Determinants of induced sputum neutrophil differential count in healthy children: role of interleukin-8 and oxidative stress

Background In healthy children, there is a wide range in the induced sputum neutrophil differential count. To date, the reason for this is unclear. In animal models, neutrophil chemoattractant cytokines and oxidative stress are major determinants of airway neutrophilia. In this study, we therefore aimed to assess the association between interleukin-8, and the oxidative stress marker 8-oxo-7, 8-dihydro-2’-deoxyguanosine (8-oxodG), with the neutrophil count in induced sputum samples from healthy children. Methods Induced sputum samples were obtained from healthy children using hypertonic saline. Concentrations of interleukin-8 and 8-oxodG were determined using ELISA. Sputum differential neutrophil count determined light microscopy and absolute neutrophil count calculated. To assess repeatability, a further sputum induction was done on a subgroup of children. Spearman's rank correlation was used to assess relationship between variables. Results The leukocyte differential was determined in 64/114 healthy children studied. The median (interquartile range) of neutrophil differential count was 20.6 % (5.67 56) and absolute neutrophil count was 0.11 (0.01 – 0.77) x 10. The induced sputum neutrophil differential and absolute counts correlated significantly with IL-8 (Rs = 0.67, p = <0.0001 and Rs = 0.60 p = <0.0001 respectively), but not with 8-oxodG (p = 0.64). The repeatability (intraclass correlation coefficient-Ri) of the neutrophil differential count was 0.58. Conclusions The normal variation in the proportion of neutrophils in the lower airway of children is driven by variation in IL-8, but not oxidative stress. The neutrophil differential count in healthy children is relatively stable over several months. Background Induced sputum (IS) is an established, non-invasive method of assessing airway inflammation in children, and normal values for the neutrophil count in IS samples in healthy children have been described. One feature of the resulting neutrophil differential is that the normal range is very wide. For example, Cai et al (reference here) reported that the interquartile range for the neutrophil differential count is 12 to 88.25% (median 35%) in children. Demographic variables may be important since a study in adults has reported that the neutrophil count increases with increasing age. However, to date no demographic variables has been reported to be associated with the neutrophil differential count in healthy children. In the diseased airway, neutrophil transmigration from the systemic circulation into the airway is mediated by the neutrophil chemoattractant cytokine Interleukin-8 (IL-8). Thus increased airway IL-8 concentrations are associated with sputum neutrophilia in asthma, chronic cough and cystic fibrosis. Thus variations in the spontaneous release of IL-8 by resident lung cells may therefore be a key determinant of neutrophil transmigration in the healthy paediatric lung. A further variable that may induce neutrophil transmigration is oxidative stress from both environmental and cellular sources. In a range of models, oxidative stress from increased free radical production, stimulates neutrophil chemoattractant release by lung cells via induction of Nuclear Factor-κB and Activator Protein –1, oxidant-sensitive transcription factors. Oxidative stress cannot be measured directly in airway samples, but oxygen radical damage to DNA and the deoxyribonucleotide pool results in, amongst others, stable products such as 8-oxo-7,8-dihydro-2’-deoxyguanosine (8-oxodG), which can be measured in biological fluids . From these data, we hypothesized that the wide range of neutrophil count in the healthy paediatric lung is a result of variations in oxidativestress, which is associated with variations in IL-8 production. We therefore sought to determine the association between both IL-8 and 8-oxodG, and the neutrophil count in induced sputum samples from healthy children. Methods Subjects and study design The study was conducted in Leicester (UK), a medium-sized city. The study protocol was approved by the Institutional Review Board (Leicestershire Research Ethics Committee). Parents of healthy children gave written, informed consent, and children gave written assent. The details of recruitment and subject characteristics are as described in previous paper. In brief, healthy children aged 815 years from nonsmoking families were recruited. We excluded children with history of respiratory symptoms, respiratory infection in last 3 months, personal smoking