Portal de Periódicos da CAPES

Aspergillus species recombinant antigens for serodiagnosis of farmer's lung disease

2012; Elsevier BV; Volume: 130; Issue: 3 Linguagem: Inglês

10.1016/j.jaci.2012.03.039

1097-6825

Laurence Millon, Sandrine Roussel, Bénédicte Rognon, Manfredo Quadroni, Karin Salamin, Gabriel Reboux, Coralie Barrera, Jean‐Marc Fellrath, John‐David Aubert, Jean‐Charles Dalphin, Michel Monod,

Interstitial Lung Diseases and Idiopathic Pulmonary Fibrosis

Farmer's lung disease (FLD) is the most frequent occupational hypersensitivity pneumonitis. In the East of France and Finland, various genera of mold (Aspergillus, Absidia = Lichtheimia, and Wallemia) have been suggested as the main causative agents of FLD.1Reboux G. Piarroux R. Mauny F. Madroszyk A. Millon L. Bardonnet K. et al.Role of molds in farmer's lung disease in Eastern France.Am J Respir Crit Care Med. 2001; 163: 1534-1539Crossref PubMed Scopus (108) Google Scholar, 2Lappalainen S. Pasanen A.L. Reiman M. Kalliokoski P. Serum IgG antibodies against Wallemia sebi and Fusarium species in Finnish farmers.Ann Allergy Asthma Immunol. 1998; 81: 585-592Abstract Full Text PDF PubMed Scopus (41) Google Scholar Immunologic studies have also shown that sera from patients with FLD specifically react with Aspergillus species from the glaucus group (Aspergillus vitis/Eurotium amstelodami).1Reboux G. Piarroux R. Mauny F. Madroszyk A. Millon L. Bardonnet K. et al.Role of molds in farmer's lung disease in Eastern France.Am J Respir Crit Care Med. 2001; 163: 1534-1539Crossref PubMed Scopus (108) Google Scholar, 3Erkinjuntti-Pekkanen R. Reiman M. Kokkarinen J.I. Tukiainen H.O. Terho E.O. IgG antibodies, chronic bronchitis, and pulmonary function values in farmer's lung patients and matched controls.Allergy. 1999; 54: 1181-1187Crossref PubMed Scopus (63) Google Scholar Several species of the Aspergillus genus have a well-characterized sexual cycle. This is the case for A vitis/E amstelodami, which are the anamorph/teleomorph, respectively, of the same fungus,4Samson R.A. Houbraken J. Thrane U. Frisvald J.C. Andersen B. Food and indoor fungi. CBS-KNAW Fungal Biodiversity Center, Utrecht (The Netherlands)2010Google Scholar and for Neosartorya fischeri, a species closely related to Aspergillus fumigatus.5Moran G.P. Coleman D.C. Sullivan D. Comparative genomics and the evolution of pathogenicity in human pathogenic fungi.Eukaryot Cell. 2011; 10: 34-42Crossref PubMed Scopus (71) Google Scholar Sexual reproduction is performed by conidia produced by conidial heads of aspergilli. Sexual reproduction is characterized by the production of ascospores. 1,3-β-Glucanase–treated protein extract (BGPE) from E amstelodami ascospores was recently demonstrated to be the most efficient antigen for differentiating between patients with FLD and healthy control subjects by using an ELISA serologic test.6Roussel S. Reboux G. Rognon B. Monod M. Grenouillet F. Quadroni M. et al.Comparison of three antigenic extracts of Eurotium amstelodami in farmer's lung disease serological diagnosis.Clin Vaccine Immunol. 2010; 17: 160-167Crossref PubMed Scopus (9) Google Scholar Developing a battery of standardized antigens known to cause FLD and making them available to clinicians for use in diagnosis is recommended in the report from the National Heart, Lung, and Blood Institute/Office of Rare Diseases Workshop.7Fink J.N. Ortega H.G. Reynolds H.Y. Cormier Y.F. Fan L.L. Franks T.J. et al.Needs and opportunities for research in hypersensitivity pneumonitis.Am J Respir Crit Care Med. 2005; 171: 792-798Crossref PubMed Scopus (160) Google Scholar The aim of the present study was to characterize immunogenic proteins of ascospores and to synthesize standardized recombinant antigens to improve the serologic diagnosis of FLD. The proteins of BGPE from N fischeri ascospores were first separated by 2-dimensional gel electrophoresis (2-DE) and revealed by means of Western blotting with sera from 6 