Neutrophil phenotypes in bronchial airways differentiate single from dual responding allergic asthmatics

Abstract

Introduction

Allergic asthmatics with both an early (EAR) and a late allergic reaction (LAR) following allergen exposure are termed ‘dual responders’ (DR), while ‘single responders’ (SR) only have an EAR. Mechanisms that differentiate DR from SR are largely unknown, particularly regarding the role and phenotypes of neutrophils. Therefore, we aimed to study neutrophils in DR and SR asthmatics.

Methods

Thirty‐four allergic asthmatics underwent an inhaled allergen challenge, samples were collected before and up to 24 h post‐challenge. Cell differentials were counted from bronchial lavage, alveolar lavage and blood; and tissue neutrophils were quantified in immune‐stained bronchial biopsies. Lavage neutrophil nuclei lobe segmentation was used to classify active (1–4 lobes) from suppressive neutrophils (≥5 lobes). Levels of transmigration markers: soluble (s)CD62L and interleukin‐1Ra, and activity markers: neutrophil elastase (NE), DNA‐histone complex and dsDNA were measured in lavage fluid and plasma.

Results

Compared with SR at baseline, DR had more neutrophils in their bronchial airways at baseline, both in the lavage (p = .0031) and biopsies (p = .026) and elevated bronchial neutrophils correlated with less antitransmigratory IL‐1Ra levels (r = −0.64). DR airways had less suppressive neutrophils and more 3‐lobed (active) neutrophils (p = .029) that correlated with more bronchial lavage histone (p = .020) and more plasma NE (p = .0016). Post‐challenge, DR released neutrophil extracellular trap factors in the blood earlier and had less pro‐transmigratory sCD62L during the late phase (p = .0076) than in SR.

Conclusion

DR have a more active airway neutrophil phenotype at baseline and a distinct neutrophil response to allergen challenge that may contribute to the development of an LAR. Therefore, neutrophil activity should be considered during targeted diagnosis and bio‐therapeutic development for DR.

Allergic asthmatics underwent an inhaled allergen challenge with samples taken from the airway and blood at baseline and after the challenge. Dual responding allergic asthmatics (DR) had more 2‐ to 3‐lobed (known as ‘active’) and less ≥5‐lobed (‘suppressive’) airway neutrophils correlating with less circulating IL‐1Ra than single responders (SR) at baseline. Following an inhaled allergen challenge, DR had more circulating NETosis factors than SR during the early allergic reaction phase (EAR).

Key Messages.

• Dual responding allergic asthmatics have more baseline airway neutrophils with an active morphology than single responders.

• Increased airway neutrophils in dual responders were associated with less antimigration signalling rather than more chemoattraction.

• Post‐challenge, dual responders release an active neutrophil response in the blood earlier than in single responders.

1. INTRODUCTION

Approximately 50% of allergic asthma patients present with both an early (EAR) and a late allergic reaction (LAR) to inhaled allergen and are termed ‘dual responders’ (DR), while those who only have an EAR are termed ‘single responders’ (SR). The LAR, occurring between 4 and 8 hours (h) after allergen exposure, is associated with more extensive airway inflammation and can last up to several weeks. ¹ Therefore, compared with SR, DR often reflects a more severe asthma clinical phenotype. ² Despite having a reproducible type‐2 immune response to inhaled allergen in both SR and DR subgroups, ¹ no baseline biomarkers related to the type‐2 immune response can presently predict the airway response in an individual (e.g. ³ ). In contrast, some studies have shown differences in neutrophil‐related biomarkers ⁴ , ⁵ and less neutrophils in the blood between DR and SR. ⁶ However, the neutrophil‐related differences within the airway compartments and how they may contribute to the dual allergic response have not been investigated.

The roles of neutrophils in asthma were attributed to the release of damage markers that attract eosinophils, ⁷ tissue repair ⁷ , ⁸ and subsequent apoptosis. ⁹ Yet, this may not be the case in a diseased state as an increased neutrophil presence in airways is often related to more severe asthma, ¹⁰ uncontrolled asthma ¹¹ or the use of oral corticosteroids. ¹² Neutrophils are also one of the first innate immune cells to be recruited to the airways of allergic asthmatics and have been found to be needed for inducing airway inflammation, play role in exacerbating allergic inflammation, but also, be dampened by type‐2 allergic inflammation ¹³ —though the mechanisms behind these roles are largely unknown.

