Abstract
A multi-omics approach revealed the underlying biological pathways in the microbiome-driven severe asthma phenotypes. This may help to elucidate new leads for treatment development, particularly for the therapeutically challenging neutrophilic asthma. https://bit.ly/3FWuJqS
To the Editor:
Asthma is a heterogeneous disease with multiple clinical presentations (phenotypes) [1]. Neutrophilic asthma is characterised by increased sputum neutrophils and generally has a poor response to corticosteroids and limited other therapeutic options. Neutrophilia originates from different factors, including the defective resolution of inflammation or bacterial infections [2]. An association between airway bacterial imbalance (disturbance) and the neutrophilic phenotype has been reported [3], suggesting that airway microbiota composition is involved in neutrophilic asthma. Rather than being a separate entity [4], neutrophilic asthma may be in part, an alliance between innate immunity and microbiota composition that prompts protective mechanisms against invading pathogens [2].
In the Unbiased BIOmarkers in PREDiction of respiratory disease outcomes (U-BIOPRED) study, adults with severe asthma were classified into two clusters (phenotypes) based on the sputum microbiome [3]. The microbiome-driven cluster 2 (C2) showed relatively worse asthma outcomes, sputum neutrophilia and microbial imbalance (reduced microbial diversity and commensals and enrichment of pathogenic bacteria, particularly Haemophilus influenzae and Moraxella catarrhalis) compared with cluster 1 (C1). We investigated the underlying molecular mechanisms of microbiome-driven phenotypes by investigating differences in sputum eicosanoids, transcriptomics, and proteomics to reveal potential targets for diagnosis and treatment.
Sputum omics, including eicosanoids, transcriptomics, and SomaScan® proteomics, were assayed as described previously [3, 5, 6]. C2 was compared with C1 patients, mild–moderate asthma patients (MM), and healthy controls (H) according to a panel of 11-eicosanoids and Shannon metagenomics α-diversity using Kruskal–Wallis H test. C2 and C1 were compared by sputum transcriptomics and proteomics using linear models for microarray data (limma) analysis after adjusting for age, sex and oral corticosteroid intake. Pathway enrichment analysis for the sputum differential transcriptome was performed by gene set enrichment analysis (Gage R package) using two databases for Homo sapiens; the Kyoto Encyclopedia of Genes and Genomes (KEGG) database and the Molecular Signatures Database (MSigDB) Hallmark gene sets. A similar analysis was also performed comparing C2 against MM and H groups. Multiple testing was adjusted for using Benjamini Hochberg false discovery rate correction and an α (q-value) <0.05 was considered significant.
The sputum 11-dehydrothromboxane B2 and prostaglandin (PG) E2 (but not leukotriene (LT) B4) levels were significantly elevated in C2 compared to C1 subjects (q<0.05) (figure 1a), consistent with sputum PTGS2 gene upregulation in C2 (log2 fold change (FC) 1.1; q<0.05). Furthermore, the levels of 15-hydroxyeicosatetraenoic acid (HETE), LTE4 and PGD2 were elevated in C2 compared with MM and/or H groups. These findings were associated with decreased Shannon alpha-diversity in C2 compared to other groups (p<0.001) (figure 1b), consistent with the increase in the inflammatory signal with the increased severity level in the groups.
A total of 2578 differentially expressed genes (DEGs; q<0.05) were found to be different between C2 and C1, of which 194 genes had a least a two-FC (figure 1c). Several upregulated DEGs were related to immune regulation and inflammation, particularly tumour necrosis factor (TNF)-α and related regulatory genes (TNFAIP-6, TNFAIP3, TNFSF10, TNFRSF10 and TNFSF14); interleukins (ILs) and related regulatory genes (IL18R1, IL18RAP, IL1R1, IL1R2, IL1B, IRAK2, IRAK3, IL6 and IL6R); Toll-like receptors (TLRs; e.g. TLR2, TLR4 and TLR10); and inflammasomes (e.g. NLRP3, NLRP12 and NLRC5) in C2 relative to C1. Similar DEGs were significantly upregulated in C2 compared to MM and H groups.
