Topological Mapping of Neutrophil Cytochrome b Epitopes with Phage-display Libraries
1995; Elsevier BV; Volume: 270; Issue: 28 Linguagem: Inglês
10.1074/jbc.270.28.16974
ISSN1083-351X
AutoresJames B. Burritt, Mark T. Quinn, Mark A. Jutila, Clifford W. Bond, Algirdas J. Jesaitis,
Tópico(s)Glycosylation and Glycoproteins Research
ResumoCytochrome b of human neutrophils is the central component of the microbicidal NADPH-oxidase system. However, the folding topology of this integral membrane protein remains undetermined. Two random-sequence bacteriophage peptide libraries were used to map structural features of cytochrome b by determining the epitopes of monoclonal antibodies (mAbs) 44.1 and 54.1, specific for the p22phox and gp91phox cytochrome b chains, respectively. The unique peptides of phage selected by mAb affinity purification were deduced from the phage DNA sequences. Phage selected by mAb 44.1 displayed the consensus peptide sequence GGPQVXPI, which is nearly identical to 181GGPQVNPI188 of p22phox. Phage selected by mAb 54.1 displayed the consensus sequence PKXAVDGP, which resembles 382PKIAVDGP389 of gp91phox. Western blotting demonstrated specific binding of each mAb to the respective cytochrome b subunit and selected phage peptides. In flow cytometric analysis, mAb 44.1 bound only permeabilized neutrophils, while 54.1 did not bind intact or permeabilized cells. However, mAb 54.1 immunosedimented detergent-solubilized cytochrome b in sucrose gradients. These results suggest the 181GGPQVNPI188 segment of p22phox is accessible on its intracellular surface, but the 382PKIAVDGP389 region on gp91phox is not accessible to antibody, and probably not on the protein surface. Cytochrome b of human neutrophils is the central component of the microbicidal NADPH-oxidase system. However, the folding topology of this integral membrane protein remains undetermined. Two random-sequence bacteriophage peptide libraries were used to map structural features of cytochrome b by determining the epitopes of monoclonal antibodies (mAbs) 44.1 and 54.1, specific for the p22phox and gp91phox cytochrome b chains, respectively. The unique peptides of phage selected by mAb affinity purification were deduced from the phage DNA sequences. Phage selected by mAb 44.1 displayed the consensus peptide sequence GGPQVXPI, which is nearly identical to 181GGPQVNPI188 of p22phox. Phage selected by mAb 54.1 displayed the consensus sequence PKXAVDGP, which resembles 382PKIAVDGP389 of gp91phox. Western blotting demonstrated specific binding of each mAb to the respective cytochrome b subunit and selected phage peptides. In flow cytometric analysis, mAb 44.1 bound only permeabilized neutrophils, while 54.1 did not bind intact or permeabilized cells. However, mAb 54.1 immunosedimented detergent-solubilized cytochrome b in sucrose gradients. These results suggest the 181GGPQVNPI188 segment of p22phox is accessible on its intracellular surface, but the 382PKIAVDGP389 region on gp91phox is not accessible to antibody, and probably not on the protein surface. The NADPH-oxidase system of neutrophils is a host-defensive plasma membrane redox system that produces superoxide anion (O)(1Badwey J.A. Karnovsky M.L. Annu. Rev. Biochem. 1980; 49: 695-726Crossref PubMed Scopus (839) Google Scholar, 2Morel F. Doussiere J. Vagnais P.V. Eur. J. Biochem. 1991; 201: 523-546Crossref PubMed Scopus (523) Google Scholar), which subsequently is converted to a variety of other toxic oxygen species that kill invading microbes and cause damage to tissue(3Smith R.M. Curnutte J.T. Blood. 1991; 77: 673-686Crossref PubMed Google Scholar). Humans lacking this enzyme system are unable to produce neutrophil-generated superoxide and suffer recurrent bacterial infections, granulomatous lesions of multiple organs, and early death (4). This condition was first reported in 1957(5Landing B.H. Shirkey H.S. Pediatrics. 