A rich microbial environment in infancy protects against asthma ,  and infections precipitate asthma exacerbations . We compared the airway microbiota at three levels in adult patients with asthma, the related condition of COPD, and controls. We also studied bronchial lavage from asthmatic children and controls.
We identified 5,054 16S rRNA bacterial sequences from 43 subjects, detecting >70% of species present. The bronchial tree was not sterile, and contained a mean of 2,000 bacterial genomes per cm2 surface sampled. Pathogenic Proteobacteria, particularly Haemophilus spp., were much more frequent in bronchi of adult asthmatics or patients with COPD than controls. We found similar highly significant increases in Proteobacteria in asthmatic children. Conversely, Bacteroidetes, particularly Prevotella spp., were more frequent in controls than adult or child asthmatics or COPD patients.
Citation: Hilty M, Burke C, Pedro H, Cardenas P, Bush A, Bossley C, et al. (2010) Disordered Microbial Communities in Asthmatic Airways. PLoS ONE5(1): e8578. https://doi.org/10.1371/journal.pone.0008578
Editor: Olivier Neyrolles, Institut de Pharmacologie et de Biologie Structurale, France
Received: October 8, 2009; Accepted: December 12, 2009; Published: January 5, 2010
Copyright: © 2010 Hilty et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The study was funded by the GABRIEL FP6 Integrated Project of the European Commission and the Wellcome Trust. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Asthma is a heterogeneous syndrome of intermittent wheeze and airway inflammation that affects 300 million individuals worldwide. Although its causes are unknown, many studies suggest a role for microbiota in its aetiology . Viral infections are important inducers of seasonal exacerbations of asthma , but there is circumstantial evidence that bacterial infections may also play a role. Asymptomatic neonates whose throats are colonized with Streptococcus pneumoniae, Haemophilus influenzae, or Moraxella catarrhalis are at increased risk for recurrent wheeze and asthma early in life . These same bacteria have consistently been associated with exacerbations of both asthma  and chronic obstructive pulmonary disease (COPD) . The response of asthmatics to antibiotics also suggests the importance of acute and chronic bacterial infections in the pathogenesis of disease . Epidemiological research has consistently indicated that a rich microbial environment in early life confers protection against the development of asthma , suggesting the need to understand the extent and nature of normal airway flora.
Apart from asthma, respiratory infections, excluding tuberculosis, cause 6% of the global burden of disease and each year 4.2 million people die of lower respiratory infections. Significantly, death in the 1918–1919, 1957 and 1968 influenza pandemics resulted most commonly from secondary bacterial pneumonia caused by organisms presumed to be previously present in the respiratory tract , .
Chronic Obstructive Pulmonary Disease (COPD) shares many features with asthma and is the fourth leading cause of death worldwide. Infectious exacerbations of the disease are a frequent cause of death, and chronic infection causes a progressive decline of lung function.
Thus, despite strong evidence to implicate bacterial infections in the course and pathogenesis of airway diseases, it is unfortunate that a systematic study of organisms in the airways has been lacking .
Only 1% of all bacteria can be cultured in the laboratory , so culture is no longer the gold standard for the diagnosis of infections. Culture-independent molecular methods have already shown that the microbiota of humans is far greater in extent than previously recognised , , . Humans are recognised to have evolved relationships with their symbiotic bacteria that are essential for health. Ecological changes altering this symbiosis can result in disease .
We therefore used molecular analysis of the polymorphic bacterial 16S-rRNA gene to characterize the composition of bacterial communities from the airways of adult subjects including patients with asthma and COPD. We sought replication of findings from these adults in an additional study of children attending clinics for therapy-resistant asthma.
Twenty-four adult subjects were studied, including 5 patients with COPD, 11 patients with asthma and 8 control subjects with no previous history of asthma or COPD and an FEV1≥95% predicted (Table 1). Sterile dry cotton-headed swabs were used to sample from the nose and the oropharynx (OP) of all subjects (n = 24), and bronchoscopy via the nasal route was used to obtain duplicate cytology brushings within the left upper (LUL) in 23 subjects. All subjects were free of clinical infection at the time of the study.
We sequenced clone libraries from the 16S rRNA PCR products for 23, 24 and 21 samples from the nose, OP and LUL respectively. Very weak amplification products were obtained from three subjects with the lowest bacterial copy numbers and we did not identify bacterial sequences from these libraries. After excluding possible chimeras, 1,081 nasal, 1,100 OP and 1068 LUL sequences were analysed further.