patients with FLD compared with serum from 1 healthy exposed control subject. We used N fischeri rather than E amstelodami because entire genome sequences were available for further bioinformatic analysis. Immunoreactive proteins were identified by using mass spectrometry. Five immune-reactive proteins found to be FLD specific were produced as recombinant antigens. ELISA was performed with the 5 recombinant antigens and sera from 4 groups of patients: patients with FLD (n = 18), exposed control subjects (n = 32), nonexposed control subjects (n = 32), and patients with interstitial lung disease (ILD; n = 11). The protocol was approved by the local ethics committee. A complete description of the patient and control groups (see Tables E1 and E2 in this article's Online Repository at www.jacionline.org), as well as descriptions of the methods for producing BGPE, identifying immune-reactive proteins by means of mass spectrometry, producing recombinant antigens (see Table E3 in this article's Online Repository at www.jacionline.org), ELISA testing, and statistical analysis, are provided in the Methods section in this article's Online Repository at www.jacionline.org. Approximately 300 individual spots could be visualized on the 2-DE gel stained with Coomassie blue. Approximately 130 spots provided a clear signal after Western blotting with the serum from FLD3 (Fig 1). Of the 130 spots, 32 were also visible on at least one of 5 supplementary membranes revealed by using FLD1, FLD2, FLD4, FLD5, or FLD8 sera. These 32 common spots were excised from the Coomassie blue–stained gel and processed for protein identification by means of mass spectrometry. In total, 26 different proteins were clearly identified from 22 different spots (Table I). Of the 22 spots selected after mass spectrometric analysis, 17 were FLD specific (ie, not revealed with the healthy exposed control serum [EC1] or were more intensely revealed with FLD sera than with the EC1 serum), suggesting that the 21 corresponding proteins might be good candidates for differentiating patients with FLD from exposed control subjects by using ELISA serologic tests. Of these 21 proteins, only 5 with isoforms identified in at least 2 distinct spots were selected to be produced as recombinant antigens: nicotinamide adenine dinucleotide (NAD)-dependent formate dehydrogenase, mannitol-1-phosphate dehydrogenase, glutamate/leucine/phenylalanine/valine (Glu/Leu/Phe/Val) dehydrogenase, enolase, and glucose-6-phosphate isomerase.Table IProteins identified by means of mass spectrometry in spots cut from the 2-DE gelSpot no.MS analysisProtein identityUNIPROT accession no.MW (kDa)Protein identification probability (%)Unique peptidesMASCOT score1MSuperoxide dismutaseA1D1W323.4992862∗FLD-specific spots.MSuperoxide dismutase (Cu-Zn)A1DER216.010042903MSuperoxide dismutaseA1D1W323.410031384∗FLD-specific spots.LCProteasome subunit α typeA1D0A30.01004153Glu/Leu/Phe/Val dehydrogenase, putativeA1CXW149.3100368Aminotransferase, putativeA1DHN549.7102555∗FLD-specific spots.LCShort-chain dehydrogenase, putativeA1CXA131.01008311Putative uncharacterized proteinA1CZ9132.31004153Dienelactone hydrolase family proteinA1DL0527.8100699Short-chain dehydrogenaseA1DBG233.11003696MAutophagic serine protease Alp2A1DER552.61005221Malate dehydrogenaseA1DCR435.810031007MMalate dehydrogenaseA1DCR435.81003156Autophagic serine protease Alp2A1DER552.610041438∗FLD-specific spots.MExtracellular lipase, putativeA1D0B231.010021469∗FLD-specific spots.?NAD-dependent formate dehydrogenaseA1DLY145.71008363Fructose-bisphosphate aldolaseA1D6Q639.81007255Aspartate transaminase, putativeA1DGQ350.41003143Cysteine-rich secreted proteinA1D8Y536.81003115Mannitol-1-phosphate dehydrogenaseA1DGY942.99827610∗FLD-specific spots.MMannitol-1-phosphate dehydrogenaseA1DGY942.9100636411LCFructose-bisphosphate aldolaseA1D6Q639.81007316NAD-dependent formate dehydrogenaseA1DLY145.71006256Mannitol-1-phosphate dehydrogenaseA1DGY942.91004220Translation