Long Noncoding RNA Interleukin 6 Antisense RNA 1 Promotes Inflammatory Effects in Lung Macrophages via Exosomes Through the S100A9/TLR4 Pathway in Chronic Obstructive Pulmonary Disease Progression

Abstract

This study investigates the role of interleukin 6 antisense RNA 1 (IL6-AS1), a highly expressed long noncoding RNA (lncRNA), in chronic obstructive pulmonary disease (COPD). An adeno-associated virus (AAV) was used to induce the expression of IL6-AS1 in mice, and they were exposed to cigarette smoke to establish a COPD model. IL6-AS1-overexpressing mice exposed to cigarette smoke demonstrated exacerbated COPD-like pathologies. Integrated with single-cell RNA sequencing analysis of COPD patients and pulmonary fibroblast-macrophage coculture system, our findings indicate that the upregulation of IL6-AS1 in fibroblasts enhances the interaction between the S100A9 protein and the AGER and TLR4 receptors on lung macrophages, thereby exacerbating pulmonary inflammation. The molecular mechanism likely involves exosome-mediated secretion, with IL6-AS1 binding to S100A9 protein. These findings suggest that IL6-AS1 may facilitate crosstalk between fibroblasts and macrophages, contributing to increased pulmonary inflammation, an effect that can be blocked by paquinimod. Mendelian randomization analysis further suggests a potential shared causal variant between IL6-AS1 and COPD risk. Taken together, this investigation provides valuable insights into the function of IL6-AS1 and its potential implications for the pathogenesis and therapeutic strategies in COPD.

Keywords: Mendelian randomization; S100A9; chronic obstructive pulmonary disease; lncRNA interleukin 6 antisense RNA 1; macrophages.

FIGURE 1

LncRNA IL6‐AS1 expression is increased in COPD lungs and negatively correlates with lung function in COPD patients. (A and B) Expression of IL6‐AS1 in the lungs of healthy smokers and COPD patients from the GSE76925 database (n = 37 smokers and n = 110 COPD patients, including n = 40 GOLD3 patients, n = 70 GOLD4 patients). (C) Correlation analysis of IL6‐AS1 expression with various parameters, including FEV₁% predicted (FEV₁%pre), FEV₁/FVC ratio, pack years, Perc15, % low attenuation area at −950 HU (%LAA950), and Pi10, based on GSE76925. (D and E) Expression of IL6‐AS1 in the lungs of control and COPD patients (D), and the correlation of IL6‐AS1 with FEV₁%pred and FEV1 (E) from the GSE8581 database (n = 18 non‐COPD patients and n = 21 COPD patients). (F and G) Expression of IL6‐AS1 in the lungs of control and COPD patients (F), including GOLD1 patients and GOLD2 patients (G) from the GSE103174 database (n = 16 non‐COPD patients and n = 34 COPD patients, including n = 7 GOLD1 patients, n = 27 GOLD2 patients). (H) Heatmap illustrating the most significantly enriched gene lists in the lungs of populations with high and low IL6‐AS1 expression, as determined by GSEA of the GO molecular function (MF) set using the GSE76925 dataset. The nominal enrichment score (ES) and p value were generated by the GSEA software, which revealed the enrichment score's significance relative to a null distribution. (I and J) IL6‐AS1 expression was analyzed in bronchial brushings of healthy smokers (n = 107) and COPD patients (n = 86), as well as in ex‐smokers (n = 116) and current smokers (n = 79)(I). Additionally, IL6‐AS1 expression was evaluated among COPD patients classified as GOLD1 (n = 24), GOLD2 (n = 48), and GOLD3/4 (n = 14), using data from GSE37147 (J). (K) Correlation analysis was conducted to investigate the relationship between IL6‐AS1 expression and clinical parameters, including FEV₁%pre, FEV₁/FVC ratio, pack years, and changes in FEV1, utilizing data from GSE37147. (L) IL6‐AS1 expression in bronchial biopsies was assessed in non‐COPD and COPD patients, as well as across subgroups including healthy smokers (n = 20), chronic bronchitis patients (GOLD0, n = 94), GOLD1/2 patients (n = 58), and GOLD3/4 patients (n = 25), using data from GSE162635. (M) Expression of IL6‐AS1 in transbronchial biopsies was evaluated during follow‐up, with patients categorized into groups based on pulmonary function: improved (better, n = 9), stable (stable, n = 100), and deteriorated (decline, n = 8). Data are presented as mean ± SD. p Values in the charts were calculated using two‐tailed Mann–Whitney tests (A, B, D, F, G, I, J, and L) and two‐tailed Pearson correlation tests (C, E, and K).

