FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2061873103> ?p ?o ?g. }

• W2061873103 endingPage "30387" @default.
• W2061873103 startingPage "30382" @default.
• W2061873103 abstract "We have established a novel method, in situ phage screening (ISPS), to identify proteins in tissue microstructures. The method is based on the selection of repertoires of phage-displayed antibody fragments with small samples of tissues microdissected using a laser. Using a human muscle frozen section with an area of 4800 μm2 as a model target, we successfully selected monoclonal antibody fragments directed against three major (myosin heavy chain, actin, and tropomyosin-α) and one minor (α-actinin 2) muscle constituent proteins. These proteins were present in the sample in amounts less than one nanogram, and the antibodies were used to visualize the proteins in situ. This shows that the use of ISPS can obtain monoclonal antibodies for histochemical and biochemical purposes against minute amounts of proteins from microstructures with no requirement for large amounts of samples or biochemical efforts. We have established a novel method, in situ phage screening (ISPS), to identify proteins in tissue microstructures. The method is based on the selection of repertoires of phage-displayed antibody fragments with small samples of tissues microdissected using a laser. Using a human muscle frozen section with an area of 4800 μm2 as a model target, we successfully selected monoclonal antibody fragments directed against three major (myosin heavy chain, actin, and tropomyosin-α) and one minor (α-actinin 2) muscle constituent proteins. These proteins were present in the sample in amounts less than one nanogram, and the antibodies were used to visualize the proteins in situ. This shows that the use of ISPS can obtain monoclonal antibodies for histochemical and biochemical purposes against minute amounts of proteins from microstructures with no requirement for large amounts of samples or biochemical efforts. in situ phage screening variable fragment of immunoglobulin single chain Fv heavy chain variable region light chain variable region colony forming units glial fibrillary acidic protein National Center of Neurology and Psychiatry horseradish peroxidase 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid immobilized pH gradient matrix-assisted laser desorption ionization time-of-flight mass spectrometry In the post-genomic era, nanoscale technologies such as femtomole-ranged mass spectrometry (1Wilm M. Shevchenko A. Houthaeve T. Breit S. Schweigerer L. Fotsis T. Mann M. Nature. 1996; 379: 466-469Crossref PubMed Scopus (1508) Google Scholar) have become indispensable for proteomic analyses to clarify changes in protein components and abnormality of protein functions in human diseases. Here, we describe a novel nanoscale method, in situ phage screening (ISPS),1 which enables us to identify protein components in microstructures seen with microscopy. The method involves the combination of phage antibody display technology (2MacCafferty J. Griffiths A.D. Winter G. Chiswell D.J. Nature. 1990; 348: 552-554Crossref PubMed Scopus (1947) Google Scholar, 3Nissim A. Hoogenboom H.R Tomlinson I.M. Flynn G. Midgley C. Lane D. Winter G. EMBO J. 1994; 13: 692-698Crossref PubMed Scopus (528) Google Scholar, 4Winter G. Griffiths A.D. Hawkins R.E. Hoogenboom H.R. Annu. Rev. Immunol. 1994; 12: 433-455Crossref PubMed Scopus (1393) Google Scholar, 5Hoogenboom H.R. de Bruine A.P. Hufton S.E. Hoet R.M. Arends J.W. Roovers R.C. Immunotechnology. 1998; 4: 1-20Crossref PubMed Scopus (367) Google Scholar) with laser microdissection technology (6Lehmann U. Bock O. Glockner S. Kreipe H. Pathobiology. 2000; 68: 202-208Crossref PubMed Scopus (38) Google Scholar, 7Simone N.L. Paweletz C.P. Charboneau L. Petricoin 3rd, E.F. Liotta L.A. Mol. Diagn. 2000; 5: 301-307Crossref PubMed Google Scholar). For these purposes, we used a library of single chain variable fragments (scFv) of antibodies (2MacCafferty J. Griffiths A.D. Winter G. Chiswell D.J. Nature. 1990; 348: 552-554Crossref PubMed Scopus (1947) Google Scholar, 3Nissim A. Hoogenboom H.R Tomlinson I.M. Flynn G. Midgley C. Lane D. Winter G. EMBO J. 1994; 13: 692-698Crossref PubMed Scopus (528) Google Scholar) fused to gene III of M13 phage and expressed at the tip of the phage particles (3Nissim A. Hoogenboom H.R Tomlinson I.M. Flynn G. Midgley C. Lane D. Winter G. EMBO J. 1994; 13: 692-698Crossref PubMed Scopus (528) Google Scholar, 8Griffiths A.D. Williams S.C. Hartley O. Tomlinson I.M. Waterhouse P. Crosby W.L. Kontermann R.E. Jones P.T. Low N.M. Allison T.J. Prospero T.D. Hoogenboom H.R. Nissim A. Cox J.P.L. Harrison J.L. Zaccolo M. Gherardi E. Winter G. EMBO J. 1994; 13: 3245-3260Crossref PubMed Scopus (879) Google Scholar). For microdissection of tissues, we developed a new laser-microdissector by combining an industrial-use laser cutter and an inverted microscope. The procedure involves isolation of the target microstructures from the surrounding tissues on glass by dissection and then incubation with the phage antibody library. Phage antibodies were isolated after a single round of selection, and their specificities were identified by immunohistochemistry and Western blotting. Target antigens were identified by immunoscreening of cDNA expression libraries using the monoclonal phage antibodies. The clinical targets of ISPS are those characterized by abnormal