Abstract
Asthma is a disease of complex origin and multiple pathologies. There are currently very few biomarkers of proven utility in its diagnosis, management or response to treatment. Recent studies have identified multiple asthma phenotypes following biofluid analysis; however, such findings may be driven by the well-characterised alterations in immune cell populations in asthma. We present a study designed to identify cell type-specific gene signatures of severe allergic asthma in peripheral blood samples. Using transcriptomic profiling of four magnetically purified peripheral blood cell types, we identify significant gene expression changes in monocytes and NK cells but not T lymphocytes in severe asthmatics. Pathway analysis indicates dysfunction of immune cell regulation and bacterial suppression in the NK cells. These gene expression changes may be useful on their own as prognostic peripheral blood cell markers of severe asthma, but also may indicate novel cell pathways for therapeutic intervention.
Introduction
Asthma is a common chronic airways disease with complex aetiology, no single causative genetic trigger and with multiple factors that appear to contribute to disease pathogenesis 1. Thus, asthma can manifest in a number of different pathologies related to severity and the degree of allergy or response to therapeutic intervention.
Currently, the principal clinical biomarkers for initial diagnosis and staging of disease severity include lung function and hyper-responsiveness tests. Sometimes FeNO is used as a surrogate for direct measurement of eosinophils. Improvement of clinical symptoms following therapeutic intervention and assessment of atopy and a history of wheeze are also useful. Monitoring of the disease is generally through symptom assessment and therapeutic dose requirements combined with periodic lung function tests. Novel approaches, such as immunotherapy, have identified the utility of additional biomarkers such as IgE or sputum inflammatory cell counts as inclusion criteria; however, in clinical practice, often the principal read-outs remain the generalised clinical ones. No single test provides definitive evidence of asthma, its severity or therapeutic efficacy of drugs.
Recent work using unbiased clustering of genomic and proteomic data from airway samples has identified multiple disease phenotypes 2. Similarly, recent work has identified mRNA signatures of severe asthma in whole blood 3. Although such studies are important, they may be confounded by known changes in inflammatory cell profiles in asthma. Thus, such phenotypic clusters may define groups of subjects on the basis of inflammatory cell predominance in biofluids and changes in blood cell sub-groups could be missed.
The aim of our study was to examine gene expression in four cell populations: monocytes, NK cells, CD4+ T cells and CD8+ T cells, magnetically enriched from peripheral blood, and to compare them between severe allergic asthmatics and matched healthy subjects in order to search for cell-subtype specific biomarkers. These biomarkers may yield new information about the status of the peripheral circulation in severe allergic asthma.
Results and discussion
To examine the effect of severe allergic asthma on peripheral blood cells, we recruited 11 severe allergic asthmatics on step 4 of BTS/Sign management, and 10 healthy control subjects (see Supplementary methods for inclusion/exclusion criteria, and Table 1 for a summary of patient characteristics).
Fresh blood was drawn and PBMCs were isolated from buffy coats (see supplementary method section for experimental details). The relative proportions of four immune cell types, monocytes, NK cells, CD4+ and CD8+ T cells were determined by flow cytometry. No difference was found in the proportion of these cell populations in the peripheral blood of healthy and severe asthmatic subjects (Figure 1, see Figure S1 for flow cytometric gating strategy). NKT cells (CD3+CD56+ cells) were not quantified as they constituted less than 1% of peripheral cells (data not shown).
PBMCs isolated from the peripheral blood of healthy and severe allergic asthmatic subjects was analysed for the presence of cells expressing CD3, CD4, CD8, CD14 and CD56 by flow cytometry. CD4 and CD8+ T cells are expressed as % of CD3+ cells. CD14 and CD56 are expressed as % of CD3-cells. N=10 healthy and 11 asthmatic subjects.
