Peroxiredoxin II functions as a signal terminator for H 2 O 2 -activated phospholipase D1 Nianzhou Xiao, Guangwei Du and Michael A. Frohman Department of Pharmacology and the Center for Developmental Genetics, University Medical Center at Stony Brook, NY, USA Mammalian phospholipase D (PLD) is a signal-trans- ducing enzyme that hydrolyzes PtdCho to generate the membrane-bound lipid signal phosphatidic acid (PA) (reviewed in [1,2]). PA is a second messenger and can be further converted into diacylglycerol. PLD is indi- rectly activated in response to cellular stimulation by various extracellular agonists including hormones, growth factors, neurotransmitters, adhesion molecules, cytokines and physical stimuli. The direct mechanism by which PLD is activated involves physical interac- tion with protein kinase C (PKC) and the ADP-ribosy- lation factor (ARF) and RhoA small GTPase families. A well-studied role for PLD in stimulation of NADPH during respiratory oxidative burst has been described by may groups [3–6] (reviewed in [7]). PLD functions both directly, by generating PA, which binds to and stimulates the p47(phox) component of the NADPH oxidase complex [5,8], and by conversion of some of the PA into diacylglycerol. Diacylglycerol recruits PKC to the plasma membrane, which is also required for NADPH activation [9,10]. Once NADPH oxidase is activated, it generates H 2 O 2 , which can func- tion to kill intracellular bacteria and play pro-apopto- tic or anti-apoptotic roles depending on the cellular context. In addition, H 2 O 2 stimulates PLD activity by a poorly understood, probably indirect, mechanism involving tyrosine kinases and PKC [11–13]. This cre- ates a runaway positive feedback cycle: PLD activation promotes H 2 O 2 production and PKC recruitment, which leads to even more PLD activity. This paper reports the identification of a cellular mechanism by which this positive feedback cycle may be regulated and terminated. In addition to the extensively documented interac- tions between PLD1 and the proteins (PKC, ARF and RhoA) that stimulate it directly [14–18], interactions involving other proteins, such as actin, protein kinase N, casein-kinase-2-like serine kinase and amphiphysin, Keywords hydrogen peroxide; peroxiredoxin II; phosphatidic acid; phospholipase D1; PMA Correspondence M. Frohman, Center for Developmental Genetics, 438 CMM, Stony Brook, NY 11794-5140, USA Fax: +1 631 632 1692 Tel: +1 631 632 1476 E-mail: michael@pharm.sunysb.edu (Received 4 May 2005, revised 3 June 2005, accepted 8 June 2005) doi:10.1111/j.1742-4658.2005.04809.x Phospholipase D1 (PLD1) is a signal-transduction regulated enzyme which regulates several cell intrinsic processes including activation of NAPDH oxidase, which elevates intracellular H 2 O 2 . Several proteins have been reported to interact with PLD1 in resting cells. We sought to identify pro- teins that interact with PLD1 after phorbol 12-myristate 13-acetate (PMA) stimulation. A novel interaction with peroxiredoxin II (PrxII), an enzyme that eliminates cellular H 2 O 2 , which is a known stimulator of PLD1, was identified by PLD1-affinity pull-down and MS. PMA stimulation was con- firmed to promote physical interaction between PLD1 and PrxII and to cause PLD1 and PrxII to colocalize subcellularly. Functional significance of the interaction was suggested by the observation that over-expression of PrxII specifically reduces the response of PLD1 to stimulation by H 2 O 2 . These results indicate that PrxII may have a signal-terminating role for PLD1 by being recruited to sites containing activated PLD1 after cellular stimulation involving production of H 2 O 2 . Abbreviations HA, hemagglutinin; MobA, molybdopterin guanine dinucleotide biosynthesis protein AI; PA, phosphatidic acid; PLD, phospholipase D; PKC, protein kinase C; PMA, phorbol-12-myristate 13-acetate; PrxII, peroxiredoxin II. FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS 3929 have also been reported [19–22]. These studies, how- ever, may only have identified a fraction of the proteins with which PLD1 interacts, as they were carried out using cells at rest (not undergoing stimulation), and we have found that PLD1 undergoes regulated transloca- tion from perinuclear membrane vesicles to the plasma membrane and back as part of the cellular response to agonist signaling [23], raising the possibility that PLD1 acquires different protein partners once it becomes acti- vated and ⁄ or as it transits through the cell. In this report, we describe one such interaction with peroxi- redoxin II (PrxII) which may serve as a mechanism for signal termination when the function of PLD1 involves stimulating increased production of H 2 O 2 . Results PLD activation pattern after stimulation with phorbol-12-myristate 13-acetate (PMA) PMA directly activates PKC and stimulates PLD1 by direct and indirect mechanisms (reviewed in [2,24]). We reported previously that stimulation of COS-7 cells with PMA for 2 h causes PLD1 to translocate from perinuclear membrane vesicles to the plasma membrane [23]. However, the timing of PLD1 activation in this system was not examined, and in other cell types, PLD1 activation can be quite rapid. To maximize the likelihood of identifying proteins newly interacting with PLD1 in its activated state, we first examined how long it takes PLD to reach peak activity after stimulation. PMA stimulation of PLD1 activation was examined using CHO cell lines stably transfected with Tet-indu- cible PLD1 expression constructs with which hemagglu- tinin (HA)-tagged PLD1 can be expressed efficiently. Butan-1-ol was added to the cultures for a 2-min win- dow to obtain brief, successive snapshots of PLD acti- vity through the accumulation of phosphatidylbutanol at different times after the initiation of the PMA stimu- lation (Fig. 1). PLD activity increased rapidly (within 5 min) after the addition of 100 nm PMA and reached peak levels at 10 min, which was subsequently taken as the standard time for which to activate PLD1 using PMA for the purpose of identifying and examining sti- mulus-dependent protein interactions. Identification of PrxII as a PMA-promoted PLD1-interacting protein HA-tagged PLD1 was induced in the CHO cells, and affinity pull-down performed before and after 10 min of 100 nm PMA stimulation. In brief, the resting stage cells and stimulated cells were harvested and exposed to anti-HA beads to pull-down HA-PLD1 and proteins potentially in complex with it. The protein samples were separated by SDS ⁄ PAGE (12% gel) followed by silver staining. Several bands exhibited different levels of intensity between the two samples (Fig. 2). Interest- ing bands were processed for MS analysis at the SUNY-Stony Brook CASM facility. The most promin- ent band ( 22 kDa; indicated by the arrow in Fig. 2) was identified as PrxII (Fig. 3). PrxII is a 22 kDa pro- tein that belongs to the peroxiredoxin family [25]. Peroxiredoxins use redox-active cysteines to reduce peroxides and thus protect many types of enzyme from oxidation [26]. Peroxiredoxin antioxidant activity is linked to many signaling pathways, including enhance- ment of natural killer cell activity, cell proliferation and differentiation, heme metabolism, immune response and apoptosis [26]. Peroxiredoxin activity and function have been associated with human disease in many contexts, in particular carcinogenesis and aging [27–29]. To rule out the possibility of inadvertent contamin- ation during the sample processing for MS analysis, the PLD1 affinity precipitation experiment was repea- ted and analyzed using western immunoblotting (Fig. 4). HA-PLD1 and PrxII were visualized using antibody to HA and a rabbit anti-PrxII IgG, respec- tively. PLD1 induction did not affect PrxII concentra- tion (Fig. 4A), and PLD1 pulled-down approximately fivefold more PrxII after PMA stimulation (Fig. 4B). The amount of PrxII pulled-down by PLD1 from unstimulated cells varied from none (as shown in Fig. 2) to small amounts (as shown in Fig. 4). In all Fig. 1. PLD activity time course upon PMA stimulation. In vivo transphosphatidylation PLD assays were performed as described in Experimental procedures. Cells were stimulated with PMA for 0, 3, 8, 28 and 58 min before the addition of butan-1-ol, followed by an additional 2 min of culture and assay termination using ice-cold methanol. The experiment was performed three times with similar results. Regulation of PLD1 by interaction with PrxII N. Xiao et al. 3930 FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS cases, however, a substantial increase in PrxII pull- down was observed after cellular stimulation. PrxII complexes with and precipitates PLD1 in an in vitro binding assay To confirm the PLD1–PrxII protein interaction revealed by HA-PLD1 affinity precipitation, we per- formed an in vitro assay using recombinant PrxII and PLD1. His 6 -tagged PrxII protein was expressed and purified from Escherichia coli, mixed with sf9 cell lysates containing baculoviral-generated GluGlu-tag- ged PLD1, pulled-down using Ni ⁄ nitrilotriacetate ⁄ agarose, and analyzed using SDS ⁄ PAGE and western blotting (Fig. 5). To address nonspecific binding, pull- down of PLD1 by Ni ⁄ nitrilotriacetate ⁄ agarose alone was determined (Fig. 5, lane 3), and also the extent to which it was pulled-down by His 6 -tagged molybdopterin guanine dinucleotide biosynthesis pro- tein A (MobA; lane 2). MobA is a 21 kDa nucleic acid-binding protein [30,31] which would not be a likely candidate to interact with PLD family members. MobA-His 6 protein was expressed and purified from E. coli (kindly provided by J. Daniels, SUNY-Stony Brook). Only a very small amount of GluGlu-PLD1 protein was pulled-down by the Ni ⁄ nitrilotriacetate ⁄ ag- arose alone. More nonspecific pull-down was observed