1957; 20: 431-438PubMed Google Scholar, 6Berendes H. Bridges R.A. Good R.A. Minn. Med. 1957; 40: 309-312PubMed Google Scholar), and is known as chronic granulomatous disease(3Smith R.M. Curnutte J.T. Blood. 1991; 77: 673-686Crossref PubMed Google Scholar, 4Curnutte J.T. Clin. Immunol. Immunopathol. 1993; 67: S2-S15Crossref PubMed Scopus (153) Google Scholar, 7Dinauer M.C. Orkin S.H. Brown R. Jesaitis A.J. Parkos C.A. Nature. 1987; 327: 717-720Crossref PubMed Scopus (260) Google Scholar). Cytochrome b (also known as flavocytochrome b, cytochrome b588, cytochrome b559, and cytochrome b−245) is the central redox component of the phagocyte NADPH-oxidase system of human neutrophils. This component is a heterodimeric integral membrane protein composed of 91-kDa (gp91phox) 1The abbreviations used are: phoxphagocyte oxidasemAbmonoclonal antibodyFACSfluorescence-activated cell scannerFITCfluorescein isothiocyanateBSAbovine serum albuminPBSphosphate-buffered salineTBSTris-buffered salinePAGEpolyacrylamide gel electrophoresis. 1The abbreviations used are: phoxphagocyte oxidasemAbmonoclonal antibodyFACSfluorescence-activated cell scannerFITCfluorescein isothiocyanateBSAbovine serum albuminPBSphosphate-buffered salineTBSTris-buffered salinePAGEpolyacrylamide gel electrophoresis.1 and 22-kDa (p22phox) subunits(8Parkos C.A. Allen R.A. Cochrane C.G. Jesaitis A.J. J. Clin. Invest. 1987; 80: 732-742Crossref PubMed Scopus (311) Google Scholar, 9Parkos C.A. Quinn M.T. Sheets S. Jesaitis A.J. Molecular Basis of Oxidative Damage by Leukocytes. CRC Press Inc., Boca Raton, FL1992Google Scholar). At least two heme groups are coordinated by these subunits(10Quinn M.T. Mullen M.L. Jesaitis A.J. J. Biol. Chem. 1992; 267: 7303-7309Abstract Full Text PDF PubMed Google Scholar), and FAD and NADPH binding activities have been demonstrated(11Koshkin V. Pick E. FEBS Lett. 1994; 338: 285-289Crossref PubMed Google Scholar, 12Rotrosen D. Yeung C.L. Leto T.L. Malech H.L. Kwong C.H. Science. 1992; 256: 1459-1462Crossref PubMed Scopus (313) Google Scholar, 13Segal A.W. West I. Wientjes F. Nugent J.H.A. Chavan A.J. Haley B. Garcia R.C. Rosen H. Scrase G. Biochem. J. 1992; 284: 781-788Crossref PubMed Scopus (289) Google Scholar). The primary structure of gp91phox includes two asparagine-linked glycosylation sites (9Parkos C.A. Quinn M.T. Sheets S. Jesaitis A.J. Molecular Basis of Oxidative Damage by Leukocytes. CRC Press Inc., Boca Raton, FL1992Google Scholar) and contains five possible transmembrane regions suggested by hydropathy analysis(14Parkos C.A. Dinauer M.C. Walker L.E. Allen R.A. Jesaitis A.J. Orkin S.H. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 3319-3323Crossref PubMed Scopus (242) Google Scholar, 15Royer-Pokora B. Kunkel L.M. Monaco A.P. Goff S.C. Newburger R.L. Baehner R.L. Cole F.S. Curnutte J.T. Orkin S.H. Nature. 1986; 322: 32-38Crossref PubMed Scopus (586) Google Scholar). p22phox contains three possible transmembrane regions, one of which includes a His-94 residue, conserved between species, that probably coordinates one of the cytochrome b heme irons. phagocyte oxidase monoclonal antibody fluorescence-activated cell scanner fluorescein isothiocyanate bovine serum albumin phosphate-buffered saline Tris-buffered saline polyacrylamide gel electrophoresis. phagocyte oxidase monoclonal antibody fluorescence-activated cell scanner fluorescein isothiocyanate bovine serum albumin phosphate-buffered saline Tris-buffered saline polyacrylamide gel electrophoresis. A number of studies have provided information about cytochrome b native structure. Electron microscopy and immunochemical analysis were used to localize cytochrome b in the neutrophil and eosinophil(16Jesaitis A.J. Buescher E.S. Harrison D. Quinn M.T. Parkos C.A. Livesey S. Linner J. J. Clin. Invest. 