Advertisement

Overexpression of Tumor Necrosis Factor-α in the Lungs Alters Immune Response, Matrix Remodeling, and Repair and Maintenance Pathways

Open AccessPublished:February 10, 2012DOI:https://doi.org/10.1016/j.ajpath.2011.12.020
      Increased production of tumor necrosis factor (TNF)-α and matrix metalloproteinases (MMPs) is a feature of inflammatory lung diseases, including emphysema and fibrosis, but the divergent pathological characteristics that result indicate involvement of other processes in disease pathogenesis. Transgenic mice overexpressing TNF-α in type II alveolar epithelial cells under the control of the surfactant protein (SP)-C promoter develop pulmonary inflammation and emphysema but are resistant to induction of fibrosis by administration of bleomycin or transforming growth factor-β. To study the molecular mechanisms underlying the development of this phenotype, we used a microarray approach to characterize the pulmonary transcriptome of SP-C/TNF-α mice and wild-type littermates. Four-month-old SP-C/TNF-α mice displayed pronounced pulmonary inflammation, airspace enlargement, increased MMP-2 and MMP-9 levels, and altered expression of 2332 probes. The functional assessment of genes with increased expression revealed enrichment of inflammatory/immune responses and proteases, whereas genes involved in protease inhibition, angiogenesis, cross-linking of basement membrane proteins, and myofibroblast differentiation were predominantly decreased. Comparison with multiple lung disease models identified a set of genes unique to the SP-C/TNF-α model and revealed that lack of extracellular matrix production distinguished SP-C/TNF-α mice from fibrosis models. Activation of inflammatory and proteolytic pathways and disruption of maintenance and repair processes are central features of emphysema in this TNF-overexpression model. Impairment of myofibroblast differentiation and extracellular matrix production may underlie resistance to induction of fibrosis.
      The pulmonary immune system protects the lungs from the harmful effects of inhaled xenobiotics and pathogens, in part through the release of pro-inflammatory mediators. These mediators activate professional phagocytes that work to clear the lungs of contaminating material. Although transient activation of the inflammatory system can be effective in responding to immune challenges in the lungs without causing lasting injury, chronic inflammatory processes can contribute to destruction of lung architecture, even after elimination of the noxious agent, ultimately leading to pathological conditions, such as emphysema. For example, increased levels of macrophages and neutrophils persist in the lungs of former smokers with chronic obstructive pulmonary disorder, even after smoking cessation, resulting in sustained increases in the production of proteolytic enzymes.
      • Owen C.A.
      Roles for proteinases in the pathogenesis of chronic obstructive pulmonary disease.
      Although an imbalance of proteases and protease inhibitors is thought to be a primary cause of the substantial remodeling and destruction of alveolar walls that occur in emphysema,
      • Churg A.
      • Wright J.L.
      Proteases and emphysema.
      other processes, including apoptosis,
      • Henson P.M.
      • Vandivier R.W.
      • Douglas I.S.
      Cell death, remodeling, and repair in chronic obstructive pulmonary disease.
      cellular senescence,
      • Tsuji T.
      • Aoshiba K.
      • Nagai A.
      Alveolar cell senescence in patients with pulmonary emphysema.
      and dysfunctional repair systems,
      • Henson P.M.
      • Vandivier R.W.
      • Douglas I.S.
      Cell death, remodeling, and repair in chronic obstructive pulmonary disease.
      • Kasahara Y.
      • Tuder R.M.
      • Taraseviciene-Stewart L.
      • Le Cras T.D.
      • Abman S.
      • Hirth P.K.
      • Waltenberger J.
      • Voelkel N.F.
      Inhibition of VEGF receptors causes lung cell apoptosis and emphysema.
      have also been postulated to play a role. It is clear that several interrelated cellular and molecular events contribute to pathological conditions resulting from a chronic inflammatory state in the lungs.
      • Wright J.L.
      • Churg A.
      Current concepts in mechanisms of emphysema.
      Understanding how diverse factors contribute collectively to the development of lung pathological conditions remains an active area of research.
      Models that allow disease phenotypes to be examined and compared are useful to gain a better understanding of processes underlying pulmonary diseases and to characterize the role of specific agents in disease pathogenesis.
      • Lewis C.C.
      • Yang J.Y.
      • Huang X.
      • Banerjee S.K.
      • Blackburn M.R.
      • Baluk P.
      • McDonald D.M.
      • Blackwell T.S.
      • Nagabhushanam V.
      • Peters W.
      • Voehringer D.
      • Erle D.J.
      Disease-specific gene expression profiling in multiple models of lung disease.
      Tumor necrosis factor (TNF)-α is a pleiotropic cytokine implicated in several lung pathological conditions, including fibrosis, emphysema, asthma, and connective tissue breakdown associated with cigarette smoking.
      • Churg A.
      • Wang R.D.
      • Tai H.
      • Wang X.
      • Xie C.
      • Wright J.L.
      Tumor necrosis factor-alpha drives 70% of cigarette smoke-induced emphysema in the mouse.
      • Mukhopadhyay S.
      • Hoidal J.R.
      • Mukherjee T.K.
      Role of TNFalpha in pulmonary pathophysiology.
      • Piguet P.F.
      • Collart M.A.
      • Grau G.E.
      • Sappino A.P.
      • Vassalli P.
      Requirement of tumour necrosis factor for development of silica-induced pulmonary fibrosis.
      To evaluate the involvement of TNF-α in inflammatory lung disease, transgenic mice that constitutively express TNF-α in alveolar type II epithelial cells under the control of the surfactant protein (SP)-C promoter were generated.
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      These mice develop chronic inflammation and airspace enlargement.
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      • Fujita M.
      • Shannon J.M.
      • Irvin C.G.
      • Fagan K.A.
      • Cool C.
      • Augustin A.
      • Mason R.J.
      Overexpression of tumor necrosis factor-alpha produces an increase in lung volumes and pulmonary hypertension.
      On the basis of the interstitial inflammation and increased collagen deposition, SP-C/TNF-α mice were initially characterized as a model of idiopathic pulmonary fibrosis.
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      However, subsequent work by Fujita et al
      • Fujita M.
      • Shannon J.M.
      • Irvin C.G.
      • Fagan K.A.
      • Cool C.
      • Augustin A.
      • Mason R.J.
      Overexpression of tumor necrosis factor-alpha produces an increase in lung volumes and pulmonary hypertension.
      demonstrated that the transgenic mice were characterized by large lung volumes and decreased elastic recoil, a phenotype more similar to emphysema. Moreover, although hydroxyproline content was increased per lung, there was no increase when values were normalized to lung weight, suggesting that the effect was related to an increase in the size of the lungs, rather than to fibrosis.
      • Fujita M.
      • Shannon J.M.
      • Irvin C.G.
      • Fagan K.A.
      • Cool C.
      • Augustin A.
      • Mason R.J.
      Overexpression of tumor necrosis factor-alpha produces an increase in lung volumes and pulmonary hypertension.
      Physiological and morphometric imaging analyses confirmed that this model displayed features of emphysema, also revealing some consolidation of airspaces consistent with fibrosis.
      • Lundblad L.K.
      • Thompson-Figueroa J.
      • Leclair T.
      • Sullivan M.J.
      • Poynter M.E.
      • Irvin C.G.
      • Bates J.H.
      Tumor necrosis factor-alpha overexpression in lung disease: a single cause behind a complex phenotype.
      Remarkably, although TNF-α has been implicated in the pathogenesis of fibrosis in humans,
      • Piguet P.F.
      • Ribaux C.
      • Karpuz V.
      • Grau G.E.
      • Kapanci Y.
      Expression and localization of tumor necrosis factor-alpha and its mRNA in idiopathic pulmonary fibrosis.
      • Grutters J.C.
      • du Bois R.M.
      Genetics of fibrosing lung diseases.
      and anti–TNF-α antibodies attenuate bleomycin- or silica-induced fibrosis in animals,
      • Piguet P.F.
      • Collart M.A.
      • Grau G.E.
      • Sappino A.P.
      • Vassalli P.
      Requirement of tumour necrosis factor for development of silica-induced pulmonary fibrosis.
      • Piguet P.F.
      • Collart M.A.
      • Grau G.E.
      • Kapanci Y.
      • Vassalli P.
      Tumor necrosis factor/cachectin plays a key role in bleomycin-induced pneumopathy and fibrosis.
      the lungs of SP-C/TNF-α mice are resistant to initiation of fibrosis by instillation of transforming growth factor (TGF)-β or bleomycin.
      • Fujita M.
      • Shannon J.M.
      • Morikawa O.
      • Gauldie J.
      • Hara N.
      • Mason R.J.
      Overexpression of TNF-alpha diminishes pulmonary fibrosis induced by bleomycin or TGF-beta.
      Together, the data suggest that factors other than TNF-α alone are important in determining the fate of injured lungs.
      Given the pleiotropic nature of TNF-α signaling and its association with several lung pathological conditions with differing phenotypes, an understanding of the molecular pathways modulated by TNF-α expression in the lungs should be of use in characterizing disease pathogenesis and identifying potential therapeutic targets for further investigation. Global screening approaches offer powerful tools to assess biological pathways that drive disease phenotypes.
      • Lewis C.C.
      • Yang J.Y.
      • Huang X.
      • Banerjee S.K.
      • Blackburn M.R.
      • Baluk P.
      • McDonald D.M.
      • Blackwell T.S.
      • Nagabhushanam V.
      • Peters W.
      • Voehringer D.
      • Erle D.J.
      Disease-specific gene expression profiling in multiple models of lung disease.
      To gain a better understanding of the molecular processes underlying the chronic inflammatory state and resistance to induction of fibrosis of the SP-C/TNF-α mice, in the present study, we used a microarray approach to compare the level of expression of approximately 21,000 genes in the lungs of TNF-overexpressing mice and wild-type littermates. We then contrasted the SP-C/TNF-α lung transcriptome with 12 other models of lung disease to identify key molecular pathways that distinguish disease phenotypes. We report that constitutive expression of TNF-α in the lungs results in several effects on critical pathways implicated in the pathogenesis of emphysema, including increased expression of genes involved in inflammation and matrix remodeling and reduced expression of genes involved in lung repair, maintenance pathways, and myofibroblast differentiation.

      Materials and Methods

      Animals

      SP-C/TNF-α male mice (C57BL/6 background)
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      were originally provided by Dr. Robert Mason (National Jewish Medical and Research Center, Denver, CO). The SP-C/TNF-α male mice were crossed with C57BL/6 female mice (Charles River Laboratories, St. Constant, QC, Canada) and maintained as a heterozygous line by repeated backcrossing. Male transgenic mice and their wild-type littermates were genotyped by PCR analysis of genomic DNA isolated from ear punches, as previously described.
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      Animals were aged 132 ± 5 days at the time of the experiment. Mice were housed in individual Plexiglas cages on wood-chip bedding under high-efficiency particulate filtered air and held to a 12-hour dark-light cycle. Food and water were provided ad libitum. All experimental protocols were reviewed and approved by the Animal Care Committee of Health Canada.

      Biological Samples

      Mice were anesthetized by administration of sodium pentobarbital (60 mg/kg, i.p.). For histological examination, lungs (n = 3 per genotype) were inflated at 25 cm H2O static pressure by intratracheal instillation of 4% paraformaldehyde in PBS. The lungs were excised, immersed in fixative, and stored at 4°C. Tissue blocks were dehydrated in ethanol and embedded in paraffin. Sections (0.75 μm thick) were stained with H&E. For microarray analyses, five mice per genotype were used, whereas for real-time PCR analyses, lavage cytological examination, and gelatin zymography, an independent set of four mice per genotype were used. Lungs were washed by bronchoalveolar lavage with saline (37°C) at 30 mL/kg body weight, flash frozen in liquid nitrogen, and stored at −80°C. Lavage fluid was centrifuged (400 × g for 10 minutes at 4°C) to remove cells and frozen at −80°C. Cells were counted using a Coulter Multisizer II (Beckman Coulter Canada Inc., Mississauga, ON, Canada), and differential cell counts were obtained from cytospin preparations.

      Gelatin Zymography

      Bronchoalveolar lavage fluid from SP-C/TNF-α and wild-type mice was evaluated for matrix metalloproteinase (MMP) activity by gelatin zymography. Equal quantities of bronchoalveolar lavage fluid (20 μL) were loaded on 10% SDS-acrylamide gels containing 1 mg/mL gelatin (Sigma-Aldrich Canada Ltd., Oakville, ON, Canada) and run for 1 hour at 200 mV. Each gel contained prestained mol. wt. markers (Bio-Rad Laboratories Canada Ltd., Mississauga, ON, Canada) and MMP-2 standards (Calbiochem, La Jolla, CA). Gels were incubated in Zymogram Renaturation Buffer (Bio-Rad Laboratories Canada Ltd.) for 30 minutes, followed by overnight incubation at 37°C in Zymogram Development Buffer (Bio-Rad Laboratories Canada Ltd.). After incubation, gels were stained in 0.5% Coomassie Blue R-250 (Sigma-Adrich Canada Ltd.) staining solution (in 40% methanol/10% acetic acid) for 1 hour and then destained in a solution of 40% methanol/10% acetic acid. Clear bands were assessed by densitometric analysis using NIH shareware. To verify MMP activity, control gels were incubated under the same conditions in buffer containing 25 mmol/L EDTA.

      RNA Isolation

      Frozen lung samples were homogenized in TRIzol reagent (Invitrogen Canada Inc., Burlington, ON, Canada), and total RNA was isolated according to the manufacturer's instructions. The isolated RNA was further purified by spin-column cleanup using RNeasy Mini Kits (Qiagen Inc., Mississauga, ON, Canada). RNA was quantified using the RiboGreen RNA Quantitation Reagent and Kit (Molecular Probes, Eugene, OR), and quality was confirmed using the RNA 6000 NanoLab Chip Kit (Agilent Technologies Canada Inc., Mississauga, ON, Canada).

