Variability and diversity of nasopharyngeal microbiota in children: a metagenomic analysis.
ABSTRACT The nasopharynx is the ecological niche for many commensal bacteria and for potential respiratory or invasive pathogens like Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis. Disturbance of a balanced nasopharyngeal (NP) microbiome might be involved in the onset of symptomatic infections with these pathogens, which occurs primarily in fall and winter. It is unknown whether seasonal infection patterns are associated with concomitant changes in NP microbiota. As young children are generally prone to respiratory and invasive infections, we characterized the NP microbiota of 96 healthy children by barcoded pyrosequencing of the V5-V6 hypervariable region of the 16S-rRNA gene, and compared microbiota composition between children sampled in winter/fall with children sampled in spring. The approximately 1,000,000 sequences generated represented 13 taxonomic phyla and approximately 250 species-level phyla types (OTUs). The 5 most predominant phyla were Proteobacteria (64%), Firmicutes (21%), Bacteroidetes (11%), Actinobacteria (3%) and Fusobacteria (1,4%) with Moraxella, Haemophilus, Streptococcus, Flavobacteria, Dolosigranulum, Corynebacterium and Neisseria as predominant genera. The inter-individual variability was that high that on OTU level a core microbiome could not be defined. Microbiota profiles varied strongly with season, with in fall/winter a predominance of Proteobacteria (relative abundance (% of all sequences): 75% versus 51% in spring) and Fusobacteria (absolute abundance (% of children): 14% versus 2% in spring), and in spring a predominance of Bacteroidetes (relative abundance: 19% versus 3% in fall/winter, absolute abundance: 91% versus 54% in fall/winter), and Firmicutes. The latter increase is mainly due to (Brevi)bacillus and Lactobacillus species (absolute abundance: 96% versus 10% in fall/winter) which are like Bacteroidetes species generally related to healthy ecosystems. The observed seasonal effects could not be attributed to recent antibiotics or viral co-infection.The NP microbiota of young children is highly diverse and appears different between seasons. These differences seem independent of antibiotic use or viral co-infection.
- SourceAvailable from: Remco Kort[Show abstract] [Hide abstract]
ABSTRACT: The variation of microbial communities associated with the human body can be the cause of many factors, including the human genetic makeup, diet, age, surroundings, and sexual behavior. In this study, we investigated the effects of intimate kissing on the oral microbiota of 21 couples by self-administered questionnaires about their past kissing behavior and by the evaluation of tongue and salivary microbiota samples in a controlled kissing experiment. In addition, we quantified the number of bacteria exchanged during intimate kissing by the use of marker bacteria introduced through the intake of a probiotic yoghurt drink by one of the partners prior to a second intimate kiss. Similarity indices of microbial communities show that average partners have a more similar oral microbiota composition compared to unrelated individuals, with by far most pronounced similarity for communities associated with the tongue surface. An intimate kiss did not lead to a significant additional increase of the average similarity of the oral microbiota between partners. However, clear correlations were observed between the similarity indices of the salivary microbiota of couples and self-reported kiss frequencies, and the reported time passed after the latest kiss. In control experiments for bacterial transfer, we identified the probiotic Lactobacillus and Bifidobacterium marker bacteria in most kiss receivers, corresponding to an average total bacterial transfer of 80 million bacteria per intimate kiss of 10 s. This study indicates that a shared salivary microbiota requires a frequent and recent bacterial exchange and is therefore most pronounced in couples with relatively high intimate kiss frequencies. The microbiota on the dorsal surface of the tongue is more similar among partners than unrelated individuals, but its similarity does not clearly correlate to kissing behavior, suggesting an important role for specific selection mechanisms resulting from a shared lifestyle, environment, or genetic factors from the host. Furthermore, our findings imply that some of the collective bacteria among partners are only transiently present, while others have found a true niche on the tongue's surface allowing long-term colonization.Microbiome. 01/2014; 2:41.
- [Show abstract] [Hide abstract]
ABSTRACT: The bacterial communities of the nasopharynx play an important role in upper respiratory tract infections (URTIs). Our study represents the first survey of the nasopharynx during a known, controlled viral challenge. We aimed to gain a better understanding of the composition and dynamics of the nasopharyngeal microbiome during viral infection.Microbiome. 01/2014; 2:22.
