Article Text

Download PDFPDF

Neutrophils and tissue damage: is hypoxia the key to excessive degranulation?
  1. Andrew E Williams,
  2. Rachel C Chambers
  1. Division of Medicine, Centre for Inflammation and Tissue Repair, UCL Respiratory, University College London, London, UK
  1. Correspondence to Professor Rachel C Chambers, Centre for Inflammation and Tissue Repair, Division of Medicine, University College London, London WC1E 6JF, UK; r.chambers{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Neutrophils are the first innate immune cells recruited into sites of infection and inflammation, and although they are critical for host defence against microbial pathogens, they have also been associated with local tissue damage. Predominant neutrophilic inflammation is a key hallmark of many respiratory diseases, and the release of proteinases and reactive oxygen species (ROS) from primed neutrophils has been implicated in mediating the damage to surrounding tissue.1 Inflammation is further accompanied by severe hypoxia (lack of oxygen) within the affected tissue, and is a characteristic of both acute and chronic lung diseases, such as acute respiratory distress syndrome (ARDS),2 COPD3 and cystic fibrosis (CF).4 Considering that tissue hypoxia is part of a normal inflammatory response, innate immune cells such as neutrophils and macrophages possess important cellular and molecular mechanisms that enable them to function effectively at low oxygen concentrations.5

There is considerable interplay between the molecular pathways that regulate hypoxia and inflammation. In previous studies, mice exposed to hypoxic conditions (5% O2 for 1 hour) demonstrated increased levels of interleukin 6 (IL-6), tumour necrosis factor and IL-1 in both serum and isolated macrophages.6 In humans, increased levels of IL-6, IL-1RA and C-reactive protein were detected in serum after volunteers experienced three overnight stays at altitude.7 Moreover, infiltrating neutrophils directly increase their oxygen consumption in order to assemble the nicotinamide adenine dinucleotide phosphate oxidase complex that generates ROS and, in conjunction with superoxide dismutase and myeloperoxidase (MPO), hydrogen peroxide (H2O2) and hypochlorous acid (HClO), thereby depleting local levels of oxygen.8 ,9 Neutrophils are also thought to influence the tissue microenvironment during episodes of hypoxia. For example, neutrophil migration across epithelia increases the transcriptional activity of hypoxia-inducible genes in epithelial cells, which in turn influence the resolution of inflammation.10

Neutrophils have, therefore, been implicated in the pathogenesis of several inflammatory lung diseases associated with tissue hypoxia.5 However, the mechanisms by which neutrophils mediate tissue damage under hypoxic conditions are less clear. Under homeostatic conditions neutrophils are short-lived cells that readily undergo apoptosis in order to avoid the release of toxic granule contents. However, under hypoxic conditions neutrophil survival is prolonged, which is thought to delay the resolution of inflammation and promote tissue damage, potentially through the release of destructive proteinases or ROS.11 The study by Hoenderdos et al in this issue of Thorax aims to explore the functional link between hypoxia and the regulation of neutrophil degranulation in the context airway inflammation.

Compared with neutrophils incubated under normoxic conditions (atmospheric O2 concentration), this study demonstrated that neutrophils incubated under hypoxic conditions (0.8% O2, 3 kPa for 4 hours) released significantly elevated levels of neutrophil elastase (NE), MPO, lactoferrin and matrix metalloproteinase-9, thereby indicating degranulation of azurophil (primary), specific (secondary) and gelatinase (tertiary) granules. This effect was largely dependent on priming with granulocyte-macrophage colony-stimulating factor (GM-CSF) and activation with formylated peptide (fMLP), as little degranulation occurred at basal levels in either normoxic or hypoxic conditions. These findings support previous studies demonstrating that neutrophils isolated from human volunteers, following a period of hypoxaemia, released increased levels of NE after fMLP stimulation,12 and neutrophils from healthy volunteers exposed to hypoxia ex vivo (3 kPA for 1–4 hours) released NE.13 However, these two studies reported opposing effects on ROS production.

Previous studies under hypoxia have sometimes given conflicting results, which is likely related to differing experimental conditions that affect neutrophil function. There may, therefore, be some merit in standardising methodologies across laboratories, to enable direct comparison of studies within the scientific community. It is also important to consider whether experimental hypoxic conditions mirror those found in the lung during episodes of inflammation. The lung is normally an oxygen-rich organ, with an alveolar PO2 ∼110 mm Hg compared with other systemic organs (3–20 mm Hg). Therefore, oxygen tensions in a hypoxic lung are higher than other organs (∼50 mm Hg compared with less than 10 mm Hg).14 The current study exposed neutrophils to hypoxic conditions of 3 kPa, which is equivalent to ∼20 mm Hg, and therefore attempts to recapitulate the physiological conditions within the airway epithelium. It is argued that similar mechanistic ex vivo studies should also reflect these unique conditions. More challenging will be to recapitulate hypoxic conditions within tissue niches or foci of inflammation, such as those that occur in the small airway epithelium in COPD and CF.