Four cell populations were sequentially enriched from 4 × 107 PBMCs from each subject using magnetic beads conjugated to monoclonal antibodies targeting different cell surface markers, namely CD14 (monocytes), CD56 (NK cells), and CD8+ and CD4+ T cells in that order. These populations were highly enriched (>99%) for each selective marker (Figure 2).
Co-expression of cell surface markers in MACS-isolated populations of cell from PBMCs. CD14+, CD56+, CD8+ and CD4+ cell populations were magnetically isolated from PBMCs. The simultaneous expression of these target cell surface markers was then measured by flow cytometry in each purified population. Data shown are from one healthy subject and are representative of typical marker distribution in these populations.
Flow cytometry indicated little co-expression of cell surface markers in each purified population, but monocytes expressed low levels of CD4, as previously demonstrated 4, and CD56+ cells also expressed some CD8 and to a lesser extent, CD4, possibly reflecting the presence of other CD56+ cell populations than NK cells, including NKT cells and NK-like cells, both of which can also express these markers 5,6. Gene expression was quantified using isolated RNA from the four enriched cell populations and analysed by a 770 gene nCounter pan cancer immune profiling array (nanostring Technologies).
Cluster analysis of peripheral blood cells
Unbiased PCA analysis of the gene expression profiles of these purified cell populations indicated robust separation of the four target cell populations (Figure 3A). There was a small degree of overlap between the CD4+ and CD8 + T cell clusters. We found little evidence of segregation between health and asthma across the cell types.
(A) PCA analysis of gene counts from all four cell types indicate clusters corresponding to each cell type. (B) Volcano plots indicating genes significantly up or down regulated in severe asthma in each cell type (p-adjusted < 0.05).
We then identified differentially expressed genes (DEGs) in asthma, adjusting for multiple tests for each cell type as appropriate (See appendix 1 for full differential gene expression results). Volcano plots indicate two genes upregulated in monocytes, and 6 upregulated and 8 downregulated in NK cells from severe asthmatics (Figure 3B).
DEGs in Monocytes
Two genes, IGF2R and C3AR1 were upregulated in monocytes from asthmatics. IGF2R upregulation has been observed in monocytes during differentiation into macrophages 7 and assists in efferocytosis. Changes in IGF2R expression by monocytes in asthma have not previously been observed, although sputum cells show upregulation 2.
C3AR1 has been previously identified as a susceptibility gene for bronchial asthma 8. The gene product acts as a G-protein coupled receptor for C3a, an anaphylotoxin released during activation of the complement system. Binding of the receptor is thought to activate chemotaxis, degranulation, superoxide production and bacterial opsonisation. C3AR1 has been proposed as a therapeutic avenue for asthma 9, as its ligand, C3a, is central to the allergen-mediated Th2 response. TH17 cell-mediated secretion of IL-17 regulates C3a secretion by epithelial cells in allergic asthma models 10. To our knowledge, no previous studies have identified elevation of the C3AR1 in monocytes of severe asthmatics
DEGs in NK-enriched cells
Cells highly enriched for CD56 expression demonstrated upregulation of six genes (LILRA5, LILRB2 and SLC11A1, PDGFRB, LYN, CD97) and downregulation of eight others (LTB, RORC, GZMK, DPP4, RPS6, FLT3LG, DOCK9 and CD28).
Upregulated genes in NK cells
Several of the gene products found in this study to be upregulated in NK cells in asthma are associated with immune cell activation, such as LILRA5 and CD97. Additionally, LYN-dependent signalling controls DC activation of NK cells, affecting the magnitude of inflammatory responses 11. CD97 is a GPCR which regulates leukocyte adhesion in inflammation and granulocyte homeostasis but has not previously been directly implicated in asthma.
Other genes are involved in pathogen resistance; for example SLC11A1 plays a role in divalent metal ion transport, thought to enable sequestration of Fe2+ and Mn2+ and thus promoting pathogen resistance. Little is known about the role of PDGFRB in immune cells in asthma.
Downregulated genes in NK cells
Downregulated gene expression in CD56+ cells included genes involved in immune cell signalling, cell activation and cytotoxic functions.