1990; 85: 821-835Crossref PubMed Scopus (104) Google Scholar, 17Ginsel L.A. Onderwater J.J.M. Fransen J.A.M. Verhoeven A.J. Roos D. Blood. 1990; 76: 2105-2116Crossref PubMed Google Scholar). In one of these studies, we found that the epitopes of cytochrome b containing the amino acid residues 547KQSISNSESGPRG559 of gp91phox and 162EARKKPSEEEAAA174 of p22phox are surface-accessible epitopes of native cytochrome b(16Jesaitis A.J. Buescher E.S. Harrison D. Quinn M.T. Parkos C.A. Livesey S. Linner J. J. Clin. Invest. 1990; 85: 821-835Crossref PubMed Scopus (104) Google Scholar). Rotrosen et al. (18Rotrosen D. Kleinberg M.E. Nunoi H. Leto T. Gallin J.I. Malech H.L. J. Biol. Chem. 1990; 265: 8745-8750Abstract Full Text PDF PubMed Google Scholar) found that synthetic peptides corresponding to the carboxyl terminus of gp91phoxinhibited NADPH-oxidase activation in electrically permeabilized cells, and antipeptide antibodies directed against this region prevented superoxide formation in a cell-free system. In addition, p22phox contains a proline-rich region in the carboxyl-terminal tail, which may provide Src homology domain binding sites, for p47phox or p47phox/p67phox complexes(19De Mendez I. Garrett M.C. Adams A.G. Leto T.L. J. Biol. Chem. 1994; 269: 16326-16332Abstract Full Text PDF PubMed Google Scholar). These data suggest functional roles for the carboxyl termini of both subunits, which are presumed to occupy cytosolic locations. Initial analysis of the folding topology of cytochrome b has been reported by Imajoh-Ohmi et al.(20Imajoh-Ohmi S. Tokita K. Ochiai H. Nakamura M. Kanegasaki S. J. Biol. Chem. 1992; 267: 180-184Abstract Full Text PDF PubMed Google Scholar), who determined accessibility of the subunits to anti-peptide antibodies and proteolytic enzymes. Two regions of gp91phox were exposed to proteolytic enzymes on the outer surface of the cell, while p22phox was not found to be sensitive to such external proteolysis(10Quinn M.T. Mullen M.L. Jesaitis A.J. J. Biol. Chem. 1992; 267: 7303-7309Abstract Full Text PDF PubMed Google Scholar). The carboxyl termini of both subunits were accessible to antibody on the internal surface of the plasma membrane(20Imajoh-Ohmi S. Tokita K. Ochiai H. Nakamura M. Kanegasaki S. J. Biol. Chem. 1992; 267: 180-184Abstract Full Text PDF PubMed Google Scholar). The protein sequence of the carboxyl-terminal half of gp91phoxshows some similarity to other NAD(P)H-oxidoreductases (12Rotrosen D. Yeung C.L. Leto T.L. Malech H.L. Kwong C.H. Science. 1992; 256: 1459-1462Crossref PubMed Scopus (313) Google Scholar, 21Taylor W.R. Jones D.T. Segal A.W. Protein Sci. 1993; 2: 1675-1685Crossref PubMed Scopus (108) Google Scholar) such as ferredoxin NADP reductase, a flavoprotein for which the crystal structure is known. Studies by Pick and co-workers (22Koshkin V. Pick E. FEBS Lett. 1994; 338: 285-289Crossref PubMed Scopus (65) Google Scholar) have shown that cytochrome b binds FAD and can function as a superoxide generating NAD(P)H-oxidase, even without added cytosolic constituents normally required for superoxide production in other cell-free systems (23, 24). These results have prompted speculation that cytochrome b is the only electron transporting component of the NADPH oxidase and that its nucleotide binding domains may resemble ferredoxin reductase (12Rotrosen D. Yeung C.L. Leto T.L. Malech H.L. Kwong C.H. Science. 