initiation factor, putativeA1CY3841.499384Quinone oxidoreductase, putativeA1D55236.510028012∗FLD-specific spots.LCMannitol-1-phosphate dehydrogenaseA1DGY942.910011594Isovaleryl-CoA dehydrogenase IvdA, putativeA1DEU146.41003111Eukaryotic translation initiation factor 3 subunit EifCh, putativeA1D37939.29824213∗FLD-specific spots.MNAD-dependent formate dehydrogenaseA1DLY145.710046714∗FLD-specific spots.MMannitol-1-phosphate dehydrogenaseA1DGY942.9100524115∗FLD-specific spots.MGlu/Leu/Phe/Val dehydrogenase, putativeA1CXW149.3100518416∗FLD-specific spots.MGlu/Leu/Phe/Val dehydrogenase, putativeA1CXW149.310047517∗FLD-specific spots.MGlu/Leu/Phe/Val dehydrogenase, putativeA1CXW149.310044818∗FLD-specific spots.MGlu/Leu/Phe/Val dehydrogenase, putativeA1CXW149.3100724719∗FLD-specific spots.MEnolaseA1DM7647.3100522020∗FLD-specific spots.MEnolaseA1DM7647.3100632721∗FLD-specific spots.MGlucose-6-phosphate isomeraseA1DGR461.4100830722∗FLD-specific spots.LCGlucose-6-phosphate isomeraseA1DGR461.410019807Glu/Leu/Phe/Val dehydrogenase, putativeA1CXW1149.399251Putative uncharacterized proteinA1D6F356.9982493-Ketosteroid–delta-1-dehydrogenase, putativeA1DDH361.499248UTP-glucose-1-phosphate uridylyltransferase Ugp1, putativeA1D96258.099246Proteins in boldface were chosen to be produced as recombinant antigens.L, Liquid chromatography–tandem mass spectrometry; M, matrix-assisted laser desorption/ionization–time-of-flight mass spectrometry; MS, mass spectrometry; MW, molecular weight.∗ FLD-specific spots. Open table in a new tab Proteins in boldface were chosen to be produced as recombinant antigens. L, Liquid chromatography–tandem mass spectrometry; M, matrix-assisted laser desorption/ionization–time-of-flight mass spectrometry; MS, mass spectrometry; MW, molecular weight. The recombinant antigens were used for immunologic investigation in ELISAs. Sensitivity, specificity, likelihood ratios, and the ELISA area under the curve results for optimal thresholds are indicated in Table E4 in this article's Online Repository at www.jacionline.org. As indicated by areas under the curve of 0.89 and 0.87, the best recombinant antigens for differentiating between patients with FLD and exposed control subjects were Glu/Leu/Phen/Val dehydrogenase and glucose-6-phosphate isomerase, with sensitivities of 72% and specificities of 94% and 84%, respectively. A score combining the positivity for these 2 antigens increased the sensitivity of the ELISA to 89%, with a specificity of 84%. Our experimental approach, combining the use of sera from patients with FLD to screen for antigenic proteins from ascospores and the use of available genomic data from Aspergillus species close to the species implicated as a causative agent of FLD, led to the production of 2 recombinant antigens that were remarkably effective in differentiating between patients with FLD and exposed control farmers by using ELISA serologic tests. Glucose-6-phosphate isomerase and Glu/Leu/Phen/Val dehydrogenase were previously identified among immunogenic proteins from A fumigatus that have been revealed by using IgE and IgG immunoblots with pooled sera from patients with allergic bronchopulmonary aspergillosis.8Singh B. Sharma G. Oellerich M. Kumar R. Singh S. Bhadoria D. et al.Novel cytosolic allergens of Aspergillus fumigatus identified from germinating conidia.J Proteome Res. 2010; 9: 5530-5541Crossref PubMed Scopus (31) Google Scholar These proteins were shown to selectively react with IgE. Interestingly, the same proteins revealed by different isotypes of immunoglobulin are useful for the serologic diagnosis of both diseases. The ability to differentiate between patients with FLD and exposed control subjects by using ELISAs with ascospore crude extract was evaluated in a previous study6Roussel S. Reboux G. Rognon B. Monod M. Grenouillet F. Quadroni M. et al.Comparison of three antigenic extracts of Eurotium amstelodami in farmer's lung disease serological diagnosis.Clin Vaccine Immunol. 