Associations of serum and bronchoalveolar immunoglobulins with lung microbiota diversity, B-cell memory phenotypes, and COPD morbidity and exacerbations.
Overview
abstract
RATIONALE: Immunoglobulins (Ig) protect against pathogens frequently implicated in COPD exacerbations. We previously demonstrated an association of low-normal serum IgA and IgG concentrations with prospective exacerbation risk, but responsible mechanisms are undefined. Here, we examined associations of lower respiratory tract bacterial diversity to Ig levels in serum and bronchoalveolar lavage (BAL) and to the memory phenotypes of blood and BAL B cells. METHODS: We analyzed data from phase I of SPIROMICS, an observational cohort study of smoking-related COPD. A subset of participants completed comprehensive research bronchoscopies, including analysis of BAL bacterial microbiota by 16 S rRNA gene (V4 region) sequencing and of blood and BAL B-cells by 12-color flow cytometry. In some participants, we also analyzed serum and BAL Ig levels by ELISA. We constructed linear regression models including either serum or BAL (albumin-corrected) Ig measurements as the independent variable and separate dependent variables, including B-cell subsets, BAL bacterial diversity metrics (Faith phylogenetic diversity, inverse Simpson, and richness indices), and clinical measures (FEV1% predicted, risk of prospective exacerbations), adjusted by age, sex, race, educational attainment, smoking status, and use of inhaled corticosteroids. RESULTS: Serum IgG and IgA (n = 66 participants) were 1,486.1 ± 510.6 mg/dL [mean ± standard deviation (SD)] and 237.7 ± 131.6 mg/dL, respectively. Albumin-corrected BAL IgG and IgA (n = 117) were 0.03 ± 0.02 mg/dL and 0.01 ± 0.01 mg/dL, respectively. B-cells (n = 82) comprised 3.5 ± 3.0% of blood leukocytes. Serum IgA was associated with higher blood switched memory (IgD- CD27+) B-cell percentages (β 6.06, p = 0.01) and inversely associated with blood double-negative (IgD-CD27-) B-cell percentages (β - 9.96, p = 0.02). Available BAL microbiome data (n = 107) showed that reduced lung bacterial diversity associated with lower serum IgG, but not with serum IgA, BAL IgA, or BAL IgG concentrations. Neither BAL IgG nor IgA were associated with lung function or exacerbations. CONCLUSIONS: These results demonstrate an association of low serum IgG with reduced lung bacterial diversity, a feature of dysbiosis that may predispose to exacerbation. Defining the role of Ig in specific anatomic compartments is relevant to designing vaccine strategies.