Chemoattractants of neutrophil migration into the tissue include interleukin‐17 (IL‐17), primarily released by T‐helper 17 cells ¹⁴ and IL‐8, released by activated monocytes/macrophages and endothelial/epithelial cells. ¹⁵ To regulate chemoattraction, IL‐1 receptor agonist (IL‐1Ra) is released by many cells, including monocytes and neutrophils, to reduce neutrophil recruitment to inflamed tissue—avoiding excessive self‐damage. ¹⁶ , ¹⁷ A major step in the migration of leukocytes, including neutrophils, from circulation to the airways is the shedding of CD62L (also called L‐selectin). ¹⁸ During inflammation, neutrophils can also undergo a process called NETosis, releasing neutrophil extracellular traps (NETs) that contain proinflammatory neutrophil elastase (NE), myeloperoxidase (MPO), self dsDNA and histone proteins. ¹⁹ The NETosis process can cause damage to the surrounding tissue and has been reported to drive exacerbation severity in a mouse asthma model. ¹⁹ In general, neutrophils responding to tissue inflammation with NETosis are usually healthy, but a prolonged neutrophil response in the tissue can be concerning.

In contrast, a newly defined suppressive phenotype of CD62L(dim) neutrophils, that is speculated to be maturing directly from unsegmented to hypersegmented nuclei (i.e. five or more lobes), ²⁰ , ²¹ , ²² has been shown to appear in the blood as early as 3 h post‐challenge. ²³ Recently, we reported an increase in blood CD62L(dim) neutrophils 23 h post‐challenge from the same blood samples as used in the present study. ²⁴ However, it is more difficult to determine whether airway CD62L(dim) neutrophils originated as naturally occurring ‘suppressive’ CD62L(dim) in circulation or whether they were activated CD62L(high) neutrophils that have shed CD62L to transmigrate (i.e. move out of circulation between the endothelial cells) in response to an allergen challenge. ²⁵ Presently, the best way to distinguish suppressive from active CD62L(dim) neutrophils in the airways is to assess the segmentation of neutrophil nuclei, where CD62L(dim) suppressive neutrophils are hypersegmented (i.e. ≥5 lobes), ²¹ while active neutrophils commonly appear unsegmented (=1) to 4 lobes. ²²

Due to the prolonged allergic response in DR, we hypothesised that there would be an altered neutrophil involvement in DR compared with SR. Therefore, we assessed the numbers and morphology of airway neutrophils and their related biomarkers in the airways and the plasma in DR versus SR before and after an inhaled bronchial allergen challenge.

2. MATERIALS AND METHODS

2.1. Subjects and study design

All subjects (n = 34) had mild to moderate allergic asthma and all were nonsmokers without or with a low dose of inhaled corticosteroids (ICS, max 400 µg of budesonide). All subjects signed a written informed consent and the Regional Ethics Review Board in Lund, Sweden, approved the study (2012/800). Each individual completed lung physiology measures and blood sampling before and at several time points following an inhaled bronchial allergen challenge. When FEV₁ dropped ≥20% compared with baseline, the challenge was considered completed (starting at the 0 h time point). Subjects included 19 SR and 15 DR, and DR were defined by the second drop of ≥12% in FEV₁ from baseline that was maintained between 4 and 8 h post‐challenge (defined by Stenberg et al, 2017), ²⁶ for subject characteristics see Table 1. The period of 0–2 h post‐challenge was considered the ‘early phase’ and from 4 to 8 h was considered the ‘late phase’. Study procedures including lung physiology (with lung reactance given as X5 and AX measured by impulse oscillometry referred to in this manuscript) and related findings have been previously reported. ²⁶ The study design for relevant sample collections and timepoints is outlined in Figure S1. Briefly, a bronchoscopy was performed before and 24 h post‐challenge, and blood was collected before and 0, 0.5, 1, 2, 4, 6, 8 and 23 h post‐challenge as previously described. ²⁷

TABLE 1.

