Pathogenic Bacteria Colonizing the Airways in Asymptomatic Neonates Stimulates Topical Inflammatory Mediator Release

Transcription

1 Pathogenic Bacteria Colonizing the Airways in Asymptomatic Neonates Stimulates Topical Inflammatory Mediator Release Nilofar V. Følsgaard 1, Susanne Schjørring 2, Bo L. Chawes 1, Morten A. Rasmussen 3, Karen A. Krogfelt 2, Susanne Brix 4, and Hans Bisgaard 1 1 Copenhagen Prospective Studies on Asthma in Childhood, Copenhagen University Hospital, Gentofte, Denmark; 2 Department of Microbiology and Infection Control, Statens Serum Institut, Copenhagen, Denmark; 3 Department of Food Science, University of Copenhagen, Frederiksberg C, Denmark; and 4 Department of Systems Biology, Technical University of Denmark, Lyngby, Denmark Rationale: Bacterial colonization of neonatal airways with the pathogenic bacterial species, Moraxella catarrhalis, Streptococcus pneumoniae, and Haemophilus influenzae, is associated with later development of childhood asthma. Objectives: To study a possible association between colonization with pathogenic bacterial strains and the immune signature of the upper airways in healthy neonates. Methods: A total of 20 cytokines and chemokines were quantified in vivo in the airway mucosal lining fluid of 662 neonates from the Copenhagen Prospective Study of Asthma in Childhood 2010 birth cohort. Colonization of the hypopharynx with M. catarrhalis, S. pneumoniae, H. influenzae, and Staphylococcus aureus was assessed simultaneously. The association between immune signatures and bacterial colonization or noncolonized controls was analyzed using conventional statistical methods supplemented by a multivariate approach for pattern identification. Measurements and Main Results: Colonization with M. catarrhalis and H. influenzae induced a mixed T helper cell (Th) type 1/Th2/Th17 response with high levels of IL-1b (M. catarrhalis, P ¼ ; H. influenzae, P ¼ ), TNF-a (M. catarrhalis, P ¼ ; H. influenzae, P¼ ), and macrophage inflammatory protein-1b (M. catarrhalis, P ¼ ; H. influenzae, P ¼ ). S. aureus colonization demonstrated a Th17-promoting profile with elevated IL-17 levels (P ¼ ). S. pneumoniae colonization was not significantly associated with any of the mediators. (Received in original form July 25, 2012; accepted in final form January 30, 2013) Copenhagen Prospective Study of Asthma in Childhood (COPSAC) is funded by private and public research funds listed online at The Lundbeck Foundation, the Danish Strategic Research Council, the Pharmacy Foundation of 1991, the Augustinus Foundation, The Danish Medical Research Council, and the Danish Pediatric Asthma Centre provided the core support for the COP- SAC research center. No pharmaceutical company was involved in the study. The funding agencies did not have any role in design or conduct of the study, collection, management, or interpretation of the data, or preparation, review, or approval of the manuscript. Author Contributions: The guarantor of the study is H.B., who has been responsible for the integrity of the work as a whole, from conception and design to acquisition of data, analysis and interpretation of data, andwriting ofthemanuscript. N.V.F. contributed to acquisition of data, data analyses and interpretation, and writing of the manuscript. S.B. performed analysis of cytokines, data interpretation, and writing of the manuscript. S.S. and K.A.K. performed analysis of the bacteria and critical final revision of the manuscript. B.L.C. contributed to data analyses, interpretation, and writing of the manuscript. M.A.R. performed statistical analyses. All authors made important intellectual contributions and critical final revision of the manuscript. Correspondence and requests for reprints should be addressed to Hans Bisgaard, M.D., D.M.Sc., Copenhagen Prospective Studies on Asthma in Childhood, Copenhagen University Hospital, Gentofte, Ledreborg Allé 34, DK-2820 Gentofte, Copenhagen, Denmark. bisgaard@copsac.com This article has an online supplement, which is accessible from this issue s table of contents at Am J Respir Crit Care Med Vol 187, Iss. 6, pp , Mar 15, 2013 Copyright ª 2013 by the American Thoracic Society Originally Published in Press as DOI: /rccm OC on January 31, 2013 Internet address: AT A GLANCE COMMENTARY Scientific Knowledge on the Subject Bacterial colonization of asymptomatic neonatal airways with the pathogenic bacterial species, Moraxella catarrhalis, Streptococcus pneumoniae, andhaemophilus influenzae, has been associated with later development of childhood asthma. What This Study Adds to the Field In this study, we show that this colonization triggers a sizable topical immune response, indicating that these bacteria are not just innocent bystanders, but induce significant stimulation of the immune profile of airways. Conclusions: M. catarrhalis and H. influenzae colonization of the airways of asymptomatic neonates is associated with an inflammatory immune response of the airway mucosa, which may result in chronic inflammation. Keywords: bacterial colonization; airways; neonatal; cytokine; chemokine The airway mucosa is constantly exposed to a complex environment, including micro-organisms, allergens and pollutants. Depending on the composition of this exposome, specific types of immune cells become activated, releasing a fingerprint of cytokines and chemokines in the airway mucosa (1 3). The immune competence of the new-born child is shaped by the penetration and impact of the exposome (3, 4). The ability to mount a balanced immune response is fundamental for managing healthy airways, whereas an imbalanced release of such inflammatory mediators may cause a deranged inflammatory response (5, 6). Bacterial colonization of neonatal airways with the pathogenic bacteria, Moraxella catarrhalis, Streptococcus pneumoniae, and Haemophilus influenzae, is associated with later development of asthma in early childhood (7). This finding led to the hypothesis that the pattern of immune response to pathogenic airway bacteria in infancy may be associated with the development of asthma in childhood. We therefore analyzed possible associations between colonization with pathogenic bacteria and the local immune signature of the upper airways in healthy neonates from the novel unselected Copenhagen Prospective Study of Asthma in Childhood (COPSAC 2010 ) cohort. METHODS Ethics The study was conducted in accordance with the guiding principles of the Declaration of Helsinki and approved by the Ethics Committee for Copenhagen (H-B ) and the Danish Data Protection

2 590 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Agency (j.nr ). Both parents gave their informed consent before enrollment of the children. The COPSAC 2010 Birth Cohort The novel COPSAC 2010 cohort is an ongoing, unselected clinical prospective birth cohort study of 700 children recruited in Zealand, Denmark during , the recruitment of which was previously described in detail (8). The key exclusion criteria for the pregnant women were endocrine, nephrologic, and cardiologic disease. The children were enrolled at 1 week of age, excluding anyone with severe congenital abnormality, and subsequently visited the clinical research unit at 1 month of age, where airway mucosal lining fluid and hypopharyngeal aspirate were sampled. Measurements of Airway Inflammatory Mediators Airway mucosal lining fluid was sampled at age 1 month with mm strips of filter paper (Accuwik Ultra, fibrous hydroxylatedpolyester sheets, cat. no.spr0730; Pall Life Sciences, Portsmouth, Hampshire, UK) inserted bilaterally into the nasal cavity against the anterior part of the inferior turbinate, and left there for absorption for 2 minutes, as illustrated in the video provided in the online supplement. After removal, the filter papers were immediately frozen to 2808C and stored until analysis. The mucosal lining fluid was analyzed for IL-12p70, CXCL10 (interferon g induced protein 10 kda [IP-10]), IFN-g, TNF-a, CCL4 (macrophage inflammatory protein [MIP]-1b), CCL2 (monocyte chemotactic protein [MCP]-1), CCL13 (MCP-4), IL-4, IL-5, IL-13, CCL11 (eotaxin-1), CCL26 (eotaxin-3), CCL17 (thymus and activation-regulated chemokine [TARC]), CCL22 (macrophage-derived chemokine [MDC]), IL-17, IL-1b, CXCL8 (IL-8), transforming growth factor (TGF)-b1, IL-10, and IL-2, as previously described in detail (8, 9). Selection of the measured cytokines and chemokines was decided a priori to represent mediators released from Th1, Th2, Th17, and regulatory T cells (Treg) of relevance in the pathogenesis of asthma (9 11). The lower limit of detection was set as the mean (13 SD) signal from blanks. Colonization of Pathogenic Bacteria in the Airways Bacterial colonization of the neonatal airway was investigated at the 1-month visit to the COPSAC clinic. Hypopharyngeal aspirate was collected with a soft-suction catheter passed through the nose into the hypopharynx after mucosal lining fluid collection, as previously described in detail (7). The aspirates were diluted in 1 ml sterile 0.9% NaCl and transported within 2 hour to our microbiology laboratory at Statens Serum Institut, where bacterial culturing was performed using standard methods (see the online supplement for details). Sampling of aspirates from the hypopharyngeal region, as well as sampling of airway mucosal lining fluid, was solely performed by the doctors employed at the COPSAC research unit. Statistical Analysis All data were collected online and captured in a dedicated database on a Novel SQL server. The database was double checked against source data and was subsequently locked. Data validity was assured by