Title: Mining the Acute Respiratory Distress Syndrome Proteome: Identification of the Insulin-Like Growth Factor (IGF)/IGF-Binding Protein-3 Pathway in Acute Lung Injury
Abstract: To obtain a more complete protein profile of the airspace milieu in acute respiratory distress syndrome (ARDS) and to identify new mediators, we analyzed bronchoalveolar lavage fluid (BALF) by shotgun proteomics. Using BALF from three patients, we identified a total of 870 different proteins, a nearly 10-fold increase from previous reports. Among the proteins identified were known markers of lung injury, such as surfactant, proteases, and serum proteins. We also identified several biologically interesting proteins not previously identified in patients with ARDS, including insulin-like growth factor-binding protein-3 (IGFBP-3). Because of the known role of IGFBP-3 in regulating cell survival, we measured IGFBP-3 levels by enzyme-linked immunosorbent assay in ARDS BALF. Normal controls had low levels of IGFBP-3, whereas patients with early ARDS had a significant increase in IGFBP-3. The IGF pathway, acting through the type 1 IGF-receptor, repressed apoptosis of lung fibroblasts but not lung epithelial cells. Furthermore, depletion of IGF in ARDS BALF led to enhanced fibroblast apoptosis. Our data suggest that the IGFBP-3/IGF pathway is involved in the pathogenesis of lung injury, illustrating the power of shotgun proteomics to catalog proteins present in complex biological fluids, such as BALF, from which hypotheses can be developed and tested. To obtain a more complete protein profile of the airspace milieu in acute respiratory distress syndrome (ARDS) and to identify new mediators, we analyzed bronchoalveolar lavage fluid (BALF) by shotgun proteomics. Using BALF from three patients, we identified a total of 870 different proteins, a nearly 10-fold increase from previous reports. Among the proteins identified were known markers of lung injury, such as surfactant, proteases, and serum proteins. We also identified several biologically interesting proteins not previously identified in patients with ARDS, including insulin-like growth factor-binding protein-3 (IGFBP-3). Because of the known role of IGFBP-3 in regulating cell survival, we measured IGFBP-3 levels by enzyme-linked immunosorbent assay in ARDS BALF. Normal controls had low levels of IGFBP-3, whereas patients with early ARDS had a significant increase in IGFBP-3. The IGF pathway, acting through the type 1 IGF-receptor, repressed apoptosis of lung fibroblasts but not lung epithelial cells. Furthermore, depletion of IGF in ARDS BALF led to enhanced fibroblast apoptosis. Our data suggest that the IGFBP-3/IGF pathway is involved in the pathogenesis of lung injury, illustrating the power of shotgun proteomics to catalog proteins present in complex biological fluids, such as BALF, from which hypotheses can be developed and tested. Acute respiratory distress syndrome (ARDS), first described in 1967 by Ashbaugh and colleagues,1Ashbaugh DG Bigelow DB Petty TL Levine BE Acute respiratory distress in adults.Lancet. 1967; 2: 319-323Abstract PubMed Google Scholar remains an important cause of morbidity and mortality in critically ill patients. ARDS is characterized by an acute pulmonary inflammatory process with epithelial apoptosis and interstitial and intra-alveolar edema, followed by fibroblast proliferation, migration, and fibrosis. The diagnosis of ARDS is based on clinical and radiographical criteria, including acute onset, bilateral infiltrates on chest radiograph, absence of congestive heart failure, and hypoxemia.2Bernard GR Artigas A Brigham KL Carlet J Falke K Hudson L Lamy M LeGall JR Morris A Spragg R Report of the American-European Consensus conference on acute respiratory distress syndrome: definitions, mechanisms, relevant outcomes, and clinical trial coordination.J Crit Care. 1994; 9: 72-81Crossref PubMed Scopus (330) Google Scholar This consensus definition has improved the standardization of clinical research and trials; however, it does not take into account the cause or mechanism of disease. Much work has focused on the identification of humoral or cellular biological markers of ARDS in hopes that such markers may provide insight into the mechanisms of ARDS and