LTB is leukotriene B4 receptor, a chemoattractant well characterised for its role in neutrophilic TH-17-mediated asthma inflammation. Elevation of LTB has been observed in lung tissue of mouse models of OVA sensitisation 12. Downregulation in peripheral blood NK cells may indicate suppression by corticosteroid use.
FLT3 ligand (FLT3LG) is a growth factor which supports DC proliferation. Application in mouse OVA models suppresses late allergen symptom onset in response to OVA application13.
RORC is a nuclear receptor thought to mediate cellular differentiation and lipid metabolism. One isoform of the protein is thought to antagonise the Th1 pathway and promote Th17 differentiation. NK cells express IL17A in mouse OVA models, contributing to viral induced asthma exacerbation through accumulation of neutrophils 14.
DPP4 is a receptor involved in the costimulatory signal for the TcR, it induces T cell proliferation and nfκb activation. DPP4 shows increased expression in NK cell activation and is thought to support NK cell cytokine production 15.
CD28 is a co-stimulatory receptor for T cells. Sputum CD8+ T cells have previously been shown to express less CD28 16 and may reflect an increase in NK-like CD8+ T cells with regulatory properties 6.
GZMK encodes granzyme K, a component of the granules of cytotoxic lymphocytes, including peripheral NK cells 17. Deficiencies in NK cell antiviral response in the form of reduced cytotoxicity of NK cells in severe asthma, indicative of exhaustion, have recently been observed 18.
Little is known about the expression of DOCK9 or RPS6 in relation to asthma or NK cells.
Summary of changes in asthma
Pathway analysis 19 using the differentially expressed genes in the NK cell-enriched population (Table 2) confirmed that severe allergic asthma was associated with upregulation of pathways involved in metabolism, divalent metal ion transport, immune response inhibition, phosphate metabolism and immune regulation, and with downregulation of lymphocyte differentiation, lymph node development, mononuclear cell differentiation and immune system development.
Overall, these data suggest that severe asthma is accompanied by NK cell dysfunctions that would impair the adaptive immune response and innate anti-microbial defences and could reflect exhaustion or generalised impairment of these cells. Additionally, monocytes in asthma appear to show increases in gene expression associated with apoptotic body clearance and chemotaxis. Thus, in our study, innate but not adaptive cell phenotypes were changed in peripheral blood cells from severe asthmatics. Given the increasing interest in innate immunity and its role in viral infections, it would be important to study the gene expression at times of acute asthma exacerbations. These findings may help to identify new therapeutic avenues for severe asthma, but also may serve as useful cell type-specific biomarkers.
As this was an exploratory study with no independent test group, further work would be required to assess the potential utility of these genes as biomarkers of disease progression or therapeutic response. The ability to assign these changes to cell populations, however, does allow us to infer some biological meaning. Although NK and NK-like cells have been examined in asthma, no fixed role for them has yet been assigned. Our study suggests that further investigation of these cells is warranted. Recent evidence indicates abnormal antiviral response in NK cells in severe asthma associated with exhaustion 18, our data suggest that other antimicrobial dysfunctions also exist.
The strengths of our study include the ability to assign gene expression changes in asthma to highly enriched individual cell populations, and the well-defined severe allergic asthma and matched healthy control cohorts. Limitations of the study include a small cohort size and the potential, when enriching on the basis of cell surface markers, that those markers themselves may be altered in asthma.
Data Availability
All data produced in the present work are contained in the manuscript
Supplementary material
Supplementary Figures
i) Size gate to exclude doublets. ii) CD3+ stain to positively identify T cells. iii) CD8 and CD4 staining of CD3+ cells identifies T lymphocyte subpopulations. iv) Gating on CD3-cells, CD56+ (predominantly NK cells) and CD14+ (monocytes) cell populations.