Patient characteristics (median ± IQR). Separated into single responders (SR) and dual responders (DR) for all patients. Allergen extract used was purified and standardized to eliminate the effects of LPS (ALK‐Abelló, Hørsholm, Denmark)

	All patients
	SR (n = 19)	DR (n = 15)	pValue
Sex (Male/Female)	11/8	6/9	.9
Age (years)	27 (26–41)	24 (22–30)	.036
BMI (kg/m2)	23 (22–26)	24 (23–26)	.7
Regular use of ICS (n)	6	10	.5
Duration of asthma (years)	19.5 (10–24)	15.0 (13–21)	.9
ACT (score)	22 (20–24)	22 (20–24)	.5
Methacholine PD20 (µg)	208 (104–550)	228 (174–917)	.3
Number of sensitizations	5 (4–6)	4 (3–6)	.3
Allergen inhaled (n)
Cat	10	8
Horse	4	2	N/A
Birch	2	2
D. pteronyssinus	2	1
Grass	1	2
FEV1% max drop 0–2 h after challenge	22.5 (21.2–24.9)	23.5 (21.1–26.3)	.5
FEV1% max drop 4–8 h after challenge	7.2 (4.7–8.7)	20.1 (14.8–24.0)	<.0001

Statistical significance between SR and DR of all patients was determined by nonparametric Mann–Whitney test (p < .05 is bolded). Body mass index (BMI), inhaled corticosteroids (ICS), asthma control test (ACT) and forced expiratory volume in one second (FEV1).

2.2. Bronchial and alveolar lavage processing

A subgroup (SR, n = 10, DR n = 8) underwent a bronchoscopy (see Table S1 for subgroup characteristics) with bronchial biopsies, bronchial lavage and alveolar lavage before and 24 h post‐ challenge. Sterile phosphate‐buffered saline (PBS) (3 × 50 ml) was instilled through a bronchoscope into the middle lobe and the first 50 ml was collected as the bronchial fraction and the remaining 100 ml was recovered as the alveolar fraction (herein referred to as bronchial lavage and alveolar lavage, respectively) similar to a previous study. ²⁸ Cells were separated from the lavage fluid by centrifugation, cytospun and either stained with Kwick–Diff (Shandon) or fixed with formaldehyde. See Supplementary Methods for a detailed description of processing.

2.3. Neutrophil nuclei segmentation differentiation

Scans (40×) of the Kwick–Diff stained lavage cytospins were assessed using 100 randomly selected neutrophils per slide identified in ImageScope (Lecia Biosystems) by a qualified technician. Each neutrophil was individually categorised according to the number of nucleus lobes, i.e. as unsegmented or two, three or four lobes or a hypersegmented nucleus of five or more lobes. The nucleus lobes had to be clearly separated to be counted as segmented (see examples in Figure S2). Percentage of each segmented neutrophil type was used to calculate the number of each segmented neutrophil subset per mL of lavage.

2.4. Blood sampling

EDTA‐treated whole blood was sampled before and at multiple time points post‐challenge and used for plasma collection and a subgroup (Table S1) was sent for a differential count of WBC at the accredited Clinical Chemistry routine lab at Skåne's University Hospital in Lund, Sweden. See Supplementary methods for processing and analysis.

2.5. Enzyme‐linked immunosorbent assay

DuoSet ELISA kits from R&D systems were used as per the manufacturer's instructions to assess all the plasma time points for CD62L (R&D Systems) and all the plasma time points and lavages for NE (R&D Systems). The cell death detection ELISA (Roche) was used to detect cytoplasmic histone‐associated DNA fragments (herein termed as ‘histones’) in lavage fluid and plasma as per the manufacture's instructions. See Supplementary methods for processing.

2.6. Protein concentrations using multiplex Luminex^®

The Magnetic Luminex^® Assay multiplex kit (bio‐techne) was used to quantify IL‐17, IL‐8 and IL‐1Ra in baseline, 6 h and 23 h plasma and lavage fluid and to quantify CD62L in lavage fluid. See Supplementary methods for reagent details and processing.

2.7. Qubit dsDNA assay

The Qubit^® dsDNA HS assay kit (Life technologies) was used to quantify dsDNA in lavage fluid and plasma samples as per the manufacturer's instructions.

2.8. Immunohistochemical staining in biopsy specimens and cytospins

Biopsies were fixed in paraffin, and prestained using an antimacrophage cocktail of anti‐CD68 (Dako) and anti‐CD163 (LEICA) (to avoid nonspecific analysis of macrophage‐based myeloperoxidase, MPO), ²⁹ labelled with DAB precipitate, then stained for neutrophils using anti‐MPO (Dako) (labelled with vina green). Slides were digitalized and all the viable airway tissue was captured using ImageScope. Images were processed to generate the percentage of neutrophil green area normalised to the total area of tissue.

A few alveolar cytospins, (5 paired samples of baseline and 24 h), were stained for CD62L (coloured with DAB) and counterstained with hematoxylin and eosin. Images of mixed lavage cells were processed using colour thresholds and cell count to produce an average CD62L stain area per cell. See Methods S1 for details on staining protocols and quantifications.