Distinct individual bacterial sequences (defined as operational taxonomic units (OTUs)) were identified with the DOTUR program. We estimated bacterial diversity, or the number of different types of bacteria (OTUs) likely to be present, with Chao 1 statistics (using a cut-off of ≥97% sequence identity). This analysis suggested the presence of 126 OTUs (95% Confidence Interval [CI] 97.7–194.8) in the nasal samples, 229 (95%CI 142.9–434.1) OTUs in the OP and 128 (95%CI 106.9–179.0) in the LUL samples. These estimates indicated that we had identified 62% of all OTUs in the nose and 68% of LUL OTUs, but only 35% of the diversity in the OP. The values in the nose and LUL are comparable with studies estimated to have identified 68.3% and 74.0% of OTUs in the human distal oesophagus  and forearm superficial skin  respectively.
We combined nose, OP and LUL sequences together to produce a global DOTUR analysis for the airways. This identified 190 OTUs of which 97 had more than a single sequence. Highly abundant OTUs belonged to the genera Prevotella, Streptococcus, Staphylococcus, Neisseria, Corynebacterium and to Haemophilus spp. (Figure 1).
The numbers of 16S rRNA gene phylotypes (OTUs) were calculated at 97% sequence identity and single-sequence OTUs omitted. OTU designations are located at the termination of each branch and represent potential organism names. Abundance of OTUs in each subject is indicated by different coloured squares (Yellow = 1 single OTU, Orange = 3–10%, Red = 10–20% and Black≥20%).
We used cladistic analysis to examine the distribution of phyla and highly abundant genera at different airway levels (Figure 2 and Table 2). Nasal specimens were highly significantly characterised by Actinobacteria (particularly Corynebacterium spp.) (Pc<10−51) and Firmicutes (particularly Staphylococcus spp.) (Pc<10−89). The most common bacteria in the OP were Bacteroidetes (particularly Prevotella spp.) (Pc<10−25), and LUL samples contained many more Haemophilus spp. than the other sites (Pc<10−10).
A whole microbial community comparison showed the nasal microbiota for all phenotypes to be clearly distinct from that of the OP and LUL (Figure 3). The OP of healthy controls clustered with OP samples of asthmatics and also with the controls' LUL. The LUL samples of patients with asthma and COPD clustered together, along with the OP samples of patients with COPD. These findings suggested that the microbiota of the healthy LUL was similar to that of the healthy OP, but the LUL microbiota differed from normal OP and LUL samples in the presence of asthma and COPD. The analysis also suggested that the microbiota in patients with COPD was altered in the OP as well as the LUL, perhaps due to exposure to cigarette smoke both regions.
We next studied the LUL samples in more detail for their relationship to airway disease. Both series of 23 specimens from the left upper lobe (LUL) were analysed in triplicate for bacterial counts. The 16S rRNA copy number estimates indicated a range of 2–7,500 bacterial genomes per 40ng total DNA, or a mean of 2.2×103 bacterial genomes per cm2 surface sampled (range 62–2.1×105) based on an estimated brushed area of 2 cm2 and one copy of the 16S rRNA gene per genome. The repeat LUL brushings gave highly reproducible counts (Spearman's rho = 0.667, P<0.001). We excluded sample-specific PCR inhibition in subjects with low copy numbers by demonstrating amplification of human DNA with equivalent intensity for all samples.
Our mean estimate of 2,000 bacterial genomes per cm2 surface sampled from LUL brushings is comparable with estimates from the upper two thirds of the small intestine . The airway counts are an order of magnitude below the 50,000 bacteria/cm2 identified by 16S rRNA amplification of scrapings from normal skin , although the maximum count of 20,000 we observed in an asthmatic subject approximates this figure.
The genome copy number estimates were significantly lower in smokers (Mann-Whitney U test Z = −3.92, P<0.001). Similar effects of smoking were observed in the duplicate samples (Z = −4.25, P<0.001) and in the nasal counts (Z = −2.38, P = 0.016). The specimens from the RLL were in the same range as for the LUL. No differences in the counts were found with respect to age or sex. It may be relevant that cigarette smoking has also been shown to influence bacterial growth in inflammatory bowel disease (IBD) .
We performed cladistic association analysis to test for differences in the frequency of particular phyla and genera between subject groups (Figure 4a and Tables 3 and 5). We found members of the phylum Proteobacteria to occur highly significantly more often in the COPD (Pc<10−14) and asthma (Pc<10−13) groups compared to controls (Table 3). In contrast, Bacteroidetes members were more common in controls compared to diseased subjects (Pc<10−4 controls versus COPD; Pc<10−2 controls versus asthma). Members of the genus Haemophilus were most strongly associated with the presence of COPD (Pc<10−4) and asthma (Pc<10−7) (Table 1d).
a) Distribution of the phyla from sheathed bronchoscopic brushings of the LUL for patients with asthma and COPD and normal subjects, subdivided into the seven most frequent genera (Croynebacterium, Prevotella, Staphylococcus, Streptococcus, Veilonella, Haemophilus and Neisseria). b) Distribution of the phyla from bronhco-alveolar lavage (BAL) in children with difficult asthma and controls.