1992; 256: 1459-1462Crossref PubMed Scopus (313) Google Scholar, 13Segal A.W. West I. Wientjes F. Nugent J.H.A. Chavan A.J. Haley B. Garcia R.C. Rosen H. Scrase G. Biochem. J. 1992; 284: 781-788Crossref PubMed Scopus (289) Google Scholar) or other redox proteins. However, the structural studies by Imajoh-Ohmi et al. suggest that major portions of the putative nucleotide binding domains are extracellular. This contention is also supported by the evidence of Umei et al. (47-49) suggesting that the putative NAD(P)H binding component of the oxidase is present in neutrophils of normal and chronic granulomatous disease patients and is thus not part of cytochrome b. These ambiguities in structure indicate that additional approaches for determining the topology of this protein are required. In this study, we have identified the epitopes bound by two monoclonal antibodies that recognize a specific subunit of cytochrome b using random peptide phage-display libraries. In addition, we present data relating the accessibility of these epitopes on native cytochrome b to their respective antibodies. Diisopropyl fluorophosphate, Tween 20, SDS, acrylamide, bisacrylamide, ammonium persulfate, TEMED, Hank's solution, FITC-labeled goat anti-mouse IgG, bovine serum albumin (BSA), and Histopaque were purchased from Sigma. Prestained protein molecular weight standards were purchased from Life Technologies, Inc. Western blots were developed with anti-mouse or anti-rabbit immunoglobulin purchased from Bio-Rad, and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium chromogen was purchased from Kirkegaard and Perry Laboratories, Inc. (Gaithersburg, MD). Cyanogen bromide-activated Sepharose CL-4B was purchased from Pharmacia Biotech Inc. Sequencing data were obtained using a Sequenase version 2.0 sequencing kit purchased from U. S. Biochemical Corp. Two random phage-display libraries were used in this study. A hexapeptide phage-display library and Escherichia coli strains K91 and MC1061 were kindly provided by Dr. George P. Smith (25Scott J. Smith G. Science. 1990; 249: 386-390Crossref PubMed Scopus (1879) Google Scholar) (University of Missouri, Columbia, MO), and a nonapeptide phage-display library was produced in our laboratory. 2J. B. Burritt, F. R. DeLeo, K. W. Doss, M. T. Quinn, C. W. Bond, and A. J. Jesaitis, unpublished data. Affinity purification of phage bearing epitopes bound by mAbs was performed as follows; 5 × 1011 phage (5 μl) from the hexapeptide phage peptide-display library (25Scott J. Smith G. Science. 1990; 249: 386-390Crossref PubMed Scopus (1879) Google Scholar) or 1 × 1012 phage (75 μl) from the nonapeptide library were combined with 1.0 ml of Sepharose beads conjugated with 4 mg of either mAb 44.1 or 54.1. The beads were mixed with the phage at 4°C for 16 h by gentle inversion. The mixture was then loaded into a 5-ml plastic column barrel (Evergreen), and unbound phage were removed by washing with 50 ml of phage buffer (50 mM Tris-Cl, pH 7.5, 150 mM NaCl, 0.5% Tween 20 (v/v), 1 mg/ml BSA). Bound phage were eluted from the column with 2.0 ml of eluting buffer (0.1 M glycine, pH 2.2), and the pH of the eluate was neutralized immediately with four drops of 2 M Trizma base(27Smith G.P. Scott J.K. Methods Enzymol. 1993; 217: 228-257Crossref PubMed Scopus (691) Google Scholar). The titer of phage (nonapeptide library) was determined for each column eluate by plaque assay according to standard procedures(28Hackett P. Fuchs J. Messing J. An Introduction to Recombinant DNA Techniques. Addison Wesley, Redding, MA1988Google Scholar). The column matrices were preserved for reuse in second and third round affinity purifications by washing with 10 ml of PBS, pH 7.0, followed by 3.0 ml of PBS containing 0.02% sodium azide. The column was stored at 4°C until the next affinity purification and was prepared for reuse by rinsing with 20 ml of phage buffer prior to mixing with amplified phage. As a control for antibody-specific selection, one column was prepared without antibody bound to the beads; all steps of affinity purification of phage were carried out on this control column, and a sample of the resulting phage was sequenced. Eluate phage were amplified in host K91 "starved" E. coli cells, which were prepared as described previously (27Smith G.P. Scott J.K. Methods Enzymol. 