Characterization of the binding of cytosolic phospholipase A2 alpha and NOX2 NADPH oxidase in mouse macrophages

Previous studies have demonstrated that cytosolic phospholipase A2α (cPLA2α) is required for NOX2 NADPH oxidase activation in human and mouse phagocytes. Moreover, upon stimulation, cPLA2α translocates to the plasma membranes by binding to the assembled oxidase, forming a complex between its C2 domain and the PX domain of the cytosolic oxidase factor, p47phox in human phagocytes. Intravenous administration of antisense against cPLA2α that significantly inhibited its expression in mouse peritoneal neutrophils and macrophages also inhibited superoxide production, in contrast to cPLA2α knockout mice that showed normal superoxide production. The present study aimed to determine whether there is a binding between cPLA2α-C2 domain and p47phox-PX in mouse macrophages, to further support the role of cPLA2α in oxidase regulation also in mouse phagocytes. A significant binding of mouse GST-p47phox-PX domain fusion protein and cPLA2α in stimulated mouse phagocyte membranes was demonstrated by pull-down experiments, although lower than that detected by the human p47phox-PX domain. Substituting the amino acids Phe98, Asn99, and Gly100 to Cys98, Ser99, and Thr100 in the mouse p47phox-PX domain (present in the human p47phox-PX domain) caused strong binding that was similar to that detected by the human p47phox-PX domain The binding between cPLA2α-C2 and p47phox-PX domains exists in mouse macrophages and is not unique to human phagocytes. The binding between the two proteins is lower in the mice, probably due to the absence of amino acids Cys98, Ser 99, and Thr100in the p47phox-PX domain that facilitate the binding to cPLA2α.