FIGURE 2

The AAV‐constructed IL6‐AS1 mice demonstrated exacerbated COPD‐like pathological changes following cigarette smoke exposure. (A) Schematic representation of the establishment of IL6‐AS1 mice, with lung function assessments and tissue sampling conducted after 6 months of cigarette smoke exposure. (B) Evaluation of lung FEV₅₀/FVC, Cchord, and Cydn in four distinct groups of mice. FEV50/FVC represents forced expiratory volume at 50 ms (FEV50) relative to forced vital capacity (FVC); Cydn indicates lung dynamic compliance; Cchord denotes lung compliance. (n = 7 for Smoke+ IL6‐AS1, n = 5 for other three groups). (C–E) Histological sections of lung tissues from the four experimental groups subjected to HE staining (C); histological sections of lung tissues from the four experimental groups assessed through HE stains, Masson staining, IHC staining for α‐SMA, PAS staining, and cell smears for inflammatory cells in BALF (D). The bar chart presents the relevant pathological statistics (E). (n = 7 for Smoke+ IL6‐AS1, n = 5 for other three groups). Data are presented as mean ± SD. p Values in charts were determined using one‐way ANOVA Bonferroni's multiple comparisons test (B, C, and D). (F and G) GSEA results for upregulated DEGs in IL6‐AS1 mice compared with wild‐type mice in the Smoke group (G) and air group (F), spanning Hallmark, biological processes (BP), and Reactome pathways. p Values are represented as log10 values. (H) qRT‐PCR analysis to assess the expression of inflammation‐associated genes (IL‐6, MCP‐1, TNF‐α) across the four groups of mice. (n = 7 for Smoke+ IL6‐AS1, n = 5 for other three groups). (I–K) Evaluations of the secretion levels of IL‐6, MCP‐1, and TNF‐α within lung tissue homogenates (I), serum (J), and BALF (K) were conducted using a CBA array. (n = 7 for Smoke+ IL6‐AS1, n = 5 for other three groups). Data are presented as mean ± SD. p Values in charts were determined by one‐way ANOVA Bonferroni's multiple comparisons test (B, E, H, I, J, and K).