compliance with good clinical practice guidelines and quality control procedures. An audit trail was run routinely. The analysis plan was based on our previous findings of association between colonization with S. pneumoniae, H. influenzae, M. catarrhalis and the development of asthma (7). Therefore, colonization with these bacteria either in coculture or as monocolonization were selected as primary end points using children without colonization as control group. In a previous study, we demonstrated older siblings as a significant risk factor for colonization with pathogen bacteria (7). In addition, we recently published an association between maternal atopic disease and the cytokine/chemokine expression pattern in the neonatal airways (8). Based on those findings, older siblings and maternal atopic disease were included as covariates in all statistical models. The primary conventional statistics approach required log transformation to achieve normality of the residuals. Generalized linear model (SAS statistics; SAS Institute, Cary, NC) was used for the unadjusted as well as the adjusted analysis. To make the results comparable across mediators irrespective of several-fold differences in molecule concentrations, results are reported as ratios of the mean mediator concentration in colonized neonates and mean mediator concentration in noncolonized neonates with 95% confidence intervals. The complementary data-driven pattern analysis was conducted by partial least squares regression to extract underlying latent variables. These latent components were derived to discriminate between colonized or noncolonized with the four species, across all mentioned bacteria simultaneously, but also aim at describing the systematic part of the variation across the 20 mediators in a few uncorrelated new/latent variables (LVs), as previously described (8, 12). In this context, we used colonized (yes/no) with the four bacteria as four dependent variables, and tried to predict these based on the 20 immune mediators. We used a loading plot of the first two most descriptive latent variables to visualize the cytokine/chemokine to cytokine/chemokine correlation structure, and, furthermore, how this structure relates to the colonization by the different bacteria. Random segmented cross validation (10 segments) was used to estimate the optimal number of components and random permutation testing with 10,000 permutations for estimation of component-wise statistical inference. Score plots of the latent variables were used to investigate differences in overall cytokine/chemokine response between the children with or without colonization with individual bacteria. The data processing was conducted using SAS version 9.2 and MATLAB R2011a v (MathWorks, Natick, MA) with the partial least squares toolbox v (Eigenvector Research Inc., Wenatchee, WA) for model estimation. Visualization was done with in-house algorithms. In partial least squares regression, the Non-Linear Iterative Partial Least Squares algorithm was used to estimate the model when dealing with missing data. Additional methodological details are given in the online supplement. RESULTS Baseline Characteristics Mucosal lining fluid was sampled at age 1 month in 662 of the 700 COPSAC 2010 participants. Of the remaining 38 children, 10 had been enrolled before the sampling of mucosal lining fluid was established, 16 did not attend the 1-month visit, 3 were lost for noncompliance, and 9 were lost for other logistic reasons (Figure E1). Baseline levels and limits of detection of cytokines and chemokines are shown in Table E1 in the online supplement. In a pilot study, the first 42 neonates were analyzed before deciding to analyze all subsequent samples. This pilot study showed a general up-regulation of mediators in association with colonization, particularly with M. catarrhalis, but also with H. influenzae and S. aureus (Table E2). The main study included the subsequent 620 neonates. The mean age was 32 days (SD ¼ 6d),51%(n ¼ 319) were boys, 49% (n ¼ 309) had atopic mothers, and 61% (n ¼ 384) had older siblings. Hypopharyngeal bacterial cultures were completed in 614 of the 620 neonates (6 were lost for logistical reasons). A total of 7.3% (n ¼ 45) were colonized with S. pneumoniae, 6.7%(n ¼ 41) with H. influenzae, 19.7% (n ¼ 121) with M. catarrhalis, and 50.3% (n ¼ 309) with S. aureus. Competition was found between S. aureus (n ¼ 309) and M. catarrhalis, (n ¼ 121), with less frequent coexistence of these two dominating species compared with their individual colonization (P, 0.001). Monocolonization (i.e., colonization with one, but not the other, targeted pathogenic bacteria) with M. catarrhalis was foundin9.5%(n ¼ 59) of individuals, H. influenzae in 2.6% (n ¼ 16), S. pneumoniae in 1.9% (n ¼ 12), and S. aureus in 38.4% (n ¼ 238). A total of 196 individuals were identified without colonization of any of the pathogenic bacteria, and were used as controls in all reported analyses. For distribution of mediators in colonized and noncolonized individuals in actual concentrations (pg/ml), see Figure E2. Only older siblings were significantly associated with colonization (P ¼ 0.03), whereas we found no association between bacterial colonization and sex, birth weight, mother s atopic