Introduction
The multi-component electron carrier, NOX-2 NADPH oxidase, transfers electrons from NADPH to molecular oxygen to form superoxides, a precursor of microbicidal oxidants. Its subunits include four cytoplasmic components, p47 phox , p67 phox , p40 phox and Rac2, and a hetero-dimeric transmembrane glycoprotein flavocytochrome b 558 composed of gp91 phox and p22 phox (for reviews [1][2][3]). The cytosolic components translocate to the plasma membrane upon stimulation and associate with the flavocytochrome b 558 to form the assembled active oxidase. In resting cells, p47 phox is found in an auto-inhibited form, thereby preventing its binding to membranes [4]. In stimulated cells, the restrictive conformation of the autoinhibitory region of p47 phox is released through phosphorylation of several critical serine residues within its polybasic region [5] and exposing the interactive SH3 domains that direct its translocation to the membranes by binding to specific targets in p22 phox [6][7][8][9]. In addition, it was reported that the PX domain of p47 phox has distinct lipid-binding sites to Phosphatidylinositol 3,4-bisphosphate (PI(3,4)P 2 ) or phosphatidic acid [8][9][10][11][12]. Residues corresponding to His-51, Lys-55, and Arg-70 contribute to forming the atypical binding site of the 47 phox -PX domain to PI(3,4)P 2 . While Lys-55 was shown to be important for recognition and the proper function of phagocyte NADPH oxidase. cPLA 2 α that hydrolyzes phospholipids containing arachidonate at the sn-2 position [13] has been implicated as the major enzyme in the formation of eicosanoids. cPLA 2 α has two functionally distinct domains: an N-terminal C2 domain necessary for Ca 2+ -dependent phospholipid binding and a C-terminal Ca 2+ -independent catalytic region [14]. It was shown that cPLA 2 α translocates from the cytosol to the nuclear membrane and the endoplasmic reticulum by an increase of cytoplasmic [Ca 2+ ] in a variety of cells [15] via its C2 domain [16] in calcium-binding regions [17]. In cPLA 2 α, the C2-domain is structurally designed to target phosphocholine (PC) -rich membrane regions to increase the catalytic domain's enzymatic efficiency, which prefers polyunsaturated PC [18]. The function for cPLA 2 α-C2 domain Tyr 96 was recently reported [19] as a critical specificity determinant for the headgroup of PC, whereas Asn65 tethers with the lipid phosphate moiety and facilitates Ca ions-PC complexation. Further enhancement of C2-domain binding to the membrane, via interaction with Arg59, Arg61, and His 62, was shown when PC-enriched membranes also contained cPLA 2 α activator, ceramide-1-phosphate [20]. It was recently reported [21] that cPLA 2 α undergoes C2 domaindependent oligomerization on membranes independently of its catalytic activity.
We have previously demonstrated an essential requirement for cPLA 2 α in the activation of the assembled phagocyte NADPH oxidase using cPLA 2 α knock out phagocytes like cells that had normal phagocytic activities but did not produce superoxide despite the presence of the assembled oxidase in the membranes after activation, and this activity could be restored by addition of arachidonic acid [22]. In an earlier study, we showed that AA increased the affinity of the assembled NADPH oxidase for NADPH, in purified membranes and in endosomes, separated from stimulated neutrophils [23]. Likewise, specific antisense against cPLA 2 α that blocked its expression and activity in vitro in various phagocytic cells and in vivo in different mouse models of diseases also inhibited the activity of NOX-2 NADPH oxidase [24][25][26]. We also showed that cPLA 2 α activity is required for the oxidase-associated H + channel [27] and oxidase-associated diaphorase activity [28]. We have demonstrated [29] that cPLA 2 α translocates to the plasma membrane by interacting with the assembled oxidase complex in addition to its translocation to nuclear membranes, in peripheral blood neutrophils and phagocyte-like PLB-985 cells. Thus, the ability of cPLA 2 α to colocalize in two different compartments in the same cells enables it to participate in both eicosanoid production and regulate NOX2-NADPH oxidase activation. The activation and translocation of cPLA 2 α by PMA in mouse macrophages [29,30] that does not induce an increase in cytoplasmic [Ca 2+ ], together with its translocation to the plasma membrane, suggesting the existence of alternative pathways for inducing translocation of cPLA 2 α that are distinct from the C2 domain phospholipid-binding mechanism. The requirement of cPLA 2 α for NOX2-NADPH oxidase activation is in line with other studies using inhibitors and antisense molecules [31][32][33][34] but stands in contrast to observations of normal superoxide production by stimulated phagocytes from cPLA 2 α-deficient mice [35]. They reported that cPLA 2 -α inhibition or gene disruption led to complete suppression of neutrophil arachidonate release and eicosanoids biosynthesis but had no effect on neutrophil NADPH oxidase activation, FcγII/III or CD11b surface expression, primary or secondary granule secretion, or phagocytosis of Escherichia coli in vitro. Surprisingly, although all phagocytic functions were normal, they reported that cPLA 2 -α inhibition or gene disruption diminished neutrophil-mediated E. coli killing in vitro. Likewise, E. coli were attenuated in cPLA 2 -α(-/-) mice than wild-type littermates following intratracheal inoculation with live E. coli in vivo.
The present study aimed to explore whether there is a binding between cPLA 2 α and the assembled NADPH oxidase in mouse macrophages, similar to that reported in human phagocytes [36] and rat microglia [24]. Such binding between these two proteins may explain the mechanism and strengthen our results [25], demonstrating that the absence of cPLA 2 α by using oligo antisense against cPLA 2 α inhibits NOX2-NADPH activity in mice phagocytes in vitro and in vivo.

Neutrophil purification
Forty ml blood with neutrophil count between 3-7X10 6 / ml was drawn from healthy volunteers with their written contest. Neutrophils at 95% purity were obtained by Ficoll/ Hipaque centrifugation, dextran sedimentation, and hypotonic lysis of erythrocytes [22]. The study was approved by the institutional Human Research Committee of the Soroka University Medical Center (No. 0370-16-SOR).

Mouse macrophage purification
4 ml of sterile thioglycolate 4% were injected intraperitoneal to ICR male mice. After 4 days, macrophages were washed from the mouse peritoneum with medium RPMI 1640. Cells were cleaned by centrifugation, and hypotonic lysis of erythrocytes was carried out [25]. The study was performed following approval by the Ben-Gurion University of the Negev committee for ethical care and use of animals in experiments, Authorization No. IL-23-05-2017 and was conducted according to the Israeli Animal Welfare Act following the Guide for Care and Use of Laboratory Animal (National Research Council, 1996).

Bacterial expression and purification of recombinant proteins
PCR was used to subclone the mice p47 phox -PX domain in-frame into the expression vector pGEX-4 T-2 using primers containing the BamHI or NotI restriction endonucleases sites (underlined): forward-5′-CGGC GGA TCC ATG GGG GACAC-3′, and reverse-5′-CATCA GCG GCC GCCAC TTT GA AGAAG-3′. The template used was cDNA from mouse macrophages. The GST fusion proteins were overexpressed and purified as described previously [29].

