Lung-targeted overexpression of the NF-κB member RelB inhibits cigarette smoke-induced inflammation

Abstract

Acute lung inflammation can be caused by a variety of respirable agents, including cigarette smoke. Long-term cigarette smoke exposure can cause chronic obstructive pulmonary disease (COPD), a serious illness that affects >10 million Americans. Cigarette smoke is a known inducer of inflammation and is responsible for approximately 90% of all COPD cases. RelB, a member of the NF-κB family, attenuates cigarette smoke-induced inflammatory mediator production in mouse lung fibroblasts in vitro. We hypothesized that overexpression of RelB in the airways of mice would dampen acute smoke-induced pulmonary inflammation. Mice received a recombinant adenovirus encoding RelB by intranasal aspiration to induce transient RelB overexpression in the lungs and were subsequently exposed to mainstream cigarette smoke. Markers of inflammation were analyzed after smoke exposure. Neutrophil infiltration, normally increased by smoke exposure, was significantly and potently decreased after RelB overexpression. Cigarette smoke-induced proinflammatory cytokine and chemokine production, cyclooxygenase-2 expression, and prostaglandin E(2) production were also significantly decreased in the context of RelB overexpression. The expression of intercellular adhesion molecule 1, an NF-κB-dependent protein, was decreased, indicating a potential mechanism through which RelB can regulate inflammatory cell migration. Therefore, increased expression and/or activation of RelB could be a novel therapeutic strategy against acute lung inflammation caused by respirable agents and possibly against chronic injury, such as COPD.

Figure 1

RelB recombinant adenovirus delivery by intranasal aspiration induces RelB overexpression in mouse lungs. A: Overexpression of human RelB was induced with an RelB recombinant adenovirus delivered by intranasal aspiration [two doses of 5 × 10⁸ plaque-forming units (pfu) per mouse delivered 24 hours apart], and RelB expression was detected in whole mouse left lung homogenates using Western blot 3, 5, and 7 days after the first adenovirus dose. Saline indicates saline treated; Control, control virus treated (two doses of 5 × 10⁸ pfu per mouse delivered 24 hours apart, both 7 days after initial treatment). B: Densitometric analysis revealed sevenfold, ninefold, and fivefold increases in RelB expression 3, 5, and 7 days, respectively, after the first treatment. RelB overexpression was confirmed 7 days after the initial treatment by IHC with an antibody specific to the human RelB isoform. C: Control virus. D: RelB virus. n = 3 for each group except saline treated, for which n = 1. *P < 0.05 for statistical significance relative to the control-treated mice (one-way analysis of variance with a Tukey posttest).

Figure 2

RelB overexpression before cigarette smoke exposure significantly affects bronchoalveolar cell populations. Mice received either the RelB or control virus (two doses of 5 × 10⁸ plaque-forming units per mouse delivered 24 hours apart) and either cigarette smoke or filtered air (two 1-hour exposures per day for 3 days). Cells present in the bronchoalveolar space were collected by lavage, counted, and analyzed after Cytospin slide preparation. Cigarette smoke exposure significantly increased the total number of cells (combined neutrophils and macrophages) present in the bronchoalveolar space (A). Significant increases in the total number and percentage of neutrophils (B and C, respectively) accounted for this increase, but RelB overexpression before smoke exposure damped these increases by more than threefold. RelB overexpression before smoke exposure also significantly increased the total number of macrophages present compared with the smoke-treated control group (D). No significant differences were seen between air-treated mice. *P < 0.05, **P < 0.01, and ***P < 0.001 (two-way analysis of variance with a Bonferroni posttest; n = 5 to 6 per group). E: Images representative of these observations are shown (black arrows indicate neutrophils; white arrows, macrophages).