      Microarray Analysis

      Archived microarray data for age-matched transgenic animals and wild-type littermates (n = 5 per genotype) from our inhalation facility
      • Thomson E.M.
      • Williams A.
      • Yauk C.L.
      • Vincent R.
      Toxicogenomic analysis of susceptibility to inhaled urban particulate matter in mice with chronic lung inflammation.
      were used for assessment of the lung transcriptome. Microarray data are available at the National Center for Biotechnology Information Gene Expression Omnibus (accession no. GSE11037; http://www.ncbi.nlm.nih.gov/projects/geo). Briefly, 2.5 μg of RNA was amplified and labeled using the Low RNA Input Fluorescent Linear Amplification Kit (Agilent Technologies Canada Inc.), according to the manufacturer's instructions. Agilent Mouse G4121A Microarrays were hybridized with 5 μg of Cy5-labeled lung RNA (one lung per array). Cy3-labeled Universal Mouse Reference RNA (Agilent Technologies Canada Inc.; 5 μg) was used as a common reference on all arrays. A randomized block design was used for the order of sample hybridizations. Arrays were scanned using a ScanArray Express scanner (Perkin-Elmer Life Sciences, Woodbridge, ON, Canada), and data were acquired with ImaGene 5.5 (BioDiscovery, El Segundo, CA). Lowess normalization
      • Yang Y.H.
      • Dudoit S.
      • Luu P.
      • Lin D.M.
      • Peng V.
      • Ngai J.
      • Speed T.P.
      Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation.
      was performed using SAS/STAT software, version 8.2 of the SAS System for Windows (1999–2001; SAS Institute Inc., Cary, NC). The logarithm base 2 relative intensities were used for subsequent analyses, and the MAANOVA library
      • Wu H.
      • Kerr M.K.
      • Cui X.
      • Churchill G.A.
      MAANOVA: a software package for the analysis of spotted cDNA microarray experiments.
      in R was used to identify differentially expressed genes. The Fs statistic
      • Cui X.G.
      • Hwang J.T.G.
      • Qiu J.
      • Blades N.J.
      • Churchill G.A.
      Improved statistical tests for differential gene expression by shrinking variance components.
      was used as a shrinkage estimator for the gene-specific variance components, and P values for all statistical tests were estimated using the permutation method (1000 permutations with residual shuffling). These P values were then adjusted for multiple comparisons using the false-discovery rate (FDR) approach,
      • Benjamini Y.
      • Hochberg Y.
      Controlling the false discovery rate: a practical and powerful approach to multiple testing.
      and genes were considered differentially expressed if they had an FDR-adjusted P < 0.05. The group means for the fold-change calculation (transgenic versus wild type) were based on the adjusted relative intensity for each gene after subtraction of nuisance factors (eg, day of hybridization) from the normalized ratio. Gene ontology enrichment analysis and functional annotation clustering were performed using the Database for Annotation, Visualization and Integrated Discovery
      • Dennis Jr, G.
      • Sherman B.T.
      • Hosack D.A.
      • Yang J.
      • Gao W.
      • Lane H.C.
      • Lempicki R.A.
      DAVID: Database for Annotation, Visualization, and Integrated Discovery.
      • Huang da W.
      • Sherman B.T.
      • Lempicki R.A.
      Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources.
      (http://david.abcc.ncifcrf.gov/home.jsp) using the array probe list as the background population. Analysis of KEGG pathways (Kyoto Encyclopedia of Genes & Genomes online resource, http://www.genome.ad.jp/kegg/pathway.html) was also performed.

      Comparison of Transcriptomes from Lung Disease Models

      Microarray data for 12 murine models of lung disease were obtained from the National Center for Biotechnology Information Gene Expression Omnibus site (accession no. GSE4231; http://www.ncbi.nlm.nih.gov/projects/geo). Because five different platforms were used in the analyses, GenBank accession identification numbers were used to identify probes common to all platforms. On arrays that contained multiple probes for a given GenBank accession identification number, the lowess normalized
      • Yang Y.H.
      • Dudoit S.
      • Luu P.
      • Lin D.M.
      • Peng V.
      • Ngai J.
      • Speed T.P.
      Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation.
      log2 of the ratios was averaged. Data were then merged across the five platforms, yielding 7148 unique GenBank accession identification numbers. Differentially expressed genes were identified as described for the SP-C/TNF-α arrays. Because of low power to detect significant effects for several of the models, a less stringent statistical filter of a 1.5-fold cutoff and an unadjusted P < 0.05 was used. Hierarchical clustering was conducted using the top 50 genes with an unadjusted P < 0.05 ranked by fold change for each disease model, excluding the SP-C/TNF-α model. The distance metric was one minus the Pearson correlation, and average linkage was used.

      Real-Time PCR

      Real-time PCR was performed on samples used for microarray experiments (n = 5 per genotype) or on independent biological samples from age-matched animals (n = 4 per genotype). cDNA, 20 ng, was combined with Quantitect (Qiagen Inc.) primer assays and real-time iQ SYBR Green supermix (Bio-Rad Laboratories Canada Ltd.), according to the product protocol, and run at an annealing temperature of 55°C on the Lightcycler 480 (Roche Diagnostics, Laval, QC, Canada). Fluorescence was monitored at every cycle during the 72°C elongation step. Postrun melt curves were routinely inspected to verify product purity. Expression was calculated relative to the TATA-box binding protein, which exhibited unchanged expression between genotypes (data not shown), using the ΔΔCT method. Statistical significance was assessed by t-test (Sigma-Stat version 3.0; SPSS Inc., Chicago, IL).

      Results

      SP-C/TNF-α Mouse Lung Phenotype

      Histological assessment of the inflated lungs confirmed that SP-C/TNF-α mice bred in our laboratory had enlarged alveolar spaces and thickened septa compared with wild-type littermates (Figure 1, A–D), as expected.
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      • Fujita M.
      • Shannon J.M.
      • Irvin C.G.
      • Fagan K.A.
      • Cool C.
      • Augustin A.
      • Mason R.J.
      Overexpression of tumor necrosis factor-alpha produces an increase in lung volumes and pulmonary hypertension.
      Increased cellularity was apparent in both the alveoli and the interstitium. Bronchoalveolar lavage recovered 10-fold higher numbers of cells in the transgenic mice compared with wild-type animals (Figure 1E), consisting of an eightfold increase in macrophages and increased numbers of neutrophils and lymphocytes (Figure 1F).
      Figure thumbnail gr1
      Figure 1Lung phenotype of SP-C/TNF-α mice. Lungs were inflated to 25 cm H2O static pressure by intratracheal instillation of 4% paraformaldehyde in PBS. Sections from the lungs of wild-type (A and C) and transgenic (B and D) mice were stained with H&E. Original magnification: ×40 (A and B); ×100 (C and D). E: Total number of cells recovered in bronchoalveolar lavage fluid (n = 4 per genotype). F: Differential count of cells recovered by bronchoalveolar lavage (BAL; n = 4 per genotype). *P < 0.05, t-test. EOS, eosinophils; LYM, lymphocytes; MAC, macrophages; M MAC, mature MAC; MNGC, multinucleated giant cells; NEU, neutrophils; WT, wild type.

      Levels of Gelatin-Degrading Enzymes

      Because MMPs are implicated in basement membrane remodeling and airspace enlargement, we examined MMP levels in bronchoalveolar lavage fluid by gelatin zymography. Clear bands representing areas of enzymatic activity were detected at 72 and 105 kDa in SP-C/TNF-α mice (Figure 2), corresponding to the expected mol. wt. of pro-MMP-2 and pro-MMP-9, respectively, whereas only a faint band at 72 kDa was observed in wild-type animals. The enzyme activity was inhibited by incubation of gels in buffer containing EDTA (data not shown).
      Figure thumbnail gr2
      Figure 2Increased gelatinase activity in the lungs of SP-C/TNF-α mice. Gelatin zymography was performed on equal volumes of bronchoalveolar lavage fluid recovered from wild-type (WT) and transgenic mice (n = 4 per genotype). A dilution series of MMP-2 standard was run alongside as a positive control. Cleared areas (light bands) indicate gelatinase activity. stnd., standard.

      Differential Gene Expression

      Comparison of the lung transcriptome of wild-type and transgenic animals resulted in the identification of 2332 differentially expressed probes (FDR-adjusted P < 0.05; see Supplemental Table S1 at http://ajp.amjpathol.org). Of the differentially expressed probes, 1283 (55%) were increased and 1049 (45%) were decreased. Fold changes ranged from 34-fold increased to sevenfold decreased expression relative to wild-type animals, with most genes (67%) exhibiting a fold change of less than two and 95% exhibiting a fold change of less than four (summarized in Table 1). Cluster analysis performed on unfiltered transcriptional profiles revealed a clear genotype effect, with samples from transgenic and wild-type animals separated on two main branches (Figure 3A). To generate a more conservative list of significant genes, a twofold cutoff filter was applied to the list of significant genes (Figure 3B). This filtering resulted in a list of 760 probes, of which 472 (62%) were increased and 288 (38%) were decreased. Probes with the greatest increase in expression in the TNF lungs included Ig genes [eg, Ig-joining chain, 34-fold; Ig κ chain variable 38, 17-fold; Ig heavy chain 1a (serum IgG2a), 16-fold; Ig λ chain, variable 1, 12-fold], consistent with the lymphocytic infiltration; the acute-phase response proteins serum amyloid A3 (30-fold) and serum amyloid A1 (10-fold); C-type lectin domain family 4, member d (17-fold); and TNF-α (17-fold). Probes with the greatest decrease in expression included procollagen C-terminal enhancer protein 2a (sevenfold), a glycoprotein that potentiates collagen cleavage; genes involved in biotransformation of endogenous and exogenous compounds, including cytochrome P450, family 1, subfamily a, polypeptide 1 (sixfold), sulfotransferase family 1D, member 1 (fivefold), and flavin-containing monooxygenase 3 (fourfold); cytochrome P450, family 2, subfamily s, polypeptide 1 (fourfold); the scaffolding protein tetraspanin 7 (fivefold); insulin-like growth factor binding proteins 2 (threefold) and 6 (threefold); and the tissue inhibitor of metalloproteinase-3 (TIMP-3; threefold). Differential expression of a panel of seven probes chosen to represent a range of fold changes (threefold decrease to fivefold increase relative to wild type) was confirmed by real-time PCR (t-test, P < 0.05 for all genes; see Supplemental Figure S1 at http://ajp.amjpathol.org). In general, the difference in gene expression between genotypes was greater according to the PCR analysis than was observed in the microarray data, as is common with microarray analyses, but the direction of change was consistent.
      Table 1Fold-Change Distribution of Differentially Expressed Probes (SP-C/TNF-α versus Wild-Type) with FDR-Adjusted P < 0.05
      Fold-change rangeTotal probesIncreased expressionDecreased expression
      1–1.5775425350
      1.5–2797386411
      2–4646369277
      4–668599
      6–826242
      8–10440
      >1016160
      Figure thumbnail gr3
      Figure 3Comparison of the lung transcriptome of wild-type (WT) and SP-C/TNF-α mice. Cy5-labeled lung RNA from WT and transgenic mice (n = 5 per genotype) was hybridized to individual oligonucleotide arrays, scanned, and analyzed as described in Materials and Methods. A: Clustering of the unfiltered data set segregated samples according to genotype. Red indicates a high level of expression relative to the reference RNA, whereas green indicates low relative expression. The sample identities are indicated at the bottom of the condition tree. B: To generate a more conservative list, a twofold cutoff filter (indicated by the two red lines) was applied, resulting in 760 probes passing this filter. Relevant non-Ig genes discussed in Results are identified. ADN, adipsin; CCL, chemokine (C-C motif) ligand; CNN, calponin; CTGF, connective tissue growth factor; CTS, cathepsin; CYP, cytochrome P450; FMO, flavin-containing monooxygenase; GSN, gelsolin; HEGFL, heparin-binding epidermal growth factor-like growth factor; PCOLCE, procollagen C-endopeptidase enhancer; RECK, reversion-inducing cysteine-rich protein with kazal motifs; SA, serum amyloid; SULT, sulfotransferase.

