Role of p85α in neutrophil extra- and intracellular reactive oxygen species generation

Drug resistance is a growing problem that necessitates new strategies to combat pathogens. Neutrophil phagocytosis and production of intracellular ROS, in particular, has been shown to cooperate with antibiotics in the killing of microbes. This study tested the hypothesis that p85α, the regulatory subunit of PI3K, regulates production of intracellular ROS. Genetic knockout of p85α in mice caused decreased expression of catalytic subunits p110α, p110β, and p110δ, but did not change expression levels of the NADPH oxidase complex subunits p67phox, p47phox, and p40phox. When p85α, p55α, and p50α (all encoded by Pik3r1) were deleted, there was an increase in intracellular ROS with no change in phagocytosis in response to both Fcγ receptor and complement receptor stimulation. Furthermore, the increased intracellular ROS correlated with significantly improved neutrophil killing of both methicillin-susceptible and methicillin-resistant S. aureus. Our findings suggest inhibition of p85α as novel approach to improving the clearance of resistant pathogens.


IntroductIon
Staphylococcus aureus make up a large proportion of human infections worldwide, causing various diseases that range from acute skin infections to life-threatening systemic toxic shock syndromes. The rise of methicillinresistant S. aureus (MRSA) and other antibiotic-resistant strains has sparked the need for new treatment strategies in both immunodeficient and immunocompetent individuals [1][2][3][4][5]. Neutrophils are part of the innate immune system and are critical for clearing S. aureus infections. They are the first responders to invading bacteria and kill microbes using reactive oxygen species (ROS) produced by the NADPH oxidase complex [1,[6][7][8][9][10]. During ingestion, neutrophils first form a phagosomal cup, which then becomes a fully internalized phagosome where microorganisms are isolated and exposed to toxic levels of superoxide (O 2 -) and other reactive oxygen species (ROS) [11]. The NADPH oxidase complex is located within the phagosome membrane and it is made up of membranebound gp91 phox and p22 phox ; and cytosolic p47 phox , p67 phox , p40 phox , and Rac2 [12,13]. Individuals with chronic granulomatous disease (CGD) demonstrate the importance of NADPH oxidase function for human health, as these patients lack a functional NADPH oxidase complex, and thus suffer from recurrent and severe bacterial and fungal infections [10,14,15].
A variety of receptors on neutrophil membranes contribute to the recognition of opsonized material. Two of the most important receptors are Fcγ receptors (FcγRs), which bind to IgG-coated pathogens, and complement receptors (CRs), which bind C3b on complement-coated pathogens [16][17][18]. FcγR or CR binding causes the cytosolic components of NADPH oxidase to translocate to gp91 phox /p22 phox at the membrane. Downstream of these receptors are many signaling molecules that regulate NADPH oxidase assembly and activation, including Class IA phosphoinositide 3-kinase (PI3K). PI3K is a heterodimer consisting of a regulatory subunit (p85α, p55α, p50α, or p85β) and a catalytic subunit (p110α, p110β, or p110δ) and phosphorylates the lipid PI(4,5)P 2 to produce PI(3,4,5)P 3 [12,13]. Furthermore, pharmacologic inhibition and genetic ablation of the catalytic subunits have been shown to decrease neutrophil ROS production in response to IgG-zymosan and Aspergillus fumigatus hyphae [16,19]. However, the specific role of p85α, the most abundant regulatory subunit of Class IA PI3K, has not been fully studied.
We previously found that a functional binding site on p47 phox for Class IA PI3K-derived phospho-lipids, PI(3,4)P 2 and PI(3,4,5)P 3, is needed for extracellular ROS production, but is dispensable for intracellular ROS production during early phagocytosis [20]. This finding is consistent with the observation that PI(3,4)P 2 and PI(3,4,5)P 3 are found on the early phagosomal cup at the location of and during the time of extracellular ROS production, but are not detected on the mature, internalized phagosome. Notably, p85α, the regulatory subunit of Class IA PI3K and thus necessary for PI(3,4,5)P 3 production, remains associated with the phagosome membrane even when PI(3,4,5)P 3 is no longer present [21]. These observations led us to hypothesize that p85α differentially influences extracellular and intracellular NADPH oxidase activity and performs a function on the internalized, sealed phagosome independent of PI(3,4,5)P 3 production.
