Exuberant Fibroblast Activity Compromises Lung Function Via ADAMTS4
Total Page:16
File Type:pdf, Size:1020Kb
Article Exuberant fibroblast activity compromises lung function via ADAMTS4 https://doi.org/10.1038/s41586-020-2877-5 David F. Boyd1, E. Kaitlynn Allen1, Adrienne G. Randolph2,3, Xi-zhi J. Guo1, Yunceng Weng4, Catherine J. Sanders1, Resha Bajracharya1, Natalie K. Lee5, Clifford S. Guy1, Peter Vogel6, Received: 16 July 2019 Wenda Guan4, Yimin Li4, Xiaoqing Liu4, Tanya Novak2,3, Margaret M. Newhams2, Accepted: 30 July 2020 Thomas P. Fabrizio5, Nicholas Wohlgemuth5, Peter M. Mourani7, PALISI Pediatric Intensive Care Influenza (PICFLU) Investigators*, Thomas N. Wight8, Stacey Schultz-Cherry5, Published online: 28 October 2020 Stephania A. Cormier9,10, Kathryn Shaw-Saliba11, Andrew Pekosz12, Richard E. Rothman11, Check for updates Kuan-Fu Chen13,14, Zifeng Yang4, Richard J. Webby5, Nanshan Zhong4, Jeremy Chase Crawford1 & Paul G. Thomas1 ✉ Severe respiratory infections can result in acute respiratory distress syndrome (ARDS)1. There are no efective pharmacological therapies that have been shown to improve outcomes for patients with ARDS. Although the host infammatory response limits spread of and eventually clears the pathogen, immunopathology is a major contributor to tissue damage and ARDS1,2. Here we demonstrate that respiratory viral infection induces distinct fbroblast activation states, which we term extracellular matrix (ECM)-synthesizing, damage-responsive and interferon-responsive states. We provide evidence that excess activity of damage-responsive lung fbroblasts drives lethal immunopathology during severe infuenza virus infection. By producing ECM-remodelling enzymes—in particular the ECM protease ADAMTS4—and infammatory cytokines, damage-responsive fbroblasts modify the lung microenvironment to promote robust immune cell infltration at the expense of lung function. In three cohorts of human participants, the levels of ADAMTS4 in the lower respiratory tract were associated with the severity of infection with seasonal or avian infuenza virus. A therapeutic agent that targets the ECM protease activity of damage-responsive lung fbroblasts could provide a promising approach to preserving lung function and improving clinical outcomes following severe respiratory infections. Respiratory infections are a leading cause of morbidity and mortal- proteases are upregulated and remodel the ECM, facilitating the migra- ity3. These infections can result in ARDS with pulmonary oedema tion of immune cells to sites of inflammation5,7. Non-immune lung cells, and hypoxia, causing mild to severe respiratory failure1. Much of including epithelial cells, endothelial cells and fibroblasts, coordi- the lung damage induced by viral infection is a result of infiltrating nate immune responses and directly mediate lung function8. Through immune cells, which kill infected and bystander cells4. Defining the their remodelling activity, specific cell populations can influence the mechanisms that alter the balance between pathogen clearance and outcome of infection and long-term sequelae, as has been described for immunopathology may enable identification of strategies to improve myofibroblasts and lung fibrosis9. The role of lung stromal cell popula- outcomes following ARDS2. Therapeutic targets include components tions in coordinating host responses to active respiratory infections of the lung ECM, which provides structural support that is critical has received less attention than their role in late-stage repair following for lung function and tissue-specific signals to coordinate immune pathogen clearance. While there is extensive literature on how immune responses5. cells regulate the host response to respiratory viral infection8, there is The ECM consists of structural proteins as well as proteases and less information on the heterogeneity and role of non-haematopoietic glycosidases that degrade or modify the ECM6. Upon lung injury, ECM cells. 1Department of Immunology, St Jude Children’s Research Hospital, Memphis, TN, USA. 2Boston Children’s Hospital, Department of Anesthesiology, Critical Care and Pain Medicine, Boston, MA, USA. 3Department of Anesthesia, Harvard Medical School, Boston, MA, USA. 4State Key Laboratory of Respiratory Diseases, National Clinical Research Center for Respiratory Disease, Guangzhou Institute of Respiratory Health, The First Affiliated Hospital of Guangzhou Medical University, Guangzhou, China. 5Department of Infectious Diseases, St Jude Children’s Research Hospital, Memphis, TN, USA. 6Veterinary Pathology Core, St Jude Children’s Research Hospital, Memphis, TN, USA. 7Section of Critical Care Medicine, Department of Pediatrics, University of Colorado School of Medicine and Children’s Hospital Colorado, Aurora, CO, USA. 8Matrix Biology Program, Benaroya Research Institute, Seattle, WA, USA. 9Department of Biological Sciences, Louisiana State University, Baton Rouge, LA, USA. 10Pennington Biomedical Research Center, Baton Rouge, LA, USA. 11Department of Emergency Medicine and Medicine, Infectious Diseases, Johns Hopkins University School of Medicine, Baltimore, MD, USA. 12Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA. 13Department of Emergency Medicine of Chang Gung Memorial Hospital at Keelung, Keelung City, Taiwan. 