      Functional Analysis

      Functional analysis was performed on the list of 760 probes with FDR-adjusted P < 0.05 and a greater than twofold difference in gene expression. Functional analysis of terms related to cellular localization revealed that the products of differentially expressed genes were predominantly associated with the extracellular space, plasma membrane, or lysosome (Table 2). The most significant enrichment of terms relating to molecular function was for chemokines, cytokines, and antigen binding (Table 3), consistent with the pronounced inflammatory/immune response. Functional analysis of biological processes showed that the 472 genes with increased expression in the lungs of transgenic mice were strongly enriched for terms related to immune system processes and inflammation (Table 4). In contrast, the 288 genes with decreased expression were enriched for terms relating to blood vessel development and cytoskeletal organization and biogenesis (Table 5).
      Table 2Functional Enrichment of Terms Related to Cellular Localization
      TermFold enrichmentNo. of genes% of totalModified Fisher exact P valueBonferroni-adjusted P value
      Extracellular region1.817023.24.4 × 10153.5 × 1012
      Extracellular space1.815421.09.6 × 10147.5 × 1011
      Extracellular region part1.816021.91.2 × 10138.9 × 1011
      Cell surface4.2375.14.5 × 10133.5 × 1010
      External side of the plasma membrane5.1283.83.7 × 10122.9 × 109
      Lysosome4304.13.3 × 10102.5 × 107
      Lytic vacuole4304.13.3 × 10102.5 × 107
      Vacuole3.7314.29.6 × 10107.4 × 107
      Plasma membrane1.712517.12.4 × 1091.8 × 106
      Membrane1.232544.43.5 × 1062.7 × 103
      Intrinsic to membrane1.225134.31.2 × 1059.1 × 103
      Integral to membrane1.225034.21.3 × 1051.0 × 102
      Plasma membrane part1.69012.31.5 × 1051.1 × 102
      MHC protein complex6.3101.41.9 × 1051.5 × 102
      Membrane part1.227137.02.4 × 1051.9 × 102
      Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation analysis was performed using a list of differentially expressed genes (760 probes and 732 unique DAVID identification numbers) with the Agilent mouse microarray G4121A probe list serving as the background. Enriched terms with a Bonferroni-adjusted P < 0.05 are displayed.
      MHC, major histocompatibility complex.
      Table 3Functional Analysis of Molecular Function
      TermFold enrichmentNo. of genes% of totalModified Fisher exact P valueBonferroni-adjusted P value
      Chemokine activity10.6172.38.7 × 10132.3 × 109
      Chemokine receptor binding10.4172.31.4 × 10123.6 × 109
      Antigen binding11152.01.5 × 10113.9 × 108
      G-protein–coupled receptor binding6.5172.33.5 × 1099.4 × 106
      Carbohydrate binding2.9334.51.6 × 1074.3 × 104
      Cytokine activity3273.79.4 × 1072.5 × 103
      Receptor binding2506.86.8 × 1061.8 × 102
      Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation analysis was performed using a list of differentially expressed genes (760 probes and 732 unique DAVID identification numbers) with the Agilent mouse microarray G4121A probe list serving as the background. Enriched terms with a Bonferroni-adjusted P < 0.05 are displayed.
      Table 4Functional Analysis for Biological Function of Genes with Higher Lung Expression in SP-C/TNF-α Mice
      TermFold enrichmentNo. of genes% of totalModified Fisher exact P valueBonferroni-adjusted P value
      Immune system process4.99220.66.2 × 10393.2 × 1035
      Immune response6.37216.17.2 × 10373.7 × 1033
      Defense response55712.85.2 × 10242.7 × 1020
      Response to wounding5.4449.91.4 × 10197.3 × 1016
      Inflammatory response6.3378.39.5 × 10194.9 × 1015
      Response to external stimulus4.25311.99.8 × 10195.1 × 1015
      Cell activation5327.22.9 × 10131.5 × 109
      Antigen processing and presentation10.2184.09.4 × 10134.9 × 109
      Response to stimulus1.912026.91.0 × 10125.2 × 109
      Leukocyte activation4.8296.59.2 × 10124.7 × 108
      Taxis7.1214.71.1 × 10115.6 × 108
      Regulation of immune response7.1214.71.1 × 10115.6 × 108
      Chemotaxis7.1214.71.1 × 10115.6 × 108
      Regulation of immune system process7214.71.5 × 10117.8 × 108
      Positive regulation of immune response7.4194.36.4 × 10113.3 × 107
      Positive regulation of immune system process7.3194.39.1 × 10114.7 × 107
      Locomotory behavior4.9265.81.2 × 10106.3 × 107
      Humoral immune response9.8153.42.1 × 10101.1 × 106
      Adaptive immune response based on somatic recombination of immune receptors built from Ig superfamily domains7.1184.04.7 × 10102.4 × 106
      Adaptive immune response7.1184.04.7 × 10102.4 × 106
      Positive regulation of multicellular organismal process5.7194.34.9 × 1092.5 × 105
      Antigen processing and presentation of exogenous peptide antigen18.392.01.1 × 1085.6 × 105
      Activation of immune response7.3153.41.3 × 1086.4 × 105
      Behavior3.5296.52.0 × 1081.0 × 104
      Lymphocyte activation4.2235.23.3 × 1081.7 × 104
      Regulation of multicellular organismal process3.4286.34.5 × 1082.3 × 104
      B-cell–mediated immunity7.3143.14.7 × 1082.4 × 104
      Response to stress2.34911.04.8 × 1082.5 × 104
      Myeloid leukocyte activation12.4102.25.7 × 1082.9 × 104
      Antigen processing and presentation of exogenous antigen14.592.09.5 × 1084.9 × 104
      Immune effector process5184.09.8 × 1085.0 × 104
      Antigen processing and presentation of peptide antigen via MHC class II18.181.81.1 × 1075.8 × 104
      Antigen processing and presentation of exogenous peptide antigen via MHC class II18.181.81.1 × 1075.8 × 104
      Antigen processing and presentation of peptide antigen11.3102.21.4 × 1077.1 × 104
      Antigen processing and presentation of peptide or polysaccharide antigen via MHC class II17.181.81.8 × 1079.3 × 104
      Ig-mediated immune response7132.92.8 × 1071.4 × 103
      Leukocyte-mediated immunity5.7153.43.1 × 1071.6 × 103
      Hemopoietic or lymphoid organ development3.5245.44.8 × 1072.4 × 103
      Mast cell activation19.371.67.2 × 1073.7 × 103
      Lymphocyte-mediated immunity5.7143.19.5 × 1074.9 × 103
      Immune system development3.2245.41.5 × 1067.6 × 103
      Immune response–regulating cell surface receptor signaling pathway9.692.03.0 × 1061.5 × 102
      Positive regulation of endocytosis11.981.83.1 × 1061.6 × 102
      B-cell activation5.3132.95.3 × 1062.7 × 102
      Acute inflammatory response5.7122.77.1 × 1063.6 × 102
      Immune response–regulating signal transduction8.592.08.4 × 1064.2 × 102
      Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation analysis was performed using a list of differentially expressed genes with higher lung expression in SP-C/TNF-α mice (472 probes and 446 unique DAVID identification numbers). The Agilent mouse microarray G4121A probe list served as the background. Enriched terms with a Bonferroni-adjusted P < 0.05 are displayed.
      MHC, major histocompatibility complex.
      Table 5Functional Analysis for Biological Function of Genes with Lower Lung Expression in SP-C/TNF-α Mice
      TermFold enrichmentNo. of genes% of totalModified Fisher exact P valueBonferroni-adjusted P value
      Anatomical structure development25719.91.2 × 1076.1 × 104
      Organ development2.34515.71.4 × 1077.1 × 104
      Enzyme-linked receptor protein signaling pathway4.5196.62.3 × 1071.2 × 103
      System development2.15017.44.7 × 1072.4 × 103
      Cytoskeleton organization and biogenesis3.2227.74.6 × 1062.3 × 102
      Blood vessel morphogenesis4.9144.95.1 × 1062.6 × 102
      Blood vessel development4.6155.25.1 × 1062.6 × 102
      Anatomical structure morphogenesis2.23712.95.1 × 1062.6 × 102
      Vasculature development4.5155.25.9 × 1063.0 × 102
      Organ morphogenesis3238.06.4 × 1063.3 × 102
      Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation analysis was performed using a list of differentially expressed genes with lower lung expression in SP-C/TNF-α mice (288 probes and 287 unique DAVID identification numbers). The Agilent mouse microarray G4121A probe list served as the background. Enriched terms with a Bonferroni-adjusted P < 0.05 are displayed.
      Functional annotation clustering of the gene ontology terms was used to reduce redundancy across the functional categories.
      • Huang da W.
      • Sherman B.T.
      • Lempicki R.A.
      Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources.
      Clustering of functional groups using medium stringency resulted in 36 clusters, with a group geometric mean of P < 0.05, according to a modified Fisher's exact test (Table 6). Broadly speaking, the clusters contained genes involved in immune/inflammatory response (clusters 2, 12, 18, and 19), Igs (clusters 4, 5, 8, and 15), cytokines/chemokines (clusters 6, 14, and 36), membranes (clusters 7, 17, and 26), carbohydrate binding (clusters 9 and 13), antigen processing (cluster 10), cell differentiation/organ development (clusters 11, 20, and 27), proteases/peptidases and inhibitors (clusters 16 and 28), cell adhesion (cluster 21), phagocytosis (cluster 22), TNF (cluster 23), blood coagulation (cluster 24), oxidoreductase activity (cluster 25), angiogenesis (cluster 29), leukocyte migration (cluster 30), cell proliferation (cluster 31), and SH2 domain (cluster 32).
      Table 6Clustering of Functional Groups
      Functional clusterGroup P valueNo. of genes
      Glycoprotein/extracellular space3.2 × 1021319
      Immune/inflammatory response1.1 × 1016165
      Lysosome5.6 × 101031
      Ig-like1.6 × 10960
      Ig C1 set4.8 × 10928
      Chemokine activity1.4 × 10898
      Membrane7.7 × 107325
      Ig1.3 × 10522
      Carbohydrate binding4.6 × 10534
      Antigen processing and presentation7.0 × 10528
      Hemopoietic or lymphoid organ development1.4 × 10429
      Immune/inflammatory response2.7 × 10476
      Carbohydrate-binding/C-type lectin2.8 × 10433
      Small chemokine4.4 × 1048
      Ig C region1.2 × 1037
      Proteases/peptidases and inhibitors2.2 × 10348
      Plasma membrane part2.5 × 10390
      Immune and inflammatory signaling pathway5.5 × 10310
      Lymphocyte proliferation and activation5.5 × 10352
      Cell differentiation/organ development6.7 × 103178
      Cell adhesion7.7 × 10339
      Phagocytosis9.2 × 10341
      TNF9.6 × 1037
      Blood coagulation1.1 × 10211
      Oxidoreductase activity1.3 × 1029
      Cell membrane1.5 × 102499
      Myeloid leukocyte differentiation1.6 × 1028
      Protease inhibitor1.9 × 10231
      Angiogenesis2.3 × 10245
      Leukocyte migration2.7 × 1027
      Cell proliferation3.1 × 10233
      SH2 domain3.3 × 10211
      Lipid binding3.3 × 10224
      Integrin complex4.5 × 10224
      Cystine knot, C terminal4.5 × 1024
      Chemokine receptor activity4.7 × 10215
      Database for Annotation, Visualization and Integrated Discovery (DAVID) functional annotation clustering analysis was performed using a list of differentially expressed genes (760 probes and 732 unique DAVID identification numbers) with the Agilent mouse microarray G4121A probe list serving as the background. The group P value is the geometric mean of all modified Fisher exact P values in a given annotation cluster. Functional groups clustered using medium stringency with a modified Fisher exact P < 0.05 are displayed.

      KEGG Pathway Analysis

      Further functional analysis was performed by scanning all KEGG pathways for involvement of genes differentially expressed in the TNF model. Sorting by number of genes with fold change >1.5-fold, top KEGG pathway hits included cytokine-cytokine receptor interaction (60 genes; 26% of the pathway), cell adhesion molecules (49 genes; 31% of the pathway), regulation of actin cytoskeleton (36 genes; 17% of the pathway), and hematopoietic cell lineage (34 genes; 39% of the pathway), consistent with the functional analysis performed using gene ontology terms. Pathways with the highest ratio of genes with increased expression to genes with decreased expression included several immune and inflammatory pathways, such as B-cell receptor signaling (27 increased and 4 decreased), antigen processing and presentation (24 increased and 0 decreased), T-cell receptor signaling pathway (23 increased and 4 decreased), Toll-like receptor signaling pathway (20 increased and 2 decreased), asthma (16 increased and 0 decreased), cell cycle (10 increased and 2 decreased), and apoptosis (9 increased and 1 decreased). Pathways with many genes with decreased expression included tight junction (3 increased and 25 decreased), drug metabolism/cytochrome P450 (0 increased and 20 decreased), glutathione metabolism (1 increased and 12 decreased), retinol metabolism (1 increased and 10 decreased), and hedgehog signaling pathway (1 increased and 7 decreased). Searchable heat maps of all KEGG pathways are provided (see Supplemental Appendix S1 at http://ajp.amjpathol.org).

      Expression of Immune and Inflammatory Mediators

      To further investigate the enrichment of immune and inflammatory genes, we examined differential expression of probes associated with the functional term inflammatory response. Comparison of expression by heat map display revealed that all but one gene exhibited higher levels of expression in transgenic mice (Figure 4A). In addition to TNF-α, the list included TNF receptor superfamily member 4, pyrin, the acute-phase response proteins serum amyloid A1 and A3, 12 chemokines, toll-like receptor-2, and several complement factors and Ig genes. Adipsin (complement factor D), a serine protease first characterized as being down-regulated in adipocytes by TNF-α,
      • Min H.Y.
      • Spiegelman B.M.
      Adipsin, the adipocyte serine protease: gene structure and control of expression by tumor necrosis factor.
      was the only gene on the list with decreased expression. To independently validate observed differences in gene expression, we measured transcript levels in a new set of age-matched animals (n = 4 per genotype) not used for the microarray experiment. Real-time PCR confirmed a dramatic increase in TNF-α mRNA in SP-C/TNF-α animals, as expected, and increased expression of serum amyloid A3, the chemokine (CXC) motif ligand 13, IL-4, and lymphotoxin A (Figure 4B).
      Figure thumbnail gr4
      Figure 4Inflammatory mediators. A: Heat map of differentially expressed genes with the functional annotation inflammatory response. Red indicates a high level of expression relative to the reference RNA, whereas green indicates low relative expression. B: Real-time PCR was conducted on independent biological samples from age-matched animals (n = 4 per genotype) to verify microarray results. LTa, lymphotoxin A; SAA3, serum amyloid A3; WT, wild type. *P < 0.05, t-test.

      Expression of Proteases and Inhibitors

      Given the pronounced airspace enlargement characteristic of this model, we examined alteration of probes associated with the functional term protease (Figure 5A). The microarray data revealed increased expression of several metalloproteinases, including MMP-9 (type IV collagenase, gelatinase), MMP-12 (macrophage elastase), and MMP-14 (membrane type 1–MMP), a protease critical for MMP maturation. Expression of cathepsins B, D, K, S, and Z was also increased, as were genes involved in pro-inflammatory caspase activity (caspases 1, 4, and 11 and CARD12). Real-time PCR conducted on independent biological samples confirmed significantly increased expression of MMPs (MMPs 2, 9, 12, and 14; P < 0.05; Figure 5B). In contrast, expression of the MMP inhibitors (reversion-inducing cysteine-rich protein with kazal motifs, TIMP-2, and TIMP-3) and the predicted MMP inhibitor procollagen C-terminal enhancer protein 2a were all significantly decreased (P < 0.05, Figure 5C). Compared with the marked decrease in TIMP-3 expression in the transgenic animals, there was less of an impact on TIMP-2 mRNA levels (P < 0.05) and no significant change in TIMP-1 expression (P > 0.05). Differential expression of cathepsin K, caspases 1 and 4, and CARD12 was also verified by real-time PCR (P < 0.05, Figure 5D).
      Figure thumbnail gr5
      Figure 5Proteases and anti-proteases. A: Heat map depicting microarray data of differentially expressed genes with the functional annotation protease. Red indicates a high level of expression relative to the reference RNA, whereas green indicates low relative expression. BD: Real-time PCR was used to examine relative expression levels in independent biological samples from age-matched animals (n = 4 per genotype). B: MMP expression. MT, membrane type. C: Inhibitors of MMPs. PCOLCE2, procollagen C–endopeptidase enhancer 2; RECK, reversion-inducing cysteine-rich protein with kazal motifs. NS, not significant; WT, wild type. D: Selected additional proteases. CARD, caspase-associated recruitment domain; CASP, caspase; CTSK, cathepsin K. *P < 0.05, t-test.

      Expression of Genes Involved in Lung Repair and Maintenance

      Given the importance of repair and maintenance processes in the control of lung architecture integrity, and in the fate of lungs after injury, we examined expression of genes associated with the terms angiogenesis, cytoskeleton organization, and biogenesis (Figure 6). Most genes associated with these terms exhibited decreased expression in SP-C/TNF-α mice. Factors implicated in endothelial cell survival and proliferation, including vascular endothelial growth factor (VEGF) and one of its receptors, FMS-like tyrosine kinase-1 (VEGF receptor-1), connective tissue growth factor, c-fos–induced growth factor (VEGFD), heparin-binding EGF-like growth factor, endoglin, and the transcription factor inhibitor of DNA binding 1 all exhibited decreased expression compared with wild-type littermates (Figure 6A). The expression of smooth muscle transcripts associated with myofibroblasts was also predominantly decreased, including calponin, myosin, troponin, tubulin, and gelsolin, a potent actin filament regulatory protein involved in the maintenance of lung vascular barrier function
      • Becker P.M.
      • Kazi A.A.
      • Wadgaonkar R.
      • Pearse D.B.
      • Kwiatkowski D.
      • Garcia J.G.
      Pulmonary vascular permeability and ischemic injury in gelsolin-deficient mice.
      (Figure 6B). TGF-β, a factor implicated in the pathogenesis of fibrosis and in conversion of fibroblasts to myofibroblasts, exhibited a modest increase. However, expression of a downstream mediator involved in myofibroblast differentiation and stimulation of fibroblast matrix production,
      • Scotton C.J.
      • Chambers R.C.
      Molecular targets in pulmonary fibrosis: the myofibroblast in focus.
      connective tissue growth factor, was significantly decreased. Real-time PCR performed on independent biological samples confirmed differential expression of TGF-β, connective tissue growth factor, and VEGF and decreased expression of lysyl oxidase, a factor critical for elastin cross-linking
      • Gao S.
      • Chen K.
      • Zhao Y.
      • Rich C.B.
      • Chen L.
      • Li S.J.
      • Toselli P.
      • Stone P.
      • Li W.
      Transcriptional and posttranscriptional inhibition of lysyl oxidase expression by cigarette smoke condensate in cultured rat fetal lung fibroblasts.
      (Figure 6C).
      Figure thumbnail gr6
      Figure 6Repair and maintenance. Heat maps were generated depicting microarray data of differentially expressed genes with the functional annotation angiogenesis (A) or cytoskeleton organization and biogenesis (B). Red indicates a high level of expression relative to the reference RNA, whereas green indicates low relative expression. C: Expression of select genes was confirmed by real-time PCR. CTGF, connective tissue growth factor; LOX, lysyl oxidase; WT, wild type. *P < 0.05, t-test.