- [Show abstract] [Hide abstract]
ABSTRACT: The use of sequencing technologies to investigate the microbiome of a sample can positively impact patient healthcare by providing therapeutic targets for personalized disease treatment. However, these samples contain genomic sequences from various sources that complicate the identification of pathogens.BMC Bioinformatics 08/2014; 15(1):262. · 2.67 Impact Factor
Variability and Diversity of Nasopharyngeal Microbiota
in Children: A Metagenomic Analysis
Debby Bogaert1*, Bart Keijser2, Susan Huse3, John Rossen4, Reinier Veenhoven5, Elske van Gils1, Jacob
Bruin6, Roy Montijn2, Marc Bonten7,8, Elisabeth Sanders1
1Department of Paediatric Infectious Diseases and Immunology, University Medical Center Utrecht-Wilhelmina Children’s Hospital, Utrecht, The Netherlands, 2Business
Unit Food and Biotechnology Innovations, Microbial Genomics Group, TNO Quality of Life, Zeist, The Netherlands, 3Marine Biological Laboratory, Josephine Bay Paul
Center for Comparative Molecular Biology and Evolution, Woods Hole, Massachusetts, United States of America, 4Laboratory of Medical Microbiology and Immunology,
St. Elisabeth Hospital, Tilburg, The Netherlands, 5Linnaeus Institute, Spaarne Hospital, Hoofddorp, The Netherlands, 6Regional Laboratory of Public Health, Haarlem, The
Netherlands, 7Department of Medical Microbiology, University Medical Center Utrecht, Utrecht, The Netherlands, 8Julius Center of Health Sciences and Primary Care,
University Medical Center Utrecht, Utrecht, The Netherlands
The nasopharynx is the ecological niche for many commensal bacteria and for potential respiratory or invasive pathogens
like Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis. Disturbance of a balanced
nasopharyngeal (NP) microbiome might be involved in the onset of symptomatic infections with these pathogens, which
occurs primarily in fall and winter. It is unknown whether seasonal infection patterns are associated with concomitant
changes in NP microbiota. As young children are generally prone to respiratory and invasive infections, we characterized the
NP microbiota of 96 healthy children by barcoded pyrosequencing of the V5–V6 hypervariable region of the 16S-rRNA gene,
and compared microbiota composition between children sampled in winter/fall with children sampled in spring. The
approximately 1000000 sequences generated represented 13 taxonomic phyla and approximately 250 species-level phyla
types (OTUs). The 5 most predominant phyla were Proteobacteria (64%), Firmicutes (21%), Bacteroidetes (11%),
Actinobacteria (3%) and Fusobacteria (1,4%) with Moraxella, Haemophilus, Streptococcus, Flavobacteria, Dolosigranulum,
Corynebacterium and Neisseria as predominant genera. The inter-individual variability was that high that on OTU level a
core microbiome could not be defined. Microbiota profiles varied strongly with season, with in fall/winter a predominance
of Proteobacteria (relative abundance (% of all sequences): 75% versus 51% in spring) and Fusobacteria (absolute
abundance (% of children): 14% versus 2% in spring), and in spring a predominance of Bacteroidetes (relative abundance:
19% versus 3% in fall/winter, absolute abundance: 91% versus 54% in fall/winter), and Firmicutes. The latter increase is
mainly due to (Brevi)bacillus and Lactobacillus species (absolute abundance: 96% versus 10% in fall/winter) which are like
Bacteroidetes species generally related to healthy ecosystems. The observed seasonal effects could not be attributed to
recent antibiotics or viral co-infection. The NP microbiota of young children is highly diverse and appears different
between seasons. These differences seem independent of antibiotic use or viral co-infection.
Citation: Bogaert D, Keijser B, Huse S, Rossen J, Veenhoven R, et al. (2011) Variability and Diversity of Nasopharyngeal Microbiota in Children: A Metagenomic
Analysis. PLoS ONE 6(2): e17035. doi:10.1371/journal.pone.0017035
Editor: Malcolm Semple, University of Liverpool, United Kingdom
Received October 27, 2010; Accepted January 11, 2011; Published February 28, 2011
Copyright: ? 2011 Bogaert 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: This work was supported by grants from The Netherlands Organisation for Scientific Research through NWO-VENI grant 91610121 (D.B.) and NWO-VICI
grant 91876611 (M.J.M.B.). These funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Authors
B.K. and R.M. are employed by TNO Quality of Life and had a role in data collection and analysis.
Competing Interests: Dr. Veenhoven reports receiving grant support from GlaxoSmith Kline and Wyeth for vaccine studies and consulting fees for
GlaxoSmithKline. Dr. Sanders reports receiving unrestricted grants from Wyeth and Baxter for research, consulting fees for Wyeth and GlaxoSmithKline, lecturing
fees from Wyeth and grant support from Wyeth and GlaxoSmithKline for vaccine studies. These grants were not received for the research described in this paper.
This does not alter the authors’ adherence to all the PLoS ONE policies on sharing data and materials. Authors employed by TNO (B.K., R.M.) have a potential
conflict of interest as their organizations may benefit from a product or patent generated on the basis of the published data. In these cases, the authors will
however not receive additional salary, additional personal income, or any form of financial support. In a dition, it does not alter the authors’ adherence to all the
PLoS ONE policies on sharing data and materials. For all other authors no potential conflicts reported.