2.9. Statistics

Statistics outlined in each figure legend were calculated from GraphPad Prism (version 9). Briefly, the Mann–Whitney U test was used to compare DR and SR; the Wilcoxon signed‐rank test was used to compare paired timepoints (of 2–3 timepoints); the 2‐way ANOVA corrected for multiple comparisons was used to compare more than 3 paired timepoints; and the Spearman's rank analysis was used for correlations. See Supplementary Methods for more details.

3. RESULTS

3.1. Dual responders have more airway neutrophils at baseline than single responders

At baseline, DR presented with significantly more neutrophils per mL in the bronchial lavage than in SR (p = .0031). In general, each bronchial lavage had consistently more neutrophils than its paired alveolar lavage (p = .0078; Figure 1A). DR also had more macrophages (p = .021) in the baseline bronchial lavage than in SR. No other differences in the cell populations were observed between SR and DR per mL lavage (see Table S2 for all medians and p values). The total number of cells per total bronchial lavage volume also had more neutrophils in DR than in SR (p = .0028, Table S3)—but not macrophages. Both the increased numbers of baseline bronchial neutrophils and macrophages correlated with an increasing maximum decrease in FEV₁% during the late phase (Figure 1B). For the complete matrix of correlations between the decrease in FEV₁% drop during the early (1–2 h) and the late (4–8 h) phase and baseline bronchial cells, see Table S4.

FIGURE 1.

Lung lavage cell differentiation for dual responders (□DR) and single responders (●SR). Data compare SR and DR at baseline (BL) and 24 h post‐challenge. (A) Number of the different bronchial and alveolar lavage cells per mL of lavage, differentiated to (i) macrophages, (ii) neutrophils, (iii) lymphocytes and (iv) eosinophils, n = 6–10. (B) Both bronchial (i) macrophages and (ii) neutrophils correlate with the late‐phase per cent drop in FEV1

The increased number of neutrophils in the DR airways at baseline was confirmed by staining for neutrophil‐specific MPO in bronchial biopsies (p = .026; Figure 2A, B). The number of neutrophils in the tissue at baseline correlated with increased lung elasticity (measured as airway reactance and might reflect more peripheral obstruction ³⁰ ) at both baseline and during the late phase (Figure 2C, D—see Table S5 for p values).

FIGURE 2.

Neutrophils in bronchial biopsies. (A) Representative images of stained biopsies, in (i) single responders (SR) and (ii) dual responders (DR), at baseline (BL) stained for neutrophil MPO (green, indicated by black arrows). (B) Number of neutrophil cells/mm² of the tissue area, n = 5–6. Biopsy neutrophil numbers at baseline were correlated with peripheral reactance readings (Ci) X5 and (Di) AX from baseline and 23 h after challenge (solid lines) including the early‐phase (dotted lines) and late‐phase times (dashed lines). Both C and D contain inserts of (ii) baseline to baseline data with the same axis as the overall graph

3.2. Airway neutrophils from dual responders have less nuclei segmentation than single responders.

To differentiate the more active (1‐ to 4‐lobed) neutrophils from the suppressive (hypersegmented) neutrophils in the lavages, we assessed the proportion of each lobed neutrophil (see examples in Figure 3A). The majority of the bronchial lavage neutrophils in DR were unsegmented (p = .0027), 2‐lobed (p = .0047) or 3‐lobed (p = .029) as compared to SR at baseline (Figure 3B). DR also maintained significantly more 3‐lobed neutrophils than SR at 24 h post‐challenge (p = .042). There was no difference in the number of 4‐lobed neutrophils between DR and SR at either time point. As numbers of hypersegmented neutrophils were very low, we merged the bronchial and alveolar counts per subject to compare total lavage occurrences (i.e. yes or no) of hypersegmented neutrophils. Overall, DR had less occurrences of hypersegmented neutrophils in the combined lavage (2/11 occurrences or 18%) than in SR (12/19 occurrences or 63%; Figure 3C).

FIGURE 3.

Morphological assessment of nuclei segmentation of lavage neutrophils. Data compare single responders (SR) and dual responders (DR) at baseline (BL) and 23 h post‐challenge. (A) Representative images of the different segmented neutrophil nuclei. (B) Number of neutrophils with each type of segmentation/ml of lavage, n(SR) = 8, n(DR) = 5. (C) Hypersegmented neutrophils (H‐neut) occurrence using combined alveolar + bronchial lavage cell numbers. (D) Correlations of 2‐ and 3‐lobed neutrophils to the number of sensitisations

The total number of alveolar neutrophils at baseline correlated with the number of sensitizations, of these specifically, the 2‐lobed and 3‐lobed neutrophils correlated with the number of sensitizations to different allergens (Figure 3D) when analysing all individuals (i.e. SR + DR), but this correlation did not occur with the hypersegmented neutrophils (see Table S6).

