CSF3R/CD114 mediates infection-dependent transition to severe asthma

The major 17q12-21 asthma susceptibility and exacerbation locus1Bisgaard H. Bonnelykke K. Sleiman P.M. Brasholt M. Chawes B. Kreiner-Moller E. et al.Chromosome 17q21 gene variants are associated with asthma and exacerbations but not atopy in early childhood.Am J Respir Crit Care Med. 2009; 179: 179-185Crossref PubMed Scopus (174) Google Scholar has been identified as the only genetic locus that is also reproducibly associated with total white blood cell count.2Soranzo N. Spector T.D. Mangino M. Kuhnel B. Rendon A. Teumer A. et al.A genome-wide meta-analysis identifies 22 loci associated with eight hematological parameters in the HaemGen consortium.Nat Genet. 2009; 41: 1182-1190Crossref PubMed Scopus (429) Google Scholar However, it is not known whether there is a common gene within this locus that links these phenotypic traits. The colony-stimulating factor-3 (CSF3) gene, alternatively known as G-CSF, resides within this locus. CSF3 binds exclusively to CSF3 receptor (CSF3R, or CD114/G-CSFR), which is highly expressed on mature neutrophils and to a lesser extent on mononuclear cells, platelets, and lung interstitial stromal cells. Although CSF3R/CD114 signaling can dictate the intensity of the host defense inflammatory response during bacterial infection by regulating neutrophil granulopoiesis and trafficking, its role in the infection-dependent transition to persistent, severe asthma has not been investigated. Neonatal colonization of the nasopharynx by potentially pathogenic bacteria including Streptococcus pneumoniae is also a risk factor for asthma development.3Bisgaard H. Hermansen M.N. Buchvald F. Loland L. Halkjaer L.B. Bonnelykke K. et al.Childhood asthma after bacterial colonization of the airway in neonates.N Engl J Med. 2007; 357: 1487-1495Crossref PubMed Scopus (757) Google Scholar The Childhood Asthma Study found that children with atopy and chronic wheeze at age 5 years were twice as likely to have been colonized with S pneumoniae as neonates.4Teo S.M. Mok D. Pham K. Kusel M. Serralha M. Troy N. et al.The infant nasopharyngeal microbiome impacts severity of lower respiratory infection and risk of asthma development.Cell Host Microbe. 2015; 17: 704-715Abstract Full Text Full Text PDF PubMed Scopus (552) Google Scholar The authors suggest that transient incursions of nasopharyngeal bacteria into the lower airways triggered by a fever-causing viral respiratory infection (respiratory syncytial virus or influenza virus) increases the risk of developing persistent asthma in atopic children. However, a plausible mechanism linking these cofactors is yet to be identified. In this study, we tested the hypothesis that CSF3-CSF3R signaling dictates the severity of infection-dependent asthma at a cellular and molecular level. To test this hypothesis, we adapted our previously published mouse model of neonatal nasopharyngeal pneumococcal colonization and acute viral infection,5FitzPatrick M. Royce S.G. Langenbach S. McQualter J. Reading P.C. Wijburg O. et al.Neonatal pneumococcal colonisation caused by Influenza A infection alters lung function in adult mice.Sci Rep. 2016; 6: 22751Crossref PubMed Scopus (3) Google Scholar and superimposed subsequent house-dust mite (HDM) aeroallergen exposure into adulthood. S pneumoniae (EF3030) colonization of the nasopharynx was not altered by influenza A virus or HDM (see Fig E1, A, in this article's Online Repository at www.jacionline.org). In contrast, pneumococcal loads in the lower airways were significantly increased in coinfected neonates chronically challenged with HDM (Fig E1, B). Histological assessment of lung pathology demonstrated that HDM increased the area positive for mucus staining in small airways (Fig E1, C and D). HDM-induced mucin levels were not further increased by inoculation with a single respiratory pathogen, whereas coinfection significantly increased the mucus area surrounding the small airways (Fig E1, D). Consistent with this finding, Muc5AC transcript levels were only