In order to confirm if the same bacteria were relevant in health and disease in a different set of patients we studied an additional 13 (5 female) asthmatic and 7 (3 female) control individuals recruited through a tertiary referral clinic for difficult asthma at the Royal Brompton Hospital (Table 4). The median age of the asthmatic children was 11.8±2.8 years and that of the controls was 11.3±5.7. None of the subjects had been taking antibiotics prior to examination and all subjects were considered free of infection. Broncho-alveolar lavage (BAL) was performed through a fibre-optic bronchoscope, DNA extracted, and clone libraries created and sequenced.
We obtained 1135 and 670 sequences respectively for the asthmatic and control individuals. Based on 97% sequence identity, this resulted in 97 (124.3 95%CI 108.3–163.2) and 68 (85.1 95%CI 73.7–119.4) OTUs for the asthma and control individuals. Taking all subjects together, 131 OTUs were detected.
We found from the distribution of the phyla that Proteobacteria were more frequent in asthmatic children compared to controls (Pc<10−13), whereas Bacterioidetes were more common in the controls than in asthmatics (Pc<10−16) (Figure 4b and Table 5). At the level of genera, the asthmatic subjects were characterised by excess of Haemophilus spp. (Pc<10−12) and Staphylococcus spp. (Pc<10−2), whereas Prevotella spp. were more frequent in controls (Pc<10−30).
Our results challenge the traditional medical teaching that the lower airways are sterile, and suggest that bronchial tree contains a characteristic microbial flora that differs between health and disease. The concept of sterility comes from an age with limited experimental access to healthy airways and was based on 120 year old culture techniques which are now recognised to detect a small percentage of the bacteria present in complex samples . Indeed, the major colonists in our control subjects were anaerobes such as Prevotella spp., which can only be grown with difficulty in culture. Microbiota are ubiquitous even in the most hostile environments, and it would be extraordinary if the lower airway able to maintain sterility in the presence of high volume airflow through a damp open communication with the oropharynx.
Our results show strong similarities in the airway microbiota between adult and paediatric asthmatics and controls (Figure 4), despite sampling by two alternative methods (bronchial brushing and BAL). Strikingly, members of the phylum Proteobacteria were very strongly associated with airway disease in both groups. The phylum contained the important potential pathogens Haemophilus, Moraxella and Neisseria spp.
These associations, although robust, do not yet establish causality between the presence of pathogens and airways disease. The numbers of subjects are small and may not be representative of most cases of asthma. Our adult controls were healthy, but the childhood controls had clinical indications for their bronchoscopy including tracheomalacia, gastro-oesophageal reflux, and symptomatic VSDs which might all cause airway mucosal abnormalities. It is also not known to what extent past antibiotic usage in response to symptoms of airway disease may have determined the patterns of bacterial species we have detected. It is also possible that the presence of pathogens in asthmatic airways is secondary to mucosal abnormalities of as yet unknown cause.
Despite these cautions, it is striking that Haemophilus, Moraxella and Neisseria spp. found by us in asthmatic airways have also been shown to carry increased risk for asthma in early life when found in neonatal throat swabs . It should therefore be considered possible that the presence of these pathogens may have some influence on chronic airway inflammation, even when below the threshold for routine culture. In addition, the presence of a low copy-number yet dominant population of potential pathogens may provide a substrate for full blown bacterial infection following otherwise self-limiting viral infections  as well as following pandemic influenza , .
We are not aware of any trials of antibiotics in asthma other than in the management of acute exacerbations , but It may be relevant that protracted course of antibiotics are required for successful therapy of the asthma overlap syndrome of chronic wet cough . The availability of effective vaccines against Haemophilus influenzae might also open new therapeutic possibilities for individuals whose airways disease is characterised by the continued presence of this pathogen.
Our data suggest that Staphylococcus spp. are also present in excess in the airways of children with difficult asthma, and it is likely that wider exploration of children and adults with asthma will identify a range of organisms associated with the disease, including Mycoplasma spp. and Chlamydia spp.  and Streptococcus spp. , and possibly defining subtypes of the syndrome. Longitudinal studies are desirable to follow changes in the microbiota during the acute exacerbations that are a prominent feature of asthma and COPD.
Bacteroidetes (particularly Prevotella spp.) were more common in control subjects than in diseased individuals. Prevotella are Gram-negative anaerobes recognised to be part of the normal oral  and vaginal flora and are the predominant anaerobic Gram-negative bacilli isolated from respiratory tract infections and their complications. They have been found in high numbers by 16S-rRNA sequencing in the lungs of children with cystic fibrosis . Our results suggest that Prevotella and Veillonella spp. are a distinctive component of the normal flora of the lung as well as other moist epithelia, and that their presence in normal airways may not have been generally recognised because of their requirement for anaerobic culture.