Upregulated KEGG pathways in C2 mainly related to pathogen fighting, inflammation and innate immune-regulation mechanisms such as natural killer cell-mediated cytotoxicity, chemokine, NOD-, Toll-, and RIG-I-like receptor signalling pathways (q<0.05) (figure 1e). Downregulated KEGG pathways in C2 were related to cell proliferation and growth, including ribosome biogenesis, DNA replication and repair; fatty acid, amino acid, nucleotide and carbohydrate metabolism; and cellular degradation and catabolism (lysosome, peroxisome and proteasome). The Hallmark gene sets supporting KEGG analysis including up-regulation of immuno- and inflammatory pathways in C2, such as TNF-α signalling via nuclear factor (NF)κB, interferon (IFN)-α, IFN-γ, inflammatory response, IL-6 JAK-STAT3 (Janus kinase-signal transducer and activator of transcription) signalling, and apoptosis (figure 1f). Downregulated pathways included oxidative phosphorylation (OXPHOS), reactive oxygen species (ROS), DNA repair, cell cycle, proliferation and growth, peroxisome, and metabolism. Similar upregulated and downregulated pathways were significant in C2 compared to MM and H groups. Additional KEGG/Hallmark upregulated pathways in C2 relative to MM and H groups were hypoxia, JAK-STAT, MAPK (mitogen-activated protein kinase), T-cell, KRAS and IL2 STAT5 signalling (data not shown).
36 differentially abundant proteins (DAPs) were significant (q<0.05) between C2 and C1, of whom 14 proteins had at least a two-FC (figure 1d). Upregulated DAPs in C2 (TSG-6, IL-17RA, IP-10, granulysin, ILT-4, PBEF, TLR4:MD-2 complex, MMP-10, NKp46, and IGFIR) were related to neutrophilia, inflammation, and/or Th17 and Th1 mediated pathways. Similar DAPs were significant when C2 patients were compared to MM and H groups. Additional enriched (pro)inflammatory DAPs were found in C2 relative to MM and/or H groups (q<0.05), including CRP, light, protein-S, azurocidin, PGRP-S, TIMP-2, IL-8, Notch1, MMP1, MMP2 and C5a.
The bacterial imbalance in C2 was associated with sputum neutrophilia, suggesting an immune-modulatory response to fight clinical/subclinical infections, potentially resulting in more inflamed/obstructed airways [7] with worse lung function parameters, as observed in C2 patients. Interestingly, sputum macrophages were significantly decreased in C2 compared to C1, consistent with a defective immune capacity to adequately eliminate pathogenic micro-organisms [8]. Most of the elevated signalling eicosanoids molecules in C2 are proinflammatory and/or induce bronchoconstriction [9], which may partly explain the reduced lung function in the relatively more severe C2 cluster. In addition, sputum PGE2 and PGD2 regulate neutrophils and/or Th17 and Th1 pathways [10], suggesting that their elevated levels in C2 patients might be driving these clinical features. PGE2 suppresses phagocytosis by airway macrophages in asthma [11], consistent with our sputum pathway enrichment analysis showing that C2 patients had down-regulated lysosome and ROS pathways, which are critical in pathogen fighting [12], relative to C1.
Many of the observed DEGs, DAPs, and pathways are implicated in immune regulation and/or are characteristic of the neutrophilic asthma signature [7] distinguishing C2 patients. TNF-α was markedly upregulated in C2, consistent with what was found in the Haemophilus-enriched cluster in patients with asthma and COPD [13]. Upregulation of the interferon pathways was observed in patients with C2, and IFN-γ has been shown to decrease epithelial barrier function and enhance neutrophil transmigration [14]. Impaired barrier function could render the airway mucosa of C2 subjects vulnerable to pathogenic bacterial infections. Our metagenomics analysis did not detect key RNA viruses, such as RV16, whose presence at a subclinical level could enhance RIG-I and IFN-γ pathways.
In contrast, downregulated pathways showed defective cell growth, proliferation, metabolism and DNA repair, suggesting an impaired capacity to repair inflamed cells and maintain homeostasis. Moreover, C2 exhibited downregulation of OXPHOS, tricarboxylic acid cycle, and ROS pathways suggesting mitochondrial dysfunction [15]. These results support previous data where neutrophilic inflammation was associated with reduced OXPHOS gene expression and mitochondrial dysfunction in bronchial epithelial cells [16].