2010; 17: 160-167Crossref PubMed Scopus (9) Google Scholar and was demonstrated as follows: sensitivity, 71%; specificity, 87%, positive likelihood ratio, 5.6; negative likelihood ratio, 0.3; and area under the curve, 0.85. Sensitivity was at 89% when a score combining the results of ELISAs with the 2 recombinant antigens was used. This was higher than the sensitivity of the serologic tests that are currently used for FLD diagnosis (sensitivity of 67% for electrosyneresis and double diffusion with crude antigenic extracts of 5 microorganisms from the agricultural environment: E amstelodami, Lichtheimia corymbifera, Wallemia sebi, Saccharopolyspora rectivirgula, and mesophilic Streptomyces species).9Fenoglio C.M. Reboux G. Sudre B. Mercier M. Roussel S. Cordier J.F. et al.Diagnostic value of serum precipitins to mould antigens in active hypersensitivity pneumonitis.Eur Respir J. 2007; 29: 706-712Crossref PubMed Scopus (26) Google Scholar Specificity was at 84%, whereas it was between 89% and 100% for routine immunoprecipitation techniques. At a time when quality procedure, control, and methodological validation of laboratory techniques are essential, the availability of large quantities of antigens produced with a standardized and reproducible procedure is a key issue. Because of the low volume of serum necessary for ELISA analysis (1 μL per test), numerous antigens can be tested for a single serum sample in 1 experiment. Testing several recombinant antigens from various environmental microorganisms known to cause FLD on the same plate can easily be considered to improve the sensitivity of serologic diagnosis. Our strategy, consisting of using sera from patients with FLD to identify immunogenic proteins from environmental microorganisms, could be applied to other microorganisms known to be causative agents of FLD, with the aim of proposing panels of recombinant antigens able to improve the sensitivity and standardization of serologic diagnosis. However, only a prospective study including patients with FLD and control subjects can determine whether ELISA with individual recombinant antigens, combinations of recombinant antigens, or both increases diagnostic performance compared with existing techniques. We thank the Mutualité Sociale Agricole and the PAPPA group (“Reseau des Pathologies Pulmonaires Agricoles”) for their help in recruiting exposed control subjects. We also thank Frances Sheppard from the Clinical Investigation Center (CIC Inserm) of Besançon for her editorial assistance. Patients with FLD were farmers who had just received a diagnosis in the pneumology departments of hospitals in Lausanne, Neuchatel (Switzerland), and Besançon (France) between September 2006 and May 2008. All 18 patients with FLD were given a diagnosis according to consensual criteria (Table E1)E1Lacasse Y. Selman M. Costabel U. Dalphin J.C. Ando M. Morell F. et al.Clinical diagnosis of hypersensitivity pneumonitis.Am J Respir Crit Care Med. 2003; 168: 952-958Crossref PubMed Scopus (484) Google Scholar and had at least 1 positive serologic test result using electrosyneresis with 1 presumed agent of FLD. Electrosyneresis on cellulose acetate was performed with a crude extract of E amstelodami (BBCM/IHEM16286), L corymbifera (BBCM/IHEM3809), W sebi (BBCM/IHEM16284), and S rectivirgula (DSMZ43747), as previously described.E2Reboux G. Piarroux R. Roussel S. Millon L. Bardonnet K. Dalphin J.C. Assessment of four serological techniques in the immunological diagnosis of farmers' lung disease.J Med Microbiol. 