The reduction of Bacteroidetes in patients with asthma and COPD compared to controls may also be relevant to disease. An investigation of patients with IBD showed a 10-fold decrease in total bacterial load and a diminished quantity of Bacteroidetes compared with controls , consistent with findings in our patients. It may be relevant that Prevotella spp. directly inhibit the growth of other bacteria . Commensal bacteria elicit tonic signals in the gut epithelium that prevent activation of innate and adaptive immune responses , , and commensal bacteria down-regulate immune responses to pathogens in the nasal mucosa . It is possible that similar mechanisms may characterise healthy airways.
Materials and Methods
Approval for the adult study was given by the Connolly Hospital Ethics Committee, and full informed consent was obtained from each patient in writing. Approval for the study of the pediatric patients was given Royal Brompton Hospital Ethics Committee, and written informed consent from parents and age-appropriate assent from children was obtained in each case.
Twenty-four adult subjects were studied, consisting of 5 patients with COPD, 11 patients with asthma and 8 control subjects with no previous history of asthma or COPD and an FEV1≥95% predicted (Table 1). Asthma severity was based on the GINA criteria (http://www.ginasthma.com/). Smoking was defined as current/recent (within the past two years), past (non-smoker for more than two years) and never. Subjects were included in the study regardless of smoking status. All patients were of European descent and from Dublin. The patients received their usual medical treatment in the 24 hours prior to this investigation. No patients were taking antibiotics at the time of the study, and all subjects were considered to be free of clinical infections.
Sterile dry cotton-headed swabs were used to sample from the nose and the oropharynx (OP) of all subjects (n = 24). Following Propofol as a short-acting sedative agent and local Lidocaine anaesthesia to the upper airway, bronchoscopy was performed via the nasal route. Olympus Endotherapy disposable sheathed cytology brushes (Model number BC-202D-5010) 5 mm diameter and 10 mm length were used to brush within the left upper (LUL) and right lower lobes (RLL). Duplicate brushings were done for the LUL, attempting in each case to reach a more distal area. One patient with severe asthma was unable to tolerate a bronchoscopy. LUL brushings were therefore performed for 23 subjects. Fourteen of the subjects (five controls; three COPD subjects and six asthma subjects) were sufficiently comfortable to allow sampling of their RLLs. Approximately 2 cm2 of bronchial mucosa was covered by each brushing. Samples were either processed immediately or stored in 2 ml tubes over night at −20°C prior to being processed.
We also studied thirteen paediatric patients with difficult asthma and seven non-asthmatic controls (Table 2). Difficult asthma was defined as symptoms requiring a rescue bronchodilator on at least three days per week, despite treatment with at least 1600 µg/day of inhaled budesonide (or equivalent), as well as regular long acting β2 agonists (or a previous unsuccessful trial of long acting β2 agonists), and/or regular prednisolone . None of the subjects were taking antibiotics at the time of the study and all were considered to be free of clinical infections. Fifteen children were Caucasian, three were of African origin and one child was of mixed origin (Table 4). The study had full ethical approval and all subjects or their parents gave written informed consent.
Bronchoscopies in the paediatric subjects were performed as previously described . For the BAL, three aliquots of normal saline (1 ml/kg, maximum 40 ml) were instilled in the right middle lobe through the bronchoscope and the fluid retrieved by mechanical suction. Between 0.4–8ml of BAL fluids were retrieved and centrifuged and the pellet kept for subsequent DNA extraction. In the case of three patients the pellets had been used for a different study. We therefore again centrifuged the remaining supernatants and processed these secondary pellets.
Samples were suspended in 450µl lysis solution (20 mg/ml lysozyme; 20 mM Tris·HCl, pH 8.0; 2 mM EDTA) and incubated for 1h at 37°C with occasional vortexing. Subsequent DNA extraction was performed as instructed by the manufacturer (DNeasy, Qiagen). The purified DNA was resuspended in 40µl of analar water and the DNA concentration determined using a Nanodrop ND-1000 spectrophotometer. For long term storage samples were kept at −20°C. Working stocks containing 20 ng/µl and 10 g/µl total DNA were prepared for samples from the LUL, RLL and nose and these stocks were stored at 4°C.
To quantify the bacterial DNA content of each sample, 2µl of each prepared working stock was used as template in PCR reactions containing 16S rRNA gene primers 63F-5′ GCAGGCCTAACACATGCAAGTC-3′ and 355R-5′ CTGCTGCCTCCCGTAGGAGT-3′ (final primer concentration 0.2 µM) with reaction and cycling conditions as previously described . A standard curve was included, amplifying serial dilutions of known quantities of Escherichia coli cells in the presence of 20 ng/µl human DNA. For the left upper and right lower lobe samples the functions used to calculate copy numbers were: Ct1 = −3.62x+35.46 with R2 = 0.994; Ct2 = −3.76x+36.30 with R2 = 0.996 and Ct3 = −3.73x+36.05 with R2 = 0.996. For the nasal swabs, the function used was: CtN = −3.70x+35.91 with R2 = 0.997. Copy numbers were calculated with E. coli having 4,500,000 bp. Initial experiments revealed no inhibition of amplification of E. coli in the presence of human DNA and confirms findings described elsewhere . Differences in genome count numbers were assessed using non-parametric tests in SPPS 16.0 for Windows.
