Inflammation Research https://doi.org/10.1007/s00011-018-1145-8

Inflammation Research

ORIGINAL RESEARCH PAPER

A large lung gene expression study identifying IL1B as a novel player in airway inflammation in COPD airway epithelial cells Gao Yi1,2 · Min Liang2 · Ming Li1,2 · Xiangming Fang1 · Jifang Liu2 · Yuxiong Lai2 · Jitao Chen1,2 · Wenxia Yao2 · Xiao Feng2 · La Hu2 · Chunyi Lin1 · Xinke Zhou1,2 · Zhaoyu Liu1,2  Received: 23 November 2017 / Revised: 22 February 2018 / Accepted: 24 March 2018 © Springer International Publishing AG, part of Springer Nature 2018

Abstract Background  Chronic obstructive pulmonary disease (COPD) is a chronic and progressive lung disease characterized by a mixture of small airway disease and lung tissue parenchymal destruction. Abnormal inflammatory responses to cigarette smoking and other noxious particles are generally thought to be responsible for causing of COPD. Since airway inflammation is a key factor in COPD progress, it is crucial to unravel its underlying molecular mechanisms. Unbiased analysis of genome-wide gene expression profiles in lung small airway epithelial cells provides a powerful tool to investigate this. Methods  Gene expression data of GSE611906, GSE20257, GSE8545 were downloaded from GEO database. All 288 lung small airway samples in these cohorts, including donors with (n = 61) and without (n = 227) COPD, were chosen for differential gene expression analysis. The gene ontology (GO) function, Kyoto Encyclopedia of Genes and Genomes pathway (KEGG) enrichment analyses, gene co-expression network analysis (WGCNA) and protein–protein interaction (PPI) network analysis were performed. Subsequently, the analyses of IL1B expression level, the Pearson correlation between IL1B and several COPD biomarkers were performed using other cohorts to validate our main findings. Results  With a change ≥ twofold and P value < 0.05 cutoff, we found 38 genes were up-regulated and 114 genes were downregulated in patients with COPD compared with health controls, while using cutoff fold change 1.5 and P value < 0.05, there were 318 genes up-regulated and 333 genes down-regulated. Among the most up-regulated genes were IL1B, CCL2, CCL23, and CXCL14, all implicated in inflammation triggering. GO, KEGG and WGCNA analysis all disclosed IL1B was highly correlated to COPD disease trait. The expression profile of IL1B was further validated using independent cohorts from COPD airway epithelium, lung tissue, sputum, and blood. We demonstrated higher IL1B gene expression in COPD small airway epithelial cells, but not in COPD lung tissue, sputum, and blood. Strong co-expression of IL1B with COPD biomarkers, such as DUOX2, MMP12, CCL2, and CXCL14, were validated in silico analysis. Finally, PPI network analysis using enriched data showed IL1B, CCL2, CCL7 and BMP7 were in the same hub node with high degrees. Conclusions  We identified IL1B was significantly up-regulated in COPD small airway epithelial cells and propose IL1B as a novel player in airway inflammation in COPD. Keywords  IL1B · COPD · Airway epithelial cells · Inflammation Responsible Editor: Liwu Li. Gao Yi, Min Liang, and Ming Li are equal contributors. Electronic supplementary material  The online version of this article (https​://doi.org/10.1007/s0001​1-018-1145-8) contains supplementary material, which is available to authorized users. * Xinke Zhou [email protected] * Zhaoyu Liu [email protected] Extended author information available on the last page of the article

Abbreviations IL1B Interleukin 1 beta IL1R2 Interleukin-1 receptor type 2 COPD Chronic obstructive pulmonary disease CCL2 Chemokine (C-C motif) ligand 2 CCL7 Chemokine (C-C motif) ligand 7 CCL23 Chemokine (C-C motif) ligand 23 HIF-1α Hypoxia inducible factor-1 alpha DEGs Differentially expressed genes GAPDH Glyceraldehyde-3-phosphate dehydrogenase ALDH3A1 Aldehyde dehydrogenase 3 family, member A1

13

Vol.:(0123456789)



DUOX2 Dual oxidase 2 MMP12 Matrix metalloproteinase 12 AKR1B10 Aldo–keto reductase family 1, member B10 CYP1B1 Cytochrome P450 family 1 subfamily B polypeptide 1 NQO1 NAD(P)H:quinoneoxidoreductase PPI Protein–protein interaction GS Gene significance MM Module membership GSEA Gene set enrichment analysis IL17 Interleukin-17 P450 Cytochrome P450 GO Gene ontology KEGG Kyoto Encyclopedia of Genes and Genomes WGCNA Weighted Gene Co-expression Network Analysis

