American Journal of Respiratory and Critical Care Medicine

Rationale: Patients with chronic obstructive pulmonary disease (COPD) experience excess cardiovascular morbidity and mortality, and exacerbations further increase the risk of such events. COPD is associated with persistent blood and airway neutrophilia and systemic and tissue hypoxia. Hypoxia augments neutrophil elastase release, enhancing capacity for tissue injury.

Objective: To determine whether hypoxia-driven neutrophil protein secretion contributes to endothelial damage in COPD.

Methods: The healthy human neutrophil secretome generated under normoxic or hypoxic conditions was characterized by quantitative mass spectrometry, and the capacity for neutrophil-mediated endothelial damage was assessed. Histotoxic protein concentrations were measured in normoxic versus hypoxic neutrophil supernatants and plasma from patients experiencing COPD exacerbation and healthy control subjects.

Measurements and Main Results: Hypoxia promoted PI3Kγ-dependent neutrophil elastase secretion, with greater release seen in neutrophils from patients with COPD. Supernatants from neutrophils incubated under hypoxia caused pulmonary endothelial cell damage, and identical supernatants from COPD neutrophils increased neutrophil adherence to endothelial cells. Proteomics revealed differential neutrophil protein secretion under hypoxia and normoxia, and hypoxia augmented secretion of a subset of histotoxic granule and cytosolic proteins, with significantly greater release seen in COPD neutrophils. The plasma of patients with COPD had higher content of hypoxia-upregulated neutrophil-derived proteins and protease activity, and vascular injury markers.

Conclusions: Hypoxia drives a destructive “hypersecretory” neutrophil phenotype conferring enhanced capacity for endothelial injury, with a corresponding signature of neutrophil degranulation and vascular injury identified in plasma of patients with COPD. Thus, hypoxic enhancement of neutrophil degranulation may contribute to increased cardiovascular risk in COPD. These insights may identify new therapeutic opportunities for endothelial damage in COPD.

Scientific Knowledge on the Subject

Chronic obstructive pulmonary disease (COPD) is characterized by persistent neutrophilia in the setting of local and systemic hypoxia and is associated with excess cardiovascular disease, even allowing for known risk factors. Neutrophils accumulating in areas of inflammation and microcirculatory impairment experience profound hypoxia, which prolongs their survival and increases their secretory responses. Thus, hypoxic neutrophils have increased potential to cause endothelial injury, but their role in mediating the increased cardiovascular risk in COPD is poorly understood.

What This Study Adds to the Field

Herein we show that hypoxia augments the ability of neutrophils to selectively secrete a subset of histotoxic proteins capable of damaging endothelial cells. Hypoxia further enhanced release of these proteins from neutrophils from patients experiencing COPD exacerbation, with elevated concentrations also detected in patient plasma. This study suggests that hypoxic enhancement of neutrophil degranulation may contribute to increased cardiovascular risk in COPD and that the hypoxic neutrophil secretome proteins may represent new therapeutic targets to alleviate endothelial dysfunction in COPD.

Chronic obstructive pulmonary disease (COPD) is characterized by neutrophilic inflammation in the setting of tissue (and often systemic) hypoxia and by increased risk of cardiovascular disease and pulmonary hypertension. Neutrophil elastase (NE) has been implicated in COPD pathogenesis (1), and we have previously shown that hypoxia markedly augments NE release from neutrophils to promote respiratory epithelial cell damage (2). However, the impact of hypoxia on the extended neutrophil secretome and the potential for “hypoxic neutrophils” to injure other disease-relevant cell types are currently unknown. Identifying novel targets implicated in driving COPD morbidity may provide new therapeutic opportunities.

Even in health, certain tissues (e.g., muscle) are hypoxic (3). This “physiological hypoxia” may be compounded during exercise, inducing neutrophil phenotypic changes (4). In disease, profound “pathological hypoxia” exists in inflamed or infected tissues and areas of microcirculatory impairment. Although patients with severe COPD are systemically hypoxic, significant tissue hypoxia (<1.3% oxygen) can occur even in mild disease, demonstrated in inflamed airways (57) and atherosclerotic vasculature (8), where neutrophils accumulate. Upregulation of hypoxia-inducible factors in neutrophils from patients with acute lung injury, including coronavirus disease (COVID-19) infection, indicates neutrophil exposure to hypoxia in vivo (9).

Neutrophil antimicrobial function depends on the fusion of preformed granules, containing cytotoxic proteins and proteases, with the pathogen-containing phagosome. However, highly activated neutrophils can release granule contents extracellularly (degranulation), with potential for collateral tissue damage (2). Platelet-activating factor (PAF), a physiological priming agent capable of substantially enhancing neutrophil degranulation in response to subsequent stimulation (2, 10), has been implicated in COPD pathogenesis and in endothelial damage and remote organ damage in the setting of hypoxia (11, 12). Bacteria release formylated peptides (N-formyl-methionyl-leucyl-phenylalanine [fMLP]), which potently activate neutrophils; these peptides are present in cigarette smoke and have been implicated in emphysema progression (13). Patients with COPD suffer recurrent infection-driven exacerbations, but neutrophilic inflammation persists even in the absence of detectable infection, correlating with disease severity and progression (14). Despite evidence of NE-induced lung injury, translation of NE inhibitors has not led to significant benefit (15), perhaps reflecting the complex array of additional neutrophil-secreted proteins with damaging potential.

Patients with COPD have increased risk of cardiovascular morbidity and mortality even after adjusting for shared risk factors, including smoking (16), particularly after exacerbation (17). Pulmonary endothelial dysfunction in patients with COPD can induce pulmonary hypertension, which correlates with hypoxia (18). Accumulating evidence indicates inflammation, oxidative stress, and vascular tissue damage as key mechanisms linking COPD and cardiovascular disease (19), with neutrophil degranulation identified as an important pathway (20). Circulating neutrophils primed for enhanced degranulation have been identified in patients experiencing COPD exacerbation (21), with potential to contribute to endothelial injury.

