Small-molecule screen identifies reactive oxygen species as key regulators of neutrophil chemotaxis

Abstract

Neutrophil chemotaxis plays an essential role in innate immunity, but the underlying cellular mechanism is still not fully characterized. Here, using a small-molecule functional screening, we identified NADPH oxidase-dependent reactive oxygen species as key regulators of neutrophil chemotactic migration. Neutrophils with pharmacologically inhibited oxidase, or isolated from chronic granulomatous disease (CGD) patients and mice, formed more frequent multiple pseudopodia and lost their directionality as they migrated up a chemoattractant concentration gradient. Knocking down NADPH oxidase in differentiated neutrophil-like HL60 cells also led to defective chemotaxis. Consistent with the in vitro results, adoptively transferred CGD murine neutrophils showed impaired in vivo recruitment to sites of inflammation. Together, these results present a physiological role for reactive oxygen species in regulating neutrophil functions and shed light on the pathogenesis of CGD.

Fig. 1.

Pharmacological inhibition of NADPH oxidase–mediated ROS production in human neutrophils leads to more frequent formation of multiple pseudopodia and reduced chemotaxis efficiency. (A) Inhibition of NADPH oxidase decreases transwell migration of human neutrophils. Human neutrophils were left untreated or pretreated with 50 μM DPI for 30 min at 37 °C, and then allowed to migrate in response to the indicated concentration of fMLP. Percentage of cells that migrated into the bottom well was recorded. Data shown are means ± SD from n = 3 wells, from one experiment representative of three. (*P < 0.05 vs. untreated neutrophils.) (B) Chemotaxing NADPH oxidase–inhibited human neutrophils display multiple pseudopodia more frequently, but do not show difference in cell speed. Representative images of human neutrophils treated (Right) or not treated (Left) with 50 μM DPI and allowed to chemotax in response to a chemoattractant gradient generated by addition of 1 μL of 100 nM fMLP in the EZ-TAXIScan device. White arrowheads specify pseudopodia in the cells. Percentage of cells that display multiple pseudopodia (bottom left, n = 20 cells; Fisher 2 × 2 test, *P < 0.05 vs. untreated) and migration speed (bottom right, mean ± SD from n = 20 cells; Student t test, *P > 0.05 vs. untreated) during the course of the EZ-TAXIScan chemotaxis assay in the DPI-treated or untreated case were quantified as described in Experimental Procedures. (C) Inhibition of NADPH oxidase leads to multiple pseudopod formation and defective directionality in fMLP-, LTB4-, and IL-8–mediated chemotaxis. Neutrophil chemotaxis in response to addition of 100 nM fMLP, 100 nM LTB4, or 10 nM IL-8 (1 μL each), with 50 μM DPI treatment (or without treatment), was analyzed (n = 20 cells) for directionality, upward directionality, and percentage of neutrophils displaying multiple pseudopodia as described earlier (n = 20 cells, *P < 0.05). (D) Pharmacological agents that inhibit ROS production cause chemotaxis defect in human neutrophils. fMLP-induced ROS production (Fig. S5 Top) in human neutrophils (5 × 10⁵) pretreated with sinomenine (10 μM), DPI (50 μM), EHNA hydrochloride (40 μM), or LE135 (14 μM), or without pretreatment, was evaluated by stimulating neutrophils with 100 nM fMLP and monitoring chemiluminescence (for 1 s) every 7 s for 280 s, in the presence of 50 μM luminol and 0.8 U HRP in a luminometer at 37 °C. Data represent maximal chemiluminescence in drug-treated neutrophils normalized to maximal chemiluminescence in untreated neutrophils (mean ± SD, n = 3 wells from one experiment representative of three). Drug-treated (and untreated) neutrophils were also exposed to an fMLP gradient generated by addition of 100 nM fMLP (1 μL) in the EZ-TAXIScan device. Images and cell tracks of migrating neutrophils were evaluated for directionality, upward directionality and percentage of neutrophils displaying multiple pseudopodia as described earlier (n = 20 cells; *P < 0.05). (E) H₂0₂ treatment of NADPH oxidase–inhibited human neutrophils rescues defect in pseudopod formation and chemotaxis. Human neutrophils were pretreated with 50 μM DPI (or without) and then treated with (or without) 100 μM H₂0₂ for 5 min and then allowed to chemotax in response to a fMLP gradient in the EZ-TAXIScan device as described earlier. Migrating neutrophils were evaluated (n = 20 cells; *P < 0.05 vs. DPI-treated neutrophils) for directionality, upward directionality, and percentage of neutrophils displaying multiple pseudopodia as described earlier.

Fig. 2.