or yet unknown protein accumulations, which closely relate with pathogenetic mechanisms, such as subcutaneous deposits or inclusion bodies and extracellular plaques often seen in neuromuscular and other degenerative diseases. ISPS is highly advantageous when the target microstructures can hardly be collected by standard biochemical methodologies; for instance, targets are very small (up to 1 μm in size) or rare (seen in few cells), or they appear in complex pathophysiological structures. Because muscle cells consist of relatively small numbers of proteins, we employed human skeletal muscle as a model target of ISPS. In this report, we describe model experiments of ISPS on human muscle, where we successfully identified the three major and one minor muscle antigens from an area of 4800-μm2 microfragments. A human synthetic scFv library “Griffin. 1” (9Goletz S. Christensen P.A. Kristensen P. Blohm D. Tomlinson I. Winter G. Karsten U. J. Mol. Biol. 2002; 315: 1087-1097Crossref PubMed Scopus (108) Google Scholar) was provided from Dr. G. Winter (Medical Research Council, UK) and amplified in our laboratory. In this library, the VH and VL sequence connected with a spacer sequence (3Nissim A. Hoogenboom H.R Tomlinson I.M. Flynn G. Midgley C. Lane D. Winter G. EMBO J. 1994; 13: 692-698Crossref PubMed Scopus (528) Google Scholar, 8Griffiths A.D. Williams S.C. Hartley O. Tomlinson I.M. Waterhouse P. Crosby W.L. Kontermann R.E. Jones P.T. Low N.M. Allison T.J. Prospero T.D. Hoogenboom H.R. Nissim A. Cox J.P.L. Harrison J.L. Zaccolo M. Gherardi E. Winter G. EMBO J. 1994; 13: 3245-3260Crossref PubMed Scopus (879) Google Scholar) was inserted into anNcoI-NotI restriction site and followed by the M13 gene III sequence in phagemid vector pHEN2. Recombinant scFv phages were rescued in Escherichia coli suppressor strain TG1 in the presence of helper phage VCSM13 (Stratagene Cloning System, La Jolla, CA) as described previously (9Goletz S. Christensen P.A. Kristensen P. Blohm D. Tomlinson I. Winter G. Karsten U. J. Mol. Biol. 2002; 315: 1087-1097Crossref PubMed Scopus (108) Google Scholar,10Coomber D.W.J. Methods Mol. Biol. 2002; 178: 133-145PubMed Google Scholar). 2Details of the Griffin. 1 library and phage protocols are also given on the Winter group home page (www.mrc-cpe.cam.ac.uk/phage/index.html). Briefly, bacteria carrying scFv phagemid were cultured in a 2× YT medium (11Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Labolatory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar) containing 100 μg/ml ampicillin and 1% glucose and infected in the log phase (absorbance at 600 nm (A600) was 0.5) with VCSM13 at a ratio of 1:20 (number of bacterial cells/phage particles) for 30 min at 37 °C without shaking. The bacteria were pelleted by centrifugation, resuspended in a 2× YT medium containing 100 μg/ml ampicillin and 25 μg/ml kanamycin (4 volumes of the original), and grown overnight at 30 °C. Phages were purified from the culture supernatant by polyethylene glycol precipitation (2MacCafferty J. Griffiths A.D. Winter G. Chiswell D.J. Nature. 1990; 348: 552-554Crossref PubMed Scopus (1947) Google Scholar) and resuspended in TBS (20 mm Tris-Cl, pH 7.6, 0.15m NaCl). The working solution of the phage library was prepared at the concentration of 1013 colony forming units (cfu)/ml. An anti-thyroglobulin scFv phage was also provided by Dr. G. Winter. Histologically normal biopsy specimens of skeletal muscle (biceps brachii) were provided by the NCNP research resource network (Kodaira, Tokyo) after obtaining the informed consent of the patients. Transverse (experiments A and B) and longitudinal (experiments C through E) frozen sections that were 2-μm thick were cut with a cryostat, attached to silanized slides (Dako), and subjected to laser microdissection. Spaces 15 μm wide surrounding 40 × 40-μm square microfragments were made with pulsed shots of the ultraviolet laser. After the sections were immersed in acetone for 5 min and air-dried, they were blocked with TBS-10% (weight/volume) skim milk for 2 h and then incubated with the phage library (5 × 1011 cfu in 50 μl of TBS in experiments A and B or in 75 μl of TBS-10% skim milk in experiments C through E) for 12 h at room temperature. After washing with TBS-0.05% Tween 20 (TBS-T) four times ×30 min, the sections were overlaid with 0.2 ml of TBS-T. Three microfragments of 40 × 40 μm were then collected into a microcentrifuge tube with a hand-made micropipette (inner diameter, 60–80 μm) together with a minimal amount of overlaid TBS-T, and the microfragments were washed three times by brief centrifugation with 0.2 ml of TBS-T each. After washing, the contents of the tube were examined microscopically to determine whether the three microfragments remained in the tube. In the direct colony recovery protocol, phages were eluted from the pelleted fragments with 100 μl of 1.4% triethylamine for 10 min at room temperature with vigorous shaking followed by neutralization with 50 μl of 1 m Tris-Cl, pH 7.4. A TG1 culture (1 ml) in the log phase (A600 = 0.5) was infected with the eluted phage solution for 30 min at 37 °C and plated on LB-agar (11Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Labolatory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar) containing 100 μg/ml ampicillin and 1% glucose. All of the ampicillin-resistant colonies obtained, which were carrying scFv phagemid, were subjected to the analysis of