For the sequencing analysis 16S rRNA genes were amplified using the primers 339F-5′-ACTCCTACGGGAGGCAGCAGT-3′  and 907R-5′-CCGTCAATTCMTTTGAGTTT-3′ . 25µl PCR reactions were set up containing 2µl of template, 2.5µl 10× buffer (Roche Applied Biosystems), and final concentrations of 1.5 mM Mg, 6% DMSO, 0.5 µM primers, 0.8 mM total dNTP (0.2 mM each) and 1U HotStart Polymerase (Roche Applied Biosystems). Cycling conditions were one cycle of 95°C for 5 min, then 32 cycles of 95°C for 30s, 55°C for 30s and 72°C for 60s followed by a final extension of 72°C for 5 min.
Following amplification, the PCR products were separated on a 1% agarose gel, and bands corresponding to the target product (∼550bp) were excised. Cloning and sequencing of the PCR products was possible for 21, 23, 24 and 20 primary LUL, nose, OP and BAL samples respectively. A low concentration of amplification product of 3 samples was attributed to low bacterial counts and prevented reliable sequencing. PCR products were extracted from the gel slices using the Qiaquick extraction kit according to the manufacturer's instructions (Qiagen) and resuspended in 40µl of analar water and stored at 4°C until further use.
Purified PCR products were cloned into the pGEM-T Easy Vector (Promega) as per the manufacturer's instructions. Between 48 and 72 colonies per ligation were picked at random. Plasmid DNA was purified using the Qiagen Plasmid Miniprep Kit according to the manufacturer's instructions and sequenced using the primer SP6. In instances when sequence from the SP6 primer was of low quality, the alternate DNA strand was sequenced using the T7 primer. DNA sequence chromatograms were uploaded to the ribosomal database (http://rdp.cme.msu.edu/), vector sequences trimmed (using the LUCY program) and remaining sequences quality control checked (using PHRED). Sequences of high quality were downloaded, primer regions additionally trimmed and only full length sequences ranging from 359 up to 906 (E. coli numbering of the 16S rRNA gene) were included for further analysis.
Sequences were aligned with NAST  and ClustalW. Alignments were manually curated and results checked for chimeras with Bellerophon Version 3  on the Greengenes website (http://greengenes.lbl.gov). Using slightly adjusted standard settings (match length to core set = 400 and window size of 200) putative chimeras were identified and excluded from subsequent analyses.
Distance matrices calculating the pair wise distances of the high quality sequences was created on the green genes website (http://greengenes.lbl.gov/cgi-bin/nph-index.cgi), masking the hyper-variable regions with the lane mask utility. OTUs were designated using DOTUR (Distance based OTU and richness determination)  by the furthest-neighbour algorithm. Chao 1 richness measures of diversity were estimated by the same program. For the creation of the phylogenetic tree, a cut off of 97% was set and each OTU was then assigned an organism name based on the phylogenetic placement using sequence match of the ribosomal database (20 best match sequences define taxanomic rank) and compared to NCBI BLAST results for the percentage of sequence identity.
Cladistic association analysis was performed as a method for association mapping . The approach was based on the observation that every edge in a phylogenetic tree defines a split of the OTUs. P-values were calculated for association between the groups and the split and for association between disease status of individuals and the split. Two-tailed Fisher's exact tests were performed for each split, followed by the application of a conservative Bonferroni correction  for multiple testing.
The UNIFRAC metric measures the distance between communities as the percentage of branch length that leads to descendants from only one of a pair of environments represented in a single phylogenetic tree, i.e. the fraction of evolution that is unique to one of the microbial communities. The metric thus reflects differences between the lineages that are adapted to live specifically in one environment or the others . A phylogenetic tree based on 16S rRNA DNA sequences was used as input to measure the difference between bacterial communities in asthma, COPD and controls by using UNIFRAC.
For Figure 1, the numbers of 16S rRNA gene phylotypes (OTUs) were calculated at 99% sequence identity with the furthest neighbour clustering in the program DOTUR. Using a single sequence for each OTU, evolutionary distances were calculated with the Jukes-Cantor (tree-Markov) model and phlyogenetic trees determined by the Neighbour-Joining method, with 1000 trees generated. Bootstrap confidence levels were shown at the tree nodes for values for ≥50%. Designations of Streptococcus mitis and Neisseria mucosa/sicca/flava were as described previously , .