3.3. Less CD62L in bronchial lavage from dual responders than single responders.

Since there is a known relationship between the leukocyte migration into the airways and CD62L, we assessed the CD62L levels in the airways (Figure 4). At baseline, soluble (s)CD62L levels in the bronchial lavage fluid were significantly lower in DR than in SR (p = .021, Figure 4A). To check whether this was due to less shedding in the lavage or whether the lavage leukocytes just had less CD62L on their surface overall, a few lavage slides were stained for CD62L on the cell surface (Figure 4B). When quantified for average stain area per cell, DR tended to have less CD62L in the cell surface than in SR at baseline, but samples were too few to statically confirm that DR leukocytes were CD62L(dim) (Figure 4C).

FIGURE 4.

CD62L levels in lung lavage and on alveolar cells. Data compare single responders (●SR) and dual responders (□DR) at baseline (BL) and 24 h post‐challenge. (A) ELISA of sCD62L in the lavage, n(SR) = 10, n(DR) = 8. (B) Representative image of CD62L stain (brown) on alveolar lavage cells. (C) CD62L stain is here as area of stain per cell, n(DR) =3, n(SR) = 2

3.4. Neutrophils comprise a large proportion of blood leukocytes that increase during the late phase

To assess the systemic immune response following an allergen challenge, WBC was differentiated from a subgroup of the subjects (SR: n = 7 and DR: n = 10) with similar characteristics as the whole study group (Table 1 and Table S1). There was a significant increase in the total blood leukocyte number from baseline to 4 h post‐challenge (p = .0005), specifically in monocytes (p = .0010), lymphocytes (p = .0010) and neutrophils (p = .0005), but not in eosinophils nor basophils (Figure 5A). An increase at 4 h was expected as this is in line with previous reports. ³¹ This was followed by a significant decrease in the total blood leukocyte number from 4 to 23 h post‐challenge (p = .0039), specifically in monocytes (p = .016), lymphocytes (p = .0039) and neutrophils (p = .016), but again, not in eosinophils nor basophils. There were no differences in total WBC counts between SR and DR at any time point (Figure S3).

FIGURE 5.

Blood cell differentiation and plasma sCD62L levels. (A) Number of WBC/L of blood at baseline (BL), 4 h and 23 h post‐challenge, n(BL) = 17, n(4h) = 12, n(23h) = 13. (B) Per cent of neutrophils relative to the total number of leukocytes at each time point, n = 6–10. (C) ELISA of sCD62L in plasma, n(SR) = 18; n(DR) = 12. (D) Correlation of BL and 23 h blood neutrophils/L and (E) of BL and 23 h blood monocytes/L to sCD62L levels in plasma at 0.5 and 23 h post‐challenge

The proportion of neutrophils among the blood leukocytes increased at 4 h in DR (p = .0050) but not in SR (Figure 5B). Concurrently, the proportion of blood lymphocytes and eosinophils dropped in DR but not SR (Figure S4). From 4 to 23 h post‐allergen challenge, the proportion of neutrophils decreased significantly (DR: p = .0018; SR: p = .023) to significantly lower numbers than at baseline (DR: p = .0078; SR: p = .063). There was no change in the proportion of monocytes and basophils throughout the study (Figure S4); however, the blood monocyte and neutrophil numbers did correlate at baseline and 4 h post‐allergen challenge, but no longer after 23 h (Figure S5).

3.5. Dual responders present with less leukocyte‐migratory sCD62L in plasma during the late phase than in single responders

Since there is a known relationship between the leukocyte migration into the airways and the shedding of CD62L, we assessed CD62L levels in the plasma (Figure 5C). Overall, DR had significantly less sCD62L than SR at all time points (p = .0054), markedly during the late phase at 6 h (p = .0076) and 8 h (p = .033) post‐challenge. Higher concentrations of blood monocytes and neutrophils at baseline correlated with higher sCD62L just 30 minutes post‐challenge (Figure 5D, E) but only the neutrophils became inversely correlated with sCD62L levels—highlighting their strong relationship to shedding CD62L and migrating out of circulation as previously reported. ³² For the complete matrix of correlations between sCD62L and blood leukocytes in this study, see Table S7.