Supplementary Methods
Ethics approvals
The study was approved by the regional Research Ethics Committee (REC), (Southampton and South West Hampshire Research Ethics Committee, LREC no: 09/H0504/109) and was conducted in accordance with Site Specific Assessment (SSA), and local Research and Development (R&D) approvals.
Sputum induction
Sputum induction was performed as previously reported1 at least 4 weeks after a respiratory tract infection. The mucous elements were selected from the expectorated sample and treated with 5 mM (final concentration) dithioerythritol in HEPES-buffered saline (HBS) to liquefy the mucus and obtain sputum cell cytospin slides. Differential cell counts were obtained using the rapid Romanowski method. Cell counts were expressed as the percentage of total respiratory cells (excluding squamous cell counts but including epithelial cells).
Purification of blood cell populations
Fresh blood drawn into heparinised tubes (60 ml) was mixed 1:1 with Dulbeccos PBS and carefully layered onto Lymphoprep. Cells were centrifuged at 800 g for 30 min at room temperature. Cells at the interface were then harvested and further diluted with PBS and centrifuged at 400 g for 10 min at room temperature to remove excess density medium. Cells were resuspended in a small volume of PBS and counted using a haemocytometer. In all cases, cell viability was greater than 95%. 4 × 107 cells were then centrifuged at 400 g for 5 min and resuspended in 320 □l of MACS isolation buffer (5% (w/v) BSA, 2 mM EDTA in PBS) and the cell slurry incubated with 80 □l of anti-CD14 magnetic beads at 4°C for 15 min. Then 2 ml of MACS buffer were added and the slurry washed by centrifuging at 400g, 4°C for 10 min. The cell pellet was resuspended in 0.5 ml of MACS buffer and applied to an LS column. Retained cells were washed with MACS buffer as previously described. Eluted cells were incubated with magnetic beads conjugated to anti-CD56, CD8 and CD4 monoclonal antibodies in succession using the same procedure. Purified cells of each type were then counted using a haemocytometer, and 1 × 106 cells of each subtype were centrifuged for 10 min, 4°C at 400 g, and each cell pellet resuspended RLT buffer containing 1% (v/v) mercaptoethanol. Cell lysates were stored at -80°C.
Purity assessment of isolation procedure
Purity of cell sub-populations isolated in this was assessed on a subset of PBMC purifications using a mixture of healthy and asthmatic subjects. Cells (1 × 105 of purified cells and 1 × 106 of starting PBMCs) were resuspended in 100 □l of MACS buffer containing 100□g of human IgG as Fc□ block. An antibody cocktail was then added, containing 10 □l each of anti-CD14/FITC, Anti-CD56/PECF594, anti-CD8/APC and anti-CD4/PerCPCy5.5 and incubated for 30 min on ice. Excess antibody was removed by the addition of 2 ml of MACS buffer and centrifugation for 10 min, 400 g at 4°C. Cells were resuspended in MACS buffer prior to analysis by flow cytometry using a FACSAria (BD) equipped with relevant lasers and filters.
RNA profile analysis
RNA from each cell type was extracted using an RNeasy mini kit (Qiagen), yielding between 0.5-1.5mg of RNA per sample, the quality of which was assessed using an RNA6000 Nano kit (Agilent). Chips were read using an Agilent 2100 Bioanalyzer equipped with relevant software. All samples gave RNA integrity numbers (RIN) >9.4. RNA quantity was assessed using Nanodrop 2000 spectrophotometer (Thermo) and concentration readings were used to aliquot 50 ng of RNA and apply to PanCancer immune profiling nCounter code set cartridges (nanoString Technologies) measuring 770 genes. Read outputs in RCC file format were analysed using Nanostring nSolver analysis software.
Bioinformatic analysis
The unnormalized gene counts for genes quantified in all samples were analysed in R using associated packages to perform principal component analysis-based clustering, differential gene expression and gene enrichment. 2-6 An adjusted p-value threshold of 0.05 was set for significantly differentially expressed genes. All other settings were package defaults.
References