in the presence of MobA, which is expected as PLD1 Fig. 3. MS-MS identification of PrxII as a PMA-stimulated PLD1- interacting protein. Protein bands were excised from the polyacryl- amide gel as described in Experimental procedures and processed for MS analysis. MS-MS identified a 22-kDa protein, PrxII, at score 103 (protein scores greater than 71 are significant. P < 0.05). The arrow denotes the PrxII score. Peptides that matched with PrxII sequences are shown in italic underline. A single hit of significant (85) but lesser significance was observed for another protein. As only one hit was observed, this candidate was not pursued further. AB Fig. 4. PMA promotes interaction of PLD1 with PrxII. CHO cells inducibly expressing HA-PLD1 and variably stimulated with PMA were immunoprecipitated using anti-HA matrix and analyzed by western blotting as described in Experimental Procedures. The HA-PLD1 and endogenous PrxII proteins were imaged using a rat monoclonal antibody to HA and rabbit anti-PrxII serum, respectively. (A) HA-PLD1 and PrxII in the whole cell lysate. (B) PLD1 (120 kDa) and PrxII (22 kDa) probed by their specific antibodies. Representa- tive of three experiments. IP, immunoprecipitation; Tet, doxy- cycline. Fig. 2. Identification of a protein that exhibits enhanced interaction with PLD1 after PMA stimulation. HA-tagged PLD1 was induced in CHO cells, and affinity pull-down performed before and after a 10-min PMA stimulation. The protein samples were analyzed by SDS ⁄ PAGE (12% gel) followed by silver staining. The arrow indi- cates the band that was determined to be PrxII in MS as described subsequently. The results are representative of three independent experiments. N. Xiao et al. Regulation of PLD1 by interaction with PrxII FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS 3931 is a hydrophobic, sticky protein [1]. After correction for the Ni ⁄ nitrilotriacetate ⁄ agarose nonspecific pull- down and normalization to the relative amounts of PrxII and MobA used in the assay (Fig. 5A, bottom panel), quantitation of the amount of PLD1 pulled- down revealed that His 6 -PrxII pulled-down PLD1 40- fold more efficiently than His 6 -MobA (Fig. 5B). PLD1 and PrxII colocalize after cellular stimulation Lysis of cells before affinity pull-down creates the opportunity for proteins normally localized in different subcellular compartments to associate and hence gen- erate false positive results [32]. We thus used immuno- fluorescent detection and confocal microscopy to examine the PLD1 and PrxII subcellular patterns of localization. PMA was used to stimulate CHO cells in which HA-PLD1 had been transfected, and the HA-PLD1 and endogenous PrxII were visualized and imaged as described in Fig. 6. In resting cells, PLD1 exhibited perinuclear membrane vesicle localization, whereas PrxII localized to both the cytoplasm and possibly some membrane vesicles of unknown identity that did not contain PLD1 (Fig. 6, top row). In con- trast, in cells stimulated with PMA for 10 min, whereas PLD1 still exhibited perinuclear localization, PrxII now localized to both the cytoplasm and the perinuclear vesicles containing PLD1 (Fig. 6, lower panel). Thus, PLD1 and PrxII partially colocalize before lysis in PMA-stimulated cells, suggesting that the interaction between them is physiological rather than an artifact of lysis. These results also support, using a visual rather than molecular biochemical approach, the proposal that PMA triggers increased interaction between PLD1 and PrxII. PrxII over-expression inhibits H 2 O 2 -stimulated PLD1 activity A priori, the interaction between PrxII and PLD1 may affect PLD1 activation directly in a general context, or PrxII effects on PLD1 activity may be restricted just to the context in which H 2 O 2 is the agonist responsible for stimulating PLD1 activation. To explore this, we examined the consequences of PrxII over-expression on PLD1 activation in the context of stimulation by PMA in comparison with stimulation by H 2 O 2 . PMA stimulates PLD1 by activating PKC, which is a direct activator of PLD1 and does not require H 2 O 2 to medi- ate its stimulation [15]. PrxII was over-expressed in CHO cells inducibly expressing PLD1. An in vivo PLD assay was per- formed in which the cells were stimulated with either PMA or H 2 O 2 for 10 min (Fig. 7). PrxII over-expres- sion did not affect basal PLD1 activity or PMA-stimu- lated PLD1 activation, but completely ablated stimulation of PLD1 by H 2 O 2 . This shows that PrxII can function as a negative regulator of PLD1 activa- tion by H 2 O 2 . Discussion We describe here the identification of a novel PLD1- interacting protein, PrxII. Although PLD1-interacting proteins have been identified previously, this is the first report of an interaction that is promoted in the context of PLD1 activation. Our original goal was to identify novel proteins that interact