Chronic bacterial pulmonary infections in advanced cystic fibrosis differently affect the level of sputum neutrophil elastase, IL‐8 and IL‐6

Abstract

Advanced cystic fibrosis (CF) lung disease is commonly characterized by a chronic Pseudomonas aeruginosa infection and destructive inflammation caused by neutrophils. However, the lack of convincing evidence from most informative biomarkers of severe lung dysfunction (SLD‐CF) has hampered the formulation of a conclusive, targeted diagnosis of CF. The aim of this study was to determine whether SLD‐CF is related to the high concentration of sputum inflammatory mediators and the presence of biofilm‐forming bacterial strains. Forty‐one patients with advanced CF lung disease were studied. The severity of pulmonary dysfunction was defined by forced expiratory volume in 1 second (FEV1) < 40%. C‐reactive protein (CRP) and NLR (neutrophil–lymphocyte ratio) were examined as representative blood‐based markers of inflammation. Expectorated sputum was collected and analysed for cytokines and neutrophil‐derived defence proteins. Isolated sputum bacteria were identified and their biofilm‐forming capacity was determined. There was no association between FEV1% and total number of sputum bacteria. However, in the high biofilm‐forming group the median FEV1 was < 40%. Importantly, high density of sputum bacteria was associated with increased concentrations of neutrophil elastase and interleukin (IL)‐8 and low concentrations of IL‐6 and IL‐10. The low concentration of sputum IL‐6 is unique for CF and distinct from that observed in other chronic pulmonary inflammatory diseases. These findings strongly suggest that expectorated sputum is an informative source of pulmonary biomarkers representative for advanced CF and may replace more invasive bronchoalveolar lavage analysis to monitor the disease. We recommend to use of the following inflammatory biomarkers: blood CRP, NLR and sputum elastase, IL‐6, IL‐8 and IL‐10.

Advanced lung cystic fibrosis is associated with P. aeruginosa biofilm infections. Sputum biomarkers are informative to monitor severity of lung inflammation and analysis of expectorated sputum analysis replace usage of more invasive bronchoscopy lavage (BAL). Neutrophil elastase and IL‐8 are key sputum inflammatory markers of lung injury, while low sputum IL‐6 level is a hallmark of lung cystic fibrosis chronic inflammation.

INTRODUCTION

Cystic fibrosis (CF) is an autosomal recessive genetic disorder caused by a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The most common mutation, delta F508, accounts for approximately 70% all mutations. The defect in the CFTR protein results in altered chloride transport, leading to impaired mucociliary clearance in the airways and to abnormalities in cells of innate immunity [1]. Mucus accumulation leads to uncontrolled bacterial adherence, chronic inflammation and respiratory failure. Patients with CF lung disease develop chronic pulmonary biofilm‐associated infections accompanied by massive infiltration of neutrophils, the major cells of innate immunity. Both components of the inflammatory response, pathogenic bacteria and neutrophils, contribute to the pathogenesis of the disease. Importantly, during the course of CF, the detrimental activities of the neutrophils usually outweigh their beneficial defensive properties [2].

Stapylococcus aureus and Pseudomonas aeruginosa species are major pathogens of CF lung disease. Initially the main colonizing bacterium detected in children with CF is methicillin‐susceptible S. aureus (MSSA) [3]. However, the incidence of MSSA declines with age and polymicrobial infections of methicillin‐resistant S. aureus (MRSA) and P. aeruginosa (PAR), as well as Burkholderia cepacia and Candida species, become more prominent [4]. The switch from Gram (+) to Gram (−) pathogens, especially to P. aeruginosa mucoid strains, within the lungs of CF patients correlates with a worsening prognosis, resulting in chronic inflammation and tissue destruction [5]. Recently, several papers have reported a supportive interaction between S. aureus (MRSA) and P. aeruginosa in the colonization and formation of a mixed biofilm in many chronic inflammatory pathologies, including CF [6, 7, 8]. Massive neutrophilic inflammation is another hallmark of CF lung disease [9]. In response to bacterial infections in lungs of healthy people neutrophils migrate to the airways, where they kill planktonic bacteria before undergoing apoptosis and removal by macrophages from the site of infection. This co‐ordinated process of innate immunity is altered in CF lungs [10]. Activated neutrophils in CF commit neutrophil extracellular traps (NET)osis (not apoptosis). Moreover, they fail to effectively clear the pathogens hidden in a biofilm. Instead, their activation leads to the accumulation of neutrophil proteases (e.g. elastase that degrades airway matrix structural proteins) and induces oxidative stress [11]. Therefore, all these changes exacerbate and maintain chronic inflammation, interfere with efficient bacterial clearance and compromise lung tissue integrity [12]. Recently, we have shown that biofilm‐associated neutrophils (BANs), exposed in vitro to components of P. aeruginosa biofilm, produce a hyperinflammatory set of neutrophil‐derived mediators [13]. Extending these in‐vitro findings, we hypothesize that the airway microenvironment of advanced CF favours biofilm‐linked infections. Consequently, it may be responsible for the BAN‐specific profile of both the CF‐related inflammatory biomarkers and severe lung dysfunction [(SLD) = forced expiratory volume in 1 second (FEV1) < 40%].

