A Ca2+-activated NADPH Oxidase in Testis, Spleen, and Lymph Nodes
2001; Elsevier BV; Volume: 276; Issue: 40 Linguagem: Inglês
10.1074/jbc.m103034200
ISSN1083-351X
AutoresBotond Bánfi, Gergely Molnár, Andrés D. Maturana, Klaus Steger, Balázs Hegedűs, Nicolas Demaurex, Karl-Heinz Krause,
Tópico(s)Neonatal Health and Biochemistry
ResumoSuperoxide and its derivatives are increasingly implicated in the regulation of physiological functions from oxygen sensing and blood pressure regulation to lymphocyte activation and sperm-oocyte fusion. Here we describe a novel superoxide-generating NADPH oxidase referred to as NADPH oxidase 5 (NOX5). NOX5 is distantly related to the gp91phox subunit of the phagocyte NADPH oxidase with conserved regions crucial for the electron transport (NADPH, FAD and heme binding sites). However, NOX5 has a unique N-terminal extension that contains three EF hand motifs. The mRNA of NOX5 is expressed in pachytene spermatocytes of testis and in B- and T-lymphocyte-rich areas of spleen and lymph nodes. When heterologously expressed, NOX5 was quiescent in unstimulated cells. However, in response to elevations of the cytosolic Ca2+concentration it generated large amounts of superoxide. Upon Ca2+ activation, NOX5 also displayed a second function: it became a proton channel, presumably to compensate charge and pH alterations due to electron export. In summary, we have identified a novel NADPH oxidase that generates superoxide and functions as a H+ channel in a Ca2+-dependent manner. NOX5 is likely to be involved in Ca2+-activated, redox-dependent processes of spermatozoa and lymphocytes such as sperm-oocyte fusion, cell proliferation, and cytokine secretion. Superoxide and its derivatives are increasingly implicated in the regulation of physiological functions from oxygen sensing and blood pressure regulation to lymphocyte activation and sperm-oocyte fusion. Here we describe a novel superoxide-generating NADPH oxidase referred to as NADPH oxidase 5 (NOX5). NOX5 is distantly related to the gp91phox subunit of the phagocyte NADPH oxidase with conserved regions crucial for the electron transport (NADPH, FAD and heme binding sites). However, NOX5 has a unique N-terminal extension that contains three EF hand motifs. The mRNA of NOX5 is expressed in pachytene spermatocytes of testis and in B- and T-lymphocyte-rich areas of spleen and lymph nodes. When heterologously expressed, NOX5 was quiescent in unstimulated cells. However, in response to elevations of the cytosolic Ca2+concentration it generated large amounts of superoxide. Upon Ca2+ activation, NOX5 also displayed a second function: it became a proton channel, presumably to compensate charge and pH alterations due to electron export. In summary, we have identified a novel NADPH oxidase that generates superoxide and functions as a H+ channel in a Ca2+-dependent manner. NOX5 is likely to be involved in Ca2+-activated, redox-dependent processes of spermatozoa and lymphocytes such as sperm-oocyte fusion, cell proliferation, and cytokine secretion. reactive oxygen species cytosolic free Ca2+ concentration dual oxidase 91-kDa glycoprotein subunit of the phagocyte NADPH oxidase kilo base nitro blue tetrazolium NADPH oxidase NADPH oxidase 5 rapid amplification of cDNA ends superoxide dismutase polymerase chain reaction reverse transcription-PCR 1,4-piperazinediethanesulfonic acid N-hydroxyethylethylenediaminetriacetic acid 4-morpholineethanesulfonic acid Over the last 10 years, our concept of the biological role of reactive oxygen species (ROS)1 has remarkably evolved. In addition to their long known toxic effects, which are involved in the killing of microorganisms, aging, and cancerogenesis, ROS are now recognized as physiologically relevant signaling molecules. They have been shown to regulate a large number of biological processes including gene expression (1Schoonbroodt S. Piette J. Biochem. Pharmacol. 