3 Følsgaard, Schjørring, Chawes, et al.: Airway Colonization Stimulates Mediator Release 591 status, use of antibiotics during pregnancy, Caesarian section, Apgar score, or mother s intake of alcohol and cigarettes (Table E3). Conventional Statistical Approach Colonization with M. catarrhalis alone or in coculture with the other pathogenic bacteria was significantly associated with upregulation of IL-12p70, IP-10, IFN-g, TNF-a, MIP-1b, MCP-1, MCP-4, IL-4, IL-5, IL-13, eotaxin-1, IL-1b, IL-17, IL-8, TGF-b1, IL-10, and IL-2. In addition, eotaxin-3 and MDC showed a nonsignificant trend of up-regulation (Table 1 and Figure 1a). H. influenzae alone or in coculture with the other pathogenic bacteria was significantly associated with up-regulation of IL-12p70, IP-10, IFN-g, TNF-a, MIP-1b, MCP-1, MCP-4, IL-4, IL-5, IL-13, eotaxin-1, eotaxin-3, MDC, IL-17, IL-1b, IL-8, IL-10, and IL-2. Furthermore, TARC and TGF-b1 showed upregulated estimates, but were not statistically significant (Table 1 and Figure 1b). S. pneumoniae alone or in coculture with the other pathogenic bacteria was significantly associated with IL-12p70, IP- 10, IFN-g, TNF-a, MIP-1b, IL-4, IL-5, IL-17, IL-8, and IL-2 (Table 1andFigure1c).S. aureus alone or in coculture with the other pathogenic bacteria was significantly associated with up-regulation of IL-12p70, IP-10, IFN-g, TNF-a, MIP-1b, MCP-4, IL-4, IL-5, IL-13, eotaxin-1, IL-17, IL-1b, IL-8, TGF-b1, IL-10, and IL-2 (Table 1 and Figure 1d). Moreover,MCP-1,eotaxin-3,TARC, and MDC all showed up-regulated estimates, but were not statistically significant. For unadjusted data see Table E4. Monocolonization with M. catarrhalis, H. influenzae, or S. aureus did not differ from colonization together with the other pathogenic bacteria, whereas monocolonization with S. pneumoniae was not significantly associated with any of the mediators (Figure 2). Data-driven Approach A score plot of the latent variables (Figure 3a) confirmed the significant general up-regulation of mediators for colonized children, and could further discriminate detailed cytokine/ chemokine patterns related to the specific colonizing species showing significantly different distribution for individuals with and without colonization by the mentioned bacteria in LV1 (S. aureus, P, ; M. catarrhalis, P, ; H. influenzae, P ¼ ; but nonsignificant for S. pneumoniae [P. 0.85]). See the online supplement for further details on the partial least squares regression analysis. Figure 3b shows the multivariate correlation structure between all the cytokines and chemokines. The first two significant components (evaluated by cross validation), latent variable (LV) 1 and LV2, explain 53.5 and 4.9% of the total variation across all 20 cytokines, respectively. Therefore, LV1 shows that all cytokine and chemokines adhere to the same one-dimensional underlying pattern of up-regulation (53.5% of the variation), and all are positively correlated. LV2 reveals finer detailed patterns of the association between bacteria and distinct immune mediators as the correlation with monocolonization is imposed on the figure. The positioning of M. catarrhalis and H. influenzae is primarily positive in LV1 and slightly negative for LV2, meaning that colonization with these bacteria shows the strongest associations with increased eotaxin-3, TARC, MDC, MCP-1, IP-10, eotaxin-1, IL-12p70, IFN-g, TNF-a, MIP-1b, IL-8, and IL-5. S. aureus is primarily described by LV2 and, to lesser extent, by LV1, where IL-17 especially discriminates the pattern induced by S. aureus from those of the other bacteria. S. pneumoniae is negative in LV1 and LV2, with no association with specific mediators. DISCUSSION Main Findings Colonization with M. catarrhalis, H. influenzae, and S. aureus of the neonatal upper airways is associated with an up-regulation of key immune mediators specific to the bacterial species. S. pneumoniae was not found to increase the immune mediators by itself. The colonization triggers a sizable topical immune response, indicating that these pathogenic bacterial strains are not just innocent bystanders, but have immune-stimulating functions. Strengths and Limitations It is a strength of this study that immune profiles were assessed in the airways of healthy neonates in a standardized manner. This noninvasive method of sampling upper airway mucosal lining TABLE 1. ASSOCIATIONS