* E-mail: firstname.lastname@example.org
According to the WHO, respiratory tract infections are still
among the leading causes of death in children and adults
worldwide . The most common pathogens like Streptococcus
pneumoniae, Haemophilus influenzae, Neisseria meningitidis and Staphylo-
coccus aureus are normal and transient residents of the nasopha-
ryngeal (NP) niche, where they are embedded in a complex
microbiota of generally presumed harmless commensals. The
human microbiome in general is assumed beneficial to the host
due to stimulation and maturation of immune systems, promotion
of mucosal structure and function and providing actual ‘coloni-
zation resistance’ against pathogen invasion . Although
colonization by the ‘‘potential pathogens’’ of the NP microbiome
is mainly asymptomatic, progression towards upper respiratory
tract infections, pneumonia or even sepsis and meningitis may
occur [3,4]. The exact mechanisms by which this occurs remain
largely unknown, although an imbalance in the composition of
microbiota, for example by acquisition of new pathogens, viral co-
infection or other host or environmental factors have been
suggested [5–8]. In addition, clear correlations between invasive
attack rates and season are observed for many of the potential
PLoS ONE | www.plosone.org1 February 2011 | Volume 6 | Issue 2 | e17035
pathogens of the upper respiratory tract [9,10], a phenomenon
that cannot be fully explained by concomitant changes in
colonization rates of the individual pathogenic bacteria [11,12].
This suggests that local containment of the colonizing pathogenic
bacteria by the host and/or the surrounding ecosystem is of major
importance in prevention of disease progression. Despite an
abundance of data on incidence, prevalence and density of
potential pathogens in NP microbiota of children and adults, the
detailed composition of the NP microbial community, both during
health and disease have not been studied. We, therefore,
performed a meta-genomic study on the detailed composition of
and variability in NP microbiota in young children sampled during
Results and Discussion
We studied the NP microbiota composition of 96 healthy 18-
months old children. Their characteristics are depicted in Table
S1. Being aware of the current discussions on the artefacts that
may be introduced by pyrosequencing [13,14], we applied a
stringent protocol for filtering and clustering of sequences. The
approx. 1 100 000 generated sequences (on average 11000
sequences per sample) yielded about 92 000 unique sequences,
representing 13 taxonomic phyla and 243 species-level phyla types
(OTUs). The data were normalized for equal numbers of reads per
sample. The 5 most predominant phyla were Proteobacteria
(64%), Firmicutes (21%), Bacteroidetes (11%), Actinobacteria (3%)
and Fusobacteria (1.4%) (Figure 1). In addition, we found
representatives of Cyanobacteria, probably reflecting plant
chloroplasts obtained through inhalation. Sporadically and/or in
low abundance we found sequences for the candidate divisions
OD1, TM7 and BRC1 and the phyla Deinococcus-Thermus,
Nitrospira, Planctomycetes and Chloroflexi. On a lower taxo-
nomic level, the most prevalent genera were Moraxella (40%),
Haemophilus (20%), Streptococcus (12%), and Flavobacterium
(10%). Other fairly common genera were Dolosigranulum (5%),
Corynebacterium (2%), Neisseria (2%) and Fusobacterium (1%).
The 30 most common OTUs representing almost 98% of all
reads, and their relative and absolute presence are shown in
Table 1 (For the complete list of OTUs; see Table S2). Although
the top 6 predominant phyla are identical to those of neighbouring
microbiota, the composition, i.e. relative contribution of each
phyla to those microbiota seems fairly different. In the oral cavity,
microbiota are dominated by Firmicutes followed by Proteobac-
teria and Bacteroidetes (overall 50% Gram-positive bacteria),
whereas the microbiome of the nostril contains more than 80%
gram-positive bacteria, mostly Actinobacteria and Firmicutes .
These data, therefore, suggest different dynamics (i.e., different
biological equilibria) in the NP microbiome.
There was a high inter-individual variability in the composition
of the microbiota up to phyla level, and in the relative abundance
of the individual bacterial inhabitants (Table 1). This resulted in a
limited core microbiome (as representing .0.1% of sequenes and
being present in all 96 children) of specific phyla only, namely
Proteobacteria and Firmicutes, however no single OTU was found
universally. Because of the observed high inter-individual
variation, we applied a less strict definition of core microbiome,
i.e. OTUs present in more than 50% of all samples and
representing .0.1% of the sequences. With this definition we
observed a core microbiome of Moraxella, Haemophilus influenzae,
Enhydrobacter (Proteobacteria), Streptococcus, Dolosigranulum
(Firmicutes), and Corynebacterium (Actinobacteria) (Table 1).