3.6. Lower levels of IL‐1Ra (but not more IL‐8 or IL‐17) in the plasma correlate with more bronchial neutrophils at baseline

Since DR had significantly less sCD62L levels in the plasma, this may indicate differences in leukocyte migration patterns, we assessed the neutrophil chemoattractants IL‐8 and IL‐17 and neutrophil antirecruitment IL‐1Ra. In plasma, DR had less IL‐8 at baseline (p = .029) but overall all IL‐8 plasma readings showed low values ³³ , ³⁴ (Figure 6Ai). Plasma IL‐17 levels tended to increase in DR from baseline to 23 h (p = 0.09, paired Wilcoxon test) (Figure 6Bi). DR tended to have less IL‐1Ra (median: 219 pg/mL) than SR (314 pg/mL) throughout with SR increasing IL‐1Ra (BL‐6h p = .019; Figure 6Ci). Meanwhile, the lavage levels of IL‐8, IL‐17 and IL‐1Ra were similar in both DR and SR (Figure 6Aii, Bii, Cii). There was a significant correlation between IL‐8 and IL‐17 in alveolar lavage showing the chemoattractant condition in the lower airways was consistent at baseline (Figure S6A).

FIGURE 6.

Levels of IL‐8, IL‐17 and IL‐1Ra within the plasma and lung lavage. Luminex assay for single responders (●SR) and dual responders (□DR) at baseline (BL), 6 and 23 h post‐challenge in the plasma for (Ai) IL‐8, (Bi) IL‐17 and (Ci) IL‐1Ra, at BL and 23 h n(SR) = 18 and n(DR) = 12; at 6 h n = 10 each. In the lavages for (Aii) IL‐8, (Bii) IL‐17 and (Cii) IL‐1Ra, n(SR) = 10 and n(DR) = 8. (D) Bronchial lavage neutrophils at baseline were correlated with each in the plasma for (i) IL‐8, (ii) IL‐17 and (iii) IL‐1Ra

Of the three factors (Figure 6D), IL‐1Ra plasma levels at baseline (but not bronchial lavage levels—see Figure S6B) correlated with the number of bronchial neutrophils at baseline (Figure 6Diii). Suggesting that DR has more bronchial neutrophils due to the lower antimigration signalling from IL‐1Ra rather than an increased attraction signalling from IL‐8 or IL‐17.

3.7. Dual responders release NETosis factors much earlier than single responders

Since DR had more neutrophils, in particular the more active 2‐ and 3‐lobed neutrophils in the airways, we assessed whether DR had more tissue‐damaging NETosis activity in the plasma and the lavage fluid. Following the allergen challenge, DR had increasing levels of histones in plasma already during the early phase (BL to 1h p = .015; Figure 7Ai). In contrast, SR had decreasing levels of histones during the early phase followed by a significant increase during the late phase (BL to 6h p = .0045). Although DR tended to have more dsDNA in the plasma throughout all time points, there was no significant difference between the two groups (Figure 7Bi). However, the NE levels followed a trend similar to the histone levels, with DR releasing more NE both during the early phase (BL to 0h p = .037, BL to 1h p = .045, BL to 2h p = .0073) and the late phase (BL to 8h p = .0044) compared with baseline, while NE in SR remained at baseline levels, only increased during the late phase (BL to 8h p = .0009; Figure 7Ci).

FIGURE 7.

Levels of NETosis factors in plasma and lung lavage. Data compare single responders (●SR) and dual responders (□DR) at baseline (BL) up to 23 h post‐challenge. (Ai) ELISA of histones relative to healthy levels, (Bi) Qubit assay of dsDNA and (Ci) ELISA of NE in plasma, n(SR) = 18, n(DR) = 12. Lavage levels of (Aii) histones, (Bii) dsDNA and (Cii) NE. The number of 3‐lobed bronchial neutrophils correlated with (D) bronchial histone levels at baseline and (E) plasma NE levels at 23/24 h post‐challenge

In the bronchial lavage (where more of the neutrophils resided) the histone levels tended to be higher 24 h post‐challenge in DR (median: 2.3‐fold) than in SR (0.5‐fold, Figure 7Aii). The NETosis levels of dsDNA and NE tended to be higher in the DR (median: 12.4 ng/ml and 3.14 ng/ml, respectively) at baseline than in SR (median: 4.4 and 1.85 ng/ml, respectively; Figure 7Bii, Cii). NETosis factors in the alveolar lavage remained relatively the same.