improve the prediction of ARDS in high risk patients and prediction of outcome in ARDS patients.3Pittet JF Mackersie RC Martin TR Matthay MA Biological markers of acute lung injury: prognostic and pathogenetic significance.Am J Respir Crit Care Med. 1997; 155: 1187-1205Crossref PubMed Scopus (524) Google Scholar To date, no single protein marker identified by traditional laboratory methods has demonstrated the specificity or sensitivity to serve as a reliable predictor of outcome. However, new proteomic methods provide the opportunity to assess the protein profile of a sample that is independent of investigator's biases and thus has the potential of identifying unsuspected mediators or pathways involved in lung injury. As a screening strategy to define the bronchoalveolar lavage fluid (BALF) proteome from ARDS patients, we used “shotgun proteomics,” consisting of digestion of proteins in BALF followed by strong-cation exchange fractionation of the peptide mixture and microcapillary-high performance liquid chromatography electrospray ionization tandem mass spectrometry analysis, and then computerized data processing.4Griffin TJ Aebersold R Advances in proteome analysis by mass spectrometry.J Biol Chem. 2001; 276: 45497-45500Abstract Full Text Full Text PDF PubMed Scopus (128) Google Scholar Using strict criteria for matching peptide tandem mass spectra to sequences in a database,5Keller A Nesvizhskii AI Kolker E Aebersold R Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search.Anal Chem. 2002; 74: 5383-5392Crossref PubMed Scopus (3931) Google Scholar, 6Nesvizhskii AI Keller A Kolker E Aebersold R A statistical model for identifying proteins by tandem mass spectrometry.Anal Chem. 2003; 75: 4646-4658Crossref PubMed Scopus (3686) Google Scholar we identified from three patients a total of 897 proteins, of which 79 were identified in all three patients. We selected several of the identified proteins for further testing based on their known functions and potential relevance to lung injury. Expression levels of the candidate proteins were analyzed by enzyme-linked immunosorbent assay (ELISA) in a large sample set of ARDS BALF. Notable among the results, we found insulin-like growth factor (IGF)-binding protein-3 (IGFBP-3) and IGF expression levels correlated with progression of ARDS. Furthermore, we showed that the IGF pathway regulates apoptosis of lung fibroblasts, but not lung epithelial cells, suggesting that the IGF pathway may contribute to the fibroproliferative response in ARDS. The protocol was approved by the Institutional Review Board, University of Washington. Written informed consent was obtained from the patient or responsible relative before patients were entered into the study. Patients with acute lung injury undergoing bronchoscopy for suspected ventilator-associated pneumonia were included in the study as the initial index patients (Table 1).Table 1Clinical Characteristics of ARDS PatientsIndex patients*None of the index patients had a diagnosis of pneumonia as a risk factor for acute lung.Retrospective cohort123At RiskARDS summaryDay 1Day 3Day 7Day 14Age28495947 ± 1543 ± 16Risk factor SepsisYes323 TraumaYes525 OtherYes013Apache score16222019 ± 822 ± 722 ± 621 ± 923 ± 819 ± 3PaO2/FiO2184271290226 ± 83162 ± 61151 ± 55162 ± 58189 ± 71178 ± 43Static compliancendndnd40 ± 1037 ± 1341 ± 1537 ± 1028 ± 935 ± 17% BAL volume recovered38713357 ± 1150 ± 1454 ± 1250 ± 1543 ± 1144 ± 19VAP?NoYesNo* None of the index patients had a diagnosis of pneumonia as a risk factor for acute lung. Open table in a new tab In addition, we retrospectively analyzed specimens obtained as part of the University of Washington Specialized Center of Research program in acute lung injury. All patients admitted to the Medical and Surgical Intensive Care Units of Harborview Medical Center between January 1994 and November 1997 were screened prospectively to identify patients who met criteria for predetermined risk criteria for trauma or sepsis “at risk”7Matute-Bello G Liles WC Radella 2nd, F Steinberg KP Ruzinski JT Jonas M Chi EY Hudson LD Martin TR Neutrophil apoptosis in the acute respiratory distress syndrome.Am J Respir Crit Care Med. 1997; 156: 1969-1977Crossref PubMed Scopus (344) Google Scholar or with established