significantly increased in coinfected mice challenged with HDM (Fig E1, E). HDM challenge also elicited eosinophilic inflammation in the bronchoalveolar (BAL) compartment, and this response was not significantly altered by coinfection (Fig E1, F). In contrast, HDM-induced neutrophilic inflammation was only significantly increased in mice coinfected as neonates (Fig E1, G). HDM challenge did not significantly alter CSF3 transcript levels, whereas the combination of neonatal coinfection and HDM significantly increased CSF3 gene expression in the lung (Fig E1, H). To assess the translational relevance of our findings, we examined CSF3 transcript expression in bronchial biopsies obtained from adults with asthma grouped into inflammatory phenotypes: neutrophilic asthma (NA) and nonneutrophilic asthma (non-NA; consisting of paucigranulocytic and eosinophilic asthma, as summarized in Table E1 in this article's Online Repository at www.jacionline.org). NA had poor control of their asthma symptoms based on Asthma Control Questionnaire scores and significantly lower lung function (median FEV1%, NA 72% vs non-NA 85%; P < .04). CSF3 and CSF3R transcript levels were significantly higher in NA compared with non-NA (Fig 1, A and B), and CSF3 expression was positively associated with BAL neutrophil numbers across the entire asthma cohort (Fig 1, C). Atopy status did not affect CSF3 transcript levels in asthma, but rather presence of positive bacterial BAL cultures resulted in higher CSF3 expression (Fig 1, D and E). In addition, CSF3 and Muc5AC transcript expression was positively associated in asthmatic bronchial biopsies (Fig 1, F). We also analyzed a transcriptomic sputum data set from the U-BIOPRED cohort,6Rossios C. Pavlidis S. Hoda U. Kuo C.H. Wiegman C. Russell K. et al.Sputum transcriptomics reveal upregulation of IL-1 receptor family members in patients with severe asthma.J Allergy Clin Immunol. 2018; 141: 560-570Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar where CSF3R expression levels were significantly increased in those with severe asthma compared with healthy volunteers (1.9-fold increase, P < .005; see Fig E3 in this article's Online Repository at www.jacionline.org). Although CSF3 was not significantly increased (1.2-fold increase, P = .09), 33 of 84 (39%) expressed high CSF3 levels (defined as above the 95th percentile of healthy control distribution), confirming that there is a subset of those with asthma who are CSF3high. We next therapeutically targeted CSF3R signaling by administering an anti-CSF3R mAb (α-CSF3R) in our mouse model (see Fig E2, A, in this article's Online Repository at www.jacionline.org). α-CSF3R administration significantly reduced BAL neutrophil numbers (Fig 2, A) and markers for NETosis including neutrophil elastase activity and double-stranded DNA in coinfected mice challenged with HDM allergen (Fig 2, B and C). Lung neutrophil numbers determined by flow cytometry (Ly6G+, Siglec F−) were significantly reduced by α-CSF3R treatment (Fig 2, D and E). Total peroxidase activity in the lung tissue was also markedly increased in coinfected mice challenged with HDM, and this response was significantly reduced by 50% with α-CSF3R treatment (Fig 2, F). Because inhibition of neutrophilic inflammation may compromise bacterial clearance, we also evaluated pneumococcal load in the nasopharynx, BAL, and lung tissue, which was not significantly altered by α-CSF3R treatment (see Fig E2, B-D). We also assessed whether α-CSF3R treatment reduced mucus levels surrounding the small airways (Fig 2, G, representative images). Elevated mucus production observed in coinfected mice challenged with HDM was significantly reduced with α-CSF3R treatment by approximately 50% (Fig 2, H). Consistent with excessive mucus obstructing the airways, methacholine-induced airway hyperreactivity (Rn) was completely reversed with α-CSF3R therapy in coinfected mice challenged with HDM (Fig 2, I). In this study, we demonstrate that the combination of HDM allergen and an acute viral infection cooperate to induce