with PLD1 as it transits from perinuclear vesicles to the plasma membrane and back through endosomes. In our previous report on A B Fig. 5. In vitro pull-down of PLD1 by PrxII. GluGlu-PLD1 and puri- fied His 6 -PrxII and His 6 -MobA proteins were prepared as described in Experimental procedures, mixed as indicated, and pulled-down using Ni ⁄ nitrilotriacetate ⁄ agarose. (A) Analysis of pull-down by SDS ⁄ PAGE (8% gel for GluGlu-PLD1; 12% gel for His 6 -PrxII and His 6 -MobA) followed by western blotting using antibodies to His 6 and GluGlu. (B) Quantification of band intensity using Odyssey soft- ware. Assay backgrounds (lane 3 in A) were subtracted from lanes 1 and 2, and then the PLD1 IP values (top row, lanes 1 and 2) nor- malized to the amounts of PrxII and MobA used for the immuno- precipitation (bottom row, lanes 1 and 2). The scale of the y-axis was set such that the amount of PLD1 immunoprecipitated by MobA is equal to 1. Representative of three independent experi- ments. IB, immunoblot. Regulation of PLD1 by interaction with PrxII N. Xiao et al. 3932 FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS PLD1 translocation, we showed that PLD1 can be found at the plasma membrane of COS-7 cells 2 h after the initiation of PMA stimulation, and that the translocation is facilitated by PLD1 activation [23]. In this study, we searched for proteins that interact with PLD1 in a stimulation-dependent manner at the peak of PLD1 activity (10 min, Fig. 1), which precedes PLD1 translocation to the plasma membrane (Fig. 6). Other stimulation-dependent interactions that occur after PLD1 translocates to the plasma membrane may occur and await discovery. What might be the functional significance of a stimu- lation-dependent PLD1–PrxII interaction? As described in the introduction and illustrated in Fig. 8, functional interaction of the NADPH oxidase complex, H 2 O 2 and PLD1 forms a positive feedback loop. When PLD1 is activated, it hydrolyzes PtdCho to produce PA. PA sti- mulates the NADPH oxidase complex to generate H 2 O 2 [5,33,34]. H 2 O 2 is a small, diffusible molecule and func- tions in part as a second messenger. Both its production and its elimination (signal termination) are important Resting PMA- Stimulated Fig. 6. PLD1 and PrxII exhibit partial colocal- ization after PMA stimulation. Immunofluo- rescent detection of HA-PLD1 and endogenous PrxII in resting and PMA-stimu- lated CHO cells (63 ·). Rabbit antibody to PrxII and an Alexa-647-labeled goat anti- rabbit IgG secondary were used to visualize PrxII. Mouse monoclonal antibody to HA and an Alexa 488-labeled goat anti-mouse secondary were used to visualize HA-PLD1. Fig. 7. PrxII over-expression inhibits PLD1 activation by H 2 O 2 but not PLD1 activation by PMA. PrxII-pCR3 was over-expressed in CHO cells inducibly expressing PLD1. A PLD in vivo assay was per- formed to record PLD activity during the first 10 min of stimulation by H 2 O 2 (0.4 mM) or PMA (0.1 mM). The experiment was performed in triplicate and is representative of three experiments with similar outcomes. Sudent’s t-test was used to establish significance. Fig. 8. Working model for PLD1–PrxII function interaction. Extracel- lular stimulation establishes a positive feedback loop wherein acti- vation of PLD1 generates PA, which leads to stimulation of NADPH oxidase complex generation of H 2 O 2 , which further stimulates PLD1 generation of PA. H 2 O 2 can participate in many signaling pathways, including both pro-apoptotic and anti-apoptotic ones. PrxII is proposed here to function as a signal terminator, eliminating H 2 O 2 through oxidation and thereby decreasing PLD1 activity in addition to inhibiting other H 2 O 2 effector pathways. N. Xiao et al. Regulation of PLD1 by interaction with PrxII FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS 3933 for many signaling cascades [12,33,35]. Among these cascades, one involves the indirect stimulation of PLD activity, as demonstrated here and reported previously, although the mechanism by which it promotes PLD activation remains incompletely defined [11–13]. Thus, PA production leads to H 2 O 2 production, which sti- mulates more PA production. As described here, the identification of PrxII as a stimulation-dependent PLD1-interacting protein suggests that it would func- tion as a signal terminator, as the oxidation of H 2 O 2 in the local region surrounding PLD1 would reduce the activity of the indirect pathways that stimulate PLD1 activity, which would then lower the concentrations of PA (which is rapidly metabolized, once made), and in turn dampen NADPH oxidase activity. Support for this hypothesis comes from the previously described obser- vation that the p29 peroxiredoxin can be found in association with NADPH in neutrophils [36]. Does PrxII interact only with PLD1? A second iso- form, PLD2 [37], can be found at the plasma mem- brane in many