2007; 56: 1317-1321Crossref PubMed Scopus (46) Google Scholar The threshold of positivity for FLD diagnosis was previously determined as 3 precipitin arcs for the E amstelodami extract and 2 precipitin arcs for the L corymbifera, W sebi, and S rectivirgula extracts.E3Fenoglio C.M. Reboux G. Sudre B. Mercier M. Roussel S. Cordier J.F. et al.Diagnostic value of serum precipitins to mould antigens in active hypersensitivity pneumonitis.Eur Respir J. 2007; 29: 706-712Crossref PubMed Scopus (71) Google Scholar The numbers of precipitin arcs obtained for patients with FLD are reported in Table E2. Exposed control subjects (n = 32) were healthy farmers recruited during occupational medicine visits organized by a social security regimen for farmers (“Mutualité Sociale Agricole”). Spirometry, serology, and standardized medical questionnaires were used to ascertain that none of the patients presented with any general or respiratory symptoms. Nonexposed control subjects (n = 32) were healthy nonfarmer volunteers living in an urban area (Paris). They were recruited during blood donations in a specialized center. A questionnaire was used to establish that control subjects were not exposed to a farming environment, birds, or molds in their dwellings. All were nonsmokers. ILDs (n = 11) were nonspecific interstitial pneumonia, idiopathic pulmonary fibrosis, desquamative interstitial pneumonia/respiratory bronchiolitis-associated ILD, and drug or connective tissue disease–associated ILD. The protocol was approved by the local ethics committee. Blood samples were taken for each subject who had provided written consent and centrifuged within 4 hours after sampling. Serum was immediately stored at −80°C. Ascospores of 20 densely grown malt agar culture media samples of N fischeri (NF9; Pasteur Institute, Paris, France) were first suspended in 60 mL of phosphate NaCl buffer (10 mmol/L Na3PO4 and 0.15 mol/L NaCl, pH 6). A 5-minute centrifugation was performed to pellet the ascospores, and then the ascospores were resuspended in 5 mL of 0.1 mol/L Tris-HCl buffer at pH 8.5. BGPE was prepared as previously describedE4Roussel S. Reboux G. Rognon B. Monod M. Grenouillet F. Quadroni M. et al.Comparison of three antigenic extracts of Eurotium amstelodami in farmer's lung disease serological diagnosis.Clin Vaccine Immunol. 2010; 17: 160-167Crossref PubMed Scopus (22) Google Scholar and purified with an SDS-PAGE clean-up kit (Roche Diagnostics, Basel, Switzerland). BGPE was then suspended in an ELISA coating buffer. Protein concentration of the purified extract was determined by using the DC Protein Assay (Bio-Rad, Marnes-la-Coquette, France), according to the manufacturer's instructions. BPBE that was generated as described above and contained 54 μg of proteins was suspended in 2-dimensional PAGE lysis buffer (5 mmol/L magnesium acetate, 7 mol/L urea, 2 mol/L thiourea, 4% [wt/vol] 3-[(3-Cholamidopropyl)dimethylammonio]-1-propanesulfonate, and 30 mmol/L Tris, pH 8.5). After sonication and centrifugation, the supernatant was diluted in an equal volume of the same buffer supplemented with 2 mg/mL dithiothreitol and 2% (vol/vol) Pharmalyte 3-10 (GE Healthcare, Freiburg, Germany) and separated by using 2-DE. Immobiline DryStrip gels (pH 3-10 NL, 11 cm; GE Healthcare) were rehydrated according to the manufacturer's instructions before first-dimension separation with a cup-loading sample application. Isoelectric focusing was carried out on an Ettan IPGphor electrophoresis system (GE Healthcare), with a total focusing of 30,000 Vh, according to the manufacturer's manual and instructions. Before the second dimension, strips were equilibrated for 15 minutes in a reducing buffer containing 6 mol/L urea, 2% (wt/vol) SDS, 30% (vol/vol) glycerol, 32 mmol/L dithiothreitol, and 50 mmol/L Tris at pH 6.8. This was followed by a 10-minute alkylation in a buffer containing 6 mol/L urea, 2% (wt/vol) SDS, 30% (vol/vol) glycerol, 240 mmol/L iodoacetamide, and 100 mmol/L Tris at pH 8. Second-dimension separation was carried out on a 7 × 9–cm 10% acrylamide gel surmounted by a 4% stacking gel. The IPG strip from the first dimension was loaded onto the slab gels together with a molecular mass size marker and run across the stacking gel at 80 V for 30 minutes and 120 V for 1 hour. After electrophoresis, proteins were stained with colloidal Coomassie blue or