Hence, it is commonly accepted that both pathogens and neutrophils are responsible for chronic inflammation in CF lung disease [14]. However, the final effect of multi‐species biofilm interaction with infiltrating neutrophils has not been adequately explained. It is also not clear whether severe CFTR gene mutation significantly contributes to the production of neutrophil chemokine interleukin (IL)‐8 at the advanced course of CF lung diseases [15]. Furthermore, it has not been determined which sputum inflammatory biomarkers are most informative and specific for advanced CF lung disease and SLD‐CF.

To clarify this problem we analysed lower airway secretions obtained from patients with advanced CF lung disease. We examined expectorated sputum samples to identify lung pathogens and to estimate the concentration of inflammatory biomarkers associated with chronic endobronchial infection in CF. Moreover, we compared the concentrations of inflammatory markers in sputum with those of blood.

MATERIALS AND METHODS

Patient selection criteria

For this study, 41 patients with advanced CF, previously P. aeruginosa‐infected, were included. The study was approved by the local bioethics committee (decision number 1072.6120.252.2018). Inclusion criteria were confirmed diagnosis of CF lung disease [i.e. respiratory clinical symptoms and positive sweat test (> 60 mmol/l) and/or genotypic confirmation of two CFTR disease‐inducing mutations] and history of chronic or periodical infection with P. aeruginosa (defined as at least one documented positive culture during the 3 years prior to study) or reasonable suspicion of presence of this pathogen. Subjects who met any of the following exclusion criteria were not enrolled into this study: haemoptysis within 4 weeks prior to study, inability to expectorate sputum (or to undergo sputum induction) or to perform reproducible spirometry infection with Burkholderia cepacia complex. Blood and sputum samples were collected from the included patients on the same day that their spirometry was performed.

Determination of CFTR mutations

Patients with a confirmed diagnosis of CF (positive sweat chloride value by quantitative pilocarpine iontophoresis test or two well‐characterized, CF‐causing gene mutations) were included into the study. Subjects were enrolled regardless of CFTR mutation. However, for later analysis, the ‘severe’ genotype was defined as homozygous for classes I–III mutations and ‘mild’ was defined as homozygous or heterozygous for classes IV–V mutations.

In the scientific literature CFTR mutations are grouped into five or six classes, based on the way the mutations affect the CFTR protein functional properties of chloride regulation in the apical membrane of epithelial cells [16]. Classes I–III mutations (also called ‘severe’ mutations including, among others: G542X, 2184insA, S466X, delF508, N1303 and G551D) are associated with pancreatic insufficiency, an earlier and steeper rate of decline in FEV1 and shorter survival time than patients who carry at least one of classes IV–VI mutations (also called ‘mild’ mutations including, for example, c.1585‐1G>A, 3849+10kbC>T). It is explained by classes IV–VI (mild allele) mutations’ phenotypical domination when occurring in combination with the severe allele (classes I–III mutations) [17].

Spirometry

Spirometry was performed according to published Polish and international guidelines [18, 19]. Participants performed at least three spirometry manoeuvres that achieve American Thoracic Society (ATS) standards. The best FEV1 was recorded and expressed as age, height and gender adjusted percentage of predicted, based on the Global Lung Function Initiative (GLI) reference values [20]. One of the criteria for severity assessment of bronchopulmonary disease was FEV1% predicted as normal (≥90% predicted); mild (70–89% predicted); moderate (40–69% predicted) or severe (< 40% predicted) [21]. Herein, FEV1 < 40% is the marker of severe lung dysfunction (SLD‐CF). If sputum was collected at the beginning of an exacerbation, FEV1 values measured at the end of the exacerbation period (after hospitalization and antibiotic therapy [22]) were used to classify the severity of the disease.

Blood‐based biomarker analysis

The following blood biomarkers were obtained from routine clinical assessment: white blood cell count [neutrophil–lymphocyte ratio (NLR)], C‐reactive protein (CRP), iron and transferrin concentration.

Sputum collection and processing

Spontaneously expectorated sputum was collected in a sterile cup and processed immediately. The sputum samples were liquefied with an equal volume of Pulmozyme (Dornase Alpha) (Roche, Indianapolis, Indiana, USA), then divided for further analysis. One part was designated for scanning electron microscopy (SEM) analysis and was not frozen, one part was transferred to a pre‐weighed tube for microbiological analysis and the last part was used for cytokine measurements (the latter two were frozen).

Bacteria identification and characterization

Samples were plated on the following media: McConkey agar (Oxoid/ThermoFisher Scientific, Fremont, California, USA) for Enterobacteriaceae, Columbia blood agar (Oxoid) with 5% sheep blood for streptococci, BBL enterococcosel agar (BD Biosciences, San Jose, California, USA) for enterococci and Pseudomonas agar (Oxoid) for Pseudomonas. Phenotypical identification of P. aeruginosa isolates from McConkey agar was conducted with a commercial identification system API20NE (BioMerieux, Marcy‐l’Étoile, France). In order to compare the samples, the colony counts were adjusted to the weight of the sample and results were shown as colony‐forming units (CFU)/g. All isolated bacterial strains were stored for further investigation in the next stage of our project.

