Investigation of the contribution of histidine 119 to the conduction of protons through human Nox2 Tosti J. Mankelow*, X. Wen Hu, Kate Adams and Lydia M. Henderson Department of Biochemistry, School of Medical Sciences, University of Bristol, UK The conduction of protons through human Nox2 has pre- viously been shown to be dependent upon His115. Align- ment of sequences for both animal and plant Nox proteins indicated that histidines 115 and 119 are both highly con- served, while His111 was co nserved among animal homo- logues of Nox1–4. T o investigate t he possible r ole that these histidine residues m ight play in the conduction of proton s through Nox2, we have introduced both paired and single mutations into these histidine residues. Each construct was used to generate a CHO cell line in which the expression of the m utated Nox2 was assessed. Nox2 was e xpressed in e ach of the C HO cell lines generated, however, the level of expression of H111/115L in CHO cells was lower and that of H111L very much reduced, compared to that of wild-type Nox2. The arachidonic acid activated proton flux was absent in the CHO cell lines expressing the mutations of H111/115L, H111/119L or H115/119L, compared to that observed for wild-type Nox2. Similarly only a small efflux of protons was observed from CHO cells expressing either H119L or H111L. In all c ases the expected proton flux was elicited through the addition of the protonophore, carbonyl cyanide m-chlorophenylhydrazone. Conclusions regarding the role of His111 in the conduction of protons cannot be drawn due to the reduced expression. We can, however, conclude th at His119, in addition to His115, is required for the conduction of protons through Nox2. His119 has been identified as a highly conserved residue for w hich no fu nction has p reviously been proposed. Keywords:gp91 phox ; membranes; NADPH oxidase; proton transport. The phagocytic white blood cells engulf, kill and digest microorganisms within the body. These cells possess a membrane bound enzyme, NADPH oxidase, which gener- ates superoxide (O Æ À 2 ) and contributes to the killing of the engulfed microorganisms within the phagosome. The NADPH oxidase is composed of two membrane subunits, gp91 phox and p22 phox , a nd three cytosolic proteins, p67 phox , p47 phox and p40 phox . The predicted binding sites for the enzyme cofactors, FAD and the two b-type heme, are located within gp91 phox [1]. A number o f proteins with amino acid sequ ences similar to that of the human gp91 phox have been described in human [2–4], mouse, rat, Caenorhabditis elegans, Dictyo- stelium,rice[5],tomatoandArabidopsis [6,7] g enomes. The identification of multiple sequences with similarity to gp91 phox within the human genome led to the need to rename and number the family of proteins from p hox (phagocyte oxidase) to Nox (NADPH oxidase-like). Of the Nox family members, Nox2 is the originally described human phagocyte protein gp91 phox . In addition, genes encoding dual oxidases (Duox), which contain both a peroxidase and N ox have been identified in a number of different genomes [8]. The generation of superoxide by the phagocytic NADPH oxidase is associated with an e fflux of protons [9]. The expression of wild-type Nox2 in CHO cells has previously been shown to b e associated with the presence o f a characterized arachidonic a cid activated proton cond uction pathway [10,11]. T he membrane topology of gp91 phox is predicted to consist of 4–6 transmembrane domains located within the N-terminal half of the protein followed by a hydrophilic, cytosolic domain [12]. The third predicted transmembrane domain of Nox2 contains a sequence of three regularly spaced histidine residues, HSAIHT- IAH(111–119). The mut ation of His115 to Leu has previously been shown to greatly reduce p roton conduction, Correspondence to L. M. Henderson, Department of Biochemistry, School of Medical Sciences, University of Bristol, University Walk, Bristol, BS8 1TD, UK. Fax: +44 117 9288274, Tel.: +44 117 9289046, E-mail: L.M.Henderson@bris.ac.uk Abbreviations: BCECF-AM, 2¢,7¢-bis-(2-carboxyethyl)-5-(and -6-)- carboxyfluorescein, acetoxymethyl ester; cPLA 2 , cytosolic phospho- lipase A 2 ; CHO91, CHO cell line resulting from the transfection of CHO cells with full-length wild-type human Nox2; CHO91H111/ 115L, CHO/Nox2 cell line with His111 and H is115 mutated to Leu; CHO91H115/119L, CHO/Nox2 cell line with His115 and His119 mutated to Leu; CHO91H111/119L, CHO/Nox2 cell line w ith His111 and His119 mutated to Leu; CHO91H119L, C HO/Nox2 cell line with single mutation His119 to Leu; CHO91H111L, CHO/Nox2 cell line with single mutation His111 to Leu. *Present address: Bristol Institute for Transfusion Sciences, Southmead Hospital, Bristol, UK. Present address : Department of Biochemistry, University of Oxford, UK. Note: T he data prese nted in t his paper have previously been pu blished as an abstract for the 56th Harden conference on ÔBiological electron and proton transferÕ Plymouth, UK, 2003. (Received 1 6 July 2004, accepted 20 August 2004) Eur. J. Biochem. 