Glycogen synthase kinase-3β modulation of glucocorticoid responsiveness in COPD

Abstract

In chronic obstructive pulmonary disease (COPD), oxidative stress regulates the inflammatory response of bronchial epithelium and monocytes/macrophages through kinase modulation and has been linked to glucocorticoid unresponsiveness. Glycogen synthase-3β (GSK3β) inactivation plays a key role in mediating signaling processes upon reactive oxygen species (ROS) exposure. We hypothesized that GSK3β is involved in oxidative stress-induced glucocorticoid insensitivity in COPD. We studied levels of phospho-GSK3β-Ser9, a marker of GSK3β inactivation, in lung sections and cultured monocytes and bronchial epithelial cells of COPD patients, control smokers, and nonsmokers. We observed increased levels of phospho-GSK3β-Ser9 in monocytes, alveolar macrophages, and bronchial epithelial cells from COPD patients and control smokers compared with nonsmokers. Pharmacological inactivation of GSK3β did not affect CXCL8 or granulocyte-macrophage colony-stimulating factor (GM-CSF) expression but resulted in glucocorticoid insensitivity in vitro in both inflammatory and structural cells. Further mechanistic studies in monocyte and bronchial epithelial cell lines showed that GSK3β inactivation is a common effector of oxidative stress-induced activation of the MEK/ERK-1/2 and phosphatidylinositol 3-kinase/Akt signaling pathways leading to glucocorticoid unresponsiveness. In primary monocytes, the mechanism involved modulation of histone deacetylase 2 (HDAC2) activity in response to GSK3β inactivation. In conclusion, we demonstrate for the first time that ROS-induced glucocorticoid unresponsiveness in COPD is mediated through GSK3β, acting as a ROS-sensitive hub.

Keywords: COPD; epithelial cells; inflammatory responses; monocytes; oxidative stress.

Fig. 1.

Phospho (p)-glycogen synthase 3β (GSK3β)-Ser9 levels are elevated in peripheral lung alveolar macrophages, peripheral blood monocytes, and primary bronchial epithelial cells (PBECs) from chronic obstructive pulmonary disease (COPD) patients. Representative images (A) and percentage of macrophages (B) positively stained for p-GSK3β-Ser9 in peripheral lung sections from nonsmokers (n = 14), smokers (n = 19), mild-moderate COPD (n = 21), and severe COPD patients (n = 7). Ratio of p-GSK3β/total GSK3β in primary monocytes (n = 6–10) with representative blots (C) and p-GSK3β/GAPDH (D) and total GSK3β/GAPDH (E) with representative blots in PBECs from nonsmokers, smokers, and Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage IV COPD patients (n = 7–8). F: representative images of staining of p-GSK3β in large airway epithelial cells in peripheral lung sections from nonsmokers, smokers, and COPD patients are shown beneath along with an isotype-stained section as a control. G: total GSK3β staining in macrophages of COPD subjects and controls. H: percentage of GSK3β-positive macrophages. P values are indicated and as tested by Kruskal-Wallis ANOVA.

Fig. 2.

GSK3β inhibition attenuates the anti-inflammatory action of glucocorticoids in monocytes. CT99021 does not affect baseline or LPS-induced granulocyte-macrophage colony-stimulating factor (GM-CSF) or CXCL8 secretion in primary monocytes from non-COPD individuals. GM-CSF (A) and CXCL8 (B) levels were measured in supernatants of primary monocytes upon pretreatment with CT99021 for 30 min followed by 24 h of LPS. Treatment of monocytes isolated from healthy subjects with CT99021 inhibits dexamethasone (Dex)-induced suppression of LPS-stimulated (C) GM-CSF and (D) CXCL8 release (means ± SE; n = 6–7). P values are indicated.

Fig. 3.

Oxidative stress induced inactivation of GSK3β is mediated via phosphatidylinositol 3-kinase (PI3K)/Akt in primary monocytes. A: primary monocytes from non—–COPD individuals were exposed to H₂O_2, which time-dependent increase in PI3K, ERK1/2, and GSK3β phosphorylation as detected by Western blotting. Representative blots of 4 independent experiments are shown. B: GSK3β phosphorylation is PI3K/Akt and ERK1/2-dependent in monocytes as indicated by pretreatment of the cells with the MEK/ERK-1/2 inhibitor U0126, the Akt inhibitor MK-2206, and the PI3Kδ inhibitor IC87114. Densitometry was performed and p-GSK3β-Ser9 levels are expressed as ratio of total GSK3β (means ± SE; n = 5).

Fig. 4.

GSK3β modulates dexamethasone function in MonoMac6 cells. A: GSK3β levels are reduced by GSK3β siRNA after 24 h of transfection. Densitometry was performed and GSK3β levels are expressed as ratio of β-actin as loading control. B: GSK3β siRNA knockdown (24 h) inhibits the concentration-dependent suppression of LPS-induced CXCL8 (left) and GM-CSF (right) release by dexamethasone (means ± SE; n = 6). C: LPS-induced GM-CSF release in mock transfected monocytes and cells transfected with the positive control pcDNA3.1 (pg/ml; means ± SE; n = 4), and effect of overexpression of the inactive mutant GSK3βK85A on dexamethasone suppression of LPS-induced GM-CSF release (%; means ± SE; n = 4). D: overexpression of the constitutively active GSK3βS9A mutant restores H₂O₂-induced dexamethasone unresponsiveness of GM-CSF release (means ± SE; n = 4).

Fig. 5.

GSK3β-regulated glucocorticoid function is histone deacetylase 2 (HDAC2)-dependent in monocytes. A: treatment of primary monocytes with CT99021 inhibits the enzymatic activity of HDAC2 (means ± SE; n = 4). B: CT99021 treatment induces p-HDAC2-Ser394 in primary monocytes. Densitometry was performed and p-HDAC2 levels are expressed as ratio of total HDAC (means ± SE; n = 4). IP, immunoprecipitation; WB, Western blot.

Fig. 6.

GSK3β inhibition attenuates the anti-inflammatory action of glucocorticoids in primary bronchial epithelial cells (PBECs). CT99021 does not affect GM-CSF or CXCL8 secretion in PBECs. GM-CSF (A) and CXCL8 (B) levels were measured in cell-free supernatants of PBECs from non-COPD individuals upon pretreatment with CT99021 for 30 min followed by 24 h TNF-α stimulation of PBECs (means ± SE; n = 4–6). Pretreatment with CT99021 (30 min) reverses budesonide (Bud)-induced suppression of TNF-α-stimulated GM-CSF (C) and CXCL8 (D) release in PBECs from non-COPD individuals (means ± SE; n = 8).

Fig. 7.

Cigarette smoke extract (CSE)-induced oxidative stress induces budesonide unresponsiveness in TNF-α-stimulated 16HBE cells. A–C: pretreatment with CSE reduces budesonide-induced suppression of TNF-α-stimulated CXCL8 release and pretreatment with N-acetyl-cysteine (NAC; 30 min) restores CSE-induced budesonide unresponsiveness (means ± SE; n = 5). Absolute values (A), values related to the TNF-α-induced control (B) and percent inhibition by budesonide (C) are shown (D) CSE induces PI3K and GSK3β phosphorylation cells, as indicated by Western blotting. Representatives of 3 independent experiments are shown. The CSE-induced increase in GSK3β phosphorylation is abrogated by NAC pretreatment. Densitometry was performed and p-GSK3β-Ser9 levels are expressed as ratio of GAPDH as loading control (means ± SE; n = 6).