the dispersion of S pneumoniae into the lower airways, which markedly increased CSF3-dependent neutrophilic inflammation. CSF3 cooperates with a number of factors including IL-6, IL-3, granulocyte-macrophage colony-stimulating factor (GM-CSF), and reactive oxygen species (ROS) to stimulate neutrophil granulopoiesis in the bone marrow to restore neutrophil homeostasis during infection. CSF3 also promotes neutrophil trafficking by modulating chemokine and adhesion receptors (CXCR2 and CD62L) on neutrophils, and in a murine arthritis model, blocking CSF3R prevented trafficking into the arthritic joint without inducing peripheral neutropenia.7Campbell I.K. Leong D. Edwards K.M. Rayzman V. Ng M. Goldberg G.L. et al.Therapeutic targeting of the G-CSF receptor reduces neutrophil trafficking and joint inflammation in antibody-mediated inflammatory arthritis.J Immunol. 2016; 197: 4392-4402Crossref PubMed Scopus (58) Google Scholar Consistent with this, blocking CSF3R did not completely deplete neutrophils, but rather selectively reduced excessive neutrophil trafficking into the lungs of coinfected/HDM-exposed mice without compromising bacterial clearance. By blocking CSF3R signaling, we also show that CSF3 promotes infection-dependent production of mucus in allergic airways. Mucin overproduction is prominent in fatal asthma because it directly contributes to severe airflow obstruction. Obstruction of the airways is exacerbated by the actions of eosinophil peroxidase, which catalyzes the generation of oxidants that cross-link mucus to form hydrogels.8Dunican E.M. Elicker B.M. Gierada D.S. Nagle S.K. Schiebler M.L. Newell J.D. et al.Mucus plugs in patients with asthma linked to eosinophilia and airflow obstruction.J Clin Invest. 2018; 128: 997-1009Crossref PubMed Scopus (242) Google Scholar Because the same pathological process is also mediated by neutrophil-derived myeloperoxidase,9Yuan S. Hollinger M. Lachowicz-Scroggins M.E. Kerr S.C. Dunican E.M. Daniel B.M. et al.Oxidation increases mucin polymer cross-links to stiffen airway mucus gels.Science Transl Med. 2015; 7: 276ra27Crossref PubMed Scopus (157) Google Scholar we identify CSF3R as a novel target to suppress pathological peroxidase activity in severe asthma. In summary, we identify CSF3 as a major effector molecule that drives infection-dependent transition to more persistent and severe asthma. Antagonising CSF3 production through targeted inhibition of CSF3R signaling represents a novel strategy to prevent this pathological inflammation and remodeling. We thank Dr Jonathan McQualter and Mr Ivan Bernardo for their technical assistance during the major experimental protocols. All human procedures were approved by The University of Newcastle Human Ethics Committee. All animal procedures were approved by the Animal Ethics Committees of the University of Melbourne and RMIT University and complied with the National Health and Medical Research Council Australian Code of Practice for the care and use of animals for scientific purposes (8th edition, 2013). Twenty-six adult participants were recruited on the basis of doctor diagnosis of asthma and demonstrated variable airflow obstruction, with a 15% or greater than 200 mL change in FEV1 following bronchodilator or airway hyperresponsiveness to mannitol challenge test. All subjects were nonsmokers with less than 5 packet-year smoking history. Participants were prescribed inhaled corticosteroids or combination inhaled corticosteroid/long-acting bronchodilator therapy and were assessed during a stable phase of disease with no change in asthma medications over the previous 4 weeks. Patient characteristics including age, sex, atopy, FEV1%, Asthma Control Questionnaire scores, and presence of bacterial cultures in BAL are presented in Table E1. The patients were grouped on the basis of inflammatory phenotypes defined using BAL differential cell count cutoff values as described below. All subjects underwent a fiberoptic bronchoscopy in accordance with standard guidelines.E1Sukkar M.B. Wood L.G. Tooze M. Simpson J.L. McDonald