To test this hypothesis, we used neutrophils lacking p85α, p55α, and p50α (encoded by Pik3r1), and distinguished production of extra-and intracellular ROS. We found that the PI3K regulatory subunits are not necessary for formation of the early phagosome cup or for production of extracellular ROS. However, we show that the loss of p85α leads to enhanced intracellular ROS, which also contributed to improved killing of methicillinsusceptible S. aureus (MSSA) and methicillin-resistant S. aureus (MRSA).
Our work provides a novel target in the regulation of enhancing neutrophil intracellular ROS, which has been shown to cooperate with anti-microbial agents to increase bacterial killing [22,23]. This is an improvement over indiscriminately increasing global ROS production, which could lead to inflammation-induced tissue injury. Using intracellular ROS to augment anti-microbial therapies may provide a novel strategy in the treatment of antibioticresistant pathogens.
To examine the function of the PI3K regulatory subunits in neutrophil ROS production, we examined both extra-and intracellular ROS production in response to various stimuli. Pik3r1 -/neutrophils had significantly reduced FcγR (hIgG-latex)-stimulated extracellular ROS (50% of WT, Figure 1D), but similar amounts of CR (SOZ)-stimulated extracellular ROS ( Figure 1D). Together with the lower expression of the PI3K catalytic subunits, this finding supports the notion that PI3K activity is required for extracellular ROS mediated by FcγR stimulation, and is consistent with our previous findings [20]. In contrast to that observed with extracellular ROS, loss of p85α/p55α/p50α led to significantly increased FcγR-and a trend of enhanced CR-stimulated intracellular ROS production ( Figure 1E). Both extra-and intracellular ROS production from heterozygous Pik3r1 +/neutrophils was similar to that of WT.
Immunoblots confirmed that activated Akt was reduced in Pik3r1 -/neutrophils; however, myeloperoxidase (MPO) levels were equal in WT and Pik3r1 -/cells ( Figure  1F), demonstrating that the increased intracellular ROS production in Pik3r1 -/neutrophils is due to a regulatory effect of p85α/p55α/p50α, rather than due to reduced MPO expression and diminished ROS consumption. Furthermore, by immunostaining with anti-F-actin and anti-Rac to visualize the phagosomes at 10min and 30min post-SOZ stimulation, we found a similar phagocytic index in WT and Pik3r1 -/neutrophils ( Figure 1G and 1H), indicating that increased FcγR-and CR-stimulated intracellular ROS levels in Pik3r1 -/neutrophils is not merely due to enhanced phagocytosis, but to a regulatory role of p85α/p55α/p50α on NADPH oxidase activity. www.impactjournals.com/oncotarget taken from WT, Pik3r1 +/-, and Pik3r1 -/embryos was assessed based on A. morphology (images taken with 40X objective) and b. Mac-1/ Gr-1 staining; C. Protein expression of PI3K catalytic subunits and NADPH oxidase subunits as measured by immunoblot; D. Extracellular and E. intracellular ROS production was measured in WT, Pik3r1 +/-, and Pik3r1 -/fetal liver-derived neutrophils stimulated with hIgG-latex (FcγR simulation) and SOZ (CR stimulation), n=10, *p<0.0001 comparing extracellular ROS production from Pik3r1 -/to WT in response to hIgG-latex, and n=10, p=0.0009 comparing intracellular ROS production from Pik3r1 -/to WT in response to hIgG-latex, statistical analyses performed by two-tailed, one-sample Student's t-test; F. Immunoblotting for phospho-Akt and MPO in WT and Pik3r1 -/fetal liver-derived neutrophils, stimulated for 0min, 10min, and 30min with SOZ; G. WT and Pik3r1 -/fetal liver-derived neutrophils were stimulated with SOZ and immunostained with anti-F-actin and anti-Rac to visualize phagosomes and quantitate phagocytic index, images taken with 100× objective; H. 10min and 30min after SOZ stimulation, phagocytic index (PI) was calculated as PI = (% of phagocytic cells containing ≥ 1 particle) × (mean number of particles/phagocytic cell containing particles).