271, 4026–4033 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04340.x whether introduced into the full-length Nox2 [13] or only the N-terminal membrane spanning domains [14]. Similarly, proton conduction was greatly reduced if all three hist idine residues were mutated to leucine [13,14]. Therefore the ability of Nox2 to act as proton conduction pathway has previously b een demonstrated to be dependent upon His115. An alignment of the amino acid sequences for Nox and Duox proteins available in the genome databases demonstrated that His115 and His119 in human Nox2 are both conserved through out the animal and plant sequences of Nox and Duox proteins (Fig. 1A). Only the sequence of Arabidopsis RbohE appeared to lack the His119 residue. However, t he introduction of a four amino acid gap in the s equence improved the sequence alignment and suggested that both His115 and 119 are also conserved in RbohE (Fig. 1B). His111 is conserved in Nox2 sequences from different animal species (Fig. 1). Histidines 101, 115, 209 and 222 have been proposed to function as the 5th and 6th coordinate ligands for the two b-type hemes due to similarities in sequence position and spacing of the histidine residues t o those of yeast ferric reductase [15]. However, no function has previ- ously been proposed for a conserved histidine residue at position 119. The contribution of His119 and/or His111 to the conduction of protons through Nox2 has not previously been investigated. To determine whether the mechanism for the conduction of protons through Nox2 requires His111 and/or His119, we have established a number of CHO cells lines which express human Nox2 containing mutations to each of the possible pair of histidine residues (111 + 115, 115 + 119 and 111 + 119) o r to individual histidine residues (111 or 119). The CHO cell lines expressing each of these mutations failed to show an arachidonic acid activated proton flux. However the mutation of His111 to Leu resulted in a reduced level of expression of full-length Nox2. Therefore, we conclude that the conduction of protons through Nox2 requires His119 as well as the previously demonstrated role for H is115. The possible role for H is111 in proton conduction through Nox2 cannot be established from the current data. Fig. 1. Amino acid seque nce a lignment o f Nox and Duox proteins with the third predicted transmembrane domain of human Nox2. Th e amino acid sequences for 40 Nox and Duox proteins were obtained from the P ubMed nucleotide and protein databases. (A) The alignment of the amino acids from each Nox and Duox protein with sequence similarity to that of residues 100–131 in human Nox2. The arrow indicates the conserved His residue (119) in addition to t he previously reported conserved histidine at position 115. T he sequences shown include the Arabidopsis (Rboh), rice (RbohAOsp), tomato (Rboh- Tomato) and induced in tomato by w hitefly (RbohTomIn), potato (Strboh) and Dictyos- telium (DictoNox2). The pig and rabbit Nox1 are incomplete partial sequences lacking the N-terminus. There are s equences for human Nox5 with and without the extended N-terminal containing the EF-hands. The residues numbers for both are shown. (B) The introduction of a gap of four amino acids into the RbohE sequence (residues 473–503) is necessary to align the histidine re sidues with those of the other Arabidopsis,plantand human Nox2. Ó FEBS 2004 Histidine 119 and conduction protons through Nox2 (Eur. J. Biochem. 