specificity. In the PCR fragment-subcloning protocol, the three pelleted microfragments were suspended in 10 μl of water, heated at 95 °C for 5 min, and then rapidly cooled in ice. The Fv sequences derived from the phages bound to the microfragments were amplified by PCR using forward (5′-CGGATAACAATTTCACACAGGAAAC-3′) and reverse (5′-CTATGCGGCCCCATTCAGATC-3′) primers. The reaction (50 μl) contained 5 μl of a 10-fold-concentrated Taq reaction buffer (Takara, Japan), 0.8 μm each of the forward and reverse primers, 0.2 mm dNTPs (Takara, Japan), the heat-treated sample (10 μl as described above), and 2.5 units ofTaq polymerase (Takara, Japan) and was kept in ice. After heating at 98 °C for 1 min, 38 cycles of PCR (60 °C for 20 s, 72 °C for 1 min, and 95 °C for 20 s) were performed and followed by incubation at 72 °C for 3 min. The PCR product (0.95 kbp) was digested with NcoI and NotI, ligated with NcoI- and NotI-digested pHEN2, and transformed into TG1. The bacteria were plated to form ampicillin-resistant colonies. It was confirmed that no PCR products were obtained by using the final wash supernatant of the muscle microfragment (see Fig. 1). A portion of the colonies was subjected to analysis of scFv specificity. The Fv sequence in each colony was amplified by PCR (30 cycles of 95 °C for 15 s, 60 °C for 15 s, and 72 °C for 1 min) using the same primer set as described above, digested with HaeIII, and electrophoresed on 4.5% agarose for fingerprinting. The monoclonal phage was rescued from the representative clones showing unique patterns on fingerprinting and analyzed immunohistochemically and by Western blotting. For immunostaining, muscle frozen-sections preblocked with TBS-10% skim milk were incubated with one to four unique phages (5 × 1010 cfu each in 100 μl of TBS-10% skim milk) simultaneously for 12 h at room temperature. After washing twice with TBS-T for 15 min, the sections were incubated with 1000-fold-diluted anti-M13 mouse IgG (Progen Biotechnik, Heidelberg) in TBS-5% skim milk for 1 h. After washing twice as above, the sections were stained by using an Envision-HRP kit (Dako) according to the manufacturer's protocol. For Western blotting, muscle sections (about 25–50 μg as protein) were dissolved in a 2-fold-concentrated sample buffer of SDS-PAGE (11Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Labolatory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar), electrophoresed on 10% preparative gel, and transferred to a nitrocellulose membrane. The membrane was cut into strips 1–3 mm wide, blocked with 10% skim milk in TBS for 1 h, and incubated with scFv phages in TBS-10% skim milk for 2 h. In another set of experiments (with Tween treatment), the membrane strips were incubated with TBS-0.1% Tween 20 for 1 h, followed by blocking and phage incubation as described above, but in the presence of 0.1% Tween 20 (12). After washing with TBS-0.1% Tween 20 for 10 min six times, the strip was incubated with 10,000-fold-diluted anti-M13 mouse IgG-HRP conjugated (Amersham Biosciences) in TBS-0.1% Tween 20 for 1 h. After washing with TBS-0.1% Tween 20 as above, the blot was visualized with an ECL Plus Western blotting detection system (Amersham Biosciences). Muscle sections (about 30 and 300 μg as protein for Western and mass analysis, respectively) were dissolved in 250 μl of a two-dimensional gel electrophoresis buffer (20 mm Tris base, 9 m urea, 2 m thiourea, 4% CHAPS, 65 mm dithioerythritol, and 0.5% immobilized pH gradient (IPG) buffer (Amersham Biosciences)) and subjected to isoelectric focusing by using an IPGphor system (Amersham Biosciences) with a 13 cm-long IPG strip (pH range, 3–10) according to the manufacturer's protocol. The IPG strip was then applied to the top of an SDS-PAGE gel (7.5%, 14 × 12 cm). For mass fingerprinting analysis, proteins on the two-dimensional gel were visualized by using a copper-staining kit (Bio-Rad). The target protein spots were then cut out, digested in gel with trypsin, and analyzed with a MALDI-TOF (matrix-assisted laser desorption ionization-time of flight) type mass spectrometer (Voyager-DEPRO, Applied Biosystems) (1Wilm M. Shevchenko A. Houthaeve T. Breit S. Schweigerer L. Fotsis T. Mann M. Nature. 1996; 379: 466-469Crossref PubMed Scopus (1508) Google Scholar, 13Kinumi T. Tobin S.L. Matsumoto H. Jackson K.W. Ohashi M. Eur. Mass Spectrom. 1997; 3: 367-378Crossref Scopus (11) Google Scholar). The peptide mass fingerprinting analysis was performed by using the program MS-FIT. Western analysis was carried out as described above. A human skeletal muscle 5′-STRETCH PLUS cDNA library (CLONTECH, Palo Alto, CA) was immunoscreened according to the manufacturer's protocol by using scFv phages. Library plaques were transferred to nitrocellulose membranes, incubated with scFv phages (5 × 1010 cfu/ml, for 2 h), and visualized by using anti-M13 mouse IgG-HRP conjugated and ECL Plus Western blotting detection reagent (Amersham Biosciences) as described above. ISPS is performed by 1) dissection and isolation of target microstructures from surrounding tissues on a tissue section, 2) incubation of the tissue section with an scFv phage library, 3) collection of targets, and 4) selection of target-specific phages (Fig. 1). A new laser microdissector was developed to cut out the microstructures under a microscope (Fig. 2A). The dissector produced a rectangular space of variable size (1–56 μm) on a tissue section with a single shot of an ultraviolet laser beam at 266 nm (Fig. 2B). Tissues surrounding the target were widely burnt off by precisely controlling the laser power. Spaces that were 15 μm wide surrounding a 40 × 40-μm2 square microfragment were usually produced in a muscle section by the dissector (Fig.2C). As a source of scFvs, we employed a phagemid library, Griffin. 