Measurement of biofilm formation

Biofilm formation capacity (total mass of bacterial polysaccharides) was measured in sterile plastic 96‐well plates with adherent surface (Greiner Bio‐One, Kremsmünster, Austria) using Congo red dye, according to a modified procedure described by Reuter et al. [23]. Briefly, 20 µl of fresh culture of the bacterial suspensions, prepared as described above, was added to each well followed with 180 µl of sterile TSB. The final concentration of the bacteria was 1/10⁷ CFU/ml. The plates were centrifuged for 10 min at 500 g to sediment bacteria at the bottom of each well. Bacteria were then incubated for 72 h (37°C, aerobic conditions). At different time‐points of the culture (0, 8, 16, 24, 48 and 72 h), the plates were centrifuged, the culture medium was gently removed from wells and 200 µl of 0.1% Congo red solution was immediately added. The plates were left for 30 min at room temperature (RT) and washed twice with buffered saline to remove unbound dye. Absorbance was measured at λ = 492 nm wavelength using a spectrophotometer (Awareness Technology, Inc., Palm City, Florida, USA)

Cytology

The cell suspension was centrifuged to prepare cytospin slides. The slides were air‐dried and stained using May–Grunwald–Giemsa.

Protein concentration measurement

The sputum samples were centrifuged at ×400 g for 10 min and supernatants were collected for analysis of protein content and determination of inflammatory biomarkers. Total protein concentrations in sputum samples were determined using a bicinchoninic acid protein assay kit (Sigma‐Aldrich, St Louis, Missouri, USA), according to the manufacturer’s protocol.

Inflammation markers determination

Concentrations of several inflammation markers in sputum samples were determined using enzyme‐linked immunosorbent assay (ELISA), according to the manufacturer’s protocol, using appropriate dilutions when necessary (initial material was already mixed with equal volume of Pulmozyne). Alpha‐1 anti‐trypsin (A1AT) ELISA kit (Abcam, Cambridge, Massachusetts, USA), human polymorphonuclear neutrophil‐elastase ELISA (Invitrogen, Carlsbad, California, USA) and prostaglandin E₂ (PGE₂) high‐sensitivity ELISA (Enzo Life Sciences, Farmingdale, New York, USA) were used. Concentrations of complement component 5a (C5a), CRP, IL‐1β, IL‐10, IL‐6, IL‐8, lactoferrin (Lf), matrix metalloproteinase 9 (MMP‐9), myeloperoxidase (MPO) and tumour necrosis factor (TNF)‐α in sputum samples were measured using the human magnetic Luminex assay (R&D Systems, Minneapolis, Minnesota, USA), according to the manufacturer’s protocol.

Scanning electron microscopy

Sputum samples were fixed using a formaldehyde 4% buffered solution. Each sample was immersed in this solution for at least 24 h at 4°C. Next, the samples were dehydrated in rising concentrations of ethyl alcohol and air‐dried. The dried specimens were mounted on plates using silver glue and sputter‐coated with gold using a JFC‐1100 sputter. The samples were examined in a JSM 35‐CF scanning electron microscope at 15 kV (Jeol, Tokyo, Japan). Images captured at various magnifications were obtained for the best possible illustration of the biofilm in sputum samples. To consolidate the results, SEM images obtained at magnifications of ×20, ×72, ×200 and ×2000 were analyzed.

Statistical analysis

Data are presented as frequency and percentage for qualitative data and as median and interquartile range (IQR) for quantitative data. Quantitative data were compared with Mann–Whitney test. Spearman’s rank correlation was used in the analysis of association between blood and sputum markers. The statistical analysis was performed in IBM SPSS Statistics version 25. The significance level was set at P < 0.05.

RESULTS

Patient demographics

Among 41 CF patients initially included, one was excluded due to infection with B. cepacia complex and another was excluded as no P. aeruginosa strains were isolated from his/her sputum. From the remaining 39 patients, 15 were classified into the SLD‐CF group; the remaining patients were classified into the non‐severe group. The SLD‐CF group was assigned to patients who had an FEV1 parameter below 40% (measured in the stable period outside CF exacerbation) or had undergone lung transplant within a year after sample collection (two patients). General characteristics of the studied group are summarized in Table 1, whereas individual parameters are listed in Table 2.

TABLE 1.

General characteristics of the cystic fibrosis patients. Data are presented as mean (range)

	CF patients (n = 39)
Male:female	21:18
Age (years)	24.6 (9.3–42.3)
BMI (kg/m2)	19.8 (12.4–28.8)
FEV1 (% predicted)	57 (13–95)
Exacerbation at the time of study (yes:no)	21:18
Cells in sputum
Neutrophils (%)	95
Macrophages (%)	1–2

TABLE 2.

Individual characteristics of the cystic fibrosis patients (n = 39). Patients 2 and 26 did not fulfil inclusion criteria and were excluded from the study