Affinity binding assay
GST or GST fusion proteins attached to glutathione-sepharose beads were added to lysates of stimulated human neutrophils and mouse macrophages and were tumbled endover-end overnight at 4 °C. The samples were centrifuged, washed six times with phosphate-buffered saline, boiled in SDS sample buffer at 95 °C for 5 min, and separated SDS-PAGE before immunoblotting.

Mutagenesis of expression vectors
pGEX-4 T-2 expression vector encoding the cDNA of mouse p47 phox -PX domain was used as a template to generate the desired mutations by the overlap extension polymerase chain reaction [37]. The PCR reactions, using appropriate complementary synthetic oligonucleotides introducing the desired mutation and two additional external primers at the ends of the p47 phox -PX, were performed with Red Load Taq Maste high yield using Thermostable DNA polymerase (LAROVA, Germany).
Val 73I mutation was done on the amplified fragment using the primers: The mutated products were digested with BamHI and NotI and cloned into a GEX-4 T-2 expression vector, digested with the same enzymes. The vectors were then transformed into Escherichia coli DH-101. The mutated fragments were sequenced using the ABI3100 Genetic Analyzer.

Statistical analysis
The mean differences were analyzed by Student's t-test.

Results
Since our previous study demonstrated that there is a binding between the C2-domain of cPLA 2 α and the PX domain of P47 phox , these domains were used to study the binding between cPLA 2 α and P47 phox in mouse macrophages. Peritoneal mouse macrophages and human neutrophils were stimulated with 50 ng/ml PMA for 3 min. An affinity binding assay with human GST-cPLA 2 α-C2 domain, which is identical to the mouse cPLA 2 α-C2 domain was performed in both lysates. Addition of GST-cPLA 2 α-C2 domain to mouse macrophage lysate resulted in binding to p47 phox , which was much lower than the binding to p47 phox in human neutrophil lyzate (Fig. 1A). These results may be due to either the low expression of mouse p47 phox or the lower affinity of the antibodies to mouse p47 phox since the expression of p47 phox is also lower in the lysate of mouse macrophages. Thus, we used another strategy and investigated an affinity binding assay, using GST-p47 phox -PX protein, to further characterize the binding between p47 phox and cPLA 2 α in mouse macrophages. Since p47 phox -PX domains in mice and humans, although very similar, are not identical, mouse GST-p47 phox -PX construct was engineered, and its efficiency to pull down cPLA 2 α was compared to that of the human GST-p47 phox -PX domain. Human GST-p47 phox -PX (PX-H) and mouse GST-p47 phox -PX (PX-M) or GST were added to either stimulated human neutrophil lysates or stimulated mouse macrophage lysates. As shown in Fig. 1B, affinity binding experiments with mouse GST-p47 phox -PX (PX-M) showed significant binding to human or mouse cPLA 2 α but was significantly (p < 0.001) lower in comparison with the binding of human GST-p47 phox -PX (PX-H) to either human or mouse cPLA 2 α although the level of cPLA 2 α is similar in both cell lysates.
To elucidate the reason for the lower binding by mouse GST-p47 phox -PX to cPLA 2 α relatively to that of human GST-p47 phox -PX, we looked for the differences in the amino acid sequences in the binding region of the P47 phox -PX domains in humans and mice. We have recently reported [36] that Ile67 in the cPLA 2 α-C2 resides in a hydrophobic pocket on the surface of the PX domain and interacts with its residues Pro114, His115 in the α4 helix, and Met59 at the end of the α1 helix. In the mouse p47 phox -PX domain, there are changes in some amino acids around the amino acids that participate in the binding ( Fig. 2A) and thus may affect the protein folding and the binding. To determine whether these amino acids affect the binding to cPLA 2 α, two constructs of the mouse p47 phox -PX domain in which the amino acids were substituted by those present in the human p47 phox -PX were engineered. In the first (PX-mut1), Glu66, His68, Thr69, and Val73 were substituted with Ala66, Asn68, Pro69, and Ile73, respectively, and in the second (PX-mut2) Phe98, Asn99, and Gly100 were substituted with Cys98, Ser99, and Thr100, respectively. Figure 2B depicts the interaction between cPLA 2 α-C2 and PX-p47 phox domains as presented in our previous study [29] and the two mutations on PX-p47 phox . As shown in Fig. 3, GST-p47 phox -PX-mut1 (expressing the human residues in mouse PX domain) was much less efficient in binding cPLA 2 α in lysates of stimulated human neutrophils (Fig. 3A) or stimulated mouse macrophages (Fig. 3B) in comparison with wild type mouse GST-p47 phox -PX, suggesting that these amino acids present in the human P47 phox -PX domain disturb the binding between the domains. GST-p47 phox -PX-mut2 was much more effective in binding cPLA 2 α in lysates of stimulated human neutrophils (Fig. 3A) or stimulated mouse macrophages (Fig. 3B) in comparison with wild type mouse GST-p47 phox -PX, suggesting that these amino acids (Cys98, Ser99, and Thr100) present in the human p47 phox -PX contribute to the binding. Moreover, the efficiency of the mouse GST-p47 phox -PX-mut2 to bind cPLA 2 α is similar to that of the human GST-p47 phox -PX domain in both stimulated human neutrophils lysate (Fig. 4A) and stimulated mouse macrophages lysate (Fig. 4B).

