Opsonic Phagocytosis in Chronic Obstructive Pulmonary Disease Is Enhanced by Nrf2 Agonists

Abstract

Rationale: Previous studies have identified defects in bacterial phagocytosis by alveolar macrophages (AMs) in patients with chronic obstructive pulmonary disease (COPD), but the mechanisms and clinical consequences remain incompletely defined.

Objectives: To examine the effect of COPD on AM phagocytic responses and identify the mechanisms, clinical consequences, and potential for therapeutic manipulation of these defects.

Methods: We isolated AMs and monocyte-derived macrophages (MDMs) from a cohort of patients with COPD and control subjects within the Medical Research Council COPDMAP consortium and measured phagocytosis of bacteria in relation to opsonic conditions and clinical features.

Measurements and main results: COPD AMs and MDMs have impaired phagocytosis of Streptococcus pneumoniae. COPD AMs have a selective defect in uptake of opsonized bacteria, despite the presence of antipneumococcal antibodies in BAL, not observed in MDMs or healthy donor AMs. AM defects in phagocytosis in COPD are significantly associated with exacerbation frequency, isolation of pathogenic bacteria, and health-related quality-of-life scores. Bacterial binding and initial intracellular killing of opsonized bacteria in COPD AMs was not reduced. COPD AMs have reduced transcriptional responses to opsonized bacteria, such as cellular stress responses that include transcriptional modules involving antioxidant defenses and Nrf2 (nuclear factor erythroid 2-related factor 2)-regulated genes. Agonists of the cytoprotective transcription factor Nrf2 (sulforaphane and compound 7) reverse defects in phagocytosis of S. pneumoniae and nontypeable Haemophilus influenzae by COPD AMs.

Conclusions: Patients with COPD have clinically relevant defects in opsonic phagocytosis by AMs, associated with impaired transcriptional responses to cellular stress, which are reversed by therapeutic targeting with Nrf2 agonists.

Keywords: antioxidant; chronic obstructive pulmonary disease; macrophage; nuclear factor erythroid 2–related factor 2 (Nrf2); phagocytosis.

Figures

Figure 1.

Chronic obstructive pulmonary disease (COPD) alveolar macrophages (AMs) show deficient opsonic bacterial phagocytosis that correlates with exacerbation frequency. (A–D) AMs (A and B) or monocyte-derived macrophages (MDMs) (C and D) from healthy or COPD (nonfrequent [NF] and frequent [F] exacerbators) were challenged with either nonopsonized (A andC) or opsonized (B andD) serotype 14 Streptococcus pneumoniae. At 4 hours after challenge, viable intracellular bacteria were assessed. n values for healthy/COPD-NF/COPD-F, respectively: (A)n = 18/27/15; (B)n = 10/19/13; (C)n = 14/18/12; (D)n = 14/15/12. ***P < 0.001, one-way ANOVA. (E and F) A pairwise comparison of phagocytosis of nonopsonized (E) and opsonized (F) bacteria in MDMs and AMs from matched donors. ns = not significant. nvalues for healthy/COPD-NF/COPD-F, respectively: (E)n = 11/12/12; (F)n = 10/11/12. *P < 0.05, **P < 0.01, ***P < 0.001, paired t test. (G andH) A pairwise comparison of phagocytosis of nonopsonized and opsonized S. pneumoniae in matched AM (G) or MDM (H) donors.n values for healthy/COPD-NF/COPD-F, respectively: (G)n = 10/20/11; (H)n = 7/8/5. *P < 0.05, pairedt test.

Figure 2.

Defects in phagocytosis in chronic obstructive pulmonary disease alveolar macrophages are associated with bacterial colonization in the lung. (A–D) Nonopsonic (A and C) and opsonic (B and D) phagocytosis was stratified into groups dependent on whether the donor had negative (−ve) or positive (+ve) sputum culture (Aand B, n = 15 andn = 14, respectively) or −ve or +ve (defined as >104 copies/ml) quantitative PCR of BAL (C and D,n = 27 andn = 26, respectively) results indicative of bacterial colonization, *P < 0.05, Student’st test. Open circles indicate frequent exacerbations; solid circles indicate nonfrequent exacerbations.

Figure 3.

Opsonic phagocytosis correlates with FEV1. (A) Nonopsonic and (B) opsonic phagocytosis rates were correlated against patient FEV1 score. Pearson’s correlation coefficients (r) and P values are shown, with correlation deemed significant if P < 0.05. Nonopsonic,n = 36; opsonic,n = 32. ns = not significant.

Figure 4.

Opsonic phagocytosis correlates with markers of chronic obstructive pulmonary disease (COPD) severity. (A,C, and E) Nonopsonic and (B, D, and F) opsonic rates of phagocytosis were correlated against patient scores for a variety of markers of COPD severity: (A andB) the St. George’s Respiratory Questionnaire (SGRQ) (nonopsonic,n = 29; opsonic,n =  27), (C andD) the COPD Assessment Test (CAT) (nonopsonic,n = 34; opsonic,n = 30), or (Eand F) the 6-minute-walk distance (6MW) (nonopsonic,n = 14; opsonic,n = 10). Values for Pearson’s (r) correlation coefficient andP values are shown, with correlation deemed significant if P < 0.05. ns = not significant.