FIGURE 3

IL6‐AS1 exhibits specific binding to S100A9 in both in vivo and in vitro settings. (A) Diagrammatic representation illustrates RNA‐seq conducted on HFL1 cells following overexpression of IL6‐AS1. (B) Metascape enrichment results for differentially upregulated genes in IL6‐AS1 mice in the air group, IL6‐AS1 mice in the smoke group, and genes upregulated following IL6‐AS1 overexpression in HFL cells. (C) qRT‐PCR for the expression of inflammation‐associated genes (IL‐6, IL‐8, ICAM‐1, CCL‐2/7, and CXCL‐3/10) in HFL1 cells after the IL6‐AS1 overexpression. (n = 4 biological replicates). (D) ChIRP assays were conducted using either an unspecific lacZ probe or probes designed for IL6‐AS1 in human lung tissue, followed by silver staining and mass spectrometric sequencing. (E) A PPI network was constructed by combining the distinct binding proteins identified in the mass spectrometry results of the target group with the differentially expressed COPD‐related genes identified in the RNA‐seq. (F) Western blot analysis of the proteins from the proteomics screen after ChIRP assays shows the specific interaction of IL6‐AS1 with S100A9 protein in lung homogenates. (G) RIP‐qPCR analysis with anti‐S100A9 antibody in wild‐type (NC) and IL6‐AS1 mice. Anti‐IgG antibody acted as a negative control (n = 5). (H) RNA immunoprecipitation followed by quantitative PCR (RIP‐qPCR) analysis using anti‐S100A9 antibody in HFL1 cells. Anti‐IgG antibody and U1 served as negative controls. (n = 3 biological replicates). (I) RIP‐qPCR analysis using anti‐S100A9 antibody in HFL1 cells after stimulation with recombinant S100A9 protein (rp‐S100A9). Anti‐IgG antibody served as a negative control. (n = 3 biological replicates). (J) Lung tissue sections from the four groups of mice were subjected to IHC staining for S100A9. (n = 7 for Smoke+ IL6‐AS1, n = 5 for other three groups). (K) The concentration of S100A9 protein in the lung tissue homogenates and serum of the four groups of mice was determined by ELISA. (n = 7 for Smoke+ IL6‐AS1, n = 5 for other three groups). (L) Western blot illustration reveals the presence of phosphorylated p65 and p38 following stimulation with rp‐S100A9 for 0, 1, 6, 12, and 24 h (n = 3 biological replicates). (M) qRT‐PCR evaluation of the expression of CCL‐2 and IL‐6 in IL6‐AS1‐augmented HFL1 cells following stimulation with rp‐S100A9 at the 1 and 24‐h time points. (n = 4 biological replicates). Data are presented as mean ± SD. p Values in charts were determined by one‐way ANOVA Bonferroni's multiple comparisons test (G, H, and I), one‐way ANOVA Tukey's multiple comparisons test (C, L, and M), and unpaired two‐tailed Student's t‐test (J and K).

FIGURE 4

Overexpression of IL6‐AS1 and stimulation with rp‐S100A9 enhance the inflammatory response in lung fibroblast cells. (A) FISH staining of IL6‐AS1 in human lung tissue sections. Red represents IL6‐AS1, and blue corresponds to DAPI. (B) Umap plot displaying single‐cell transcriptomes of lung tissues from healthy controls and COPD patients from the GSE136861 dataset. (C) Dotplot of the expression of S100A9, AGER and TLR4 in different cell types from GSE136861. (D) Enrichment results for Reactome, KEGG, and GO pathways in DEGs between IL6‐AS1 positive and negative macrophages. (E) Bubble chart illustrating the enrichment of related inflammatory genes (including CCL‐2/18 23, IL‐6, CXCL‐2/9, and TNF) in the distribution and expression profiles between IL6‐AS1 ⁺ and IL6‐AS1⁻ macrophages, as identified in single‐cell transcriptomes (GSE136861). (F) Bubble chart illustrating the enrichment of related inflammatory genes (including CCL‐2/18/23, IL‐6, and CXCL8) in the distribution and expression profiles between S100A9^‐IL6‐AS1⁻ , S100A9⁺IL6‐AS1⁻ , and S100A9⁺IL6‐AS1⁺ macrophages from GSE136861. (G) Secretory‐type intercellular interactions among various cell types within IL6‐AS1⁺ and IL6‐AS1⁻ macrophage clusters were analyzed, including the degree of enrichment of incoming and outgoing signaling pathways, as identified in single‐cell transcriptomes. Blue frames indicate significantly enriched pathways with distinct differences. (H) Network graphs comparing pulmonary cell interactions in IL6‐AS1⁻ (left) and IL6‐AS1⁺ (right) macrophages as ligand cells. Each vertex represents a cellular subpopulation; edges signify ligand–receptor interactions. The thickness of the edges quantifies the cumulative expression of ligand–receptor genes, while the size of each vertex reflects Kleinberg centrality, indicating the cell's role in signaling. Cellular subpopulations are differentiated by color and number. (I) Comparison of IL‐6 (IL‐6R+IL‐6ST) signaling interactions between IL6‐AS1 ⁺ and IL6‐AS1⁻ macrophages and other cell subpopulations.