Introduction Chronic obstructive pulmonary disease (COPD) is currently the fourth leading global cause of death worldwide and will become the third leading cause of death by 2030 [1]. Pathologically, COPD is characterized by a mixture of small airway disease (obstructive bronchitis) and lung tissue parenchymal destruction (emphysema) [2, 3]. Although exposure to smoke is considered to be a common dominant risk factor for development of COPD, quitting smoking just can reduce the accelerated lung function decline in some but not all COPD patients and it does not restore lost lung function [4]. There is no effective medicine for curing COPD and current drugs are mainly effective in improving and alleviating symptoms but generally do not reverse the progression of the disease. Abnormal inflammatory response to inhaled noxious particles and gases, resulting in structural changes in the airways (process termed “remodeling”), is considered to be the main mechanism leading to COPD [5, 6], whereas the specific mechanisms have remained incompletely understood. Small airways are generally defined as non-cartilaginous airways with an internal diameter less than 2 mm, which include airways from about the eighth generation down to the terminal bronchioles and respiratory bronchioles. In normal individuals, small airways contribute only a little to airway resistance, while studies have shown that the small airways are the primary site of pathologic changes, and to be the major position of airflow limitation in COPD [7–9]. The small airway component of airway limitation is characterized by remodeling of airways (goblet cell hyperplasia and small airway squamous metaplasia, etc.), obstruction of airway lumen by mucus and other inflammatory exudates, infiltration of inflammatory cells, increased airway muscle and peribronchiolar

13

G. Yi et al.

fibrosis [3]. All of these pathological features in small airway are associated with chronic inflammation. Small airway inflammatory reaction that results from exposure to cigarette smoking and other noxious particles plays an indispensable role in the original and progressive stages of COPD, which even occurs before the development of airway obstruction and fibrosis. Many studies have been done on airway inflammation in COPD [10], whereas, there still are many unknown mechanism required to be resolved. More research works are needed to elucidate the role of other inflammatory genes and pathways that are important in COPD. Unbiased analysis of high-throughput gene expression in small airway epithelial cells affords a powerful tool to investigate the inflammatory mechanism in COPD. However, studies using this approach have been impeded by the limited available of expression data from public databases. In this study, we have performed an unbiased analysis of genome-wide gene expression profiles in small airway epithelial cells obtained by bronchoscopic brushing from a large number (N = 288) of well characterized patients with (N = 61) and without (N = 227) COPD.

Materials and methods Gene expression data The gene expression profiles of GSE11906, GSE20257, GSE8545, GSE64614, GSE76925, GSE56766, and GSE22148 were download from GEO database. GSE11906, GSE20257, GSE8545, and GSE64614 were sequencing data using small airway samples, while GSE76925, GSE56766, and GSE22148 were from lung tissue, blood and induced sputum, respectively. Please see the references and online data for additional information about these cohorts [11–16].

Identification of DEGs The raw data files used for the analysis included CEL files (Affymetrix platform). The analysis was carried out using R language (version 3.3.3) and Bioconductor. Gene expression normalization was performed using the MAS 5 algorithm and every probe set as normalized. Differential expression genes were calculated using an R/Bioconductor package “DESeq”, and mRNA with a P value ≤ 0.05 and estimated absolute log2-fold change ≥ 1 in the sequence were considered to be significantly differentially expressed. Volcano plots were drawn using R software package “ggplot2”, and heatmaps for the DEGs were drawn using the R software package “Pheatmap”.

A large lung gene expression study identifying IL1B as a novel player in airway inflammation…

Gene ontology and pathway enrichment analysis of DEGs Gene ontology (GO) analysis is a common useful method for high-throughput genome or transcriptome data for identifying characteristic biological attributes. KEGG is a knowledge base for systematic analysis of gene functions, linking genomic information with higher-order functional information. Comprehensively, mapping of user’s gene to the relevant biological annotation in the DAVID database is an essential functional analysis. GO biological process (BP) and KEGG analysis were conducted using the R software package “clusterProfiler”, P < 0.01 was considered statistically significant.

Weighted gene co‑expression network analysis (WGCNA) in COPD We used WGCNA to identify co-expression modules in COPD. A more detailed description can be found at the website (https​://labs.genet​ics.ucla.edu/horva​th/Coexp​ressi​ onNetw ​ ork/Rpacka​ ges/WGCNA/​ ). In short, a weighted adjacency matrix containing pair-wise connection strengths was constructed by using the soft-thresholding approach (β  =  14) on the matrix of pair-wise correlation coefficients. A connectivity measure (k) per gene was calculated by summing the connection strengths with other genes. Modules were defined as branches of a hierarchical clustering tree using a dissimilarity measure (1—topological overlap matrix). Each module is subsequently assigned a color. Preservation of module structure was assessed using the module preservation R function. The gene expression profiles of each module were summarized by the module eigengene (defined as the first principal component of the module expression levels). Each module eigengene was regressed on disease status (COPD) using the linear model in the limma R package.

Integration of protein–protein interaction (PPI) network Search Tool for the Retrieval of Interacting Genes (STRING) database is online tool designed to evaluate the protein–protein interaction (PPI) information. STRING (version 10.5) covers 9,643,763 proteins from 2031 organisms. To evaluate the interactive relationships among enriched pathway data, we mapped them to STRING, and only experimentally validated interactions with a combined score > 0.4 were selected as significant.

were performed using IBM SPSS Statistics 22 and R software (version 3.3.3). Differences in IL1B expression levels in COPD and control samples were assessed using Mann–Whitney U tests and a two-sided P value < 0.05 was considered significant. The relation between mRNA expression of IL1B, DUOX2, MMP12, CCL2 and CXCL14 was assessed by Spearman’s correlation. A P value < 0.05 was considered significant. Availability of data and materials  Essential cohorts supporting the conclusion are included in this article.

