complexed with granule contents, were identified as an alternative mechanism of pathogen killing. The hypothesis that NET-mediated killing of P. aeruginosa is impaired in the context of the CF airway was tested. P. aeruginosa induced NET formation by neutrophils from healthy donors in a bacterial density dependent fashion. When maintained in suspension through continuous rotation, P. aeruginosa became physically associated with NETs. Under these conditions, NETs were the predominant mechanism of killing, across a wide range of bacterial densities. Peripheral blood neutrophils isolated from CF patients demonstrated no impairment in NET formation or function against P. aeruginosa. However, isogenic clinical isolates of P. aeruginosa obtained from CF patients early and later in the course of infection demonstrated an acquired capacity to withstand NET-mediated killing in 8 of 9 isolates tested. This resistance correlated with development of the mucoid phenotype, but was not a direct result of the excess alginate production that is characteristic of mucoidy. Together, these results demonstrate that neutrophils can kill P. aeruginosa via NETs, and in vitro this response is most effective under non-stationary conditions with a low ratio of bacteria to neutrophils. NET-mediated killing is independent of CFTR function or bacterial opsonization. Failure of this response in the context of the CF airway may occur, in part, due to an acquired resistance against NET-mediated killing by CF strains of P. aeruginosa.

Figure 1. P. aeruginosa stimulates NET formation in nonadherent neutrophils.

Human neutrophils suspended in media were untreated or treated with 25 nM PMA, 10 µM DPI, or P. aeruginosa at MOI of 0.1, 1, 10, or 100. NET formation was measured at 30-minute intervals by DNA released (ng per 10⁶ PMN). Neutrophils were isolated from healthy donors (n = 3) with samples performed in duplicate; error bars represent SEM. ***  =  p<0.001 by two-way ANOVA with Bonferroni's post-test compared to untreated control.

Figure 2. NET formation and binding of P. aeruginosa by nonadherent neutrophils.

Panels A and B: Bacterial cells of strain PAO1 labeled with polymyxin B-BODIPY were incubated with isolated human neutrophils (MOI of 10) for 2 hours with (Panel B) or without (Panel A) DNase. NETs were then stained with the cell-impermeant DNA binding dye, Sytox Orange. In the presence of DNase, NETs are seen to be completely degraded. Panels C and D present the results of a similar experiment performed at a higher ratio of bacteria to neutrophils (MOI of 100), to highlight the physical association of bacteria with NETs.

Figure 3. NET-mediated killing of P. aeruginosa is most effective for nonadherent neutrophils.

Panel A: Non-opsonized or opsonized bacterial cells of strain PA01 were juxtaposed to neutrophils on the surface of a plate by centrifugation (MOI of 0.01). Neutrophils were untreated or exposed to 25 nM PMA for 2 hours to maximize NET formation prior to the addition of bacteria. ***  =  p<0.001 and NS  =  p>0.05 by ANOVA with Bonferroni's post-test. Panel B: This killing assay was performed with continuous rotation of the neutrophils. Under these conditions, virtually all killing is attributable to NETs. Neutrophils were isolated from healthy donors (n = 5) with samples performed in duplicate; error bars represent SEM. All comparisons were non-significant, with p>0.05 by ANOVA with Bonferroni's post-test.

Figure 4. NETs kill P. aeruginosa in suspension across a wide range of MOI.

Panel A: The effect of MOI on NET-mediated killing was tested with neutrophils and PAO1 stationary on a plate, as depicted in Figure 3A. Over a 3-log range of MOI, NET-mediated killing represented a very small fraction of total killing. ***  =  p<0.001 and NS  =  p>0.05 by ANOVA with Bonferroni’s post-test. Panel B: With PAO1 and neutrophils in suspension, NET-mediated killing accounted for nearly all killing over a 5-log range of MOI. The lowest MOI tested trended towards greater efficiency of NET-mediated killing. Neutrophils were isolated from healthy donors (n = 5) with samples performed in duplicate. Differences in total killing and NETs killing at each MOI were not significant.

Figure 5. Neutrophils isolated from CF patients form NETs that effectively kill P. aeruginosa.

Panel A: Neutrophils from healthy volunteers (filled boxes) or CF patients (hatched boxes) were unstimulated (PMN), treated with an inhibitor of NET formation (DPI), treated with a stimulant of maximal NET formation (PMA), or exposed to PAO1 at the indicated MOI. Released NET-associated DNA was quantitated after a 120-minute exposure. There were no significant differences in NET formation between neutrophils isolated from healthy volunteers and CF patients. Panel B: CF neutrophil NET-mediated killing of bacterial cells of strain PAO1 was performed in suspension as in Figure 3B. Effective killing of PAO1 was observed in the absence of additional stimulation, and was significantly enhanced by pre-treatment with PMA to stimulate maximal NET formation. For both panels, neutrophils were isolated from CF patients (n = 5) with samples performed in duplicate; error bars represent SEM. *  =  p<0.05 and **  =  p<0.01 by ANOVA with Bonferroni’s post-test.

Figure 6. P. aeruginosa acquires resistance to NET-mediated killing in the CF airway.

Panel A. Neutrophils were stimulated with PMA to induce maximal NET formation and were then exposed to isogenic clinical strains of P. aeruginosa isolated from CF patients around the time of the first positive culture (“Early”) or a mean of 10.6 years later (“Late”). *  =  p<0.05, **  =  p<0.01, and *** = P<0.001 by Student’s t-test comparing early strain killing versus late strain killing for each pair of isolates. Neutrophils were isolated from healthy donors (n  =  ≥4 for each set) with samples performed in duplicate and error bars represent SEM. Total and NET-mediated killing were determined as in Figure 3. Panel B. Aggregate analysis of the data in Panel A demonstrates a significant resistance to NET-mediated killing acquired by the Late isolates when compared to Early isolates, or laboratory-adapted strain PAO1. *  =  p<0.05, **  =  p<0.01, and *** = P<0.001 by ANOVA with Bonferroni’s post-test. Panel C. Isolated human neutrophils were stimulated with PMA as for Panel A, and exposed to P. aeruginosa PAO1 or two independently derived mucA mutants of PAO1, PW2387 (University of Washington Pseudomonas transposon mutant library) or MV mucA (targeted gentamicin cassette disruption from the laboratory of Michael Vasil) with assessment of total and NET-mediated killing. Neutrophils were isolated from healthy donors (n  =  ≥5) with samples performed in duplicate. Differences between samples in NET-mediated or total killing were not significant by Student’s t-test.