FIGURE 5

IL6‐AS1 on fibroblasts can facilitate the binding of S100A9 protein to TLR4 and AGER receptors on macrophages. (A) Schematic representation of the coculture system involving macrophages and fibroblasts. (B) qRT‐PCR assay determining CD86 and CD11b expression in THP‐1 cells following coculture, after stimulation with PMA solely or in conjunction with rp‐S100A9 (n = 4 biological replicates). (C) qRT‐PCR assay measuring the expression of IL6‐AS1 in HFL1 and Thp‐1 cells following coculture. (n = 4 biological replicates). (D) qRT‐PCR assay analyzing the expression of inflammation‐associated genes (IL‐6, IL‐8, CCL‐2/7, and CXCL‐3/10) in THP‐1 cells after coculture (n = 4 biological replicates). (E) qRT‐PCR assay of IL‐10, CD206, CD86, and TNF‐α expression in Thp‐1 cells following coculture (n = 4 biological replicates). (F) Co‐IP analysis of Thp‐1 cocultured with IL6‐AS1‐overexpressed HFL1 using anti‐S100A9 antibody. Western blot was used to verify the co‐IP results with anti‐S100A9, anti‐TLR4, and anti‐AGER antibodies (n = 3 biological replicates). (G and H) IF double staining was performed using anti‐S100A9/anti‐TLR4 antibody (G) or anti‐S100A9/anti‐AGER antibody (H) in mouse lung tissue sections extracted from 4 distinct groups of mice. (n = 4). (I) Western blot revealing the presence of TLR4 and AGER in four distinct mouse groupings, utilizing Actin as a control reference. (n = 5, n = 7 [Smoke+IL6‐AS1]). Data shown mean ± SD. p Values shown in charts are determined by multiple two‐tailed Students’ t‐tests (B, C, D, E, and F) and one‐way ANOVA Bonferroni's multiple comparisons test (G and I).

FIGURE 6

Coculturing IL6‐AS1 overexpressed fibroblasts with macrophages can enhance the secretion of inflammation‐related factors in both cell types. (A and B) Western blot illustrates the phosphorylation levels of p65 and p38 in HFL1 (A) and THP‐1 (B) after 24 h of cocultivation and stimulation with rp‐S100A9. (n = 4 biological replicates). (C) IF assay detection of phos‐p38 and phos‐p65 double‐positive immune cells in the lungs of the four mouse groups. Red: phos‐p65; green: phos‐p38; blue: DAPI. (n = 4 in each group). (D and E) ELISA was performed to detect the secretion of CCL2 and IL‐6 in HFL1 cells (D) and CCL2, IL‐6, and TNF‐α in THP‐1 cells (E) following cocultivation and stimulation with rp‐S100A9. (n = 4 biological replicates. (F) Schematic delineation provides insight into the RNA secondary structure of IL6‐AS1 and the prospective binding sites interacting with the S100A9 protein. Red lines indicated potential binding sites. (G) RIP‐qPCR analysis with an anti‐S100A9 antibody after transfection with either wild‐type or truncated IL6‐AS1 (49–333aa, 334–568aa, 692–1268aa) in HFL1 cells offers evidence of binding dynamics (n = 4 biological replicates). (H and I) qRT‐PCR assay of the expression of inflammation‐related genes (IL‐6, IL‐8, CCL2, CCL7, CXCL3, CXCL10, and TNF) in HFL1 cells and Thp‐1 cells following coculture, posttransfection with either wild‐type vector or truncation vector of IL6‐AS1. (n = 4 biological replicates). (J) The secretion of IL‐6 and TNF‐α in THP‐1 cells following coculture after transfection with wily‐type vector or truncation vector of IL6‐AS1 were detected by ELISA assay. (n = 4 biological replicates). Data are presented as mean ± SD. p Values in charts were determined by one‐way ANOVA Tukey's multiple comparisons test (A, B, D, E, F, G, and J), one‐way ANOVA Bonferroni's multiple comparisons test (C), and multiple two‐tailed Student's t‐test (H and I).