V.M. Gibson P.G. et al.Soluble RAGE is deficient in neutrophilic asthma and COPD.Eur Respir J. 2012; 39: 721-729Crossref PubMed Scopus (106) Google Scholar To minimize variability, the bronchoscope was inserted into the third- or fourth-generation airway of the subject where sterile saline was instilled and collected. Cytospins were prepared from the recovered BAL, and NA (n = 12) was defined as greater than or equal to 65% neutrophils of total cell count, eosinophilic asthma (n = 8) as greater than or equal to 3.5% BAL eosinophils of total cell count, and paucigranulocytic asthma (n = 6) as less than 65% neutrophils and less than 3.5% eosinophils of total cell count. The presence of bacteria in the BAL was determined by qualitative and quantitative microbiology.E2Wilkinson T.M. Donaldson G.C. Johnston S.L. Openshaw P.J. Wedzicha J.A. Respiratory syncytial virus, airway inflammation, and FEV1 decline in patients with chronic obstructive pulmonary disease.Am J Respir Crit Care Med. 2006; 173: 871-876Crossref PubMed Scopus (168) Google Scholar Bronchial biopsies were then obtained from the same areas using biopsy forceps applied under direct vision and were used for RT-quantitative PCR analysis.Table E1Patient characteristicsCharacteristicNA (n = 12)Non-NA (n = 14)Paucigranuloctyic (n = 6)Eosinophilic (n = 8)Age (y), median (range)65.5 (42-78)55.5 (38-57)61 (42-76)Sex (M:F)4:84:23:5Atopy, n (%)4 of 12 (33)3 of 6 (50)7 of 8 (88)Median FEV1% (range)72 (59-97)84 (71-104)86 (58-108)Median ACQ (range)2 (1-3.3)0.65 (0-1.3)2.15 (1.1-3.5)+ Bacterial culture, n (%)8 of 12 (67)1 of 6 (17)5 of 8 (63)%Neutrophil (range)88 (67-96)38 (1.5-58)26 (18-64)%Eosinophils (range)1.1 (0.3-2.8)1.4 (0.5-3.3)10.5 (3.8-47.3)ACQ, Asthma Control Questionnaire. Open table in a new tab ACQ, Asthma Control Questionnaire. We analyzed a subset of the U-BIOPRED adult baseline transcriptomics data. The U-BIOPRED cohort comprises nonsmoker severe asthmatic (n = 84) and nonasthmatic healthy control subjects (n = 16). Study participants had undergone detailed phenotypic characterization using established standard operating procedures, as described previously,E3Shaw D.E. Sousa A.R. Fowler S.J. Fleming L.J. Roberts G. Corfield J. et al.Clinical and inflammatory characteristics of the European U-BIOPRED adult severe asthma cohort.Eur Respir J. 2015; 46: 1308-1321Crossref PubMed Scopus (355) Google Scholar and details of sputum transcriptomics methodology are detailed in Kuo et al.E4Kuo C.S. Pavlidis S. Loza M. Baribaud F. Rowe A. Pandis I. et al.T-helper cell type 2 (Th2) and non-Th2 molecular phenotypes of asthma using sputum transcriptomics in U-BIOPRED.Eur Respir J. 2017; 49 (pii: 1602135. https://doi.org/10.1183/13993003.02135-2016)Crossref Scopus (232) Google Scholar Advanced pregnant BALB/c dams (Animal Resources Centre, Western Australia, Australia) were housed separately and monitored for birth with minimal disruption. Upon birth, dams were housed with their litters until weaning at age 3 weeks. Mice were housed at 22°C under normal 12-hour: 12-hour light: dark cycle, and given free access to a normal diet and water. To control for potential variability caused by sex and litter size, only female mice born to litter sizes of 5 to 7 pups were used for the experimental protocol. Neonatal mice were infected with S pneumoniae and/or influenza A virus as previously describedE5FitzPatrick M. Royce S.G. Langenbach S. McQualter J. Reading P.C. Wijburg O. et al.Neonatal pneumococcal colonisation caused by influenza A infection alters lung function in adult mice.Sci Rep. 2016; 6: 22751Crossref PubMed Scopus (4) Google Scholar with the following minor modifications. At age 8 days, BALB/c infant mice were inoculated intranasally without anesthesia with S pneumoniae (serotype 19F strain EF3030, 2 × 103 CFU) in a volume of 3 μL sterile saline or saline alone. At age 15 days, mice were inoculated by the intranasal route with influenza A virus (strain HKx31, H3N2, 500 PFU) in a volume of 3 μL sterile saline or vehicle. HDM extract (Dermatophagoides