transferred onto the nitrocellulose membrane for Western blotting with sera from FLD3, FLD1, FLD2, FLD4, FLD5, or FLD8 and with 1 serum from a healthy exposed control subject (EC1). These sera were selected by using previous 1-dimensional Western blot experiments (data not shown): FLD3 because it yielded the highest number of colored bands and the weakest background; FLD1, FLD2, FLD4, FLD5, or FLD8 because they showed a majority of bands common to patients with FLD; and EC1 because it was representative of most of the bands present in exposed control subjects. Spot detection and alignment between gel and Western blot imaging were performed by using Melanie 6.0 software (Genebio, Geneva, Switzerland). Before incubation with the serum, proteins transferred onto the nitrocellulose membrane were stained with Ponceau red. This made it possible to select and mark intense reference spots on the nitrocellulose membrane and facilitated the alignment of the spot patterns between the Coomassie blue–colored gel and the Western blot membrane. Gel spots were excised from SDS-PAGE and transferred to special 96-well plates (PerkinElmer Life Sciences, Boston, Massachusetts). Tryptic digestion and protein identification by means of tandem mass spectrometry on a MALDI-TOF/TOF instrument (Applied Biosystems 4700 Proteome Analyzer; Applied Biosystems, Framingham, Mass) was performed essentially as previously described.E5Giddey K. Monod M. Barblan J. Potts A. Waridel P. Zaugg C. et al.Comprehensive analysis of proteins secreted by Trichophyton rubrum and Trichophyton violaceum under in vitro conditions.J Proteome Res. 2007; 6: 3081-3092Crossref PubMed Scopus (45) Google Scholar When no identification was obtained or the results were ambiguous, the samples were reanalyzed by using nanoflow liquid chromatography–tandem mass spectrometry on a Thermo Scientific LTQ-Orbitrap XL essentially as described.E6Sriranganadane D. Waridel P. Salamin K. Reichard U. Grouzmann E. Neuhaus J.M. et al.Aspergillus protein degradation pathways with different secreted protease sets at neutral and acidic pH.J Proteome Res. 2010; 9: 3511-3519Crossref PubMed Scopus (39) Google Scholar Tandem mass spectrometric datasets were searched by using Mascot (version 2.2.0; Matrix Science, London, United Kingdom) against the set of 10,426 protein sequences from N fischeri extracted from release 14.4 of the UNIPROT database. Mass tolerance for precursor peptide matching was at 20 and 10 ppm for MALDI-TOF/TOF and Orbitrap data, respectively. MASCOT database search results were imported into Scaffold software (Proteome Software, Portland, Ore) and filtered based on protein identification probability with a threshold at 95%. With the aim of simplifying our procedure, we used available cDNA from A fumigatus to amplify sequences coding for the proteins of interest. The coverage of the 5 selected proteins was 100% with corresponding proteins in A fumigatus (Af293, GenBank). Identity with A fumigatus Af293 sequences was 99% for Glu/Leu/Phe/Val dehydrogenase, glucose-6-phosphate isomerase, and enolase; 98% for NAD-dependent formate dehydrogenase Aci/FdH; and 97% for mannitol-1-phosphate dehydrogenase. Large peptides corresponding to 5 sequenced proteins were produced by using the pET expression system from Novagen (Darmstadt, Germany). The P1/P2, P3/P4, P5/P6, P7/P8, and P9/P10 pairs of sense/antisense primers (Table E3) were used to amplify sequence coding for the proteins of interest from A fumigatus cDNA.E7Monod M. Paris S. Sarfati J. Jaton-Ogay K. Ave P. Latgé J.P. Virulence of alkaline protease-deficient mutants of Aspergillus fumigatus.FEMS Microbiol Lett. 1993; 106: 39-46Crossref PubMed Scopus (111) Google Scholar PCR products were digested with NcoI/BamHI or Rca/BglII and cloned into the NcoI and BamHI sites of pET-11aH6, which is a derivative of pET-11a.E8Reichard U. Le'chenne B. Asif A.R. Streit F. Grouzmann E. Jousson O. et al.Sedolisins, a new class of secreted proteases from Aspergillus fumigatus with endoprotease or tripeptidyl-peptidase activity at acidic pHs.Appl Environ Microbiol. 