Patient no.	Severe CFTR mutation	Exacerbation	FEV1 (%)	NLR	CRP (mg/l)	Total bacteria count (CFU/g)	P. aeruginosa strains	High biofilm‐forming P. aeruginosa strains	Other strains
01	+	+	21	5.2	100.7	4.1/108	3	+	C. albicans
03	−	−	34	2.9	4.1	2.9/105	1	+	C. lusitaniae
04	+	+	65	2.4	3.2	3.7/104	1	−	MSSA, C. albicans
05	−	−	50	5.4	3.0	7.9/105	3	+	C. lusitaniae
06	−	−	88	3.0	2.0	5.2/106	3	+	MSSA, C. albicans
07	+	+	72	5.1	0.4	7.3/105	2	−	MSSA
08	+	−	41	5.8	51.3	2.1/109	3	−	MSSA, C. albicans
09	+	+	60	1.3	0.9	9.5/107	2	+	MSSA
10	+	+	74	4.2	6.4	8.8/105	1	−	MSSA, C. albicans
11	+	−	29	2.7	3.1	1.2/108	2	+	MSSA, S. pneumoniae, C. tropicalis
12	−	+	66	2.6	8.0	2.0/107	3	+	C. albicans
13	−	+	77	2.4	5.7	7.6/109	1	−	MSSA, C. glabrata
14	+	−	64	0.6	10.7	1.4/109	2	+	MSSA, C. albicans
15	+	+	24	7.3	133.2	8.6/105	2	+	C. albicans
16	−	+	36	11.4	58.2	1.8/107	1	+	MSSA
17	+	+	13	5.3	52.3	9.6/109	3	+	MSSA
18	+	+	30	5.0	36.5	7.1/106	2	+	‐
19	+	+	63	2.1	1.0	3.6/108	2	−	MSSA
20	+	−	57	2.7	0.3	5.0/107	2	+	MSSA
21	+	−	83	1.3	1.7	1.9/109	1	−	MSSA, C. glabrata
22	+	−	67	4.0	0.4	2.5/107	1	−	MSSA
23	+	−	54	1.8	3.9	4.2/105	1	−	MRSA, C. glabrata
24	+	+	79	5.0	0.4	4.0/107	2	+	MRSA, C. albicans
25	+	+	21	1.3	8.4	4.6/107	2	−	MSSA
27	+	−	26	2.0	1.8	3.3/108	3	+	MSSA
28	+	+	83	0.9	0.2	2.0/102	1	−	‐
29	−	−	36	5.1	32.3	5.3/105	2	+	MSSA
30	+	−	14	19.1	30.1	2.7/105	2	−	C. albicans
31	+	+	74	1.3	16.3	2.0/102	1	−	C. albicans
32	+	−	43	3.2	11.2	2.1/108	2	−	MSSA
33	−	+	26	1.3	4.0	1.5/108	1	+	MRSA
34	+	−	66	2.0	1.7	2.5/107	1	+	MRSA, MRSE
35	+	+	56	3.0	1.9	1.9/107	1	−	MRSA
36	−	−	71	1.2	0.4	2.0/102	1	−	C. albicans
37	−	−	95	1.5	1.1	1.5/108	2	+	MRSA
38	+	+	83	1.6	0.1	1.4/109	1	−	MSSA
39	+	−	59	3.4	9.9	1.8/109	2	−	MRSA. C. albicans
40	−	+	51	3.8	5.3	1.9/109	2	−	MSSA
41	+	+	39	9.2	5.7	2.7/1010	2	+	MSSA, MRSE, C. albicans, C. glabrata

CFU = colony‐forming units; CFTR = cystic fibrosis transmembrane conductance regulator; FEV1 = forced expiratory volume in 1 second; NLR = neutrophil–lymphocyte ratio; CRP = C‐reactive protein; MSSA = methicillin‐susceptible Staphylococcus aureus; MRSA = methicillin‐resistant S. aureus; MRSE = methicillin‐resistant S. epidermidis. FEV1 values for patients in SLD‐CF group were shown in bold.

Association between blood inflammatory biomarkers and lung function

Two common blood inflammatory biomarkers, CRP concentration and NLR values, were compared between the cohorts (FEV1 = < 40% versus FEV1 = > 40%). Both biomarkers differed significantly between patients with non‐severe and SLD‐CF. Importantly, the median concentration of CRP in the severe CF (30.0 mg/l) was markedly elevated above physiological levels (Figure 1A). NLR in both non‐severe and severe CF groups (median = 2.6 and 5.1, respectively) (Figure 1B) was markedly above the standard value (1.6 for healthy people) [24, 25].

FIGURE 1.

Blood C‐reactive protein (CRP) concentrations and neutrophil–lymphocyte ratio (NLR) values in the non‐severe (n = 24) and severe lung dysfunction‐cystic fibrosis (SLD‐CF) (n = 15) cohort. Median with interquartile range as well as minimum and maximum values are shown. **P < 0.01, ***P < 0.001 (Mann–Whitney test)

Association between the genotype, CF exacerbation, prevalence of bacteria, biofilm formation and lung function (FEV1%)

We did not find any statistically significant relationship between FEV1% and severe CFTR mutation, disease exacerbation and total number of sputum bacteria (data not shown). There was no difference in the density of sputum bacteria between patients with clinically defined CF exacerbation and those with stable CF (median = 3.7/10⁷ in stable CF versus 3.9/10⁷ CFU/g in exacerbation). Furthermore, there was no difference in the total number of bacteria in sputum (median = 1.8/10⁷ for non‐severe mutation cohort versus 4.8/10⁷ CFU/g for severe mutation cohort), lung function (defined by FEV1%, median = 50 versus 58% for non‐severe versus severe mutations) as well as inflammatory biomarkers (blood CRP: medians = 4.1 versus 3.6 mg/l and NLR = 3.0 versus 2.9 for the non‐severe versus severe mutation cohorts, respectively), between cohorts with severe and non‐severe CFTR mutations.