pteronyssinus) was obtained from Greer Laboratories (Charlotte, NC). At age 20 to 21 days, weaned female mice were sensitized intranasally with HDM aeroallergen (10 μg in 10 μL saline) or vehicle (10 μL saline alone, VEH) under isoflurane anesthesia for 5 consecutive days per week over 3 weeks. All outcomes were assessed 24 hours after the final HDM treatment (or VEH) when mice were approximately aged 6 weeks. In a separate experiment, 100 μg of anti–colony-stimulating factor-3 receptor (α-CSF3R) antibody or isotype control antibody (CSL Limited, Parkville, Victoria, Australia) was administered to coinfected/HDM-exposed mice via intraperitoneal injection every second day during the last week of HDM sensitization. Uninfected mice that received VEH only during the HDM exposure were used as control (saline alone). All outcomes were assessed 24 hours after the final HDM/antibody treatment when mice were approximately aged 6 weeks. In vivo airway reactivity was measured using Flexivent (SCIREQ, Montreal, Quebec, Canada) as previously published.E5FitzPatrick M. Royce S.G. Langenbach S. McQualter J. Reading P.C. Wijburg O. et al.Neonatal pneumococcal colonisation caused by influenza A infection alters lung function in adult mice.Sci Rep. 2016; 6: 22751Crossref PubMed Scopus (4) Google Scholar Briefly, mice were anesthetized with ketamine (125 mg/kg) and xylazine (25 mg/kg) before tracheotomy was performed and a cannula was inserted. In vivo airway responsiveness was assessed in response to nebulized PBS and methacholine (MCh, 100 mg/mL). BAL was performed and total and differential BAL cell counts were determined as previously published.E6Anthony D. Seow H.J. Uddin M. Thompson M. Dousha L. Vlahos R. et al.Serum amyloid A promotes lung neutrophilia by increasing IL-17A levels in the mucosa and gammadelta T cells.Am J Respir Crit Care Med. 2013; 188: 179-186Crossref PubMed Scopus (54) Google Scholar Nasopharyngeal tissue was collected and homogenized. Serial dilutions of BAL fluid and nasopharyngeal tissue homogenates were cultured on horse blood agar plates with gentamycin (5 μg/mL) as previously described.E5FitzPatrick M. Royce S.G. Langenbach S. McQualter J. Reading P.C. Wijburg O. et al.Neonatal pneumococcal colonisation caused by influenza A infection alters lung function in adult mice.Sci Rep. 2016; 6: 22751Crossref PubMed Scopus (4) Google Scholar RT-quantitative PCR was used to quantify S pneumoniae in lung tissue using a commercial kit from Qiagen, Hilden, Germany. Briefly, bacterial DNA was isolated by homogenizing lung tissue in Trizol using TissueLyser (Qiagen), and a 2-step PCR cycle as per manufacturer's instructions (Qiagen) was used to quantify pneumococci from a standard curve. RNA was purified from snap-frozen lung tissue using RNeasy kit as per manufacturer's instructions and cDNA was prepared as previously described.E7Bozinovski S. Vlahos R. Zhang Y. Lah L.C. Seow H.J. Mansell A. et al.Carbonylation caused by cigarette smoke extract is associated with defective macrophage immunity.Am J Respir Cell Mol Biol. 2011; 45: 229-236Crossref PubMed Scopus (44) Google Scholar RT-quantitative PCR was performed using bioinformatically validated Taqman primer/probes. All threshold cycle values (Ct) were normalized to a reference gene (glyceraldehyde phosphate dehydrogenase) and the relative fold change determined by the ΔΔCt method as previously published.E7Bozinovski S. Vlahos R. Zhang Y. Lah L.C. Seow H.J. Mansell A. et al.Carbonylation caused by cigarette smoke extract is associated with defective macrophage immunity.Am J Respir Cell Mol Biol. 2011; 45: 229-236Crossref PubMed Scopus (44) Google Scholar Antibodies were purchased from BD Biosciences, San Jose, Calif (fluorescein isothiocyanate–conjugated CD45, phycoerythrin-conjugated Siglec F, and allophycocyanin/Cy7-conjugated Ly6G). Flow cytometric analysis was performed on single-cell suspension preparation from lung homogenates using cell-specific surface markers (CD45, Ly6G, and Siglec F, as previously publishedE8Wang H. Anthony D. Yatmaz S. Wijburg O. Satzke C. Levy B. et al.Aspirin-triggered resolvin D1 reduces pneumococcal lung infection and inflammation in a viral and bacterial coinfection pneumonia model.Clin Sci (Lond). 