In chronic obstructive pulmonary disease (COPD), lung natural killer cells (NKs) lyse autologous lung epithelial cells in vitro, but underlying mechanisms and their relationship to epithelial cell apoptosis in vivo are undefined. Although this cytolytic capacity of lung NKs depends on priming by dendritic cells (DC), whether priming correlates with DC maturation or is limited to a specific DC subset are also unknown. We recruited ever-smokers (≥10 pack-years) (n=96) undergoing clinically-indicated lung resections. We analyzed lung NKs for cytotoxic molecule transcripts and for cytotoxicity, which we correlated with in situ detection of activated Caspase-3/7+ airway epithelial cells. To investigate DC priming, we measured lung DC expression of CCR2, CCR7, and CX3CR1, and co-cultured peripheral blood NKs with autologous lung DC, either matured using LPS (non-obstructed smokers) or separated into conventional DC type-1 (cDC1) versus cDC type-2 (cDC2) (COPD). Lung NKs in COPD expressed more perforin (p<0.02) and granzyme B (p<0.03) transcripts; inhibiting perforin blocked in vitro killing by lung NKs. Cytotoxicity in vitro correlated significantly (Sr=0.68, p=0.0043) with numbers of apoptotic epithelial cells per airway. In non-obstructed smokers, LPS-induced maturation enhanced DC-mediated priming of blood NKs, reflected by greater epithelial cell death. Although CCR7 expression was greater in COPD in both cDC1 (p<0.03) and cDC2 (p=0.009), only lung cDC1 primed NK killing. Thus, rather than being intrinsic to those with COPD, NK priming is a capacity of human lung DC that is inducible by recognition of bacterial (and possibly other) danger signals and restricted to the cDC1 subset.

No abstract

Apoptosis of lung epithelial cells is central to the development of emphysema, the phenotype of chronic obstructive pulmonary disease (COPD) most closely associated with mortality. We previously showed that human lung natural killer (NK) cell-mediated cytotoxicity against autologous lung epithelial cells is increased in COPD [PMID:29676596]. Regulatory T cells (Tregs) are known to inhibit NK-mediated cytotoxicity, but there is evidence of Treg deficiency in COPD. Whether restoring Tregs could restrict NK cytotoxicity in COPD is not known. Methods: We collected lung tissue and peripheral blood from consented smokers undergoing clinicallyindicated resections (n=3 with COPD). We isolated blood Tregs (CD4+ CD25+), lung NKs (CD56+), and lung epithelial cells (EpCAM+). NKs and epithelial cells were cultured without or with autologous peripheral blood Tregs at a ratio of 1 NK to 1 Treg. After 4 hours, we isolated cells, stained for Annexin-V and 7-AAD, then analyzed viability by flow cytometry. In a murine model, lung NKs and lung epithelial cells isolated from mice after 8-week of cigarette (CS) exposure were cultured for 4 hours without or with Tregs (isolated from spleens of naïve mice using CD4+ CD25+ selection). We evaluated epithelial cell apoptosis as described above. In separate experiments, mice exposed to CS for 8 weeks received adoptive transfer by the intranasal route of either vehicle or splenic Tregs from naïve mice. After 7 days, we collected lungs, isolated NKs and epithelial cells, and evaluated NK cytotoxicity towards epithelial cells in vitro. Results: In the presence of Tregs, human lung NKs from COPD subjects killed fewer autologous epithelial cells (specific cytotoxicity 10% vs. 38%, p<0.05). In the murine experiments, co-culture of splenic Tregs isolated from naïve mice with NKs from CS-exposed mice resulted in 2fold decreased cytotoxicity towards lung epithelial cells. Co-culture of Tregs in the absence of NKs had no effect on epithelial cells. We further found that adoptive transfer of splenic CD4+ CD25+ Tregs into CS-exposed mice modestly decreased NK cytotoxicity at 7 days after adoptive transfer (11% vs. 17% specific cytotoxicity). Conclusions: These data show for the first time that Tregs can modulate the cytotoxicity of lung NKs from COPD patients in vitro, and both in vitro and in vivo in a murine CS-exposure model. The agreement between human and murine data implies that the latter could be used to examine mechanisms and to optimize delivery strategies. Therapeutic use of Tregs is unstudied in COPD but warrants additional investigation.

No abstract

No abstract