Systemic inflammatory response to smoking in chronic obstructive pulmonary disease: evidence of a gender effect

Abstract

Background: Tobacco smoking is the main risk factor of chronic obstructive pulmonary disease (COPD) but not all smokers develop the disease. An abnormal pulmonary and systemic inflammatory response to smoking is thought to play a major pathogenic role in COPD, but this has never been tested directly.

Methods: We studied the systemic biomarker and leukocyte transcriptomic response (Affymetrix microarrays) to smoking exposure in 10 smokers with COPD and 10 smokers with normal spirometry. We also studied 10 healthy never smokers (not exposed to smoking) as controls. Because some aspects of COPD may differ in males and females, and the inflammatory response to other stressors (infection) might be different in man and women, we stratified participant recruitment by sex. Differentially expressed genes were validated by q-PCR. Ontology enrichment was evaluated and interaction networks inferred.

Results: Principal component analysis identified sex differences in the leukocyte transcriptomic response to acute smoking. In both genders, we identified genes that were differentially expressed in response to smoking exclusively in COPD patients (COPD related signature) or smokers with normal spirometry (Smoking related signature), their ontologies and interaction networks.

Conclusions: The use of an experimental intervention (smoking exposure) to investigate the transcriptomic response of peripheral leukocytes in COPD is a step beyond the standard case-control transcriptomic profiling carried out so far, and has facilitated the identification of novel COPD and Smoking expression related signatures which differ in males and females.

Figure 1. Median (and IQR) values of blood carboxy-hemoglobin (panel A), circulating leukocytes (panel B), C-reactive protein serum levels (panel C) and IL-6 (panel D) in COPD patients, smoker (S) and non-smoker (NS) controls, before (T0) and 30 min.

(T30) or 180 min. (T180) later. For further explanations, see text.

Figure 2

Panel A: Principal component analysis (PCA) of those genes with the largest variability of expression values across the experimental groups (COPD patients, smokers (S) and non-smokers (NS)) at baseline (T0). Because the sex effect observed, further analysis was stratified by sex (right and left columns.) Panels B1 (females) and B2 (males) present the number of differentially expressed (DE) genes from T0 to T180 in the three groups of participants studied. Red and green figures indicate up and down-regulation, respectively. Panels C1 (females) and C2 (males) show a BioVenn diagram (www.cmbi.ru.nl/cdd/biovenn) with the number of DE genes shared between COPD, S and NS controls. For further explanations, see text.

Figure 3. Expression correlation networks (Pearson r≥0.8) of COPD related genes in females (Panels A and B) and males (Panels C and D).

Red and green figures indicate up and down-regulation of expression, respectively. Arrows point genes forming functional clusters identified in the network. For further explanations, see text.

Figure 4. Expression correlation networks (Pearson r≥0.8) of Smoking genes in females.

Arrows indicate genes identified by gene ontology analysis in the network. Red and green indicate up and down-regulation, respectively. For further explanations, see text.

Figure 5. Putative interaction networks identified by Ingenuity Pathway Analysis (IPA), and their corresponding scores, in female smoker controls COPD related in females (Panel A), Smoking in females (Panel B) and COPD related in males (Panels C and D).

For further explanations, see text.