1993; 217: 228-257Crossref PubMed Scopus (691) Google Scholar, 29Tzagloff H. Pratt D. Virology. 1964; 24: 372-380Crossref PubMed Scopus (43) Google Scholar) to maximize phage attachment and infection. The entire volume of the first eluate, minus a small amount used for titering, was added to 200 μl of starved K91 cells and incubated 15 min at room temperature without shaking. Two ml of LB broth (30Miller J.H. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1972Google Scholar) with tetracycline at 0.2 μg/ml or kanamycin at 0.75 μg/ml (depending on the library used) was added to the cells, which were then incubated with aeration at 37°C for 45 min. The infected cells were spread with a sterile glass rod on the surface of LB agar containing tetracycline (40 μg/ml) or kanamycin (75 μg/ml) in a sterile 9 × 24-inch glass baking dish. After 24-48 h of incubation at 37°C, the antibiotic-resistant colonies were suspended in 25 ml of TBS (50 mM Tris-HCl, pH 7.5, 150 mM NaCl) by gentle scraping with a bent glass rod. Phage were harvested from the culture as described(27Smith G.P. Scott J.K. Methods Enzymol. 1993; 217: 228-257Crossref PubMed Scopus (691) Google Scholar), and diluted in 1.0 ml of phage buffer and allowed to mix with the column matrix, which was also resuspended in 1 ml of phage buffer. These steps were repeated so that a total of three purifications and two amplifications were used to select and amplify adherent phage from the library. One hundred μl of phage from the final column eluate were used to infect starved K91 cells as described above. Serial 100-fold dilutions of infected cells were used to inoculate LB agar plates containing the appropriate dilution, which were then incubated overnight at 37°C. Isolated colonies were used to inoculate 2 ml of 2 × YT (30Miller J.H. Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1972Google Scholar) containing the appropriate antibiotic, and the minipreps were incubated overnight at 37°C in a shaking water bath. DNA from the isolated phage was prepared and sequenced according to the directions of the Sequenase version 2.0 kit (U. S. Biochemical Corp.). An oligonucleotide primer with the sequence 5′-GTT TTG TCG TCT TTC CAG ACG-3′ was used to determine the nucleotide sequence of the unique region of the phage, and autoradiographs were viewed using a Molecular Dynamics model 400E PhosphorImager. Some phage bearing hexapeptides of interest (Fig. 2, sequences 3 and 27-29) were propagated by infecting 750 μl of mid-log phase K91 cells with 20 μl of phage supernatant saved from sequencing minipreps. The phage were produced in a method similar to amplifications described above, except the infected cells were grown in 100 ml of 2 × YT with 40 μg/ml tetracycline, and the purified phage were resuspended in 400 μl of TBS. Human neutrophils were purified from citrated blood using Histopaque gradients as described by Boyum(31Boyum A. J. Clin. Invest. 1968; 21: 77-89Google Scholar). Purified cells were incubated on ice for 15 min with 2 mM diisopropyl fluorophosphate to inhibit serine proteases. 