Herein we show that hypoxia promotes neutrophil degranulation and neutrophil-induced endothelial damage. Proteomic analysis reveals hypoxia-driven secretion of highly histotoxic proteins from healthy neutrophils, and a subset of these are further increased from COPD neutrophils. Supernatants from hypoxic COPD neutrophils enhance neutrophil–endothelial adhesion. Finally, we identify increased concentrations of corresponding hypoxic neutrophil histotoxic granule proteins in COPD plasma, together with endothelial injury biomarkers. Some results have been reported previously in abstract form (2224).

Ethics Statement

Written informed consent was obtained from all healthy volunteers and patients with COPD (06/Q0108/281, 08/H0308/281, 18/WM/0097, and 20/SS/0085). All studies complied with the Declaration of Helsinki. All animal experiments were approved in accordance with the Animals (Scientific Procedures) Act 1986.

Human Neutrophils

Venous blood neutrophils were isolated by centrifugation over plasma-Percoll gradients (25). Neutrophils were resuspended in normoxic (atmospheric 21% O2) or hypoxic (0.8% O2 equating to media Po2 of 3 kPa [26], 5% CO2, Baker Ruskinn or Whitley hypoxia workstation) Iscove’s modified Dulbecco’s medium (IMDM). At 4 hours, neutrophils (11.1 × 106/ml) were treated with PAF (1 μM, 5 min) and then fMLP (100 nM, 10 min) (2). PI3K inhibitors were added before incubation: PI3Kγ-selective (AS605240, 3 μM) and PI3Kδ-selective (CAL-101, 100 nM).

Murine Neutrophils

Femoral bone marrow neutrophils were isolated by negative immunomagnetic selection from PI3Kδ-hyperactived E1020K heterozygote (PPL 80/2248 and P4802B8AC) (27), PI3Kδ-kinase-dead D910A homozygote (PPL 70/7661) (28), or PI3Kγ−/− (PPL 70/8100) mice, alongside age- and strain-matched wild-type control mice (E1020K/D910A: C57BL/6J, and PI3Kγ−/−: C57BL/E129). Neutrophils were resuspended in normoxic or hypoxic IMDM before treatment at 4 hours with cytochalasin B (5 μg/ml, 5 min) and then fMLP (10 μM, 10 min).

Protein Secretion

Neutrophil supernatant and plasma protein content were measured by ELISA, chemiluminescence immunoassay, or activity assay. Plasma NE- and PR3-specific fibrinogen cleavage products were measured (1, 29).

Neutrophil Secretome Preparation for Tandem Mass Tag–Mass Spectrometry

Neutrophils were resuspended in normoxic or hypoxic IMDM (4 h) containing ethylenediaminetetraacetic acid (1 mM) and sivelestat (10 μM) and treated with PAF and fMLP as above. Concentrated protein supernatants underwent tandem mass tag–mass spectrometry (TMT-MS).

Endothelial Cell Survival

Confluent human pulmonary artery endothelial cells (hPAECs) (Lonza) or human pulmonary microvascular endothelial cells (hPMECs) (Promocell) were treated with neutrophil supernatants, with or without alpha-1-antitrypsin (α1AT, 46 μg/ml, 10 min). Cell detachment of rhodamine phalloidin- and DAPI-stained fixed hPAECs was assessed by immunofluorescence. Viability and apoptosis of unfixed hPAECs or hPMECs was assessed by MTT assay or annexin V positivity.

Endothelial–Neutrophil Interaction

Confluent hPMECs were treated with neutrophil supernatants. At 4 hours, hPMECs were perfused with neutrophils (1 × 106 cells/ml, 4 min, 0.1 Pa shear stress) and neutrophil adhesion or transmigration assessed.

Statistical Analysis

Data were analyzed using GraphPad Prism version 9 software, reported as mean ± SEM from (n) independent experiments. Gaussian data were analyzed by t test or two-way ANOVA with Sidak’s correction. Non-Gaussian data were analyzed by Mann-Whitney test. TMT-MS–generated P values were adjusted by Benjamini-Hochberg procedure. A P value of less than 0.05 was considered statistically significant.

For further information, see Expanded Materials and Methods in the online supplement.

Patients Experiencing COPD Exacerbation Have Increased Circulating Evidence of Neutrophil Protease Activity

Although neutrophil proteases are implicated in COPD lung parenchymal destruction, the extent of systemic release of neutrophil granule proteins and their potential role in endothelial injury are unclear. We show that patients experiencing COPD exacerbation (see Table E1 in the online supplement) have higher plasma concentrations of fibrinogen cleavage products AαVal360 and AαVal541 than age- and sex-matched healthy control subjects (Figures 1A and 1B). These footprints specifically indicate increased activity of neutrophil azurophil granule proteases NE and proteinase 3, respectively, secreted on neutrophil activation (which may occur in the circulation, during adherence to vascular endothelium, or after migration into tissues) before inactivation by circulating antiproteases, such as α1AT. Given this systemic signature of enhanced neutrophil protease activity during COPD exacerbation, established evidence of pathological hypoxia during inflammation, and our previous results demonstrating hypoxia-augmented NE release from GM-CSF (granulocyte–macrophage colony–stimulating factor)–primed neutrophils (2), we next examined the ability of inflammatory mediators relevant to COPD and hypoxia to influence neutrophil degranulation.