Correlation analysis showed that higher numbers of 3‐lobed active neutrophils in the bronchial lavage also had high levels of histone at baseline (Figure 7D). In addition, higher baseline numbers of 1‐ to 4‐lobed active neutrophils, in contrast with hypersegmented neutrophils, correlated well with the level of histone in the lavage at 24 h post‐challenge (see the matrix of correlations in Table S8A), whereby only 1‐ to 2‐lobed neutrophils at 24 h maintained this correlation to histone levels at 23 h in the bronchial lavage (see the matrix of correlations in Table S8B). Of the more active 1‐ to 4‐lobed neutrophils, the bronchial NE levels were correlated with unsegmented neutrophils numbers at baseline (Table S9A) and plasma NE levels were correlated with the bronchial 3‐lobed (Figure 7E) and 4‐lobed neutrophils 23/24 h post‐challenge (see the matrix of correlations in Table S9B).

4. DISCUSSION

Our data show that DR present with higher numbers of neutrophils within the airways at baseline compared with SR. These neutrophils are likely active neutrophils as they present with mostly 2‐ or 3‐lobed segmented nuclei, that correlated with higher NETosis levels. The combination of less suppressive neutrophils and less IL‐1Ra with no difference in chemoattractant levels suggests that DR may have had less antitransmigration signalling rather than more attractant signalling of neutrophils to the airways. After the allergen challenge, an early increase in NETosis factors and a later decrease of sCD62L indicate that DR had a differently timed neutrophil response to the allergen challenge than SR. These findings show, in agreeance with our hypothesis, that neutrophils and their response take part in differentiating DR from SR responses in the airway environment and the overall response to an allergen challenge.

The higher number of airway neutrophils at baseline in DR was specially found in the bronchial lavage but not in the alveolar lavage. This location‐specific increase is similar to the previously reported general increase proportion of neutrophils in the bronchial compared with the alveolar compartment in both asthmatics and healthy individuals. ²⁸ Furthermore, the strong correlation of central biopsy neutrophils to the DR defining physiology during the late phase, provided a strong case for the more elevated airway neutrophils to become a distinguishing feature of DR asthma specifically before an allergen challenge. As allergen type was evenly distributed between SR and DR, the levels of natural allergen‐associated LPS ³⁵ are not likely to play a differentiating role in the levels of baseline neutrophils. ³⁶

Morphological characterisation showed that the more abundant neutrophils in DR had lower nuclei segmentation. In general, 2‐ to 4‐lobed neutrophils have previously been suspected to be mature enough to be active. ²² Here, the 3‐lobed neutrophils were likely the best representation of active neutrophils, correlating with both NETosis factors and atopic sensitizations. NETosis factors, such as NE and histones, are commonly known to be released by active neutrophils and are often increased in severe asthma, ³⁷ and more specifically neutrophilic asthma, ³⁸ but are also increased during exacerbation. ¹⁹

Interestingly, higher numbers of airway neutrophils and more sensitizations are key traits that distinguish more severe/uncontrolled asthma from milder asthma. ⁷ , ¹¹ , ³⁹ The LAR response is representative of more severe disease, ² suggesting that DR (mild/moderate) asthmatics are biologically closer to more severe asthmatics than their SR (mild/moderate) counterparts. This theory was further supported by the strong correlation, at both baseline and during the late phase, to clinical markers of peripheral obstruction (decreased X5 and increased AX) that has been reported in uncontrolled asthma with airway neutrophilia. ³⁰ Our previous report on this same cohort also shows a high involvement of the peripheral airways in the DR. ²⁶ Based on these similarities, one would think that the more promising treatments for severe or uncontrolled asthma based on targeting non‐type‐2 mechanisms may also be of benefit to DR. ⁴⁰

In general, differences in neutrophil population numbers in the tissue can come from either increased migration into the tissue, decreased migration out of the tissue (e.g. less reverse transmigration ¹⁷ ) or decreased cell death and clearance. ⁴¹ Based on the low CD62L levels in the fluid and lavage cells in DR, the increased number of neutrophils seem to be newly migrated into the bronchial airways. Future flow cytometry of lavage and/or biopsy cells should be completed to confirm the migration status of the neutrophils. Nonetheless, increased migration could be due to increased chemoattraction either from specific signals or general tissue damage (such as the aforementioned obstruction), decreased antimigration signalling or increased permeabilization through the tissue. ⁴¹