Bacterial community comparisons of PCR libraries were determined with weighted UniFrac and the relationships are displayed by the Unweighted Pair Group Method with Arithmetic mean (UPGMA). Sequences of individuals with the same phenotype and clinical site were pooled and sequence numbers annotated. The tree with 100% identity was used as input. Only individuals from whom all 3 sites resulted in bacterial sequences are included (20 of 24 individuals; 2,871 of 3,259 sequences).
Sequences are stored as GenBank accession GQ360090-GQ365143.
We thank Jan Henson at the National Jewish Hospital in Denver for her encouragement to pursue this study.
Conceived and designed the experiments: MH CB LP MFM WOCC. Performed the experiments: MH CB PC CB JCD AE LP. Analyzed the data: MH HP PC LP MFM WOCC. Contributed reagents/materials/analysis tools: CB AB CB JCD. Wrote the paper: MH MFM WOCC.
- 1. Eder W, Ege MJ, von Mutius E (2006) The asthma epidemic. N Engl J Med 355: 2226–2235.W. EderMJ EgeE. von Mutius2006The asthma epidemic.N Engl J Med35522262235
- 2. Cookson WO, Moffatt MF (1997) Asthma: an epidemic in the absence of infection? [comment]. Science 275: 41–42.WO CooksonMF Moffatt1997Asthma: an epidemic in the absence of infection? [comment].Science2754142
- 3. Sykes A, Johnston SL (2008) Etiology of asthma exacerbations. J Allergy Clin Immunol 122: 685–688.A. SykesSL Johnston2008Etiology of asthma exacerbations.J Allergy Clin Immunol122685688
- 4. Cookson W (2004) The immunogenetics of asthma and eczema: a new focus on the epithelium. Nat Rev Immunol 4: 978–988.W. Cookson2004The immunogenetics of asthma and eczema: a new focus on the epithelium.Nat Rev Immunol4978988
- 5. Johnston S, Pattemore P, Sanderson G, Smith S, Lampe F, et al. (1995) Community study of role of viral infections in exacerbations of asthma in 9–11 year old children. BMJ 310: p1225–1229.S. JohnstonP. PattemoreG. SandersonS. SmithF. Lampe1995Community study of role of viral infections in exacerbations of asthma in 9–11 year old children.BMJ310p12251229
- 6. Bisgaard H, Hermansen MN, Buchvald F, Loland L, Halkjaer LB, et al. (2007) Childhood asthma after bacterial colonization of the airway in neonates. N Engl J Med 357: 1487–1495.H. BisgaardMN HermansenF. BuchvaldL. LolandLB Halkjaer2007Childhood asthma after bacterial colonization of the airway in neonates.N Engl J Med35714871495
- 7. Kraft M (2000) The role of bacterial infections in asthma. Clin Chest Med 21: 301–313.M. Kraft2000The role of bacterial infections in asthma.Clin Chest Med21301313
- 8. Sethi S, Evans N, Grant BJ, Murphy TF (2002) New strains of bacteria and exacerbations of chronic obstructive pulmonary disease. N Engl J Med 347: 465–471.S. SethiN. EvansBJ GrantTF Murphy2002New strains of bacteria and exacerbations of chronic obstructive pulmonary disease.N Engl J Med347465471
- 9. Blasi F, Johnston SL (2007) The role of antibiotics in asthma. Int J Antimicrob Agents 29: 485–493.F. BlasiSL Johnston2007The role of antibiotics in asthma.Int J Antimicrob Agents29485493
- 10. Morens DM, Taubenberger JK, Fauci AS (2008) Predominant role of bacterial pneumonia as a cause of death in pandemic influenza: implications for pandemic influenza preparedness. J Infect Dis 198: 962–970.DM MorensJK TaubenbergerAS Fauci2008Predominant role of bacterial pneumonia as a cause of death in pandemic influenza: implications for pandemic influenza preparedness.J Infect Dis198962970
- 11. Hussell T, Wissinger E, Goulding J (2009) Bacterial complications during pandemic influenza infection. Future Microbiol 4: 269–272.T. HussellE. WissingerJ. Goulding2009Bacterial complications during pandemic influenza infection.Future Microbiol4269272
- 12. Sethi S, Murphy TF (2001) Bacterial infection in chronic obstructive pulmonary disease in 2000: a state-of-the-art review. Clin Microbiol Rev 14: 336–363.S. SethiTF Murphy2001Bacterial infection in chronic obstructive pulmonary disease in 2000: a state-of-the-art review.Clin Microbiol Rev14336363
- 13. Staley JT, Konopka A (1985) Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats. Annu Rev Microbiol 39: 321–346.JT StaleyA. Konopka1985Measurement of in situ activities of nonphotosynthetic microorganisms in aquatic and terrestrial habitats.Annu Rev Microbiol39321346