Figure 5.

Transcriptional response of alveolar macrophages (AMs) reveals less differential gene expression in chronic obstructive pulmonary disease (COPD) in response to infection. AMs from healthy subjects or patients with COPD were challenged with opsonized serotype 14 Streptococcus pneumoniae(n = 3 in each group). At 4 hours after challenge, cell total RNA was collected for transcriptional analysis. (A) Venn diagram showing the number of probes differentially expressed in response to infection (moderatedt test, <0.05; false discovery rate, <0.05). (B) Plots represent the top 10 enriched Gene Ontology (GO) biological process terms and the cellular response to stress terms; in addition, the response to oxidative stress term is plotted in the healthy AMs. The x-axis represents enrichment by a hypergeometric test [−log10(P value)]. The size of the circle and color represent the number of differentially expressed genes of interest in that term. Figures were generated using NIPA (available from: https://github.com/ADAC-UoN/NIPA). (C) Volcano plots represent the probe sets identified from the transcriptomic analysis. In the left plot, which displays findings for healthy AMs, the red triangles are the differentially expressed probes related to the “Cellular response to stress term” with some representative terms named. In blue are the terms associated with the Nrf2 (nuclear factor erythroid 2–related factor 2) pathway in the analysis of healthy AMs. In the right plot, which displays COPD AMs, the red triangles are the differentially expressed probes related to the “Cellular response to stress” GO term. In blue are the terms associated with the Nrf2 pathway seen in the healthy AM analysis. GOI = gene of interest.

Figure 6.

Treatment with the Nrf2 (nuclear factor erythroid 2–related factor 2) agonist sulforaphane increases nonopsonic and opsonic phagocytosis in chronic obstructive pulmonary disease (COPD) alveolar macrophages (AMs) but not in monocyte-derived macrophages (MDMs). (A) AMs and (B) MDMs were pretreated with the designated dose of sulforaphane (Sulf) for 16 hours before cells were lysed and probed for expression of heme oxygenase (HO)-1 and actin (n = 3). (C–F) AMs (C and D) or MDMs (E and F) from healthy (H) or COPD nonfrequent (NF) or frequent (F) exacerbators were pretreated with vehicle (Sulf−) or sulforaphane (Sulf+) for 16 hours before cells were challenged with nonopsonized (C andE) or opsonized (D andF) serotype 14 Streptococcus pneumoniae. Four hours after challenge, numbers of intracellular viable bacteria were measured, and nvalues for H/COPD-NF/COPD-F, respectively, were as follows: (C)n = 11/19/14; (D)n = 8/10/4; (E)n = 9/9/5; (F)n = 8/9/5. *P < 0.05, pairedt test. (G and H) AMs (G) and MDMs (H) from patients with COPD or healthy (H) donors were pretreated with sulforaphane before being challenged with nontypeable Haemophilus influenzae (NTHi). Four hours after challenge, the numbers of intracellular viable bacteria were measured, and nvalues for H/COPD, respectively, were as follows: (G)n = 3/4; (H)n = 3/3. *P < 0.05, pairedt test. (I) COPD AMs were pretreated with sulforaphane (+Sulf) before being challenged with nonopsonized serotype 14 S. pneumoniae for 4 hours before extracellular bacteria were killed by the addition of antimicrobials. At the designated time after addition of antimicrobials, viable bacteria in duplicate wells were measured (n = 3), with no significant difference between vehicle and sulforaphane. ns = not significant.

Figure 7.

The Nrf2 (nuclear factor erythroid 2–related factor 2) agonist Compound 7 also increases phagocytosis in chronic obstructive pulmonary disease (COPD) alveolar macrophages (AMs). (A andB) COPD monocyte-derived macrophages (MDMs) (A) or AMs (B) were pretreated with the Nrf2 agonist Compound 7 for 16 hours at the designated dose before cells were lysed and probed for the expression of heme oxygenase (HO)-1, GCLC (glutamate-cysteine ligase catalytic subunit), or NQO1 (NADPH:quinone oxidoreductase 1) by Western blotting. (C and D) Healthy donor and COPD AMs were pretreated with Compound 7 at 5 × IC50 (0.065 μM) for 16 hours before being challenged with nonopsonized (n = 3 healthy;n = 5 COPD) (C) or opsonized (n = 3 healthy;n = 10 COPD) (D) serotype 14 Streptococcus pneumoniae for 4 hours, after which numbers of intracellular viable bacteria were assessed. **P < 0.01, pairedt test. (E and F) Healthy donor and COPD MDMs were pretreated with Compound 7 and challenged with nonopsonized (E) or opsonized (F) S. pneumoniae as was done for AMs. All n = 4. **P < 0.01, ***P < 0.001, paired t test. IC50 = half-maximal inhibitory concentration.