      Comparison of SP-C/TNF-α Lung Transcriptome and Other Models of Lung Disease

      To gain further insight into the relative importance of differentially expressed genes in defining the phenotype of transgenic animals, data from the present study were compared with 12 other mouse models of lung disease.
      • Lewis C.C.
      • Yang J.Y.
      • Huang X.
      • Banerjee S.K.
      • Blackburn M.R.
      • Baluk P.
      • McDonald D.M.
      • Blackwell T.S.
      • Nagabhushanam V.
      • Peters W.
      • Voehringer D.
      • Erle D.J.
      Disease-specific gene expression profiling in multiple models of lung disease.
      These included a model of IL-13 overexpression, two models of bleomycin-induced injury (9 and 21 days after treatment), two models of lipopolysaccharide exposure (inhalation and i.p. injection), two allergy models involving challenge with ovalbumin, and five models of bacterial infection. Of the 7148 genes common to all platforms used in these studies, 736 were differentially expressed (FDR-adjusted P < 0.05 and >1.5-fold change in expression) in the SP-C/TNF-α model. Unsupervised clustering of the expression profiles generated using the top 50 differentially expressed genes, according to fold change from each model, showed that the SP-C/TNF-α model clustered with the two bleomycin models (Figure 7). Similar clusters were seen using the top 25 or 100 differentially expressed genes (data not shown), indicating that the relationships among models were fairly robust. Given that bleomycin treatment leads to fibrosis, comparison of differentially expressed genes common to the SP-C/TNF-α model or unique to the bleomycin models might serve to identify processes critical for disease pathogenesis. Functional assessment of differentially expressed genes with ≥1.5-fold change common to the SP-C/TNF-α model and either the 9- or the 21-day bleomycin model revealed enrichment of genes involved in inflammation/immune response. Similar pathways were found for all comparisons of models with SP-C/TNF-α mice (Table 7), suggesting that these processes constitute common features of lung disease. Indeed, of the 23 transcripts found to be differentially expressed in common across most (at least 9 of 12) models
      • Lewis C.C.
      • Yang J.Y.
      • Huang X.
      • Banerjee S.K.
      • Blackburn M.R.
      • Baluk P.
      • McDonald D.M.
      • Blackwell T.S.
      • Nagabhushanam V.
      • Peters W.
      • Voehringer D.
      • Erle D.J.
      Disease-specific gene expression profiling in multiple models of lung disease.
      and present on the array used in the present study, 19 were similarly altered in SP-C/TNF-α mice (see Supplemental Table S2 at http://ajp.amjpathol.org).
      • Lewis C.C.
      • Yang J.Y.
      • Huang X.
      • Banerjee S.K.
      • Blackburn M.R.
      • Baluk P.
      • McDonald D.M.
      • Blackwell T.S.
      • Nagabhushanam V.
      • Peters W.
      • Voehringer D.
      • Erle D.J.
      Disease-specific gene expression profiling in multiple models of lung disease.
      Genes differentially expressed in the bleomycin models, but not the SP-C/TNF-α model, were enriched for extracellular matrix genes, including increased expression of collagens, elastin, and fibronectin transcripts. Connective tissue growth factor mRNA was elevated in the 21-day bleomycin model of fibrosis (1.7-fold) and not in any other model. Functional assessment of genes differentially expressed in other models did not reveal similar enrichment for extracellular matrix genes (Table 7), indicating that this difference was specific to the bleomycin models.
      Figure thumbnail gr7
      Figure 7Comparison of multiple lung disease models. Lung disease models were compared by hierarchical clustering according to the most differentially expressed transcripts for each model. Transcripts with unadjusted P < 0.05 were ranked by fold change, and the top 50 transcripts for each disease model were selected, resulting in 331 unique transcripts used for clustering. LPS, lipopolysaccharide; M. pulmonis, Mycoplasma pulmonis; M. tuberculosis, Mycobacterium tuberculosis; N. brasiliensis, Nippostrongylus brasiliensis; OVA, ovalbumin; P. aeruginosa, Pseudomonas aeruginosa.
      Table 7Functional Comparison of Gene Profiles across Mouse Models of Lung Disease
      ModelFunctional analysis of common genes
      The number in common is given in parentheses.
      Functional analysis of unique genes
      The number unique or discordant is given in parentheses.
      Bleomycin
       Acute (7 days)Glycoprotein, immune/inflammatory response (n = 53)Extracellular matrix, collagen (n = 126)
       Fibrosis (21 days)Glycoprotein, lysosome, chemokine, chemotaxis (n = 119)Glycoprotein, cell adhesion, extracellular matrix, collagen (n = 232)
      Aspergillus extractGlycoprotein, inflammatory response, chemotaxis (n = 29)Thyroglobulin type-1, heat shock protein 70, lipid transport (n = 146)
      Nippostrongylus brasiliensisGlycoprotein, immune/inflammatory response, lysosome (n = 193)Inflammatory response, glycoprotein, cell adhesion (n = 467)
      IL-13 overexpressionGlycoprotein, lysosome, immune/inflammatory response (n = 207)Glycoprotein, inflammatory response, chemotaxis (n = 441)
      Ovalbumin
       BALBcGlycoprotein, immune/inflammatory response, chemokine (n = 319)DNA replication, purine ribonucleotide-binding, chromatin modification (n = 736)
       C57BL6Glycoprotein, inflammatory response, chemokine (n = 182)Inflammatory response, glycoprotein, acute phase response (n = 307)
      LPS
       AerosolizedGlycoprotein, immune/inflammatory response, chemokine (n = 158)Basic leucine zipper, protein dimerization activity, toll-like receptor signaling (n = 748)
       i.p.Immune/inflammatory response, signal peptide, small chemokine (n = 229)Basic leucine zipper, wound healing, lysosome (n = 1208)
      Mycoplasma pulmonisGlycoprotein, immune response, chemokine (n =358)Wound healing, glycoprotein, glycolysis (n = 957)
      Mycobacterium tuberculosisImmune/inflammatory response, glycoprotein, lysosome (n = 304)Cytoskeleton organization, Golgi apparatus, hypertrophic cardiomyopathy (n = 845)
      Pseudomonas aeruginosaGlycoprotein, immune/inflammatory response, regulation of cytokine production (n = 159)Protein dimerization, basic leucine zipper, GTP binding (n = 738)
      Common genes were defined as being differentially expressed by at least 1.5-fold in both SP-C/TNF-α (FDR-adjusted P < 0.05) and each disease model (unadjusted P < 0.05). Unique or discordant (expressed in both models, but increased in one and decreased in the other) genes composed the second group. Functional terms listed represent the most enriched clusters following functional annotation clustering using the Database for Annotation, Visualization and Integrated Discovery. The 7148 GenBank identification numbers common to all platforms under comparison were used as the reference population for enrichment analysis.
      LPS, lipopolysaccharide.
      low asterisk The number in common is given in parentheses.
      The number unique or discordant is given in parentheses.
      To identify genes unique to the SP-C/TNF-α model, genes were selected that were differentially expressed in SP-C/TNF-α mice (FDR-adjusted P < 0.05 and twofold or greater change in expression) and not differentially expressed (unadjusted P < 0.05 or >1.5-fold change in expression) in any other model. This relatively stringent approach identified 29 genes (Table 8). Functional assessment did not detect any significant enrichment of gene terms, and broadening the list to include unique genes with ≥1.5-fold change in gene expression (n = 80) did not significantly alter results (data not shown). A PubMed search for each gene and the term emphysema identified three of the top eight genes, ranked by fold change, as having previously been linked to the disease. These genes were triggering receptor expressed on myeloid cells 2 (13.1-fold), transcription factor Spi-C (5.3-fold), and caspase-1 (3.1-fold). Also included in the top 10 genes were the TNF receptor superfamily member 17 (3.5-fold) and the aryl hydrocarbon receptor repressor (3.0-fold), consistent with TNF-α signaling and repression of the aryl hydrocarbon receptor pathway, respectively.
      Table 8Gene Expression Changes Unique to the SP-C/TNF-α Model
      GenBank no.Gene symbolGene descriptionFold change
      NM_031254Trem2Triggering receptor expressed on myeloid cells 213.1
      AK0085512010309G21RikRIKEN cDNA 2010309G21 gene9.5
      NM_011461SpicSpi-C transcription factor (Spi-1/PU.1 related)5.3
      NM_007793CstbCystatin B3.7
      NM_011608Tnfrsf17TNF receptor superfamily, member 173.5
      NM_021319Pglyrp2Peptidoglycan recognition protein 23.5
      NM_009267Spt1Salivary protein 13.2
      NM_009807Casp1Caspase 13.1
      NM_009644AhrrAryl-hydrocarbon receptor repressor3.0
      NM_013546Hebp1Heme-binding protein 12.9
      NM_024177Mrpl38Mitochondrial ribosomal protein L382.4
      NM_008509LplLipoprotein lipase2.3
      NM_023383AadacAryl acetamide deacetylase (esterase)2.3
      NM_013533Gpr162G-protein–coupled receptor 1622.2
      NM_009518Wnt10aWingless-related MMTV integration site 10a2.1
      NM_008964Ptger2Prostaglandin E receptor 2 (subtype EP2)2.1
      AK014074Katnal2Katanin p60 subunit A-like 2−2.0
      NM_015775Tmprss2Transmembrane protease, serine 2−2.0
      NM_013569Kcnh2Potassium voltage-gated channel, subfamily H (eag related), member 2−2.1
      NM_054074Defb6Defensin β6−2.1
      NM_013593MbMyoglobin−2.1
      NM_021447Trim54Tripartite motif-containing 54−2.2
      NM_011902Tekt2Tektin 2−2.2
      NM_0261364930449I24RikRIKEN cDNA 4930449I24 gene−2.2
      NM_007775CrygcCrystallin, γ C−2.3
      AK016963Dydc2DPY30 domain-containing 2−2.3
      NM_008664Myom2Myomesin 2−2.4
      NM_011449Spa17Sperm autoantigenic protein 17−2.5
      NM_011569Tekt1Tektin 1−2.6
      Genes were screened for inclusion according to the following criteria: significant differential expression (twofold change or greater and FDR-adjusted P < 0.05) in the SP-C/TNF-α model and not differentially expressed (1.5-fold change or greater and unadjusted P < 0.05) in any of the 12 lung disease models.