271) 4027 Materials and methods The composition o f the salt solutions used are as described previously [11,13]. Sequence alignment The PubMed protein and nucleotide databases contain sequence data for a number o f Nox and Duox proteins from human, mouse, rat, cow, bison, pig, dolphin, rabbit, guinea- pig, Japanese puffer fish, Dictyostelium, Arabidopsis, rice, tomato, potato and C. elegans. Alignment o f these sequences in species groups identified a region of sequence similarity including and in a ddition to the previously described conserved histidine at position 115 in human Nox2. The amino acids with sequence homology to residues 100–131 of human Nox2 were identified in each Nox and Duox sequences and aligned using GENEDOC and DNASTAR .The histidine residues 115 and 119 (human Nox2) are conserved throughout the 40 sequences currently available. Amino acids 100–129 of human Nox2 are predicted to form the third transmembrane domain. Mutagenesis of His111, 115 and/or 119 to Leu in full-length Nox2 cDNAs encoding full-length Nox2 containing each of the following mutations: His111 & 115 to Leu (H111/115L); His115 & 119 to Leu (H115/119L); His111 & 119 to Leu (H111/119L); His111 to Leu (H111L) and His119 to Leu (H119L) wer e generated from two overlapping PCR products as described p reviously [14], using the specific oligonucleotide primers contained in Table 1. The two fragments were annealed and joined by extension from the overlap region before final P CR amplification using ÔFull- length Nox2 fw dÕ and ÔFull-length Nox2 revÕ oligonucleotide primers (Table 1). These PCR products were inserted as a HindIII/BamHI fragment into the multiple cloning site of the p lasmid pMEP4 (Invitrogen). Oligonucleotides were synthesized by MWG-Biotech AG. Construction and maintenance of transfected CHO cell lines Stable CHO cell lines expressing each of the mutant forms of full-length human Nox2 (CHO91) were established following transfection by electroporation (230 V, 975 lF) of the constructs described above, followed by s election with 100 lgÆmL )1 hygromycin b as described previously [11,14]. The following stable ce ll lines were established: CHO91H111/115L (Nox2 with histidines 111 and 115 mutated to leucine); CHO91H115/119L (Nox2 with histi- dines 115 and 119 mutated to leucine); CHO91H111/119L (Nox2 with histidines 111 and 119 mutated to leucine); CHO91H111L (Nox2 with histidine 111 only mutated to leucine) and CHO91H119L (Nox2 w ith histidine 119 only mutatedtoleucine). All CHO cell lines were maintained in Ham’s F-12 nutrient mixtu re with GlutaMAX-I, 10% (v/v) fetal bovine serum, 50 UÆmL )1 penicillin and 50 lgÆL )1 streptomycin. The c ells were divided 1 : 2 once a week following trypsinization [11]. Expression of the mutant Nox2 proteins Verification of the expression of each of the mutant Nox2 proteins and determination of their cellular l ocalization were assessed by i mmunostaining a nd confocal microscopy as previously described [11,13]. The immunocytochemistry was performed with an antipeptide [PRGVHFIFN KENF(558–570)] polyclonal antibody [11,16] upo n each of the CHO91 mutant cell lines. Expression of Nox2 was driven from the inducible metallothionein promoter in pMEP4 through the preincubation of the cells with 10 l M Cd 2+ for 16–24 h. The CHO cell lines were grown on r ound cover slips for 40–48 h prior to fixation in 4% (v/v) formaldehyde for 10 min and permeabilization with 0.2% (v/v) Triton in phosphate buffered s aline (2 min). Binding of the antibody raised against Nox2 was detected with a fluorescein isothiocyanate-labelled anti-rabbit Ig. An image of an optical plane through the cells was collected with a Table 1. Sequences of the oligonucleotide primers used for the introduction of mutations in full-length Nox2. ÔFull-length Nox2 fwdÕ contains a HindIII restriction site (underlined) for the directionalinsertionintopMEP4andcoverstheN-terminal initiating Met codon (bold) of Nox2. ÔFull-length Nox2 revÕ contains a BamHI restriction site (underlined) for the d irectional insertion into th e expression vector, pMEP4 and covers the stop codon of the C-terminus of Nox2 (bo ld). Full-length No x2 mu tants we re genera ted fro m two ove rlapping PCR products (nucleotide 1–352 and 336–1713) which we re annealed and joined by extension before final PCR amplification using ÔFull-length Nox2 fwdÕ and ÔFull-length Nox2 revÕ.Bold underlined single nucleotides are the s ubstitutions in the m utant and groups of three n ucleotid es underlined show the positions of the h istidine residues. Primer name Primer sequence Mutants constructed Full-length Nox2 fwd CTCAGCT AAGCTTGCCACCATGGGGAAC All full length constructs Full-length Nox2 rev AGCTGA GGATCCCTACCCACGTACAATTCG All full length constructs His111Leu rev CAATGGTGTGAATCGCAG AGAGAAGTG His111/119Leu and His111Leu His115Leu fwd TGCGATTC TCACCATTGCACATCTATTT His111/115Leu His115Leu rev CAATGGTG AGAATCGCAGAGTGAAGTGC His115/119Leu His119Leu fwd TGCGATTCACACCATTGCA CTTCTATT His111/119Leu and His119Leu His111/115wt rev CAATGGT GTGAATCGCAGTGAGAAGTG His119Leu His115/119wt fwd TGCGATT CACACCATTGCACATCTATT His111Leu His115/119Leu fwd TGCGATTC TCACCATTGCACTTCTATTT His115/119Leu His111/115Leu rev CAATGGTG AGAATCGCAGAGAGAAGTGC His111/115Leu 4028 T. J. Mankelow et al.