and passive smoking (confirmed by salivary cotinine). Sputum induction and processing Lung function was recorded using a Vitalograph 2120 spirometer (Vitalograph Ltd, Buckingham England) with Vitalograph 2120 Spirotrac IV software (Vitalograph Ltd.) as described previously . Sputum induction was done by a standard methodology using nebulised 4.5% saline via an ultrasonic nebuliser (Sonix 2000 nebuliser, Clement Clarke International, Harlow, UK) in sequential 5 min inhalations. Induced sputum was processed by a standard technique . To assess the stability (repeatability) of the neutrophil differential count, repeat sample was obtained in a subgroup of children after six months. Interleukin-8 Interleukin-8 in induced sputum supernatants was analyzed according to an established ELISA assay, using a BD OptEIA set for human IL-8 (BD Biosciences Pharmingen, San Diego, CA, US) . IL-8 was expressed as ng/ml; sensitivity level of the assay was 0.8 x 10 ng/ml. 8-oxo-7,8-dihydro-2’deoxyguanosine (8-oxodG) The 8-oxodG in sputum supernatant samples was analysed by competitive ELISA according to the manufacturer's protocol (Japan Institute for the Control of Aging, Fukuroi City, Japan). The range of the assay’s calibration curve was 0.5 – 200 ng/mL (1.77 – 706 pmol/mL). 0.1% Dithiothreitol (DTT) was added to standards and incubated as for samples to assess effect of DTT on the assay. The addition of DTT (0.1%) in the same concentration as in the sputum samples did not interfere with the ELISA for 8-oxodG (data not shown). Statistics All data were summarised as the median and interquartile range (IQR; Q1, Q3). Data were tested for normality by Kolmogorov-Smirnov test. Spearman's rank correlation was used to assess relationship between neutrophil count i) IL-8, ii) 8-oxodG and iii) demographic factors. The p values <0.05 were considered statistically significant. The repeatability on two occasions was determined by intraclass correlation coefficient (Ri) and represented graphically by plotting the difference against the mean as suggested by Bland and Altman. Statistical analysis was done using SPSS (version 12.0.1 for Windows). Results Adequate samples were obtained from 64 (35 boys) /114 children meeting the inclusion criteria. The summary of demographic, lung function variables are same as previously reported for a study of the association between airway macrophage carbon content and lung function. The median and (IQR, Q1-Q3) of squamous cell contamination (%) was 2.6 (0 8.48), indicating adequate quality sampling. As expected, the interquartile range of the neutrophil differential count was large (5.67 to 56, Table 1). There was no association between any of the demographic variables (age, height, weight and percent predicted FEV1) and the neutrophil absolute or differential count or with the IS IL8 concentration (Table 2). However, the induced sputum neutrophil differential and absolute neutrophil count correlated significantly with IL-8 (Rs = 0.67, p = <0.0001 and Rs = 0.60 p = <0.0001 respectively) (Table 2 and Figure 1). In contrast, there was no association between 8oxodG and either IL-8 concentration or neutrophil count (Table 2). The intraclass correlation coefficient (Ri) for neutrophil % was 0.58 (Figure 2), indicating a moderate degree of repeatability (stability) in the neutrophil differential count in samples taken several months apart. Discussion This is the first study to examine the determinants of neutrophil count in healthy children, and the first to measure 8-oxodG in IS samples from healthy children. The median values and range for the neutrophil differential count in our study are similar to those previously reported for normal children, suggesting that these data can be generalised to all the healthy children. In contrast to one previous study in adults , we did not find an association between age and the differential count. Although we cannot rule out an age effect, it is not significant over the small age range of subjects recruited into the present study. The key finding of this study is that, as hypothesised, the sputum neutrophil count is associated with the concentrations of IL-8 in the healthy paediatric airway. IL-8 is a potent neutrophil chemotactic factor responsible for recruitment of neutrophils in to lungs. IL-8 mediated neutrophil recruitment has been demonstrated in inflammatory conditions in children and adults ., although its role in the healthy lung is unclear. In the only previous study Gibson et al 7 reported an association between the neutrophil count and IL-8 in IS samples for 8 healthy adults. Our data therefore supports and extends this