Results Identification of DEGs in COPD small airway epithelial cells A total number of samples analyzed were 61 COPD samples and 227 normal samples. All small airway epithelial samples in cohort GSE11906, GSE20257, and GSE8545 were separately analyzed using R software and its extension packages. Based on the analysis, using P < 0.05 and absolute log2fold change ≥ 1 criteria, a total of 152 genes were identified after the analyses of GSE11906, GSE20257 and GSE8545; of which 38 genes were up-regulated and 114 genes were down-regulated. While using cutoff 1.5-fold change and P value < 0.05, there were 318 genes up-regulated and 333 genes down-regulated. A volcano plot of all probe sets is shown in Fig. 1, and a heatmap of 70 most changed genes, including 35 top up-regulated genes and 35 top down-regulated genes, is shown in Fig. 2. The complete list of all symbol genes and statistics is shown in online supplementary table S1. Among the most up-regulated genes were SPRR3 (small proline-rich repeat protein 3), cytochrome P4501A1 (CYP1A1), IL1B and IL1R2. From the up-regulated genes, IL1B is most interesting in the context of COPD pathogenesis. IL1B is a critical inflammatory factor that plays an indispensable role in a variety of inflammatory-related diseases. Meta-analysis researches have shown IL1B, IL1RN polymorphisms associated with COPD risk, and serum IL1B is highly statistically significant difference between COPD and healthy donors. Moreover, two additional genes implicated in IL1 pathway were found to be among the most up-regulated genes in COPD, which is IL1R2 and IL1RN. Give the potential novel roles for IL1B in airway inflammation in COPD, we decided to further focus on this gene in this article.

Statistics

GO term enrichment analysis among genes up‑regulated in COPD small airway epithelial cells

The expression levels of IL1B, DUOX2, MMP12, CCL2 and CXCL14 were normalized to GAPDH. Statistical analyses

We subsequently assessed the enrichment of biological processes and predicted phenotypes among the 38 genes

13



G. Yi et al.

Fig. 1  The volcano plot shows the up-regulated and down-regulated genes in COPD small airway epithelium. The horizontal axis represents the fold change between non-COPD and COPD small airway epithelium. The vertical axis represents the P value of the t test for

the differences between non-COPD and COPD small airway epithelium. The genes most relevant for COPD and this manuscript are highlighted in red. a GSE11906; b GSE20257; c GSE8545. (Color figure online)

up-regulated in patients with COPD at P < 0.01. GO enrichment analysis results showed that up-regulated DEGs were significantly enriched in biological process, including reactive oxygen species biosynthetic process, response to toxic substance, cellular response to interleukin-1, response to xenobiotic stimulus, positive regulation of inflammatory response, and regulation of cytokine secretion (Table 1). Several terms related to cytokine secretion, inflammation and IL1 pathway activation are enriched among genes upregulated in COPD small airway epithelial cells.

expression = 8.253526). The selection of the 3075 most variable transcripts was used in a hierarchical clustering procedure to identify groups of co-expressed genes, termed “modules”. These correspond to the branches of the resulting clustering tree. Each module is assigned a unique color label, which is visualized in the color band underneath the cluster tree (Fig. 3). We identified six modules range in size from 1169 genes in the Turquoise module to 43 in the Red module. We also defined an improper (grey) module with genes not belonging to any of the six proper modules. The expression profiles of transcripts inside a given module were summarized by their first principal component (referred to as module eigengene). The module eigengene is a weighted (quantitative) average of the module gene expression profiles.

Co‑expression network construction in small airway epithelial cells A weighted gene co-expression network was constructed using small airway epithelial cells expression data from cohort GSE11906, in dependent of disease (COPD) status, as described in the Methods section. For this, a total of 3075 most variable transcripts (1.5-fold) were selected. The variance of the expression of these genes range from ALDH3A1 (average expression = 22846.83) to LOC101928298 (average

13

Co‑expression modules related to COPD status To identify modules related to COPD disease status, we regressed each of the six module eigengenes on disease (COPD) status. Strikingly, all six modules were

A large lung gene expression study identifying IL1B as a novel player in airway inflammation…

Fig. 2  Heatmap of differentially expressed genes in COPD small airway epithelium. The heatmap shows differential gene expression (P < 0.05) between COPD and non-COPD control small airway epithelium for the three patients cohorts from GSE11906, GSE20257 and GSE8545. The colored, grey and red bars above the heatmap indicate the different groups of small airway epithelium. Genes with a

higher expression in COPD small airway epithelium are shown in the upper part of the heatmap, genes with a lower expression in the lower part. Blue represents lower and red higher relative expression. The genes most relevant for COPD and this manuscript are highlighted in red on the left side of the heatmap. (Color figure online)

significantly (P value < 0.05) associated with COPD status (Fig. 4; Table 2). Four modules positively correlated to COPD disease status: Turquoise (cor = 0.67, p = 4E−14), Red (cor = 0.54, p = 4e−09), Green (cor = 0.47, p = 1e−06) and Brown (cor = 0.4, p = 3e−05), containing mostly genes over-expressed in COPD. In contrast, Yellow (cor = − 0.35, p = 3E−04) and Blue (cor = − 0.34, p = 6E−04) modules were negatively correlated with COPD, meaning that the genes in these modules are predominantly under-expressed in COPD cases. Of these six gene modules, Turquoise, Red, Green and Brown modules, which were positively associated with COPD status, attracted our most attentions. Turquoise module mainly contained all the same genes previously described with detoxification and oxidative stress (e.g., AKR1B10, CYP1B1, NQO1, ALDH3A1). Red, Green and Brown Module contained genes that encode proteins related to leukocyte chemotaxis and airway inflammation, including CCL2, CCL7, IL1B, IL1R2, etc. subsequently, WGCNA was further used to calculate gene significance (GS) vs. Turquoise (and other three positive) module membership (MM). we found that Turquoise, Red, Green and Brown genes most significantly associated with COPD characteristic traits (GS) were also the most important elements of modules (MM), as demonstrated by the upper right genes in the plots of Fig. 5. In addition, IL1B, IL1R2,

CCL2, CCL7, CXCL14 and other inflammatory-related genes are found in the upper right in Fig. 5d, indicating they are both key components to the underlying biological process, and highly correlated to the trait of interest (COPD).