1 (see “Experimental Procedures”). The scFv sequences were constructed with synthetic human V-gene sequences of immunoglobulin, VH and VL (3Nissim A. Hoogenboom H.R Tomlinson I.M. Flynn G. Midgley C. Lane D. Winter G. EMBO J. 1994; 13: 692-698Crossref PubMed Scopus (528) Google Scholar, 8Griffiths A.D. Williams S.C. Hartley O. Tomlinson I.M. Waterhouse P. Crosby W.L. Kontermann R.E. Jones P.T. Low N.M. Allison T.J. Prospero T.D. Hoogenboom H.R. Nissim A. Cox J.P.L. Harrison J.L. Zaccolo M. Gherardi E. Winter G. EMBO J. 1994; 13: 3245-3260Crossref PubMed Scopus (879) Google Scholar). When the phages were rescued by using helper phages derived from M13 (VCSM13), the V-gene was fused with the gene III of the M13 phage and expressed as an scFv at the tip of the phage particle (2MacCafferty J. Griffiths A.D. Winter G. Chiswell D.J. Nature. 1990; 348: 552-554Crossref PubMed Scopus (1947) Google Scholar). The Griffin. 1 library contained about 109 independent unique clones (9Goletz S. Christensen P.A. Kristensen P. Blohm D. Tomlinson I. Winter G. Karsten U. J. Mol. Biol. 2002; 315: 1087-1097Crossref PubMed Scopus (108) Google Scholar). We first determined the optimal conditions for phage antibody reaction with muscle microfragments. Anti-human α-actin and anti-human glial fibrillary acidic protein (GFAP) phages (designated Actin89 and GFAP53, respectively) were obtained from the Griffin. 1 library by the conventional panning method using polystyrene tubes (10Coomber D.W.J. Methods Mol. Biol. 2002; 178: 133-145PubMed Google Scholar). Actin89 phage was used for the evaluation of specific binding to muscle microfragments, and GFAP53 phage was used for that of nonspecific binding, because GFAP is an astrocyte-specific protein. After reacting these phages with a muscle section, 3–5 microfragments (40 × 40-μm2 square) were collected and washed three times to eliminate contamination by phages from residual tissues other than the microfragments. The phages bound to the microfragments were eluted with 1.4% triethylamine and allowed to infect E. coli TG1. The phage titer was measured as the number of ampicillin-resistant colonies (cfu). Skim milk was employed as a blocking agent because of findings in studies of muscle immunostaining with phage antibody in which 10% skim milk, but not 5%, worked well as a blocking agent. We checked the blocking effect of 10% skim milk on the nonspecific binding of GFAP53 phage to muscle fragments (Table I). As seen, preincubation with 10% skim milk or its presence during phage incubation was optimal in preventing the binding of unrelated phage GFAP53 to muscle fragments. Elevated amounts of nonspecific phage binding were only seen in the absence of the blocking agent throughout the experiment. Interestingly, the binding of Actin89 phage to muscle fragments was not affected either in the presence or absence of 10% skim milk (Table I).Table IEffect of skim milk on specific or non-specific phage binding to muscle microfragmentsSkim milkPhage recoveryPreblockIncubationActin89GFAP53cfu/1000 μm2++11 ± 30.5 ± 0.2−+ND1-aND, not determined.0.6 ± 0.1+−ND0.2 ± 0.2−−9 ± 68 ± 41-a ND, not determined. Open table in a new tab We next checked the relation between input phage concentrations and amounts of phage binding to muscle fragments (Table II). As seen, the numbers of Actin89 phage bound to muscle microfragments were linearly related to the concentrations of input Actin89 phage (108–1012 cfu/ml) under optimal conditions. Similar linearity was observed when Actin89 was reacted with purified α-actin bound to an enzyme-linked immunosorbent assay plate (input 105–1012 cfu/ml, data not shown). When the input phage concentration was 1012 cfu/ml, the amount of Actin89 bound to the plate was 13,000 cfu/1000 μm2. An anti-thyroglobulin phage was examined as an unrelated phage.Table IIRelation between input phage concentration and phage recovery from muscle microfragmentsPhage inputPhage recoveryActin89GFAP53Anti-thyroglobulincfu/mlcfu/1000 μm21012850 ± 3002-an=5.0.5 ± 0.22-bn=3.1.0 ± 0.42-bn=3.101173 ± 312-an=5.ND2-cND, not determined.ND101015 ± 142-an=5.NDND1090.8 ± 0.42-bn=3.NDND10802-bn=3.NDND2-a n=5.2-b n=3.2-c ND, not determined. Open table in a new tab We also investigated the effect of the duration of incubation on the binding of Actin89 phage to a muscle microfragment. No significant difference was observed among the numbers of the phages collected after incubation for 3, 6, 12, and 60 h. The antigen retrieval effect of solvents on muscle microfragments was evaluated by quantifying the binding capacity of Actin89 to the microfragment. Among the solvents tested (acetone, methanol, ethanol, isopropanol, chloroform, sodium dodecyl sulfate (0.01 and 0.1%), Tween 20 (1%), and paraformaldehyde (2 and 4%)), acetone and ethanol pretreatment enhanced binding capacity the most, namely, three times more than in the absence of treatment. Acetone pretreatment was employed in the standard protocol for ISPS on human muscle. We produced various size microfragments on a transverse section of human muscle by laser-dissection and incubated them with the Griffin. 