BETWEEN MEDIATOR LEVELS AND COLONIZATION WITH INDIVIDUAL PATHOGENIC BACTERIA Moraxella catarrhalis Haemophilus influenzae Streptococcus pneumoniae Staphylococcus aureus Mediator Ratio (CI) P Value Ratio (CI) P Value Ratio (CI) P Value Ratio (CI) P Value IL-12p ( ) ( ) ( ) ( ) IP ( ) ( ) ( ) ( ) IFNg 2.25 ( ) ( ) ( ) ( ) TNFa 3.37 ( ) ( ) ( ) ( ) MIP-1b 3.89 ( ) ( ) ( ) ( ) MCP ( ) ( ) ( ) ( ) 0.12 MCP ( ) ( ) ( ) ( ) IL ( ) ( ) ( ) ( ) IL ( ) ( ) ( ) ( ) IL ( ) ( ) ( ) ( ) Eotaxin ( ) ( ) ( ) ( ) 0.01 Eotaxin ( ) ( ) ( ) ( ) 0.25 TARC 1.01 ( ) ( ) ( ) ( ) 0.20 MDC 1.17 ( ) ( ) ( ) ( ) 0.24 IL ( ) ( ) ( ) ( ) IL-1b 5.67 ( ) ( ) ( ) ( ) IL ( ) ( ) ( ) ( ) TGF-b ( ) ( ) ( ) ( ) IL ( ) ( ) ( ) ( ) IL ( ) ( ) ( ) ( ) Definition of abbreviations: CI ¼ confidence interval; IP ¼ interferon g induced protein; MCP ¼ monocyte chemotactic protein; MDC ¼ macrophage-derived chemokine; MIP ¼ macrophage inflammatory protein; TARC ¼ thymus and activation-regulated chemokine; TGF ¼ transforming growth factor. All analyses are adjusted for maternal atopic disease and older siblings in the home.

4 592 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL Figure 1. Ratios (95% confidence interval) between mean values of cytokines and chemokines in the upper airway mucosal lining fluid in healthy colonized neonates (alone or in coculture with the other pathogenic bacteria) compared with the noncolonized neonates (n [colonized with one or more pathogenic bacteria] ¼ 418; n [noncolonized] ¼ 196). All analyses are adjusted for maternal atopic disease and older siblings in the home. (a) Neonates colonized with M. catarrhalis (19.7%) versus no M. catarrhalis. (b) Neonates colonized with H. influenzae (6.7%) versus no H. influenzae. (c) Neonates colonized with S. pneumoniae (7.3%) versus no S. pneumoniae. (d) Neonates colonized with S. aureus (50.3%) versus no S. aureus. IP ¼ interferon g induced protein; MCP ¼ monocyte chemotactic protein; MDC ¼ macrophage-derived chemokine; MIP ¼ macrophage inflammatory protein; TARC ¼ thymus and activation-regulated chemokine; TGF ¼ transforming growth factor. fluid in neonates allowed us to assess in vivo the innate and adaptive immune responses in the airway mucosa, as reflected by a panel of cytokines and chemokines. It is an advantage of this method that mediators are measured directly in undiluted samples after in vivo collection, thereby avoiding the excessive response accompanying challenge models and the problem of dilution inherent to nasal lavage (13, 14). These in vivo measures of the mucosal immune response are more complete assessments of the integrated collaborative immune signature than a response measured in vitro, as the mediator concentrations determined in vivo are the product of networks of interactions between the host mucosal cell types reacting with the exposome. The study is highly powered with sampling in 662 neonates. This large number of study subjects allowed us to scrutinize the immune signature in neonates only colonized with one of the pathogenic bacteria (22% of the colonized neonates were colonized with more than one pathogenic bacterium). It strengthens the conclusion that the findings in the pilot study of 42 children were replicated internally in the succeeding main study. The statistical evidence is strong, with conventional statistics confirmed in the data-driven partial least squares analysis (i.e., the potential issue of multiple testing by univariate analyses of each of the 20 mediators is mitigated by the overall significance in the first latent variables; Figure 3a). Partial least squares simultaneously analyzes variation across all cytokines and chemokines, and identifies patterns and possible immune pathways elicited by the individual bacteria (Figure 3b). It is a strength that the biological sampling of mucosal lining fluid was from one of the target organs of allergic disease, although it is a limitation that the sampling was done in the upper airways, and not further down in the bronchial tree. It is also a limitation of the study that colonization status was only explored as dichotomized values (colonization: yes/no). We expect to complement the analyses with 16sRNA sequencing in the future. Another limitation to our study is that S. aureus could represent contamination from vestibulum nasi. However, in our previous birth cohort study (COPSAC 2000 ), we found S. aureus in aspirates from hypopharynx in 61% of neonates at 1 month of age, but only in 51% of cultures from vestibulum nasi (7, 15). At 12 months of age, the prevalence of S. aureus in pharyngeal aspirates from the same cohort was only 11% (15), suggesting that, although S. aureus later in life homes in on skin surfaces, it is also prevalent in the airways in the first months of life. Further limitations to the interpretation of our study are that we limited the cytokines and chemokines to 20, excluding other potentially interesting mediators, such as IL-9, IL-22, and IL-23 (16 18), and that we only assessed the immune signature at the 1-month time point, limiting conclusions on chronicity. Interpretation The specific patterns of immune activation against the four pathogenic bacteria, M. catarrhalis, H. influenzae, S. pneumoniae, and S. aureus, differ as illustrated by the differential signatures of