cell types [37,38] and has been reported to stimulate NADPH in vascular smooth muscle cells [39]. Although we have not examined whether PLD2 and PrxII physically interact, this observation suggests that PrxII may play a role in the regulation of both isoforms. On the other hand, PLD1 appears to be the relevant isoform in other cell types, such as neutrophils stimulated through the Fcc receptor [40]. Having established that PrxII can function as a sig- nal terminator in an artificial situation, i.e. CHO cells inducibly over-expressing PLD1 in the presence of PrxII, the next step will be to demonstrate that endo- genous concentrations of PrxII function as PLD signal terminators in relevant cell types, such as neutrophils. Experimental procedures General reagents Cell culture media (Opti-MEM-I, Dulbecco’s modified Eagle’s medium and F-12) were obtained from Gibco-BRL (Gaithersburg, MD, USA), fetal bovine serum was from Clontech (Mountain View, CA, USA), and complete Grace’s Medium, LipofectAmine Plus and Cellfectin rea- gent for cell transfection from Invitrogen (Carlsbad, CA, USA). Antibiotics were obtained as follows: doxycycline (Sigma, St Louis, MO, USA), gentamicin (Fisher, Pitts- burgh, PA, USA), penicillin ⁄ streptomycin (Cellgro, Hern- don, VA, USA), and zeocin and blasticidine (Invitrogen). l-Dipalmitoyl PtdCho [choline-methyl- 3 H] ([ 3 H]PtdCho) was purchased from American Radiolabeled Chemicals, Inc. (St Louis, MO, USA), and protease inhibitor cocktail from Roche (Basel, Switzerland). The following antibodies were used: 3F10 antibody (rat monoclonal antibody to HA tag) from Roche Molecular Biochemicals (Indianapolis, IN, USA); goat anti-mouse IgG conjugated with Alexa 488 and 568 from Molecular Probes (Eugene, OR, USA). Rabbit polyclonal antibody to PrxII was a gift from S.G. Rhee (NIH, Bethesda, MD, USA). MobA protein (generated using bacterial expression and purified using nickel resin affinity chromatography and HPLC) was a gift from J. Daniels (Stony Brook University). All other reagents were of analytical grade unless otherwise specified. Cell culture and transfection Cell cultures (except Sf9 cells) were maintained in a humi- dified atmosphere containing 5% (v ⁄ v) CO 2 at 37 °C as des- cribed previously [23]. CHO T-REx PLD1 stable cell lines were maintained in conditioned complete medium [F-12 with 10% (v ⁄ v) tetracycline-free fetal bovine serum, 300 lgÆmL )1 zeocin, 10 lgÆmL )1 blasticidine, 100 UÆmL )1 penicillin and 100 lgÆmL )1 streptomycin]. For transient transfection, cells at 80% confluence were switched into Opti-MEM I, and transfected with 1 lg DNA per (3–4) · 10 5 cells using Lipo- fectAmine Plus. At 3 h after transfection, the medium was replaced with conditioned complete medium. Sf9 cells were maintained in a humidified atmosphere at 27 °C in complete Grace’s medium as described previously [1]. To transfect the cells, 9 · 10 5 Sf9 cells were seeded in 35-mm tissue culture plate in fresh medium. A mixture of PLD1-GluGlu-pFASTBAC DNA and Cellfectin reagent was added to the cells, which were then incubated at 27 °C for 5 h. At 72 h after transfection, virus-containing medium was collected, and the titer determined. In vivo PLD assay Nearly confluent PLD1-induced CHO stable cells cultured in 35 mm plates were labeled with 4 lCiÆmL )1 [ 3 H]myristic acid and serum-starved overnight (F-12 medium), using previously described methods [41]. At 24 h after labeling, PMA (100 nm) was added to the cells for various lengths of time. The medium was then spiked with butan-1-ol to a final concentration of 0.3%, incubated for another 2 min, and collected into 600 lL ice-cold methanol. Total cellular lipids were then extracted and analyzed on Whatman LK5DF silica gel 150A TLC plates using previously pub- lished protocols [23]. PLD activity was expressed as the ratio of [ 3 H]phosphatidylbutanol to total 3 H-labeled lipids. Affinity pull-down assay Nearly confluent PLD1-induced CHO stable cells were grown on 150-mm plates and serum-starved overnight. The cells were stimulated with PMA (100 nm) for 10 min, harvested, and lysed in a mixture containing 60 mm Regulation of PLD1 by interaction with PrxII N. Xiao et al. 3934 FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS n-octyl-b-glucopyranoside, 0.5 mm EDTA, 50 mm Tris ⁄ HCl, pH 7.5, 150 mm NaCl, and 1 · protease inhib- itor cocktail at 37 °C for 20 min. The lysate was spun down at 50 000 g for 15 min. The supernatant was mixed with pre-equilibrated anti-HA affinity matrix and rocked at 4 °C for 2 h. The lysate ⁄ matrix was pelleted and washed three times with the lysis buffer. To elute the matrix-bound pro- teins, 1 mgÆmL )1 HA peptide (Roche) was incubated with the matrix at 37 °C for 20 min, and the supernatant collec- ted after brief centrifugation. Then 2 · SDS ⁄ PAGE sam- ple-loading buffer containing urea (8 m) was added for subsequent analysis