Re-introduction of p85α corrects the levels of extra-and intracellular ROS production in Pik3r1 -/neutrophils
We next examined the effect of re-introducing p85α (Figure 2A) on extra-and intracellular ROS production from Pik3r1 -/fetal liver-derived neutrophils [27]. We found comparable neutrophil differentiation (Mac-1, Gr-1) between Pik3r1 -/neutrophils and Pik3r1 -/neutrophils upon re-introduction of p85α ( Figure 2B). Protein levels of PI3K catalytic subunit p110δ were increased upon re-introduction of p85α, and concordantly, Akt phosphorylation was normalized ( Figure 2C). Moreover, p85α restored hIgG-latex-stimulated extracellular ROS levels and inhibited hIgG-latex-stimulated intracellular ROS compared to Pik3r1 -/neutrophils ( Figures 2D, 2E, and 2F). Consistent with a dispensable role of p85 α, p55α, and p50α on SOZ-stimulated extracellular ROS ( Figure  1D), re-introduction of p85α did not alter SOZ-stimulated extracellular ROS production from Pik3r1 -/neutrophils ( Figure 2G); however, SOZ-stimulated intracellular ROS was inhibited by re-introduction of p85α. These findings demonstrate that of the PI3K regulatory subunits, p85α uniquely is able to negatively regulate hIgG-latexand SOZ-stimulated intracellular ROS production. Furthermore, as the C-terminal portion of p85α (nSH2, iSH2, and cSH2 domains, the shared domains between the p85α, p55α, and p50α regulatory proteins) is critical for promoting PI3K activity, these findings suggest that the N-terminus of p85α (SH3 and BH domains) functions to negatively regulate intracellular ROS.

Elevated intracellular ROS in Pik3r1 -/neutrophils enhances bacterial killing
Based on our observation that Pik3r1 -/fetal liverderived neutrophils have increased intracellular ROS production in response to Fcγ receptor and CR stimulation, we predicted that Pik3r1 -/neutrophils would demonstrate enhanced S. aureus-stimulated ROS production and enhanced bacterial killing compared to WT neutrophils. Consistent with our hypothesis, Pik3r1 -/fetal liverderived neutrophils produce more intracellular ROS in response to serum-opsonized MSSA (Wood 46), while the extracellular ROS levels were not affected ( Figures  3A and 3B).
To determine the antimicrobial function of Pik3r1 -/neutrophils, we performed a bacterial killing assay. WT and Pik3r1 -/fetal liver-derived neutrophils were incubated with serum-opsonized methicillin-sensitive S. aureus (MSSA) (Wood 46) over 60 min, and at various time points, neutrophil samples were quenched in ice-cold LB-saponin, sonicated to liberate ingested bacteria, and surviving bacteria were enumerated by plating on LBagar. Surviving MSSA was reduced when incubated with Pik3r1 -/neutrophils compared to WT neutrophils at all time points, with a trend toward statistical significance at 60 minutes ( Figure 3C).