Figure 3

RelB overexpression reduces cigarette smoke–induced neutrophilic inflammation. Mice were treated with either the RelB or control virus (two doses of 5 × 10⁸ plaque-forming units per mouse delivered 24 hours apart) and were exposed to either cigarette smoke or filtered air (two 1-hour exposures per day for 3 days). Unlavaged left mouse lungs were formalin fixed, sectioned, stained for neutrophils (red), and counterstained with hematoxylin. Few neutrophils were present in the lungs of air-treated mice (A and B), but neutrophil infiltration dramatically increased after cigarette smoke exposure (C). However, this increase was dramatically reduced after RelB overexpression (D). Black arrows indicate neutrophils.

Figure 4

RelB overexpression reduces cigarette smoke–induced MPO activity. Mice were treated with either the RelB or control virus (two doses of 5 × 10⁸ plaque-forming units per mouse delivered 24 hours apart) and with either smoke or filtered air (two 1-hour exposures per day for 3 days). MPO activity was detected in whole right lung homogenates, as described. MPO activity was significantly increased in smoke-treated mice. This was blocked by RelB overexpression. *P < 0.05 and **P < 0.01 (two-way analysis of variance with a Bonferroni posttest; n = 4 per group).

Figure 5

RelB overexpression reduces cigarette smoke–induced proinflammatory cytokine production in mouse lungs. Mice received either the control or RelB virus (two doses of 5 × 10⁸ plaque-forming units per mouse delivered 24 hours apart) and either air or smoke (two 1-hour exposures per day for 3 days). Cytokine production was detected in the BAL fluids by either enzyme-linked immunosorbent assay (A–E) or cytokine multiplex assay (F–H). Cigarette smoke exposure significantly increased production of each cytokine, whereas RelB overexpression reduced the production of IL-6 (A), MIP-2 (B), KC (C), JE (D), TNF-α (E), granulocyte-macrophage colony-stimulating factor (F), LIX (G), IFN-γ (H), and IL-10 (I). Only MIP-2 (B) and LIX (G) did not achieve statistical significance. Basal production of JE was also significantly reduced on RelB overexpression (D). *P < 0.05, **P < 0.01, and ***P < 0.001 (two-way analysis of variance with a Bonferroni posttest; n = 5 to 6 per group).

Figure 6

RelB overexpression reduces cigarette smoke–induced Cox-2 expression and PGE₂ production. Mice were treated with either the control or RelB virus (two doses of 5 × 10⁸ plaque-forming units per mouse delivered 24 hours apart) and were exposed to either smoke or air (two 1-hour exposures per day for 3 days). A: Cox-2 expression was detected in homogenates of both the left and right lungs using Western blot (n = 3 mice per group). B: Densitometric analysis of both blots confirmed that cigarette smoke significantly increased Cox-2 expression by 4.5-fold and that RelB overexpression before smoke exposure completely dampened this increase (n = 6 per group). C: PGE₂ production, detected in the BAL fluids of the same mice, was significantly increased on smoke exposure, but RelB overexpression dampened this increase. *P < 0.05, **P < 0.01, and ***P < 0.001, two-way analysis of variance with a Bonferroni posttest (n = 5 to 6 per group).

Figure 7

RelB overexpression reduces cigarette smoke–induced ICAM-1 expression in mouse lungs. Mice were treated with either the RelB or control virus (two doses of 5 × 10⁸ plaque-forming units per mouse delivered 24 hours apart) and exposed to filtered air or smoke (two 1-hour exposures per day for 3 days). A: Unlavaged left lungs were formalin fixed, sectioned, and probed for ICAM-1 expression. Cigarette smoke increased ICAM-1 expression (red), but RelB overexpression before smoke exposure dampened this increase. Nuclei were detected by DAPI staining (blue). B: ICAM-1 was also detected in whole left lung homogenates using Western blot. C: Densitometric analysis confirmed that smoke exposure significantly increased ICAM-1 expression and that RelB overexpression significantly dampened this increase. RelB overexpression did not significantly affect basal ICAM-1 expression. **P < 0.01, two-way analysis of variance with a Bonferroni posttest (n = 3 per group).