1 library. The bound phages were directly eluted from microfragments and used to infect TG1 (“direct colony recovery” protocol), or the scFv sequences of bound phages were amplified by PCR and subcloned into phagemid vector pHEN2 (“PCR fragment subcloning” protocol). Phages were rescued from ampicillin-resistant colonies, and after unique clones were picked up by HaeIII-fingerprinting, their specificity for muscle antigens was checked by both immunohistochemistry and Western blotting of muscle antigens. The PCR fragment subcloning protocol yielded no phage clones from an area of 100 (10 × 10) μm2 and three phage clones from 1100 (33 × 33) μm2. The three phage clones from the 1100 μm2 microfragment showed no specificity for muscle antigens, as judged by immunostaining. When the reaction area was enlarged to 4800 μm2 (3 microfragments of 40 × 40 μm), we obtained a number of unique phage clones in two independent trials both by direct colony recovery and PCR fragment subcloning protocols (Table III, experiments A and B). The PCR fragment subcloning protocol gave us larger numbers of unique clones than the direct colony recovery protocol (Table III), because less than 1 cfu of phagemid was amplified under our PCR conditions. About 8 phage particles were equivalent to 1 cfu. 3M. Furuta, T. Ito, C. Eguchi, T. Tanaka, E. Wakabayashi-Takai, and K. Kaneko, unpublished result. The two trials, A and B, yielded three positive clones, A28, B6, and B85, which specifically reacted with 37-, 100-, and 42-kDa muscle antigens, respectively (Tables III and IV and Fig.3A). In PCR fragment subcloning experiments, the collected phages (round 1) were amplified and applied to the second round of selection (round 2) as in the same manner. In experiment A, the positive clone obtained from round 1 (A28) was again recovered in round 2; however, in experiment B, the positive clone in round 1 (B85) was not found in round 2 (Table III). This suggested that repeated panning steps sometimes lost positive clones that were present in the earlier round of phage pool.Table IIIProfile of in situ phage screening on 4800 μm2 of muscle microfragmentsExperimentNumber of clonesAnalyzedUniquePositiveADirect330PCR round 172291PCR round 22481BDirect22141PCR round 172471PCR round 272230C PCR48132D PCR48111E PCR4893 Open table in a new tab Table IVFeatures of anti-muscle scFv clonesCloneWesternAntigenIHC4-aIHC, immunohistochemistry; IS, immunoscreening; ND, not determined.ISkDaA2837Tropomyosin-α++B6100α-Actinin 2++B8542α-Actin−NDC2242α-Actin−NDC43210Myosin heavy chain IIa++D242α-Actin−NDE2210Myosin heavy chain β++E642α-Actin+−E3042α-Actin−ND4-a IHC, immunohistochemistry; IS, immunoscreening; ND, not determined. Open table in a new tab We planned another series of experiments by using a longitudinal muscle section to obtain a homogeneous distribution at the reaction surface of muscle antigens which localized in muscle-specific, repeating zone structures such as the A- and I-bands. The screening was also performed on 4800-μm2 microfragments by using the Griffin. 1 library. Three independent trials (experiments C, D, and E) yielded scFv clones reactive with 210-kDa protein (C43 and E2) and 42-kDa protein (C22, D2, E6, and E30) by the PCR fragment subcloning protocol (Tables III and IV). The results of one-dimensional Western analysis using the representative phage clones (A28, B6, E2, and E6) are demonstrated in Fig. 3A. As seen, the reactivity of these scFvs was markedly enhanced by Tween treatment of the blot membranes (12Van Dam A.P. Van den Brink H.G. Smeenk R.J.T. J. Immunol. Methods. 1990; 129: 63-70Crossref PubMed Scopus (19) Google Scholar). Because these scFv phages were selected on a frozen-section mildly fixed by acetone, they may have preferentially reacted with renatured antigens on the membrane. A considerable number of phages with low specificity that reacted with multiple bands on muscle Western analysis were obtained through experiments A through E (data not shown). Two-dimensional Western blotting of muscle proteins was performed by using positive scFv phages, and their corresponding antigens were assigned on a Coomassie-stained two-dimensional gel (Fig.3, B–D). The 42-kDa protein recognized by several clones (B85, C22, D2, E6, and E30) had a pI of 5.2 and was classified as α-actin (Fig. 3B, Ac), which accounts for about 20% of the total protein in muscle (14Ohtsuki I. Maruyama K. Ebashi S. Adv. Protein Chem. 1986; 38: 1-67Crossref PubMed Scopus (204) Google Scholar). Similarly, the 210-kDa antigen had a pI of 5.8 and was classified as a myosin heavy chain (Fig. 3B, My), which weighed about 40% of the total proteins in muscle (14Ohtsuki I. Maruyama K. Ebashi S. Adv. Protein Chem. 1986; 38: 1-67Crossref PubMed Scopus (204) Google Scholar), although it was not efficiently displayed on our two-dimensional gel system when compared with the results of a previous researcher who used agarose gel for the first dimension (15Hirabayashi T. Electrophoresis. 