Principal component analysis identified 3 distinct clusters of
microbiota profiles correlating strongly with a predominance
(.50% of sequences per sample) of single OTU’s, i.e. Moraxella
(OTU 1), Haemophilus influenzae (OTU 2), and Streptococci (OTU
3), respectively, connected by a group of community profiles
representing mixed microbiota (Figure 2). Additionally, we
observed transition zones for microbiota profiles between the
Haemophilus- and Moraxella-dominated clusters, but not between
Haemophilus- or Moraxella-dominated clusters and the Strepto-
cocci-dominated cluster, which might implicate potential interac-
tions between microbiota profiles.
Since respiratory and invasive infections are associated with
fall/winter season, we analysed our samples concordantly. When
distinguished by the time of sampling (fall/winter versus spring)
groups of children did not differ significantly in demographics or
life style characteristics, infectious symptoms, medical history or
environmental parameters (Table S1), reducing the likelihood of
internal confounders as a cause of potential seasonal correlations
with microbiota profiles. However, with respect to microbiota
profiles, we observed marked differences between samples
obtained in fall/winter versus samples from spring (Table 2). In
samples obtained in late fall and winter, we observed a
predominance of Proteobacteria (relative abundance (% of all
sequences): 75% versus 51% in spring), Fusobacteria (absolute
abundance (% of children): 14% versus 2% in spring), and
Cyanobacteria (absolute abundance: 64% versus 30% in spring;
relative abundance: 0.27% versus 0.09% in spring) were
significantly more abundant compared to spring, whereas
Bacteroidetes were more frequently present in samples obtained
in spring (relative abundance: 19% versus 3% in fall/winter,
absolute abundance: 91% versus 54% in fall/winter) (Figure 3a).
On OTU level we observed amongst others more Bacillus,
Brevibacillus and Lactobacillus species, and Flavobacterium and
B. fragilis (both Bacteroidetes) in samples from spring compared to
fall/winter. In addition we found less a-Proteobacteria, Oxalo-
bacteriaceae, Microbacteriaceae, Ralstonia, Pseudomonas and
Acidovorax (all Proteobacteria), Cyanobacteria, and Porphyromonas
catoniae (Bacteroidetes) in samples from spring compared to fall/
winter (Figure 3b). When re-evaluating the core microbiome per
individual season (i.e. OTUs present in more than 50% of samples
of a certain season), we observed an additional core of
Proprionibacterium and Cyanobacteria for fall/winter and an
additional core of Flavobacteria, Brevibacillus and Bacillus (almost
exclusively) for spring. The latter groups of bacteria, i.e.
Bacteroidetes and (Brevi)bacillus and Lactobacillus species, are
generally related to protection against overgrowth of pathogenic
species due to the production of bacteriocins and other inhibitory
substances [7,16]. In other microbiota like the gastrointestinal and
vaginal tract they are highly related to maintenance of a balanced
microbiome as well [17–19]. Since infections with respiratory
pathogens, especially pneumonia, are strongly related to fall and
winter season [9,10], the presence and abundance of these
bacteria in respiratory microbiota in spring might therefore
suggest in general a more balanced respiratory microbiome in
this specific season as well protecting against onset of respiratory or
We tested all samples for the presence of respiratory viruses by
q-PCR methods and detected one or more viruses in 67% of
samples (Table S3). We found no evidence for associations
between the observed seasonal shift in microbiota and the overall
presence of respiratory viruses, nor for any of the individual viruses
when tested by SAM analysis (Figure 3b). Although these data do
not exclude an effect of respiratory viruses on microbiota
composition, they do suggest other season-related factors like
environmental factors (temperature, humidity, smoke exposure,
crowding), or nutrient- or vitamine-related effects, or a combina-
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org2 February 2011 | Volume 6 | Issue 2 | e17035
tion of factors might be important for the observed shifts in
microbiota profiles [3,16,20]. Interestingly, day-care attendance or
smoke exposure could not be related to the observed shifts in
microbiota, although the latter was encountered very rarely. This
further underlines that different or more complex effect may be
responsible for the observed phenomenon. In addition, no
association was observed between seasonal changes in microbiota
and recent antibiotic use, which was rather limited in this
population (Table 1, Figure 3b). Because of the explicit correlation
between season and microbiota profiles, correlations between
other environmental determinants and microbiota could not be
accurately tested, which underlines that in future studies one needs
to control and power for seasonal effects.
