RAGE-mediated functional DNA methylated modification contributes to cigarette smoke-induced airway inflammation in mice

Chronic obstructive pulmonary disease (COPD) is characterized by persistent respiratory symptoms and airflow limitation that are associated with persistent airway inflammation induced by cigarette smoke (CS), a major causative factor for COPD [1].

Receptor for advanced glycation end-products (RAGE), a membrane protein from the immunoglobulin superfamily, has been implicated in the pathogenesis of COPD [2]. Overexpression of RAGE contributed to airway inflammation in CS-associated COPD [3]. Our previous study further indicated knockout (KO) of RAGE gene significantly attenuated CS-induced airway inflammation in a mouse model [4]. However, the mechanisms regarding the effects of RAGE on airway inflammation in COPD remain not clear. Recently, some evidences implied DNA methylation, an epigenetic regulation, played important roles in RAGE-mediated inflammatory responses in various diseases [5–7], although the role in DNA methylated modification mediated by RAGE in airway inflammation in COPD was not reported.

In consequence, we did the present study, using the established mouse model of CS-induced airway inflammation, to explore the underlying mechanisms of RAGE-mediated DNA methylated modification in airway inflammation in COPD.

The animal model of CS-induced airway inflammation has been established in our former study [4]. C57BL/6 mice (7–9 weeks old, 20–22 g weight) were used to generate RAGE KO mice through CRISPR/Cas9 gene targeting technology by Bioray Biotechnology (Shanghai, China). The four experimental groups (n=3 mice per group) were included in the present study, as follows: (i) wildtype (WT) group, (ii) CS+WT group, (iii) CS+KO group, (iv) WT+KO group. All mice were specific pathogen-free and kept on a 12-h light/12-h dark cycle, at a room temperature of 22 ± 2°C, with free access to food and water. WT and RAGE KO mice were exposed to mainstream CS or room air for 2 h twice daily, 6 days per week for consecutive 4 weeks. After 4-week CS exposure, the mice were anesthetized intraperitoneally with pentobarbital sodium and killed by femoral artery transection. The animal study was approved by the Panel on Laboratory Animal Care of West China Hospital of Sichuan University and took place at the Experimental Animal Center of West China Hospital of Sichuan University.

The total DNA of lung tissues was extracted and purified by DNeasy Blood Tissue Kit (Qiagen), according to the manufacturer’s instructions. Purified DNA was then quantified by NanoDrop2000 Spectrophotometer (Thermo) and agarose gel electrophoresis. Only DNA samples with A_260/280 ratio between 1.8 and 2.0 were used for further experiments.

Genomic DNA (1 μg per sample) was randomly fragmented into approximately 200–300 bp by sonication. After purification, the DNA fragments were repaired in the blunt and phosphorylated ends, which were subsequently 3′ adenylated and then ligated to the methylated adapter using the SureSelect^XT Mouse methyl-seq Library Prep Kit (Agilent). After that, the DNA hybridization was performed using the SureSelect™ Methyl-SeQ Hybridization Kit (Agilent), which covered 100 Mb of mouse genomic regions, including CpG islands, Gencode promoters, tissue-specific differential methylated regions (DMRs) and DNase I hypersensitive sites. The hybridized DNA was subsequently bisulfite-treated using the EZ DNA Methylation-Gold™ Kit (Zymo Research) to convert unmethylated cytosine into uracil according to the manufacturer’s instructions. Finally, the treated DNA was amplified by polymerase chain reaction (PCR) and sequenced on Illumina Novaseq PE150.

The Fastp software (version 1.2.1) was used to process methylated raw data and remove the low quality reads, including (i) contaminated sequences; (ii) the Q-value of 3′ end is less than 20; (iii) reads with less than 15 bp; (iv) reads with 40% base Q-value less than 15; (v) reads containing N bases greater than 5 (Q = −10*log10(P), P is the probability of error) [8]. The clean reads were aligned to the reference genome using Bismark software (version 0.19.0) with bowtie2 (version 2.3.4.2) to attain the methylated type, status and proportion [9]. DMRs were subsequently analyzed using the R package methylKit (version 1.6.1) [10] and eDMR [11], with an adjusted P-value <0.05 and absolute differential methylation levels (absolute meth. diff > 5%). Finally, related differential methylated genes (DMGs) were located and annotated in the DMRs by the ChIPseeker software.

The data of gene expression by cDNA microarray have been obtained in our former study [4]. In the present study, differentially expressed genes (DEGs) were identified by fold-change, and only genes that at least 1.2-fold up-regulated or down-regulated were analyzed.

To identify the candidate genes with methylated modification mediated by RAGE, a novel intersection model was performed (Figure 1). Briefly, CS-associated (CS+WT vs WT) and RAGE-associated (CS+KO vs CS+WT plus WT+KO vs WT) DMGs intersected to get the overlapped DMGs, and so did CS-associated and RAGE-associated DEGs (overlapped DEGs). Then, the overlapped DMGs and DEGs intersected again to select the candidate genes.

The candidate genes with functional methylated modification, also called functional methylated genes, were identified according to the reversed effects of methylation in promoters on gene transcription. To further identify the hub genes that were most correlated with other genes, protein–protein interaction (PPI) network analysis among the functional methylated genes was performed using the online database STRING [12] and displayed using Cytoscape [13].