5 × 105 cells were used for each sample to determine antibody binding by FACS analysis. Some cell samples were permeabilized on ice for 10 min by adding 500 μl of saponin solution (0.01% saponin, 0.1% gelatin in Dulbecco's PBS) and pelleted by centrifugation as before. Permeabilized cells were incubated on ice for 30 min with 80 μl of the primary antibody (usually 50 μg/ml in saponin solution), then washed once with 3.0 ml of saponin solution, centrifuged to collect, and resuspended in 80 μl of the FITC-conjugated goat anti-mouse antibody diluted 1:150 in saponin solution, and incubated on ice for 30 min. Permeabilized cells were washed once with 3.0 ml of saponin solution containing propidium iodide at 10 μg/ml, pelleted as before, and resuspended in 500 μl of FACS buffer (Dulbecco's PBS containing 10% rabbit serum). In separate experiments, 100 μl of mAb 44.1 at 10 μg/ml was incubated at 37°C for 30 min with ∼2 × 1010 transducing units (27) phage expressing phage sequence 3 or 27 (Fig. 2) prior to incubation with the saponin-permeabilized cells to determine if the phage-expressed peptide could compete with the natural epitope for binding by the mAb. Control samples in all experiments consisted of cells not incubated with either primary or secondary mAb, cells not incubated with primary mAb, and cells incubated with both an isotype-matched primary mAb and the labeled secondary mAb. Staining of some samples of intact cells was performed as above without treatment with saponin solution. Fluorescence intensity of the FITC-labeled cells was determined on a Becton Dickinson FACScan model FACS analyzer with a 15-milliwatt argon-ion laser using CONSORT 30 and LYSYS software according to the manufacturer's directions. Immunoaffinity-purified phage bearing peptides resembling a region of cytochrome b were isolated and grown as indicated above. Approximately 5 × 1011 purified phage in 20 μl of TBS were heated in a boiling water bath for 5 min with an equal volume of SDS sample buffer (3.3% (w/v) SDS, 167 mM Tris-Cl, pH 6.8, 33% (v/v) glycerol, 0.03% (w/v) bromphenol blue, 0.035% (v/v) 2-mercaptoethanol). Heparin-Ultrogel-purified cytochrome b, prepared as described(8Parkos C.A. Allen R.A. Cochrane C.G. Jesaitis A.J. J. Clin. Invest. 1987; 80: 732-742Crossref PubMed Scopus (311) Google Scholar), was combined with an equal volume of SDS sample buffer without heating. Protein samples were separated by SDS-PAGE at room temperature on 12% (w/v) polyacrylamide gels as described(8Parkos C.A. Allen R.A. Cochrane C.G. Jesaitis A.J. J. Clin. Invest. 1987; 80: 732-742Crossref PubMed Scopus (311) Google Scholar, 10Quinn M.T. Mullen M.L. Jesaitis A.J. J. Biol. Chem. 1992; 267: 7303-7309Abstract Full Text PDF PubMed Google Scholar), and electrophoretic mobility of sample proteins were compared to prestained protein standards. Following electrophoresis, protein samples were transferred to nitrocellulose as described previously(8Parkos C.A. Allen R.A. Cochrane C.G. Jesaitis A.J. J. Clin. Invest. 1987; 80: 732-742Crossref PubMed Scopus (311) Google Scholar). Monoclonal antibodies 44.1 and 54.1 were diluted to 2 μg/ml in diluting buffer (3% (v/v) goat serum, 1% (w/v) BSA, 0.2% (v/v) Tween 20, 0.1% (w/v) thimerosal in PBS) and incubated with separate regions of the blot for 1 h at room temperature with continuous rocking. The blot was developed using alkaline phosphatase-conjugated goat anti-mouse IgG and a 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium phosphatase substrate system. Purified cytochrome b (10 μg) was diluted in 100 μl of relax buffer (8Parkos C.A. Allen R.A. Cochrane C.G. Jesaitis A.J. J. Clin. Invest. 