(Eur. J. Biochem. 271) Ó FEBS 2004 Bio-Rad MRC 600 inverted confocal microscope and presented as a Kalman average of five successively collected scans, with excitation at 488 nm. The expression levels of all the mutants were compared to that observed for wild-type Nox2 in CHO91 cells (positive control). N onspecific binding of the antibodies was assessed in nontransfected CHO cells ± 10 l M Cd 2+ (negative co ntrol). Transmembrane proton flux in mutant CHO91 cell lines The conduction of protons through the arachidonate activated, NADPH oxidase associated H + channel was assessed as change in pH i with 2¢,7¢-bis-(2-carbox yethyl)-5- (and -6-)-carboxyfluorescein, acetoxymethyl e ster (BCECF- AM) for all m utant CHO91 cell lines, a s described previously [11,13,14]. The pH of the external solution was rapidly altered by the addition of HEPES (pH 6.6) or Tris (pH 8.3). The addition of the potassium ionophore, valinomycin, equilibrates K + ions across the plasma membrane, d epolarizing t he membrane potential of cells in a high K + solution and hyperpolarizing the membrane potential if cells are in a Na + solution (low K + ). Therefore, the addition of Tris and v alinomycin to cells in a high K + medium establishes a strong electrochemical gradient for the efflux of protons. A gradient favouring the influx of protons is established following the addition of HEPES and valinomyc in to cells in a Na + solution (low K + ). The predicted third transmembrane helix for Nox2 A predicted structure for amino acids 98–129 of human Nox2 was modelled by C. Dempsey (Department of Biochemistry, University of Bristol, UK) based on the assumption that the a mino acids of t he third predicted transmembrane domain adopt an a-helix conformation. Results The conduction of protons through Nox2 was previously observed to be greatly redu ced in CHO cells expressing Nox2 in which His115 h ad been mutated to L eu and when all three histidines, 111, 115 and 119, were mutated to Leu [13,14]. The alignment of Nox protein sequences indicates that His119 is highly conserved and His111 is conserved in Nox2 proteins from different animal species. However, the contribution of histidines 111 and 119 to the conduction of protons through Nox2 has not previously been invest igated despite their p ossible location on the same face of an a-helix as His115. To determine th e contribution of histidines 111 and 119 to the conduction of protons through human Nox2, we have established CHO cells that express Nox2, into which mutations to either paired or single histidine residues within the third transmembrane domain have been intro- duced. The exp ression of the r esulting protein and condu c- tion of protons in re sponse to arachidonic acid was assessed for each of the cell lines. Expression of Nox2 histidine mutations in CHO cells Table 1 contains the sequences o f the o ligonucleotide PCR primers used to introduce the mutations of His fi Leu for the three histidine residues of interest within the third transmembrane domain of Nox2. Each of the three possible combinations of pairs of histidine residues were mutated to leucine ( 111/115, 115/119 and 111/119) and used to establish stable CHO cell lines, CHO91H111/115L, C HO91H115/ 119L and C HO91H111/119L. Transcription and t ransla- tion of the Nox2 gene is under the control of an inducible metalothionine promoter (10 l M Cd 2+ for 16 h). Immuno- staining, using an antibody raised against the Nox2 C-terminal, was used to assess the inducible expression of Nox2proteinineachofthecellslines. The immunostaining was observed to be greater for CHO91 (Fig. 2A), CHO91H115/119L (Fig. 2C), CHO91H111/115L (Fig. 2E) and C HO91H111/119L (Fig. 2G) when the cells had be en g rown in the presence of 10 l M Cd 2+ compared to the same CHO cell lines grown in the absence of Cd 2+ (Figs 2B,D,F,H, respectively). Non- transfected CHO cells showed low l evels of immunostaining whether grown in the presence (Fig. 2I) or absence (Fig. 2J) of Cd 2+ . Therefore, each of the CHO cell lines show an induced expression of protein, which was dependent upon transfection with Nox2 cDNA. The level of expression observed in CHO91H111/115L cells, in which histidines 111 and 115 were both mutated to leucines (Fig. 2E) was lower than that observed for CHO91 with the wild-type Nox2 (Fig. 2A), while the expression