5 Følsgaard, Schjørring, Chawes, et al.: Airway Colonization Stimulates Mediator Release 593 Figure 2. Ratios between mean values of cytokine and chemokine in the upper airway mucosal lining fluid in healthy monocolonized neonates compared with the noncolonized group (n ¼ 196). All analyses are adjusted for maternal atopic disease and older siblings in the home. (a) M. catarrhalis monocolonized (9.5%; n ¼ 59). (b) H. influenzae monocolonized (2.6%; n ¼ 16). (c) S. pneumoniae monocolonized (1.9%; n ¼ 12). (d) S. aureus monocolonized (38.4%; n ¼ 238). IP ¼ interferon g induced protein; MCP ¼ monocyte chemotactic protein; MDC ¼ macrophagederived chemokine; MIP ¼ macrophage inflammatory protein; TARC ¼ thymus and activation-regulated chemokine; TGF ¼ transforming growth factor. individual cytokine and chemokine levels. The profile of mediators driven by the two intracellular gram-negative bacteria, M. catarrhalis and H. influenzae, was different from that of the extracellular grampositive bacterium, S. aureus, whereas the gram-positive bacterium, S. pneumoniae, showed no significant effects on individual mediators in monocolonized neonates. The immune mediators measured in this study are produced by tissue cells as well as innate and adaptive immune cells. IL-12p70, IP-10, IFN-g, TNF-a, MIP-1b, MCP-1, and MCP-4 are involved in the clearance of intracellular pathogens or pathogenic gram-negative bacteria (Th1). IL-4, IL-5, IL-13, eotaxin-1, eotaxin-3, TARC, MDC, MCP-1, and MCP-4 are Th2 markers that link to atopic disorders, and are important for immunity against parasites. IL-17, IL-1b, and IL-8 are involved in immunity against extracellular bacteria and fungi (Th17), whereas TGF-b1 and IL-10 are known to be immunosuppressive. Lastly, IL-2 is important for replication of the adaptive immune cell compartment. M. catarrhalis and H. influenzae induce similar patterns, with up-regulation of all mediators, and with no specific amplification of T cell subsets, but a mixed Th1/Th2/Th17 type response, with high levels of IL-1b, TNF-a, andalsomip-1b. The presence of both Th2 and Th17 mediators suggests that colonization with M. catarrhalis and H. influenzae recruits and activates eosinophils (IL-5) and neutrophils (IL-17, IL-8), which may counteract the Th1 response needed to eradicate these intracellular bacteria (19 21). This seems an inadequate response to colonization with intracellular bacteria (22), which may lead to the inception of a chronic inflammation. S. pneumoniae caused a weaker (cocolonized) or no (monocolonized) response of the inflammatory cytokines and chemokines, which was confirmed in the score plot analysis model, where the response of S. pneumoniae was not significantly different from that of the noncolonized individuals (Figure 3a). Previous studies have reported that colonization with gram-negative bacteria of the airways induces a stronger response of cytokines in the mucosal lining fluid than gram-positive bacteria, supporting our S. pneumoniae results (20). The lack of response to a known pathogenic bacterium may hamper the infant s ability to clear infections with S. pneumoniae. Thus, the deficient immune activation against S. pneumoniae might allow immune evasion by S. pneumoniae (23, 24). S. aureus, on the other hand, demonstrated a strong neutrophilassisted type of immune response, dominated by IL-17, IL-1b, and IL-8 (IL-17 concentration was 5.4-fold higher in colonized compared with noncolonized neonates; P ¼ ), but less up-regulation of factors involved in expansion of Th1, Th2, or regulatory T cells, or of innate cell mediators linked to these T cell subsets. This neutrophil-directed immune response against the extracellular, gram-positive bacterium, S. aureus, seems appropriate (25, 26), and aligns with our previous observation that the prevalence of S. aureus in the airways drops after the first months of life and with no association with the long-term risk of asthma during childhood (7). The species-dependent differences in the immune profiles mimic our previous clinical findings, where M. catarrhalis and H. influenzae colonization in neonates showed a strong association with the development of asthma later in childhood, whereas S. pneumoniae showed similar, albeit weak, association, and S. aureus was not associated with asthma development (7).