      Discussion

      Depending on the context, TNF-α expression can elicit a variety of biological responses, including provoking the influx of inflammatory cells, exerting an effect on resident cells, and increasing production of proteolytic factors; cell death, proliferation, and migration may also occur. Although it remains unclear to what extent TNF-α is implicated in the development of emphysema in humans, TNF-null mice exhibit significantly reduced inflammation and airspace enlargement in response to 6 months of exposure to cigarette smoke,
      • Churg A.
      • Wang R.D.
      • Tai H.
      • Wang X.
      • Xie C.
      • Wright J.L.
      Tumor necrosis factor-alpha drives 70% of cigarette smoke-induced emphysema in the mouse.
      suggesting an important role in disease pathogenesis. The dominant feature of the lung transcriptome of TNF-overexpressing mice was increased expression of genes involved in immune and inflammatory processes, consistent with the pro-inflammatory role of TNF-α and with the pronounced inflammation evident in the lungs. In a conditional TNF-overexpression model, only mild inflammatory changes were observed and TNF-α mRNA was itself not detected
      • Vuillemenot B.R.
      • Rodriguez J.F.
      • Hoyle G.W.
      Lymphoid tissue and emphysema in the lungs of transgenic mice inducibly expressing tumor necrosis factor-alpha.
      ; in the present model, TNF-α mRNA levels were several hundredfold higher according to real-time PCR analysis as a result of the constitutive expression of the TNF-α transgene. The robust inflammatory response observed in the lungs and bronchoalveolar lavage fluid were reflected in the enrichment of genes that play a role in inflammation. Chemokines, critical for the influx of lymphocytes in the lungs, were highly expressed, as were transcripts known to be expressed by lymphocytes. Also increased were genes involved in the processing of cytokines to their mature forms, including inflammatory caspases with known pro-inflammatory roles.
      • Wang J.
      • Lenardo M.J.
      Roles of caspases in apoptosis, development, and cytokine maturation revealed by homozygous gene deficiencies.
      Lungs of transgenic mice also exhibited decreased expression of genes involved in xenobiotic metabolism, consistent with known effects of inflammation on the aryl hydrocarbon receptor signaling pathway
      • Ke S.
      • Rabson A.B.
      • Germino J.F.
      • Gallo M.A.
      • Tian Y.
      Mechanism of suppression of cytochrome P-450 1A1 expression by tumor necrosis factor-alpha and lipopolysaccharide.
      ; tight junction and adhesion factors, consistent with alteration of lung barrier function by TNF-α
      • Mazzon E.
      • Cuzzocrea S.
      Role of TNF-alpha in lung tight junction alteration in mouse model of acute lung inflammation.
      ; and glutathione metabolism, consistent with impacts of inflammation on antioxidant status.
      • Rahman I.
      Regulation of glutathione in inflammation and chronic lung diseases.
      The degradation of alveolar wall structure in emphysema is thought to be due, in large part, to protease/anti-protease imbalance, exemplified by increased levels and activity of proteases, such as MMPs; increased expression of factors that enhance protease expression and activity; and decreased levels of inhibitors. The functional significance of protease/anti-protease imbalance is demonstrated in several animal models. MMP-12−/− mice are protected from emphysema, resulting from long-term exposure to cigarette smoke
      • Hautamaki R.D.
      • Kobayashi D.K.
      • Senior R.M.
      • Shapiro S.D.
      Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice.
      ; and airspace enlargement in mice, resulting from overexpression of IL-13 in the airways, is abrogated somewhat by inhibition or knockout of MMP-9 or MMP-12.
      • Lanone S.
      • Zheng T.
      • Zhu Z.
      • Liu W.
      • Lee C.G.
      • Ma B.
      • Chen Q.
      • Homer R.J.
      • Wang J.
      • Rabach L.A.
      • Rabach M.E.
      • Shipley J.M.
      • Shapiro S.D.
      • Senior R.M.
      • Elias J.A.
      Overlapping and enzyme-specific contributions of matrix metalloproteinases-9 and -12 in IL-13-induced inflammation and remodeling.
      TIMP-3−/− mice, but not TIMP-1−/− or TIMP-2−/− mice, exhibit airspace enlargement characteristic of pulmonary emphysema.
      • Leco K.J.
      • Waterhouse P.
      • Sanchez O.H.
      • Gowing K.L.
      • Poole A.R.
      • Wakeham A.
      • Mak T.W.
      • Khokha R.
      Spontaneous air space enlargement in the lungs of mice lacking tissue inhibitor of metalloproteinases-3 (TIMP-3).
      Functions of MMPs and their inhibitors extend beyond matrix remodeling, with several MMPs capable of processing pro-TNF to the active form,
      • Gearing A.J.
      • Beckett P.
      • Christodoulou M.
      • Churchill M.
      • Clements J.
      • Davidson A.H.
      • Drummond A.H.
      • Galloway W.A.
      • Gilbert R.
      • Gordon J.L.
      • Leber T.M.
      • Mangan M.
      • Miller K.
      • Nayee P.
      • Owen K.
      • Patel S.
      • Thomas W.
      • Wells G.
      • Wood L.M.
      • Woolley K.
      Processing of tumour necrosis factor-alpha precursor by metalloproteinases.
      whereas TIMP-3 can inhibit the TNF-converting enzyme
      • Amour A.
      • Slocombe P.M.
      • Webster A.
      • Butler M.
      • Knight C.G.
      • Smith B.J.
      • Stephens P.E.
      • Shelley C.
      • Hutton M.
      • Knauper V.
      • Docherty A.J.
      • Murphy G.
      TNF-alpha converting enzyme (TACE) is inhibited by TIMP-3.
      and, thus, serve as an important regulator of inflammation. Our data indicate that TNF-α overexpression leads to profound disruption at several levels of protease regulation. Serum amyloid A3, which was among the most increased transcripts, is a TNF-responsive
      • Son D.S.
      • Roby K.F.
      • Terranova P.F.
      Tumor necrosis factor-alpha induces serum amyloid A3 in mouse granulosa cells.
      acute-phase response factor that can regulate MMP activity.
      • Vallon R.
      • Freuler F.
      • Desta-Tsedu N.
      • Robeva A.
      • Dawson J.
      • Wenner P.
      • Engelhardt P.
      • Boes L.
      • Schnyder J.
      • Tschopp C.
      • Urfer R.
      • Baumann G.
      Serum amyloid A (apoSAA) expression is up-regulated in rheumatoid arthritis and induces transcription of matrix metalloproteinases.
      Transgenic animals displayed increased transcription of MMPs 2, 9, and 12, each of which cleaves basement membrane proteins, including elastin, and increased expression of membrane type 1–MMP, a transmembrane MMP critical for the activation of pro-MMP forms. Increased cathepsin expression is consistent with their role in matrix remodeling in the lungs.
      • Wolters P.J.
      • Chapman H.A.
      Importance of lysosomal cysteine proteases in lung disease.
      In contrast, metalloproteinase inhibitors (reversion-inducing cysteine-rich protein with kazal motifs, TIMP-2, and TIMP-3) exhibited decreased expression relative to levels in the lungs of wild-type animals. Procollagen C-terminal enhancer protein 2a, a protein antagonistic to metalloproteases that provokes a net increase in deposition of fibrous matrix,
      • Steiglitz B.M.
      • Keene D.R.
      • Greenspan D.S.
      PCOLCE2 encodes a functional procollagen C-proteinase enhancer (PCPE2) that is a collagen-binding protein differing in distribution of expression and post-translational modification from the previously described PCPE1.
      • Baker A.H.
      • Edwards D.R.
      • Murphy G.
      Metalloproteinase inhibitors: biological actions and therapeutic opportunities.
      • Overall C.M.
      • Dean R.A.
      Degradomics: systems biology of the protease web: pleiotropic roles of MMPs in cancer.
      was the transcript most decreased in transgenic animals. Thus, genes with the greatest differential expression in this model are regulators of MMP activity. Overall, there was a clear shift at the transcriptional level in the protease/anti-protease balance toward increased proteolytic activity.
      Although the protease/anti-protease hypothesis has been the dominant view of emphysema pathogenesis, impacts of vascular abnormalities have also been considered.
      • Voelkel N.F.
      • Douglas I.S.
      • Nicolls M.
      Angiogenesis in chronic lung disease.
      VEGF is required for endothelial cell survival, and decreased signaling through its receptors is associated with apoptosis and airspace enlargement,
      • Kasahara Y.
      • Tuder R.M.
      • Taraseviciene-Stewart L.
      • Le Cras T.D.
      • Abman S.
      • Hirth P.K.
      • Waltenberger J.
      • Voelkel N.F.
      Inhibition of VEGF receptors causes lung cell apoptosis and emphysema.
      suggesting that capillary endothelial cells are required for maintenance of alveolar structure. Reduced VEGF expression was previously observed in the SP-C/TNF-α model and was associated with pulmonary hypertension at altitude in Denver, CO,
      • Fujita M.
      • Mason R.J.
      • Cool C.
      • Shannon J.M.
      • Hara N.
      • Fagan K.A.
      Pulmonary hypertension in TNF-alpha-overexpressing mice is associated with decreased VEGF gene expression.
      although the effects on both VEGF expression and hypertension were attenuated at sea level in Japan without any reduction in the emphysema.
      • Fujita M.
      • Ikegame S.
      • Ye Q.
      • Harada E.
      • Ouchi H.
      • Inoshima I.
      • Watanabe K.
      • Mason R.J.
      • Nakanishi Y.
      Attenuation of pulmonary hypertension, but not emphysematous change, by breeding emphysema model mice at sea level.
      Herein, we extend this observation by showing that transcript levels of several genes involved in angiogenesis are lower in the transgenic mice, suggesting that dysfunction of blood vessel development is a feature of the SP-C/TNF-α model, as is seen in emphysema in human lungs.
      • Voelkel N.F.
      • Douglas I.S.
      • Nicolls M.
      Angiogenesis in chronic lung disease.
      Decreased expression of the transcription factor inhibitor of DNA binding 1, an endothelial progenitor cell marker that regulates angiogenesis, may be an important driver of these effects.
      • Gao D.
      • Nolan D.J.
      • Mellick A.S.
      • Bambino K.
      • McDonnell K.
      • Mittal V.
      Endothelial progenitor cells control the angiogenic switch in mouse lung metastasis.
      Pulmonary hypertension is often detected in patients with emphysema and is thought to be mainly due to hypoxia, inflammation, and loss of capillaries.
      • Chaouat A.
      • Naeije R.
      • Weitzenblum E.
      Pulmonary hypertension in COPD.
      Although pulmonary hypertension was not assessed in the present work, the observation of decreased expression of VEGF and genes involved in angiogenesis in our laboratory in Ottawa, ON, Canada (approximately 70 m above sea level), suggests that effects on this system may not be strictly dependent on altitude.
      Although the SP-C/TNF-α mouse was initially proposed as a model of pulmonary fibrosis,
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      4-month-old mice were resistant to the development of fibrosis after treatment with the profibrotic agent bleomycin or TGF-β.
      • Fujita M.
      • Shannon J.M.
      • Morikawa O.
      • Gauldie J.
      • Hara N.
      • Mason R.J.
      Overexpression of TNF-alpha diminishes pulmonary fibrosis induced by bleomycin or TGF-beta.
      The presence of functional repair mechanisms is a critical determinant of lung fate after injury, with connective tissue synthesis during the healing process a prerequisite for fibrosis.
      • Niewoehner D.E.
      • Hoidal J.R.
      Lung fibrosis and emphysema: divergent responses to a common injury.
      Myofibroblasts are the primary cell type responsible for connective tissue synthesis during repair processes, and although they play a central role in the pathogenesis of fibrosis,
      • Scotton C.J.
      • Chambers R.C.
      Molecular targets in pulmonary fibrosis: the myofibroblast in focus.
      there is evidence of disruption of myofibroblast structure and function in emphysema.
      • Sirianni F.E.
      • Milaninezhad A.
      • Chu F.S.
      • Walker D.C.
      Alteration of fibroblast architecture and loss of Basal lamina apertures in human emphysematous lung.
      Despite a modest increase of TGF-β mRNA in SP-C/TNF-α animals, we observed decreased expression of connective tissue growth factor, an important downstream mediator of TGF-β–regulated myofibroblast differentiation and extracellular matrix production,
      • Scotton C.J.
      • Chambers R.C.
      Molecular targets in pulmonary fibrosis: the myofibroblast in focus.
      and lower expression of genes involved in cytoskeleton organization, biogenesis, and myofibroblast differentiation. TNF-α can down-regulate connective tissue growth factor
      • Yu F.
      • Chou C.W.
      • Chen C.C.
      TNF-alpha suppressed TGF-beta-induced CTGF expression by switching the binding preference of p300 from Smad4 to p65.
      and antagonize TGF-β–driven myofibroblast differentiation,
      • Liu X.
      • Kelm Jr, R.J.
      • Strauch A.R.
      Transforming growth factor beta1-mediated activation of the smooth muscle alpha-actin gene in human pulmonary myofibroblasts is inhibited by tumor necrosis factor-alpha via mitogen-activated protein kinase kinase 1-dependent induction of the Egr-1 transcriptional repressor.
