COPD-associated proteins as biomarkers and drug targets: Insights from genetic and proteomic analyses

Figure 1.. Flowchart illustrates the screening process for plasma proteins associated with COPD. The study utilizes a 2-sample MR analysis on data from Decode/UKB-PPP to investigate the causal relationship between circulating proteins and COPD. Co-localization analysis was performed to confirm shared genetic variants affecting protein levels and COPD susceptibility. This was followed by pathway and tissue-specific differential expression analyses to uncover the biological roles of identified proteins. A PPI and drug target analysis were conducted to prioritize proteins as potential therapeutic targets. Finally, a systematic MR approach examined the influence of lifestyle factors on COPD-related proteins, suggesting possible lifestyle-based intervention strategies. COPD = chronic obstructive pulmonary disease, MR = Mendelian randomization, PPI = protein–protein interaction, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project

Figure 2.. Positive causal relationships between circulating proteins and COPD after FDR correction. (A) Volcano plot showing significant protein associations with COPD identified in the Decode project after FDR correction, with upregulated proteins in red and downregulated proteins in blue. (B) Volcano plot for significant associations from the UKB-PPP cohort, following FDR correction, displaying the direction and magnitude of effects. (C) Venn diagram indicating the overlap of proteins associated with COPD across Decode and UKB-PPP datasets, showing 4 shared protein markers, emphasizing consistency in results across cohorts. COPD = chronic obstructive pulmonary disease, FDR = false discovery rate, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project.

Figure 3.. Analysis of causal relationships between circulating proteins and COPD after FDR correction. (A) Forest plot depicting odds ratios and confidence intervals (CIs) for proteins from the Decode dataset with significant associations to COPD. (B) Forest plot for the UKB-PPP dataset, illustrating significant protein associations with COPD, supporting the Decode findings and reinforcing the validity of these proteins as potential COPD biomarkers. CI = confidence interval, COPD = chronic obstructive pulmonary disease, FDR = false discovery rate, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project.

Figure 4.. Co-localization analysis of circulating proteins and COPD with strong evidence of causal association (PPH4 > 0.7). (A) KLC1 in the Decode dataset (PPH4 = 0.848), demonstrating a significant association with COPD. (B) MMP12 in the Decode dataset (PPH4 = 0.802), showing strong co-localization with COPD susceptibility loci. (C) Nephronectin (NPNT) in the Decode dataset (PPH4 = 0.744), indicating a potential causal role in COPD. (D) SNX1 in the Decode dataset (PPH4 = 0.908), with high co-localization probability for COPD. (E) APOE in the UKB-PPP dataset (PPH4 = 0.885), supporting its involvement in COPD. (F) DAG1 in the UKB-PPP dataset (PPH4 = 0.888), providing further evidence of COPD association. (G) MMP12 in the UKB-PPP dataset (PPH4 = 0.885), reinforcing its causal link to COPD. (H) TIE1 in the UKB-PPP dataset (PPH4 = 0.892), highlighting its potential relevance to COPD pathology. APOE = apolipoprotein E, COPD = chronic obstructive pulmonary disease, MMP12 = matrix metalloproteinase-12, NPNT = NPNT = nephronectin, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project.

Figure 5.. Differential expression analysis of key genome-wide significant loci and protein-coding genes identified in Decode and UKB-PPP. This figure displays the expression profiles of significant genes across 54 GTEx v8 tissues. Red bars indicate tissues where the gene set exhibits differential expression, potentially highlighting tissue-specific relevance in COPD pathology. COPD = chronic obstructive pulmonary disease, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project.

Figure 6.. Heatmap of gene expression for significant proteins identified in Decode and UKB-PPP datasets. This figure provides a comparative view of expression levels across various tissues, illustrating the differential expression patterns of genes such as APOE, CTSH, DAG1, IL17RD, KLC1, NPNT, SERPINA1, SNX1, and TIE1, which may be critical in lung-related health and COPD progression. APOE = apolipoprotein E, COPD = chronic obstructive pulmonary disease, NPNT = nephronectin, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project.

Figure 7.. Pathway enrichment analysis for unique proteins identified in Decode and UKB-PPP. This figure emphasizes the involvement of these genes in lung function-associated pathways, including those linked to respiratory mechanics, immune response, and inflammation regulation, underscoring their potential as therapeutic targets in COPD and other respiratory diseases. COPD = chronic obstructive pulmonary disease, UKB-PPP = UK Biobank Pharmaceutical Proteomics Project.

Figure 8.. PPI network analysis of existing drug targets and newly predicted protein targets. This network illustrates the interactions between selected drug targets and COPD-associated proteins identified through our analysis. Notable connections include the N-acetylcysteine target GRIN2B and the dupilumab targets IL13 and IL4, which exhibit direct interactions with predicted protein targets, suggesting their potential as therapeutic intervention points. In contrast, targets for other drugs displayed minimal to no direct interactions with identified candidate proteins, underscoring the specificity of certain proteins, such as GRIN2B, IL13, and IL4, in COPD-related pathways. COPD = chronic obstructive pulmonary disease, PPI = protein–protein interaction.

Figure 9.. The impact of lifestyle factors on predicted COPD-related protein targets. This figure presents the Mendelian randomization analysis examining causal relationships between various lifestyle factors and specific COPD-related protein loci. For the UKB-PPP dataset, the analysis was restricted to chromosome-classified data, focusing on chromosomes with relevant protein genes. In the Decode dataset, a comprehensive analysis was conducted across all data. Significant findings include causal associations between oily fish intake and TIE1 on chromosome 1; battered fish intake and never smoking with MMP12 on chromosome 11; and cooked vegetable intake with APOE on chromosome 19. Additionally, crisp intake was linked to multiple proteins, including KLC1, MMP12, NPNT, PLG, and SNX1. Other associations were observed between cheese intake, dried fruit intake, water intake, and proteins such as MMP12, NPNT, and PLG. These results highlight the potential of dietary and lifestyle interventions in modulating COPD-related proteins, providing insights into managing COPD risk factors. APOE = apolipoprotein E, COPD = chronic obstructive pulmonary disease, MMP12 = matrix metalloproteinase-12, NPNT =…