1987; 80: 732-742Crossref PubMed Scopus (311) Google Scholar) containing 50 μg of mAb (44.1, 54.1, irrelevant, or none) and then incubated overnight at 4°C. The pretreated cytochrome b was then loaded onto a 1.36-ml continuous 5-20% sucrose gradient in relax buffer and centrifuged at 53,000 RPM in a Beckman TLS-55 rotor for seven hours at 4°C. The gradient was manually fractionated into 12 120-μl samples from the top. A 20-μl sample of each fraction was separated by SDS-PAGE and transferred to nitrocellulose membranes as described above. Cytochrome b was detected by Western blot using a rabbit anti-p22phox polyclonal primary antibody. Western blots were digitized and quantitated using an image analysis system as described(32Quinn M.T. Evans T. Loetterle L.R. Jesaitis A.J. Bokoch G.M. J. Biol. Chem. 1993; 268: 20983-20987Abstract Full Text PDF PubMed Google Scholar). To determine the structure of immunogenic surface regions of cytochrome b and acquire additional information about its membrane topology, monoclonal antibodies were produced against the Triton X-100-solubilized wheat germ agglutinin and heparin-Ultrogel-purified cytochrome b protein(8Parkos C.A. Allen R.A. Cochrane C.G. Jesaitis A.J. J. Clin. Invest. 1987; 80: 732-742Crossref PubMed Scopus (311) Google Scholar). Hybridoma supernatants were screened for neutrophil-specific IgG-producing clones that recognized intact or saponin-permeabilized neutrophils and either subunit of heparin-purified cytochrome b. Numerous clones were identified, and two were chosen that recognized either the light or heavy chain of cytochrome b. Epitope mapping of cytochrome b was carried out using these cytochrome b-specific antibodies and phage random peptide-display library technology(25Scott J. Smith G. Science. 1990; 249: 386-390Crossref PubMed Scopus (1879) Google Scholar, 27Smith G.P. Scott J.K. Methods Enzymol. 1993; 217: 228-257Crossref PubMed Scopus (691) Google Scholar, 33Cwirla S. Peters E. Barrett R. Dower W. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 6378-6382Crossref PubMed Scopus (851) Google Scholar, 34Devlin J. Panganiban L. Devlin P. Science. 1990; 249: 404-406Crossref PubMed Scopus (817) Google Scholar). FACS and immunosedimentation analysis were then used to confirm accessibility of the epitope on native cytochrome b to antibody, and placement of the epitope relative to the plasma membrane. Finally, confirmation of the specificity of the mAb for both the cytochrome b subunit and the phage peptide-bearing pIII protein selected by the mAb was demonstrated by Western blot analysis. To identify the amino acid sequence of the epitopes of cytochrome b recognized by the reactive mAbs, a nonapeptide phage-display library capable of binding to the mAbs was created. Using this library, mAb-binding epitopes were selected from a collection of 5 × 108 unique nine-residue sequences of all 20 amino acids. The epitopes thus mimic the original immunogenic cytochrome b epitope. By sequencing the relevant region of the phage genome, the original cytochrome b epitope was deduced(35Burritt J.B. Quinn M.T. Jutila M.A. Bond C.W. Doss K.W. Jesaitis A.J. Mol. Biol. Cell. 1994; 5 (abstr.): 121aGoogle Scholar). Confirmation of the epitope selection was achieved using a second epitope library kindly provided by George P. Smith at the University of Missouri, Columbia(27Smith G.P. Scott J.K. Methods Enzymol. 1993; 217: 228-257Crossref PubMed Scopus (691) Google Scholar). Three cycles of immunoaffinity purification and amplification were used to select phage expressing peptides bound by either mAb 44.1 or 54.1. Fig. 1 shows selection and amplification of phage bound by mAb columns. An increase of about 5 logs was observed for the population of binding phage for each mAb, and about 3 logs for phage that interacted with the control column without mAb. The random insert region was sequenced in 34 phage selected by mAb 44.1 from the nonapeptide library and 27 from the hexapeptide library as described under "Materials and Methods."Fig. 2 shows 33 of the 37 sequences (89%) from the nonapeptide library, which exhibited an obvious consensus pattern matching a region of cytochrome b shown in Fig. 3. This consensus peptide sequence, GGPQVXPI, closely