Journal list menu
The antimicrobial peptide parabutoporin competes with p47phox as a PKC-substrate and inhibits NADPH oxidase in human neutrophils
Abstract
We investigated parabutoporin (PP), an antimicrobial scorpion peptide, to understand its inhibition on NADPH oxidase in human PMN. We show that PP is a good substrate for all PKC-isotypes, implicated in the activation of NADPH oxidase, and acts as a potent competitive inhibitor of in vitro p47phox-phosphorylation by PKC-α, -βI, -βII and -δ, but not PKC-ζ. In PMN, PP also inhibits the PMA-stimulated phosphorylation of p47phox and its subsequent translocation. In contrast, PP affects the PKC-independent activation to a much lesser degree. This indicates that PP inhibits the activation of NADPH oxidase at submicromolar concentrations in a strongly PKC-dependent manner.
1 Introduction
We have previously isolated and sequenced bioactive peptides from the venom of south African scorpions and most of them either affect excitable cells by interaction with ion channels [1] or cause pore formation in appropriate target cells [2].
One of them, parabutoporin (PP), a 45-mer amphipathic α-helical antimicrobial peptide isolated from the venom of Parabuthus schlechteri scorpions, interacts with polymorphonuclear neutrophils (PMN), thereby stimulating chemokinesis and inducing reversible Ca2+-release from intracellular stores [3]. It also strongly inhibits the activity of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase [4], which is responsible for the generation of superoxide anions 
and functions as a vital primary host defence mechanism against invading microorganisms [5].
In its inactive state, the multiprotein NADPH oxidase is composed of a transmembrane component, cytochrome b 558 (a 91 kDa glycoprotein and a 22 kDa protein) and cytosolic proteins (p47phox, p67phox, and p40phox) and the small G-protein Rac2 bound to its inhibitor RhoGDI [6-8].
Key to the assembly and activation of NADPH oxidase is the generation of activated Rac, which is formed by exchanging GDP for GTP after dissociation from its inhibitor, and also the multiple phosphorylation of p47phox, leading to the translocation of the cytosolic complex p47phox/p67phox/p40phox to the membrane. Certain of these phosphorylations are also implicated in the acquisition of full catalytic activity [9, 10].
Various stimuli such as N-formyl peptide (fMLP), phorbol esters (PMA) and opsonized particles lead to phosphorylation of p47phox [11, 12], mainly catalysed by PKC [13-15], of which PMN express five different isotypes: three conventional Ca2+-dependent PKC-isotypes (-α, -βI and -βII), one novel PKC-δ, which is activated by phosphatidylserine (PS) and diacylglycerol (DAG) but not Ca2+, and the atypical PKC-ζ, which is Ca2+-independent and not activated by DAG [6, 7].
PP is known to activate heterotrimeric G-proteins and subsequently trigger the conversion of Rac to its GTP bound form [4], which is a known activator of NADPH oxidase [8]. Nevertheless, PP strongly inhibits 
production in PMN with an IC50 of about 0.3 μM [4], and the mechanism of this unexpected inhibition remains to be clarified. Here, we have accumulated evidence for a leading role of PKC in these events.
2 Materials and methods
2.1 Parabutoporin and PP double mutant PP-S5A/S13A
PP was initially isolated from the venom of Parabuthus schlechteri scorpions. It was purified by HPLC, sequenced and the synthetic 45 amino acid peptide [16] was used throughout this study. PP-S5A/S13A was synthesized by GLM Biochem (Shanghai, China).
2.2 PMN isolation
Human neutrophils were obtained from the blood of healthy volunteers, purified after centrifugation on Ficoll-Paque (Pharmacia, Sweden) and hypotonic lysis of contaminating red blood cells [16].
2.3 MALDI-TOF analysis
Matrix-assisted laser desorption/ionization (MALDI) time-of-flight (TOF) mass spectrometry was performed as described previously [2].