2000; 21: 446-451Crossref PubMed Scopus (13) Google Scholar). These assignments of the 42-kDa and 210-kDa spots were confirmed both by Western analysis of the same two-dimensional blots using anti-actin or anti-myosin heavy chain mouse monoclonal antibodies (data not shown). Clone A28 recognized a doublet protein spot of 36–37 kDa with a pI of 4.7, suggesting that it reacted with more than one isoform of the antigen (Fig. 3C). Clone B6 recognized a single spot of 100 kDa with a pI of 5.3 (Fig. 3D). These antigens were also assigned to spots on the two-dimensional gel (Fig.3B, A28 and B6, respectively). Because the scFv phages A28 and B6 were highly specific, immunoscreening of a human muscle expression library (cloned in λ TriplEx vector, CLONTECH Laboratories) was performed to identify the corresponding antigens (Fig. 3E). We found 80 positive plaques for A28 out of 1.2 × 104λ clones and 3 positive plaques for B6 out of 1.2 × 105 λ clones. For A28, four of the five λ inserts had a sequence that matched the human cDNA sequence of tropomyosin-α (TPM3) 100%, and the other λ insert matched the other isoform of tropomyosin-α (TPM1). For B6, all three sequences of the λ inserts completely matched the human α-actinin 2 sequence. The observed molecular masses and pI values for the two protein spots were consistent with those calculated for tropomyosin-α (33 kDa and 4.6) and α-actinin 2 (104 kDa and 5.3). To confirm the results of immunoscreening, the 37- and 100-kDa protein spots visualized by copper-staining (Bio-Rad) were cut out, digestedin gel with trypsin (1Wilm M. Shevchenko A. Houthaeve T. Breit S. Schweigerer L. Fotsis T. Mann M. Nature. 1996; 379: 466-469Crossref PubMed Scopus (1508) Google Scholar), and subjected to mass fingerprinting analysis with a MALDI-TOF mass-spectrometer (13Kinumi T. Tobin S.L. Matsumoto H. Jackson K.W. Ohashi M. Eur. Mass Spectrom. 1997; 3: 367-378Crossref Scopus (11) Google Scholar). The masses of the tryptic peptides obtained for the 36- to 37-kDa protein best matched those of α- and β-tropomyosin. Similarly, a mass fingerprint of the 100-kDa protein closely matched that of α-actinin 2. Histochemical examination revealed that scFv phages A28 (anti-tropomyosin) and B6 (anti-α-actinin) diffusely stained muscle cells (Fig. 4). Anti-myosin heavy chain phage E2 stained muscle cells in an isoform-dependent manner (Fig. 4), and another anti-myosin heavy chain phage, C43, also showed an isoform-dependent staining pattern (data not shown). As expected, the isoforms of the myosin heavy chain recognized by the phage clones C43 and E2 were identified by immunoscreening to be the myosin heavy chains IIa and β (MYH7), respectively. By contrast, only one of the five clones of anti-actin scFv phage, E6, weakly stained the muscle cells, possibly because of the low affinity of anti-actin scFvs for F-actin. The features of the anti-muscle scFvs are listed in Table IV. We performed ISPS on a 4800-μm2 microarea of muscle. The 40 × 40-μm area corresponded to about 0.5–1.0 transverse sections of muscle cells (see Fig. 2C). A single round of selection was fruitful in all of our trials (experiments A through E) to obtain specific scFvs against minute amounts of antigens present in a microarea, resulting in the identification of the muscle proteins by immunoscreening. Most of the specific scFv phages obtained here were for major muscle antigens, i.e. myosin heavy chain, α-actin, and tropomyosin-α, present in the reaction area. Myosin and actin are the two major constituents of muscle and account for about 40 and 20%, respectively, of the total weight of muscle protein (14Ohtsuki I. Maruyama K. Ebashi S. Adv. Protein Chem. 1986; 38: 1-67Crossref PubMed Scopus (204) Google Scholar). One or more tropomyosins comprise the third major protein in muscle (accounting for about 5% of the total (14Ohtsuki I. Maruyama K. Ebashi S. Adv. Protein Chem. 1986; 38: 1-67Crossref PubMed Scopus (204) Google Scholar)). Because the total amount of proteins in a 4-mm2 muscle section that was 2-μm thick was measured 2.5 μg (n = 4), the amount of tropomyosin in a 4800-μm2 microfragment was 150 pg; therefore, 31 fg = 6 × 105 molecules of tropomyosin was estimated to exist in 1-μm2microfragments (2-μm thick). To our surprise, we also succeeded in obtaining an scFv phage against α-actinin 2, which accounts for only 1% of muscle protein (14Ohtsuki I. Maruyama K. Ebashi S. Adv. Protein Chem. 1986; 38: 1-67Crossref PubMed Scopus (204) Google Scholar); hence, the total weight in 4800 μm2 is 30 pg (similarly, 6.3 fg = 4 × 104 molecules of α-actinin/μm2microfragments). The reason why the scFv for such a minor component (α-actinin 2) was selected may have been the strong antigenicity of the antigen or the high abundance of one or more anti-α-actinin scFvs in the Griffin. 1 library. Another possibility is that a portion of the Z-disk, in which α-actinin is localized (16Masaki T. Endo M. Ebashi S. J. Biochem. 1967; 62: 630-632Crossref PubMed Scopus (183) Google Scholar), was efficiently exposed to the reaction surface on the transverse sections. To select scFvs against minor antigens existing in target structures, a single round, rather than multiple rounds, of selection was recommended, because positive clones could be lost during repeated rounds of selection. Presumably, changes in phage population in a phage pool through the steps of panning are affected not only by the reactivity of scFvs displayed on the phage particles but also by biological factors such as the efficiency of phage production or the growth rates of E. coli in each clone. High throughput analysis of the initial round of clones, which involves picking up unique clones from a large pool of redundant colonies and producing each clone phage, is a key for the further advance of ISPS. We noted the high quality of the phage antibodies (scFvs) selected against such minute amounts of antigens in situ from the Griffin. 