As previously mentioned, the majority of the children (87%) had
a predominant Gram-negative NP community profile (.50%
Gram-negative bacteria). This is probably due to predominance of
Gram-negative Moraxella and Haemophilus species known to
reside specifically at this body site [3,4]. On average, 76% of the
overall NP microbiome in children was composed of Gram-
negative bacteria, however with a wide range of 9–99% Gram-
negative bacteria per sample. In addition, there was a higher
contribution of Gram-negative bacteria to microbiota obtained in
fall/winter (81%) compared to spring (72%) (Independent samples
t-test: p=0.044). This could potentially explain some of the
observed differences between gram-negative ratios at the NP
microbiome and at other human microbiota, where seasonal
changes in composition have so far not been studied.
Finally, we studied the inter-individual diversity in NP
microbiota overall and in relation to season. We observed highly
diverse microbiota, with on average 40 OTUs per sample, and a
high inter-individual diversity with 20–87 OTUs per individual.
With respect to season, there was no significant difference in
diversity between fall/winter (average: 38 OTUs, range: 20–77)
and spring (average: 43 OTUs, range: 20–87) (independent
samples T-test: p=0.083).
As internal control, we compared conventional culturing results
for the potential pathogens S. pneumoniae (71% positive), H.
influenzae (69% positive) and M. catarrhalis (88% positive) with
sequencing results for the OTUs Streptococcus (OTU 3, OTU 23,
OTU 31), H. influenzae (OTU 2, OTU 10), and Moraxella (OTU
1, OTU 6) respectively (Figure S1) and found strong correlations
between S. pneumoniae and Streptococcus OTU 3 and 31
(p,0.0001) but not OTU 23, which is probably another
Streptococcus species, and H. influenzae and H. influenzae OTU 2
and OTU 10 (p,0.0001). For M. catarrhalis, we were only able to
find a positive correlation between M. catarrhalis and Moraxella
OTU 6 with independent samples t-test (p=0.003) but not with
Spearman’s Correlation, which may be explained by the low
number of Moraxella negative individuals, making a comparison
between binary and quantitative data difficult. Also the presence of
other Moraxella species with high sequence homology might
interfere with a strong correlation between these results.
In conclusion, to our knowledge this is the first report describing
in detail the composition of and the variability within the human
NP microbiota assessed at the depth of next generation
sequencing. In line with other human body habitats, we found a
complex, diverse and highly variable microbiota with a relatively
limited core microbiome. There is considerable seasonal variation
in NP microbiota. This implies the time of sampling should be
considered when describing or comparing NP microbiomes, and
preferably controlled for when other potential determinants like
the impact of viruses or antibiotics on microbiota profiles or the
correlation between microbiota profiles and diseases will be tested.
Whether these seasonal changes in composition of the NP
microbome are causally related to seasonal occurrence of
respiratory tract infections remains to be determined, though
seems relevant for further understanding of pathogenesis of
infectious diseases and in the long run potentially for understand-
ing of effects of current and design of future preventive measures.
We randomly selected 150 NP samples from a cohort of 330
healthy children 18 months of age who had participated in a
randomised controlled trial studying the effect of reduced-dose
schedules of 7-valent pneumococcal conjugate vaccine (PCV-7)
performed in a general community in the Western part of The
Netherlands where the control children received PCV-7 only after
the trial was finished at the age of 24 months . An
Figure 1. Relative abundance of all bacterial phyla found in the NP microbiota of 96 infants 18 months of age. A cut-off of 0.1% is used
for visual differentiation between predominant and less dominant phyla.
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org3 February 2011 | Volume 6 | Issue 2 | e17035
Table 1. Thirty most common OTUs or ‘species-level’ phylotypes (ranked by predominance, i.e. absolute presence among the approx. 1 100 000 reads).
(% of reads)
(. .0.1% of
Nr. of samples (of total of 96 samples) containing each OTU in .0% or .0.1% of the reads is stated. Core microbiome: OTUs found in .50% of the samples in .0.1% of reads per sample (All: OTU found in .50% of samples; Spring
and Fall/Winter: OTU found in .50% of samples obtained in spring or fall/winter, respectively). NA: not assigned.
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org4 February 2011 | Volume 6 | Issue 2 | e17035
acknowledged national ethics committee from the Netherlands
(Stichting Therapeutische Evaluatie Geneesmiddelen, http://
www.stegmetc.org) approved the study protocol. The trial was
undertaken in accordance with the European Statements for Good
Clinical Practice, which includes the provisions of the Declaration
of Helsinki of 1989. Written informed consent was provided by the
parents or their legal guardians.
Nasopharyngeal swabs (Transwab Pernasal Plain (Catalogue
MW173P), Medical Wire & Equipment Co, Ltd, Corsham,
Wiltshire, England) were collected between November 2007 and
June 2008 during home visits after written informed consent was
provided by study participants and/or their legal guardians. The
swabs were obtained by approaching the nasopharynx transna-
sally, transported to the laboratory in Transwab (modified Amies)
medium (room temperature) and plated within 24 hours on
selective agar media. After plating, the cotton swabs were
consecutively rinsed in 1 ml of saline and stored at 280uC until
The nasopharyngeal swabs were plated onto a 5% sheep blood
agar plate, a 5% sheep blood agar plate with 5 mg/L gentamicin,
a chocolate agar plate and a Haemophilus chocolate agar plate.