      which may provide a critical break in the TGF-β signaling required for fibrosis. The transcriptional changes suggest a pulmonary environment resistant to myofibroblast differentiation and are in sharp contrast with the up-regulation of smooth muscle–related transcripts observed after treatment of fibroblasts with TGF-β.
      • Wicks J.
      • Haitchi H.M.
      • Holgate S.T.
      • Davies D.E.
      • Powell R.M.
      Enhanced upregulation of smooth muscle related transcripts by TGF beta2 in asthmatic (myo) fibroblasts.
      Although fibroblast proliferation was reported in the initial characterization of the SP-C/TNF-α mouse, the cells were negative by immunocytochemistry to anti-smooth muscle actin-α antibody,
      • Miyazaki Y.
      • Araki K.
      • Vesin C.
      • Garcia I.
      • Kapanci Y.
      • Whitsett J.A.
      • Piguet P.F.
      • Vassalli P.
      Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
      a myofibroblast marker. The physiological significance of decreased myofibroblast-associated gene expression in determining lung fate is substantiated in other animal models. For example, expression of gelsolin is required for the development of pulmonary fibrosis in bleomycin-treated mice,
      • Oikonomou N.
      • Thanasopoulou A.
      • Tzouvelekis A.
      • Harokopos V.
      • Paparountas T.
      • Nikitopoulou I.
      • Witke W.
      • Karameris A.
      • Kotanidou A.
      • Bouros D.
      • Aidinis V.
      Gelsolin expression is necessary for the development of modelled pulmonary inflammation and fibrosis.
      whereas inhibition of lysyl oxidase in guinea pigs exposed to CdCl2 resulted in emphysema, rather than the fibrotic phenotype observed in the presence of lysyl oxidase.
      • Niewoehner D.E.
      • Hoidal J.R.
      Lung fibrosis and emphysema: divergent responses to a common injury.
      Together, the data describe a pulmonary environment characterized by transcriptional changes associated with emphysema, coupled with transcriptional changes that suggest resistance to fibrosis, consistent with the phenotype described herein and elsewhere.
      • Fujita M.
      • Shannon J.M.
      • Morikawa O.
      • Gauldie J.
      • Hara N.
      • Mason R.J.
      Overexpression of TNF-alpha diminishes pulmonary fibrosis induced by bleomycin or TGF-beta.
      Analysis of whole lung homogenates provides a global view of transcriptional changes in the lungs but may be insensitive to contributions from specific cell types or lung regions. Despite the dramatic transcriptional changes presented herein, another level of complexity is related to the extent to which TNF-α and the resulting injury alter alveolar structure and cell-cell interactions. Maintenance of alveolar architecture depends on the integrity of the scaffold and cellular relationships. After injury, successful lung repair relies on complex interactions of type II epithelial cells, fibroblasts, endothelial cells, and connective tissue governed by growth factors and cytokines, as well as by direct cell-cell and cell-matrix contacts. The high level of TNF-α produced by type II cells may interfere with the local cellularity and architectural integrity that guide epithelial and fibroblast proliferation and differentiation
      • Sirianni F.E.
      • Milaninezhad A.
      • Chu F.S.
      • Walker D.C.
      Alteration of fibroblast architecture and loss of Basal lamina apertures in human emphysematous lung.
      • Adamson I.Y.
      • Hedgecock C.
      • Bowden D.H.
      Epithelial cell-fibroblast interactions in lung injury and repair.
      and, thus, contribute to remodeling and loss of alveolar structure. As in any analysis of whole tissues, transcript levels will be influenced by changes in cell composition in addition to strict transcriptional activation. For example, we observed increased levels of Ig genes, consistent with the infiltration of B cells into the lungs. Nevertheless, the abundant changes in gene expression measured herein, whether from transcriptional activation or through import by inflammatory cells, reflect ongoing processes in the lungs that contribute to the observed pathological features and, therefore, are useful in understanding the underlying causes of disease.
      Comparison across multiple models of lung disease using global screening approaches can be used to identify gene profiles that distinguish disease phenotypes. We analyzed similarities and differences between the SP-C/TNF-α mice and 12 models of lung disease previously analyzed by gene expression profiling.
      • Lewis C.C.
      • Yang J.Y.
      • Huang X.
      • Banerjee S.K.
      • Blackburn M.R.
      • Baluk P.
      • McDonald D.M.
      • Blackwell T.S.
      • Nagabhushanam V.
      • Peters W.
      • Voehringer D.
      • Erle D.J.
      Disease-specific gene expression profiling in multiple models of lung disease.
      This rich data set of transcriptomes provides an ideal resource because of the breadth of pulmonary disease processes covered, encompassing models of acute lung injury, allergic inflammation, and bacterial infection, and also because many of the models share the same genetic background as the SP-C/TNF-α model (C57BL/6). The small set of genes uniquely altered in the SP-C/TNF-α model is likely enriched for genes involved in disease pathogenesis, providing candidates for further study. Indeed, three of the top eight genes identified have previously been associated with emphysema, including caspase-1, which was recently shown to be an important factor in cigarette smoke–induced emphysema through its role as an IL-1–converting enzyme
      • Churg A.
      • Zhou S.
      • Wang X.
      • Wang R.
      • Wright J.L.
      The role of interleukin-1beta in murine cigarette smoke-induced emphysema and small airway remodeling.
      ; Spi-C, a transcription factor implicated in transcriptional activation of MMP-12 previously identified in an emphysema model
      • Lian X.
      • Yan C.
      • Qin Y.
      • Knox L.
      • Li T.
      • Du H.
      Neutral lipids and peroxisome proliferator-activated receptor-{gamma} control pulmonary gene expression and inflammation-triggered pathogenesis in lysosomal acid lipase knockout mice.
      ; and triggering receptor expressed on myeloid cells 2, a factor critical for macrophage migration found to be increased in former smokers with chronic obstructive pulmonary disease.
      • Koth L.L.
      • Cambier C.J.
      • Ellwanger A.
      • Solon M.
      • Hou L.
      • Lanier L.L.
      • Abram C.L.
      • Hamerman J.A.
      • Woodruff P.G.
      DAP12 is required for macrophage recruitment to the lung in response to cigarette smoke and chemotaxis toward CCL2.
      Remarkably, the SP-C/TNF-α model clustered most closely with two disease models representing responses to the profibrotic agent, bleomycin, used to model acute lung injury (7 days after exposure) or fibrosis (21 days after exposure). Functional analysis of genes common to SP-C/TNF-α and bleomycin models indicated that the similarity was due primarily to the inflammatory/immune response common to the models, processes observed across all lung models that likely play a role in common features of disease pathogenesis. However, the gene expression profiles of the bleomycin models showed enrichment of genes involved in production and deposition of extracellular matrix, a key feature of fibrosis not present in the SP-C/TNF-α model. Connective tissue synthesis appears to be important in determining lung fate after injury,
      • Niewoehner D.E.
      • Hoidal J.R.
      Lung fibrosis and emphysema: divergent responses to a common injury.
      and induction of connective tissue growth factor is required for the development of lung fibrosis.
      • Bonniaud P.
      • Martin G.
      • Margetts P.J.
      • Ask K.
      • Robertson J.
      • Gauldie J.
      • Kolb M.
      Connective tissue growth factor is crucial to inducing a profibrotic environment in “fibrosis-resistant” BALB/c mouse lungs.
      Because SP-C/TNF-α mice exhibit decreased expression of genes implicated in myofibroblast differentiation and extracellular matrix production, including connective tissue growth factor, it seems plausible that the resistance to induction of fibrosis is due at least in part to an inability of bleomycin or TGF-β to elicit significant synthesis and deposition of extracellular matrix in these animals. The data suggest that this impaired repair process is an important feature distinguishing the SP-C/TNF-α model from the bleomycin models of acute lung injury and fibrosis.
      The dysfunctional immune processes, imbalance of proteases and their inhibitors, and disruption of maintenance and repair systems revealed by the gene expression profile of mice overexpressing TNF-α may all contribute to the pathogenesis of emphysema.
      • Henson P.M.
      • Vandivier R.W.
      • Douglas I.S.
      Cell death, remodeling, and repair in chronic obstructive pulmonary disease.
      • Wright J.L.
      • Churg A.
      Current concepts in mechanisms of emphysema.
      Although direct testing of the relative importance of each observed change was beyond the scope of this work, the literature
      • Kasahara Y.
      • Tuder R.M.
      • Taraseviciene-Stewart L.
      • Le Cras T.D.
      • Abman S.
      • Hirth P.K.
      • Waltenberger J.
      • Voelkel N.F.
      Inhibition of VEGF receptors causes lung cell apoptosis and emphysema.
      • Hautamaki R.D.
      • Kobayashi D.K.
      • Senior R.M.
      • Shapiro S.D.
      Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice.
      • Lanone S.
      • Zheng T.
      • Zhu Z.
      • Liu W.
      • Lee C.G.
      • Ma B.
      • Chen Q.
      • Homer R.J.
      • Wang J.
      • Rabach L.A.
      • Rabach M.E.
      • Shipley J.M.
      • Shapiro S.D.
      • Senior R.M.
      • Elias J.A.
      Overlapping and enzyme-specific contributions of matrix metalloproteinases-9 and -12 in IL-13-induced inflammation and remodeling.
      • Leco K.J.
      • Waterhouse P.
      • Sanchez O.H.
      • Gowing K.L.
      • Poole A.R.
      • Wakeham A.
      • Mak T.W.
      • Khokha R.
      Spontaneous air space enlargement in the lungs of mice lacking tissue inhibitor of metalloproteinases-3 (TIMP-3).
      • Niewoehner D.E.
      • Hoidal J.R.
      Lung fibrosis and emphysema: divergent responses to a common injury.
      • Oikonomou N.
      • Thanasopoulou A.
      • Tzouvelekis A.
      • Harokopos V.
      • Paparountas T.
      • Nikitopoulou I.
      • Witke W.
      • Karameris A.
      • Kotanidou A.
      • Bouros D.
      • Aidinis V.
      Gelsolin expression is necessary for the development of modelled pulmonary inflammation and fibrosis.
      • Tang K.
      • Rossiter H.B.
      • Wagner P.D.
      • Breen E.C.
      Lung-targeted VEGF inactivation leads to an emphysema phenotype in mice.
      • Maki J.M.
      • Sormunen R.
      • Lippo S.
      • Kaarteenaho-Wiik R.
      • Soininen R.
      • Myllyharju J.
      Lysyl oxidase is essential for normal development and function of the respiratory system and for the integrity of elastic and collagen fibers in various tissues.
      • Sonnylal S.
      • Shi-Wen X.
      • Leoni P.
      • Naff K.
      • Van Pelt C.S.
      • Nakamura H.
      • Leask A.
      • Abraham D.
      • Bou-Gharios G.
      • de Crombrugghe B.
      Selective expression of connective tissue growth factor in fibroblasts in vivo promotes systemic tissue fibrosis.
      • Nishida T.
      • Kawaki H.
      • Baxter R.M.
      • Deyoung R.A.
      • Takigawa M.
      • Lyons K.M.
      CCN2 (connective tissue growth factor) is essential for extracellular matrix production and integrin signaling in chondrocytes.
      provides ample evidence of the impact of disruption or overexpression of several genes identified in the present study (Table 9). Although manipulation of individual genes has provided insight into their roles in responding to exogenous stimuli or in lung maintenance and repair, it is clear that a better understanding of how these diverse players interact as a system is needed. This is particularly true in understanding disease states because effective treatment may require a therapeutic approach that targets multiple processes. Our data indicate that the emphysema phenotype exhibited by the transgenic SP-C/TNF-α mice is characterized by effects on several pathways (Figure 8). It may be that perturbation of any one of the highly interdependent processes that govern alveolar architecture could result in the emphysema phenotype through subsequent impacts on all or a subset of these pathways. Application of systems biology approaches to ascertain to what extent this is true in other models of emphysema should prove useful in defining the range and specificity of biological processes that play a causal role in disease pathogenesis.
      Table 9Functional Analysis of Select Genes Altered in SP-C/TNF-α Implicated in Emphysema and Fibrosis
      GeneAccession no.Expression in SP-C/TNF mouse
      Expression is represented by the fold change, according to microarray data. Data from real-time PCR are indicated in parentheses, if available.
      ModelRef. no.
      MMP-9NM_0135994.0 (12.4)IL-13–induced airspace enlargement reduced in the absence of MMP-9
      • Lanone S.
      • Zheng T.
      • Zhu Z.
      • Liu W.
      • Lee C.G.
      • Ma B.
      • Chen Q.
      • Homer R.J.
      • Wang J.
      • Rabach L.A.
      • Rabach M.E.
      • Shipley J.M.
      • Shapiro S.D.
      • Senior R.M.
      • Elias J.A.
      Overlapping and enzyme-specific contributions of matrix metalloproteinases-9 and -12 in IL-13-induced inflammation and remodeling.
      MMP-12NM_0086053.8 (36.4)Knockout mice protected from cigarette smoke–induced emphysema
      • Hautamaki R.D.
      • Kobayashi D.K.