IL1B pathway are enriched among genes up‑regulated in COPD small airway epithelium We used gene set enrichment analysis (GSEA) to assess the enrichment of biological processes among the upregulated and down-regulated genes in COPD small airway epithelium at p < 0.01. Enrichment was found for biological processes involved in metabolic pathways, cytokine–cytokine receptor interaction, metabolism of xenobiotics by cytochrome P450, arachidonic acid metabolism, IL17 signaling pathway, etc. (Table 3). The genes enriched in cytokine–cytokine interaction pathway were subsequently evaluated the protein–protein interaction (PPI) network by STRING database. As shown in Fig. 6, IL1B and several other cytokine genes relevant for COPD pathogenesis (e.g., CCL2, CCL7, and CCL23) are clustered together in the center of the network, indicating IL1B may also as a potential inflammatory gene in COPD small airway.

13

Table 1  Pathway enrichment based on gene up-regulated in COPD small airway epithelium

G. Yi et al. ID

Term (biological processes)

P value

GO:1903409 GO:0010573 GO:0010817 GO:0071675 GO:0009636 GO:0045765 GO:0072593 GO:1901342 GO:0006766 GO:0018149 GO:0097529 GO:0009914 GO:1905039 GO:0071674 GO:1903825 GO:0030595 GO:0071347 GO:0030073 GO:0002688 GO:0071621 GO:0001819 GO:0050709 GO:0006493 GO:0002687 GO:0042493 GO:0009410 GO:0050729 GO:0050921 GO:0015833 GO:0032103 GO:0046683 GO:0044344 GO:0015849 GO:0071774 GO:0050707 GO:0014074 GO:0048514 GO:0043491

Reactive oxygen species biosynthetic process Vascular endothelial growth factor production Regulation of hormone levels Regulation of mononuclear cell migration Response to toxic substance Regulation of angiogenesis Reactive oxygen species metabolic process Regulation of vasculature development Vitamin metabolic process Peptide cross-linking Myeloid leukocyte migration Hormone transport Carboxylic acid transmembrane transport Mononuclear cell migration Organic acid transmembrane transport Leukocyte chemotaxis Cellular response to interleukin-1 Insulin secretion Regulation of leukocyte chemotaxis Granulocyte chemotaxis Positive regulation of cytokine production Negative regulation of protein secretion Protein O-linked glycosylation Positive regulation of leukocyte migration Response to drug Response to xenobiotic stimulus Positive regulation of inflammatory response Positive regulation of chemotaxis Peptide transport Positive regulation of response to external stimulus Response to organophosphorus Cellular response to fibroblast growth factor stimulus Organic acid transport Response to fibroblast growth factor Regulation of cytokine secretion Response to purine-containing compound Blood vessel morphogenesis Protein kinase B signaling

3.12E−05 4.82E−05 4.85E−05 5.28E−05 6.90E−05 8.29E−05 0.000125 0.000133 0.000176 0.000237 0.000304 0.000422 0.000498 0.000518 0.000561 0.000629 0.000809 0.000813 0.001083 0.001186 0.001205 0.001371 0.00141 0.001531 0.001532 0.001573 0.001934 0.002031 0.002174 0.002332 0.002682 0.003049 0.003379 0.003447 0.003516 0.003586 0.003719 0.003729

Verification of IL1B gene expression in COPD small airway and several tissues relevant to COPD The analysis of IL1B expression level was performed on COPD small airway and several other tissues relevant to COPD. We found significantly higher IL1B expression in COPD small airway epithelium compared with nonCOPD controls (Fig. 7a, including 37 patients with COPD and 136 non-COPD controls from GSE64614). In contrast, there was no difference of IL1B expression level in lung tissues and blood cells from patients with COPD and

13

non-COPD controls (Fig. 7b, c). In addition, there was also no difference of IL1B expression level in sputum from different COPD stages (Fig. 7d, including COPD stages 2, 3, 4 patients from GSE22148). Subsequently, Pearson correlation between IL1B and several COPD biomarker genes were performed. Significant positive correlations were found between IL1B and several genes relevant to COPD pathogenesis, including DUOX2, MMP12, CCL2, and CXCL14 (Fig. 8). These results show the robustness of our finding and validate IL1B as a novel player in small airway inflammation in COPD.