preliminary observation to healthy children. We have not, however, identified the source of IL-8 in paediatric IS samples. A potent proinflammatory resident lung cell is the alveolar macrophage, which has the capacity to both spontaneously release IL-8 and up regulate IL-8 release by epithelial cells in response to environmental stimuli 22 . We therefore speculate that the normal variation in the neutrophil count results from variations in spontaneous release of IL-8 from alveolar macrophages. The putative environmental or genetic variables associated with variations in spontaneous release therefore merit further study. We found no association between 8-oxodG and either the concentration of IL-8, or the neutrophil differential and absolute count. Thus our hypothesis that variations in IL-8 are driven by oxidative stress is not supported. Oxidative stress certainly has the capacity to induce neutrophilia through oxidant sensitive chemokines . 8-oxodG was detected in IS samples from healthy children, using a robust and reliable ELISA technique. Furthermore we excluded the possibility that DTT might interfere with the 8-oxodG assay. The presence of 8-oxodG in IS from all the healthy children studied suggests that oxidative stress, is a normal feature of the healthy paediatric airway. The origin of the 8-oxodG in IS samples remains unclear. The normal lung epithelial lining fluid contains very high levels of antioxidants, and has the capacity to rapidly neutralise free radicals. It is therefore somewhat surprising that 8-oxodG would be at a level detectable in all IS samples. We speculate therefore that this marker reflects normal intracellular oxidant generation and background level of biomolecule oxidation and repair, and hence does not represent a pathological process. There are limitations to our study. It is perhaps unclear as to whether 8-oxodG in IS indicates systemic oxidative stress, or only in the pulmonary microenvironment. The levels of 8-oxodG in extracellular matrices (e.g. plasma and urine) are considered valid markers of oxidative stress . Generally these measures are thought of a reflective of ‘whole body’ stress, although in some matrices, such as cerebrospinal fluid (Olinski ref Clin Chem) such measures may be more reflective of oxidative stress in a more localised environment. We are of the opinion that, like CSF, measurement of 8-oxodG in sputum reflects oxidative stress in the lung. We cannot be certain if other oxidative stress markers may correlate with neutrophil count. However of all the oxidative stress markers measured in extracellular matrices, 8oxodG is perhaps the best characterised, and most studied whereas, protein and lipid oxidation appearing to be less well established (Biomarkers review). Furthermore 8oxodG appears to be a sensitive biomarker of oxidative stress and eminently stable (Loft 2006) A longitudinal evaluation would strengthen the conclusions of this cross sectional study. However repeated short-term sampling of healthy children will be difficult to achieve. In conclusion, we have shown, in a large group of healthy children, that the airway neutrophil differential count is associated with IL-8. For the first time, we have demonstrated that 8-oxodG is detectable in sputum supernatant, but we found no correlation was found between neutrophil differential count and this marker of oxidative stress. Conflict of interest statement There are no conflicts of interest to disclose from all authors (NK, MSC, JG). AcknowledgementsThe project was part-funded by the Health Effects Institute, Boston USA (ResearchAgreement # 02-1). We would like to acknowledge Lucy Woodman (Institute forLung Health, Glenfield General Hospital, Leicester) and Ananth Tellabathi(University of Leicester) for help with analysis of IL-8. We acknowledge the supportof local education authorities, head teachers, and especially the parents and children. REFERENCES(1) Gibson PG, Henry RL, Thomas P. Noninvasive assessment of airwayinflammation in children: induced sputum, exhaled nitric oxide, and breathcondensate. Eur Respir J 2000; 16:1008-1015. (2) Cai Y, Carty K, Henry RL, Gibson PG. Persistence of sputum eosinophilia inchildren with controlled asthma when compared with healthy children. EurRespir J 1998; 11:848-853. (3) Zihlif N, Paraskakis E, Tripoli C, Lex C, Bush A. Markers of airwayinflammation in primary ciliary dyskinesia studied using exhaled breathcondensate. Pediatr Pulmonol 2006; 41:509-514. (4) Gibson PG, Simpson JL, Hankin R, Powell H, Henry RL. Relationshipbetween induced sputum eosinophils and the clinical pattern of childhoodasthma. Thorax 2003; 