2.4 Preparation and purification of GST-p47phox
Recombinant GST-p47phox fusion protein was expressed in Escherichia coli (kindly provided by Dr. J. El Benna), purified by affinity chromatography on glutathione-Sepharose beads (Invitrogen, USA) as described previously [17].
2.5 In vitro PKC assay
PKC activity was assayed using a PKC-kit (Upstate, USA). γ-32P-ATP was obtained from MP Biomedicals (USA). Total PKC was derived from PMN resuspended in extraction buffer (25 mM Tris–HCl, pH 7.4, 0.5 mM EDTA, 0.5 mM EGTA, 0.05% Triton-X 100, 5 mM β-mercapto-ethanol, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 0.5 mM PMSF) and activated according to the kit's instructions. Purified PKC-isotypes were obtained from Sigma, USA. For K m studies, reaction mixtures were spotted onto p81 paper (Whatman, UK), washed with a 1% H3PO4-solution and with acetone. Radioactivity was quantified in a liquid scintillation counter (TriCarb 1900 Packard, USA). The reactions for kinetic studies were subjected to 10% SDS–PAGE, stained with Coomassie Brilliant Blue. After excision of the band of interest, 32P-incorporation was quantified as described above.
2.6 PMN fractionation
Isolated PMN were resuspended at 108 cells/ml in relaxation buffer (100 mM KCl, 3 mM NaCl, 3.5 mM MgCl2, 15 mM HEPES, pH 7.3, 1 mM PMSF, 10 μg/ml pepstatin A, 10 μg/ml chymostatin, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 1.25 mM EGTA) and disrupted by sonication (4 × 5 s). After cell debris removal, the supernatant was loaded on a discontinuous sucrose gradient (15–40% sucrose), and centrifuged at 100 000 × g for 60 min at 4 °C in a L5 Ultracentrifuge (Beckman, USA). Cytosolic and membrane fractions were collected, aliquoted and stored at −80 °C.
2.7 NADPH oxidase activation in a cell-free system
PMN membranes and cytosol were mixed in relaxation buffer supplemented with 10 μM FAD, 40 μM GTP, 50 μM lucigenin, in presence or absence of 300 U superoxide dismutase (SOD) (Sigma, USA). Before activation, 75 μM SDS was added. 
production was initiated by adding 200 μM NADPH, and measured in terms of lucigenin-amplified chemiluminescence using a Biolumat 9505 apparatus (Berthold Co, Germany). As the kinetics of activation were the same for all reaction mixtures, we used the height of the peaks to express the results, minus the height of the peak in de SOD reference well.
2.8 Immunoprecipitation of p47phox in 32P labelled PMN
PMN (5 × 106) were incubated in phosphate-free buffer containing 0.2 mCi of 32P orthophosphoric acid (MP Biomedicals, USA). p47phox was immunoprecipitated with anti-p47phox antibody (dilution 1:50) (Santa Cruz, USA) complexed with proteinA-TSK beads (Affiland, Belgium), subjected to 10% SDS–PAGE, blotted and detected by autoradiography.
3 Results
3.1 PP is a good substrate for all PKC-isotypes of PMN
PP contains three serine residues, two of which (Ser 5 and Ser 13) are located in a PKC consensus sequence [18] and were therefore expected to be targets of PKC (Fig. 1 A). This was confirmed by MALDI-TOF analysis, showing that PP (mass 4991 Da) yields two additional peaks at 5071 and 5151 Da, respectively (Fig. 1B) after incubation with PKC, not present in PKC-untreated samples. These results were obtained when either crude PKC or pure PKC-isotype were used. A double mutant of PP, with Serines 5 and 13 substituted by Ala, did not incorporate 32P, confirming that these two residues are the only PKC-targets (data not shown). To further determine to which extent PP could influence the activities of the different PKC-isotypes of PMN, the affinities of PP for each PKC-isotype were quantified (Table 1 ). PP seems to be a good in vitro substrate for all PKC-isotypes, as indicated by the similar K m values ranging from 4.5 to 7.5 μM.
| PKC-α | 5.5 ± 1.5 μM |
| PKC-βI | 7.5 ± 0.9 μM |
| PKC-βII | 6.9 ± 0.3 μM |
| PKC-δ | 5.3 ± 0.9 μM |
| PKC-ζ | 4.5 ± 0.6 μM |
- a Each PKC-isotype was incubated with increasing amounts of PP (ranging from 1 to 20 μM) for time spans during which phosphorylation was time-dependent and linear. K m values are expressed as means ± S.D. (n = 3).