of Nox2 in each of the remaining m utant C HO ce ll lines (Fig. 2 C,G) was compar- able to that observed for CHO91 cells (Fig. 2A) as observed previously [11,13,14]. Arachidonic acid activated proton conduction The arachidonic acid activated proton conduction pathway has previously been reported to be greatly reduced by the mutation of His115 to Leu compared to that observed for wild-type Nox2 [13,14]. To assess the contribution of histidines 119 and 111 to the conduction of p rotons th rough Nox2, the ability of CHO91H111/115L, CHO91H115/119L and CHO91H111/119L cells to conduct protons in response to stimulation by arachidonic acid was assessed. Electro- chemical gradients for either proton influx (Fig. 3A–C) or proton efflux (Fig. 3D,E) were imposed upon the cells prior to the addition of arachidonic acid. CHO91H115/119L (Fig. 3C) cells showed little change in pH i in response to arachidonic acid compared with the response of CHO91 ce lls (Fig. 3A) expressing human Nox2. A similar low conduction of protons following the addition of sodium arachidonate was observed for CHO91H111/115L (Fig. 3E). The observed very low conduction of protons for both CHO91H111/119L and CHO9H115/119L cells is as would b e pred icted and may be the result s olely of the mutation of His115 to Leu in both of these cell lines. However, CHO91H111/119L cells failed to exhibit an arachidonic acid activated conduction of protons whether assessed as a proton influx (Fig. 3B) or as a proton efflux (Fig. 3D). The subsequent addition of the protonophore, carbonyl cyanide m-chlorophenylhydrazone, resulted i n the observed pH i change expected for the established proton electrochemical gradient in all cases (Fig. 3A–E). CHO91, CHO91H111/115L, C HO91H115/119L and CHO91H111/ 119L all demonstrated a similar pH i change in response to a sodium acetate i mposed acid pulse (Fig. 3F) indicating that Ó FEBS 2004 Histidine 119 and conduction protons through Nox2 (Eur. J. Biochem. 271) 4029 Fig. 2. Expression of wild-type and double histidine mutations o f Nox2 in CHO cell lines. The expression of full-length Nox2 was determined by immunocytochemistry as described in the Materials and methods. The inducible e xpression of Nox2 was determined for CHO91 (A and B), CHO91H115/119L (C and D), C HO91H111/115L (E and F), CHO91H111/119L (G and H) and for nontransfected CHO cells (I and J). The CHO cells were incubated in the presence of 10 l M Cd 2+ for 16 h prior to fixation and staining (A, C, E, G and I). The images are Kalman averages of five successively collected scans. The intensity o f th e fl uorescen ce is represented by a pseudocolour scale wh ere r ed is highest and blu e is lowest. 4030 T. J. Mankelow et al.(Eur. J. Biochem. 271) Ó FEBS 2004 the cytoplasm in e ach of these cell lines has a similar buffering capacity, with that of CHO91H111/119L a little lower. The calibration of BCECF-AM response to pH i was similar in all four cell lines (Fig. 3G). We therefore conclude that the absence of an arachidonate activated proton conduction in CHO91H111/119L cells suggests t hat His111 and/or His119 are r equired, in addition to His115, for the conduction of protons through Nox2. Mutation of individual histidine residues in human Nox2 To investigate the individual roles of His111 and His119, in the conduction of protons through Nox2, two CHO cells lines were constructed, CHO91H111L and CHO91H119L, in which the single His fi Leu mutations had been intro- duced. The expression of Nox2 in which the histidine at position 119 had been mutated to leucine was inducible in CHO91H119L cells (Fig. 4A) and the protein was expressed at a level comparable with that observed for wild-type Nox2 (Fig. 2A). However, the expression of Nox2 in CHO91H111L cells, in which the histidine at position 111 had been mutated to leucine, although enhanced by activation of the inducible promoter, was observed t o be significantly lower than that of w ild-type Nox2 (Figs 4B and 2A). CHO91H119L cells in the presence of a transmembrane gradient for the efflux of protons exhibited a very small conduction of protons following the addition of arachido- nate (Fig. 4E). The efflux of protons was similar to that observed for CHO91H111/119L cells (Fig. 3D). The efflux of protons was only observed following the subsequent addition of carbonyl cyanide m-chlorophenylhydrazone (Fig. 4E). The reduction in proton efflux observed for CHO91H119L cells is not due to a reduction in the level of protein expression as a result of the introduction of this mutation into