ARDS.7Matute-Bello G Liles WC Radella 2nd, F Steinberg KP Ruzinski JT Jonas M Chi EY Hudson LD Martin TR Neutrophil apoptosis in the acute respiratory distress syndrome.Am J Respir Crit Care Med. 1997; 156: 1969-1977Crossref PubMed Scopus (344) Google Scholar, 8Steinberg KP Milberg JA Martin TR Maunder RJ Cockrill BA Hudson LD Evolution of bronchoalveolar cell populations in the adult respiratory distress syndrome.Am J Respir Crit Care Med. 1994; 150: 113-122Crossref PubMed Scopus (339) Google Scholar, 9Greene KE Wright JR Steinberg KP Ruzinski JT Caldwell E Wong WB Hull W Whitsett JA Akino T Kuroki Y Nagae H Hudson LD Martin TR Serial changes in surfactant-associated proteins in lung and serum before and after onset of ARDS.Am J Respir Crit Care Med. 1999; 160: 1843-1850Crossref PubMed Scopus (368) Google Scholar Patients at risk for ARDS did not meet either radiographical or oxygenation criteria for ARDS. Patients were screened prospectively for the onset of ARDS using the following criteria: 1) PaO2/FiO2 < 150 mm Hg or < 200 mm Hg on ≥ 5 cm of H2O positive end-expiratory pressure, 2) diffuse parenchymal infiltrates, 3) pulmonary artery wedge pressure < 18 mm Hg or no clinical evidence of congestive heart failure, and 4) no other obvious explanation for these findings. All patients with ARDS met the criteria of the American-European Consensus Conference definition.2Bernard GR Artigas A Brigham KL Carlet J Falke K Hudson L Lamy M LeGall JR Morris A Spragg R Report of the American-European Consensus conference on acute respiratory distress syndrome: definitions, mechanisms, relevant outcomes, and clinical trial coordination.J Crit Care. 1994; 9: 72-81Crossref PubMed Scopus (330) Google Scholar Day 1 was defined as the first 24 hours after meeting the above criteria for ARDS. The clinical characteristics of the patient groups are shown in Table 1 and have been previously described.7Matute-Bello G Liles WC Radella 2nd, F Steinberg KP Ruzinski JT Jonas M Chi EY Hudson LD Martin TR Neutrophil apoptosis in the acute respiratory distress syndrome.Am J Respir Crit Care Med. 1997; 156: 1969-1977Crossref PubMed Scopus (344) Google Scholar, 8Steinberg KP Milberg JA Martin TR Maunder RJ Cockrill BA Hudson LD Evolution of bronchoalveolar cell populations in the adult respiratory distress syndrome.Am J Respir Crit Care Med. 1994; 150: 113-122Crossref PubMed Scopus (339) Google Scholar, 9Greene KE Wright JR Steinberg KP Ruzinski JT Caldwell E Wong WB Hull W Whitsett JA Akino T Kuroki Y Nagae H Hudson LD Martin TR Serial changes in surfactant-associated proteins in lung and serum before and after onset of ARDS.Am J Respir Crit Care Med. 1999; 160: 1843-1850Crossref PubMed Scopus (368) Google Scholar As controls, BALF was obtained from six normal volunteers. Bronchoalveolar lavage (BAL) was performed as previously described.7Matute-Bello G Liles WC Radella 2nd, F Steinberg KP Ruzinski JT Jonas M Chi EY Hudson LD Martin TR Neutrophil apoptosis in the acute respiratory distress syndrome.Am J Respir Crit Care Med. 1997; 156: 1969-1977Crossref PubMed Scopus (344) Google Scholar, 8Steinberg KP Milberg JA Martin TR Maunder RJ Cockrill BA Hudson LD Evolution of bronchoalveolar cell populations in the adult respiratory distress syndrome.Am J Respir Crit Care Med. 1994; 150: 113-122Crossref PubMed Scopus (339) Google Scholar, 9Greene KE Wright JR Steinberg KP Ruzinski JT Caldwell E Wong WB Hull W Whitsett JA Akino T Kuroki Y Nagae H Hudson LD Martin TR Serial changes in surfactant-associated proteins in lung and serum before and after onset of ARDS.Am J Respir Crit Care Med. 1999; 160: 1843-1850Crossref PubMed Scopus (368) Google Scholar Briefly, five separate 30-ml aliquots of 0.89% sterile saline were instilled into the right middle lobe or lingula. The BAL recovery averaged 75 ml (49% return) and was not statistically different between the different ARDS groups by one-way analysis of variance (P > 0.05). BALF was centrifuged immediately after collection, and cell-free supernatants were aliquoted into polypropylene tubes and stored at −70°C. Total protein measurements were made on aliquots of supernatants using a modified Lowry method.10Lowry O Rosebrough N Farr A Randall R Protein measurement with the Folin-Phenol reagents.J