Hypoxia Augments NE Release from PAF-primed Neutrophils in a PI3Kγ-Dependent Manner

Neutrophils treated with PAF and fMLP in combination (but not alone) released up to threefold more active NE when incubated under hypoxia (0.8% O2, 3 kPa) compared with normoxia (21 kPa) (Figure 2A). This enhanced secretion was not reversed by reoxygenation (Figure 2B). Given the known role of the PI3K pathway in the hypoxic upregulation of degranulation from GM-CSF–primed neutrophils, and the aberrant chemotaxis of neutrophils from patients with COPD, which could be corrected by PI3K inhibition (30), we explored whether inhibition of PI3K signaling pathways modulated the hypoxic response of PAF-primed neutrophils using PI3K isoform–selective small molecule inhibitors. PI3Kγ-selective inhibition abrogated the hypoxic uplift of NE release from PAF-primed neutrophils; this effect was not seen with the PI3Kδ-selective inhibitor (Figure 2C). These results were replicated in a cohort of patients with COPD (Table E2 and Figure E1). As PI3Kγ inhibition also markedly inhibited NE release from stimulated normoxic cells, we further explored whether PI3K signaling was simply essential for overall degranulation or had a specific role in hypoxia-mediated degranulation, using transgenic mice with abolished or enhanced activity of PI3Kγ/δ isoforms. As PAF did not elicit a detectable priming response in murine neutrophils (data not shown), these cells were treated with cytochalasin B and fMLP to liberate NE. The hypoxic increase in NE release from murine neutrophils deficient in PI3Kγ was abolished, with preserved ability to degranulate under normoxia (Figure 2D). In contrast, the hypoxic enhancement of NE release from murine neutrophils with either activating or kinase-dead PI3Kδ mutations was unaffected (Figures 2E and 2F). Together, these data indicate that the augmented degranulation observed under hypoxia from human or murine neutrophils requires PI3Kγ but not PI3Kδ activity.

Hypoxia Increases the Capacity for Neutrophil Supernatants to Damage Endothelial Cells

As patients with COPD suffer increased cardiovascular morbidity compared with healthy control subjects and display a footprint of increased circulating protease activity (Figure 1), we investigated whether hypoxia increases the potential for neutrophil-mediated endothelial damage. We incubated supernatants from normoxic versus hypoxic activated (PAF/fMLP) neutrophils with hPAEC and hPMEC monolayers in the presence or absence of α1AT and assessed cell integrity and survival. Supernatants from activated hypoxic neutrophils induced more endothelial detachment (Figures 3A and 3B) and death (Figures 3C, 3D, and E2) than their normoxic counterparts, which was not completely rescued by coincubation with α1AT (Figure 3D).

Hypoxia Differentially Regulates Protein Release from Activated Neutrophils

Because the antiprotease strategy did not completely mitigate neutrophil-induced endothelial damage (Figure 3D), and multiple neutrophil-derived granule products have potentially damaging actions, we investigated the effect of hypoxia on the total detectable proteome released by activated neutrophils. Mass spectrometry characterization of the normoxic versus hypoxic neutrophil secretome revealed clear separation by principal component analysis (Figure 4A). TMT-MS identified 1,245 proteins, 717 of which were present in all samples. Of these 717 proteins, 199 had a false discovery rate < 0.01, and 63 were differentially regulated (adjusted P value < 0.05) between normoxia and hypoxia (Figure 4B). Of these 63 proteins, 35 were more abundant in normoxic (Table 1) and 28 were increased in hypoxic (Table 2) neutrophil supernatants. The majority of proteins upregulated in hypoxic supernatants were granule proteins, whereas those more abundant in normoxic supernatants were predominantly cytoplasmic. However, some granule-associated proteins were increased in normoxic samples (e.g., leukocyte-specific protein 1), and certain cytoplasmic (e.g., cyclophilin A) and nuclear (e.g., histone H4) proteins were increased by hypoxia. Selected hypoxia-upregulated protein targets with the potential to play a role in endothelial damage were biochemically validated using supernatants from independent healthy donors. Concentrations of the azurophil granule protein resistin (Figure 4C), the specific granule protein NGAL (neutrophil gelatinase-associated lipocalin); (Figure 4D), and the cytoplasmic protein cyclophilin A (Figure 4E) were significantly elevated in hypoxic versus normoxic neutrophil supernatants.

Table 1. Proteins Significantly Increased in Normoxic Neutrophil Supernatants

AccessionDescriptionAdjusted P ValueFCLocation
Q5TCU8Tropomyosin β chain0.00111.855CYT
E7EX2914-3-3 Protein zeta/delta0.0153.844S/G
P52566Rho GDP-dissociation inhibitor 20.0043.802CYT
O00299Chloride intracellular channel protein 10.0302.918?
P1102178-kDa glucose-regulated protein0.0092.873CYT
P62993Growth factor receptor-bound protein 20.0042.478CYT
Q32MZ4Leucine-rich repeat flightless-interacting protein 10.0042.337NUC/CYT
P06737Glycogen phosphorylase, liver form0.0252.217?
P32942Intercellular adhesion molecule 30.0152.149S/G
O15144Actin-related protein 2/3 complex subunit 20.0312.086CYT
P06702Protein S100-A90.0132.084CYT
P522096-Phosphogluconate dehydrogenase, decarboxylating0.0131.950CYT
P33241Leukocyte-specific protein 10.0131.878S/G
P30740Leukocyte elastase inhibitor0.0061.813A
Q96C19EF-hand domain-containing protein D20.0271.794?
A6NIZ1Ras-related protein Rap-1b-like protein0.0091.749SV
Q29963HLA class I histocompatibility antigen, Cw-6 α chain0.0391.627PM
P6124740S ribosomal protein S3a0.0281.528?
P02042Hemoglobin subunit delta0.0491.492?
P08133Annexin A60.0351.458?
Q15907Ras-related protein Rab-11B0.0151.447G
P4678140S ribosomal protein S90.0441.438CYT
P46940Ras GTPase-activating-like protein IQGAP10.0491.411G
P39687Acidic leucine-rich nuclear phosphoprotein 32 family member A0.0431.358?

Definition of abbreviations: A = azurophil granules; CYT = cytoplasm; FC = fold change; G = gelatinase granules; NUC = nucleus; PM = plasma membrane; S = specific granules; SV = secretory vesicles.

Proteins significantly increased in supernatants from normoxic neutrophils (adjusted P value < 0.05), which were present in all 10 samples with a false-discovery rate < 0.01, are listed in order of the magnitude of the FC. Location data were compiled from Reference 50 and the Uniprot database ( For some proteins, the location within neutrophils is currently uncertain or unknown (?).