Given the promising results using fetal liver-derived neutrophils and Wood 46 MSSA, we turned our attention to a more relevant pathogen, MRSA. Hypervirulent MRSA, such as LAC MRSA USA300 strain, has become an important public health problem due to the sheer number of infections and widespread antibiotic resistance [3]. Consistent with results observed using MSSA, we observed enhanced intracellular ROS in fetal liver-derived Pik3r1 -/neutrophils compared to WT neutrophils in response to serum-opsonized MRSA (USA300), and no difference in extracellular ROS production ( Figures 3D  and 3E). We next conducted bacterial killing assays using a more physiologic source of neutrophils directly isolated from the bone marrow using a percoll gradient. Since global knockout of Pik3r1 induces perinatal lethality, we utilized a murine model bearing a conditionally targeted Pik3r1 allele (Pik3r1 flox/flox ) crossed with Mx1-Cre. Pik3r1 flox/flox ;Mx1Cre + and Pik3r1 flox/flox ;Mx1Crelittermate controls were treated with polyI:polyC to induce recombination of the Pik3r1 allele. Animals were permitted to recover from polyI:polyC treatment for at least 12 weeks prior to isolation of bone marrow neutrophils. Phenotypically and morphologically, bone marrow neutrophils isolated directly from the Pik3r1 flox/ flox ;Mx1Cre + and Pik3r1 flox/flox ;Mx1Crebone marrow were similar (data not shown). To control for phagocytosis, neutrophils were incubated with GFP-expressing MRSA for 2 hours, followed by quenching of extracellular GFP using trypan blue. To measure bacterial killing, a second plate was incubated for an additional 2 hours and then quenched with trypan blue. Following quenching, intracellular GFP levels were read immediately on a fluorometer to measure phagocytosis and bacterial killing, respectively. GFP-expressing MRSA were phagocytized equally ( Figure 3F); however, significantly reduced GFP levels, previously shown to correlate with bacterial survival [28], was observed in Pik3r1 flox/flox ;Mx1Cre + neutrophils compared to Pik3r1 flox/flox ;Mx1Creneutrophils ( Figure 3G). Collectively, these findings suggest that increased intracellular ROS production may provide enhanced bacterial killing, in particular of MSSA and MRSA.
Our findings demonstrate that genetic disruption of Pi3kr1 differentially regulates NADPH oxidase activity on the plasma membrane (extracellular ROS production) v. the phagosome membrane (intracellular ROS production). These novel findings are consistent with other studies demonstrating that the regulation of NADPH oxidase activity differs between the plasma and phagosome membranes, and highlights the varied environments of these two compartments. An example relevant to the current work is the Class III PI3K product, PI3P, which is a strong positive regulator of intracellular ROS production liver-derived neutrophils upon transduction with p85α (either in tandem with EGFP or tagged with YFP); C. Immunoblotting for Akt, PI3K catalytic subunit p110δ, and NADPH oxidase subunits in Pik3r1 -/and Pik3r1 -/-+ p85α neutrophils, stimulated for 0 and 23min with SOZ; D. Representative extracellular and E. representative intracellular ROS production measured in Pik3r1 -/and Pik3r1 -/-+ p85α neutrophils stimulated with hIgG-latex; F. Quantitative assessment of hIgG-latex-stimulated extra-and intracellular ROS production, n = 6, *p = 0.04 comparing intracellular ROS production from Pik3r1 -/to Pik3r1 -/-+ p85α; G. Quantitative assessment of SOZ-stimulated extra-and intracellular ROS production, n = 7, **p = 0.05 comparing intracellular ROS production from Pik3r1 -/to Pik3r1 -/-+ p85α; statistical analyses performed by two-tailed, one-sample Student's t-test. www.impactjournals.com/oncotarget but plays no role in extracellular ROS production. A key molecule increasing Class III PI3K activity and increased PI3P production on the phagosome is the Rab GTPase, Rab5. Down-regulation of Rab5 reduces the capacity of S. aureus-containing phagosomes to fuse with endocytic organelles resulting in poorer bacterial killing [29]. Notably, p85α is known to interact with Rab5 and to bear GTPase Activating Protein (GAP) activity towards Rab5-GTP, which is localized to the p85α BH domain (Figure 2A). These considerations suggest that a feasible mechanism underlying the negative regulatory effect of p85 α on intracellular ROS may be downregulating Rab5a-GTP levels via its GAP function [30,31], resulting in reduced Class III PI3K-derived PI3P. Thus, while our current work cannot exclude the possibility that p85α negatively regulates intracellular ROS production in an indirect manner due to altered expression of the Class IA PI3K catalytic subunits, a thought-provoking consideration is that p85 α functions in a Class IA PI3K catalytic subunit-independent manner to regulate NADPH oxidase activity on the phagosome membrane.