These findings reveal potential targets for the C2 microbiome-driven asthma phenotype, which might constitute a selected subtype within corticosteroid-resistant asthmatics. Therapeutic options for this cluster might be antibiotics (e.g. macrolides) or phage therapy. In addition, the dysregulated neutrophilic and cytokines/chemokines might be diagnostic/therapeutic targets in C2 patients [7]. Differing biological pathways and mediators involved in these patients suggest that pharmacotherapies targeting multiple mechanisms might be required. This also suggests neutrophilic asthma is not a single phenotype [4] and underlying mechanisms should be considered to optimise treatment.
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-02603-2021.Shareable
Footnotes
Author contributions: M.I. Abdel-Aziz performed the analysis and drafted the initial version of the manuscript. M.I. Abdel-Aziz, A.H. Maitland-van der Zee, I.M. Adcock, K.F. Chung and P.J. Sterk have contributed to the design of the analysis plan. All co-authors have contributed to the acquisition of data, interpretation of the analysis, revision, drafting, critical appraisal and ensuring accuracy and integrity of the analysis. All co-authors have provided final approval of the version to be published. A.H. Maitland-van der Zee is the corresponding author. In addition, the U-BIOPRED is a consortium effort and we wish to acknowledge the help and expertise of the individuals and groups in the U-BIOPRED study group.
Conflicts of interest: M.I. Abdel-Aziz, S.J.H. Vijverberg, A.H. Neerincx, P. Brinkman, A.H. Wagener, A.R. Sousa, Å.M. Wheelock, AT. Bansal, M. Caruso, J. Corfield, N. Krug, J. Musial, K. Sun, D.E. Shaw, P. Montuschi, R. Lutter, P. Howarth, M. Sanak, I.M. Adcock, P. Chanez and AD. Kraneveld have no conflicts of interest to disclose. J.H. Riley was an employee and a shareholder of GlaxoSmithKline. S. Bates is an employee of Johnson & Johnson and a former employee of GlaxoSmithKline, and holds stock in both companies. S.S. Wagers reports consulting fees from Kings College Hospital NHS Foundation Trust, Academic Medical Research, Aladdin Healthcare Technologies Ltd, AMC Medical Research BV, AMC Medical Research BV P402, Asthma UK, AstraZeneca AB, Athens Medical School, Autobedrijf Verfaillie V-CO BVBA, Boehringer Ingelheim International GmbH, Breathomix BV, Chiesi Farmaceutici S.p.A., CHU de Toulouse, CIRO, Consorcio Centro de Investigación Biomédica en Red, Dawi Iman Automobile, DS Biologicls Ltd, École Polytechnique Fédérale de Lausanne, European Federation of Allergy, European Respiratory Society, European Society for Swallowing Disorders, F. Hoffmann-La Roche AG, FISEVI, Fluidic Analytics Ltd, Foundation for Research and Technology Hellas, Fraunhofer IGB, Fraunhofer ITEM, Fraunhofer-Institut für Produktionstechnologie IPT, GlaxoSmithKline Research & Dev Ltd, Gossamer Bio, Inc, Holland & Knight, Imperial College, Institute for Computer Science and Control, Karolinska Institutet Fakturor, KU Leuven, Longfonds, Maastricht University, MedImmune LLC, Merck Sharp & Dohme Corp., National Heart & Lung Institute, Ninja Hoen, Novartis Pharma AG, OncoRadiomics S.A., Owlstone Medical Limited, PExA AB, Pulmonary Fibrosis Foundation, RISE Research Institutes of Sweden AB, Sanofi US Services Inc., TEVA Branded Pharmaceutical Products R&D Inc., The North West Lung Centre Charity, Three Lakes Foundation Trust, TopMD Precision Medicine Ltd, UCB Biopharma S.P.R.L., UCB Biosciences GmbH, UGent, Umeå University, Univ. Hospital Southampton NHS Foundation Trust, Università Campus Bio-Medico di Roma, Universita Cattolica Del Sacro Cuore, Universität Ulm, Universitätsklinikum Würzburg, University College London, University Hospital Southampton, University of Bern, University of Edinburgh, University of Hull, University of Leicester, University of Loughborough, University of Luxembourg, University of Manchester, University of Nottingham, VITO NV, Vlaams Brabant, Dienst Europa, Kings College Hospital NHS