Association between sputum microbiota biofilm formation capacity, lung function (FEV1%) and inflammatory biomarkers

We have isolated and identified 70 various P. aeruginosa strains, all of which were able to produce biofilm in vitro. Among these, 33 were classified as high biofilm‐forming strains in comparison to reference strains of that species. In Figure 2 we show the representative SEM image of sputum bacterial biofilm collected from patient 17 who presented extremely severe symptoms of CF (severe mutation in the CFTR; FEV1 = 13%; NLR = 5.4; blood CRP = mg/l; also deceased 6 months after the sample collection).

FIGURE 2.

The scanning electron microscopy (SEM) image of the sputum sample (sputum isolated from patient with severe symptoms of cystic fibrosis (CF) infected with high‐biofilm forming P. aeruginosa strain) shows an infiltrating mass of neutrophils, bacterial cells hidden in a matrix of biofilm and dehydrated sticky CF mucus. Magnification ×2000

The presence of sputum high biofilm‐forming P. aeruginosa strains was not associated with SLD‐CF (Table 3). However, in the high biofilm‐forming group the mean FEV1 was markedly lower (37%) than that of the low biofilm‐forming group (60%). Moreover, the median of sputum concentration of many inflammatory biomarkers was not significantly but markedly higher in the high biofilm‐forming group. Namely, CRP (82.9 versus 3.6 mg/l); IL‐1 (2012.4 versus 1117.1 ng/g of protein); IL‐8 (455.9 versus 293.1 ng/g of protein); elastase (70.8 versus 38.6 µg/g of protein); MMP‐9 (1399.4 versus 355.8 ng/g of protein); MPO (18.4 versus 10.4 µg/g of protein).

TABLE 3.

The impact of the high biofilm‐forming airway pathogens on FEV1 and the level of inflammatory biomarkers. FEV1 as well as blood (CRP, NLR) and sputum biomarkers’ concentrations (CRP, elastase, A1AT, PGE₂, C5a, IL‐1, IL‐6, IL‐8, IL‐10, MPO, MMP‐9) were compared between cohorts classified into high and low biofilm formation. Patients with at least one bacterial strain capable of producing high amounts of biofilm (corresponding to absorbance > 0.8 in biofilm formation assay) were classified into a high biofilm‐forming group

	Biofilm formation		P
	Low n = 25	High n = 14
FEV1 (% predicted)	60 (43–72)	37.5 (26–69.3)	0.185
CRP blood (mg/l)	5.3 (0.9–11.2)	4.0 (1.8–53.0)	0.333
NLR	2.6 (1.3–4.0)	5 (2–6.2)	0.067
Sputum agents:	n = 16	n = 8
CRP (µg/g of protein)	3.6 (3.2–5.0)	82.9 (3.7–186.9)	0.095
Elastase (µg/g of protein)	38.6 (19.9–103.6)	70.8 (25.1–143)	0.388
A1AT (µg/g of protein)	1406.6 (766.5–2056.1)	1832.4 (1192.2–2627.2)	0.487
PGE2 (ng/g of protein)	774.2 (579.7–1659.2)	1035 (609.1–1561.2)	0.474
C5a (µg/g of protein)	38.5 (28.8–59)	42 (33.5–52)	0.846
IL‐1 (ng/g of protein)	1117.1 (670.5–2181.7)	2012.4 (1569.4–2049)	0.095
IL‐6 (ng/g of protein)	132.6 (119.9–163.5)	137.7 (119.8–179.2)	0.803
IL‐8 (ng/g of protein)	293.1 (222–454.3)	455.9 (405.8–827.3)	0.057
IL‐10 (ng/g of protein)	122.1 (87.6–135)	99.8 (83.8–130.9)	0.357
MPO (µg/g of protein)	10.4 (4.3–16.4)	18.4 (15.1–27.4)	0.014
MMP‐9 (ng/g of protein)	355.8 (242.6–1319.5)	1399.4 (431.5–2042.2)	0.136

FEV1 = forced expiratory volume in 1 second; NLR = neutrophil–lymphocyte ratio; CRP = C‐reactive protein; A1AT = alpha‐1‐anti‐trypsin; PGE₂ = prostaglandin E₂; C5a = complement component 5a; IL = interleukin; MPO = myeloperoxidase; MMP‐9 = matrix metalloproteinase‐9.

Association between sputum inflammatory biomarkers and lung function

Concentrations of inflammatory biomarkers were assayed in the sputum samples (normalized to the total protein concentration) and compared between the non‐severe and SLD‐CF cohorts. Analysis revealed that concentrations of CRP, NE and MPO in sputum were associated with the severity of lung dysfunction (FEV1%) (Figure 3). In contrast, the concentrations of IL‐1β, IL‐6 and IL‐8 did not differ statistically between the groups. However, the concentration of IL‐8 – the major neutrophil chemokine – was greater than 400 ng/g of protein in most patients with the SLD‐CF. This observation prompted further analysis of relationship between the concentration of IL‐8 and concentrations of inflammatory biomarkers. Interestingly, we noted that patients with the high IL‐8 exhibited higher concentrations of CRP (both blood and sputum) and sputum NE, IL‐1β, MPO and MMP‐9 (Figure 4).

FIGURE 3.