3.2 PP inhibits the in vitro phosphorylation of GST-p47phox by PKC
Because both GST-p47phox and PP are phosphorylated by PKC, the PKC assay in which the global 32P-incorporation is measured cannot give quantitative results concerning the inhibitory effect of PP on the GST-p47phox-phosphorylation. Therefore, reaction mixtures containing GST-p47phox and increasing amounts of PP were subjected to SDS–PAGE (Fig. 2 A), and then 32P-incorporation in GST-p47phox was quantified. Fig. 2B shows a typical result of a dose-dependent inhibition by PP on p47phox-phosphorylation by PKC-α. From these results an apparent IC50 for all five PKC-isotypes was calculated (Table 2 ). These data indicate that PP is a potent competitive inhibitor of in vitro p47phox-phosphorylation by PKC-α and -βI, while the inhibition by PP of p47phox-phosphorylation by PKC-βII and -δ was found to be slightly less efficient, and the activity of PKC-ζ in phosphorylating p47phox was only poorly inhibited by PP.
| Total neutrophil PKC | 2.4 ± 0.3 μM |
| PKC-α | 2.7 ± 0.2 μM |
| PKC-βI | 3.8 ± 0.9 μM |
| PKC-βII | 6.5 ± 0.3 μM |
| PKC-δ | 6.8 ± 0.9 μM |
| PKC-ζ | >9 μM |
- a IC50 values are derived from in vitro PKC assays, and represent the concentration at which PP inhibits 50% of 32P-incorporation in 5 μM GST-p47phox. The results are expressed as means ± S.D. (n = 3).
3.3 PP inhibits superoxide production weakly under PKC-independent conditions
To further correlate the inhibitory effect of PP on 
production in PMN with its property as a PKC inhibitor, we studied the effect of PP on 
production in a cell-free system, activated with SDS. Under such conditions, 
production is independent of PKC activity, as indicated by the inability of ATP-depletion and PKC-inhibitors to inhibit this activation [19].
Our results show that at concentrations below 1 μM, there is a weak PKC-independent inhibition on 
production by PP (Fig. 3
), compared to the more profound inhibition observed in PMN [4]. In contrast, above 1 μM, we notice a stronger PKC-independent inhibition. Since pore forming multimers are known to be generated at these higher concentrations, these results suggest that the PP-multimers also inhibit NADPH oxidase, but in a more PKC-independent manner.
production in a cell free system was assayed as described. Maximum rates of 
production are expressed as a percentage of the amounts of 
production without PP ± S.D. (n
= 3).3.4 PP inhibits the PMA-induced translocation of p47phox in PMN
We next addressed the effect of PP on p47phox-translocation in PMN. PMA-activation results in an increased recruitment of p47phox to the membrane and a corresponding decrease in cytosolic p47phox, relative to unstimulated cells (Fig. 4
). Importantly, after PMA-stimulation, PP inhibits p47phox-translocation in the same concentration range in which it inhibits 
production in PMN.
3.5 PP inhibits the PMA-induced phosphorylation of p47phox in PMN
To further correlate the inhibition of 
production and p47phox-translocation by PP at submicromolar concentrations with PKC inhibition, we next investigated the effect of PP on the PMA-induced phosphorylation of p47phox in PMN (Fig. 5
). Stimulation with PMA leads to a severe increase in p47phox-phosphorylation, which is significantly inhibited after pre-treatment with PP.