2000; 60: 1075-1083Crossref PubMed Scopus (215) Google Scholar), kinase activation (2Lo Y.Y. Wong J.M. Cruz T.F. J. Biol. Chem. 1996; 271: 15703-15707Abstract Full Text Full Text PDF PubMed Scopus (455) Google Scholar), oxygen sensing (3Ebert B.L. Bunn H.F. Blood. 1999; 94: 1864-1877Crossref PubMed Google Scholar), regulation of vascular diameter (4Munzel T. Hink U. Heitzer T. Meinertz T. Ann. N. Y. Acad. Sci. 1999; 874: 386-400Crossref PubMed Scopus (82) Google Scholar), bone resorption (5Key L.L. Wolf W.C. Gundberg C.M. Ries W.L. Bone. 1994; 15: 431-436Crossref PubMed Scopus (113) Google Scholar), cell growth (6Burdon R.H. Free Radic. Biol. Med. 1995; 18: 775-794Crossref PubMed Scopus (1076) Google Scholar), and even mediation of sperm-oocyte fusion (7Aitken R.J. Paterson M. Fisher H. Buckingham D.W. van Duin M. J. Cell Sci. 1995; 108: 2017-2025PubMed Google Scholar). Superoxide generation by phagocytes plays an important role in the killing of microorganisms during host defense. It is catalyzed by the phagocyte NADPH oxidase, an enzyme complex that has been known and studied for a long time (8Clark R.A. J. Infect. Dis. 1999; 179: S309-17Crossref PubMed Scopus (151) Google Scholar, 9Babior B.M. Blood. 1999; 93: 1464-1476Crossref PubMed Google Scholar, 10Segal A.W. Abo A. Trends Biochem. Sci. 1993; 18: 43-47Abstract Full Text PDF PubMed Scopus (587) Google Scholar). Upon assembly, the NADPH oxidase generates electron currents (11Schrenzel J. Serrander L. Banfi B. Nusse O. Fouyouzi R. Lew D.P. Demaurex N. Krause K.H. Nature. 1998; 392: 734-737Crossref PubMed Scopus (168) Google Scholar) that flow from intracellular NADPH to extracellular (or phagosomal) oxygen resulting in superoxide generation. The main subunit of the phagocyte NADPH oxidase, gp91phox forms a transmembrane heterodimer with p22phox and functions as an electron transport chain containing four NADPH binding regions, an FAD binding site, and two heme groups anchored by four histidines (12Rotrosen D. Yeung C.L. Leto T.L. Malech H.L. Kwong C.H. Science. 1992; 256: 1459-1462Crossref PubMed Scopus (354) Google Scholar, 13Finegold A.A. Shatwell K.P. Segal A.W. Klausner R.D. Dancis A. J. Biol. Chem. 1996; 271: 31021-31024Abstract Full Text Full Text PDF PubMed Scopus (166) Google Scholar). However, association of cytosolic proteins, including p47phox, p67phox, and Rac2, with the transmembrane subunits is crucial for the activation of the electron flow (9Babior B.M. Blood. 1999; 93: 1464-1476Crossref PubMed Google Scholar). Over the last several years it has become apparent that nonphagocytic cells (e.g. vascular smooth muscle (4Munzel T. Hink U. Heitzer T. Meinertz T. Ann. N. Y. Acad. Sci. 1999; 874: 386-400Crossref PubMed Scopus (82) Google Scholar), endothelial (4Munzel T. Hink U. Heitzer T. Meinertz T. Ann. N. Y. Acad. Sci. 1999; 874: 386-400Crossref PubMed Scopus (82) Google Scholar), mesangial (14Jaimes E.A. Galceran J.M. Raij L. Kidney Int. 1998; 54: 775-784Abstract Full Text Full Text PDF PubMed Scopus (254) Google Scholar), thyroid (15Dupuy C. Virion A. Ohayon R. Kaniewski J. Deme D. Pommier J. J. Biol. Chem. 1991; 266: 3739-3743Abstract Full Text PDF PubMed Google Scholar), fibroblasts (16Meier B. Radeke H.H. Selle S. Younes M. Sies H. Resch K. Habermehl G.G. Biochem. J. 1989; 263: 539-545Crossref PubMed Scopus (595) Google Scholar), lymphocytes (17Lee J.R. Koretzky G.A. Eur. J. Immunol. 1998; 28: 4188-4197Crossref PubMed Scopus (74) Google Scholar), (18Pani G. Colavitti R. Borrello S. Galeotti T. Biochem. J. 2000; 347: 173-181Crossref PubMed Scopus (96) Google Scholar), and spermatozoa (19Aitken J. Fisher H. Bioessays. 