Relationship of neutrophil‐derived inflammatory mediators with severe cystic fibrosis (CF) lung dysfunction [forced expiratory volume in 1 second (FEV1) < 40%)]. Sputum C‐reactive protein (CRP), NE, alpha‐1‐anti‐trypsin (A1AT), interleukin (IL)‐1β, IL‐6, IL‐8, IL‐10, myeloperoxidase (MPO) and matrix metalloproteinase‐9 (MMP‐9) concentrations were compared in non‐severe CF (n = 16) versus severe lung dysfunction‐CF (SLD‐CF) (n = 8) groups. *P < 0.05 (Mann–Whitney test)

FIGURE 4.

Association between sputum interleukin (IL)‐8 and inflammatory markers. Blood parameters [C‐reactive protein (CRP) and neutrophil–lymphocyte ratio (NLR)] as well as sputum CRP, elastase, alpha‐1‐anti‐trypsin alpha‐1‐anti‐trypsin (A1AT), PGE₂ = prostaglandin E₂; C5a = complement component 5a IL‐1β, IL‐10, myeloperoxidase (MPO), matrix metalloproteinase‐9 (MMP‐9) and lactoferrin (Lf) concentrations were compared in groups of patients with high (400 ng/g of protein or more, n = 11) or low (n = 13) concentrations of IL‐8 in sputum. *P < 0.05; **P < 0.01 (Mann‐Whitney test)

Performed analysis has revealed several interesting correlations between the tested parameters (Table 4). FEV1 was found positively related to the IL‐10 concentrations (Spearman’s correlation coefficient = 0.41), while CRP correlated positively with NLR, A1AT and IL‐8 (Spearman’s correlation coefficients = 0.49, 0.56 and 0.46, respectively). At the same time, the total bacteria number was directly related to neutrophil elastase concentrations (Spearman’s correlation coefficient = 0.38). Moreover, FEV1 was shown to be inversely related to both blood markers: CRP and NLR as well as P. aeruginosa number (Spearman’s correlation coefficients = −0.62, −0.44 and −0.40, respectively). At the same time, total bacteria density correlated negatively with IL‐6 and IL‐10 (Spearman’s correlation coefficients = −0.54 and −0.47, respectively).

TABLE 4.

Analysis of the correlation between FEV1, blood CRP, NLR, sputum bacteria number and sputum neutrophil‐derived inflammatory agents. *Bold type indicates significant correlation (P < 0.05)

	P. aeruginosa number (CFU/g)	Total bacteria number (CFU/g)	FEV1 (%)	CRP (mg/l)	NLR	Elastase (µg/g of protein)	A1AT (µg/g of protein)	IL‐6 (ng/g of protein)	IL‐8 (ng/g of protein)	IL‐10 (ng/g of protein)
P. aeruginosa number (CFU/g)	1.00
Total bacteria number (CFU/g)	0.27	1.00
FEV1 (%)	−0.40*	−0.10	1.00
Blood CRP (mg/l)	0.23	0.10	−0.62*	1.00
NLR	0.22	0.05	−0.44*	0.49*	1.00
Elastase (µg/g of protein)	0.15	0.38*	−0.26	0.31	0.23	1.00
A1AT (µg/g of protein)	0.15	0.07	−0.19	0.56*	0.35*	0.36*	1.00
IL‐6 (ng/g of protein)	−0.16	−0.54*	0.18	0.06	−0.03	−0.16	0.09	1.00
IL‐8 (ng/g of protein)	0.16	0.13	−0.22	0.46*	0.38	0.38	0.49*	−0.03	1.00
IL‐10 (ng/g of protein)	−0.28	−0.47*	0.41*	−0.34	−0.16	−0.39	−0.17	0.72*	−0.35	1.00

FEV1 = forced expiratory volume in 1 second; NLR = neutrophil–lymphocyte ratio; CRP = C‐reactive protein; CFU = colony‐forming units.

Profiles of CF sputum neutrophil‐derived cytokines and defense proteins

A number of previous reports demonstrated distinct sputum cytokine profiles in CF and other chronic inflammatory lung diseases. Unique for CF was an extremely low concentration of IL‐6 associated with a massive secretion of other cytokines [26]. Analysis of sputum samples in the studied cohort (24 CF patients) confirmed this observation (Figure 5). Sputum samples from the CF patients exhibited high concentrations of proinflammatory IL‐1β and TNF‐α as well as chemotactic IL‐8 (medians = 2868, 1493 and 744 pg/ml, respectively). In contrast, very low concentrations of IL‐6 and IL‐10 (medians = 290 and 208 pg/m, respectively) were observed in the tested samples. Importantly, it resulted in the high ratio of IL‐1 concentration to both cytokines. IL‐1/IL‐6 (median = 11.9) and IL‐1/IL‐10 (median = 14.2).

FIGURE 5.

Concentrations of cytokines in sputum samples taken from cystic fibrosis (CF) patients (n = 24). Median with interquartile range as well as minimum and maximum values are shown

DISCUSSION

Pulmonary biomarkers of infection and inflammation are commonly used to monitor the clinical status of CF lung disease. For many years a bronchoscopy with bronchoalveolar lavage (BAL) were considered to be the optimal source of pathogens, inflammatory cells and soluble mediators [27]. However, at the time of the COVID‐19 pandemic, due to the difficulties in obtaining BAL samples and the high risk of infection (to transmit the virus), less invasive methods for obtaining lower airway secretions are recommended [28]. One of these methods is the collection of expectorated sputum. It has been demonstrated that measurements of CF sputum inflammatory biomarkers are fairly reproducible, but the diversity of the concentration of sputum biomarkers among CF patients is very high [29]. Importantly, some previous studies have shown that the content of cytokines in sputum, but not serum, are consistent with the inflammatory environment in the lung, where the effect of cytokines is primarily localized [30].