1 library. Among the phage antibodies obtained here, anti-myosin heavy chain phage antibodies are definitely specific for one or more isoforms on immunohistochemistry (Fig. 4) and immunoscreening, indicating that they are reliable for immunological examinations. As expected, many of the antibodies obtained by ISPS in this study can be used for Western analyses, immunohistochemistry, and immunoscreening (Table IV). Interestingly, the obtained phage antibodies showed higher reactivity against renatured antigens (Fig.3A), suggesting that “conformation-dependent” phage antibodies can be selected by ISPS. Thus, phage antibodies that are useful for a wide variety of immunological and biochemical analyses of the target antigens could be obtained by ISPS in combination with the Griffin. 1 library. Theoretically, ISPS can also be applied to structures consisting of non-protein components, such as complex sugars and lipids. When monoclonal antibodies are obtained against non-protein components, the antigens can be identified by other methods besides immunoscreening, for example, by immunoaffinity purification. A major technical problem encountered in this study was that the collected target fragments tended to stick to the inside of glass capillaries. This was partially prevented by the addition of detergent to the buffer overlaid on tissues or by careful manipulation of the capillaries. Siliconization of capillaries was not effective to solve this problem. In this case, fragments collected from paraffin sections were not very sticky to glass capillaries. When the target structures were very small (less than 10 μm), we found another problem, namely that substantial amounts of the collected targets were lost during washing by centrifugation. This problem remains to be solved. On the other hand, when the target structures were rather large (100 μm or more), no such problems were encountered through the experimental steps. Relatively large targets could be collected directly into PCR reactions by using a needle under a stereomicroscope after tissue sections were washed on glass and dried. In this method, it was not necessary to overlay a buffer on tissue sections during collection or to wash the collected targets by centrifugation. There are four factors that may affect the sensitivity of ISPS: 1) the avidity of scFv phages for their target antigen (which depends on theKd and the valence of the displayed scFvs), 2) the phage concentration, 3) the amount of antigens involved in the reaction, and 4) the reactivity of the antigens (affected by treatment of tissue section, such as by fixation, antigen retrieval treatment, drying, or others). The proteins in paraffin sections, the most common source of tissue samples, seemed to be less reactive with scFvs, possibly due to the lower reactivity of the antigens (data not shown). It is important to find an effective method of antigen retrieval for a variety of histopathological structures when paraffin sections are used for ISPS. The scFv library can be modified to further improve the sensitivity of ISPS, because higher concentrations of individual phages are needed to detect lower amounts of antigens. Libraries with different diversities may be applied to a target structure depending on the amount of antigens to be identified. In the last decade, applications of mass spectrometry (MS) to peptide profiling (17Jimenez C.R., Li, K.W. Dreisewerd K. Spijker S. Kingston R. Bateman R.H. Burlingame A.L. Smit A.B. van Minnen J. Geraerts W.P.M. Biochemistry. 1998; 37: 2070-2076Crossref PubMed Scopus (92) Google Scholar), protein profiling (18Chaurand P. Stoeckli M. Caprioli R.M. Anal. Chem. 1999; 71: 5263-5270Crossref PubMed Scopus (247) Google Scholar), and imaging of protein distribution (19Stoeckli M. Chaurand P. Hallahan D.E. Caprioli R.M. Nature Med. 2001; 7: 493-496Crossref PubMed Scopus (1009) Google Scholar) in tissue samples have been developed. The literature does not describe the amount of tissue needed for MS-based protein identification, although as little as 5 ng of protein has been identified by MS after separation on two-dimensional polyacrylamide gel electrophoresis (1Wilm M. Shevchenko A. Houthaeve T. Breit S. Schweigerer L. Fotsis T. Mann M. Nature. 1996; 379: 466-469Crossref PubMed Scopus (1508) Google Scholar). It should be pointed out, however, that ISPS was the only approach that enabled us to obtain highly specific monoclonal antibodies by using such a minute amount of tissue components. We were able to collect more than 100 target microstructures in 1 day, depending on their abundance in a tissue section. ISPS may enable the identification of even minor components of bodies, plaques, or microdeposits of unknown components in various organs or subcutaneous lesions in focal and systemic disorders. Some of these microstructures seen in neuromuscular and systemic disorders are now under investigation. We thank Dr. Greg Winter (Medical Research Council, Cambridge, UK) for providing the Griffin. 1 library and useful comments on this work. We also acknowledge R. Kamata, K. Inugami, M. Shin, and D. Nishikiori for technical assistance. We are grateful to Drs. I. Nishino and I. Nonaka, NCNP, for arrangement of muscle samples, and Drs. S. Takashima, M. Imamura, and T. Sasaoka, NCNP, for helpful suggestions." @default.