1994; 16: 259-267Crossref PubMed Scopus (490) Google Scholar)) also generate superoxide. The subsequent search for nonphagocyte NADPH oxidases led to the discovery of two families of gp91phoxhomologues. The NADPH oxidase (NOX) family members have approximately the same length as gp91phox (i.e.∼560–580 amino acids), while the dual oxidase (DUOX) family members are markedly longer (∼1550 amino acids) because of their N-terminal extensions consisting of two EF hand (i.e. presumed Ca2+ binding) motifs, an additional transmembrane helix, and a peroxidase homology domain (20Lambeth J.D. Cheng G. Arnold R.S. Edens W.A. Trends Biochem. Sci. 2000; 25: 459-461Abstract Full Text Full Text PDF PubMed Scopus (337) Google Scholar). The NOX family includes NOX1 (initially referred to as Mox1 or NOH-1, Refs. 21Suh Y.A. Arnold R.S. Lassegue B. Shi J. Xu X. Sorescu D. Chung A.B. Griendling K.K. Lambeth J.D. Nature. 1999; 401: 79-82Crossref PubMed Scopus (1299) Google Scholar and 22Banfi B. Maturana A. Jaconi S. Arnaudeau S. Laforge T. Sinha B. Ligeti E. Demaurex N. Krause K.H. Science. 2000; 287: 138-142Crossref PubMed Scopus (257) Google Scholar), which is predominantly expressed in colon, NOX3 (also referred to as gp91–3, Ref. 23Kikuchi H. Hikage M. Miyashita H. Fukumoto M. Gene (Amst.). 2000; 254: 237-243Crossref PubMed Scopus (129) Google Scholar) cloned from fetal kidney, and NOX4 found in kidney cortex (also termed Renox, Refs. 24Geiszt M. Kopp J.B. Varnai P. Leto T.L. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8010-8014Crossref PubMed Scopus (731) Google Scholar and 25Shiose A. Kuroda J. Tsuruya K. Hirai M. Hirakata H. Naito S. Hattori M. Sakaki Y. Sumimoto H. J. Biol. Chem. 2001; 276: 1417-1423Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar) and in osteoclasts (26Yang S. Madyastha P. Bingel S. Ries W. Key L. J. Biol. Chem. 2001; 276: 5452-5458Abstract Full Text Full Text PDF PubMed Scopus (187) Google Scholar). The DUOX family includes DUOX1 (initially termed Thox1, Refs. 27De Deken X. Wang D. Many M.C. Costagliola S. Libert F. Vassart G. Dumont J.E. Miot F. J. Biol. Chem. 2000; 275: 23227-23233Abstract Full Text Full Text PDF PubMed Scopus (517) Google Scholar and 28Dupuy C. Ohayon R. Valent A. Noel-Hudson M.S. Deme D. Virion A. J. Biol. Chem. 1999; 274: 37265-37269Abstract Full Text Full Text PDF PubMed Scopus (385) Google Scholar) expressed in the thyroid gland and DUOX2 (also designated as Thox2, Ref. 29Dupuy C. Pomerance M. Ohayon R. Noel-Hudson M.S. Deme D. Chaaraoui M. Francon J. Virion A. Biochem. Biophys. Res. Commun. 2000; 277: 287-292Crossref PubMed Scopus (59) Google Scholar), which is found in thyroid, small intestine, and colon. NOX1 and NOX4 have been shown to produce low amounts of superoxide in a constitutively active manner (21Suh Y.A. Arnold R.S. Lassegue B. Shi J. Xu X. Sorescu D. Chung A.B. Griendling K.K. Lambeth J.D. Nature. 1999; 401: 79-82Crossref PubMed Scopus (1299) Google Scholar, 24Geiszt M. Kopp J.B. Varnai P. Leto T.L. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8010-8014Crossref PubMed Scopus (731) Google Scholar, 25Shiose A. Kuroda J. Tsuruya K. Hirai M. Hirakata H. Naito S. Hattori M. Sakaki Y. Sumimoto H. J. Biol. Chem. 2001; 276: 1417-1423Abstract Full Text Full Text PDF PubMed Scopus (454) Google Scholar). Hitherto no functional evidence for superoxide generation by NOX3, DUOX1, and DUOX2 has been obtained. The novel gp91phox homologues have been proposed to regulate cell growth (NOX1), participate in host defense (NOX1), mediate in oxygen sensing (NOX4), or play a role in thyroid hormone synthesis (DUOX1 and -2); however, so far no conclusive evidence concerning these proposals exists. Activation of DUOX enzymes by elevations of the cytosolic free Ca2+ concentration, [Ca2+]c, has been suggested based on their EF hand motifs and the previously described Ca2+-activated superoxide generation in thyroid cells (30Nakamura Y. Makino R. Tanaka T. Ishimura Y. Ohtaki S. Biochemistry. 