- 14. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, et al. (2007) The human microbiome project. Nature 449: 804–810.PJ TurnbaughRE LeyM. HamadyCM Fraser-LiggettR. Knight2007The human microbiome project.Nature449804810
- 15. Ahmed S, Macfarlane GT, Fite A, McBain AJ, Gilbert P, et al. (2007) Mucosa-associated bacterial diversity in relation to human terminal ileum and colonic biopsy samples. Appl Environ Microbiol 73: 7435–7442.S. AhmedGT MacfarlaneA. FiteAJ McBainP. Gilbert2007Mucosa-associated bacterial diversity in relation to human terminal ileum and colonic biopsy samples.Appl Environ Microbiol7374357442
- 16. Grice EA, Kong HH, Renaud G, Young AC, Bouffard GG, et al. (2008) A diversity profile of the human skin microbiota. Genome Res 18: 1043–1050.EA GriceHH KongG. RenaudAC YoungGG Bouffard2008A diversity profile of the human skin microbiota.Genome Res1810431050
- 17. Dethlefsen L, McFall-Ngai M, Relman DA (2007) An ecological and evolutionary perspective on human-microbe mutualism and disease. Nature 449: 811–818.L. DethlefsenM. McFall-NgaiDA Relman2007An ecological and evolutionary perspective on human-microbe mutualism and disease.Nature449811818
- 18. Pei Z, Bini EJ, Yang L, Zhou M, Francois F, et al. (2004) Bacterial biota in the human distal esophagus. Proc Natl Acad Sci U S A 101: 4250–4255.Z. PeiEJ BiniL. YangM. ZhouF. Francois2004Bacterial biota in the human distal esophagus.Proc Natl Acad Sci U S A10142504255
- 19. Gao Z, Tseng CH, Pei Z, Blaser MJ (2007) Molecular analysis of human forearm superficial skin bacterial biota. Proc Natl Acad Sci U S A 104: 2927–2932.Z. GaoCH TsengZ. PeiMJ Blaser2007Molecular analysis of human forearm superficial skin bacterial biota.Proc Natl Acad Sci U S A10429272932
- 20. Wang M, Ahrne S, Jeppsson B, Molin G (2005) Comparison of bacterial diversity along the human intestinal tract by direct cloning and sequencing of 16S rRNA genes. FEMS Microbiol Ecol 54: 219–231.M. WangS. AhrneB. JeppssonG. Molin2005Comparison of bacterial diversity along the human intestinal tract by direct cloning and sequencing of 16S rRNA genes.FEMS Microbiol Ecol54219231
- 21. Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, et al. (2007) Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proc Natl Acad Sci U S A 104: 13780–13785.DN FrankAL St AmandRA FeldmanEC BoedekerN. Harpaz2007Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases.Proc Natl Acad Sci U S A1041378013785
- 22. Didierlaurent A, Goulding J, Hussell T (2007) The impact of successive infections on the lung microenvironment. Immunology 122: 457–465.A. DidierlaurentJ. GouldingT. Hussell2007The impact of successive infections on the lung microenvironment.Immunology122457465
- 23. Donnelly D, Critchlow A, Everard ML (2007) Outcomes in children treated for persistent bacterial bronchitis. Thorax 62: 80–84.D. DonnellyA. CritchlowML Everard2007Outcomes in children treated for persistent bacterial bronchitis.Thorax628084
- 24. Sutherland ER, Martin RJ (2007) Asthma and atypical bacterial infection. Chest 132: 1962–1966.ER SutherlandRJ Martin2007Asthma and atypical bacterial infection.Chest13219621966
- 25. Aas JA, Paster BJ, Stokes LN, Olsen I, Dewhirst FE (2005) Defining the normal bacterial flora of the oral cavity. J Clin Microbiol 43: 5721–5732.JA AasBJ PasterLN StokesI. OlsenFE Dewhirst2005Defining the normal bacterial flora of the oral cavity.J Clin Microbiol4357215732
- 26. Tunney MM, Field TR, Moriarty TF, Patrick S, Doering G, et al. (2008) Detection of anaerobic bacteria in high numbers in sputum from patients with cystic fibrosis. Am J Respir Crit Care Med 177: 995–1001.MM TunneyTR FieldTF MoriartyS. PatrickG. Doering2008Detection of anaerobic bacteria in high numbers in sputum from patients with cystic fibrosis.Am J Respir Crit Care Med1779951001
- 27. Murray PR, Rosenblatt JE (1976) Bacterial interference by oropharynegeal and clinical isolates of anaerobic bacteria. J Infect Dis 134: 281–285.PR MurrayJE Rosenblatt1976Bacterial interference by oropharynegeal and clinical isolates of anaerobic bacteria.J Infect Dis134281285
- 28. Artis D (2008) Epithelial-cell recognition of commensal bacteria and maintenance of immune homeostasis in the gut. Nat Rev Immunol 8: 411–420.D. Artis2008Epithelial-cell recognition of commensal bacteria and maintenance of immune homeostasis in the gut.Nat Rev Immunol8411420