by SDS ⁄ PAGE. Immunofluorescent analysis Cells were cultured on 35 mm tissue culture plates contain- ing coverslips. After transfection and ⁄ or treatment with reagent, the cells were washed five times with ice-cold NaCl ⁄ P i , fixed in 2% (v ⁄ v) formaldehyde for 10 min, and permeabilized with 0.1% (v ⁄ v) Triton X-100 in NaCl ⁄ P i for 10 min at room temperature. The fixed cells were washed three times in NaCl ⁄ P i , blocked with 5% (w ⁄ v) BSA and 5% (v ⁄ v) goat serum for 1 h, and then incubated with pri- mary antibody in NaCl ⁄ P i containing 5% (v ⁄ v) goat serum for 1 h. After being washed three times, the cells were stained with secondary antibodies in NaCl ⁄ P i with 5% (v ⁄ v) goat serum for 1 h in the dark. After another wash, the cells were mounted on slides using mounting medium (Vector, Burlingame, CA, USA). Images were captured using confocal microscopy (Leica Microsystems Inc.). Recombinant protein expression in E. coli A recombinant PrxII-pQE construct containing a His 6 tag at the C-terminus was generously given by Professor Rong- Nan Huang, National Central University, Taiwan [42]. The PrxII-His 6 protein was transformed into E. coli XL10-Gold and induced with 1 mm isopropyl thio-b-d-galactoside at room temperature for 4 h, D  0.6. The cells were harves- ted by centrifuging and lysed in 5 mL lysis buffer (10 mm Tris ⁄ HCl, pH 7.5, 300 mm NaCl, 10 mm imidazole, 5mgÆmL )1 lysozyme, 10 lgÆmL )1 RNase, 5 lgÆmL )1 DNase) per 50 mL culture cells. After a brief sonication, the lysate was centrifuged at 10 000 g for 20 min, and the supernatant recovered. All steps were performed at 4 °C. Purification of PrxII-His 6 protein The supernatant recovered as described above was mixed with pre-equilibrated Ni ⁄ nitrilotriacetate ⁄ agarose (Qiagen, Valen- cia, CA, USA), and rocked at 4 °C for 1 h. The lysate ⁄ Ni ⁄ nitrilotriacetate bead mixture was transferred to a poly prep chromatography column and the agarose packed. The column was washed twice with wash buffer (10 mm Tris ⁄ HCl, pH 7.5, 300 mm NaCl, 20 mm imidazole), and PrxII-His 6 protein was eluted using 10 mm Tris ⁄ HCl, pH 7.5, containing 300 mm NaCl and 250 mm imidazole. Protein concentration and integrity were confirmed by SDS ⁄ PAGE (12% gel). Baculoviral production of PLD1 The PLD1-GluGlu-pFASTBAC construct was used to pre- pare PLD1 protein as described previously [1]. In brief, recombinant baculoviruses were generated in the Bac-to-Bac baculovirus expression system (Life Science). To express PLD1, exponentially growing Sf9 cells (200–300 mL cells at a density of 1 · 10 6 ÆmL )1 ) were seeded in a 500-mL spinner flask and infected with recombinant baculoviruses at a mul- tiplicity of 10. The infected cells were grown for 48 h and pelleted at 2000 g for 5 min. After being washed, the pellet was lysed with 5 mL lysis buffer (1% Nonidet P40, 20 mm Tris ⁄ HCl, pH 7.5, 1 mm EDTA, 1 mm dithiothreitol, 20 lm leupeptin, 0.1 mm phenylmethanesulfonyl fluoride, 0.1 mm benzamidine) by incubation on ice for 30 min. The lysate was centrifuged at 50 000 g for 30 min, and the supernatant recovered for use in the in vitro PLD1 and PrxII interaction assays. All steps were performed at 4 °C. In vitro PLD1 and PrxII binding assay Purified His 6 -PrxII, His 6 -MobA, and PLD1 were prepared as described above. PrxII or MobA were added to aliquots of PLD1-containing lysate and mixed well. Ni ⁄ nitrilotriace- tate ⁄ agarose (Qiagen) was added to the protein mixture and rocked for 1 h. Elution steps were performed as des- cribed above. All steps were performed at 4 °C. Western blotting Protein samples were separated by SDS ⁄ PAGE (8% or 12% gel) and transferred to nitrocellulose membrane in semidry transfer buffer (25 mm Tris, 250 mm glycine, 15% methanol) using the Pantherä Semidry Electroblotter apparatus. The blots were incubated with Odyssey blocking buffer [1% (w ⁄ v) casein in Tris-buffered saline] and primary antiserum [in 1% (w ⁄ v) casein in Tris-buffered saline ⁄ Tween 20], fol- lowed by washing with Tris-buffered saline ⁄ Tween 20 four times, 5 min each time. The blot was then incubated in fluor- escent secondary antibody in the dark. After a wash with Tris-buffered saline ⁄ Tween 20 and NaCl ⁄ P i , blots were scanned using an Odyssey machine (LI-COR Inc., Lincoln, NE, USA) to image the resulting signals. Silver staining After SDS ⁄ PAGE, gels were washed briefly with double- distilled water and fixed in 50% (v ⁄ v) ethanol, 10% (v ⁄ v) N. Xiao et al. Regulation of PLD1 by interaction with PrxII FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS 3935 acetic acid and 40% (v ⁄ v) double-distilled water for 2 h at room temperature. The gels were then incubated in 30% ethanol, double-distilled water and sensitization solution [1% (w ⁄ v) ProteoSilver Sensitizer in double-distilled water] each for 10 min, respectively, and washed twice with dou- ble-distilled