Collectively, our findings show that neutrophils lacking the PI3K regulatory subunit p85α produce significantly more intracellular ROS without affecting phagocytosis. Furthermore, this correlates to significantly increased killing of both MSSA and MRSA. These results suggest a new strategy for combating the growing threat of resistant microorganisms.

neutrophil differentiation
Fetal liver cells were collected from embryos at day 14 of gestation, genotyped, and differentiated into neutrophils in α-minimum essential medium with 20% heat-inactivated FCS, 1% penicillin/streptomycin, 50ng/ ml G-CSF and 50units/ml IL-3. Every 2 days, cells were counted and replated at a concentration of 0.5×10 6 /ml in fresh differentiation medium. Activity was analyzed on days 6 and 7 of differentiation [20].

Reintroduction of p85α
p85α cDNA was cloned upstream of an internal entry site (IRES) and the enhanced green fluorescence protein (EGFP) in the bicistronic retroviral plasmid, pMIEG3 [37]. Alternatively, p85α was tagged on the C-terminal end with yellow fluorescent protein (YFP) and cloned into the retroviral plasmid, pMSCV (Clontech). Ecotropic retrovirus containing the vectors was used to transduce Pik3r1 -/fetal liver cells, cells were sorted for EGFP or YFP positivity, and differentiated into neutrophils. Data using both constructs are combined for statistical analyses, as the YFP tag did not alter the www.impactjournals.com/oncotarget function of p85α (data not shown).

neutrophil isolation from bone marrow and MRSA (USA300 LAC) killing
Mice were sacrificed, bone marrow cells were collected from the pelvis, femur, and tibia, and neutrophils were isolated using a 62% and 55% percoll gradient. Neutrophils were washed in Hanks' Balanced Salt solution (Sigma, St. Louis, MO, USA), resuspended in Iscove's Modified Dulbecco's Medium (Life Technologies, Carlsbad, CA, USA), and plated at 2x10 5 cells per well into a 96well plate coated with celltak (Corning, Corning, NY, USA). MRSA (USA300 strain LAC) was incubated with neutrophils for 2h at 37°C, and then plates were washed with warm PBS. For phagocytosis analysis, 50uL of 500mg/mL trypan blue was added immediately to PBS-washed plates to quench extracellular fluorescence. For bacterial survival analysis, PBS-washed plates were incubated for an additional 2h at 37°C followed by the addition of 50uL of 500mg/mL trypan blue. For both phagocytosis and bacterial killing analyses, plates were read on a fluorometer to measure intracellular GFP fluorescence promptly following the addition of trypan blue. Fluorescence intensity correlates to bacterial survival based on the previously defined correlation of HOClbleaching of superfolded GFP to bacterial viability. Abbreviations CGD = chronic granulomatous disease, CR = complement receptor, FcγR = Fcγ receptor, hIgG = Human IgG-opsonized latex beads, HRP = horseradish peroxidase, MPO = myeloperoxidase, MRSA = methicillin-resistant S. aureus, MSSA = methicillin-susceptible S. aureus, PI3K = phosphoinositide 3-kinase, PMN = polymorphonuclear neutrophil, ROS = reactive oxygen species, S. aureus = Staphylococcus aureus, SOD = superoxide dismutase, SOZ = serum opsonized zymosan