58:116-121. (5) Thomas RA, Green RH, Brightling CE, Birring SS, Parker D, Wardlaw AJ,Pavord ID. The Influence of Age on Induced Sputum Differential Cell Countsin Normal Subjects. Chest 2004; 126:1811-1814. (6) Liu L, Mul FP, Kuijpers TW, Lutter R, Roos D, Knol EF. Neutrophiltransmigration across monolayers of endothelial cells and airway epithelialcells is regulated by different mechanisms. Ann N Y Acad Sci 1996; 796:21-29. (7) Gibson PG, Simpson JL, Saltos N. Heterogeneity of Airway Inflammation inPersistent Asthma : Evidence of Neutrophilic Inflammation and IncreasedSputum Interleukin-8. Chest 2001; 119:1329-1336. (8) Jatakanon A, Lalloo UG, Lim S, Chung KF, Barnes PJ. Increased neutrophilsand cytokines, TNF-alpha and IL-8, in induced sputum of non-asthmaticpatients with chronic dry cough. Thorax 1999; 54:234-237. (9) Sagel SD, Kapsner R, Osberg I, Sontag MK, Accurso FJ. Airwayinflammation in children with cystic fibrosis and healthy children assessed bysputum induction. Am J Respir Crit Care Med 2001; 164:1425-1431. (10) Drost EM, Skwarski KM, Sauleda J, Soler N, Roca J, Agusti A, MacNee W.Oxidative stress and airway inflammation in severe exacerbations of COPD.Thorax 2005; 60:293-300. (11) Rahman I, MacNee W. Regulation of redox glutathione levels and genetranscription in lung inflammation: therapeutic approaches. Free RadicalBiology and Medicine 2000; 28:1405-1420. (12) Cooke MS, Evans MD, Herbert KE, Lunec J. Urinary 8-oxo-2'-deoxyguanosine--source, significance and supplements. Free Radic Res 2000;32:381-397. (13) Kulkarni N, Pierse N, Rushton L, Grigg J. Carbon in airway macrophages andlung function in children. N Engl J Med 2006; 355:21-30. (14) Cataldo D, Foidart JM, Lau L, Bartsch P, Djukanovic R, Louis R. Inducedsputum: comparison between isotonic and hypertonic saline solutioninhalation in patients with asthma. Chest 2001; 120:1815-1821. (15) Brightling CE, Ward R, Woltmann G, Bradding P, Sheller JR, Dworski R,Pavord ID. Induced sputum inflammatory mediator concentrations ineosinophilic bronchitis and asthma. Am J Respir Crit Care Med 2000;162:878-882. (16) Brightling CE, Monteiro W, Ward R, Parker D, Morgan MD, Wardlaw AJ,Pavord ID. Sputum eosinophilia and short-term response to prednisolone inchronic obstructive pulmonary disease: a randomised controlled trial. Lancet2000; 356:1480-1485. (17) Bland JM, Altman DG. Statistical methods for assessing agreement betweentwo methods of clinical measurement. Lancet 1986; 1:307-310. (18) Geiser T, Dewald B, Ehrengruber MU, Clark-Lewis I, Baggiolini M. Theinterleukin-8-related chemotactic cytokines GRO alpha, GRO beta, and GROgamma activate human neutrophil and basophil leukocytes. J Biol Chem 1993;268:15419-15424. (19) Norzila MZ, Fakes K, Henry RL, Simpson J, Gibson PG. Interleukin-8secretion and neutrophil recruitment accompanies induced sputum eosinophilactivation in children with acute asthma. Am J Respir Crit Care Med 2000;161:769-774. (20) Malerba M, Ricciardolo F, Radaeli A, Torregiani C, Ceriani L, Mori E,Bontempelli M, Tantucci C, Grassi V. Neutrophilic inflammation and IL-8levels in induced sputum of alpha-1-antitrypsin PiMZ subjects. Thorax 2006;61:129-133. (21) Yamamoto C, Yoneda T, Yoshikawa M, Fu A, Tokuyama T, Tsukaguchi K,Narita N. Airway inflammation in COPD assessed by sputum levels ofinterleukin-8. Chest 1997; 112:505-510. (22) Jimenez LA, Drost EM, Gilmour PS, Rahman I, Antonicelli F, Ritchie H,MacNee W, Donaldson K. PM10-exposed macrophages stimulate aproinflammatory response in lung epithelial cells via TNF-alpha. Am J PhysiolLung Cell Mol Physiol 2002; 282:L237-L248. (23) Cooke MS, Lunec J, Evans MD. Progress in the analysis of urinary oxidativeDNA damage. Free Radic Biol Med 2002; 33:1601-1614. (24) Mudway IS, Stenfors N, Duggan ST, Roxborough H, Zielinski H, MarklundSL, Blomberg A, Frew AJ, Sandstrom T, Kelly FJ. An in vitro and in vivoinvestigation of the effects of diesel exhaust on human airway lining fluidantioxidants. Archives of Biochemistry and Biophysics 2004; 423:200-212. (25) Rogan MP, Geraghty P, Greene CM, O'Neill SJ, Taggart CC, McElvaney NG.Antimicrobial proteins and polypeptides in pulmonary innate defence. RespirRes 2006; 7:29. Figure LegendsFigure 1. The scatter plot showing the significant correlation (p < 0.0001) betweensputum A) absolute neutrophil count (x 10 6 /ml) and B) neutrophil % and interleukin-8 (ng/ml). Figure 2. Bland and Altman plot of neutrophil percentage in induced sputumperformed six months apart. The limits of agreement (mean difference of twomeasurements [heavy solid line] ± 2SD) are represented by dotted lines. Ri is theintraclass correlation coefficient.