A large lung gene expression study identifying IL1B as a novel player in airway inflammation… Fig. 3  Network construction identifies distinct modules of co-expressed genes. The network was constructed using small airway epithelial cell expression data of GSE11906. The dendrogram was produced by average linkage hierarchical clustering of genes using 1— topological overlap as dissimilarity measure (see “Materials and methods” section). Modules of co-expressed genes were assigned colors corresponding to the branches indicated by the horizontal bar beneath the dendrogram

Discussion

Fig. 4  Weighted gene co-expression network analysis heat map: using the default parameter setting and 1.5-fold change genes (n = 3075), 6 gene modules were identified using WGCNA that were correlated to COPD disease trait. Each row corresponds to a module eigengene (given on the left graph margin), column to a clinical trait (COPD status). Positive correlations are red, and negative correlations are blue. (Color figure online)

Table 2  Correlation of module eigengene with COPD disease trait WGCNA modules

Blue Brown Green Red Turquoise Yellow grey

#Genes

856 213 106 43 1169 189 499

COPD dataset Correlation

P value

− 0.34 0.4 0.47 0.54 0.67 − 0.35 − 0.34

6e−04 3e−05 1e−06 9e−09 4e−14 3e−04 5e−04

The modules that were found by WGCNA in the dataset are listed together with the number of genes they contain (shown in the second column). All 6 gene modules were correlated to COPD disease trait, and correlation coefficients and P value are listed in the third and fourth column

Using genome-wide gene expression analyses on a large cohort of small airway specimens from patients with COPD and non-COPD controls, we identified a clear gene signature for inflammagenesis in COPD small airway epithelium, with increased expression of CCL2, CCL7, IL1B, and IL1R2. Among the up-regulated inflammatory genes in COPD small airway epithelium, CCL2, CCL7 are well-known inflammatory factors and have been identified as important factors involved in COPD pathogenesis [17–20]. In this study, we validated IL1B and IL1B related signaling pathway were significantly up-regulated in COPD small airway epithelial cells and found a positive correlation between the IL1B expression and COPD disease status. In addition, we showed high correlations between the gene expressions of IL1B and several COPD biomarker genes, including DUOX2, MMP12, CCL2, and CXCL14. Finally, we found the expression level of IL1B gene was only up-regulated in COPD airway epithelial cells, but not in lung tissues and blood cells. Accordingly, we suggest that IL1B may be one of the important pro-inflammatory genes involved in COPD pathogenesis, based on the fact that IL1B mRNA was over-expressed in small airway epithelium in COPD and that the expression level of IL1B was significantly with COPD disease trait. Chronic inflammation is a hallmark of COPD [21, 22]. The specific mechanism of chronic inflammation in COPD is currently unknown. It generally believed that the chronic inflammation in airway is resulted from the abnormal responses to inhaled toxic particles and gases, particularly in tobacco smoke [23, 24]. Our results showed several genes and pathways involved in xenobiotics metabolism, reactive oxygen species biosynthesis and inflammagenesis were upregulated in COPD small airway epithelium. These results are consistent with those of other works reported in previous

13



G. Yi et al.

Fig. 5  COPD disease trait absolute gene significance (GS) vs. module membership (MM). WGCNA calculation of gene significance (GS) to sample traits (COPD) vs. module membership (MM). In oversimplified terms, MM is a measure of how “tight” genes cluster within the module, or mathematically, how close gene expression is to the module eigenvalue. A gene with high MM and GS identifies hub genes that are both key components to the underlying biological Table 3  Gene set enrichment analysis of differentially expressed genes associated with COPD

Category ID

Category name

Enrichment

p value

hsa01100 hsa04060 hsa00980 hsa00590 hsa04610 hsa01230 hsa04657 hsa05204 hsa00830 hsa01200 hsa00190 hsa05012

Metabolic pathways Cytokine–cytokine receptor interaction Metabolism of xenobiotics by cytochrome P450 Arachidonic acid metabolism Complement and coagulation cascades Biosynthesis of amino acids IL-17 signaling pathway Chemical carcinogenesis Carbon metabolism Carbon metabolism Oxidative phosphorylation Parkinson’s disease

0.310522 0.41508 0.47897 0.51738 0.537833 0.55407 0.422948 0.447403 0.523712 0.413379 0.414334 0.411904

0.000999 0.001002 0.001027 0.00104 0.00104 0.001056 0.002024 0.002055 0.002114 0.004049 0.004053 0.004065

study [25–28]. Considering the fact that the pivotal roles of inflammatory reaction played in COPD pathogenesis, subsequently, we focused our attention on the up-regulated of inflammatory genes and inflammatory pathways in COPD small airway epithelial cells. As expectedly, a plenty of genes involved in pro-inflammatory and inflammatory reaction, such as CCL2, CCL7, IL1B, and IL1R2, were significantly up-regulated in COPD small airway epithelial cells.

13

process, and highly correlated to the trait of interest. In the figure, GS for COPD disease trait was plotted against a Turquoise module membership; b red module membership; c brown module membership; d green module membership. AKR1B10, CYP1B1, NQO1, ALDH3A1, IL1B, IL1R2, CCL2, CCL7 and CXCL14 are found in the top right corner of each graph. (Color figure online)

Current strategies of therapy for COPD are relatively ineffective, as there are no drugs available for curing COPD. Long-acting bronchodilators are the mainstay of current forms of COPD therapy [29, 30]. Although these drugs, β2adrenergic receptor agonists, produce effective bronchodilation, they fail to treat the underlying inflammatory disease in patients with COPD [31]. Inhaled corticosteroids, which are effective at treating airway inflammation in patients with

A large lung gene expression study identifying IL1B as a novel player in airway inflammation…