14Clinical Informatics and Medical Statistics Research Center of Chang Gung University, Taoyuan, Taiwan. *A list of authors and their affiliations appears at the end of the paper. ✉e-mail: [email protected] 466 | Nature | Vol 587 | 19 November 2020 a Inuenza A PR8 b Mouse lung cells Fig. 1 | Single-cell gene-expression profiling of Single•cell T cells − suspension Type 1 epithelial Type 2 epithelial CD45 cells during severe influenza A virus C57BL/6 mice Mast 1 dpi 3 dpi 6 dpi cells Pericytes Smooth infection. a, Schematic of sample collection for 0 dpi 25 muscle cells scGEX. SSC, side scatter. b, t-distributed stochastic Ciliated B cells Collect whole lungs epithelial neighbour embedding (t-SNE) projection of mouse – Sort CD45 cells Run scGEX Myeloid Label individual 0 lung cells based on scGEX. Data from all time points 250K (5′ 10× Genomics) Club cells lung samples 200K Platelets Lymphatic and mice were aggregated (n = 5 (0, 1 and 3 dpi), and 150K -SNE2 t endothelial C SS 100K n = 4 (6 dpi)). Approximately 40,800 individual cells 50K –25 Neutrophils 0 are represented in the figure. c, t-SNE projection of 10–3 0 103 104 105 Fibroblasts Vascular endothelial CD45 Mesothelial mouse mesenchymal cells expressing Col1a2. −25 025 t-SNE1 d, t-SNE projection of mouse mesenchymal cells at c Mouse mesenchymal cells d Days post-infection e different time points after infection. e, Summary of 50 Damage- ECM- ECM synthesis 0 dpi responsive Wnt + βcat sig. SMCs (2) GSEA comparing all cells in each cluster (cluster 1, 50 1 dpi synthesizing Resting 10 (4) TGFβ sig. ESFibs (8) 3 dpi (8) n = 1,188; cluster 2, n = 1,123; cluster 4, n = 892; Inammatory 13 Myogenesis DRFibs (4) 25 6 dpi Notch sig. IRFibs (6) 25 cluster 6, n = 749; cluster 8, n = 681). βcat, β-catenin; 0 8 ROS Col13a1 (1) 4 7 Hypoxia resp., response; ROS, reactive oxygen species; sig., 5 0 Complement 0 11 3 signalling; SMCs, smooth muscle cells. -SNE2 6 -SNE2 Coagulation t 16 t 14 Angiogenesis f, Expression of key genes in mouse lung 2 15 17 1 TNFα via NF-κβ –25 12 18 –25 IFNα resp. mesenchymal cells. g, Kinetics of fibroblast Interferon- Inammatory resp. 19 9 Col13a1 responsive IL-6 JAK STAT3 activation states. Representative flow cytometry (6) –50 –50 –5 0510 –40 –20 0 20 –40 –20 020 for ITGA5, CD9 and BST2 staining. h, Frequency of Normalized enrichment score t-SNE1 t-SNE1 DRFib and IRFib activation states during infection – – – fgGated on CD45 EpCAM CD31 h Damage-responsive (0 dpi, n = 5; 3 dpi, n = 4; 6 dpi, n = 3; 9 dpi, n = 4; 12 dpi, Bst2 Eln DRFibs hi lo ) – 50 Il33 Lum 0 dpi 12 dpi (ITGA5 CD9 ) n = 4; 21 dpi, n = 5). Data are mean ± s.e.m. Irf7 Dcn lo 40 105 105 Ifnar2 Timp3 CD31 – M 30 Tnfrsf1a Vcan 4 104 CD9 10 hi Il1r1 5 Timp1 20 Cxcl1 Cd9 103 103 Resting lo Il6 Itga5 (ITGA5 ITGA 10 Tnfaip6 Mmp23 0 0 CD9hi) Nfkb1 Lox (% of EpCA 0 036912 21 Tgfbr3 0 103 104 105 0 103 104 105 Adamts4 Time after infection (d) Tgfbi Adamts5 ITGA5–APC Interferon-responsive ) ) ) CD9–PE ) 50 – 105 105 hi 40 ) (2 SMC ) (2 SMC IRFib (6 IRFib IRFib (6) IRFib IRFibs CD31 ) (8 ESFib (8 ESFib DRFib (4 DRFib ) DRFib (4 DRFib 4 4 BST2 10 10 hi hi – ) (5 Resting ) (5 Resting (ITGA5 CD9 hi M 30 hi 9 Per cent Average Per cent Average 3 3 BST2 ) 10 10 CD 20 expressed expression expressed expression hi 25 2 25 2 0 0 10 1 50 1 50 ITGA5 0 75 0 3 4 5 3 4 5 (% of EpCA 75 BST2–FITC 0 10 10 10 0 10 10 10 0 –1 100 –1 036912 21 EpCAM–BV421 Time after infection (d) including NF-κB signalling and hypoxia, whereas IRFibs were enriched Stromal responses to influenza infection for type-I-interferon-responsive pathways (Fig. 1e, f, Extended Data We sorted live, CD45− lung cells from mice 0, 1, 3 and 6 days after infec- Fig. 2b). Analysis of five human lung biopsies identified fibroblast acti- tion with influenza A virus and performed single-cell gene-expression vation states analogous to those defined in mouse lungs (Extended Data profiling (scGEX) identifying three main populations: fibroblasts, epi- Fig. 2c) with clusters enriched for ECM synthesis, NF-κB signalling and thelial cells and endothelial cells (Fig. 1a, b, Extended Data Fig. 1a, b). type-I interferon signalling (Extended Data Fig. 2c, d). High levels of viral mRNA, indicative of productive infection, were To validate these fibroblast activation states, we identified genes detected primarily in type I pneumocytes, ciliated epithelial cells and encoding surface-protein markers that were enriched in DRFibs and type II pneumocytes (Extended Data Fig. 1c). Fibroblasts were particu- IRFibs in mouse and human lungs (Fig. 1f, Extended Data Figs. 2d, larly dynamic, with multiple transcriptional states emerging following 3a, b). Upregulation of Itga5 and downregulation of Cd9 defined the infection (Fig. 1c, d, Extended Data Fig.