Fig. 1
Binding between cPLA 2 α and p47 phox . A Affinity-binding assay between GST-C2 domain and P47 phox : GST-C2 domain fusion protein and GST attached to glutathione beads were added to lysates of stimulated human neutrophils and stimulated mouse macrophages and subjected to Western blot analysis for detection of p47 phox . The samples were separated on 10% SDS gel and subjected to western blot analysis with anti-p47 phox or anti-GST antibodies. The last lanes in the blot indicate the location and expression of p47 phox in the gel. Although GST by itself pulled down p47 phox in neurophil lysate it was significantly lower than that pulled down by GST-C2 domain. The results are from a representative experiment out of three. B Affinitybinding assay between GST-PX domain and cPLA 2 α Human and mouse GST-PX domain fusion proteins and GST attached to glutathione beads were added for 24 h to lysates of stimulated human neutrophils and stimulated mouse macrophages (with 50 ng/ml PMA). The samples were separated on 10% SDS gel and subjected to western blot analysis with anti-cPLA 2 α or anti-GST antibodies. The last lanes in the blot indicate the location and expression of cPLA 2 α and actin. The results are from a representative experiment. **p < 0.01, ***p < 0.001 Shown significance between PX-M and GST and between PX-H and PX-M

Discussion
The results of the present study show that there is a binding between cPLA 2 α and p47 phox mediated by cPLA 2 α-C2 domain and p47 phox -PX in mouse macrophages, similar to that of human phagocytes although with a lower affinity. The binding between the two proteins may provide the mechanism by which arachidonic acid can be released in the oxidase milieu and exert its activation effect. These results are in line with and support our previous results [25], demonstrating that inhibition of cPLA 2 α expression in mice by intravenous injections of specific antisense against cPLA 2 α, caused inhibition of the superoxide production by peritoneal macrophages. The lower binding of the mouse GST-p47 phox -PX (Fig. 1B) is probably attributed to the expression of Phe98, Asn99, and Gly100 instead of Cys98, Ser99, and Thr100, which are expressed in human p47 phox -PX, since their substitution to the human 98-100 amino acid forming the GST-p47 phox -PX-mut2 had increased the binding to cPLA 2 α in both human neutrophil and mouse macrophage lysates (Fig. 3) and was similar to the binding of human GST-p47 phox -PX domain (Fig. 4). Residues 98-100 in the p47 phox -PX domain are not conserved, and there are some differences in other species. Interestingly, when the residues 98-100 are identical in species such as in humans and cows or mice and rats, residues 66,67,69,73 are also identical. Our previous study [29] suggested that Ile67, the first amino acid of the β4 strand of the cPLA 2 α-C2 domain, resides in a hydrophobic pocket on the surface of the p47 phox -PX domain Fig. 2 Comparison between human and mouse p47 phox -PX domains. A Amino acids that are different are not marked with stars, and amino acids that participate in the binding between the p47 phox -PX domain and cPLA 2 α C2-domain in humans are marked with squares. B A model structure depicting the interaction between p47 phox -PX and cPLA 2 α-C2 domains and the two mutations. The p47 phox -PX domain is shown as a ribbon diagram in green and the cPLA 2 α C2-domain in light blue (cyan). The side chains of residues that form the binding (as suggested in our previous study, [29]) are labeled in blue. The side chains residues of mutation 1 are marked in purple and mutation 2 in red. The amino acids residues that interact with the membranes lipids p47 phox -PX domain (His-51, Lys-55, and Arg-70) and the amino acids in the cPLA 2 α-C2 domain that targets PC headgroup (Tyr96 and Asn65) are labeled in grey. (Color figure online) and interacts with its residues Pro114, His115 in the α4 helix, and Met59 at the end of the α1-helix (as depicted in Fig. 2B). The expression of the aromatic Phe98 and the imido Asn99 in the mouse p47 