1991; 30: 4880-4886Crossref PubMed Scopus (33) Google Scholar). But hitherto no direct experimental evidence exists. [Ca2+]c elevations do not activate NOX4 (24Geiszt M. Kopp J.B. Varnai P. Leto T.L. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8010-8014Crossref PubMed Scopus (731) Google Scholar), and there is no evidence that they would activate NOX1 or NOX3. gp91phox can be activated by Ca2+, although most likely not because of a direct effect of Ca2+ on gp91phox but rather because of the activation of Ca2+-sensitive second messenger systems (31Christiansen N.O. Larsen C.S. Esmann V. Biochim. Biophys. Acta. 1988; 971: 317-324PubMed Google Scholar, 32Elzi D.J. Bjornsen A.J. MacKenzie T. Wyman T.H. Silliman C.C. Am. J. Physiol. 2001; 281: C350-C360Crossref PubMed Google Scholar, 33Nakamura T. Suchard S.J. Abe A. Shayman J.A. Boxer L.A. J. Leukoc. Biol. 1994; 56: 105-109Crossref PubMed Scopus (12) Google Scholar) and Ca2+-dependent exocytosis of gp91phox-containing granules. Electron export by an NADPH oxidase has important electrophysiological consequences: loss of negative charges and accumulation of H+ ions lead to plasma membrane depolarization (34Jankowski A. Grinstein S. J. Biol. Chem. 1999; 274: 26098-26104Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar, 35Banfi B. Schrenzel J. Nusse O. Lew D.P. Ligeti E. Krause K.H. Demaurex N. J. Exp. Med. 1999; 190: 183-194Crossref PubMed Scopus (119) Google Scholar) and to cytosolic acidification (36Grinstein S. Furuya W. Am. J. Physiol. 1986; 251: C55-C65Crossref PubMed Google Scholar). There is now growing evidence that, in addition to their electron transport function, gp91phox and its homologues may also function as H+ channels, allowing charge and pH compensation (35Banfi B. Schrenzel J. Nusse O. Lew D.P. Ligeti E. Krause K.H. Demaurex N. J. Exp. Med. 1999; 190: 183-194Crossref PubMed Scopus (119) Google Scholar, 37Henderson L.M. Meech R.W. J. Gen. Physiol. 1999; 114: 771-786Crossref PubMed Scopus (57) Google Scholar). Moreover, there is an alternatively spliced short form of NOX1, which is devoid of elements of the electron transport chain. For this protein, H+ conduction may even be the only physiological role (22Banfi B. Maturana A. Jaconi S. Arnaudeau S. Laforge T. Sinha B. Ligeti E. Demaurex N. Krause K.H. Science. 2000; 287: 138-142Crossref PubMed Scopus (257) Google Scholar). Here we describe a distant relative of the members of NOX and DUOX families, NADPH oxidase 5 (NOX5). NOX5, which is found in testis and lymphoid organs, contains an N-terminal extension with three EF hands and is able to generate superoxide and to conduct H+ ions in response to cytosolic free [Ca2+] elevations. BLAST nucleotide searches were conducted in the High Throughput Genomic Sequences data base of GenBankTM with the C terminus of human gp91phox. The exons of the found gene, including the one containing the stop codon, were predicted either based on their homology with gp91phox or with the GENSCAN software (genes.mit.edu/oldGENSCAN.html). Primers were designed from the first predicted exon and from the exon encoding the predicted stop codon (5′-GCT CCG GTG GGT GAC TCA GCA GTT T-3′ and 5′-ATG GTG TGG ACT TGG GGC AGA GCT T-3′), and PCR was performed on several human cDNAs prepared from different organs (Multiple Tissue cDNA Panel II,CLONTECH, Basel, Switzerland) with Taqpolymerase using the “Q solution” (Qiagen, Basel, Switzerland) to help the melting of the GC-rich regions. The 5′-ends of the cDNAs were obtained by 5′-RACE PCR with the outer primer 5′-CTC CTG GAG GGT GAT GGT