6 594 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL and (3) colonization with S. aureus caused a strong Th17 response. Therefore, colonization of the airways of asymptomatic neonates with pathogenic bacteria induces species-specific and significant stimulations of the immune response of the airway mucosa, indicating that particularly M. catarrhalis and H. influenzae, and also S. pneumoniae, are not just innocent bystanders. Author disclosures are available with the text of this article at Acknowledgment: The authors thank the children and parents participating in the Copenhagen Prospective Study of Asthma in Childhood (COPSAC) 2010 cohort, as well as the COPSAC study team. Figure 3. (a) Score plot from partial least squares (PLS) model. Each point corresponds to one child where the colors indicate colonization with specific bacteria. The distribution of the specific colonizations is shown as ellipses with coverage of 47% (1 SD in two dimensions). Differences are registered between noncolonized and monocolonized Staphylococcus aureus, Haemophilus influenzae, and Moraxella catarrhalis, whereas the pattern of noncolonized and Streptococcus pneumoniae monocolonized children are much alike. (b) Loading plot from PLS regression model predicting colonization status for four bacteria (M. catarrhalis, H. influenzae, S. pneumoniae, ands. aureus) markedwith red dots. The different cytokines are marked and colored according to their regulatory functions (T helper cell [Th] type 1, Th2, Th17, and regulatory T cells [Treg]). Latent variable (LV) 1 and LV2 explain 53.5 and 4.9% of the total cytokine/chemokine variation, respectively. IP ¼ interferon g induced protein; MCP ¼ monocyte chemotactic protein; MDC ¼ macrophage-derived chemokine; MIP ¼ macrophage inflammatory protein; TARC ¼ thymus and activation-regulated chemokine; TGF ¼ transforming growth factor. Conclusions Colonization of the neonatal airways with pathogenic bacteria showed three distinctly different profiles of immune response: (1) colonization with M. catarrhalis and H. influenzae promoted a mixed Th1/Th2/Th17 type response, which may suggest induction of a chronic inflammation; (2) monocolonization with S. pneumoniae caused no activation of mediators allowing immune evasion, which may hamper eradication of colonization; References 1. Lappalainen MHJ, Hyvärinen A, Hirvonen M-R, Rintala H, Roivainen J, Renz H, Pfefferle PI, Nevalainen A, Roponen M, Pekkanen J, et al. High indoor microbial levels are associated with reduced Th1 cytokine secretion capacity in infancy. Int Arch Allergy Immunol 2012; 159: Chromy BA, Fodor IK, Montgomery NK, Luciw PA, McCutchen- Maloney SL. Cluster analysis of host cytokine responses to biodefense pathogens in a whole blood ex vivo exposure model (WEEM). BMC Microbiol 2012;12: Martino D, Prescott S. Epigenetics and prenatal influences on asthma and allergic airways disease. Chest 2011;139: Prescott SL. Early origins of allergic disease: a review of processes and influences during early immune development. Curr Opin Allergy Clin Immunol 2003;3: Liu AH, Murphy JR. Hygiene hypothesis: fact or fiction? J Allergy Clin Immunol 2003;111: Martinez FD, Holt PG. Role of microbial burden in aetiology of allergy and asthma. Lancet 1999;354:SII12 SII Bisgaard H, Hermansen MN, Buchvald F, Loland L, Halkjaer LB, Bønnelykke K, Brasholt M, Heltberg