Agar plates were incubated at 35uC for 48 h; the blood agar plate
aerobically, the blood agar plate with gentamicin and the
chocolate agar plates with raised CO2. Identification of S.
pneumoniae, H. influenzae, M. catarrhalis and S. aureus was based on
colony morphology and conventional methods of determination.
Bacterial DNA isolation
One 200 ml aliquot of swab ‘‘rinse’’solution wasdistributed intwo
separate sterile screw-cap Eppendorf tubes, each containing 0.25 ml
lysis buffer (AGOWA mag Mini DNA Isolation Kit, catalgue 40410,
AGOWA, Berlin, Germany). Then 0.3 g zirconium beads (diame-
ter, 0.1 mm, catalogue 11079101z, Biospec Products, Bartlesville,
OK 74005. USA) and 0.2 ml phenol (Phenol solution BioUltra, TE-
saturated, catalogue P4557, Sigma-Aldrich, St. Louis, MO, USA)
Figure 2. Principal component analysis of the individual NP communities. We observed three individual clusters/axes surrounding a centre
of profiles, with in green individual profiles depicting predominantly (.50% of sequences) Moraxella OTU 1, in blue microbiota profiles depicting
predominantly H. influenzae OTU 2, and in red microbiota profiles depicting predominantly Streptococcus OTU 3. Mixed phyla profiles (no single OTU
is representing .50% of sequences) cluster in the centre of this PCA plot.
Table 2. Relative abundance of individual phyla is depicted per season.
Proteobacteria75.39 23.811.91 98.4150 51.13 23.778.0998.89 46
Firmicutes 17.2216.58 0.30 67.6450 24.7625.220.7891.18460.097
Bacteroidetes3.398.99 0.00 43.4527 19.31 16.620.0055.4242
Actinobacteria1.883.36 0.0522.69 493.675.910.02 29.4042NS
Fusobacteria1.75 8.140.0048.007 0.986.620.00 44.881 NS
Cyanobacteria 0.27 0.440.002.54 320.090.130.000.74140.029
OD10.0670.0860.0000.38 120.0460.0540.0000.199 NS
TM70.0450.3050.0002.1510.023 0.151 0.0001.021 NS
Deinococcus-Thermus0.0150.0300.0000.143 0.0080.0140.0000.060 NS
Nitrospira0.0050.0120.000 0.0500.0010.0050.0000.020 0.103
Planctomycetes 0.0010.008 0.0000.060 0.0020.0130.0000.080 NS
Chloroflexi 0.001 0.005 0.0000.0300.0010.0050.0000.040 NS
Mean, SD, and Range of each phyla per group of samples are depicted for the samples obtained in fall/winter (n=50) versus samples obtained in spring (n=46).
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org5 February 2011 | Volume 6 | Issue 2 | e17035
were added to each sample. The samples were homogenized with a
Mini-beadbeater (Mini-beadbeater 16, catalogue 607EUR, Biospec
Products, Bartlesville, OK 74005. USA) for 2 min. The released
DNA was purified with the AGOWA mag Mini DNA Isolation Kit
according to the manufacturer’s recommendations. To maximize
recovery, the DNA binding step was performed twice for each
sample. Following, the DNA for each sample was eluted in a total
volume of 40 ml milliQ. The integrity of the DNA was inspected by
agar gel electrophoresis. DNA was quantified on the NanoDrop
spectrophotometer (Thermo Scientific NanoDrop 1000 Spectro-
photometer, Thermo Scienific, Wilmington, DE 19810 USA).
Real time PCR for bacterial DNA
The total bacterial load of the samples was established by
quantitative PCR. The primer-probe set targeting the bacterial 16S
rDNA gene comprised of forward primer 16S-F1 (59-CGA AAG
CGT GGG GAG CAA A -39), reverse primer 16S-R1 (59-GTT
CGT ACT CCC CAG GCG G-39) and probe 16S-P1 (FAM- ATT
AGA TAC CCT GGT AGT CCA –MGB). The PCR mixture
consisted of 15 ml of 2x master mix (Universal Mastermix,catalogue
GMO-UN-A100, Europe Diagenode sa, Lie `ge, Belgium), 1 ml of
each primer (10 mM), 1 ml of the probe (5 mM), 9.5 ml DNA free
water and 2.5 ml of template DNA. Amplifications were performed
using a 7500 Fast Real-Time PCR System (Applied Biosystems,
catalogue 4351107, Foster City, CA 94404 USA) under the
following conditions: 2 min at 50uC and 10 min at 95uC, followed
by 45 cycles of 15 s at 95uC and 1 min at 60uC.