Disruption of NADPH oxidase leads to chemotaxis defects. (A–E) Neutrophils from CGD mice do not produce ROS in response to chemoattractant stimulation, display multiple pseudopodia, and show loss of directionality during chemotaxis. (A) ROS production in neutrophils (5 × 10⁵) from WT or CGD mice after stimulation with 1 μM fMLP was evaluated by monitoring chemiluminescence (for 1 s) every 7 s for 280 s, in the presence of 50 μM luminol and 0.8 U HRP in a luminometer at 37 °C. Data represents mean ± SD from three wells from one experiment representative of three. (B) Neutrophils from WT and CGD mice (3,000 cells) were plated into the EZ-TAXIScan device and exposed to a shallow chemoattractant gradient generated by addition of 1 μL LTB4 (100 nM). Cell tracks of migrating WT (Left) and CGD (Right) neutrophils (cells that move ≥65 μm from the bottom of the channel; n = 20) were traced from captured images, realigned such that all cells started from the same starting point (0,0) and plotted. Chemoattractant concentration increases in the positive y direction. (C) Representative images of WT (Left) and CGD migrating neutrophils (Right) are also shown; white arrowheads specify pseudopodia. (D) Migrating neutrophils were evaluated (n = 20 cells; *P < 0.05 vs. WT neutrophils) for directionality, upward directionality, percentage of neutrophils displaying multiple pseudopodia, and migration speed as described earlier. (E) Transwell migration of CGD mice neutrophils. WT or CGD murine neutrophils were allowed to migrate in response to the indicated concentration of LTB4. Percentage of cells that migrated into the bottom well was recorded. Data shown are means ± SD from three wells from one experiment representative of three. (*P < 0.05 vs. WT neutrophils.) (F–H) Knocking down p22phox via siRNA results in impaired cell migration and chemotaxis. HL60 cells were differentiated with 1.75% DMSO for 1 d, transfected with 1 μM control siRNA or p22phox siRNA, and further differentiated until d 5 or 6. (F) Knockdown of p22phox in dHL60 cells. At d 5 (Left) or d 6 (Right), dHL60 cells were lysed and probed with p22phox antibody to evaluate knockdown and GAPDH antibody to evaluate loading. (G) Decreased ROS production in p22phox-knockdown dHL60 cells. ROS production in control siRNA or p22phox siRNA transfected dHL60 cells (2 × 10⁵, 5 d of differentiation) after stimulation with 10 nM C5a was evaluated by monitoring chemiluminescence (for 1 s) every 30 s for 300 s, in the presence of 50 μM luminol and 0.8 U HRP in a luminometer at 37 °C. (H) Control siRNA or p22phox siRNA transfected dHL60 cells (d 5) were exposed to a chemoattractant gradient generated by addition of 25 nM or 100 nM C5a (1 μL) in the EZ-TAXIScan device and imaged every 0.5 min for 20 min. Cell tracks of migrating dHL60 cells (n = 15) were traced from the captured images, realigned to start from the same point (0,0), and plotted (Left). Migration paths of the dHL60 cells were evaluated for a 0- to 20-min time frame (n > 15 cells, *P < 0.005 vs. control siRNA dHL60 cells) for directionality, upward directionality, and migration speed as described in Experimental Procedures (Right).

Fig. 3.

Neutrophils isolated from the CGD patient are defective in ROS production and chemotactic migration. (A) Decreased ROS production in neutrophils from a CGD patient. ROS production in neutrophils (5 × 10⁵) from a CGD patient or a healthy volunteer after addition of HBSS (Top) or 100 nM fMLP (Bottom) was evaluated by monitoring chemiluminescence (for 1 s) every 20 s for 360 s, in the presence of 50 μM isoluminol and 0.8 U HRP in a luminometer at 37 °C. Data represent mean ± SD from three wells. (B) Neutrophils from the CGD patient or healthy volunteer were exposed to a chemoattractant gradient generated by addition of 100 nM fMLP (1 μL) in the EZ-TAXIScan device and imaged every 0.5 min for 20 min. Cell tracks of migrating neutrophils (n = 20) were traced from the captured images, realigned to start from the same point (0,0), and plotted (Top). Images of chemotaxing neutrophils from the CGD patient or healthy volunteer are shown (Middle). White arrowheads specify pseudopodia in the cells. Neutrophils were evaluated (n = 20 cells, *P < 0.01 vs. WT neutrophils) for directionality, upward directionality, percentage of neutrophils displaying multiple pseudopodia, and migration speed as described earlier (Bottom).

Fig. 4.

CGD neutrophils exhibit an intrinsic defect in recruitment to sites of inflammation. (A) Recruitment of adoptively transferred neutrophils in the TG-induced peritonitis model. Neutrophils isolated from WT and CGD mice were labeled with intracellular fluorescent dye CFSE (final concentration, 1 μM; Molecular Probes) or 5- (and 6-) chloromethyl SNARF-1 acetate (final concentration, 1 μM; Molecular Probes) at 37 °C for 10 min. Labeled cells were mixed (1:1) as indicated and then injected i.v. (via tail vein) into WT mice that have been challenged with 1 mL 3% TG for 2.5 h. Peritoneal fluids were harvested 1.5 h after the injection of cell mixture. The amount of adoptively transferred neutrophils recruited to the peritoneal cavity was analyzed using a BD FACSCanto II flow cytometer (Becton Dickinson) and BD FACSDiva software. Relative recruitment of neutrophil was calculated as the ratio of indicated populations in the peritoneal cavity. (B) Recruitment of adoptively transferred neutrophils to a preformed air pouch. The dorsal air pouch was generated on WT recipient mice as described in Experimental Procedures. Neutrophil recruitment to the pouch was induced by TNF-α, which was directly injected into the pouch 2.5 h before the neutrophil injection. The pouch was flushed 1.5 h after the injection of cell mixture. The relative recruitment of WT and CGD neutrophil was calculated as described earlier.