Functional enrichment analyses on the functional methylated genes, as well as the hub genes, using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases were performed with the clusterProfiler package using P-value <0.05 was set as the threshold.

Based on the liquid hybridization capture-based bisulfite sequencing (LHC-BS) method, 15 Gbp raw sequence data were generated on average for each sample. More than 79% were mapped to at less one genomic position, covering ∼77% of the target regions, with an average of 57× sequencing depth per CpG and 12.5% duplication data, and the duplicated sequence reads were filtered for the subsequent analyses.

After RAGE KO, as reported in our previous study, CS-induced airway inflammation was significantly improved [4]. Meanwhile, 90 overlapping candidate genes were identified via the intersection of DMGs with DEGs (Supplementary Table S1). It is well-known that DNA methylation occurs almost in CpG islands that are primarily located in the regions of promoter and are negatively correlated with gene expression [14]. As a result, 14 genes from the 90 candidate genes, with functional methylated modification in promoters were identified (Table 1), among which, three genes, including chemokine (C–X–C motif) ligand 1 (CXCL1), Toll-like receptor 6 (TLR6) and oncostatin M (OSM) with hypomethylation in promoters were finally selected as the hub genes by PPI network analysis (Figure 2). Furthermore, GO and KEGG analyses indicated the 14 functional methylated genes, including the 3 hub genes, were significantly enriched in immune-inflammatory responses, especially mitogen-activated protein kinase (MAPK), tumor necrosis factor (TNF), TLRs, interleukin (IL)-6 and IL-17 signaling pathways (Supplementary Tables S2 and S3).

DNA methylation is closely associated with COPD susceptibility, exacerbation and lung function decline [15,16]. In this process, RAGE may play a regulatory role in DNA methylated modification in COPD, whereas the mechanisms remain unexplained. Consequently, in the present study, the 90 DMGs, regarding RAGE-mediated airway inflammation induced by CS exposure, were initially selected using a novel intersection model, and the functional DNA methylated modification in 14 targeted genes were subsequently identified, especially hypomethylation in CXCL1, TLR6 and OSM promoters, which might significantly contribute to RAGE-mediated airway inflammation in COPD via a network of signaling pathways, such as MAPK, TNF, TLRs, ILs, etc.

According to the GO and KEGG analyses, the majority of the candidate genes, especially the 14 functional methylated genes widely participated in CS-associated inflammatory-immune responses. In particular, the three hub genes were documented to play important roles in the inflammatory process in COPD. CXCL1 is a member of chemokine subfamily of CXC [17] and the increased level of CXCL1 was detected in the lungs of COPD [18]. CXCL1 served as a chemoattractant for neutrophils migrating from circulation to respiratory tracts, which contributed to neutrophilic inflammation of COPD [19]. TLR6 belongs to the Toll-like receptor family [20], which initiates innate immune responses in airway epithelial cells and triggers inflammatory responses [21]. Furthermore, TLRs activation could increase CXCL1 production in human pulmonary macrophages [22]. OSM, a member of IL-6 subfamily [23], participates in a variety of inflammatory diseases with a high level [24–26]. In COPD, the inflammatory mechanisms for OSM may be stimulating IL-6 production and IL-6 related inflammation, which is positively correlated with pulmonary function decline [23,27]. Noticeably, as suggested in our study, CXCL1, TLR6 and OSM have been documented to be targeted genes of DNA methylated modification in other diseases, such as schizophrenia [28], type 1 diabetes [29] and Richter syndrome [30].

However, some limitations in the present study should be considered. First, the sample size of mice in each group was relatively small, although the minimum requirement for biological repeat was reached. Second, the novel intersection model might be theoretically imperfect. Third, validation was thus needed in future studies.

In summary, the present study performed intersections of DMGs with DEGs in a CS-exposed mouse model and indicated RAGE could mediate functional methylated modification in multiple targeted genes, especially CXCL1, TLR6 and OSM, which might significantly contribute to airway inflammation in COPD.

The data presented in this manuscript are available from the corresponding author (Lei Chen) on reasonable request.

The authors declare that there are no competing interests associated with the manuscript.

This work was supported by the National Natural Science Foundation of China [grant number 81970040].

Ping Li: Software, Investigation, Visualization, Methodology, Writing—original draft. Tao Wang: Software, Investigation, Visualization, Methodology. Mei Chen: Software, Investigation, Visualization, Methodology. Jun Chen: Software, Investigation. Yongchun Shen: Data curation, Supervision, Validation, Writing—review and editing. Lei Chen: Conceptualization, Formal Analysis, Writing—review and editing, Funding acquisition.

COPD

chronic obstructive pulmonary disease

CS

cigarette smoke

CXCL1

chemokine (C–X–C motif) ligand 1

DEG

differentially expressed gene

DMG

differential methylated gene

DMR

differential methylated region

GO

Gene Ontology

IL

interleukin

KEGG

Kyoto Encyclopedia of Genes and Genomes

KO

knockout

MAPK

mitogen-activated protein kinase

OSM

oncostatin M

PPI

protein–protein interaction

RAGE

receptor for advanced glycation end-product

TLR

Toll-like receptor

TNF

tumor necrosis factor

WT

wildtype