Table 2. Proteins Significantly Increased in Hypoxic Neutrophil Supernatants

AccessionDescriptionAdjusted P ValueFCLocation
P62805Histone H40.0043.587NUC
A6NC48ADP-ribosyl cyclase/cyclic ADP-ribose hydrolase 20.0252.468SV
O75083WD repeat-containing protein 10.0292.382CYT
Q92820γ-Glutamyl hydrolase0.0202.143S
A5A3E0POTE ankyrin domain family member F0.0152.074CYT
Q0VD83Apolipoprotein B receptor0.0251.993S/G
A0A087WXL1Folate receptor γ0.0291.877S
P11215Integrin α-M0.0041.872S
P16035Metalloproteinase inhibitor 20.0341.853G
X6R8F3Neutrophil gelatinase-associated lipocalin0.0041.842S
V9GYM3Apolipoprotein A-II0.0491.769?
G3V3D1Epididymal secretory protein E10.0071.765A
P10153Nonsecretory RNase0.0291.650?
P05107Integrin β-20.0181.580G
P01024Complement C30.0321.556?
P78324Tyrosine-protein phosphatase nonreceptor type substrate 10.0181.449SV
J3KNB4Cathelicidin antimicrobial peptide0.0391.446S/G
P62937Peptidyl-prolyl cis-trans isomerase A/Cyclophilin A0.0351.438CYT
P30086Phosphatidylethanolamine-binding protein 10.0491.406?
A0A075B6H6Ig kappa chain C region0.0301.358?
A0A075B6K9Ig lambda-2 chain C regions0.0491.305?

Definition of abbreviations: A = azurophil granules; CYT = cytoplasm; FC = fold change; G = gelatinase granules; NUC = nucleus; PM = plasma membrane; S = specific granules; SV = secretory vesicles.

Proteins significantly increased in supernatants from hypoxic neutrophils (adjusted P value < 0.05), which were present in all 10 samples with a false discovery rate < 0.01, are listed in order of the magnitude of the FC. Location data were compiled from Reference 50 and the Uniprot database ( For some proteins, the location within neutrophils is currently uncertain or unknown (?).

Investigation of Cytoplasmic and Nuclear Protein Secretion

As cyclophilin A is cytoplasmic rather than granule associated, we examined the release of neutrophil-derived microvesicles (NMVs), which contain components derived from parent cells, as a potential source of protein release in addition to degranulation. However, we found no difference in NMV numbers in hypoxic versus normoxic supernatants (Figure E3A) or in plasma from healthy patients versus patients with COPD (Figure E3B). Furthermore, there was no difference in the content of cyclophilin A between NMVs generated from normoxic versus hypoxic cells, and cyclophilin A was also detected in microvesicle-depleted neutrophil supernatants (Figures E3C and E3D).

Because the nuclear protein, histone H4, was increased in the hypoxic supernatant proteome (although other histones were not likewise increased), we also investigated the release of neutrophil extracellular traps (NETs), which release both nuclear and granule proteins into the extracellular space. However, there was no difference in NETosis from normoxic versus hypoxic neutrophils (Figure E4). Overall, our data do not support a contribution of NMVs or NETs to the differential spectrum of cytoplasmic or nuclear proteins released from neutrophils under hypoxia.

Hypoxic Neutrophils from Patients Experiencing COPD Exacerbation Release More Histotoxic Proteins

During COPD exacerbations (which are associated with excess cardiovascular morbidity), neutrophils are subject to intensified local and systemic hypoxia in addition to a markedly proinflammatory microenvironment. Because hypoxia enables even “healthy” neutrophils to release multiple proteins capable of causing endothelial damage, we studied the effect of hypoxia on neutrophils isolated from patients experiencing COPD exacerbation (Table E1). COPD neutrophils were not basally shape-changed (indicating no priming or activation) and responded identically to healthy control neutrophils after fMLP stimulation (Figure 5A). Furthermore, there was no difference in the release of NE from unstimulated neutrophils obtained from patients experiencing COPD exacerbation versus age- and sex-matched healthy control subjects under normoxic or hypoxic conditions (Figure 5B). Together, these data indicate that, in our cohort 1, circulating neutrophils from patients with COPD were not primed during exacerbations. Despite this, when incubated under hypoxia, stimulated COPD neutrophils released up to threefold more active NE than equivalent healthy control cells (Figure 5B). Likewise, secretion of selected granule and cytoplasmic protein targets identified by proteomics NGAL (Figure 5C) and cyclophilin A (Figure 5D) was increased 1.5- and 5-fold, respectively, from stimulated hypoxic neutrophils from patients with COPD versus healthy control subjects, with a similar pattern demonstrated for resistin release (Figure 5E). In contrast, although secretion of the azurophil granule protein MPO (myeloperoxidase) was consistently increased in hypoxia versus normoxia, it was not further enhanced when comparing COPD and healthy control neutrophils (Figure 5F).

Hypoxia Promotes Endothelial–Neutrophil Interaction Induced by COPD Patient Neutrophil Supernatants

To investigate whether neutrophils from patients with COPD have increased capacity for endothelial cell injury and/or activation, we applied supernatants from normoxic or hypoxic activated COPD versus healthy neutrophils to hPMEC monolayers and assessed neutrophil recruitment (rolling, adhesion, and transmigration) in a biologically relevant in vitro flow system (Figures 6A and 6B). Treatment with hypoxic COPD neutrophil supernatants resulted in a marked increase in neutrophil rolling and adhesion compared with both normoxic COPD supernatants and hypoxic healthy supernatants (Figure 6C).