2006; 72: 1739-1748Crossref PubMed Scopus (56) Google Scholar The plasmid pET-11aH6 was generated by incorporating the sequence 5′-CATGCACCATCACCATCACCATGGTAAGGATC-3′ encoding a Met-His6 amino acid sequence between the unique NheI and BamHI cloning sites of pET 11a. The sixth His residue in pET-11aH6 was encoded by a unique NcoI cloning site. The plasmids obtained after cloning were termed pAg1 to pAg5. Heterologous 6×His-tagged peptides were produced in Escherichia coli BL21 transformed with pAg1 to pAg5. Cells were grown at 37°C to an OD of 600 nm (OD600nm) of 0.6, and 6×His-tagged peptide expression was induced by adding isopropyl β-D-1-thiogalactopyranoside to a final concentration of 0.1 mmol/L. Incubation was continued for an additional 4-hour period at 37°C. Cells were collected by means of centrifugation (4500g at 4°C for 15 minutes), and 6×His-tagged peptides were extracted with guanidine hydrochloride buffer and Ni-NTA resin (Qiagen, Hilden, Germany) columns, as recommended by the manufacturer. Protein concentrations were determined spectrophotometrically measuring OD280nm. ELISAs were performed with 93 sera (18 from patients with FLD, 32 from exposed control subjects, 32 from nonexposed control subjects, and 11 from patients with ILD) to look for IgG against the 5 recombinant antigens synthesized. The wells of 96-well plates (PolySorp Immunomodule; Nalge Nunc, Rochester, NY) were coated with 200 μL of 1 μg/mL recombinant antigen solution in 50 mmol/L K2HPO4 buffer, pH 8.5, at 4°C for 48 hours. ELISAs for specific IgG were then conducted as previously described.E4Roussel S. Reboux G. Rognon B. Monod M. Grenouillet F. Quadroni M. et al.Comparison of three antigenic extracts of Eurotium amstelodami in farmer's lung disease serological diagnosis.Clin Vaccine Immunol. 2010; 17: 160-167Crossref PubMed Scopus (22) Google Scholar Well intensities were read with a spectrophotometer at 450 nm (Titertek Multiskan, Helsinki, Finland), and the results were expressed in OD. Sera were deposited in triplicate on ELISA plates, and variation coefficients were calculated. Sera for which the variation coefficient was superior to 20% for an OD mean of greater than 0.050 were tested again. Means of ODs validated by variation coefficient were used for statistical analysis by using receiver operating characteristic curves with STATA9 software (StataCorp, College Station, Tex). Sensitivity and specificity tables were calculated for each OD, and the value that showed the best percentage of correctly classified patients and control subjects was chosen as the threshold. When several thresholds were equivalent, the one with the best sensitivity was chosen. Receiver operating characteristics allowed recombinant antigens to be compared by examining the area under the curve. The area under the curve was considered significant when the 95% CI did not include the value 0.5. An ELISA test with recombinant antigen was considered high performance when the area under the curve was greater than 0.8.Table E1Characteristics of patients with FLDPatient no.Age (y)SexSmokerHospitalClinical presentationRadiologic dataBronchoalveolar lavageCell/mm3Lymphocytes (%)FLD148MNoBesançonAcute recurrentGround-glass opacities76047FLD248MNoBesançonSubacuteGround-glass opacitiesAir trapping46058FLD353MNoBesançonChronicGround-glass opacities32060FLD451MNoBesançonAcute recurrentGround-glass opacities98082FLD547MNoBesançonSubacute recurrentMosaic featureGround-glass opacities68059FLD651FNoBesançonChronicDiffuse ground-glass opacities101083FLD747MNoBesançonAcuteGround-glass opacities27057FLD859MNoBesançonSubacuteSmall nodulesGround-glass opacities45045FLD951MNoBesançonSubacuteSmall nodulesGround-glass opacities34061FLD1051FNoBesançonSubacute severeSmall nodulesGround-glass opacitiesMosaic feature41065FLD1143MNoBesançonChronicDiffuse