This study was performed to determine if sputum inflammatory biomarkers are associated with advanced CF lung disease and relevant to the detrimental role of neutrophils in the pathogenesis of CF. We also analyzed the impact of the CFTR gene mutation and the magnitude of infections on the severity of lung dysfunction (FEV1). To address this issue, samples of expectorated sputum were collected from a cohort of selected patients, all with confirmed CF lung disease and P. aeruginosa airways infection, but with different levels of lung dysfunction. The severity of lung dysfunction was defined by FEV1% relative to normal; the spirometry criterion is commonly used in CF [21]. Based on this, all tested patients with FEV1 = < 40 % were classified as patients with a severe lung dysfunction (SLD‐CF). We did not include samples from healthy volunteers or patients with other chronic lung diseases, for which abundant historical data already exist [26, 31, 32].

Previously, it has been demonstrated that at the early stages of CF, severe mutations of the CFTR gene (delta F508 deletion) are associated with an elevated concentration of sputum IL‐8 and an excessive inflammatory response, even before bacterial challenge [33]. IL‐8 is a potent chemokine and activator of neutrophils and is considered to be an important proinflammatory cytokine in the pathogenesis of CF. During the early stages of CF, IL‐8 is produced mainly by airway epithelial cells, non‐stimulated by exogenous stimuli [34]. Our data suggest that, in advanced CF, in the presence of infectious stimuli, the severe CFTR protein mutation has no impact on lung dysfunction, the prevalence of high biofilm‐forming bacterial strains and the production of sputum inflammatory biomarkers, including IL‐8. Moreover, the high concentration of sputum IL‐8 was associated with elevated concentrations of CRP, the major systemic inflammatory biomarker and neutrophil‐derived proteins, such as elastase (NE), MPO, MMP‐9 and lactoferrin. The relationship between IL‐8 and NE confirms other observations, suggesting a self‐perpetuating inflammatory process in the CF airways where NE released by neutrophils induces epithelial cells to secrete IL‐8 which, in turn, recruits additional neutrophils to the bronchial surface [35].

Microbiological studies showed a mixed composition of sputum pathogens containing high and low biofilm‐forming strains. The increase in P. aeruginosa population numbers, a major colonizer of the respiratory tract, was not associated with an elevation in the sputum inflammatory biomarkers. Moreover, the concentration of some cytokines (IL‐1, IL‐8) was not significantly elevated, but remained markedly higher in the high biofilm‐forming group than that of the low biofilm group. Previously, it has been demonstrated that bacterial DNA, the component of the P. aeruginosa biofilm, induces human respiratory epithelial cells to secrete IL‐8 [36]. The data presented here also suggest that the presence of additional pathogens (MSSA, MRSA, C. albicans) and total bacteria number is not associated with SLD‐CF (FEV1). These findings are in an agreement with other reports, and suggest that a decline in lung function in advanced CF does not generally result from increased bacterial density within the airways. Alternative models of CF pulmonary exacerbation regarding neutrophil‐conducted chronic inflammation have been discussed [37]. In our opinion, this information suggests that the biofilm–neutrophil interactions may contribute more to CF lung injury than general bacteria numbers. This is in agreement with our hypothesis that the airway microenvironment of advanced CF favours biofilm‐linked infections. Consequently, it may be responsible for the BAN‐specific profile of CF inflammatory biomarkers. To support this hypothesis we have collected high biofilm‐forming P. aeruginosa strains from patient with SLD‐CF (patient 17: severe mutation; FEV1 = 13%; NLR = 5.4; CRP 54 = mg/l; deceased 6 months after the sample collection). In further studies we will investigate the effect of biofilm components isolated from selected P. aeruginosa strains based on neutrophil and macrophage polarization and their capacity to kill bacteria.

Herein, to confirm the usefulness of systemic biomarkers of inflammation in monitoring CF lung disease, the relationship between blood inflammatory biomarkers and clinical status of CF patients was investigated. We observed that blood concentrations of CRP and NLR values (> 3.5) were associated with the severity of lung dysfunction (FEV1 = < 40%). These results are in agreement with other reports showing that NLR negatively correlates with FEV1 of CF patients [38]. Both elevated blood neutrophil count and massive neutrophil infiltration of airways in CF suggest that neutrophils may be an important contributor to lung dysfunction, but increased CRP and NLR are widely observed in many inflammatory diseases, as well as cancer [39, 40]. We therefore looked in more detail at a broader panel of inflammatory mediators found in sputum, which may reflect more precisely the immunopathology of advanced CF.

Our data expressly show that the CRP concentration in sputum is significantly higher than that of serum in these CF patients. Sputum CRP positively correlated with FEV1, the major lung function index. This suggests that CRP may be secreted locally from the respiratory tract and is a highly informative marker of airways inflammation. In contrast, we did not detect any significant associations between the concentration of proinflammatory cytokines (IL‐1β, TNF‐α) and SLD‐CF. Moreover, our results have shown that the sputum cytokine profile in advanced CF is unique and distinct from that found in many other chronic inflammatory airway diseases [26]. Specifically, CF is characterized by high concentrations of IL‐1β, IL‐8 and TNF‐α but very low concentrations of IL‐6 and IL‐10. The low concentration of IL‐6 in sputum taken from patients with severe chronic lung inflammation is especially surprising. It raises the question of why the production/degradation of IL‐6 differs from that of the other proinflammatory cytokines. However, both beneficial and detrimental roles have been ascribed to IL‐6 in the context of CF [30, 41, 42], but the mechanism leading to low IL‐6 concentrations remains unclear.