• W2061873103 created "2016-06-24" @default.
• W2061873103 creator A5049952021 @default.
• W2061873103 creator A5051396519 @default.
• W2061873103 creator A5054746242 @default.
• W2061873103 creator A5071368896 @default.
• W2061873103 creator A5074513326 @default.
• W2061873103 creator A5076536479 @default.
• W2061873103 creator A5082340729 @default.
• W2061873103 date "2002-08-01" @default.
• W2061873103 modified "2023-09-29" @default.
• W2061873103 title "In Situ Phage Screening" @default.
• W2061873103 cites W1599833391 @default.
• W2061873103 cites W1606431180 @default.
• W2061873103 cites W1627325189 @default.
• W2061873103 cites W1966809971 @default.
• W2061873103 cites W1968421302 @default.
• W2061873103 cites W1970209157 @default.
• W2061873103 cites W1974511915 @default.
• W2061873103 cites W19835385 @default.
• W2061873103 cites W1987643769 @default.
• W2061873103 cites W1999830169 @default.
• W2061873103 cites W2011261569 @default.
• W2061873103 cites W2031154947 @default.
• W2061873103 cites W2033967455 @default.
• W2061873103 cites W2148721874 @default.
• W2061873103 cites W2175687390 @default.
• W2061873103 cites W2418850862 @default.
• W2061873103 cites W4248038669 @default.
• W2061873103 doi "https://doi.org/10.1074/jbc.m203547200" @default.
• W2061873103 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/12048201" @default.
• W2061873103 hasPublicationYear "2002" @default.
• W2061873103 type Work @default.
• W2061873103 sameAs 2061873103 @default.
• W2061873103 citedByCount "26" @default.
• W2061873103 countsByYear W20618731032012 @default.
• W2061873103 countsByYear W20618731032013 @default.
• W2061873103 countsByYear W20618731032014 @default.
• W2061873103 countsByYear W20618731032015 @default.
• W2061873103 countsByYear W20618731032016 @default.
• W2061873103 countsByYear W20618731032017 @default.
• W2061873103 countsByYear W20618731032021 @default.
• W2061873103 crossrefType "journal-article" @default.
• W2061873103 hasAuthorship W2061873103A5049952021 @default.
• W2061873103 hasAuthorship W2061873103A5051396519 @default.
• W2061873103 hasAuthorship W2061873103A5054746242 @default.
• W2061873103 hasAuthorship W2061873103A5071368896 @default.
• W2061873103 hasAuthorship W2061873103A5074513326 @default.
• W2061873103 hasAuthorship W2061873103A5076536479 @default.
• W2061873103 hasAuthorship W2061873103A5082340729 @default.
• W2061873103 hasBestOaLocation W20618731031 @default.
• W2061873103 hasConcept C153911025 @default.
• W2061873103 hasConcept C178790620 @default.
• W2061873103 hasConcept C185592680 @default.
• W2061873103 hasConcept C2777822432 @default.
• W2061873103 hasConcept C70721500 @default.
• W2061873103 hasConcept C86803240 @default.
• W2061873103 hasConcept C89423630 @default.
• W2061873103 hasConceptScore W2061873103C153911025 @default.
• W2061873103 hasConceptScore W2061873103C178790620 @default.
• W2061873103 hasConceptScore W2061873103C185592680 @default.
• W2061873103 hasConceptScore W2061873103C2777822432 @default.
• W2061873103 hasConceptScore W2061873103C70721500 @default.
• W2061873103 hasConceptScore W2061873103C86803240 @default.
• W2061873103 hasConceptScore W2061873103C89423630 @default.
• W2061873103 hasIssue "33" @default.
• W2061873103 hasLocation W20618731031 @default.
• W2061873103 hasOpenAccess W2061873103 @default.
• W2061873103 hasPrimaryLocation W20618731031 @default.
• W2061873103 hasRelatedWork W1531601525 @default.
• W2061873103 hasRelatedWork W2021904675 @default.
• W2061873103 hasRelatedWork W2026653168 @default.
• W2061873103 hasRelatedWork W2323003408 @default.
• W2061873103 hasRelatedWork W2607424097 @default.
• W2061873103 hasRelatedWork W2748952813 @default.
• W2061873103 hasRelatedWork W2899084033 @default.
• W2061873103 hasRelatedWork W2948807893 @default.
• W2061873103 hasRelatedWork W2951142470 @default.
• W2061873103 hasRelatedWork W2778153218 @default.
• W2061873103 hasVolume "277" @default.
• W2061873103 isParatext "false" @default.
• W2061873103 isRetracted "false" @default.
• W2061873103 magId "2061873103" @default.
• W2061873103 workType "article" @default.