PtdIns3P binding to the PX domain of p40phox is a physiological signal in NADPH oxidase activation

Abstract

The production of reactive oxygen species by the NADPH oxidase complex of phagocytes plays a critical role in our defence against bacterial and fungal infections. The PX domains of two oxidase components, p47(phox) and p40(phox), are known to bind phosphoinositide products of PI3Ks but the physiological roles of these interactions are unclear. We have created mice which carry an R58A mutation in the PX domain of their p40(phox) gene, which selectively prevents binding to PtdIns3P. p40(phoxR58A/R58A) embryos do not develop normally but p40(phoxR58A/-) mice are viable and neutrophils from these animals exhibit significantly reduced oxidase responses compared to those from their p40(phox+/-) siblings (e.g. 60% reduced in response to phagocytosis of Staphylococcus aureus). Wortmannin inhibition of the S. aureus oxidase response correlates with inhibition of phagosomal PtdIns3P accumulation and overlaps with the reduction in this response caused by the R58A mutation, suggesting PI3K regulation of this response is substantially dependent on PtdIns3P-binding to p40(phox). p40(phoxR58A/-) mice are significantly compromised in their ability to kill S. aureus in vivo, defining the physiological importance of this interaction.

Figure 1

p40^phoxR58A targeting strategy. (A) Schematic of p40^phoxR58A targeting strategy. The intron–exon structure of the p40^phox gene (ncf4), the extent and overall organisation of the targeting construct, and relevant restriction sites and fragment sizes are shown. Positions of Southern blot probes are shown as diamonds in dotted boxes. Drawing not to scale. (B) Mutations introduced by the targeting strategy are highlighted. (C) Examples of Southern blots of ES cell DNA digested with SpeI or EcoRI and probed with the Neo, 5′ or 3′ probes. T, targeted; U, untargeted. (D) Example of PCR screening of mouse tail genomic DNA of progeny from p40^phoxR58A/+ matings. Primers flanking exon 3 generate an 850 bp fragment that is XhoI-cleaved to 525 and 325 bp in targeted alleles.

Figure 2

Blood cell counts and phox protein expression in p40^phoxR58A/− mice. (A) Blood cell counts from p40^phox+/− and p40^phoxR58A/− animals (WBC, white blood cells; RBC, red blood cells; Plt, platelets, Lymph, lymphocytes; Mono, monocytes; Gran, granulocytes). Data represented are mean % of p40^phox+/− counts±s.e. (n=12). (B–D) BMNs from p40^phox+/+, ^+/−, ^R58A/− and ^−/− animals were sonicated into SDS sample buffer, subjected to SDS–PAGE and immunoblotted for p40^phox, p47^phox and p67^phox. Graphs represent mean % of p40^phox+/+ samples ±s.e. (n=10–13) from four independent experiments, using 3–6 mice per preparation.

Figure 3

p40^phoxR58A/− neutrophils have defects in ROS production in response to TNFα/fMLP and IgG-latex beads. Duplicate wells of BMNs were stimulated in a luminometer and ROS production was followed over time. Response kinetics (line graph; mean±range of one representative experiment), and total integrated responses as % of p40^phox+/− (bar graph; mean±s.e.; n=22 from 11 independent experiments, and n=11 from five independent experiments for TNFα/fMLP and IgG-latex beads, respectively). (A) fMLP-stimulated production of extracellular ROS in mTNFα-primed BMNs from p40^phoxR58A/− and ^+/− animals. (B) IgG-latex bead-stimulated production of intracellular ROS in primed BMNs from p40^phox+/− and p40^phoxR58A/− animals.