phox -PX domain instead of three polar-uncharged amino, at the end of α2 helix near the α4-helix containing the amino acids Pro114 and His115 that participate in the binding, probably affected the mouse p47 phox -PX structure and reduced its affinity binding to cPLA 2 α-C2 domain. It seems that the expression of amino acids Glu66, His68, Thr69, and Val73 in the mouse p47 phox -PX, found in the free region, are more effective in forming the binding to cPLA 2 α-C2 domain than the amino acids Ala66, Asn68, Pro69, and Ile73 expressed in human p47 phox -PX since their substitution to the amino acids present in the human p47 phox -PX domain (mut1), significantly reduced the affinity binding (Fig. 3). The expression of the negative charged Glu66 in the mouse p47 phox -PX domain instead of Ala66, and the expression of hydrophobic Thr69 instead of Pro69, probably caused significant changes in the three-dimensional structure of the protein and affected the location of the α helices. Thus, mouse amino acids 98-100 in the p47 phox -PX domain caused a reduction in the binding to cPLA 2 α-C2 domain while the mouse amino acids 66,67,69,73 increased the binding to cPLA 2 α yet the whole mouse p47 phox -PX domain, although showed binding to cPLA 2 α-C2 domain, it is lower compared with the human p47 phox -PX domain (Fig. 1B). The existence of the binding between mouse p47 phox -PX and cPLA 2 α-C2 domains enables the translocation of cPLA 2 α to the plasma membranes and binding to the assembled oxidase via p47 phox , similar to other types of phagocytes. These results align with our previous study Comparison between the binding of the human p47 phox -PX domain and mouse mutant 2 p47 phox -PX domain to cPLA 2 α. Affinity-binding assay-human GST p47 phox -PX domain (PX-H) or mouse mutant 2 GST p47 phox -PX domain (PX-Mut2) and GST alone attached to glutathione beads were added to stimulated human neutrophil lysates (A) or mouse macrophage lysates (B). The samples were separated on 10% SDS gel and subjected to western blot analysis with anti-cPLA 2 α or anti-GST antibodies. The last lane in each blot indicates the location of cPLA 2 α. The results are from a representative experiment. The bar graphs represent the means ± SEM of the density ratio of cPLA 2 α to the different GST-PX bands of three experiments in primary rat microglia [24], demonstrating that cPLA 2 α was bound to p47 phox in membranes of activated cells and regulated NOX2 NADPH oxidase activity.
In agreement with our results, it was reported [38] that during phagocytosis of zymosan by mouse peritoneal macrophages, cPLA 2 α translocates in a Ca 2+ -independent manner to form phagosomes with kinetics similar to its translocation to the plasma membrane and before phagolysosome fusion. F4/80, a cell surface macrophage protein highly expressed on resident peritoneal macrophages, was used as a marker to monitor plasma membrane internalization during uptake of zymosan. They showed that F4/80 fluorescence was found around the zymosan particle and on extensions of the plasma membrane adjacent to the forming phagosome, and its localization entirely overlapped with GFP-cPLA 2 α. The translocation of cPLA 2 α to the zymosan forming phagosomes was also demonstrated; colocalization with 5-lipoxygenase, 5-lipoxygenase-activating protein, and leukotriene C4 synthase was found in mouse peritoneal macrophages [39].
In conclusion, the binding between the cPLA 2 α-C2 and p47 phox -PX domains is not unique to human phagocytes; it was demonstrated in mouse macrophages, although to a lower extent. Specific amino acids 98-100 in the human p47 phox -PX domain facilitate the binding to cPLA 2 α, and their absence in the mouse p47 phox -PX domain reduced the binding. But amino acids 66-69and 73 are more efficient in mice than in the human p47 phox -PX domain for the binding formation. The binding between both enzymes suggests that mouse NOX2-NADPH oxidase is also regulated by cPLA 2 α and may provide the molecular mechanism by which cPLA 2 α can activate the assembled oxidase.