water, for 10 min each time. The gels were placed in silver solution [1% (w ⁄ v) silver solution in dou- ble-distilled water] and incubated for another 10 min, fol- lowed by brief rinsing in double-distilled water and incubation in developer [5% (v ⁄ v) ProteoSilver Developer I, 0.1% ProteoSilver Developer 2, in double-distilled water] until the desired staining intensity was achieved. Proteo- Silver Stop reagent was used to stop the reaction. Acknowledgements We are grateful to laboratory members for helpful dis- cussions and critical reading of the manuscript, to J. Daniels (Stony Brook University) for the gift of MobA protein, and to S G. Rhee (NIH) for the gift of several critical reagents including antiserum to PrxII. This work was supported by NIH GM60452 and DK64166 to M.A.F. References 1 Du G, Morris AJ, Sciorra VA & Frohman MA (2002) G-protein-coupled receptor regulation of phospholipase D. Methods Enzymol 345, 265–274. 2 McDermott M, Wakelam MJ & Morris AJ (2004) Phos- pholipase D. Biochem Cell Biol 82, 225–253. 3 Regier DS, Greene DG, Sergeant S, Jesaitis AJ & McPhail LC (2000) Phosphorylation of p22phox is mediated by phospholipase D-dependent and -indepen- dent mechanisms. Correlation of NADPH oxidase activ- ity and p22phox phosphorylation. J Biol Chem 275, 28406–28412. 4 Agwu DE, McPhail LC, Sozzani S, Bass DA & McCall CE (1991) Phosphatidic acid as a second messenger in human polymorphonuclear leukocytes. Effects on acti- vation of NADPH oxidase. J Clin Invest 88, 531–539. 5 Palicz A, Foubert TR, Jesaitis AJ, Marodi L & McPhail LC (2001) Phosphatidic acid and diacylglycerol directly activate NADPH oxidase by interacting with enzyme components. J Biol Chem 276, 3090–3097. 6 Dewald B, Payne TG & Baggiolini M (1984) Activation of NADPH oxidase of human neutrophils. Potentiation of chemotactic peptide by a diacylglycerol. Biochem Biophys Res Commun 125, 367–373. 7 Gomez-Cambronero J & Keire P (1998) Phospholipase D: a novel major player in signal transduction. Cell Signal 10, 387–397. 8 Park JW (1996) Phosphatidic acid-induced translocation of cytosolic components in a cell-free system of NADPH oxidase: mechanism of activation and effect of diacylglycerol. Biochem Biophys Res Commun 229, 758–763. 9 Ohtsuka T, Okamura N & Ishibashi S (1986) Involve- ment of protein kinase C in the phosphorylation of 46 kDa proteins which are phosphorylated in parallel with activation of NADPH oxidase in intact guinea-pig poly- morphonuclear leukocytes. Biochim Biophys Acta 888, 332–337. 10 Bey EAet al. (2004) Protein kinase C delta is required for p47phox phosphorylation and translocation in acti- vated human monocytes. J Immunol 173, 5730–5738. 11 Servitja JM, Masgrau R, Pardo R, Sarri E & Picatoste F (2000) Effects of oxidative stress on phospholipid sig- naling in rat cultured astrocytes and brain slices. J Neurochem 75, 788–794. 12 Natarajan V, Taher MM, Roehm B, Parinandi NL, Schmid HH, Kiss Z & Garcia JG (1993) Activation of endothelial cell phospholipase D by hydrogen peroxide and fatty acid hydroperoxide. J Biol Chem 268, 930–937. 13 Min DS, Kim EC & Exton JH (1998) Involvement of tyrosine phosphorylation and protein kinase C in the activation of phospholipase D by H2O2 in Swiss 3T3 fibroblasts. J Biol Chem 273, 29986–29994. 14 Hammond SM, Altshuller YM, Sung TC, Rudge SA, Rose K, Engebrecht J, Morris AJ & Frohman MA (1995) Human ADP-ribosylation factor-activated phos- phatidylcholine-specific phospholipase D defines a new and highly conserved gene family. J Biol Chem 270, 29640–29643. 15 Hammond SM, Jenco JM, Nakashima S, Cadwallader K, Gu Q, Cook S, Nozawa Y, Prestwich GD, Frohman MA & Morris AJ (1997) Characterization of two alter- nately spliced forms of phospholipase D1. Activation of the purified enzymes by phosphatidylinositol 4,5-bisphos- phate, ADP-ribosylation factor, and Rho family mono- meric GTP-binding proteins and protein kinase C-alpha. J Biol Chem 272, 3860–3868. 16 Du G, Altshuller YM, Kim Y, Han JM, Ryu SH, Morris AJ & Frohman MA (2000) Dual requirement for rho and protein kinase C in direct activation of phospholipase D1 through G protein-coupled receptor signaling. Mol Biol Cell 11, 4359–4368. 17 Zhang Y, Altshuller YM, Hammond SM, Hayes F, Morris AJ & Frohman MA (1999) Loss of receptor reg- ulation by a phospholipase D1 mutant unresponsive to protein kinase C. EMBO J 18, 6339–6348. 18 Yamazaki Met al. (1999) Interaction of the small G protein RhoA with the C terminus of human phospho- lipase D1. J Biol Chem 274, 6035–6038. 19 Kusner DJ, Barton JA, Wen KK, Wang X, Rubenstein PA & Iyer SS (2002) Regulation of phospholipase D activity by actin. Actin exerts bidirectional modulation of mammalian phospholipase D activity in a polymeri- Regulation of PLD1 by interaction with PrxII N. Xiao et al. 3936 FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS zation-dependent, isoform-specific manner. J Biol Chem 277, 50683–50692. 