Hypoxia May Synergize with Inflammatory Mediators to Promote Upregulation of Circulating Histotoxic Neutrophil Granule Proteins in Patients with COPD

Because hypoxia enhanced histotoxic protein release from COPD neutrophils (Figure 5), and supernatants from these cells promoted neutrophil–endothelial interaction (Figure 6), we examined whether there was a circulating signature of hypoxia-induced neutrophil protein secretion. We detected significantly increased concentrations of the neutrophil granule proteins NE (Figure 7A), MPO (Figure 7B), and NGAL (Figure 7C) in patients experiencing COPD exacerbation versus healthy control plasma (derived from an independent cohort 3) (Table E3), although there was no difference in the plasma content of resistin (Figure 7D). Despite the observation of increased release of the cytoplasmic protein cyclophilin A from isolated COPD neutrophils (Figure 5E), unexpectedly, the plasma content of cyclophilin A was higher for healthy control subjects than patients with COPD (Figure 7E). Biomarkers of vascular injury or activation ICAM-1 (intercellular adhesion molecule-1) (Figure 7F) and VCAM-1 (vascular cell adhesion molecule-1) (Figure 7G) and inflammation (Figures E5A and E5B) were increased in plasma from patients experiencing COPD exacerbation. In contrast, the COPD plasma content of angiogenesis biomarkers was predominantly unchanged compared with that of healthy control subjects (Figures E5C–E5J), suggesting a damage phenotype that may enhance endothelial–neutrophil interaction but without a corresponding increase in vascular regeneration ability.

Our work demonstrates that hypoxic neutrophils display a hypersecretory phenotype with enhanced capacity to both activate and injure cultured endothelial cells. Hypoxia-driven release of histotoxic proteins was observed from healthy donor neutrophils and translated to a pathophysiologically relevant cohort of patients experiencing COPD exacerbation, confirming even further augmented histotoxic protein secretion under hypoxia, together with higher circulating concentrations of selected neutrophil-derived proteins and a plasma signature of increased neutrophil protease activity and vascular injury.

Our data from human neutrophils treated with PI3K isoform-selective inhibitors and from transgenic murine cells support a nonredundant role for PI3Kγ in the hypoxic augmentation of NE release. Dysregulated PI3K signaling has previously been associated with COPD, with the impaired neutrophil migratory accuracy improved by PI3Kγ/δ inhibition (30), although we found no role for the δ isoform in hypoxic degranulation. Our results suggest that PI3Kγ is required for hypoxia-potentiated neutrophil degranulation, whether in response to tyrosine kinase– (2) or G-protein–coupled agonists, such as PAF. Consistent with a role for PI3Kγ-dependent PAF/hypoxia-mediated neutrophil degranulation in vascular insults, PI3Kγ/δ inhibition limited both PAF-induced hindlimb inflammation and ischemic cardiac infarct size (31), and PI3Kγ-null mice had improved cardiac recovery after ischemia (32). Hence, this signaling pathway could conceivably be targeted to mitigate neutrophil-mediated endothelial damage in COPD and other diseases underpinned by hypoxia, inflammation, and vascular damage.

Hypoxia enhances neutrophil degranulation, but, surprisingly, hypoxia-upregulated proteins identified by proteomic analysis did not segregate precisely with discrete granule populations. We have previously shown that hypoxia promotes differential secretion from eosinophil granules (26), with similar results reported using mast cells (33). Our results imply a comparable “differential degranulation” may be occurring from neutrophils in the setting of hypoxia. A limited number of studies analyzing the neutrophil secretome generated under normoxic conditions have shown variation in protein content according to the inciting stimulus (34), suggesting that the (potentially hypoxic) inflammatory environment can influence the precise composition of secreted granule proteins. This may also explain our observation that release of the azurophilic granule proteins NE, resistin, and MPO from hypoxic COPD neutrophils did not fully mirror each other.

In addition to enhanced degranulation, our proteomic data further suggest active secretion of selected cytoplasmic and nuclear proteins under hypoxia. We provide the first description of neutrophilic secretion of the proinflammatory cytoplasmic protein cyclophilin A, aligning to a previous study demonstrating its hypoxia-driven secretion from cardiac myocytes (35). Our data also demonstrated increased release of the nuclear protein histone H4 (although no other histones) in hypoxic supernatants. The source of this protein remains unclear, as we did not observe any difference in NETosis between normoxic and hypoxic neutrophils.

We have established that peripheral blood neutrophils from patients experiencing COPD exacerbation display markedly greater hypoxic release of NE, NGAL, and cyclophilin A relative to matched healthy control subjects. These proteins have been previously implicated in endothelial dysfunction or atherosclerosis (36, 37), with raised circulating levels of NGAL associated with cardiovascular events and with hypoxemia in COPD (38, 39). Although we found no relationship between admission or venesection oxygen levels and protein release (data not shown), this may have been confounded by prior exposure to supplemental oxygen. In keeping with a potential in vivo role for hypoxic augmentation of neutrophil degranulation, we found higher levels of the granule proteins NE, MPO, and NGAL and a footprint of increased NE and PR3 activity in COPD versus healthy plasma, aligning with recent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) studies, where NGAL in particular was associated with mortality (40). Our results suggest that enhanced neutrophil degranulation in the setting of hypoxia may contribute to systemic endothelial injury in patients experiencing COPD exacerbation and potentially other conditions, such as SARS-CoV-2 infection.

Primed circulating neutrophils with heightened potential for cellular damage have been identified in patients with COPD (21). However, we did not observe significant shape change or enhanced degranulation (both features of priming) from unstimulated COPD neutrophils, suggesting that the enhanced hypoxic release of granule proteins from these cells is not simply a consequence of priming and may “substitute” for this process in promoting damage. A potential explanation for enhanced protein release is a change in protein abundance in COPD neutrophils, with one study showing increased NE activity in COPD versus healthy neutrophil lysates (41). However, this result may represent a protease/antiprotease imbalance, as leukocyte elastase inhibitor was reduced, a finding consistent with our proteomic data and that may explain the discrepancy between these results and the NE activity assay.