IL‐6 is the key cytokine in pathogenesis of the cytokine storm and is the most frequently reported cytokine to be increased in the fatal outcome of COVID‐19 [43]. Mortality in these patients has been linked to uncontrolled production of proinflammatory cytokines that leads to pulmonary hyperinflammation (‘cytokine storm’) and acute respiratory failure [3]. Therefore, patients having pre‐existing lung problems, such as chronic obstructive pulmonary disease (COPD), severe asthma and CF lung disease, should be at high risk of developing severe symptoms of COVID‐19 [44]. Surprisingly, the preliminary data collected by the European Cystic Fibrosis Society on the response of CF patients to COVID‐19 suggest that the course of the disease may be milder than expected [45]. We have recently reported our hypothesis that low concentrations of IL‐6 present in the inflamed airway tract of CF patients might not only be the hallmark of CF lung inflammation, but also an indicator of the inhibition of the cytokine storm and the predictor of diseases severity in the course of COVID‐19 superinfection [46].

In addition to measuring the key cytokines which recruit and orchestrate the inflammatory response in CF we measured a number of key neutrophil proteins, including MPO, NE and MMP‐9. All tested proteins were found in higher concentrations in sputa from individuals with worse lung function (SLD‐CF) and their concentrations were associated with high concentrations of IL‐8. These neutrophil proteins are released into the lung fluid of CF patients as a result of neutrophil degranulation or death by NETosis or necrosis [47]. All these proteins have known detrimental activities. NE may attack a number of host extracellular proteins, including lung elastin and fibronectin causing tissue injury at the site of inflammation. MPO, through the generation of cytotoxic hypochlorous acid, also contributes to tissue destruction. MMP‐9 degrades the extracellular matrix [31, 48]. Moreover, MMP‐9 cleaves pulmonary collection SP‐D (surfactant protein D), and this cleavage leads to a loss of its innate immune functions [49]. All these proteins also play a key role in the innate immune system and may contribute to the biofilm destruction [50, 51]. The observation that higher concentrations of these neutrophil proteins are found to be associated with more severe CF lends support to the underlying hypothesis that lung dysfunction is associated with an exaggerated and ultimately ineffective neutrophil response.

Another interesting observation was that high concentrations of all these neutrophil‐derived mediators are associated with high concentrations of sputum IL‐8. IL‐8 is a major neutrophil chemoattractant and may therefore play a key mechanistic role in driving airway neutrophil infiltration, leading ultimately to neutrophil death and the uncontrolled release of neutrophil proteins into the airway microenvironment [33, 52].

CONCLUSIONS

In summary, our results, taken together with other previous reports, indicate that expectorated sputum is a representative and highly informative source of CF biomarkers, reflecting key molecular processes which lead to CF lung dysfunction. Hence, we recommend the use of expectorated sputum samples instead of more invasive and potentially contagious BAL samples in the diagnosis of CF. This recommendation is in accordance with the ruling sanitary regime of the COVID‐19 pandemic.

These data also suggest that a specific cytokine profile characterized by low concentrations of both IL‐6 and IL‐10 may reflect the unique pathogenesis of CF lung disease. In addition, elevated concentrations of key neutrophil effector proteins, especially NE, may be useful as markers associated with the disease severity. Thus, we suggest that NE, the well‐known destructive enzyme contributing to lung injury, is a potential therapeutic target in CF. Therefore, from a wide spectrum of commonly tested sputum inflammatory biomarkers we recommend measurement of NE, IL‐8, IL‐6 and IL‐10 concentrations. Moreover, in contrast to early CF, in the advanced CF and chronic P. aeruginosa infections, the genotype of the CFTR mutation has no impact on pulmonary exacerbation (FEV1) and cannot be used as a predictor of the disease progression. Further studies in larger cohorts of patients are necessary to support our observations.

CONFLICT OF INTERESTS

The authors declare that there are no conflicts of interest regarding the publication of this paper.

AUTHOR CONTRIBUTIONS

Conceptualization and funding acquisition: J.M.; methodology: H.M., M.S., J.M.; resources: H.M., A.P., M.S., J.M.; supervision: J.M.; project administration: M.C.L., J.M.; investigation: G.M., M.C.L., H.M., R.S., E.G.; data curation: G.M., M.C.L., E.G.; formal analysis: G.M., M.C.L., H.M., A.P.; visualization: G.M., A.P.; writing – original draft: G.M., M.C.L., J.M.; writing – review and editing: H.M., A.P., M.S.

ETHICS APPROVAL AND PATIENT CONSENT STATEMENT

Ethical committee approval was received for the studies and the informed consent of all participating subjects was obtained.

Majka G, Mazurek H, Strus M, Ciszek‐Lenda M, Szatanek R, Pac A, et al. Chronic bacterial pulmonary infections in advanced cystic fibrosis differently affect the level of sputum neutrophil elastase, IL‐8 and IL‐6. Clin Exp Immunol. 2021;205:391–405. 10.1111/cei.13624

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.