      • Senior R.M.
      • Shapiro S.D.
      Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice.
      IL-13–induced airspace enlargement reduced in the absence of MMP-12
      • Lanone S.
      • Zheng T.
      • Zhu Z.
      • Liu W.
      • Lee C.G.
      • Ma B.
      • Chen Q.
      • Homer R.J.
      • Wang J.
      • Rabach L.A.
      • Rabach M.E.
      • Shipley J.M.
      • Shapiro S.D.
      • Senior R.M.
      • Elias J.A.
      Overlapping and enzyme-specific contributions of matrix metalloproteinases-9 and -12 in IL-13-induced inflammation and remodeling.
      TIMP-3NM_011595−2.2 (−5.2)Knockout mice develop airspace enlargement
      • Leco K.J.
      • Waterhouse P.
      • Sanchez O.H.
      • Gowing K.L.
      • Poole A.R.
      • Wakeham A.
      • Mak T.W.
      • Khokha R.
      Spontaneous air space enlargement in the lungs of mice lacking tissue inhibitor of metalloproteinases-3 (TIMP-3).
      VEGFNM_009505−2.9 (−3.2)Conditional knockout results in emphysema
      • Tang K.
      • Rossiter H.B.
      • Wagner P.D.
      • Breen E.C.
      Lung-targeted VEGF inactivation leads to an emphysema phenotype in mice.
      Flt-1NM_010228−3.7Inhibition of VEGF receptors resulted in apoptosis and emphysema
      • Kasahara Y.
      • Tuder R.M.
      • Taraseviciene-Stewart L.
      • Le Cras T.D.
      • Abman S.
      • Hirth P.K.
      • Waltenberger J.
      • Voelkel N.F.
      Inhibition of VEGF receptors causes lung cell apoptosis and emphysema.
      LoxNM_010728−3.3 (−5.7)Inhibition of lysyl oxidase resulted in emphysema, rather than fibrosis, after CdCl2 treatment
      • Niewoehner D.E.
      • Hoidal J.R.
      Lung fibrosis and emphysema: divergent responses to a common injury.
      Knockout mice cannot cross-link collagen or elastin, embryonic lethality, embryonic lungs exhibit emphysema
      • Maki J.M.
      • Sormunen R.
      • Lippo S.
      • Kaarteenaho-Wiik R.
      • Soininen R.
      • Myllyharju J.
      Lysyl oxidase is essential for normal development and function of the respiratory system and for the integrity of elastic and collagen fibers in various tissues.
      GsnNM_146120−2.4Required for the development of fibrosis after bleomycin treatment
      • Oikonomou N.
      • Thanasopoulou A.
      • Tzouvelekis A.
      • Harokopos V.
      • Paparountas T.
      • Nikitopoulou I.
      • Witke W.
      • Karameris A.
      • Kotanidou A.
      • Bouros D.
      • Aidinis V.
      Gelsolin expression is necessary for the development of modelled pulmonary inflammation and fibrosis.
      CtgfNM_010217−1.6 (−2.5)Overexpression in fibroblasts results in fibrosis
      • Sonnylal S.
      • Shi-Wen X.
      • Leoni P.
      • Naff K.
      • Van Pelt C.S.
      • Nakamura H.
      • Leask A.
      • Abraham D.
      • Bou-Gharios G.
      • de Crombrugghe B.
      Selective expression of connective tissue growth factor in fibroblasts in vivo promotes systemic tissue fibrosis.
      Knockout has impaired extracellular matrix deposition
      • Nishida T.
      • Kawaki H.
      • Baxter R.M.
      • Deyoung R.A.
      • Takigawa M.
      • Lyons K.M.
      CCN2 (connective tissue growth factor) is essential for extracellular matrix production and integrin signaling in chondrocytes.
      low asterisk Expression is represented by the fold change, according to microarray data. Data from real-time PCR are indicated in parentheses, if available.
      Figure thumbnail gr8
      Figure 8Simplified schematic of major pathways exhibiting differential expression in the lungs of mice with constitutive expression of TNF-α. Constitutive overexpression of TNF-α under the control of the SP-C promoter results in chronic lung inflammation and emphysema. Increased cytokine and chemokine production and tight junction alterations promote the influx of inflammatory cells into the lungs. Increased protease production and activation, coupled with reduced expression of protease inhibitors, increases the processing and destruction of basement membrane proteins and cytokine maturation. The disruption of repair and maintenance functions of endothelial cells and fibroblasts contributes to the alveolar wall damage and impairs the capacity to repair. Together, effects on pathways critical to the maintenance of lung structure result in airspace enlargement characteristic of emphysema. CTGF, connective tissue growth factor; Flt, FMS-like tyrosine kinase (VEGF receptor); Id1, inhibitor of DNA binding 1; LTA, lymphotoxin A; PCOLCE, procollagen C-endopeptidase enhancer; PDGF, platelet-derived growth factor; RECK, reversion-inducing cysteine-rich protein with kazal motifs; SAA, serum amyloid A.
      Interspecies comparisons of specific genes implicated in disease processes are complicated by several factors, including the possibility that gene products may serve different functions. A more robust approach compares biological processes, because these may be more conserved across species. Gene expression profiling conducted in human lung tissue distinguished between severe emphysema and mild or no emphysema, through detection of increased expression of genes involved in oxidative stress, inflammation, and matrix remodeling, and decreased expression of endothelium–related genes,
      • Spira A.
      • Beane J.
      • Pinto-Plata V.
      • Kadar A.
      • Liu G.
      • Shah V.
      • Celli B.
      • Brody J.S.
      Gene expression profiling of human lung tissue from smokers with severe emphysema.
      consistent with the profiles observed in the SP-C/TNF-α transgenic mice. Nevertheless, there is some evidence supporting involvement of specific genes in emphysema. Patients with chronic obstructive pulmonary disorder exhibit high levels of TNF-α in the lungs,
      • Keatings V.M.
      • Collins P.D.
      • Scott D.M.
      • Barnes P.J.
      Differences in interleukin-8 and tumor necrosis factor-alpha in induced sputum from patients with chronic obstructive pulmonary disease or asthma.
      and knockout of TNF receptors in mice affords some protection from cigarette smoke–induced emphysema.
      • Churg A.
      • Wang R.D.
      • Tai H.
      • Wang X.
      • Xie C.
      • Wright J.L.
      Tumor necrosis factor-alpha drives 70% of cigarette smoke-induced emphysema in the mouse.
      Emphysematous lungs exhibit increased MMP-2, MMP-9, MMP-12, and membrane type 1–MMP.
      • Ohnishi K.
      • Takagi M.
      • Kurokawa Y.
      • Satomi S.
      • Konttinen Y.T.
      Matrix metalloproteinase-mediated extracellular matrix protein degradation in human pulmonary emphysema.
      • Molet S.
      • Belleguic C.
      • Lena H.
      • Germain N.
      • Bertrand C.P.
      • Shapiro S.D.
      • Planquois J.M.
      • Delaval P.
      • Lagente V.
      Increase in macrophage elastase (MMP-12) in lungs from patients with chronic obstructive pulmonary disease.
      In animal models, knockout of MMP-9 or MMP-12 reduces the airspace enlargement resulting from IL-13 overexpression,
      • Lanone S.
      • Zheng T.
      • Zhu Z.
      • Liu W.
      • Lee C.G.
      • Ma B.
      • Chen Q.
      • Homer R.J.
      • Wang J.
      • Rabach L.A.
      • Rabach M.E.
      • Shipley J.M.
      • Shapiro S.D.
      • Senior R.M.
      • Elias J.A.
      Overlapping and enzyme-specific contributions of matrix metalloproteinases-9 and -12 in IL-13-induced inflammation and remodeling.
      knockout of MMP-12 protects against cigarette smoke–induced emphysema,
      • Hautamaki R.D.
      • Kobayashi D.K.
      • Senior R.M.
      • Shapiro S.D.
      Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice.
      and knockout of TIMP-3 produces emphysema.
      • Leco K.J.
      • Waterhouse P.
      • Sanchez O.H.
      • Gowing K.L.
      • Poole A.R.
      • Wakeham A.
      • Mak T.W.
      • Khokha R.
      Spontaneous air space enlargement in the lungs of mice lacking tissue inhibitor of metalloproteinases-3 (TIMP-3).
      Lower levels of VEGF have been measured in human emphysematous lungs,
      • Kasahara Y.
      • Tuder R.M.
      • Cool C.D.
      • Lynch D.A.
      • Flores S.C.
      • Voelkel N.F.
      Endothelial cell death and decreased expression of vascular endothelial growth factor and vascular endothelial growth factor receptor 2 in emphysema.
      and inhibition of VEGF receptors in rats leads to airspace enlargement through endothelial and epithelial apoptosis.
      • Kasahara Y.
      • Tuder R.M.
      • Taraseviciene-Stewart L.
      • Le Cras T.D.
      • Abman S.
      • Hirth P.K.
      • Waltenberger J.
      • Voelkel N.F.
      Inhibition of VEGF receptors causes lung cell apoptosis and emphysema.
      It is possible that emphysematous changes in the lungs are the result of interactions and combined actions of several pathways. Similarly, the observed resistance to acute treatment with fibrotic agents
      • Fujita M.
      • Shannon J.M.
      • Morikawa O.
      • Gauldie J.
      • Hara N.
      • Mason R.J.
      Overexpression of TNF-alpha diminishes pulmonary fibrosis induced by bleomycin or TGF-beta.
      and retinoic acid rescue
      • Fujita M.
      • Ye Q.
      • Ouchi H.
      • Nakashima N.
      • Hamada N.
      • Hagimoto N.
      • Kuwano K.
      • Mason R.J.
      • Nakanishi Y.
      Retinoic acid fails to reverse emphysema in adult mouse models.
      may well be the result of several effects that persist as a result of the constitutive expression of TNF-α, including impaired repair and maintenance functions. Involvement of TNF-α in fibrosis is controversial. Bleomycin or silica-induced fibrosis can be reverted using TNF-α antagonists,
      • Piguet P.F.
      • Vesin C.
      Treatment by human recombinant soluble TNF receptor of pulmonary fibrosis induced by bleomycin or silica in mice.
      indicating the critical involvement of TNF-α in these experimental models. However, because fibrosis can proceed in the absence of inflammation, and anti-inflammatory treatments have not proved effective in treating the human disease, there has been a shift in the view of fibrosis as an inflammatory disease to one of disordered wound healing, with TGF-β signaling playing a prominent role.
      • Hardie W.D.
      • Glasser S.W.
      • Hagood J.S.
      Emerging concepts in the pathogenesis of lung fibrosis.
      Emphysema and fibrosis have been detected in the same lungs,
      • Cottin V.
      • Nunes H.
      • Brillet P.Y.
      • Delaval P.
      • Devouassoux G.
      • Tillie-Leblond I.
      • Israel-Biet D.
      • Court-Fortune
      • Valeyre D.
      • Cordier J.F.
      Groupe D'Etude et de Recherche sur les Maladies Orphelines Pulmonaires (GERM O P)
      Combined pulmonary fibrosis and emphysema: a distinct underrecognised entity.
      suggesting that common processes initiated by cigarette smoking or other agents can lead to establishment of traits of both pathological conditions in susceptible individuals. Remarkably, it was recently shown that the presence of detectable interstitial lung abnormalities in smokers was inversely associated with the degree of emphysema.
      • Washko G.R.
      • Hunninghake G.M.
      • Fernandez I.E.
      • Nishino M.
      • Okajima Y.
      • Yamashiro T.
      • Ross J.C.
      • Estepar R.S.
      • Lynch D.A.
      • Brehm J.M.
      • Andriole K.P.
      • Diaz A.A.
      • Khorasani R.
      • D'Aco K.
      • Sciurba F.C.
      • Silverman E.K.
      • Hatabu H.
      • Rosas I.O.
      Lung volumes and emphysema in smokers with interstitial lung abnormalities.
      It is tempting to speculate that factors driving emphysema may inhibit the development of fibrosis, as is implied by the SP-C/TNF-α lung transcriptome analyses or, conversely, that fibrotic changes inhibit emphysema, in which the local pulmonary environment is permissive to fibrosis.
      In summary, mice with increased expression of TNF-α in type II alveolar epithelial cells exhibit pronounced transcriptional changes in key pathways that control alveolar wall integrity. We found striking increases in the expression of genes involved in inflammatory and proteolytic processes, in line with the pronounced inflammation and alveolar wall destruction. We also detected decreased expression of genes involved in anti-protease and repair and maintenance processes, suggesting that regulatory systems that control protease activities were impaired as a result of constitutive expression of TNF-α. The decreased expression of factors involved in myofibroblast differentiation, matrix deposition, and cross-linking of basement membrane proteins may contribute to the loss of connective tissue scaffolding and to the resistance to induction of fibrosis observed in this model. Our data indicate that the resulting pathological condition, although arising from overexpression of a single gene, is characterized by critical effects on several pathways relevant to the control of alveolar wall homeostasis (Figure 8). Overall, the transcriptional changes correspond well with the observed phenotype and provide evidence supporting involvement of key genes and pathways in disease pathogenesis. Understanding the mechanisms that govern normal and dysfunctional lung responses to injury has the potential to uncover therapeutic opportunities to address chronic lung diseases. The present work identifies several key transcriptional changes that can serve as targets for future assessment using available molecular biology tools and provides a data set that can be interrogated for hypothesis generation. Comparison of disease models using transcriptome-wide screening should prove useful for the identification of genes, pathways, and processes that warrant further investigation in diseases of unknown cause.