Biol Chem. 1951; 193: 265-275Abstract Full Text PDF PubMed Google Scholar BALF proteins were concentrated by ice-cold acetone precipitation. BALF containing 2 mg of protein underwent digestion with trypsin (20 μg, sequencing grade; Promega, Madison, WI) overnight at 37°C to allow complete digestion. To prepare for strong-cation exchange chromatography and to reduce the salt concentration, the resulting peptide solutions were diluted eightfold with running buffer (5 mmol/L KH2PO4, 25% acetonitrile, pH 3), and their pH was reduced to ∼2.9 with phosphoric acid (H3PO4). The peptide solutions were passed over a 2.1 × 200 mm, 5-μm particle, 300-Å pore Polysulfoethyl A column (PolyLC; Columbia, MD), washed with running buffer, and then eluted with a 50-minute biphasic gradient of 0 to 25% elution buffer (running buffer plus 350 mmol/L potassium chloride) in 0 to 30 minutes followed by 25 to 100% elution buffer in 30 to 50 minutes. Flow rate was constant at 0.2 ml/minute. Sixteen 2-minute (0.4-ml) fractions were collected. Fractions from strong-cation exchange chromatography were completely dried down in a Speed-Vac (Thermo-Savant, Milford, MA) and redissolved in 0.1% trifluoroacetic acid. To desalt, fractions were loaded onto Oasis mixed-mode cation-exchange cartridges (Waters, Milford, MA), washed with 0.1% tri-fluoroacetic acid, and eluted with 0.1% trifluoroacetic acid, 80% acetonitrile solution. The samples were again dried down and redissolved in 0.2% acetic acid and transferred to autosampler vials for liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.8Steinberg KP Milberg JA Martin TR Maunder RJ Cockrill BA Hudson LD Evolution of bronchoalveolar cell populations in the adult respiratory distress syndrome.Am J Respir Crit Care Med. 1994; 150: 113-122Crossref PubMed Scopus (339) Google Scholar, 23Pala L Giannini S Rosi E Cresci B Scano G Mohan S Duranti R Rotella CM Direct measurement of IGF-I and IGFBP-3 in bronchoalveolar lavage fluid from idiopathic pulmonary fibrosis.J Endocrinol Invest. 2001; 24: 856-864Crossref PubMed Scopus (23) Google Scholar Briefly, an LCQ DECA ion trap mass spectrometer (Thermo Finnigan, San Jose, CA) outfitted with a microelectrospray source (Brechbuehler, Houston, TX) and an HP1100 solvent delivery system (Agilent, Palo Alto, CA) was used to analyze the samples. The samples were automatically delivered by a FAMOS autosampler (LC Packings, San Francisco, CA) to a 100-μm internal diameter, fused silica capillary precolumn packed with 2 cm of 200-Å pore-size Magic C18AQ material (Michrom Bioresources, Auburn, CA) as described elsewhere.11Yi EC Lee H Aebersold R Goodlett DR A microcapillary trap cartridge-microcapillary high-performance liquid chromatography electrospray ionization emitter device capable of peptide tandem mass spectrometry at the attomole level on an ion trap mass spectrometer with automated routine operation.Rapid Commun Mass Spectrom. 2003; 17: 2093-2098Crossref PubMed Scopus (95) Google Scholar The samples were washed with solvent A (0.1% formic acid, 5% acetonitrile) on the precolumn and then eluted with a gradient of 10 to 35% solvent B (100% acetonitrile) over 128.5 minutes to a 75-μm × 14-cm fused silica capillary column packed with 100-Å pore-size Magic C18AQ material and then into the mass spectrometer at a constant column-tip flow rate of ∼300 nL/minute. Peptides entering the mass spectrometer were selected for collision induced dissociation by data-dependent methods, and resultant tandem mass spectra were used to obtain protein matches. SEQUEST12Link AJ Eng J Schieltz DM Carmack E Mize GJ Morris DR Garvik BM Yates 3rd, JR Direct analysis of protein complexes using mass spectrometry.Nature Biotechnol. 