The release of neutrophil proteinases has been strongly implicated in promoting tissue damage within inflamed lung.15 In this study, hypoxic supernatants from GM-CSF-stimulated and fMLP-stimulated neutrophils increased the detachment and death of epithelial cells compared with normoxic neutrophils. Treatment with α-1-anti-trypsin further revealed that the toxic effects were mediated by proteinases, supporting a role for activated neutrophils in mediating tissue damage. The enhanced toxic effects of hypoxic neutrophil supernatants were apparent when epithelial cells were cultured in two-dimensional (2D) monolayers. However, recapitulating these findings in 3D culture systems or, more importantly in vivo, remain a considerable challenge.

It is important to fully appreciate the significant effect that hypoxia has on downstream signalling pathways in both tissue-resident cells and infiltrating leukocytes. Eukaryotic cells have evolved several oxygen-sensing pathways, of which the principal regulators are hypoxia-inducible transcription factor-1 and transcription factor-2 (HIF-1 and HIF-2).16 Under hypoxic conditions HIFs are stabilised, allowing them to translocate to the nucleus and regulate HIF response genes.17 Interestingly, the current study reports that hypoxia-induced degranulation of neutrophils is not primarily mediated by HIF-1. Treatment of neutrophils with either dimethyloxalyl glycine or desferrioxamine, which are both HIF-1-stabilising agents, did not affect NE release. Furthermore, hypoxia did not increase the expression of NE but did increase the expression of BCL2/Adenovirus E1B 19kDa Interacting Protein 2, both of which are putative HIF-1 response genes. Taken together, these data suggest that NE release following hypoxia-induced degranulation may be HIF-1 independent, although the contribution of HIF-2 was not directly explored in this study.

Neutrophil degranulation is also thought to be dependent on actin cytoskeletal rearrangements. Indeed, hypoxia caused focal actin polymerisation in fMLP-stimulated neutrophils. However, inhibition of actin polymerisation using cytochalasin B increased NE release, while treatment with jasplakinolide (which promotes actin polymerisation) had no significant effect on NE release. Actin remodelling is necessary for neutrophil degranulation, as well as polarisation and migration in response to exogenous stimuli such as fMLP and IL-8.18 Cytoskeletal rearrangements are thought to be a prerequisite for the intracellular movement of granules within the cytoplasm, while actin disassembly is necessary for access to the inner surface of the plasma membrane. It is, therefore, plausible that cytochalasin B relaxes the actin cytoskeleton and thereby promotes granule fusion with the plasma membrane, in a similar way that immune synapses form during polarised secretion of cytotoxic T-cell granules.19

Degranulation is a tightly regulated process involving the activation of G-protein-coupled receptors, intracellular calcium mobilisation and ATP-dependent phosphorylation of downstream signalling molecules.18 Treatment with a phospholipase C inhibitor or ethylene glycol tetraacetic acid, in order to block intracellular Ca2+, inhibited NE release under hypoxic conditions. Furthermore, treatment with thapsigargin, which is a Ca2+ ATPase inhibitor that elevates intracellular Ca2+ levels, caused an increase in NE release. These studies, therefore, confirm that Ca2+ mobilisation is an important signalling event during neutrophil degranulation under hypoxic conditions. However, the specific downstream targets of Ca2+ during neutrophil degranulation remain largely unknown.

Considering that hypoxia-induced neutrophil degranulation was independent of HIF signalling, an alternative phosphoinositide 3-kinase (PI3K) signalling pathway was investigated. A pan PI3K inhibitor reduced neutrophil survival under hypoxic conditions, while a pan PI3K inhibitor and a PI3Kγ inhibitor completely abolished NE release from GM-CSF-stimulated and fMLP-stimulated neutrophils. Inhibition of PI3Kγ also inhibited the downstream phosphorylation of Akt, suggesting that the PI3K-Akt pathway may play a role in hypoxia-mediated neutrophil degranulation. It will be interesting to explore whether other known regulators of neutrophil degranulation, such as Src family kinases, guanosine triphosphatases, such as Ras and Rho, and MAP kinases, are also regulated by hypoxia. In addition, hypoxia, PI3K/Akt and nuclear factor-κB (NF-κB) pathways interact to promote proinflammatory and HIF-target gene expression. Although not explored in the current study, the interplay between hypoxia-mediated signalling pathways and NF-κB-mediated pathways is likely to be functionally important in driving neutrophilic inflammation.11

In summary, the current study by Hoenderdos et al in this issue of Thorax highlights the importance of hypoxia in regulating neutrophil degranulation. The continued release of neutrophil-derived proteinases is likely to contribute to the tissue damage observed in acute and chronic lung diseases, such as ARDS, CF and COPD. Studying neutrophil function in hypoxic conditions ex vivo may therefore reveal novel, drug-targetable cellular and molecular processes. The future challenge will be to reproduce the physiological conditions that occur within an injured lung as closely as possible.


View Abstract


  • Twitter Follow Andrew Williams and CITR at @AndrewEWilliams and @LungLab

  • Contributors AEW and RCC wrote and edited the manuscript.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.

Linked Articles

  • Respiratory infection
    Kim Hoenderdos Katharine M Lodge Robert A Hirst Cheng Chen Stefano G C Palazzo Annette Emerenciana Charlotte Summers Adri Angyal Linsey Porter Jatinder K Juss Christopher O'Callaghan Edwin R Chilvers Alison M Condliffe