2017; 131: 2347-2362Crossref PubMed Scopus (41) Google Scholar). Neutrophils were gated as intermediate or high forward scatter-area and side scatter-area, further classified as CD45Hi, Siglec F−, and Ly6G+. All cells were permeabilized and fixed using a Cell Fixation & Cell Permeabilization Kit (Life Technologies, Carlsbad, Calif) before being analyzed on a BD FACSARIA III flow cytometer. Data were then analyzed with FlowJo v7.6.5 (Tree Star, Ashland, Ore). The left lobe of lung was removed post mortem and immediately fixed in 10% w/v neutral-buffered formalin. Tissues were paraffin-embedded and cut at a thickness of 4 μm. Sections were stained with Alcian blue-periodic acid Schiff for assessment of goblet cell transdifferentiation. Whole-lung sections were scanned using a slide scanner (Olympus VS-120, Tokyo, Japan), and morphometric analysis was performed using CellSens Dimensions software from Olympus. Morphometric evaluation of lung tissue sections was analyzed on a minimum of 4 bronchi per mouse section selected according to size (100-350 μm luminal diameter). Briefly, a 30 μm band was selected around the subepithelial layer of randomly selected bronchioles per section and the percent positive stain areas analyzed from Alcian blue-periodic acid Schiff–stained slides using standardized threshold values. Peroxidases from neutrophils and eosinophils were extracted by homogenizing ground lung tissue (20 mg) in 0.4 mL extraction buffer (50 mM potassium phosphate monobasic pH 6.0, 0.5% w/v hexa-decyl-trimethyl ammonium bromide and 10 mM EDTA). Following centrifugation, 10 μL lung lysate was incubated with 90 μL reaction buffer (50 mM potassium phosphate pH 6.0, 0.167 mg/mL o-Dianisidine [Fast Blue B, Sigma], and 0.005% H2O2). The change in absorbance (A460) (between 2-minute and 3-minute postincubation) resulted from the peroxidase-mediated decomposition of H2O2, and oxidation of o-Dianisidine was measured using a Clariostar plate reader. Neutrophil elastase activity was measured on BAL fluid with an EnzChek Elastase Assay Kit according to manufacturer's instructions (Life Technologies). Content of double-stranded DNA in the BAL fluid was also measured using Quant-iT PicoGreen double-stranded DNA reagent (Life Technologies), according to the manufacturer's instructions. Data from the experimental model was presented as the mean ± SEM with normal distribution or median ± interquartile range with nonnormal distribution. All data were statistically analyzed using GraphPad Prism 7.0 (GraphPad, San Diego, Calif). Where detailed and appropriate, 1-way or 2-way ANOVA with Tukey post hoc tests were performed. P less than .05 was considered to be statistically significant. Nonparametric data from the clinical cohort were reported as the median and interquartile range and by Kruskal-Wallis test followed by Dunns correction or the Mann-Whitney U test or Spearman correlation .Fig E2Neutralizing CSF3R did not affect pneumococcal nasal or lung load. A, In the last week of HDM challenge, mice were treated with 100 μg anti-CSF3R antibody (α-CSF3R) or isotype control antibody (ISO) by intraperitoneal injection every second day. Twenty-four hours after final HDM/antibody administration, S pneumoniae load in the (B) nasopharynx, (C) BAL fluid, and (D) lung tissue was determined (n = 9-10 per group). IAV, Influenza A virus; CFU, colony forming unit; PFU, plaque forming unit; SP, Streptococcus pneumoniae. Data are expressed as median ± interquartile range.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3CSF3 and CSF3R expression in the U-BIOPRED cohort. Baseline transcript expression of (A) CSF3 and (B) CSF3R in sputum cells analyzed by microarray profiling of those with severe asthma (nonsmokers) and healthy controls enrolled within the U-BIOPRED cohort. ES, Enrichment Score.View Large Image Figure ViewerDownload Hi-res image Download (PPT)