Nox2. Therefore, we can conclude that the reduced proton efflux is as a c onsequence of the mutation of His119 to Leu, and therefore that His119 is required for the conduction of protons through Nox2. We note that proton efflux was also not observed following the addition of arachidonate to CHO91H111L cells suspended in a K + medium (not shown). However, due to the reduced level of expression of H111L Nox2 in CHO compared with wild-type and the other mutations described here and reported p reviously [13,14], no conclu- sions can be draw regarding the role o f His111 in the conduction of protons through Nox2. Therefore, we can conclude that the conduction of protons through human Nox2 requires both His119 and His115. Fig. 3. Arachidonic acid stimulated proton flux. The CHO cell lines were inc ubated with 10 l M Cd 2+ for 16 h prior to the assay. Cells from each of the CHO cell lines were harvested and lo aded with BCECF-AM as described in t he Materials and me thods. Following the addition of t he potassium ionophore, valinomycin (2.7 l M ), t he transmembrane potassium gradient dictat es the membr ane potential. T he addition of 5 m M HEPES rapidly decreases the external pH to 6.6, or to pH 8.3 following th e addition of 5 m M Tris. This established an electrochemical gradient for the influx (A–C) or effl ux (D and E ) of protons. T he influx of p rotons was assessed for control C HO91 cells expressing wild-t ype gp91 phox (A), CHO91H111/119L (B) and CHO91H115/119L (C), all suspended in the Na + medium. In addition the efflux of protons w as assessed in both CHO91H111/119L (D) and CHO91H111/115L (E) with c ells suspended i n a K + medium. (A–E) The opening of the proton conduction pathway and the t ransport of protons was monitored as an alteration in i nternal pH, following the addition of 10 l M arachidonic a cid (AA). The maximum change in pH i was determined following the addition of the protonophore, carbonyl cyanide m-chlorophenylhydrazone (CCCP; 66 l M ) where indictated. The requirement for a s econd addition of CCCP in (E), in order to elicit a full pHi change, probably results from a l arger number of cells. The buffering capacity was asse ssed fo llowing the addition of 26 m M NaAc (F) to CHO91H115/119L, CHO91H111/119L, CHO91, CHO91H111/115L (left to right) and the response of the BCECF-AM to pH was calibrated (G) in CHO91 (j), CHO91H111/119L (m), CHO91H115/119L (d)and CHO91H111/115L ( r). Ó FEBS 2004 Histidine 119 and conduction protons through Nox2 (Eur. J. Biochem. 271) 4031 Discussion It has previously been shown that His115 is required for the conduction of proton s through human Nox2, in either the full-length protein [13] or the N-terminal membrane domain alone [14]. The third transmembrane domain of Nox2 contains three histidine residues separated from each other by three amino acid residues [14]. In this paper we have generated CHO cell lines expressing human Nox2 into which double and single mutations to histidines 111, 115 and 119 had been introduced. Mutation of His119 to Leu severely reduced the arachidonic acid activated conduction of protons. We therefore conclude that His119, as well as His115, is required for the conduction of protons through human Nox2. The role of His111 in proton cond uction cannot be firmly established f rom the current data. T he alignment of sequences for Nox proteins indicates that His119 is a highly conserved residue and yet no role for the His119 has previously been proposed. The M2 protein is an integral membrane, coat protein of the influenza A virus. It has previously been reported to function as a proton channel [17–21]. The virus enters the cells via endocytosis. The M2 protein proton channel is opened by the acidic pH environment of the endosomes, resulting in a fall in pH of the v iral particle, l eading to dissociation of the viral coat and release of the viral genome [17,22]. Amantidine, an anti-influenza drug, acts as an inhibitor of the influenza A M2 proton channel and hence prevents the uncoating of the viral particle and release of its genome into the cell [18,23]. Therefore, the influx of protons through the M2 protein is important in the life cycle of the virus. The M2 protein has a histidine residue, His37, located in the transmembrane domain, which is required for th e open ing of the channel by acidic pH and for the influx of protons through the channel [22,24]. A possible cation–p interaction between the prot- onated imidazole ring of His37 and the indole r ing o f Trp41 has been identified. It has