Fig. 6  Protein–protein interaction (PPI) network of the genes enriched in cytokine–cytokine interaction pathway. Gene set enrichment analysis (GSEA) shows cytokine–cytokine interaction pathway is enriched in COPD small airway epithelium at P < 0.01. The enrichment method assesses whether a priori defined set of genes shows statistically significant, concordant differences between two biological states. PPI network shows IL1B and several COPD-related cytokines are clustered together in the center of the network

asthma, are largely ineffective as an anti-inflammatory therapy in COPD [32]. In our studies, the levels of many inflammatory cytokines are elevated in patients with COPD, which may provide some of the most promising new therapeutic targets for COPD therapy. Among the up-regulated cytokines, many studies have been reported that CCL2 and CCL7 were up-regulated in COPD, and played a vital role in recruiting inflammatory cells to the airway. Although IL1B is an important proinflammatory cytokine involved in several diseases, nothing is known regarding expression of IL1B in COPD airway epithelium. As a potent pro-inflammatory cytokine, IL1B plays a crucial role in many inflammation related diseases [33]. Several studies have shown the polymorphisms of IL1B and IL1B promotor associated with COPD risk [34–36], and serum IL1β is also evaluated in COPD patients [37, 38]. However, the expression level of IL1B in COPD is not clear. Our results showed that the expression of IL1B level was significantly up-regulated about 2- to 3-fold in COPD small airway epithelial cells. IL1B is mainly produced by stimulated monocytes, macrophages neutrophils, and several stromal cell types, including epithelial cells, keratinocytes and endothelial cells [39]. Exogenous stimuli, such as endotoxin and cigarette smoke extract (CSE), have been known

to increase IL1B expression and IL1β secretion [40, 41]. We suggest that the progressive inflammation in the airway may promote the gene of IL1B expression in airway epithelium, whereas the mechanism is unclear. Several recent studies have reported that inflammatory signal can induces IL1β expression through HIF-1α [42], and inflammatory cytokines can change the DNA methylation status at key CpG sites [43], resulting in long-term induction of IL1B expression. Subsequently, we want to determine whether the expression level of IL1B was also up-regulated in COPD lung tissue and COPD blood cells. Intriguingly, the expression level of IL1B did not change significantly in COPD lung tissue and COPD blood cells. The reasons for the discrepancy in different tissue are not clear, but several possible explanations exist. Firstly, as the primary site of pathologic changes in COPD, the small airway epithelial cells may more sensitive to inflammatory stimuli. Secondly, Airway epithelial cells have been reported as one of the potent sources for cytokine IL1β, whereas, there was no information about IL1B expression in lung alveolar cells. Finally, although there are some studies reported the serum IL1β was increased in COPD, the mRNA expression level of IL1B was not clear. A recently sequencing data revealed no transcripts in the whole blood that distinguishes COPD patients and control smokers (GSE56766, https:​ //www.ncbi.nlm.nih.gov/ geo/query/​ acc.cgi?acc=GSE567​ 66). In addition, the correlation between IL1B expression level and COPD disease trait also was calculated using WGCNA matrix. As we expect, the expression of IL1B level was significant with COPD disease trait, which was located in the upper right in Module membership vs. gene significance figure (Fig. 5).

Conclusion In the current study, we found a clear inflammagenesis gene signature in COPD small airway epithelium, with increased gene expression of CCL2, CCL7, IL1B and IL1R2 in lung small airway epithelial cells. We provide evidence for IL1B as a novel player in inflammation. We suggest that increased IL1B mRNA expression in COPD small airway epithelial cells may be due to a response to exogenous stimuli and inflammatory microenvironment. Further studies are required to address the question of whether IL1B is a dominant pro-inflammatory cytokine in COPD. A detailed understanding of the mechanisms of the regulation IL1B expression may provide an insight into chronic airway inflammation in COPD and lead to therapeutic opportunities for treating COPD.

13



Fig. 7  IL1B expression in different tissues relevant to COPD. Validation IL1B mRNA expression relative to housekeeping gene expression in small airway epithelium (a), lung tissue (b), blood (c) and

13

G. Yi et al.

sputum (d) from patients with COPD and non-COPD controls. The results of the Mann–Whitney U analyses are depicted in the figure

A large lung gene expression study identifying IL1B as a novel player in airway inflammation…

Fig. 8  The Pearson correlation between the expression of IL1B and several biomarkers relevant to COPD in the airway epithelium. The results of the Spearman correlation is depicted in the top right corner of each graph Acknowledgements  We thank all members from department of central laboratory, the Fifth Affiliated Hospital of Guangzhou Medical University for their invaluable help. Author contributions  ZYL, XKZ and JFL designed the study; ML (Min Liang), ML (Ming Li) and XMF performed data collection; GY and YXL analyzed the data; ZYL and GY wrote the manuscript. All authors read and approved the final manuscript. Funding  This work was supported by the National Natural Science Foundation of China (Grant Number. 81400013), Science and Technology Planning Project of Guangdong Province, China (Grant Number. 2014A20212329) and Department of education of GuangDong Province, China (Grant Number. 2016KTSCX110).

Compliance with ethical standards  Conflict of interest  The authors declare that they have no competing interests. Ethical approval and consent to participate  In the current study, all analyses were based on publicly available data, and this article does not contain any studies with human participants and animals performed by any of the authors. Consent for publication  Not applicable.