Foundation Trust, and Academic Medical Research paid to his company (BioSci Consulting); outside the submitted work. B. De Meulder and C. Auffray have received grants from Innovative Medicines Initiative (IMI U-BIOPRED grant number 115010 and IMI eTRIKS grant number 115446), during the conduct of the study. M. Uddin is an employee of AstraZeneca and holds shares in the company. I. Horvath has received personal honoraria for lectures from AstraZeneca, Chiesi, GSK, Novartis and MSD, and had unpaid leadership or fiduciary roles in European Respiratory Society, Hungarian Respiratory Society, and Hungarian Association of Medical Societies. T. Sandström has received fees from lecturing for Boehringer Ingelheim with payments to clinic/institution. S.J. Fowler has received honoraria for lectures from AstraZeneca, Novartis, Teva, Boehringer Ingelheim and Chiesi, and has received support to attend ERS and BTS meetings from AstraZeneca and Chiesi, and participated in the advisory board of Chiesi. R. Djukanovic is a co-founder of and has stocks with Synairgen, a University of Southampton spin out company, which is funding a phase 3 trial of nebulised interferon beta for treatment of hospitalised COVID-19 patients and a community based study of PCR positive individuals to see if treatment can improve clinical outcomes, and is a member (without receiving funding) of the Urgent Public Health (UPH) panel that approves studies under a priority badge, and has received fees for consultancy services from TEVA Pharmaceuticals, Sanofi, Boehringer and Novartis, and has received grants (paid to his institution) from IMI funded EU project, UBIOPRED, MRC funded project on COPD: COPD-MAP and Novartis to Study of mechanisms of action of Omalizumab, and has received fees for lectures on mechanisms of action of Xolair for Novartis and a lecture in symposium on mechanisms of asthma for Teva. P. Skipp is a shareholder in TopMD Precision Medicine. K.F. Chung reports funding during the conduct of the study paid to his institution from the Innovative Medicines Initiative (IMI) Joint Undertaking under grant agreement number 115010. P.J. Sterk reports funding during the conduct of the study paid to his institution from Innovative Medicines Initiative (IMI), which is public-private funding provided by the EU and the European Federation of Pharmaceutical Industries and Associations (EFPIA), and is a scientific adviser and has an officially non-substantial interest in the start-up company Breathomix BV, The Netherlands; outside the submitted work. A.H. Maitland-van der Zee has received grants from Health Holland and she is the PI of a P4O2 (Precision Medicine for more Oxygen) public private partnership sponsored by Health Holland involving many private partners that contribute in cash and/or in kind (Boehringer Ingelheim, Breathomix, Fluidda, Ortec Logiqcare, Philips, Quantib-U, Smartfish, SODAQ, Thirona, TopMD and Novartis), received unrestricted research grants from GSK, Boehringer Ingelheim and Vertex, received consulting fees paid to her institution from Boehringer Ingelheim and AstraZeneca, and received honoraria for lectures paid to her institution from GlaxoSmithKline; outside the submitted work.
Support statement: U-BIOPRED has received funding from the Innovative Medicines Initiative (IMI) Joint Undertaking under grant agreement no. 115010, resources of which are composed of financial contributions from the European Union's Seventh Framework Programme (FP7/2007–2013) and European Federation of Pharmaceutical Industries and Associations (EFPIA) companies’ in-kind contributions (www.imi.europa.eu). The researcher M.I. Abdel-Aziz is funded by a full scholarship from the Ministry of Higher Education of the Arab Republic of Egypt. Funding information for this article has been deposited with the Crossref Funder Registry.
- Received July 27, 2021.
- Accepted October 11, 2021.
- Copyright ©The authors 2022. For reproduction rights and permissions contact permissions{at}ersnet.org