Figure 4

p40^phoxR58A/− neutrophils exhibit reduced intracellular ROS production in response to phagocytosis of S. aureus, which equates with a wortmannin-sensitive component of the response. Primed BMNs from p40^phoxR58A/− and ^+/− animals were analysed for S. aureus-induced intracellular ROS production. Where appropriate, cells were pretreated with indicated wortmannin concentrations for 10 min prior to addition of S. aureus. (A) Time course of ROS production in the presence of increasing concentrations of wortmannin. Data are means±s.d. from a single experiment, performed in triplicate, representative of three independent experiments. (B, C) Total integrated ROS responses for the first 16.5 min of each condition (as indicated by dashed lines in A) were calculated and expressed as a percentage of p40^phox+/− response (B), or as a percentage of relevant control response for p40^phox+/− and p40^phoxR58A/− (C). Data are mean±s.e. (n⩾8). (D) Phagocytosis of S. aureus by p40^phox+/− and p40^phoxR58A/− BMN and the effects of wortmannin. Data shown are mean±s.e. from at least 100 cells at each condition, expressed as a percentage of p40^phox+/− phagocytosis.

Figure 5

p40^phoxR58A/− neutrophils have reductions in intracellular but not extracellular ROS production in response to PMA. Primed BMNs from p40^phoxR58A/− and ^+/− animals were analysed for PMA-induced ROS production, using luminol-dependent chemiluminescence either in the presence of HRP (extracellular, A) or absence of HRP (intracellular, B–D) and either in the presence or absence of wortmannin (100 nM). Kinetics of ROS production are shown as means±s.d. from a single experiment performed in triplicate, representative of three independent experiments. Total integrated ROS produced during the initial phase of PMA stimulated ROS production (indicated by dashed line in (C, D)) are shown in the bar graphs (A, B) as a percentage of the relevant p40^phox+/− response; means±s.e. (n=9).

Figure 6

PtdIns3P accumulation and NADPH oxidase activation in response to S. aureus are inhibited in parallel by wortmannin. (A) 5 × 10⁴ differentiated GFP-iPX PLB-985 cells were pretreated with DMSO (vehicle control) or 100 nM wortmannin prior to incubation with RITC-labelled, serum-opsonised S. aureus for 5 min. Samples were cytospun onto glass slides, fixed and mounted, and GFP-positive phagosomes and phagocytosed bacteria visualised by fluorescence microscopy; the scale bar represents 1 μm. (B) Intracellular ROS production in response to S. aureus, enumeration of GFP-positive phagosomes encapsulating phagocytosed S. aureus, and number of phagocytosed S. aureus/cell, in the presence of the indicated wortmannin concentrations, are expressed as a percentage of the relevant control (DMSO) responses. The wortmannin insensitive component of ROS production (approximately 22%) was subtracted from the ROS responses shown; data are means±s.e., n=6. Data for phagocytic indices are means±s.e.; at least 60 cells were analysed for each condition.

Figure 7

p40^phoxR58A/− neutrophils are deficient in killing of S. aureus in vitro and in vivo. (A) Primed BMNs from p40^phoxR58A/− and ^+/− animals were incubated with serum-opsonised S. aureus (approximate ratio of 1 bacterium to 4 neutrophils), in the presence or absence of 3 μM DPI, with rapid mixing. After 15 min, the neutrophils were lysed and surviving bacteria were counted. A control with no neutrophils (input bacteria) is included. Data are means±s.e. (n=3) and are representative of 2–5 experiments, using 3–6 mice per preparation. (B) Mice were injected intraperitoneally with approximately 5 × 10⁷ live S. aureus. After 4 or 24 h, mice were killed, their peritoneums were flushed and surviving bacteria enumerated. Total numbers of surviving bacteria in four independent experiments are shown for the 4 h time point only; data are means±s.e. (n=3 mice per experiment). Two two-way analyses of variance were carried out on the data (one for the data at 4 h and one for the data at 24 h) to test whether the observed differences in bacterial survival were significant, taking into account the variability between the four experiments. The increased survival of S. aureus in the p40^phoxR58A/− mice was significant at 4 h (F=7.664, df=3, P<0.0001) and 24 h (F=717.700, df=3, P<0.0001).