resembles 181GGPQVNPI188 of p22phox (Fig. 3A). The remaining four sequences did not resemble any region of cytochrome b or each other (data not shown). Nineteen of the 27 sequences (70%) from the hexapeptide library showed similarity to the same cytochrome b epitope, and the same nucleotide sequence coding for the PQVRPI peptide was recovered in 17 of the 19 cases. The remaining eight hexapeptide sequences showed no consensus (data not shown). Phage expressing the hexapeptide sequences PQVRPI, FKRGVD, LRRGID, and PKGAYD (Fig. 2; sequences 3 and 27-29, respectively) were isolated and propagated for further study by Western blot and FACS analysis. Phage selected by mAb 54.1 expressed the consensus amino acid sequence PKXAVDGP (the GP is adjacent in the constant pIII region in all phage except phage sequence 34), which is similar to 382PKIAVDGP389 of gp91phox (Fig. 3B). All phage sequenced that were selected from each library by mAb 54.1 suggested a match to this putative gp91phoxepitope. Peptides of phage selected on the column without antibody suggested no match to cytochrome b, or to each other (data not shown). Western blotting analysis was used to show specificity of the mAb for both the cytochrome b subunit and the unique hexapeptide expressed on the phage. As shown in Fig. 4, mAb 54.1 specifically recognized gp91phox (lane A), which migrates between the 68- and 97-kDa molecular size markers, and a 20-kDa proteolytic fragment. The appearance of this 20-kDa immunoreactive fragment can be enhanced in cytochrome b samples treated with V8 protease (10). This mAb also bound to the phage pIII protein expressing the FKRGVD peptide (Fig. 2, sequence 27), which migrates at about 64 kDa (lane C). mAb 54.1 also recognized pIII proteins of phage expressing the LRRGID and PKGAYD peptides (Fig. 2, sequences 28 and 29, respectively) by Western blot with equal intensity (data not shown). Phage expressing the PQVRPI peptide (Fig. 2, sequence 3) was not recognized by mAb 54.1 (lane B), confirming the specificity of mAb 54.1 for the former sequences. As seen in Fig. 4, mAb 44.1 bound to a band migrating at 22 kDa in lane D (p22phox) and a less intense band at 44 kDa (subunit dimer). The pIII protein of phage containing the PQVRPI peptide (Fig. 2, sequence 3) was also recognized by mAb 44.1 (lane F), but not the pIII protein of phage expressing the FKRGVD peptide (Fig. 2, sequence 27) in lane E. The pIII protein migrates at an apparent molecular mass of 64 kDa. This mobility appears slow, considering the size of the protein is 406 amino acids(36Ebright R. Dong Q. Messing J. Gene (Amst.). 1992; 114: 81-83Crossref PubMed Scopus (12) Google Scholar); however, our migration rate compares favorably with other reported PAGE mobilities for this protein (37-40). Binding of mAb 44.1 and 54.1 to both the specific cytochrome b subunit and phage pIII protein confirmed the presence of a similar epitope on each. In addition, mAbs did not bind the pIII protein on phage displaying irrelevant peptides in the random region. Therefore, the sequence of the peptide alone, expressed in the variable region simulates the natural epitope recognized by the mAb. Although the variable peptide of the selected phage may not represent the complete and exact epitope, it clearly contains the residues sufficient for recognition by the mAb. In order to determine the accessibility of the identified epitopes to mAb on native cytochrome b in the cell, FACS analysis was performed on purified leukocytes. While saponin-permeabilized cells stained strongly with mAb 44.1 at 50 μg/ml (Fig. 5, trace B), intact cells show background staining when incubated with either mAb 44.1 (Fig. 5, trace A) or an irrelevant mAb (data not shown). A similar background le
Referência(s)