GCC ACT T-3′ and then with the nested primer 5′-GGT GAT GGT GCC ACT TCT ATC GGA G-3′ using the SMART RACE cDNA Amplification kit (CLONTECH), but random primers were applied instead of oligo(dT) for the reverse transcription of spleen and testis mRNAs. Human Multiple Tissue Northern Blots (CLONTECH) were hybridized with a32P-labeled NOX5β cDNA fragment corresponding to amino acids 460–625. The hybridization was performed under high stringency conditions using the ExpressHyb hybridization solution (CLONTECH). For in situ hybridization experiments the in vitro transcription of digoxigenin-labeled cRNA was performed using the RNA-DIG Labeling Mix (Roche Molecular Biochemicals) and RNA polymerases T7 and SP6. Prior to cRNA synthesis, the vector containing the NOX5β insert had been digested with ApaI or PstI (New England Biolabs, Frankfurt, Germany) for the production of sense cRNA or antisense cRNA. In situ hybridization was performed as reported previously (38Steger K. Pauls K. Klonisch T. Franke F.E. Bergmann M. Mol. Hum. Reprod. 2000; 6: 219-225Crossref PubMed Scopus (115) Google Scholar). Briefly, 7-μm sections from Bouin-fixed and paraffin-embedded tissue samples of human testis, spleen, and lymph node were mounted on slides coated with 2% aminopropyltriethoxysilane (Sigma). Deparaffinized and rehydrated tissue sections were digested with proteinase K (20 μg/ml of 1× phosphate-buffered saline) for 30 min at 37 °C and prehybridized in 20% glycerol for 30 min. Sections were then incubated with the digoxigenin-labeled sense and antisense cRNA probes at a dilution of 1:100 in hybridization buffer containing 50% deionized formamide, 10% dextran sulfate, 2× SSC (0.3m NaCl, 0.03 m trisodium citrate), 1× Denhardt's solution, 10 μg/ml salmon sperm DNA, and 10 μg/ml yeast tRNA. Hybridization was performed overnight at 37 °C in a humid chamber containing 50% formamide in 2× SSC. Posthybridization washes were performed as reported previously (39Lewis F.A. Wells M. Wilkinson D.G. In Situ Hybridization—A Practical Approach. Oxford University Press, Oxford1992: 121-135Google Scholar). Tissue samples were incubated overnight at 4 °C with an anti-digoxigenin Fab antibody conjugated to alkaline phosphatase (Roche Molecular Biochemicals). Staining was visualized with NBT/5-bromo-4-chloro-3-indolyl phosphate (KPL, Gaithersburg, MD) in a humid chamber protected from light. Finally, sections were mounted in glycerol gel and examined under bright-field microscopy. COS-7, HEK293, and HeLa cells were maintained in Dulbecco's modified Eagle's medium, and HL-60 and PLB-985 cells were maintained in RPMI 1640 medium. Both culture media were supplemented with 10% fetal calf serum, penicillin (100 units/ml), streptomycin (100 μg/ml), and 4 mmol/literl-glutamine. HL-60 and PLB-985 cells were differentiated for 5 days with 1.2% Me2SO and 0.5% dimethylformamide, respectively. Human neutrophils were purified as described previously (40Mocsai A. Banfi B. Kapus A. Farkas G. Geiszt M. Buday L. Farago A. Ligeti E. Biochem. Pharmacol. 1997; 54: 781-789Crossref PubMed Scopus (43) Google Scholar). For transfection, the coding regions of human NOX5β cDNA and mouse NOX4 cDNA (both with an inserted Kozak sequence) were subcloned into pcDNA3.1 (Invitrogen, Groningen, Netherlands). COS-7, HEK293, and HeLa cells were transfected with those constructs using the Effectene transfection system (Qiagen). To obtain stable clones, NOX5β-transfected HEK293 cells were selected with 400 μg/ml G418 starting on the 2nd day after the transfection, and 13 surviving colonies were isolated 10 days after the transfection. Total RNA of the cells were prepared with Trizol reagent (Life Technologies, Inc.). The RT-PCR for NOX5 was performed with GeneAmp (PerkinElmer Life Sciences) using the