To generate the PCR amplicon libraries, the small subunit
ribosomal RNA gene V5–V6 hypervariable region was amplified
for each individual sample independently. Of the 150 samples
tested, 96 samples contained more than 1,3*10‘3 fg/ml DNA
and were included in sequence analysis. PCR was performed
using the forward primer 785F (59-GGA TTA GAT ACC CBR
GTA GTC-39) and the reverse primer 1061R (59-TCA CGR
CAC GAG CTG ACG AC-39). The primers were fitted with the
454 Life Sciences Adapter A (forward primer) and B (reverse
primer), fused to the 59 end of the 16S rDNA bacterial primer
sequences. The reverse primer also included a unique tetranu-
cleotide sample identification key. The amplification mix
contained 2 units of Pfu Ultra II Fusion HS DNA polymerase
(Stratagene, La Jolla, CA, USA) and 1x PfuUltra II reaction
buffer (Stratagene), 200 mM dNTP PurePeak DNA polymerase
Mix(Pierce Nucleic Acid Technologies, catalogue NU606001
Milwaukee, WI, USA), and 0.2 mM of each primer. After
denaturation (94uC; 2 min), 30 cycles were performed that
consisted of denaturation (94uC; 30 sec), annealing (50uC;
40 sec), and extension (72uC; 80 sec). DNA was isolated by
means of the MinElute kit (Qiagen, catalogue 28006, Hilden,
Germany). The quality and the size of the amplicons were
analyzed on the Agilent 2100 Bioanalyser with the DNA 1000
Chip kit (Agilent Technologies, catalogue 5067–1504, Santa
Clara, CA, USA) and quantified using Nanodrop ND-1000
spectrophotometer. The amplicons of the individual samples
were pooled in equimolar amounts in four libraries. The four
libraries were sequenced unidirectionally in the reverse direction
(B-adaptor) during two 454 Genome Sequencer FLX (GS-FLX,
454 Life Sciences (Roche), Branford, CT 06405 USA) runs.
Sequences are available at the Short Read Archive of the
National Center for Biotechnology Information (NCBI) [NCBI
Figure 3. Seasonal differences between microbiota profiles of 50 children sampled in fall-winter and 46 children in spring. The
samples are marked by the week number they were obtained (week 48 until week 23).Infigure 3a thephyla showingsignificant association withseason of
sampling by SAM analysis are depicted. In figure 3b the OTUsshowing significant association with season are depicted. The samples are marked by season
(blue; fall, green;winter, red; spring),antibiotic use (,1 month: green), presenceof viruses (positive: green), presenceof multiple viruses (green: 1 virus, red:
are depicted with separate colours; Yellow: Firmicutes, Orange: Proteobacteria, Green: Bacteroidetes, Blue: Actinobacteria, Pink: Cyanobacteria.
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org6February 2011 | Volume 6 | Issue 2 | e17035
Real-time PCR for viruses
One 200 ml aliquot of swab ‘‘rinse’’ solution was used to extract
viral nucleic acids using the MagNA Pure LC total nucleic acid
isolation kit (Roche Diagnostics, catalogue 03 038 505 001, Basel,
Switzerland) as described previously . Detection of viral
pathogens was performed in parallel, using real-time PCR assays
for bocavirus (HBoV), polyomaviruses (WUPyV and KIPyV),
respiratory syncytial virus (RSV) A and B, influenzavirus (IV) A
and B, para-influenzavirus (PIV) 1–4, human rhinoviruses (HRV),
adenoviruses, human coronavirus OC43, NL63, HKU and 229E,
and human metapneumovirus (hMPV). Real-time PCR proce-
dures were performed as described previously  Briefly, samples
were assayed in duplicate in a 25 ml reaction mixture containing
10 ml (c)DNA, 12.5 ml 2 6TaqMan Universal PCR Master Mix
(Applied Biosystems, catalogue 4304437, Foster City, CA 94404
USA), 300–900 nmol/l of the forward and reverse primers and
75–200 nmol/l of each of the probes. All samples had been spiked
before extraction with an internal control virus (phocine distemper
virus [RNA virus] and phocine herpes virus [DNA virus]) to
monitor for efficient extraction and amplification.