- 29. Macpherson AJ, Harris NL (2004) Interactions between commensal intestinal bacteria and the immune system. Nat Rev Immunol 4: 478–485.AJ MacphersonNL Harris2004Interactions between commensal intestinal bacteria and the immune system.Nat Rev Immunol4478485
- 30. Henriksson G, Helgeland L, Midtvedt T, Stierna P, Brandtzaeg P (2004) Immune response to Mycoplasma pulmonis in nasal mucosa is modulated by the normal microbiota. Am J Respir Cell Mol Biol 31: 657–662.G. HenrikssonL. HelgelandT. MidtvedtP. StiernaP. Brandtzaeg2004Immune response to Mycoplasma pulmonis in nasal mucosa is modulated by the normal microbiota.Am J Respir Cell Mol Biol31657662
- 31. Payne D, McKenzie SA, Stacey S, Misra D, Haxby E, et al. (2001) Safety and ethics of bronchoscopy and endobronchial biopsy in difficult asthma. Arch Dis Child 84: 423–426.D. PayneSA McKenzieS. StaceyD. MisraE. Haxby2001Safety and ethics of bronchoscopy and endobronchial biopsy in difficult asthma.Arch Dis Child84423426
- 32. Green SJ, Michel FC Jr, Hadar Y, Minz D (2004) Similarity of bacterial communities in sawdust- and straw-amended cow manure composts. FEMS Microbiol Lett 233: 115–123.SJ GreenFC Michel JrY. HadarD. Minz2004Similarity of bacterial communities in sawdust- and straw-amended cow manure composts.FEMS Microbiol Lett233115123
- 33. DeSantis TZ Jr, Hugenholtz P, Keller K, Brodie EL, Larsen N, et al. (2006) NAST: a multiple sequence alignment server for comparative analysis of 16S rRNA genes. Nucleic Acids Res 34: W394–399.TZ DeSantis JrP. HugenholtzK. KellerEL BrodieN. Larsen2006NAST: a multiple sequence alignment server for comparative analysis of 16S rRNA genes.Nucleic Acids Res34W394399
- 34. Huber T, Faulkner G, Hugenholtz P (2004) Bellerophon: a program to detect chimeric sequences in multiple sequence alignments. Bioinformatics 20: 2317–2319.T. HuberG. FaulknerP. Hugenholtz2004Bellerophon: a program to detect chimeric sequences in multiple sequence alignments.Bioinformatics2023172319
- 35. Schloss PD, Handelsman J (2005) Introducing DOTUR, a computer program for defining operational taxonomic units and estimating species richness. Appl Environ Microbiol 71: 1501–1506.PD SchlossJ. Handelsman2005Introducing DOTUR, a computer program for defining operational taxonomic units and estimating species richness.Appl Environ Microbiol7115011506
- 36. Kittles RA, Long JC, Bergen AW, Eggert M, Virkkunen M, et al. (1999) Cladistic association analysis of Y chromosome effects on alcohol dependence and related personality traits. Proc Natl Acad Sci U S A 96: 4204–4209.RA KittlesJC LongAW BergenM. EggertM. Virkkunen1999Cladistic association analysis of Y chromosome effects on alcohol dependence and related personality traits.Proc Natl Acad Sci U S A9642044209
- 37. Durrant C, Zondervan KT, Cardon LR, Hunt S, Deloukas P, et al. (2004) Linkage disequilibrium mapping via cladistic analysis of single-nucleotide polymorphism haplotypes. Am J Hum Genet 75: 35–43.C. DurrantKT ZondervanLR CardonS. HuntP. Deloukas2004Linkage disequilibrium mapping via cladistic analysis of single-nucleotide polymorphism haplotypes.Am J Hum Genet753543
- 38. Lozupone C, Hamady M, Knight R (2006) UniFrac–an online tool for comparing microbial community diversity in a phylogenetic context. BMC Bioinformatics 7: 371.C. LozuponeM. HamadyR. Knight2006UniFrac–an online tool for comparing microbial community diversity in a phylogenetic context.BMC Bioinformatics7371
- 39. Paster BJ, Boches SK, Galvin JL, Ericson RE, Lau CN, et al. (2001) Bacterial diversity in human subgingival plaque. J Bacteriol 183: 3770–3783.BJ PasterSK BochesJL GalvinRE EricsonCN Lau2001Bacterial diversity in human subgingival plaque.J Bacteriol18337703783
- 40. Smith NH, Holmes EC, Donovan GM, Carpenter GA, Spratt BG (1999) Networks and groups within the genus Neisseria: analysis of argF, recA, rho, and 16S rRNA sequences from human Neisseria species. Mol Biol Evol 16: 773–783.NH SmithEC HolmesGM DonovanGA CarpenterBG Spratt1999Networks and groups within the genus Neisseria: analysis of argF, recA, rho, and 16S rRNA sequences from human Neisseria species.Mol Biol Evol16773783