20 Oishi K, Takahashi M, Mukai H, Banno Y, Nakashima S, Kanaho Y, Nozawa Y & Ono Y (2001) PKN regu- lates phospholipase D1 through direct interaction. J Biol Chem 276, 18096–18101. 21 Ganley IG, Walker SJ, Manifava M, Li D, Brown HA & Ktistakis NT (2001) Interaction of phospholipase D1 with a casein-kinase-2-like serine kinase. Biochem J 354, 369–378. 22 Lee C, Kim SR, Chung JK, Frohman MA, Kilimann MW & Rhee SG (2000) Inhibition of phospholipase D by amphiphysins. J Biol Chem 275, 18751–18758. 23 Du G, Altshuller YM, Vitale N, Huang P, Chasserot- Golaz S, Morris AJ, Bader MF & Frohman MA (2003) Regulation of phospholipase D1 subcellular cycling through coordination of multiple membrane association motifs. J Cell Biol 162, 305–315. 24 Jenkins GM & Frohman MA (2005) Phospholipase D: a lipid centric review. Cell Mol Life Sci, in press. 25 Lim MJ, Chae HZ & Rhee SG, Yu DY, Lee KK & Yeom YI (1998) The type II peroxiredoxin gene family of the mouse: molecular structure, expression and evolu- tion. Gene 216, 197–205. 26 Wood ZA, Poole LB & Karplus PA (2003) Peroxire- doxin evolution and the regulation of hydrogen per- oxide signaling. Science 300, 650–653. 27 Neumann CA, Krause DS, Carman CV, Das S, Dubey DP, Abraham JL, Bronson RT, Fujiwara Y, Orkin SH & Van Etten RA (2003) Essential role for the peroxi- redoxin Prdx1 in erythrocyte antioxidant defence and tumour suppression. Nature 424, 561–565. 28 Kim SH, Fountoulakis M, Cairns N & Lubec G (2001) Protein levels of human peroxiredoxin subtypes in brains of patients with Alzheimer’s disease and Down syndrome. J Neural Transm Suppl 223–235. 29 Noh DY, Ahn SJ, Lee RA, Kim SW, Park IA & Chae HZ (2001) Overexpression of peroxiredoxin in human breast cancer. Anticancer Res 21, 2085–2090. 30 McLuskey K, Harrison JA, Schuttelkopf AW, Boxer DH & Hunter WN (2003) Insight into the role of Escherichia coli MobB in molybdenum cofactor bio- synthesis based on the high resolution crystal structure. J Biol Chem 278, 23706–23713. 31 Guse A, Stevenson CE, Kuper J, Buchanan G, Schwarz G, Giordano G, Magalon A, Mendell RR, Lawson DM & Palmer T (2003) Biochemical and structural analysis of the molybdenum cofactor biosynthesis protein MobA. J Biol Chem 278, 25302–25307. 32 Sambrook J & Russell DW (2001) Molecular Cloning: A Laboratory Manual. 3rd Edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 33 Rhee SG, Chang TS, Bae YS, Lee SR & Kang SW (2003) Cellular regulation by hydrogen peroxide. JAm Soc Nephrol 14, S211–S215. 34 Bokoch GM & Diebold BA (2002) Current molecular models for NADPH oxidase regulation by Rac GTPase. Blood 100, 2692–2696. 35 Bae YS, Kang SW, Seo MS, Baines IC, Tekle E, Chock PB & Rhee SG (1997) Epidermal growth factor (EGF)- induced generation of hydrogen peroxide. Role in EGF receptor-mediated tyrosine phosphorylation. J Biol Chem 272, 217–221. 36 Leavey PJ, Gonzalez-Aller C, Thurman G, Kleinberg M, Rinckel L, Ambruso DW, Freeman S, Kuypers SA & Ambruso DR (2002) A 29-kDa protein associated with p67phox expresses both peroxiredoxin and phos- pholipase A2 activity and enhances superoxide anion production by a cell-free system of NADPH oxidase activity. J Biol Chem 277, 45181–45187. 37 Colley WC, Sung TC, Roll R, Jenco J, Hammond SM, Altshuller Y, Bar-Sagi D, Morris AJ & Frohman MA (1997) Phospholipase D2, a distinct phospholipase D isoform with novel regulatory properties that provokes cytoskeletal reorganization. Curr Biol 7, 191–201. 38 Du G, Huang P, Liang BT & Frohman MA (2004) Phospholipase D2 localizes to the plasma membrane and regulates angiotensin II receptor endocytosis. Mol Biol Cell 15, 1024–1030. 39 Shaw SS, Schmidt AM, Banes AK, Wang X, Stern DM & Marrero MB (2003) S100B-RAGE-mediated augmen- tation of angiotensin II-induced activation of JAK2 in vascular smooth muscle cells is dependent on PLD2. Diabetes 52, 2381–2388. 40 Melendez AJ, Bruetschy L, Floto RA, Harnett MM & Allen JM (2001) Functional coupling of FcgammaRI to nicotinamide adenine dinucleotide phosphate (reduced form) oxidative burst and immune complex trafficking requires the activation of phospholipase D1. Blood 98, 3421–3428. 41 Morris AJ, Frohman MA & Engebrecht J (1997) Mea- surement of phospholipase D activity. Anal Biochem 252, 1–9. 42 Chang KN, Lee TC, Tam MF, Chen YC, Lee LW, Lee SY, Lin PJ & Huang RN (2003) Identification of galec- tin I and thioredoxin peroxidase II as two arsenic-bind- ing proteins in Chinese hamster ovary cells. Biochem J 371, 495–503. N. Xiao et al. Regulation of PLD1 by interaction with PrxII FEBS Journal 272 (2005) 3929–3937 ª 2005 FEBS 3937 . Peroxiredoxin II functions as a signal terminator for H 2 O 2 -activated phospholipase D1 Nianzhou Xiao, Guangwei Du and Michael A. Frohman Department. Oishi K, Takahashi M, Mukai H, Banno Y, Nakashima S, Kanaho Y, Nozawa Y & Ono Y (2001) PKN regu- lates phospholipase D1 through direct interaction. J