A, Vissing NH, Thorsen SV, et al. Childhood asthma after bacterial colonization of the airway in neonates. N Engl J Med 2007;357: Følsgaard NV, Chawes BL, Rasmussen MA, Bischoff AL, Carson CG, Stokholm J, Pedersen L, Hansel TT, Bønnelykke K, Brix S, et al. Neonatal cytokine profile in the airway mucosal lining fluid is skewed by maternal atopy. AmJRespirCritCareMed2012;185: Chawes BLK, Edwards MJ, Shamji B, Walker C, Nicholson GC, Tan AJ,FølsgaardNV,BønnelykkeK,BisgaardH,HanselTT.Anovel method for assessing unchallenged levels of mediators in nasal epithelial lining fluid. J Allergy Clin Immunol 2010;125: e Prescott SL. New concepts of cytokines in asthma: is the Th2/Th1 paradigm out the window? J Paediatr Child Health 2003;39: Ngoc PL, Ngoc LP, Gold DR, Tzianabos AO, Weiss ST, Celedón JC. Cytokines, allergy, and asthma. Curr Opin Allergy Clin Immunol 2005;5: Rasmussen MA, Skov J, Bladbjerg E-M, Sidelmann JJ, Vamosi M, Jespersen J. Multivariate analysis of the relation between diet and warfarin dose. Eur J Clin Pharmacol 2012;68: Bisgaard H, Krogsgaard OW, Mygind N. Measurement of secretion in nasal lavage. Clin Sci 1987;73: Pipkorn U, Proud D, Lichtenstein LM, Kagey-Sobotka A, Norman PS, Naclerio RM. Inhibition of mediator release in allergic rhinitis by pretreatment with topical glucocorticosteroids. N Engl J Med 1987; 316: Skov L, Halkjaer LB, Agner T, Frimodt-Møller N, Jarløv JO, Bisgaard H. Neonatal colonization with Staphylococcus aureus is not associated with development of atopic dermatitis. Br J Dermatol 2009;160: Akdis M, Palomares O, Van de Veen W, Van Splunter M, Akdis CAT. TH17 and TH22 cells: a confusion of antimicrobial response with tissue inflammation versus protection. J Allergy Clin Immunol 2012; 129: Miossec P, Korn T, Kuchroo VK. Interleukin-17 and type 17 helper T cells. N Engl J Med 2009;361:

7 Følsgaard, Schjørring, Chawes, et al.: Airway Colonization Stimulates Mediator Release Korn T, Bettelli E, Oukka M, Kuchroo VK. IL-17 and Th17 Cells. Annu Rev Immunol 2009;27: Zhu J, Yamane H, Paul WE. Differentiation of effector CD4 T cell populations. Annu Rev Immunol 2010;28: De Dooy J, Ieven M, Stevens W, Schuerwegh A, Mahieu L. Endotracheal colonization at birth is associated with a pathogen-dependent pro- and antiinflammatory cytokine response in ventilated preterm infants: a prospective cohort study. Pediatr Res 2004;56: King PT, Hutchinson PE, Johnson PD, Holmes PW, Freezer NJ, Holdsworth SR. Adaptive immunity to nontypeable Haemophilus influenzae. Am J Respir Crit Care Med 2003;167: Nakayamada S, Takahashi H, Kanno Y, O Shea JJ. Helper T cell diversity and plasticity. Curr Opin Immunol 2012;24: Gamez G, Hammerschmidt S. Combat pneumococcal infections: adhesins as candidates for protein-based vaccine development. Curr Drug Targets 2012;13: Izoré T, Contreras-Martel C, El Mortaji L, Manzano C, Terrasse R, Vernet T, Di Guilmi AM, Dessen A. Structural basis of host cell recognition by the pilus adhesin from Streptococcus pneumoniae. Structure 2010;18: Milner JD, Brenchley JM, Laurence A, Freeman AF, Hill BJ, Elias KM, Kanno Y, Spalding C, Elloumi HZ, Paulson ML, et al. Impaired T(H) 17 cell differentiation in subjects with autosomal dominant hyper-ige syndrome. Nature 2008;452: O Connor W Jr, Zenewicz LA, Flavell RA. The dual nature of T(H)17 cells: shifting the focus to function. Nat Immunol 2010;11: