Transcriptional profiles of peripheral eosinophils in chronic obstructive pulmonary disease and asthma—An exploratory study

2.1. Patients

Initially, 25 patients were recruited, but 16 samples were excluded due to low RNA concentration and unsuccessful library generation. The final study group included five patients with asthma and four patients with COPD. All subjects were White. Among the assessed variables, only smoking status was significantly different between these two groups (Table S1). We found that among patients with inadequate RNA quality the majority (62.5%) were current smokers, whereas among patients with adequate RNA quality, current smokers constituted only 11%. This suggests that current smoking might have impacted eosinophil RNA quality and integrity. It is well known that smoking significantly impacts pattern of mRNA expression and induces cytotoxic, mutagenic and oxidative stress effects on cells. ¹⁶ Cigarette smoke contains over 7000 chemicals, including well‐characterized ones such as acrolein, formaldehyde, acetaldehyde, which could impair RNA stability and integrity. ¹⁶ Data on RNA quality, concentration and eosinophil numbers in each subject are presented in Table S2. The subjects' characteristics are presented in Table 1. At the time of the study, none of the COPD patients were receiving ICS, three of the five asthma patients were on low‐dose ICS, and none of the studied patients received biological treatment or oral steroids. The COPD patients were all but one ex‐smokers, the median smoking load of 40 (IQR 30–80) pack‐years (Table 1).

2.2. RNA‐seq analysis

The RNA‐Seq data analysis identified 26 differentially expressed genes (DEGs) between COPD and asthma (according to adjusted p‐value). In total, 6 genes were upregulated and 20 were downregulated in eosinophils of COPD patients compared to asthma (Table 2).

The most strongly upregulated genes in COPD (2.40‐fold change and 2.35‐fold change, respectively) were CCL3L1 and CCL4L2, which encode chemokines. Dot plots showing the transcripts per million (TPM) values of CCL3L1 and CCL4L2 in COPD and asthma patients are presented in Figure 1. CCL3L1 and CCL4L2 expression is largely driven by copy number variation which is dependent on race. ¹⁷ However, in the present study we found that all COPD subjects had increased expression levels of both genes compared to asthma. Interestingly, we found that CCL3L1, but not CCL3, was upregulated in eosinophils of COPD patients. Both CCL3L1 and CCL3 are isoforms of macrophage inflammatory protein 1α (MIP‐1α), with the former exhibiting a sixfold greater affinity to the receptor CCR5. ¹⁷ MIP‐1α‐CCR5 binding is linked to the processes associated with tight junction injury in the airway epithelium in COPD. ¹⁸ CCR5 also serves as a receptor for MIP‐1β encoded by i.a. CCL4L2 which in our study was found to be upregulated in COPD patients. Both MIP‐1α and MIP‐1β are monocyte and macrophage chemoattractants. It has been shown that MIP‐1α plays a negligible role as an eosinophil chemoattractant, ¹⁹ unlike MIP‐1β. ²⁰ Increased expression of MIP‐1α in COPD patients has been linked to cigarette smoke exposure. ¹⁸ However, the MIP‐1β isoform encoded by CCL4L2 has an unknown affinity to CCR5 and its biological function has not been studied. ²¹ The upgraded mRNA expression of CCL3L1 in COPD eosinophils found in our study may suggest that in COPD eosinophils attract macrophages into the lungs rather than drive local eosinophilic inflammation. This finding underscores the different function of eosinophils in asthma and COPD pathobiology and might explain different ICS‐sensitivity of patients with COPD and asthma. Keeping in mind a small sample size in this analysis, these results should be interpreted carefully and confirmed in larger cohorts.

Other genes upregulated in COPD eosinophils in our study include protease inhibitor (SERPINB2), serine protease (PRSS21) and G‐protein coupled receptor (GPR82). Of these, GPR82 has previously been identified as an eosinophil signature gene, ²² but in fact, its role in the pathogenesis of eosinophilic inflammation remains unknown. SERPINB2 upregulation in COPD patients found in our study is surprising. SERPINB2 is typically induced by IL‐13 and elevated in eosinophilic asthma patients. ²³ , ²⁴ In COPD however, the SERPINB2 expression was found not to be different between eosinophilic and non‐eosinophilic COPD patients. ²⁵ Although SERPINB2 is known to be downregulated with corticosteroid treatment, ²⁴ additional analysis did not reveal significant differences in expression level between asthmatics receiving ICS and those without ICS (p = 0.39). However, these results must be interpreted with caution due to the low sample size included in our study.

The two most strongly upregulated genes (over 3‐fold) in asthmatic eosinophils revealed in our analysis, JUN and IFITM3, have already been reported to be involved in asthma pathogenesis. IFITM3 belongs to the IFITM family which encodes homologous proteins localized in the plasma and endolysosomal membranes. It has been shown that IFITM deficiency reduces allergic inflammation in a murine model of asthma. ²⁶ Moreover, increased expression of JUN in PBMC, the transcription factor known to upregulate other inflammatory genes has been previously observed in patients with severe asthma compared to mild‐to‐moderate asthma or healthy controls. ²⁷ The protein encoded by JUN is involved in MAPK signalling cascades ²⁸ and might contribute to severe asthma pathogenesis by causing corticosteroid resistance. ²⁹ Interestingly, we found upregulation of DUSP1 in asthma patients which might suggest an expectation of downregulated JUN expression, as DUSP1 attenuates MAPK signalling cascade. Our study implies that MAPK signalling in eosinophils is complex and dependent on many signal inducers. Moreover, DUSP1 upregulation is known to be induced by ICS treatment. ²⁸ Although DUSP1 expression trended higher in ICS‐users (n = 3) compared to ICS‐non‐users (n = 6), it did not achieve statistical significance (p = 0.09), probably due to the small sample size. Other genes downregulated in eosinophils of COPD compared to asthma patients in our study encoded membrane proteins (GNG7, CNTNAP3, SCARF1) and transcription factors (ZNF107, BCL6, JUNB, ZNF467). A volcano plot was used to visually present the detailed analysis of differentially expressed genes between blood eosinophils from COPD and asthma patients (Figure S1).

2.3. Enrichment and pathway analyses

The Gene Ontology (GO) enrichment analysis of the downregulated genes in COPD versus asthma eosinophils revealed association with negative regulation of transcription by RNA polymerase II (GO:0000122), positive regulation of RNA metabolic process (GO:0051254) and cell differentiation (GO:0030154) (Figure 2A). The molecular function terms of the downregulated genes in COPD versus asthma eosinophils were associated with DNA binding (GO:0003677), including: DNA‐binding transcription factor activity, RNA polymerase II‐specific (GO:0000981), RNA polymerase II cis‐regulatory region sequence‐specific DNA binding (GO:0000978) and RNA polymerase II transcription regulatory region sequence‐specific DNA binding (GO:0000977) (Figure 2B). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis did not reveal any significantly overrepresented pathways. However, REACTOME pathway analysis indicated five overrepresented pathways (three for the upregulated and two for the downregulated genes) in COPD compared to asthma eosinophils (Figure 2C). Significant associations were found with pathways involving Rhodopsin‐like receptors, G protein‐coupled receptors (GPCR) ligand binding and metabolism of proteins (for the upregulated genes) and signalling of IL‐6 family, IL‐4 and IL‐13 (for the downregulated genes) in COPD eosinophils. IL‐4 and IL‐13, alongside IL‐5 are the key cytokines in type 2 (Th2) inflammation. Our results confirm that Th2 inflammation is an important pathway in asthma pathophysiology but to a lesser extent in COPD. Similar conclusions were made by other authors suggesting that the differences in the transcriptional profiles in COPD and asthma patients might reflect different mechanisms driving eosinophilic inflammation in COPD. ²⁰ On the other hand, the promising results of a recent dupilumab trial among severe eosinophilic COPD patients ³⁰ suggests greater involvement of IL‐4/IL‐13 genes and therefore of the Th2 pathway in COPD pathobiology.