4 Discussion
The strong inhibitory effect that PP exerts on fMLP- or PMA-induced 
production in PMN [16] has been difficult to explain. To our knowledge, only few other natural peptides are known to inhibit 
production in neutrophils: PR-39 and mastoparan, with IC50 values of about 5 and 2 μM, respectively [20, 21]. PR-39, a Pro-Arg rich peptide isolated from porcine PMN, was reported to bind to the SH3-domains of p47phox, thereby blocking the association between p47phox and cytochrome b
558 [20]. Mastoparan, a 14-mer cationic peptide derived from wasp venom, was shown to bind cytosolic components of NADPH oxidase, especially p67phox [21].
Although binding studies with PP also indicated affinity of PP for some NADPH components, especially for p47phox [4], the IC50 on 
production in PMN is about 10–15 times lower for PP than for mastoparan and PR-39, suggesting that PP might interfere with NADPH oxidase in multiple ways. The inhibition of NADPH oxidase by PP at submicromolar concentrations is certainly not caused by cell lysis or damage, as evidenced by Trypan Blue staining [16] and the observation that PP induces a reversible Ca2+-release from intracellular stores [3].
Here, we demonstrate in vitro that PP is a good substrate for all PKC-isotypes expressed in PMN. When comparing this data with the similar K
m values of each PKC-isotype for p47phox [14], we need to be careful, since p47phox has up to 10 target serine residues of which many are phosphorylated by PKC-α, -βII and -δ, but significantly less by PKC-ζ. In contrast to p47phox, PP contains only two phosphorylation sites for PKC. Therefore, PP would be the best substrate for PKC-α and -βI, a good substrate for PKC-βII and -δ, and a weak substrate for PKC-ζ, relative to p47phox. Consistent with these findings were the results of experiments in which we investigated the effect of PP as a competitive inhibitor on p47phox-phosphorylation by PKC-isotypes. Again, PKC-ζ is deviant from the other PKC-isotypes. The fact that PKC-ζ activity is less inhibited by PP as evidenced by its higher IC50 value is in agreement with the observation that – in neutrophils – PKC-ζ is implicated in chemotaxis and chemokinesis [22], activities which are stimulated by PP [4]. It has also been shown that inhibition of PKC-ζ expression does not affect PMA-induced 
production in HL-60 cells [23].
Importantly, PP inhibits p47phox-translocation and p47phox-phosphorylation at submicromolar concentrations which also inhibit 
production in PMN. In the same concentration range, the SDS-stimulated activation of NADPH oxidase is only inhibited to a small extent. On the other hand, PP seems to be much more efficient in preventing 
production, p47phox-phosphorylation and p47phox-translocation in PMN than inhibiting PKC in vitro, when comparing the respective IC50 values of PP in these assays. In an attempt to explain the difference between ex vivo and in vitro IC50 on p47phox-phosphorylation by PP, we note that PP can increase its concentration considerably by inserting itself in the membrane, as shown previously by confocal microscopy [4], and could thus inhibit the activity of PKC-α and -βII, which have been reported to translocate with p47phox to the membrane [15, 24].
Scorpion venom contains many peptides, most of which act on ion channels. The antimicrobial activity of PP protects the venom gland of Parabuthus scorpions against infection, while pore formation properties enhance the effect of ion channel toxins. It is still unclear whether the inhibition of PKC-driven p47phox-phosporylation by PP plays an evolutionary role in the action of scorpion venom peptides, and further experiments have to show whether the interaction of PP-like peptides with different PKC-isoforms in mammalian cells is more than an epiphenomenon and may serve still unknown functions in the defensive (catching of prey) and offensive (deterring agressors) strategy of scorpions. In conclusion, this study shows that the property of PP as a competitive PKC inhibitor largely contributes to the impaired superoxide production in PMN at submicromolar concentrations.
Acknowledgements
Elke Clynen is a postdoctoral fellow of the Fund for Scientific Research-Flanders (FWO-Vlaanderen). We thank Elke Wybaillie for technical assistance and the ‘Dienst voor het Bloed, Rode Kruis – Vlaanderen’ for blood supplies.
