Myeloperoxidase is required for neutrophil extracellular trap formation: implications for innate immunity

Abstract

The granule enzyme myeloperoxidase (MPO) plays an important role in neutrophil antimicrobial responses. However, the severity of immunodeficiency in patients carrying mutations in MPO is variable. Serious microbial infections, especially with Candida species, have been observed in a subset of completely MPO-deficient patients. Here we show that neutrophils from donors who are completely deficient in MPO fail to form neutrophil extracellular traps (NETs), indicating that MPO is required for NET formation. In contrast, neutrophils from partially MPO-deficient donors make NETs, and pharmacological inhibition of MPO only delays and reduces NET formation. Extracellular products of MPO do not rescue NET formation, suggesting that MPO acts cell-autonomously. Finally, NET-dependent inhibition of Candida albicans growth is compromised in MPO-deficient neutrophils. The inability to form NETs may contribute in part to the host defense defects observed in completely MPO-deficient individuals.

Figure 1

MPO is required for NET formation. Neutrophils were activated with PMA for 4-6 hours to stimulate NET formation, then stained with the DNA dye Sytox green (A-C) or (D) immunostained with antibodies against the histone H2A/H2B/DNA complex (red) and neutrophil elastase (green). (A) Neutrophils from a control donor formed NETs after PMA stimulation, and naive cells had small, condensed nuclei. (B) Neutrophils from completely MPO-deficient donors did not form NETs, as indicated by their condensed nuclei similar to those of unstimulated neutrophils. (C) Partially MPO-deficient neutrophils made NETs. (D) Extracellular DNA released from control neutrophils stained with antibodies against chromatin and neutrophil elastase, 2 known NET components. In completely MPO-deficient neutrophils, as in naive neutrophils, neutrophil elastase was localized in granules and chromatin remained condensed. Experiments with CD neutrophils were repeated at least 6 times with 3 independent, unrelated donors. Experiments with PD neutrophils were repeated at least 3 times with 3 independent, unrelated donors. Scale bar, 25 μm.

Figure 2

Quantitation of NET formation by MPO-deficient neutrophils. Nuclear areas of neutrophils stimulated with PMA as in Figure 1 are plotted against the percentage of Sytox-positive cells corresponding to a given nuclear area range. Cells were unfixed (A-B,E) or fixed with paraformaldehyde (C-D) before Sytox staining. (Ai) By 4 hours after PMA stimulation, neutrophils from a control donor made NETs, indicated by the broad range of nuclear areas. Naive cells were still alive and did not stain with Sytox. (Aii) In contrast, neutrophils from a completely MPO-deficient donor did not form NETs. Instead, the cells died by 6 to 8 hours after PMA stimulation, but their nuclei remained small. (B-C) Neutrophils from 2 other completely MPO-deficient donors did not make NETs after 4-6 hours of PMA stimulation. (D) Neutrophils from 3 partially MPO-deficient donors made NETs. (E) Neutrophils from a control donor pretreated with the MPO inhibitor ABAH displayed a 2-hour delay and significant decrease in NET formation, reaching peak nuclear areas at 6 hours instead of 4 hours after PMA stimulation. Color scheme: gray, naive cells; black, control or untreated cells; red, completely MPO-deficient cells; blue, partially MPO-deficient cells; green, ABAH-treated cells. Statistical analysis: ***P < .0001; n.s., differences not significant. (A) Control versus donor 1 (6 hours)***; control versus donor 1 (8 hours)***; donor 1 (6 hours) versus donor 1 (8 hours), n.s. (B) Control versus donor 2***. (C) Control versus naive***; control versus donor 3***; donor 3 versus naïve, n.s. (D) Control versus donor 4***; control versus donor 5, n.s.; control versus donor 6***. (E) 4-hour PMA versus 4-hour PMA plus ABAH***; 4-hour PMA versus 6-hour PMA plus ABAH***; 4-hour PMA plus ABAH versus 6-hour PMA plus ABAH, n.s. For each sample, 100-500 cells were evaluated.

Figure 3

NET formation in response to C albicans is impaired in MPO-deficient neutrophils. Neutrophils were co-incubated with the yeast form of C albicans at an MOI of 10, or activated with PMA, for 2 hours, and DNA was stained with Sytox. (A-D) Images of Sytox fluorescence. (E-F) Quantitation of nuclear areas. Neutrophils from a healthy donor formed NETs upon stimulation with C albicans (A,E) and PMA (C,F), but neutrophils from a completely MPO-deficient donor (donor 1) did not form NETs in response to C albicans (B,E) or PMA (D,F). Scale bar, 25 μm.

Figure 4

NET formation by MPO-deficient cells is not rescued by extracellular products of MPO. (A) Neutrophils from a healthy and a completely MPO-deficient donor (donor 1) were incubated either separately or co-incubated in a transwell system. Upon PMA activation, control neutrophils formed NETs. MPO-deficient neutrophils did not form NETs, even in the presence of control cells on a transwell insert. (B) Neutrophils from a healthy and a completely MPO-deficient donor (donor 1) were either incubated separately or mixed at a 1:1 ratio and stimulated with PMA. Before mixing, cells were differentially stained using MitoTracker Green and Deep Red dyes to distinguish the 2 populations. NET formation by MPO-deficient neutrophils was not significantly increased in the presence of control neutrophils. (C) A representative image of the mixed cell culture from panel B. Red, MitoTracker Deep Red–stained neutrophils from a healthy donor; green, MitoTracker Green–stained neutrophils from donor 1 (CD); blue, DNA (Hoechst). Left panel, merge of all 3 colors; middle panel, merge of red and blue channels; right panel, merge of green and blue channels. Scale bar, 25 μm. (D-E) Histamine monochloramine and histamine dichloramine were generated using an in vitro enzymatic system, then applied to neutrophils from a completely MPO-deficient donor (donor 1) at the indicated concentrations. Neither chloramine tested rescued NET formation by the MPO-deficient cells. Statistical analysis: ***P < .0001; *P = .01-05; n.s., differences not significant. (A) Control versus donor 1***; control versus donor 1 plus control transwell***; donor 1 versus donor 1 plus control transwell, n.s. (D) Control, PMA versus donor 1, PMA***; control, PMA versus donor 1, 100μM ***; control, PMA versus donor 1, 1mM*; donor 1, PMA versus donor 1, 100μM, n.s.; donor 1, PMA versus donor 1, 1mM, n.s.; donor 1, 100μM versus donor 1, 1mM, n.s. (E) Control, PMA versus donor 1, PMA***; control, PMA versus donor 1, 100μM***; control, PMA versus donor 1, 1mM***; donor 1, PMA versus donor 1, 100μM, n.s.; donor 1, PMA versus donor 1, 1mM, n.s.; donor 1, 100μM versus donor 1, 1mM, n.s. For each sample, 100-500 cells were evaluated.

Figure 5

NET-dependent inhibition of C albicans growth is compromised in MPO-deficient neutrophils. Neutrophils from a healthy donor and a completely MPO-deficient donor (donor 1) were stimulated to make NETs using glucose oxidase for 2 hours or were left unstimulated. After NET formation, cells were left untreated or cytochalasin D was added to prevent phagocytic killing by neutrophils that had not made NETs. The yeast form of C albicans was added to the cells, and fungal growth was measured after 10 hours using XTT, a chromogenic metabolic substrate. Measurements were normalized to growth of C albicans in the absence of neutrophils. Control neutrophils efficiently made NETs and restricted fungal growth both in the presence and in the absence of cytochalasin D. MPO-deficient neutrophils did not restrict fungal growth in the presence of cytochalasin D.