Our study has some limitations. Although our patients and control subjects were age and sex matched, the patients had a significant smoking history and more comorbidities than the healthy volunteers. The patients with COPD were studied during exacerbations, a high-risk period for acute cardiovascular events (17), but with variation in terms of exacerbation etiology and severity; however, the use of two separate cohorts in different institutions mitigates this limitation. It is possible that acute or long-term medications could affect neutrophil function in our patient cohorts. For example, all patients with COPD in cohort 1 were taking inhaled combination corticosteroid and long-acting β-agonists and were treated with oral prednisolone during exacerbation with varying duration before venesection. However, although glucocorticoids are known to delay neutrophil apoptosis, this effect is lost in hypoxia (42), and inhaled corticosteroids have been shown not to affect neutrophil protein secretion (43). Given the variation in comorbidities (Table E4), and thus treatments, within our cohorts, there is unlikely to be any consistent impact on neutrophil function.

Although our results are predominantly consistent with previous studies, we note that despite reports of enhanced circulating levels of resistin and cyclophilin A in patients with COPD versus healthy control subjects (44, 45), we did not detect such an increase in our COPD cohort. These discrepant results may reflect differences in the timing of sampling, medications, patient heterogeneity, or obscuration of a neutrophil-specific signal, because both proteins have multiple cellular sources (46, 47). Both cyclophilin A and its receptor have been detected at high levels in atherosclerotic plaques (48). Hence, in COPD, cyclophilin A may already be membrane bound, preventing its detection in plasma. As patients with COPD have increased atherosclerotic burden, hypoxia within plaques may promote enhanced local release of cyclophilin A from neutrophils in this and other microenvironments. This would be consistent with our in vitro data and with a previous study demonstrating increased cyclophilin A expression in the lung tissue of patients with COPD compared with either smoking or nonsmoking control subjects (49).

Overall, our data demonstrate that patients experiencing COPD exacerbation have an enhanced footprint of circulating neutrophil protease activity and that neutrophils from these patients exhibit a hypoxia-driven hypersecretory phenotype with enhanced capacity for endothelial damage. We provide the first description of the ability of hypoxia to augment the secretion of histotoxic proteins from COPD neutrophils in vitro and have identified a corresponding increase in the plasma concentrations of selected granule proteins and markers of vascular injury in patients with COPD. Our findings may illuminate novel therapeutic targets in treatment-recalcitrant neutrophil-mediated inflammatory diseases such as COPD.

The authors thank Professor Klaus Okkenhaug and Dr. Anne-Katrien Stark (Department of Pathology, University of Cambridge) for kindly providing E1020K heterozygote and D910A homozygote mice; Dr. Helen Marriott (Department of Infection, Immunity, and Cardiovascular Disease, University of Sheffield) for kindly providing E1020K heterozygote mice; Dr. Len Stephens and Dr. Sabine Suire (Babraham Institute, Cambridge) for kindly providing PI3Kγ−/− mice; Dr. Mairi MacLeod, Professor Gavin Donaldson, and Professor Jadwiga Wedzicha (NHLI, Imperial College London) for facilitating access to blood samples from patients with COPD at the Imperial College London NIHR Biomedical Research Unit; Rosalind Simmons (Department of Medicine, University of Cambridge) for coordinating donor recruitment; Dr. Mike Deery and Dr. Marco Chiapello (University of Cambridge Centre for Proteomics) for TMT-MS; Keith Burling and NIHR Cambridge BRC Core Biochemistry Facility; Simon McCallum and NIHR Cambridge BRC Core Flow Cytometry; and all blood donors. Patients with COPD and healthy control subjects recruited from Birmingham were done so with the support of the NIHR Clinical Research Facility within UHB NHS Foundation Trust.