recently been proposed that the Trp41 side chain blocks t he pore of the channel and that the in teraction between protonated His37 and Trp41 pulls the Trp41 side chain out of the pore of the channel, thereby opening the M2 proton channel [22,25]. A tryptophan o ccurs at position 125 of human Nox2, located within the transmembrane domain that also contains His111, 115, and 119. The model of this region as an a-h elix places Trp125 two turns of the helix above His119 and not on the same face o f the helix. Therefore, the distance and the orientation of His119 to Trp125 i s too great to permit a cation–p interaction similar to that described for the pH stimulated gating of the M2 protein by the Trp41–His37 interaction [22,25]. A role f or cytosolic phospholipase A 2 (cPLA 2 )inthe opening of the NADPH oxidase associated proton channel has previously been demonstrated through the development and use of a cPLA 2 deficient PLB-985 cell line (PLB-D cells) [26–28]. Gating o f the proton channel formed by either full- length human Nox2 [27] or just the N-terminal membrane domain alone [28] have both been reported to be deficient in PLB-D cells. The requirement for cPLA 2 for both the activation of superoxide generation [26] and the gating of the proton channel [27,28] can be over come by the addition of arachidonate acid. The site of interaction with and the mechanism by which arachidonate results in the activation of both the NADPH oxidase and the gating of the proton channel is at present unknown. In this paper we have demonstrated a requirement for His119 in the conduction of protons through human Nox2. This, combined with the previously described role for His115 in proton conduction, suggests a mechanism that involves passage of the p roton from one histidine residue to the next. Acknowledgements The au thors wish to thank Dr Chris Dempsey, Department of Biochemistry for the modelling of the transmembrane domain as an a-helix. The work was supported by Grant No. HO604 from the Arthritis Research Campaign and a MRC PhD studentship to T.J.M. E AB DC Fig. 4. Expression of the single mutations, H119L and H111L, in CHO cells and determination of proton e fflux from CHO91H119L cells. The expression of mutant Nox2 in CHO91H119L (A and B) and CHO91H111L cells (C and D) was assessed as described i n the Materials a nd methods. The CHO91H119L (A) and CHO91H111L (C) were g rown in the presence (A,C) or absence (B,D) of 10 l M Cd 2+ for 16 h prior to the assay. The efflux of protons from CHO91H119L cells preincubated with 10 l M Cd 2+ (E) was determined as described in the Materials and methods. The CHO91H119L cells were resuspended in a K + medium and additions of 2.7 l M valinomycin, 5 m M Tris, 10 l M arachidonate (AA) and 6 6 l M carbonyl c yanide m-chlorophe- nylhydrazone (CCCP) were m ade where indicated i n the figure (E). 4032 T. J. Mankelow et al.(Eur. J. Biochem. 271) Ó FEBS 2004 References 1. Segal, A.W., West, I., Wient jes, F., Nugent, J .H.A., Chavan, A.J., Haley, B., Garcia, R.C., Rosen, H. & Scrace, G. (1992) Cyto- chrome b-245 is a flavocytochrome containing FAD and the NADPH-binding s ite of the microbicidal oxidase of phagocytes. Biochem. J. 284, 781–788. 2. Lambeth, J.D., Cheng, G., Arnold, R.S. & Edens, W.A. (2000) Novel hom ologs of gp91 phox . Trends in Bioc hem. Sci 25, 459–461. 3. Banfi, B., Molnar, G., Maturana, A., Steger, K., Hegedus, B., Demaurex, N. & Krause, K.H. (2001) A Ca 2+ -activated NADPH oxidase in testis, spleen, and lymph nodes. J. Biol. Chem. 276, 37594–37601. 4. Cheng, G., C ao, Z., Xu, X., Meir, E.G. & Lambeth, J.D. (2001) Homologs of gp91 phox : cloning and tissue expression of Nox3, Nox4, and Nox5. Gene 269, 131–140. 5. Groom, Q.J., Torres, M.A., Fordham- Skelton, A.P., Ha mmond- Kosack, K.E., Robins on, N.J. & Jones, J.D. (1996) rbohA,arice homologue of the mammalian gp91 phox respiratory burst oxidase gene. Plant J. 10, 515–522. 6. Keller, T., Damude, H.G., Werner,D.,Doerner,P.,Dixon,R.A. & L amb, C. (1998) A plant homolog of the neutrophil NADPH oxidase gp91 phox subunit ge ne encodes a plasma membrane pro- tein with Ca 2+ binding motifs. Plant Cell 10, 255–266. 7. Torres, M.A., Onouchi, H., Hamada, S., Machida, C., Ham- mond-Kocak, K.E. & Jones, J.D.G. (1998) Six Arabidop sis thali- ana homologues of the human respiratory burst oxidase (gp91 phox ). Plant J. 14, 365–370. 8. Dupuy, C., Pomerance, M., Ohayon, R., Noel-Hudson , M.S., Deme, D., Chaaraoui, M., Franc on, J . & Virion, A. (2000) Thyroid oxidase (THOX2) gene expression in the rat thyroid cell line FRTL-5. Biochem. Biophys. Res. Commun. 