1999; 17: 676-682Crossref PubMed Scopus (2077) Google Scholar was used to screen tandem mass spectra for matches to peptide sequence by searching against the human International Protein Index database (European Bioinformatics Institute, Cambridge, UK). PeptideProphet and ProteinProphet were used to verify correctness of peptide and protein assignments,5Keller A Nesvizhskii AI Kolker E Aebersold R Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search.Anal Chem. 2002; 74: 5383-5392Crossref PubMed Scopus (3931) Google Scholar respectively, and those that displayed Prophet scores of >0.9 were considered identified. To compare different experiments, data were imported into SBEAMS, a relational database management systems that allows comparison across multiple experiments.13Baliga NS Pan M Goo YA Yi EC Goodlett DR Dimitrov K Shannon P Aebersold R Ng WV Hood L Coordinate regulation of energy transduction modules in Halobacterium sp. analyzed by a global systems approach.Proc Natl Acad Sci USA. 2002; 99: 14913-14918Crossref PubMed Scopus (116) Google Scholar Gene Ontology classifications were used for functional annotations of described genes.14Hosack DA Dennis Jr, G Sherman BT Lane HC Lempicki RA Identifying biological themes within lists of genes with EASE.Genome Biol. 2003; 4: R70Crossref PubMed Google Scholar Cytokine measurements were performed in duplicate by ELISA using commercially available kits (IGFBP-3 and heparin-binding EGF-like growth factor (HB-EGF) ELISA; R&D Systems, Minneapolis, MN, and total and free IGF ELISA; Diagnostic Standards Laboratory, Webster, TX). Human serum albumin, β2-microglobulin (Alpha Diagnostic International, Inc., San Antonio, TX), surfactant-D and Clara cell protein (Biovendor LLC, Candler, NC), and fibrinogen (DiaPharma Group, Inc., West Chester, OH) were also detected with commercially available ELISA kits. Concentrations were extrapolated from simultaneously run standard curves. Differences between experimental conditions and normal controls were assessed with the Mann-Whitney Test using VasserStats software. The Spearman rank order correlation coefficient was determined for ELISA concentrations and total protein concentrations (VasserStats). All tests were two-tailed, and P values of <0.05 were considered significant. To detect proteolytic fragments of IGFBP-3, equal volumes of BALF samples were separated by 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to Immobilon, and blocked for 1 hour. Blots were incubated with polyclonal IGFBP-3 antibody (Diagnostic Standards Laboratory), which recognizes the major proteolytic fragments of IGFBP-3, for 1 hour, followed by peroxidase-conjugated secondary antibody for 1 hour, and then developed with ECL. To quantitate relative band intensities, gels were captured with Photoshop (version 7; Adobe Systems Inc., San Jose, CA) and then imported into ImageJ (version 1.30; National Institutes of Health) for analysis.15Rasband WS: ImageJ. http://rsb.info.nih.gov/ij/. Bethesda, MD, US National Institutes of Health, 1997–2005Google Scholar Primary normal human lung fibroblasts (Clonetics, Cambrex Bioscience, Rockland, ME), primary distal lung human epithelial cells (Clonetics) or A549 (American Type Culture Collection, Manassas, VA) were seeded in a 96-well tissue culture plate (2 × 104 cells/well) overnight and serum-starved for 24 hours. Function blocking antibody to the human type 1 IGF receptor (A12), a generous gift from ImClone Systems (New York, NY),16Burtrum D Zhu Z Lu D Anderson DM Prewett M Pereira DS Bassi R Abdullah R Hooper AT Koo H Jimenez X Johnson D Apblett R Kussie P Bohlen P Witte L Hicklin DJ Ludwig DL A fully human monoclonal antibody to the insulin-like growth factor I receptor blocks ligand-dependent signaling and inhibits human tumor growth in vivo.Cancer Res. 2003; 63: 8912-8921PubMed Google Scholar, 17Wu JD Odman A Higgins LM Haugk K Vessella R Ludwig DL Plymate SR In vivo effects of the human type I insulin-like growth factor receptor antibody A12 on androgen-dependent and androgen-independent xenograft human prostate tumors.Clin Cancer Res. 2005; 11: 3065-3074Crossref PubMed Scopus (143) Google Scholar was added to media at indicated concentration for 24 hours. In some experiments, soluble Fas ligand (Alexis Biochemicals, Lausen, Switzerland) or IGFBP-3 (R&D Systems) was added to cells. To assess the contribution of IGF to fibroblast survival in ARDS, BALF was diluted 1:10 with Dulbecco's modified Eagle's medium and then incubated with a neutralizing polyclonal antibody (5 ng/ml) to human IGF-I (R&D Systems) or preimmune goat serum for 30 minutes at 4°C before incubation with normal human lung fibroblast. After 48 hours, apoptosis was measured. To prevent detached cells from being aspirated, plates were centrifuged at 200 × g for 10 minutes, and