FIGURE 7

Paquinimod delays the COPD‐like changes induced by smoke exposure in IL6‐AS1 mice. (A) Schematic illustration delineating the construction of IL6‐AS1 mice and the administration of paquinimod. (B) Lung FEV₅₀/FVC, Cchord, Cydn, FRC, and RI were conducted across the four groups of mice (n = 7). (C) Samples of lung tissue from the four distinct groups of mice analyzed using HE staining (n = 7). (D) Flow cytometric cell sorting to quantify the populations of macrophages and eosinophils (n = 7). (E) Tissue sections from the lungs of the four groups of mice underwent hematoxylin and eosin staining, Masson's trichrome staining, Immunohistochemical staining for α‐SMA, and PAS staining. (n = 7). (F) Sections of lung tissues from the four groups of mice were subjected to IHC staining for S100A9. (n = 7). (G) Western blot showing the phosphorylation level of p65 and the phosphorylation level of p38 in 4 groups of mice, with Actin as the control. (n = 7). (H) qRT‐PCR analysis of IL‐6 and TNF‐α in 3 groups of mice. (n = 7). (I) The secretion of IL‐6 and TNF‐α in BALF detected by ELISA. (n = 7). Data are presented as mean ± SD. p Values in charts were determined by one‐way ANOVA Bonferroni's multiple comparisons test (B, C, D, E, F, G, H, and I).

FIGURE 8

MR analysis indicates that the expression of IL6‐AS1 in lung and whole blood was associated with COPD progression. (A) Schematic diagram outlines the steps in the MR analysis between IL6‐AS1 and COPD outcomes (including doctor‐diagnosed COPD, FEV₁, FVC, and FEV₁/FVC). The analysis is based on three assumptions: (1) IVs are associated with the exposure, (2) IVs are not influenced by confounders, and (3) IVs affect the outcome exclusively through the exposure, excluding alternative pathways. (B) Forest plot of SMR results of IL6‐AS1 in lung and whole blood with doctor diagnosis COPD, FEV₁, FVC and FEV₁/FVC. (C and D) Forest plots display two‐sample MR analyses of IL6‐AS1 expression in the lung (C) and whole blood (D) with COPD outcomes after adjusting for the effects of SNPs on other cell types. Effect sizes (beta, 95% CI) are shown as standard deviation changes in COPD outcomes per standard deviation increase in IL6‐AS1 expression. The points represent effect size estimates, and the whiskers indicate 95% confidence intervals (CIs). (E) Correlation analysis of allelic and replacement bases at rs2069832 with IL6‐AS1 expression in the lung and whole blood, based on data from the GTEx database. (F) Chromatin locations of rs1474348 and rs2069832, along with histone modification levels (H3K27ac, H3K4me3, and H3K4me1) at the IL6‐AS1 gene locus, referenced through the UCSC database. (G) Correlation analysis between IL6‐AS1 and transcription factors ELF1, EHF, FEV, RFX5, SPI1, and NR3C1 in healthy controls and COPD patients, based on the GSE76925 dataset. (H) The schematic representation suggests that fibroblast‐derived IL6‐AS1, transported via exosomes, interacts with the S100A9 protein, stabilizing it and promoting its binding to the TLR4/AGER receptor on macrophages. This interaction activates the TLR4/AGER‐mediated NF‐κB pathway, increasing the expression of downstream inflammatory factors in macrophages. Data are presented as mean ± SD. p Values in charts were determined by one‐way ANOVA Bonferroni's multiple comparisons test (E) and Pearson correlation two‐tailed (G).