277, 287–292. 9. Henderson, L.M., Chappell, J.B. & Jones, O.T. (1987) The superoxide-generating NADPH oxidase of h uman neutrophils is electrogenic and a ssociated w ith a n H + channel. Biochem. J. 246, 325–329. 10. Henderson, L.M. & Chappell, J.B. (1992) The NADPH-oxidase- associated H + channel is opened by arachidonate. Biochem. J. 283, 171–175. 11. Henderson, L.M., Banting, G. & Chappell, J.B. (1995) The ara- chidonate-activable, NADPH oxidase-associated H + channel. Evidence that gp91 phox functions as an essential part of t he channel. J. Biol. Chem. 270, 5909–5916. 12. Hend erson, L.M., Thomas, S., Banting, G. & Chap pell, J.B. (1997) The arachidonate-activatable, NADPH oxidase-associated H + channel is contained within the multi-membrane-spanning N-terminal region of gp 91 phox . Biochem. J. 325, 701–705. 13. Mankelow, T.J. & Henderson, L. M. (2003) Proton conduction through full-length gp91 phox requires histidine 115. Protoplasma 221, 101–108. 14. Henderson, L.M. (1 998) NADPH Oxidase associated H + chan- nel: role for histidines identified by mutagenesis. J. Biol. Chem. 273, 33216–33223. 15. Finegold, A.A., Shat well, K.P., Segal, A.W., Klausner, R.D. & Dancis, A. (1996) Intramembrane bis-heme motif for trans- membrane electron transport conserved in a y east iron reductase and the human NADPH oxidase. J. Biol. Chem. 271, 31021– 31024. 16. Rotrosen, D., Kleinberg, M.E., Nunoi, H., Leto, T., Gallin, J.I. & Malech, H.L . (19 90) Ev idence f or a f unc tional cy topl asmic domain of phagocyte oxidase cytochrome b558. J. Biol. Ch em. 265, 8745–8750. 17. Pinto, L.H., Holsinger, L.J. & Lamb, R.A. (1992) Influenza virus M2 protein h as ion channel activity. Cell 69, 517–528. 18. Duff, K.C. & Ashley, R.H. (1992) The transmembrane domain of influenza A M 2 p rotein forms amantadine sensitive proton channels in planar lipid bilayers. Virology 190, 485–489. 19. Tosteson,M.T.,Pinto,L.H.,Holsinger,L.J.&Lamb,R.A.(1994) Reconstitution of the influenza virus M2 ion channel in lipid bilayers. J. M embr. Biol. 42, 117–126. 20. Wang, C., Lamb, R.A. & Pinto, L.H. (1994) Direct measurement of th e influenza A virus M2 protein io n channel activity in mammalian cells. Virology 205 , 133–140. 21. Chizhmakov, I.V., Geraghty, F.M., Ogden, D.C., Hayhurst, A., Antoniou, M. & Hay, A.J. (1996) Selective proton permeability and pH regulation of the influenza virus M2 channel expressed in mouse erythroleukaemia c ells. J. Physiol. 494, 329–336. 22. Okada, A., Miura, T. & Takeuchi, H. (2001) Protonation of histidine and histidine–tryptophan interaction in the activation of the M2 ion channel from influenza A virus. Biochemistry 40, 6053–6060. 23. Wang, C., Takeuchi, K., Pinto, L.H. & Lamb, R.A. (1993) Ion channel a ctivity of influenza A virus M2 protein: c haracterizatio n of the amantidine block. J. Virol. 67, 5 585–5594. 24. Wang, C., Lamb, R .A. & Pinto, L.H. (1995) Activation of the M2 ion channel of influenza virus: a ro le for the transmembrane domain histidine residue. Biophys. J. 69, 1363–1371. 25. Tang, Y., Zaitseva, F., Lamb, R.A. & Pinto, L.H. (2002) The gate of the influe nza virus M2 proton channel is formed by a single tryptophan residue. J. Biol. Chem. 277, 39880–39886. 26. Dana, R., Leto, T.L., Male ch, H .L. & Levy, R. (1998) Essential requirement of cytosolic phospholipase A 2 for activation o f the phagocyte NADPH oxidase. J. Biol. Chem. 273, 441–445. 27. Lowenthal, A. & Levy, R. (1999) Essential requirement of cyto- solic phospholipase A 2 for a ctivation of the H + channel in pha- gocyte-like cells. J. Biol. Chem. 274, 21603–21608. 28.Mankelow,T.J.,Pessach,E.I.,Levy,R.&Henderson,L.M. (2003) The requirement of cytosolic phospholipase A 2 for the PMAactivationofprotonefflux through the N-terminal 230- amino-acid fragment of gp91 phox . Biochem. J. 374, 315–319. Supplementary material The following material is available f rom http://www. blackwellpublishing.com/products/jour nals/suppmat/ EJB/EJB4340/EJB4340sm.htm Table S1. The Nox and Duox gene sequences used to construct the alignment in Fig. 1. Fig. S1. The third predicted t ransmembrane helix of Nox2. Ó FEBS 2004 Histidine 119 and conduction protons through Nox2 (Eur. J. Biochem. 271) 4033 . Investigation of the contribution of histidine 119 to the conduction of protons through human Nox2 Tosti J. Mankelow*, X. Wen. regarding the role o f His111 in the conduction of protons through Nox2. Therefore, we can conclude that the conduction of protons through human Nox2 requires