13



References 1. Diaz-Guzman E, Mannino DM. Epidemiology and prevalence of chronic obstructive pulmonary disease. Clin Chest Med. 2014;35:7–16. 2. Hogg JC, Timens W. The pathology of chronic obstructive pulmonary disease. Ann Rev Pathol 2009;4:435–59. 3. McDonough JE, Yuan R, Suzuki M, Seyednejad N, Elliott WM, Sanchez PG, et al. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med. 2011;365:1567–75. 4. Tonnesen P. Smoking cessation. and COPD. Eur Respir Rev Off J Eur Respir Soc. 2013;22:37–43. 5. Vogelmeier CF, Criner GJ, Martinez FJ, Anzueto A, Barnes PJ, Bourbeau J, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive lung disease 2017 report: GOLD executive summary. Eur Respir J. 2017;49(3). https​://doi. org/10.1183/13993​003.00214​-2017. 6. Goncalves RB, Coletta RD, Silverio KG, Benevides L, Casati MZ, da Silva JS, et al. Impact of smoking on inflammation: overview of molecular mechanisms. Inflamm Res. 2011;60:409–24. 7. Kim WD, Ling SH, Coxson HO, English JC, Yee J, Levy RD, et  al. The association between small airway obstruction and emphysema phenotypes in COPD. Chest. 2007;131:1372–8. 8. Hogg JC, McDonough JE, Suzuki M. Small airway obstruction in COPD: new insights based on micro-CT imaging and MRI imaging. Chest. 2013;143:1436–43. 9. Hogg JC, Pare PD, Hackett TL. The contribution of small airway obstruction to the pathogenesis of chronic obstructive pulmonary disease. Physiol Rev. 2017;97:529–52. 10. Angelis N, Porpodis K, Zarogoulidis P, Spyratos D, Kioumis I, Papaiwannou A, et al. Airway inflammation in chronic obstructive pulmonary disease. J Thorac Dis. 2014;6(Suppl 1):S167–72. 11. Yang J, Zuo WL, Fukui T, Chao I, Gomi K, Lee B, et al. Smoking-dependent distal-to-proximal repatterning of the adult human small airway epithelium. Am J Respir Crit Care Med. 2017;196:340–52. 12. Shaykhiev R, Otaki F, Bonsu P, Dang DT, Teater M, StruloviciBarel Y, et al. Cigarette smoking reprograms apical junctional complex molecular architecture in the human airway epithelium in vivo. Cell Mol Life Sci CMLS 2011;68:877–92. 13. Raman T, O’Connor TP, Hackett NR, Wang W, Harvey BG, Attiyeh MA, et al. Quality control in microarray assessment of gene expression in human airway epithelium. BMC Genom. 2009;10:493. 14. Ammous Z, Hackett NR, Butler MW, Raman T, Dolgalev I, O’Connor TP, et al. Variability in small airway epithelial gene expression among normal smokers. Chest. 2008;133:1344–53. 15. Morrow JD, Zhou X, Lao T, Jiang Z, DeMeo DL, Cho MH, et al. Functional interactors of three genome-wide association study genes are differentially expressed in severe chronic obstructive pulmonary disease lung tissue. Sci Rep. 2017;7:44232. 16. Singh D, Fox SM, Tal-Singer R, Plumb J, Bates S, Broad P, et al. Induced sputum genes associated with spirometric and radiological disease severity in COPD ex-smokers. Thorax. 2011;66:489 – 95. 17. Davies C, Rhodes JA, Barnes P, Donnelly L. Elevated CCL2 responses in COPD and attenuation by selective chemokine receptor antagonists. Eur Respir J. 2015;46(suppl 59):PA3900. 18. Abdel-Halim M, Darwish SS, ElHady AK, Hoppstadter J, Abadi AH, Hartmann RW, et al. Pharmacological inhibition of protein kinase C (PKC)zeta downregulates the expression of cytokines involved in the pathogenesis of chronic obstructive pulmonary disease (COPD). Eur J Pharm Sci Off J Eur Fed Pharm Sci. 2016;93:405–9.