      Acknowledgments

      We thank Pat Goegan and Erica Blais for conducting the necropsies and Lorraine Casavant for performing differential cell counts and histological analysis.

      Supplementary data

      • Supplemental Appendix S1

        Heat maps of microarray gene expression for all KEGG pathways. Heat maps were generated for all KEGG pathways using microarray data for wild-type and SP-C/TNF-α mice (n = 5 per genotype). Red indicates a high level of expression relative to the reference RNA, whereas green indicates low relative expression. The heat maps can be interrogated for genes (using the gene symbol) and pathways (using the KEGG pathway name) of interest using search functions.

      • Supplemental Figure S1

        Validation of selected microarray probe intensity ratios [SP-C/TNF-α:wild type (WT)] by real-time PCR. Asterisks indicate probes within background for the wild-type samples according to microarray analysis, which likely accounts for the difference in fold change between microarray and PCR approaches.

      References

        • Owen C.A.
        Roles for proteinases in the pathogenesis of chronic obstructive pulmonary disease.
        Int J Chron Obstruct Pulmon Dis. 2008; 3: 253-268
        • Churg A.
        • Wright J.L.
        Proteases and emphysema.
        Curr Opin Pulm Med. 2005; 11: 153-159
        • Henson P.M.
        • Vandivier R.W.
        • Douglas I.S.
        Cell death, remodeling, and repair in chronic obstructive pulmonary disease.
        Proc Am Thorac Soc. 2006; 3: 713-717
        • Tsuji T.
        • Aoshiba K.
        • Nagai A.
        Alveolar cell senescence in patients with pulmonary emphysema.
        Am J Respir Crit Care Med. 2006; 174: 886-893
        • Kasahara Y.
        • Tuder R.M.
        • Taraseviciene-Stewart L.
        • Le Cras T.D.
        • Abman S.
        • Hirth P.K.
        • Waltenberger J.
        • Voelkel N.F.
        Inhibition of VEGF receptors causes lung cell apoptosis and emphysema.
        J Clin Invest. 2000; 106: 1311-1319
        • Wright J.L.
        • Churg A.
        Current concepts in mechanisms of emphysema.
        Toxicol Pathol. 2007; 35: 111-115
        • Lewis C.C.
        • Yang J.Y.
        • Huang X.
        • Banerjee S.K.
        • Blackburn M.R.
        • Baluk P.
        • McDonald D.M.
        • Blackwell T.S.
        • Nagabhushanam V.
        • Peters W.
        • Voehringer D.
        • Erle D.J.
        Disease-specific gene expression profiling in multiple models of lung disease.
        Am J Respir Crit Care Med. 2008; 177: 376-387
        • Churg A.
        • Wang R.D.
        • Tai H.
        • Wang X.
        • Xie C.
        • Wright J.L.
        Tumor necrosis factor-alpha drives 70% of cigarette smoke-induced emphysema in the mouse.
        Am J Respir Crit Care Med. 2004; 170: 492-498
        • Mukhopadhyay S.
        • Hoidal J.R.
        • Mukherjee T.K.
        Role of TNFalpha in pulmonary pathophysiology.
        Respir Res. 2006; 7: 125
        • Piguet P.F.
        • Collart M.A.
        • Grau G.E.
        • Sappino A.P.
        • Vassalli P.
        Requirement of tumour necrosis factor for development of silica-induced pulmonary fibrosis.
        Nature. 1990; 344: 245-247
        • Miyazaki Y.
        • Araki K.
        • Vesin C.
        • Garcia I.
        • Kapanci Y.
        • Whitsett J.A.
        • Piguet P.F.
        • Vassalli P.
        Expression of a tumor necrosis factor-alpha transgene in murine lung causes lymphocytic and fibrosing alveolitis: a mouse model of progressive pulmonary fibrosis.
        J Clin Invest. 1995; 96: 250-259
        • Fujita M.
        • Shannon J.M.
        • Irvin C.G.
        • Fagan K.A.
        • Cool C.
        • Augustin A.
        • Mason R.J.
        Overexpression of tumor necrosis factor-alpha produces an increase in lung volumes and pulmonary hypertension.
        Am J Physiol Lung Cell Mol Physiol. 2001; 280: L39-L49
        • Lundblad L.K.
        • Thompson-Figueroa J.
        • Leclair T.
        • Sullivan M.J.
        • Poynter M.E.
        • Irvin C.G.
        • Bates J.H.
        Tumor necrosis factor-alpha overexpression in lung disease: a single cause behind a complex phenotype.
        Am J Respir Crit Care Med. 2005; 171: 1363-1370
        • Piguet P.F.
        • Ribaux C.
        • Karpuz V.
        • Grau G.E.
        • Kapanci Y.
        Expression and localization of tumor necrosis factor-alpha and its mRNA in idiopathic pulmonary fibrosis.
        Am J Pathol. 1993; 143: 651-655
        • Grutters J.C.
        • du Bois R.M.
        Genetics of fibrosing lung diseases.
        Eur Respir J. 2005; 25: 915-927
        • Piguet P.F.
        • Collart M.A.
        • Grau G.E.
        • Kapanci Y.
        • Vassalli P.
        Tumor necrosis factor/cachectin plays a key role in bleomycin-induced pneumopathy and fibrosis.
        J Exp Med. 1989; 170: 655-663
        • Fujita M.
        • Shannon J.M.
        • Morikawa O.
        • Gauldie J.
        • Hara N.
        • Mason R.J.
        Overexpression of TNF-alpha diminishes pulmonary fibrosis induced by bleomycin or TGF-beta.
        Am J Respir Cell Mol Biol. 2003; 29: 669-676
        • Thomson E.M.
        • Williams A.
        • Yauk C.L.
        • Vincent R.
        Toxicogenomic analysis of susceptibility to inhaled urban particulate matter in mice with chronic lung inflammation.
        Part Fibre Toxicol. 2009; 6: 6
        • Yang Y.H.
        • Dudoit S.
        • Luu P.
        • Lin D.M.
        • Peng V.
        • Ngai J.
        • Speed T.P.
        Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation.
        Nucleic Acids Res. 2002; 30: e15
        • Wu H.
        • Kerr M.K.
        • Cui X.
        • Churchill G.A.
        MAANOVA: a software package for the analysis of spotted cDNA microarray experiments.
        in: Parmigiani G. Garrett E.S. Irizarry R.A. Zeger S. The Analysis of Gene Expression Data: Methods and Software. Springer-Verlag, New York2003: 313-431
        • Cui X.G.
        • Hwang J.T.G.
        • Qiu J.
        • Blades N.J.
        • Churchill G.A.
        Improved statistical tests for differential gene expression by shrinking variance components.
        Biostatistics. 2005; 6: 59-75
        • Benjamini Y.
        • Hochberg Y.
        Controlling the false discovery rate: a practical and powerful approach to multiple testing.
        J R Stat Soc. 1995; 57: 289-300
        • Dennis Jr, G.
        • Sherman B.T.
        • Hosack D.A.
        • Yang J.
        • Gao W.
        • Lane H.C.
        • Lempicki R.A.
        DAVID: Database for Annotation, Visualization, and Integrated Discovery.
        Genome Biol. 2003; 4: P3
        • Huang da W.
        • Sherman B.T.
        • Lempicki R.A.
        Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources.
        Nat Protoc. 2009; 4: 44-57
        • Min H.Y.
        • Spiegelman B.M.
        Adipsin, the adipocyte serine protease: gene structure and control of expression by tumor necrosis factor.
        Nucleic Acids Res. 1986; 14: 8879-8892
        • Becker P.M.
        • Kazi A.A.
        • Wadgaonkar R.
        • Pearse D.B.
        • Kwiatkowski D.
        • Garcia J.G.
        Pulmonary vascular permeability and ischemic injury in gelsolin-deficient mice.
        Am J Respir Cell Mol Biol. 2003; 28: 478-484
        • Scotton C.J.
        • Chambers R.C.
        Molecular targets in pulmonary fibrosis: the myofibroblast in focus.
        Chest. 2007; 132: 1311-1321
        • Gao S.
        • Chen K.
        • Zhao Y.
        • Rich C.B.
        • Chen L.
        • Li S.J.
        • Toselli P.
        • Stone P.
        • Li W.
        Transcriptional and posttranscriptional inhibition of lysyl oxidase expression by cigarette smoke condensate in cultured rat fetal lung fibroblasts.
        Toxicol Sci. 2005; 87: 197-203
        • Vuillemenot B.R.
        • Rodriguez J.F.
        • Hoyle G.W.
        Lymphoid tissue and emphysema in the lungs of transgenic mice inducibly expressing tumor necrosis factor-alpha.
        Am J Respir Cell Mol Biol. 2004; 30: 438-448
        • Wang J.
        • Lenardo M.J.
        Roles of caspases in apoptosis, development, and cytokine maturation revealed by homozygous gene deficiencies.
        J Cell Sci. 2000; 113: 753-757
        • Ke S.
        • Rabson A.B.
        • Germino J.F.
        • Gallo M.A.
        • Tian Y.
        Mechanism of suppression of cytochrome P-450 1A1 expression by tumor necrosis factor-alpha and lipopolysaccharide.
        J Biol Chem. 2001; 276: 39638-39644
        • Mazzon E.
        • Cuzzocrea S.
        Role of TNF-alpha in lung tight junction alteration in mouse model of acute lung inflammation.
        Respir Res. 2007; 8: 75
        • Rahman I.
        Regulation of glutathione in inflammation and chronic lung diseases.
        Mutat Res. 2005; 579: 58-80
        • Hautamaki R.D.
        • Kobayashi D.K.
        • Senior R.M.
        • Shapiro S.D.
        Requirement for macrophage elastase for cigarette smoke-induced emphysema in mice.
        Science. 1997; 277: 2002-2004
        • Lanone S.
        • Zheng T.
        • Zhu Z.
        • Liu W.
        • Lee C.G.
        • Ma B.
        • Chen Q.
        • Homer R.J.
        • Wang J.
        • Rabach L.A.
        • Rabach M.E.
        • Shipley J.M.
        • Shapiro S.D.
        • Senior R.M.
        • Elias J.A.
        Overlapping and enzyme-specific contributions of matrix metalloproteinases-9 and -12 in IL-13-induced inflammation and remodeling.
        J Clin Invest. 2002; 110: 463-474
        • Leco K.J.
        • Waterhouse P.
        • Sanchez O.H.
        • Gowing K.L.
        • Poole A.R.
        • Wakeham A.
        • Mak T.W.
        • Khokha R.
        Spontaneous air space enlargement in the lungs of mice lacking tissue inhibitor of metalloproteinases-3 (TIMP-3).
        J Clin Invest. 2001; 108: 817-829
        • Gearing A.J.
        • Beckett P.
        • Christodoulou M.
        • Churchill M.
        • Clements J.
        • Davidson A.H.
        • Drummond A.H.
        • Galloway W.A.
        • Gilbert R.
        • Gordon J.L.
        • Leber T.M.
        • Mangan M.
        • Miller K.
        • Nayee P.
        • Owen K.
        • Patel S.
        • Thomas W.
        • Wells G.
        • Wood L.M.
        • Woolley K.
        Processing of tumour necrosis factor-alpha precursor by metalloproteinases.
        Nature. 1994; 370: 555-557
        • Amour A.
        • Slocombe P.M.
        • Webster A.
        • Butler M.
        • Knight C.G.
        • Smith B.J.
        • Stephens P.E.
        • Shelley C.
        • Hutton M.
        • Knauper V.
        • Docherty A.J.
        • Murphy G.
        TNF-alpha converting enzyme (TACE) is inhibited by TIMP-3.
        FEBS Lett. 1998; 435: 39-44
        • Son D.S.
        • Roby K.F.
        • Terranova P.F.
        Tumor necrosis factor-alpha induces serum amyloid A3 in mouse granulosa cells.
        Endocrinology. 2004; 145: 2245-2252
        • Vallon R.
        • Freuler F.
        • Desta-Tsedu N.
        • Robeva A.
        • Dawson J.
        • Wenner P.
        • Engelhardt P.
        • Boes L.
        • Schnyder J.
        • Tschopp C.
        • Urfer R.
        • Baumann G.
        Serum amyloid A (apoSAA) expression is up-regulated in rheumatoid arthritis and induces transcription of matrix metalloproteinases.
        J Immunol. 2001; 166: 2801-2807
        • Wolters P.J.
        • Chapman H.A.
        Importance of lysosomal cysteine proteases in lung disease.
        Respir Res. 2000; 1: 170-177
        • Steiglitz B.M.
        • Keene D.R.
        • Greenspan D.S.
        PCOLCE2 encodes a functional procollagen C-proteinase enhancer (PCPE2) that is a collagen-binding protein differing in distribution of expression and post-translational modification from the previously described PCPE1.
        J Biol Chem. 2002; 277: 49820-49830
        • Baker A.H.
        • Edwards D.R.
        • Murphy G.
        Metalloproteinase inhibitors: biological actions and therapeutic opportunities.
        J Cell Sci. 2002; 115: 3719-3727
        • Overall C.M.
        • Dean R.A.
        Degradomics: systems biology of the protease web: pleiotropic roles of MMPs in cancer.
        Cancer Metastasis Rev. 2006; 25: 69-75
        • Voelkel N.F.
        • Douglas I.S.
        • Nicolls M.
        Angiogenesis in chronic lung disease.
        Chest. 2007; 131: 874-879
        • Fujita M.
        • Mason R.J.
        • Cool C.
        • Shannon J.M.
        • Hara N.
        • Fagan K.A.
        Pulmonary hypertension in TNF-alpha-overexpressing mice is associated with decreased VEGF gene expression.
        J Appl Physiol. 2002; 93: 2162-2170
        • Fujita M.
        • Ikegame S.
        • Ye Q.
        • Harada E.
        • Ouchi H.
        • Inoshima I.
        • Watanabe K.
        • Mason R.J.
        • Nakanishi Y.
        Attenuation of pulmonary hypertension, but not emphysematous change, by breeding emphysema model mice at sea level.
        Cytokine. 2008; 41: 286-292
        • Gao D.
        • Nolan D.J.
        • Mellick A.S.
        • Bambino K.
        • McDonnell K.
        • Mittal V.
        Endothelial progenitor cells control the angiogenic switch in mouse lung metastasis.
        Science. 2008; 319: 195-198
        • Chaouat A.
        • Naeije R.
        • Weitzenblum E.
        Pulmonary hypertension in COPD.
        Eur Respir J. 2008; 32: 1371-1385
        • Niewoehner D.E.
        • Hoidal J.R.
        Lung fibrosis and emphysema: divergent responses to a common injury.
        Science. 1982; 217: 359-360
        • Sirianni F.E.
        • Milaninezhad A.
        • Chu F.S.
        • Walker D.C.
        Alteration of fibroblast architecture and loss of Basal lamina apertures in human emphysematous lung.
        Am J Respir Crit Care Med. 2006; 173: 632-638
        • Yu F.
        • Chou C.W.
        • Chen C.C.
        TNF-alpha suppressed TGF-beta-induced CTGF expression by switching the binding preference of p300 from Smad4 to p65.
        Cell Signal. 2009; 21: 867-872
        • Liu X.
        • Kelm Jr, R.J.
        • Strauch A.R.
        Transforming growth factor beta1-mediated activation of the smooth muscle alpha-actin gene in human pulmonary myofibroblasts is inhibited by tumor necrosis factor-alpha via mitogen-activated protein kinase kinase 1-dependent induction of the Egr-1 transcriptional repressor.
        Mol Biol Cell. 2009; 20: 2174-2185
        • Wicks J.
        • Haitchi H.M.
        • Holgate S.T.
        • Davies D.E.
        • Powell R.M.
        Enhanced upregulation of smooth muscle related transcripts by TGF beta2 in asthmatic (myo) fibroblasts.
        Thorax. 2006; 61: 313-319
        • Oikonomou N.
        • Thanasopoulou A.
        • Tzouvelekis A.
        • Harokopos V.
        • Paparountas T.
        • Nikitopoulou I.
        • Witke W.
        • Karameris A.
        • Kotanidou A.
        • Bouros D.
        • Aidinis V.
        Gelsolin expression is necessary for the development of modelled pulmonary inflammation and fibrosis.
        Thorax. 2009; 64: 467-475
        • Adamson I.Y.
        • Hedgecock C.
        • Bowden D.H.
        Epithelial cell-fibroblast interactions in lung injury and repair.
        Am J Pathol. 1990; 137: 385-392
        • Churg A.
        • Zhou S.
        • Wang X.
        • Wang R.
        • Wright J.L.
        The role of interleukin-1beta in murine cigarette smoke-induced emphysema and small airway remodeling.
        Am J Respir Cell Mol Biol. 2009; 40: 482-490
        • Lian X.
        • Yan C.
        • Qin Y.
        • Knox L.
        • Li T.
        • Du H.
        Neutral lipids and peroxisome proliferator-activated receptor-{gamma} control pulmonary gene expression and inflammation-triggered pathogenesis in lysosomal acid lipase knockout mice.
        Am J Pathol. 2005; 167: 813-821
        • Koth L.L.
        • Cambier C.J.
        • Ellwanger A.
        • Solon M.
        • Hou L.
        • Lanier L.L.
        • Abram C.L.
        • Hamerman J.A.
        • Woodruff P.G.
        DAP12 is required for macrophage recruitment to the lung in response to cigarette smoke and chemotaxis toward CCL2.
        J Immunol. 2010; 184: 6522-6528
        • Bonniaud P.
        • Martin G.
        • Margetts P.J.
        • Ask K.
        • Robertson J.
        • Gauldie J.
        • Kolb M.
        Connective tissue growth factor is crucial to inducing a profibrotic environment in “fibrosis-resistant” BALB/c mouse lungs.
        Am J Respir Cell Mol Biol. 2004; 31: 510-516
        • Tang K.
        • Rossiter H.B.
        • Wagner P.D.
        • Breen E.C.
        Lung-targeted VEGF inactivation leads to an emphysema phenotype in mice.
        J Appl Physiol. 2004; 97: 1559-1566
        • Maki J.M.
        • Sormunen R.
        • Lippo S.
        • Kaarteenaho-Wiik R.
        • Soininen R.
        • Myllyharju J.
        Lysyl oxidase is essential for normal development and function of the respiratory system and for the integrity of elastic and collagen fibers in various tissues.
        Am J Pathol. 2005; 167: 927-936
        • Sonnylal S.
        • Shi-Wen X.
        • Leoni P.
        • Naff K.
        • Van Pelt C.S.
        • Nakamura H.
        • Leask A.
        • Abraham D.
        • Bou-Gharios G.
        • de Crombrugghe B.
        Selective expression of connective tissue growth factor in fibroblasts in vivo promotes systemic tissue fibrosis.
        Arthritis Rheum. 2010; 62: 1523-1532
        • Nishida T.
        • Kawaki H.
        • Baxter R.M.
        • Deyoung R.A.
        • Takigawa M.
        • Lyons K.M.
        CCN2 (connective tissue growth factor) is essential for extracellular matrix production and integrin signaling in chondrocytes.
        J Cell Commun Signal. 2007; 1: 45-58
        • Spira A.
        • Beane J.
        • Pinto-Plata V.
        • Kadar A.
        • Liu G.
        • Shah V.
        • Celli B.
        • Brody J.S.
        Gene expression profiling of human lung tissue from smokers with severe emphysema.
        Am J Respir Cell Mol Biol. 2004; 31: 601-610
        • Keatings V.M.
        • Collins P.D.
        • Scott D.M.
        • Barnes P.J.
        Differences in interleukin-8 and tumor necrosis factor-alpha in induced sputum from patients with chronic obstructive pulmonary disease or asthma.
        Am J Respir Crit Care Med. 1996; 153: 530-534
        • Ohnishi K.
        • Takagi M.
        • Kurokawa Y.
        • Satomi S.
        • Konttinen Y.T.
        Matrix metalloproteinase-mediated extracellular matrix protein degradation in human pulmonary emphysema.
        Lab Invest. 1998; 78: 1077-1087
        • Molet S.
        • Belleguic C.
        • Lena H.
        • Germain N.
        • Bertrand C.P.
        • Shapiro S.D.
        • Planquois J.M.
        • Delaval P.
        • Lagente V.
        Increase in macrophage elastase (MMP-12) in lungs from patients with chronic obstructive pulmonary disease.
        Inflamm Res. 2005; 54: 31-36
        • Kasahara Y.
        • Tuder R.M.
        • Cool C.D.
        • Lynch D.A.
        • Flores S.C.
        • Voelkel N.F.
        Endothelial cell death and decreased expression of vascular endothelial growth factor and vascular endothelial growth factor receptor 2 in emphysema.
        Am J Respir Crit Care Med. 2001; 163: 737-744
        • Fujita M.
        • Ye Q.
        • Ouchi H.
        • Nakashima N.
        • Hamada N.
        • Hagimoto N.
        • Kuwano K.
        • Mason R.J.
        • Nakanishi Y.
        Retinoic acid fails to reverse emphysema in adult mouse models.
        Thorax. 2004; 59: 224-230
        • Piguet P.F.
        • Vesin C.
        Treatment by human recombinant soluble TNF receptor of pulmonary fibrosis induced by bleomycin or silica in mice.
        Eur Respir J. 1994; 7: 515-518
        • Hardie W.D.
        • Glasser S.W.
        • Hagood J.S.
        Emerging concepts in the pathogenesis of lung fibrosis.
        Am J Pathol. 2009; 175: 3-16
        • Cottin V.
        • Nunes H.
        • Brillet P.Y.
        • Delaval P.
        • Devouassoux G.
        • Tillie-Leblond I.
        • Israel-Biet D.
        • Court-Fortune
        • Valeyre D.
        • Cordier J.F.
        • Groupe D'Etude et de Recherche sur les Maladies Orphelines Pulmonaires (GERM O P)
        Combined pulmonary fibrosis and emphysema: a distinct underrecognised entity.
        Eur Respir J. 2005; 26: 586-593
        • Washko G.R.
        • Hunninghake G.M.
        • Fernandez I.E.
        • Nishino M.
        • Okajima Y.
        • Yamashiro T.
        • Ross J.C.
        • Estepar R.S.
        • Lynch D.A.
        • Brehm J.M.
        • Andriole K.P.
        • Diaz A.A.
        • Khorasani R.
        • D'Aco K.
        • Sciurba F.C.
        • Silverman E.K.
        • Hatabu H.
        • Rosas I.O.
        Lung volumes and emphysema in smokers with interstitial lung abnormalities.
        N Engl J Med. 2011; 364: 897-906