If the increased migration was due to increased chemoattraction, we would expect an increase in IL‐8 and IL‐17, especially at the tissue site to attract the neutrophils; however, this is not the case here. Instead, we find DR plasma has less IL‐1Ra than SR similar to neutrophilic asthma. ⁴² This compliments our findings as the IL‐1R (IL‐1Ra target—not assessed here) has previously been reported to be a positive biomarker of airway neutrophilia and airway obstruction. ⁴³ Glucocorticoids have been shown to decrease the amount of IL‐1Ra released by neutrophils, which could be a contributing factor since more DR used ICS than SR in this study. ⁴⁴ Based on these results, DR may benefit more from the nonchemoattractant targets tested as non‐type‐2 biotherapeutics than chemoattractant targets. ⁴⁰

DR had less hypersegmented suppressive neutrophils to regulate immunoactivity—similar to how asthmatics have less T regulatory cells compared with healthy ⁴⁵ —adding to the theory that the immune response in DR is less regulated. To support our suppressive neutrophil findings, since a number of occurrences were low, we found a recent study that shows higher CC16 levels correlates with less suppressive neutrophils in a mouse model. ⁴⁶ This coincides with our previously reported findings that these DR subjects had significantly higher CC16 levels than SR post‐challenge. ²⁷ With the current assessments, our study suggests that DR may lack sufficient cell types and signalling that could block neutrophil transmigration at baseline that could lead to a type of neutrophilic asthma.

After the allergen challenge, DR had less sCD62L in plasma than in SR—a state known to relate to lung injury inflammation. ⁴⁷ Less sCD62L has been reported to occur when there is less leukocyte migration ¹⁸ into the lungs ⁴¹ while an acute rise (like in SR) can self‐limit the neutrophil transmigration. ⁴⁸ Neutrophils are also often required for monocyte recruitment, ⁴⁹ a cell in excess found to be associated with fatal asthma airway tissue ⁵⁰ and were also increased here in DR airways, so they should be assessed further alongside neutrophils.

Using correlation analysis at baseline and 23 h post‐challenge, we reason that sCD62L levels in the plasma might have given a good indication for neutrophil migration in these subjects. The pattern of sCD62L levels relative to baseline indicated that DR neutrophils may not have transmigrated during the early reaction and even less during the late phase, meaning that DR neutrophils stayed in the blood long after the allergen challenge while SR neutrophils may have started migrating out before the late phase. This coincides with the early‐phase release of NETosis factors in DR plasma but not SR. Suggesting that SR may have a more ‘healthy’ neutrophil response to the allergen challenge that may prevent the LAR from occurring or that the more active neutrophil response to the allergen challenge may give rise to the LAR.

5. CONCLUSION

In conclusion, DR had more neutrophils within the airways that were mostly 2‐ or 3‐lobed active phenotype at baseline than at SR. Correlation analysis indicates that an increase in neutrophils in DR airways comes from either more peripheral tissue obstruction and/or less antimigration signalling. Following an allergen challenge, DR released more NETosis factors during the early phase and shed less transmigrating sCD62L during the late phase than SR. Overall, these findings indicate that the neutrophil response is differently regulated in DR than in SR, and this may be related to the LAR occurring in DR but not in SR. In the future, clinically, it is important to also target the excessive population of active neutrophils together with the type‐2 inflammation in more difficult types of asthma such as the DR.

CONFLICT OF INTEREST

All authors declare no conflict of interest.

AUTHOR CONTRIBUTION

NV, HS, ZD, LB and ET involved in conceptualization; NV, SE, DB, LB and ET involved in methodology; NV and SE involved in formal analysis; NV, HS, SE and DB involved in investigation; HS, JA, L‐OC, LB, JE and ET involved in resources; NV involved in data curation and visualisation; NV and ET involved in writing‐drafts and project administration; NV, HS, SE, ZD, SKG, L‐OC, LB, JE and ET involved in writing‐revisions; ZD, JA, SKG, L‐OC, LB, JE and ET involved in supervision; NV, L‐OC, LB, JE and ET involved in funding acquisition.

Supporting information

Fig S1

Fig S2

Fig S3

Fig S4

Fig S5

Fig S6

Supplementary Material

Table S1‐S9

van der Burg N, Stenberg H, Ekstedt S, et al. Neutrophil phenotypes in bronchial airways differentiate single from dual responding allergic asthmatics. Clin Exp Allergy. 2023;53:65–77. doi: 10.1111/cea.14149