13

G. Yi et al. 19. Qin S, Huleihel L, Lucht L, Clarke A, Ries JW, Kessinger C, et al. Alterations of inflammatory chemokine and matrix metalloproteinase mRNA levels in BAL cells from HIV-infected COPD patients. Am J Resp Crit Care. 2015;191:A4716. 20. Frankenberger M, Eder C, Hofer TP, Heimbeck I, Skokann K, Kassner G, et al. Chemokine expression by small sputum macrophages in COPD. Mol Med. 2011;17:762–70. 21. Eapen MS, Myers S, Walters EH, Sohal SS. Airway inflammation in chronic obstructive pulmonary disease (COPD): a true paradox. Expert Rev Respir Med. 2017;11:827–39. 22. Mehta H, Nazzal K, Sadikot RT. Cigarette smoking and innate immunity. Inflamm Res Off J Eur Histamine Res Soc [et al]. 2008;57:497–503. 23. Vogelmeier C, Koczulla R, Fehrenbach H, Bals R. Pathogenesis of chronic obstructive pulmonary disease. Der Internist. 2006;47(6):885–6 (888–90, 892–4). 24. Szulakowski P, Mroz RM, Pierzchala W, Chyczewska E, MacNee W. Pathogenesis of chronic obstructive pulmonary disease. Molecular mechanisms (part II). Wiad Lek. 2006;59:250–4. 25. McGuinness AJ, Sapey E. Oxidative stress in COPD: sources, markers, and potential mechanisms. J Clin Med. 2017;6(2):21. 26. Choudhury G, MacNee W. Role of inflammation and oxidative stress in the pathology of ageing in COPD: potential therapeutic interventions. Copd. 2017;14:122–35. 27. Malic Z, Topic A, Francuski D, Stankovic M, Nagorni-Obradovic L, Markovic B, et al. Oxidative stress and genetic variants of xenobiotic-metabolising enzymes associated with copd development and severity in serbian adults. Copd. 2017;14:95–104. 28. Meihua G, Jian W, Nanshan Z. [esearch progress of the role of oxidative stress in the pathogenesis of COPD. Zhonghua jie he he hu xi za zhi = Zhonghua jiehe he huxi zazhi = Chin J Tuberc Respir Dis. 2015;38:222–4. 29. Vestbo J, Hurd SS, Agusti AG, Jones PW, Vogelmeier C, Anzueto A, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med. 2013;187:347–65. 30. Barnes PJ. New anti-inflammatory targets for chronic obstructive pulmonary disease. Nat Rev Drug Discov. 2013;12:543–59. 31. Vogelmeier CF, Criner GJ, Martinez FJ, Anzueto A, Barnes PJ, Bourbeau J, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive lung disease 2017 report. gold executive summary. Am J Respir Crit Care Med. 2017;195:557–82. 32. Barnes PJ. Inhaled corticosteroids in COPD: a controversy. Respir Int Rev Thorac Dis. 2010;80:89–95. 33. Gabay C, Lamacchia C, Palmer G. IL-1 pathways in inflammation and human diseases. Nat Rev Rheumatol. 2010;6:232–41. 34. Xie ZK, Huang QP, Huang J, Xie ZF. Association between the IL1B, IL1RN polymorphisms and COPD risk: a meta-analysis. Sci Rep. 2014;4:6202. 35. Mei JJ, Liang Y, Shen N, He B. Association between interleukin1B polymorphisms and chronic obstructive pulmonary disease: a meta-analysis. Zhonghua yi xue za zhi. 2013;93:910–5. 36. Danilko KV, Korytyna GF, Akhmadishina LZ, Yanbaeva DG, Zagidullin SZ, Victorova TV. Association of polymorphisms of cytokine genes (IL1B, IL1RN, TNFA, LTA, IL6, IL8, and IL10) with chronic obstructive pulmonary disease. Mol Biol. 2007;41:22–31. 37. Mahajan B, Vijayan VK, Agarwal MK, Bansal SK. Serum interleukin-1beta as a marker for differentiation of asthma and chronic obstructive pulmonary disease. Biomarkers. 2008;13:713–27. 38. Hammad DR, Elgazzar AG, Essawy TS, Sameie SA. Evaluation of serum interleukin-1 beta as an inflammatory marker in copd patients. Egypt J Chest Dis Tuberc. 2015;64:347–52. 39. Dinarello CA. Immunological and inflammatory functions of the interleukin-1 family. Ann Rev Immunol. 2009;27:519–50.

A large lung gene expression study identifying IL1B as a novel player in airway inflammation… 40. Skerrett SJ, Liggitt HD, Hajjar AM, Ernst RK, Miller SI, Wilson CB. Respiratory epithelial cells regulate lung inflammation in response to inhaled endotoxin. Am J Physiol Lung Cell Mol Physiol. 2004;287:L143–52. 41. Hellermann GR, Nagy SB, Kong X, Lockey RF, Mohapatra SS. Mechanism of cigarette smoke condensate-induced acute inflammatory response in human bronchial epithelial cells. Respir Res. 2002;3:22.

42. Tannahill GM, Curtis AM, Adamik J, Palsson-McDermott EM, McGettrick AF, Goel G, Frezza C, et al. Succinate is an inflammatory signal that induces il-1β through hif-1α. Nature. 2013;496(7444):238–42. 43. Hashimoto K, Oreffo RO, Gibson MB, Goldring MB, Roach HI. DNA demethylation at specific CpG sites in the IL1B promoter in response to inflammatory cytokines in human articular chondrocytes. Arthritis Rheum. 2009;60:3303–13.

Affiliations Gao Yi1,2 · Min Liang2 · Ming Li1,2 · Xiangming Fang1 · Jifang Liu2 · Yuxiong Lai2 · Jitao Chen1,2 · Wenxia Yao2 · Xiao Feng2 · La Hu2 · Chunyi Lin1 · Xinke Zhou1,2 · Zhaoyu Liu1,2  Gao Yi [email protected]

Xiao Feng [email protected]

Min Liang [email protected]

La Hu [email protected]

Ming Li [email protected]

Chunyi Lin [email protected]

Xiangming Fang [email protected]

1



Department of Respiratory Medicine, State Key Laboratory of Respiratory Disease, The Fifth Affiliated Hospital of Guangzhou Medical University, Guangzhou 510700, China



Department of Center Laboratory, The Fifth Affiliated Hospital of Guangzhou Medical University, Guangzhou 510700, China

Jifang Liu [email protected] Yuxiong Lai [email protected] Jitao Chen [email protected]

2

Wenxia Yao [email protected]

13