apoptosis was measured using the Cell Death Detection ELISA-plus System (Roche Applied Science, Penzberg, Germany), which detects cytosolic histone-complexed DNA fragments. All experiments were done in triplicate and repeated at least twice. The data are reported as the mean absorbance of triplicate wells mean ± SE or as apoptosis index, defined as the ratio of the mean absorbance of triplicate wells in the experimental condition OD405 nm/control (media alone) OD405 nm. BALF samples from three patients with acute lung injury were analyzed by shotgun proteomics, and a total of 870 unique proteins was identified (downloadable data, including spectra and searchable SEQUEST files, are available at http://www.peptideatlas.org/contributors, authors L.M. Schnapp and S. Donohoe). The numbers of identifications from individual samples were 226, 291, and 659 (Figure 1). Of these, 79 proteins were identified in all three samples (Supplementary Table 1 at http://ajp.amjpathol.org). These identifications represent approximately 10-fold increase in the proteins previously identified in BALF.18Noel-Georis I Bernard A Falmagne P Wattiez R Database of bronchoalveolar lavage fluid proteins.J Chromatogr B Analyt Technol Biomed Life Sci. 2002; 771: 221-236Crossref PubMed Scopus (115) Google Scholar Our approach identified similar classes of proteins to those previously reported using two-dimensional electrophoresis (2DE).18Noel-Georis I Bernard A Falmagne P Wattiez R Database of bronchoalveolar lavage fluid proteins.J Chromatogr B Analyt Technol Biomed Life Sci. 2002; 771: 221-236Crossref PubMed Scopus (115) Google Scholar Of the 79 proteins common to all three patients, proteins from all cellular compartments were identified, including membrane proteins, cytosolic proteins, nuclear proteins, and cytoskeletal proteins, as well as extracellular and secreted proteins. As expected, we identified albumin in all three samples.19Hirsch J Hansen KC Burlingame AL Matthay MA Proteomics: current techniques and potential applications to lung disease.Am J Physiol. 2004; 287: L1-L23PubMed Google Scholar However, the number of peptides corresponding to albumin varied widely in the patients ranging from 1 to 23 to 79, suggesting that the degree of serum protein leakage varied widely even among these three patients with clinically diagnosed ARDS. We also found extensive coverage of other abundant serum proteins, such as ceruloplasmin (average number of peptides: 13) and fibrinogen α chain (average number of peptides: 117), and other acute phase reactant proteins, such as α1 chymotrypsin (average number of peptides: 31), α2-HS-glycoprotein (average number of peptides: 58), and anti-trypsin inhibitor (average number of peptides: 77). We also identified a number of intracellular proteins, as has been previously reported in lung BAL,18Noel-Georis I Bernard A Falmagne P Wattiez R Database of bronchoalveolar lavage fluid proteins.J Chromatogr B Analyt Technol Biomed Life Sci. 2002; 771: 221-236Crossref PubMed Scopus (115) Google Scholar presumably due to increased cellular turnover and death in lung compartment during lung injury. We confirmed the presence of several serum proteins (albumin, fibrinogen, and β2-microglobulin) and pulmonary proteins (surfactant D and Clara cell protein) by ELISA. Surfactant A2 was also identified, as previously reported by 2DE.18Noel-Georis I Bernard A Falmagne P Wattiez R Database of bronchoalveolar lavage fluid proteins.J Chromatogr B Analyt Technol Biomed Life Sci. 2002; 771: 221-236Crossref PubMed Scopus (115) Google Scholar, 20Bowler RP Duda B Chan ED Enghild JJ Ware LB Matthay MA Duncan MW Proteomic analysis of pulmonary edema fluid and plasma in patients with acute lung injury.Am J Physiol. 