ground-glass opacitiesAir trapping36064FLD1253MYesBesançonSubacuteDiffuse ground-glass opacities62079FLD1334MNoBesançonAcuteDiffuse ground-glass opacitiesMosaic feature97063FLD1445MYesLausanneSubacuteSmall nodulesGround-glass opacitiesMosaic feature57068FLD1561MNoNeuchatelAcute exacerbation of a chronic formDiffuse ground-glass opacitiesAir trappingMosaic feature70015FLD1654MNoNeuchatelAcute exacerbation of a chronic formEmphysemaMosaic feature40064FLD1753MNoNeuchatelSubacute severeMosaic featureAir trapping122544FLD1854MNoBesançonChronicGround-glass opacitiesAir trapping39047F, Female; M, male. Open table in a new tab Table E2Serologic diagnosis using crude antigen extracts from Eurotium amstelodami, Lichtheimia corymbifera, Wallemia sebi, and Saccharopolyspora rectivirgulaCase no.Electrosyneresis on cellulose acetateNo. of arcsE amstelodami (threshold ≥3 arcs)L corymbifera (threshold ≥2 arcs)W sebi (threshold ≥2 arcs)S rectivirgula (threshold ≥2 arcs)FLD12240FLD23121FLD31143FLD44141FLD51050FLD642115FLD70204FLD80130FLD92131FLD101127FLD111260FLD122200FLD131157FLD141141FLD151043FLD167677FLD1743105FLD183264Positive results are indicated in boldface. Open table in a new tab Table E3List of primers usedTarget genePrimer sequence (5′-3′)NAD-dependent formate dehydrogenase FPTCTTCGGTACGTCCAGGGCCC RPGTTGGATCCTTACTTCTGTCTCTGGCCGTAMannitol-1-phosphate dehydrogenase FPGTTCCCATGGTATGGGAAAGAAGGCTATCCAG RPGTTGGATCCCTACTTGCTGTCCTTCTGCACCGlu/Leu/Phe/Val dehydrogenase FPGTTCCCATGGTATGTCTAACCTTCCTCACGAG RPGTTGGATCCTTACCACCAGTCACCCTGCTCEnolase FPGTTCCATGGGCCTATCTCCAAGATCCAC RPCTTGGATCCTTACAGGTTGACGGCAGTGCGGlucose-6-phosphate isomerase FPGTTTCATGAGATGCCTGGTTTTTCGCAGGC RPGTTAGATCTTTAAGCCAGGTTGGCCTTCTTFP, Forward primer; RP, reverse primer. Open table in a new tab Table E4Sensitivity, specificity, positive and negative likelihood ratios, and area under the curve obtained for the 5 recombinant antigens with sera from patients with FLD and control subjectsProtein nameSensitivity (% [95% CI])Specificity (% [95% CI])LR+ (95% CI)LR− (95% CI)AUC (95% CI)Patients with FLD (n = 18) vs exposed control subjects (n = 32) NAD-dependent formate dehydrogenase61 (36-83)84 (67-95)3.9 (2.6-5.8)0.5 (0.2-1.2)0.72 (0.55-0.89) Mannitol-1-phosphate dehydrogenase56 (31-79)69 (60-84)1.8 (1.1-2.9)0.6 (0.3-1.3)0.62 (0.45-0.78) Glu/Leu/Phe/Val dehydrogenase72 (47-90)94 (79-99)11.6 (8.6-15.6)0.3 (0.1-1.4)0.89 (0.80-1) Enolase33 (13-59)84 (67-95)2.1 (1.1-4.2)0.8 (0.3-1.9)0.53 (0.35-0.70) Glucose-6-phosphate isomerase72 (47-90)84 (67-95)4.6 (3.3-6.4)0.3 (0.1-1.0)0.87 (0.78-0.97)Patients with FLD (n = 18) vs nonexposed control subjects (n = 32) NAD-dependent formate dehydrogenase61 (36-83)91 (75-98)6.5 (4.4-9.6)0.4 (0.1-1.5)0.93 (0.85-1) Mannitol-1-phosphate dehydrogenase56 (31-79)47 (29-65)1.0 (0.6-1.8)0.9 (0.5-1.7)0.51 (0.33-0.68) Glu/Leu/Phe/Val dehydrogenase72 (47-90)84 (67-95)4.6 (3.3-6.4)0.3 (0.1-1.0)0.88 (0.79-0.98) Enolase33 (13-59)88 (71-96)2.7 (1.4-5.2)0.8 (0.3-2.0)0.67 (0.52-0.82) Glucose-6-phosphate isomerase72 (47-90)63 (44-79)1.9 (1.3-2.9)0.4 (0.2-1.1)0.75 (0.60-0.89)Patients with FLD (n = 18) vs patients with ILD (n = 11) NAD-dependent formate dehydrogenase61 (36-83)91 (59-100)6.7 (4.4-10.2)0.4 (0.1-3.0)0.72 (0.51-0.93) Mannitol-1-phosphate dehydrogenase56 (31-79)82 (48-98)3.1 (1.9-5.0)0.5 (0.1-2.1)0.73 (0.54-0.92) Glu/Leu/Phe/Val dehydrogenase72 (47-90)64 (31-89)2.0 (1.2-3.4)0.4 (0.1-1.3)0.72 (0.53-0.91) Enolase33 (13-59)91 (59-100)3.7 (1.9-7.2)0.7 (0.1-4.9)0.57 (0.33-0.80) Glucose-6-phosphate isomerase72 (47-90)72 (39-94)2.6 (1.7-4.2)0.4 (0.1-1.3)0.70 (0.48-0.91)AUC, Area under the curve; LR+, positive likelihood ratio; LR−, negative likelihood ratio; Se, sensitivity; Sp, specificity. Open table in a new tab F, Female; M, male. Positive results are indicated in boldface. FP, Forward primer; RP, reverse primer. AUC, Area under the curve; LR+, positive likelihood ratio; LR−, negative likelihood ratio; Se, sensitivity; Sp, specificity.