following primers: 5′-CAC TAT AGA CCT GGT GAC TA-3′ and 5′-CAT GCT CAG AGG CAA AGA T-3′. Either the cDNAs for the PCRs were purchased fromCLONTECH (Human Blood Fractions Panel) or total RNA was reverse transcribed with the Superscript reverse transcriptase (Life Technologies, Inc.). NOX5 was amplified with the following primers: 5′-ATG AGT GGC ACC CCT TCA CCA TCA G-3′ and 5′-GTC AGC AGG CTC ACA AAC CAC TCG AA-3′. ROS generation was measured by the peroxidase-dependent luminol-amplified chemiluminescence technique on a Luminometer Wallac 1420 Multilabel Counter (PerkinElmer Life Sciences). Cells were plated 18 h before the measurement in 96-well plates. Measurements were performed in Hanks' balanced salt solution (containing 1.26 mm Ca2+) supplemented with 6 units/ml horseradish peroxidase and 250 μm luminol. To elevate intracellular Ca2+ concentration 0.1–3 μmionomycin (Sigma) was used. Chemiluminescence was measured once per minute at 37 °C. After the measurements cells were counted, and the results were normalized to 20,000 cells. Extracellular superoxide production was measured in 96-well microplates at 550 nm as the SOD-sensitive reduction of 100 μmferricytochrome C. The O⨪2 production was calculated using an absorption coefficient of 21.1 mm−1cm−1 and normalized to 1 min and 106 cells (40Mocsai A. Banfi B. Kapus A. Farkas G. Geiszt M. Buday L. Farago A. Ligeti E. Biochem. Pharmacol. 1997; 54: 781-789Crossref PubMed Scopus (43) Google Scholar). To measure both intra- and extracellular superoxide generation a quantitative NBT test was used (41Rook G.A. Steele J. Umar S. Dockrell H.M. J. Immunol. Methods. 1985; 82: 161-167Crossref PubMed Scopus (325) Google Scholar). The cells were plated on 48-well plates using 200,000 polymorphonuclear granulocytes or HL-60 cells/well or 450,000 HEK293 cells/well and incubated in Hanks' balanced salt solution containing 0.5 mg/ml NBT with or without stimuli (100 nm phorbol 12-myristate 13-acetate or 1 μmionomycin) and with or without 800 units/ml SOD. After 12 min the cells were fixed and washed with methanol to remove the nonreduced NBT. The reduced formazan was then dissolved in 230 μl of 2 mpotassium hydroxide and in 280 μl of dimethyl sulfoxide, and the absorption was measured at 630 nm. The absorption of dissolved formazan reduced by the stimulated cells was regarded as 100%, while the absorption of dissolved formazan reduced by the resting cells was regarded as 0%. Whole-cell patch-clamp recordings were performed as described previously (22Banfi B. Maturana A. Jaconi S. Arnaudeau S. Laforge T. Sinha B. Ligeti E. Demaurex N. Krause K.H. Science. 2000; 287: 138-142Crossref PubMed Scopus (257) Google Scholar) using an Axopatch 1D amplifier (Axon Instruments, Foster City, CA) in the voltage clamp mode. Patch pipettes were pulled from borosilicate glass with a P-87 Brown-Flaming Micropipette Puller (Sutter Instrument Company, Novato, CA). Pipette resistance ranged between 3 and 7 megaohms, seal resistance ranged between 2 and 20 gigaohms, and the mean access resistance varied between 10 and 30 megaohms. Cells were voltage-clamped at a holding potential of −60 mV and depolarized to various test potentials as indicated. The currents were filtered at 1 kHz and sampled at 200 Hz using the pClamp 6.0 software. Leak currents were small compared with the whole-cell H+ currents and were subtracted only to allow calculation of the whole-cell conductance. The bath solution contained (in mm): CsCl 75, HEPES 100, CsOH 50, MgCl2 1, EGTA 0.2, HEDTA 0.5, pH = 7.5; the pipette solution contained (in mm): CsCl 35, PIPES 100, CsOH 132, MgCl2 1, EGTA 0.2, HEDTA 0.5, and 2 mm ATP, pH = 6.5, free [Ca2+] <10 nm. The high Ca2+ solutions contained 0.55 and 0.65 mm CaCl2 in the bath and pipette, respectively, to