1. Carter RI, Ungurs MJ, Mumford RA, Stockley RA. Aα-Val360: a marker of neutrophil elastase and COPD disease activity. Eur Respir J 2013;41:3138.
2. Hoenderdos K, Lodge KM, Hirst RA, Chen C, Palazzo SG, Emerenciana A, et al. Hypoxia upregulates neutrophil degranulation and potential for tissue injury. Thorax 2016;71:10301038.
3. Nolte D, Steinhauser P, Pickelmann S, Berger S, Härtl R, Messmer K. Effects of diaspirin-cross-linked hemoglobin (DCLHb) on local tissue oxygen tension in striated skin muscle: an efficacy study in the hamster. J Lab Clin Med 1997;130:328338.
4. van Staveren S, Ten Haaf T, Klöpping M, Hilvering B, Tinnevelt GH, de Ruiter K, et al. Multi-dimensional flow cytometry analysis reveals increasing changes in the systemic neutrophil compartment during seven consecutive days of endurance exercise. PLoS One 2018;13:e0206175.
5. Mall MA, Harkema JR, Trojanek JB, Treis D, Livraghi A, Schubert S, et al. Development of chronic bronchitis and emphysema in β-epithelial Na+ channel-overexpressing mice. Am J Respir Crit Care Med 2008;177:730742.
6. Belton M, Brilha S, Manavaki R, Mauri F, Nijran K, Hong YT, et al. Hypoxia and tissue destruction in pulmonary TB. Thorax 2016;71:11451153.
7. Polosukhin VV, Cates JM, Lawson WE, Milstone AP, Matafonov AG, Massion PP, et al. Hypoxia-inducible factor-1 signalling promotes goblet cell hyperplasia in airway epithelium. J Pathol 2011;224:203211.
8. Joshi FR, Manavaki R, Fryer TD, Figg NL, Sluimer JC, Aigbirhio FI, et al. Vascular imaging with 18F-fluorodeoxyglucose positron emission tomography is influenced by hypoxia. J Am Coll Cardiol 2017;69:18731874.
9. McElvaney OJ, McEvoy NL, McElvaney OF, Carroll TP, Murphy MP, Dunlea DM, et al. Characterization of the inflammatory response to severe COVID-19 illness. Am J Respir Crit Care Med 2020;202:812821.
10. Kitchen E, Rossi AG, Condliffe AM, Haslett C, Chilvers ER. Demonstration of reversible priming of human neutrophils using platelet-activating factor. Blood 1996;88:43304337.
11. Bitencourt CS, Bessi VL, Huynh DN, Ménard L, Lefebvre JS, Lévesque T, et al. Cooperative role of endogenous leucotrienes and platelet-activating factor in ischaemia-reperfusion-mediated tissue injury. J Cell Mol Med 2013;17:15541565.
12. Shukla SD, Muller HK, Latham R, Sohal SS, Walters EH. Platelet-activating factor receptor (PAFr) is upregulated in small airways and alveoli of smokers and COPD patients. Respirology 2016;21:504510.
13. Cardini S, Dalli J, Fineschi S, Perretti M, Lungarella G, Lucattelli M. Genetic ablation of the fpr1 gene confers protection from smoking-induced lung emphysema in mice. Am J Respir Cell Mol Biol 2012;47:332339.
14. Stănescu D, Sanna A, Veriter C, Kostianev S, Calcagni PG, Fabbri LM, et al. Airways obstruction, chronic expectoration, and rapid decline of FEV1 in smokers are associated with increased levels of sputum neutrophils. Thorax 1996;51:267271.
15. Kuna P, Jenkins M, O’Brien CD, Fahy WA. AZD9668, a neutrophil elastase inhibitor, plus ongoing budesonide/formoterol in patients with COPD. Respir Med 2012;106:531539.
16. Ye C, Younus A, Malik R, Roberson L, Shaharyar S, Veledar E, et al. Subclinical cardiovascular disease in patients with chronic obstructive pulmonary disease: a systematic review. QJM 2017;110:341349.
17. Kunisaki KM, Dransfield MT, Anderson JA, Brook RD, Calverley PMA, Celli BR, et al.; SUMMIT Investigators. Exacerbations of chronic obstructive pulmonary disease and cardiac events: a post hoc cohort analysis from the SUMMIT randomized clinical trial. Am J Respir Crit Care Med 2018;198:5157.
18. Dinh-Xuan AT, Higenbottam TW, Clelland CA, Pepke-Zaba J, Cremona G, Butt AY, et al. Impairment of endothelium-dependent pulmonary-artery relaxation in chronic obstructive lung disease. N Engl J Med 1991;324:15391547.
19. Eickhoff P, Valipour A, Kiss D, Schreder M, Cekici L, Geyer K, et al. Determinants of systemic vascular function in patients with stable chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2008;178:12111218.
20. Shi Y, Yang S, Luo M, Zhang W-D, Ke Z-P. Systematic analysis of coronary artery disease datasets revealed the potential biomarker and treatment target. Oncotarget 2017;8:5458354591.
21. Oudijk E-JD, Gerritsen WBM, Nijhuis EHJ, Kanters D, Maesen BLP, Lammers J-WJ, et al. Expression of priming-associated cellular markers on neutrophils during an exacerbation of COPD. Respir Med 2006;100:17911799.
22. Lodge K, Hoenderdos K, Robbins A, Storisteanu D, Chilvers E, Li W, et al. Hypoxia upregulates PI3Kinase-dependent neutrophil degranulation and neutrophil-mediated tissue injury [abstract]. Thorax 2016;71:A27.
23. Lodge K, Hoenderdos K, Chilvers E, Li W, Condliffe A. Hypoxia upregulates PI3Kinase-dependent neutrophil degranulation and neutrophil-mediated tissue injury [abstract]. Am J Respir Crit Care Med 2017;195:A7058.
24. Lodge K, Hoenderdos K, Robbins A, Chilvers E, Li W, Condliffe A. Hypoxia drives neutrophil-mediated endothelial damage in COPD [abstract]. Thorax 2017;72:A69A70.
25. Haslett C, Guthrie LA, Kopaniak MM, Johnston RB Jr, Henson PM. Modulation of multiple neutrophil functions by preparative methods or trace concentrations of bacterial lipopolysaccharide. Am J Pathol 1985;119:101110.
26. Porter LM, Cowburn AS, Farahi N, Deighton J, Farrow SN, Fiddler CA, et al. Hypoxia causes IL-8 secretion, Charcot Leyden crystal formation, and suppression of corticosteroid-induced apoptosis in human eosinophils. Clin Exp Allergy 2017;47:770784.
27. Stark A-K, Chandra A, Chakraborty K, Alam R, Carbonaro V, Clark J, et al. PI3Kδ hyper-activation promotes development of B cells that exacerbate Streptococcus pneumoniae infection in an antibody-independent manner. Nat Commun 2018;9:3174.
28. Okkenhaug K, Bilancio A, Farjot G, Priddle H, Sancho S, Peskett E, et al. Impaired B and T cell antigen receptor signaling in p110δ PI 3-kinase mutant mice. Science 2002;297:10311034.
29. Newby PR, Crossley D, Crisford H, Stockley JA, Mumford RA, Carter RI, et al. A specific proteinase 3 activity footprint in α1-antitrypsin deficiency. ERJ Open Res 2019;5:00095-02019.
30. Sapey E, Stockley JA, Greenwood H, Ahmad A, Bayley D, Lord JM, et al. Behavioral and structural differences in migrating peripheral neutrophils from patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2011;183:11761186.
31. Doukas J, Wrasidlo W, Noronha G, Dneprovskaia E, Fine R, Weis S, et al. Phosphoinositide 3-kinase γ/δ inhibition limits infarct size after myocardial ischemia/reperfusion injury. Proc Natl Acad Sci USA 2006;103:1986619871.
32. Alloatti G, Levi R, Malan D, Del Sorbo L, Bosco O, Barberis L, et al. Phosphoinositide 3-kinase γ-deficient hearts are protected from the PAF-dependent depression of cardiac contractility. Cardiovasc Res 2003;60:242249.
33. Möllerherm H, Branitzki-Heinemann K, Brogden G, Elamin AA, Oehlmann W, Fuhrmann H, et al. Hypoxia modulates the response of mast cells to Staphylococcus aureus infection. Front Immunol 2017;8:541.
34. Snäll J, Linnér A, Uhlmann J, Siemens N, Ibold H, Janos M, et al. Differential neutrophil responses to bacterial stimuli: Streptococcal strains are potent inducers of heparin-binding protein and resistin-release. Sci Rep 2016;6:21288.
35. Seko Y, Fujimura T, Taka H, Mineki R, Murayama K, Nagai R. Hypoxia followed by reoxygenation induces secretion of cyclophilin A from cultured rat cardiac myocytes. Biochem Biophys Res Commun 2004;317:162168.
36. Tian-Tian Z, Jun-Feng Z, Heng G. Functions of cyclophilin A in atherosclerosis. Exp Clin Cardiol 2013;18:e118e124.
37. Wen G, An W, Chen J, Maguire EM, Chen Q, Yang F, et al. Genetic and pharmacologic inhibition of the neutrophil elastase inhibits experimental atherosclerosis. J Am Heart Assoc 2018;7:e008187.
38. Lindberg S, Pedersen SH, Mogelvang R, Jensen JS, Flyvbjerg A, Galatius S, et al. Prognostic utility of neutrophil gelatinase-associated lipocalin in predicting mortality and cardiovascular events in patients with ST-segment elevation myocardial infarction treated with primary percutaneous coronary intervention. J Am Coll Cardiol 2012;60:339345.
39. Eagan TM, Damås JK, Ueland T, Voll-Aanerud M, Mollnes TE, Hardie JA, et al. Neutrophil gelatinase-associated lipocalin: a biomarker in COPD. Chest 2010;138:888895.
40. Abers MS, Delmonte OM, Ricotta EE, Fintzi J, Fink DL, de Jesus AAA, et al.; NIAID COVID-19 Consortium. An immune-based biomarker signature is associated with mortality in COVID-19 patients. JCI Insight 2021;6:e144455.
41. Shahriary A, Mehrani H, Ghanei M, Parvin S. Comparative proteome analysis of peripheral neutrophils from sulfur mustard-exposed and COPD patients. J Immunotoxicol 2015;12:132139.
42. Marwick JA, Dorward DA, Lucas CD, Jones KO, Sheldrake TA, Fox S, et al. Oxygen levels determine the ability of glucocorticoids to influence neutrophil survival in inflammatory environments. J Leukoc Biol 2013;94:12851292.
43. Culpitt SV, Maziak W, Loukidis S, Nightingale JA, Matthews JL, Barnes PJ. Effect of high dose inhaled steroid on cells, cytokines, and proteases in induced sputum in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1999;160:16351639.
44. Kumor-Kisielewska A, Kierszniewska-Stępień D, Pietras T, Kroczyńska-Bednarek J, Kurmanowska Z, Antczak A, et al. Assessment of leptin and resistin levels in patients with chronic obstructive pulmonary disease. Pol Arch Med Wewn 2013;123:215220.
45. Zhang M, Tang J, Yin J, Wang X, Feng X, Yang X, et al. The clinical implication of serum cyclophilin A in patients with chronic obstructive pulmonary disease. Int J Chron Obstruct Pulmon Dis 2018;13:357363.
46. Park HK, Ahima RS. Resistin in rodents and humans. Diabetes Metab J 2013;37:404414.
47. Bukrinsky M. Extracellular cyclophilins in health and disease. Biochim Biophys Acta 2015;1850:20872095.
48. Wang C, Jin R, Zhu X, Yan J, Li G. Function of CD147 in atherosclerosis and atherothrombosis. J Cardiovasc Transl Res 2015;8:5966.
49. Hu R, Ouyang Q, Dai A, Tan S, Xiao Z, Tang C. Heat shock protein 27 and cyclophilin A associate with the pathogenesis of COPD. Respirology 2011;16:983993.
50. Lominadze G, Powell DW, Luerman GC, Link AJ, Ward RA, McLeish KR. Proteomic analysis of human neutrophil granules. Mol Cell Proteomics 2005;4:15031521.
Correspondence and requests for reprints should be addressed to Katharine M. Lodge, M.B. B.Chir., Ph.D., National Heart and Lung Institute, Imperial College London, ICTEM Building, Level 5 Vascular Science, Du Cane Road, London W12 0NN, UK. E-mail: .