2004; 286: L1095-L1104Crossref PubMed Scopus (90) Google Scholar In addition, surfactant B2 (two patients) and surfactant D (one patient) were also identified here but were not found in previous 2DE analysis of BALF. Previous reports speculated that 2DE failed to identify surfactant B2, because of its hydrophobicity, and surfactant D, because of its relative underexpression compared to surfactant A2. Representative spectra are shown in Figure 2. From the long list of proteins identified by LC-MS/MS shotgun proteomics, we decided to focus on secreted proteins, because they may represent mediators of lung injury. This category included identifications of pre-B-cell colony-enhancing factor in two patients, HB-EGF in one patient and IGFBP-3 and the acid labile subunit (ALS) in two patients, both components of the IGF signaling complex (Figure 2). Pre-B-cell colony-enhancing factor was recently described as an inhibitor of apoptosis that was expressed by neutrophils from septic patients.21Jia SH Li Y Parodo J Kapus A Fan L Rotstein OD Marshall JC Pre-B cell colony-enhancing factor inhibits neutrophil apoptosis in experimental inflammation and clinical sepsis.J Clin Invest. 2004; 113: 1318-1327Crossref PubMed Scopus (541) Google Scholar Previous reports showed elevated levels of IGFBP-3 in BALF from patients with idiopathic pulmonary fibrosis (IPF)22Aston C Jagirdar J Lee TC Hur T Hintz RL Rom WN Enhanced insulin-like growth factor molecules in idiopathic pulmonary fibrosis.Am J Respir Crit Care Med. 1995; 151: 1597-1603Crossref PubMed Scopus (76) Google Scholar, 23Pala L Giannini S Rosi E Cresci B Scano G Mohan S Duranti R Rotella CM Direct measurement of IGF-I and IGFBP-3 in bronchoalveolar lavage fluid from idiopathic pulmonary fibrosis.J Endocrinol Invest. 2001; 24: 856-864Crossref PubMed Scopus (23) Google Scholar and sarcoidosis.24Allen JT Bloor CA Knight RA Spiteri MA Expression of insulin-like growth factor binding proteins in bronchoalveolar lavage fluid of patients with pulmonary sarcoidosis.Am J Respir Cell Mol Biol. 1998; 19: 250-258Crossref PubMed Scopus (39) Google Scholar In contrast, HB-EGF has not been previously associated with acute lung injury. Proteomics results do not distinguish between reproducible changes and sampling variability during the comparison of data from three different patients. To assess the potential significance of secreted BALF proteins identified by LC-MS/MS shotgun proteomics, we measured expression levels HB-EGF, IGFBP-3, and IGF-I by ELISA in a large BALF sample set that includes patients at different time points in ARDS progression. BALF samples from normal subjects (n = 6), patients at risk for development of ARDS (n = 8), and established ARDS at day 1 (n = 26), day 3 (n = 20), day 7 (n = 10), and day 14 (n = 5) were analyzed. HB-EGF is a potent mitogen and chemotactic factor for fibroblasts.25Raab G Klagsbrun M Heparin-binding EGF-like growth factor.Biochim Biophys Acta. 1997; 1333: F179-F199PubMed Google Scholar Thus, we hypothesized that it might play a role in the fibroproliferative response during acute lung injury. However, ELISA results revealed very low levels of HB-EGF in ARDS BALF and in normal BALF (Figure 3). Furthermore, we did not observe a correlation between HB-EGF levels and progression of lung injury in ARDS. Failure to detect changes in HB-EGF levels in ARDS BALF does not necessarily preclude a role for HB-EGF in lung injury. For instance lack of correlation by ELISA may be due to complex tissue distribution of the multiple forms of HB-EGF26Iwamoto R Mekada E Heparin-binding EGF-like growth factor: a juxtacrine growth factor.Cytokine Growth Factor Rev. 2000; 11: 335-344Crossref PubMed Scopus (187) Google Scholar making it inaccessible in BALF. However, these data illustrate the importance of independently verifying proteins identified by any proteomic screen. Because of its role in regulating cell survival,27LeRoith D Roberts Jr, CT The insulin-like growth factor system and cancer.Cancer Lett. 2003; 195: 127-137Crossref PubMed Scopus (940) Google Scholar IGFBP-3 is also a potential candidate protein relevant to the pathogenesis of ARDS. While we found very low levels of IGFBP-3 in BALF from normal controls by ELISA, we detected a marked increase in IGFBP-3 concentration in patients at risk for ARDS and in those with established ARDS (Figure 4A). Because earlier work demonstrated elevated levels of IGFBP-3 in BALF from patients with IPF and sarcoidosis,23Pala L Giannini S Rosi E Cresci B Scano G Mohan S Duranti R Rotella CM Direct measurement of IGF-I and IGFBP-3 in bronchoalveolar lavage fluid from idiopathic pulmonary fibrosis.J Endocrinol Invest. 2001; 24: 856-864Crossref PubMed Scopus (23) Google Scholar, 24Allen JT Bloor CA Knight RA Spiteri MA Expression of