achieve a free [Ca2+] of 75 μm. For H+ selectivity experiments, the pipette pH was buffered to pH = 5.7 with 100 mm MES or pH = 7.5 with 100 mm HEPES. Recordings were performed 15 min after achieving the whole-cell configuration to allow the equilibration of the cytosol with the pipette solution. A search of the High Throughput Genomic Sequences data base (www.ncbi.nlm.nih.gov/BLAST/) with the gp91phox sequence yielded a human gene (GenBankTM accession number AC026512) encoding possible exons of a novel NADPH oxidase. From the predicted exons, we designed PCR primers and screened cDNAs of several tissues by PCR. Prominent bands around 2 kb were obtained in spleen and testis. Sequencing of the PCR products demonstrated the expected homology with gp91phox and an in-frame stop codon. A 495-base pair fragment of the PCR product was then used as a probe for Northern blot analysis of tissue distribution of the novel NADPH oxidase. The Northern blot experiment showed an ∼2.7-kb transcript in spleen and an ∼2.9-kb mRNA in testis (Fig.1A). Additional Northern blot experiments (not shown) did not detect the mRNA of the novel NADPH oxidase in the following tissues: brain, heart, skeletal muscle, kidney, liver, placenta, and lung. The Northern blots yielded the following information. (i) No hybridization was detected in peripheral blood leukocytes, colon (Fig. 1A), and kidney (data not shown), suggesting a restricted tissue distribution and no cross-hybridization of the probe with gp91phox, NOX1, or NOX4 mRNAs. (ii) The length of the transcripts was longer than the mRNAs of previously described NOX family members (∼2.4–2.5 kb) but shorter than the transcripts of DUOX family members (∼5.7–6.4 kb) (27De Deken X. Wang D. Many M.C. Costagliola S. Libert F. Vassart G. Dumont J.E. Miot F. J. Biol. Chem. 2000; 275: 23227-23233Abstract Full Text Full Text PDF PubMed Scopus (517) Google Scholar). (iii) The length of the transcripts of the novel NADPH oxidase was different in testis and in spleen. The mRNA of the novel NADPH oxidase was present in spleen but was undetectable by Northern blot in peripheral blood leukocytes. To verify this result, we applied the more sensitive PCR technique to detect possible transcripts in different subsets of lymphocytes derived from blood and in myeloid cell lines. However, even after 35 PCR cycles no transcripts could be detected in circulating blood lymphocytes or in two myeloid cell lines, HL-60 and PLB-985 (Fig. 1B). PCR amplification, however, detected transcripts in testis, uterus, bone marrow, and vascular smooth muscle (Fig. 1B). To obtain the start codon of the novel gp91phox homologue, primers were designed from the 5′-end of the 2-kb-long initial PCR product. The 5′-RACE PCR performed with those primers demonstrated a 220-base pair difference of the 5′-ends of the testis and spleen transcripts (Fig.1B). Thus, the difference observed with the 5′-RACE PCR was in accordance with the length difference of the testis and spleen mRNAs detected by Northern blot. To amplify the coding region of the detected mRNAs we designed primers from the sequence located in the 5′ direction from the start codon (identified by the 5′-RACE PCR) and the 3′ direction from the stop codon (identified by the initial PCR). The PCR products amplified with those primers demonstrated that, with the exception of the 5′-regions, the cDNA sequences of the spleen and the testis isoforms were identical, strongly suggesting that both transcripts were the products of the same gene. The 220-base pair difference at the 5′-end of the RNAs was mainly due to different 5′-untranslated regions. In addition, however, the sequence analysis predicts that the shorter RNA of spleen encodes a protein 18 amino acids longer than the protein encoded by