Role of macrophages in healing the fibrotic lung: pan hydroxylase inhibition as a potential therapeutic mechanism

Abstract

Pulmonary fibrosis is a common consequence of lung inflammation, leading to organ dysfunction and significant morbidity and mortality. Macrophages, through their diverse functions associated with polarisation status, play a role in lung homeostasis and alternatively activated (M2) macrophages have been associated with lung fibrosis. Prolyl hydroxylases (PHDs) are the main oxygen sensors and regulators of hypoxia inducible factors (HIFs). The PHD/HIF pathway is known to play a role in tissue inflammation and fibrosis, but their role in macrophage polarisation is not fully understood. Aim To study the role of the PHD/HIF pathway in macrophage polarisation and lung fibrosis, and specifically in Idiopathic Pulmonary Fibrosis (IPF). Hypothesis It was hypothesised that pan hydroxylase inhibition alters macrophage polarisation and modulates lung inflammation and fibrosis. Methods A combination of pharmacological (pan hydroxylase inhibitors DMOG and FG41) and genetic (HIF and PHD-null) tools were used to manipulate the PHD/HIF pathway. The bleomycin induced lung fibrosis model was used to define the effect of pan hydroxylase inhibition during the early, inflammatory or the late, fibrotic phase of this model. Murine bone marrow derived macrophages (BMDM), human monocyte derived macrophages and alveolar macrophages obtained from patients with lung fibrosis were used to study the effect of pan hydroxylase inhibition on macrophage polarisation. Bronchoalveolar lavage fluid (BALF) from patients was used to define the association between lung CCL18, an M2 associated chemokine, and disease progression in IPF. Results DMOG therapy during the early phase of the bleomycin model significantly reduced lung fibrosis at day 24. In contrast, late phase pan hydroxylase inhibition enhanced lung fibrosis at day 24. In both instances there was evidence of enhanced alveolar macrophage M2-like polarisation following pan hydroxylase inhibition. Reduced fibrosis after early pan hydroxylase inhibition was not a consequence of reduced acute lung inflammation or direct inhibition of collagen synthesis. In BMDM, pan hydroxylase inhibition resulted in an ‘augmented M2-like’ macrophage. Using LysM-Cre HIF-1α, HIF-2α and PHD-3 KO mice as well as chetomin, a potent inhibitor of HIF-1α and HIF-2α mediated gene expression, the HIF-dependent and HIF-independent polarisation markers were defined. PHD-3 deficiency was not sufficient to enhance M2 skewing. In contrast to murine BMDM, in human monocyte derived macrophages and alveolar macrophages from healthy volunteers and patients with interstitial lung disease including IPF, pan hydroxylase inhibition did not augment M2 polarisation and indeed significantly inhibited macrophage CCL18 expression. CCL18 studies in clinical BALF samples confirmed that CCL18 was elevated in the lungs of patients with IPF and other ILDs compared to controls. However, baseline BALF CCL18 concentrations did not correlate with disease severity or with disease progression, suggesting this is not a useful biomarker in IPF. Further, a unique study of serial BAL in IPF patients showed no association between 12-month change in CCL18 and disease progression over the same period. Indeed CCL18 concentrations mostly fell over 12 months in patients that did progress, strongly suggesting that CCL18 does not play a major pathogenic role in IPF. Concluding, it was shown that in both BMDM and murine lung pan hydroxylase inhibition promoted an ‘augmented M2-like’ polarisation. Pharmacological pan hydroxylase inhibition during the late fibrotic phase of injury enhanced fibrosis but it is not known if there was a causal association between M2 macrophages and lung fibrosis. Similarly, the functional relevance of finding enhanced M2 polarisation observed during early DMOG therapy, which subsequently resulted in attenuated fibrosis, is not known. In human macrophages, pan hydroxylase inhibition unexpectedly attenuated CCL18 production, a chemokine associated with an M2-like phenotype in man whilst other M2 markers were unchanged. However, there was no evidence to support a pathogenic role for CCL18 in IPF, and therefore there is little potential for using pan hydroxylase inhibition to target CCL18 and treat IPF.