*Co–senior authors.

Supported by Wellcome Trust Research Training Fellowship 102706/Z/13/Z (K.M.L.), British Medical Association Foundation for Medical Research Josephine Lansdell 2017 grant (K.M.L.), Academy of Medical Sciences Starter Grant for Clinical Lecturers SGL024\1086 (K.M.L.), Cambridge National Institute for Health Research Biomedical Research Centre–GlaxoSmithKline Experimental Medicine Initiative Clinical Research Training Fellowship (A.V.), British Heart Foundation project grant PG/19/75/34686 (K.P.), National Institute for Health Research Imperial Biomedical Research Centre, Medical Research Council, and Imperial College National Heart and Lung Institute Foundation.

Author Contributions: K.M.L., E.R.C., W.L., and A.M.C. conceptualized the project and designed the methodology. K.M.L. recruited patients, collected samples, performed experiments, and conducted data analysis. A.V., B.L., M.L., Z.T., P.R.N., and D.A.-J. performed experiments. C.E.G., K.B.R.B., and E.S. recruited patients and collected plasma samples. K.P., V.C.R., R.A.S., E.S., C.S., and A.S.C. provided resources and supervised experiments. W.L. and A.M.C. supervised the project. K.M.L., W.L., and A.M.C. wrote the original manuscript. K.M.L., A.S.C., E.R.C., W.L., and A.M.C. revised the manuscript. All authors reviewed and edited the manuscript.

This article has a related editorial.

This article has an online supplement, which is accessible from this issue’s table of contents at

Originally Published in Press as DOI: 10.1164/rccm.202006-2467OC on January 19, 2022

Author disclosures are available with the text of this article at

Comments Post a Comment

New User Registration

Not Yet Registered?
Benefits of Registration Include:
 •  A Unique User Profile that will allow you to manage your current subscriptions (including online access)
 •  The ability to create favorites lists down to the article level
 •  The ability to customize email alerts to receive specific notifications about the topics you care most about and special offers
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record