the longer mRNA of testis (Fig.2A). The comparison of the amino acid sequences with gp91phox revealed that the regions crucial for NADPH oxidation and electron transport (NADPH and FAD binding sequences, presumed heme-anchoring histidines) and the all-over structure of the protein (predicted with TMpred at www.ch.embnet.org/software/TMPRED_form.html) were conserved (Fig. 2,A and B). However, the sequence identity with gp91phox was relatively low (27%, as compared with 56% for NOX1, and 37% for NOX4), and unlike all the other NOX enzymes the novel NADPH oxidase showed an extended N terminus containing three EF hand (e.g. presumed Ca2+binding) motifs. Two EF hand motifs were also found in DUOX1 and DUOX2 but hitherto not in any of the NOX family members. In contrast to the DUOX enzymes, however, the novel enzyme does not have a peroxidase homology domain. Thus, we will refer to the enzyme as NOX5. The spleen and testis isoforms will be referred to as NOX5α and NOX5β, respectively (GenBankTM accession numbers AF353088 andAF325189). Further sequencing revealed two additional splice variants referred to as NOX5γ and NOX5δ (GenBankTM accession numbers AF353089 and AF325190, respectively). Note also that there are presently five additional GenBankTM entries of NOX5-related sequences: (i) two human cDNA sequences that do not show the N-terminal EF hand regions (accession numbers AAG33638 as NOX5 (42Cheng G. Cao Z. Xu X. Meir E.G. Lambeth J.D. Gene (Amst.). 2001; 269: 131-140Crossref PubMed Scopus (724) Google Scholar) andAK026011 as an unnamed protein), (ii) two short porcine expressed sequence tag clones (accession numbers BF444388 and AW416159), and (iii) a Drosophila melanogaster protein sequence predicted from the gene (accession number AE003807) that shows a high degree of similarity with human NOX5, including the presence of the N-terminal EF hands. Based on the genomic sequence found in the GenBankTM data base we analyzed the genomic organization of NOX5. The NOX5 gene has a size of at least 30 kb (the genome fragment encoding the NOX5 gene contains gaps in its sequence) and contains at least 19 exons. Exons 3–19 generate the NOX5α mRNA, while exons 1, 2, and 4–19 encode the NOX5β mRNA (Fig. 2C). The predicted start codon for NOX5α is in exon 3, while NOX5β is predicted to use a start codon in exon 4. This explains why the shorter mRNA is predicted to generate a larger protein. The exon-intron boundaries of NOX5 are very different from members of the NOX and DUOX families. This is in striking contrast with the conserved exon-intron boundaries within the NOX and DUOX families (for example, see Ref. 22Banfi B. Maturana A. Jaconi S. Arnaudeau S. Laforge T. Sinha B. Ligeti E. Demaurex N. Krause K.H. Science. 2000; 287: 138-142Crossref PubMed Scopus (257) Google Scholar). Thus, there appears to be a considerable evolutionary distance between NOX5 and other members of the NOX and the DUOX families. We next localized NOX5 by in situ hybridization. In testis, the antisense NOX5 probe strongly labeled pachytene spermatocytes (Fig.3A, arrows) and more weakly labeled round spermatids (Fig. 3A,arrowheads). The sense probe did not give any signal proving the specificity of the antisense hybridization (Fig. 3B). In spleen, the antisense NOX5 probe strongly hybridized with the mantle zone surrounding the germinal centers and also with the periarterial lymphoid sheaths of the white pulp (Fig. 3C), while the sense probe did not produce any signal (Fig. 3D). Thus, areas rich in mature B-lymphocytes (mantle zone) as well as areas rich in T-lymphocytes (periarterial lymphoid sheaths) were labeled. To investigate whether in other lymphoid organs NOX5 mRNA has similar localization, human lymph node sections were hybridized with the
Referência(s)