GS-FLX sequencing data were processed as previously described
. In brief, we trimmed data by removing primer sequences and
low-quality data, sequences that did not have an exact match to the
or that were shorter than 50 nt after trimming. We then used the
GAST algorithm  to calculate the percent difference between
each unique sequence and its closest match in a database of 69816
unique Eubacterial and 2779 unique Archaeal V5–V6 sequences,
representing 323499 SSU rRNA sequences from the SILVA
database . Taxa were assigned to each full-length reference
sequence using several sources including Entrez Genome entries,
cultured strain identities, SILVA, and the Ribosomal Database
Project Classifier . In cases where reads were equidistant to
multiple V5–V6 reference sequences, and/or where identical V5–
V6 sequences were derived from longer sequences mapping to
least two-thirds of the sequences. The operational taxonomic units
(OTUs) were created by aligning unique sequences and calculating
distance matrices as previously described  and using DOTUR
 to create clusters at the 3% level.
Only those sequences that were found at least 5 times were
included in the analyses. This strict and conservative approach was
chosen to preclude inclusion of sequences from potential
contamination or sequencing artefacts. To compare the relative
abundance of OTUs among samples, the data were normalized
for number of sequenced reads obtained for each sample. To
reduce the influence of abundant taxa on principal component
analyses, the normalized abundance data were log2 transformed.
Unsupervised data analysis, Principle Component Analysis, and
hierarchical clustering was performed using MeV software
package as part of TM4 microarray software suite .
Seasonal differences in phyla distribution were studied by using
Mann Whitney U test (SPSS software Version 15.0). Seasonal
differences in OTU patterns as well as potential correlations
between respiratory viruses and OTU patterns were studied by the
Significant Analysis of Microarrays (SAM analysis) - a non-
parametric statistical technique for finding significant differences
between microarray data of groups based on experimental
conditions , implemented in the MeV software package
. To determine significant differences between microbiota
profiles, we used Pearson’s correlation with average linkage
clustering method and a FDR significance criterion of ,0.05.
Independent samples T-test, and Spearman correlation coeffi-
cients(SPSS softwareVersion 15.0) wereused for testing correlations
between conventional cultures and pyrosequencing data. Indepen-
dent samples t-test was also used to compare the contribution of
gram negative and positive bacteria to the microbiota in different
seasons and to test for differences in diversity between seasons.
culture results for S. pneumoniae, H. influenzae, and M. catarrhalis on
the x-axis (absent/present) and sequencing results (as % of total
microbiota profile) for Streptococcus (OTU 3), H. influenzae (OTU
2 plus 10), and Moraxella (OTU 6) on the y-axis.
Graphs showing correlation between conventional
subdivided per season.
Population characteristics for all samples and
study samples. Here all 243 taxa (species or more inclusive taxa
when sequences could not be confidently classified to species level)
and their absolute presence per child (yes/no) and relative
abundance (% of all sequences) in NP microbiomes are listed.
NA- not assigned.
Full list of and relative abundance of taxa in the 96
viruses in the 96 nasopharyngeal samples.
Results from q-PCR for detection of respiratory
We gratefully acknowledge the members of the Spaarne Hospital Research
Center, Hoofddorp, for their dedication and work which made this project
possible and the participating children and their families for their time and
commitment to the studies. We thank Hakim Rahaoui for the excellent
technical assistance. We thank Prof. dr. Marc Lipsitch for his constructive
comments on the manuscript.
Conceived and designed the experiments: DB BK ES. Performed the
experiments: BK JB JR RM. Analyzed the data: DB BK SH. Contributed
reagents/materials/analysis tools: DB BK SH RV ES. Wrote the paper:
DB MB ES. Collected the clinical materials and epidemiological data:
1. (2007) Pneumococcal conjugate vaccine for childhood immunization–WHO
position paper. Wkly Epidemiol Rec 82: 93–104.
2. Blaser MJ, Falkow S (2009) What are the consequences of the disappearing
human microbiota? Nat Rev Microbiol 7: 887–94.
3. Bogaert D, De Groot R, Hermans (2004) PW. Streptococcus pneumoniae
colonisation: the key to pneumococcal disease. Lancet Infect Dis 4:
4. Garcia-Rodriguez JA, Fresnadillo Martinez MJ (2002) Dynamics of nasopha-
ryngeal colonization by potential respiratory pathogens. J Antimicrob Che-
mother 50(Suppl S2): 59–73.
5. Sevillano D, Aguilar L, Alou L, Gimenez MJ, Gonzalez N, et al. (2008) Beta-
lactam effects on mixed cultures of common respiratory isolates as an approach
to treatment effects on nasopharyngeal bacterial population dynamics. PLoS
One 3: e3846.
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org7 February 2011 | Volume 6 | Issue 2 | e17035
6. Brogden KA, Guthmiller JM, Taylor CE (2005) Human polymicrobial
infections. Lancet 2005 365: 253–5.
7. Brook I (1999) Bacterial interference. Crit Rev Microbiol 25: 155–72.
8. Brook I, Gober AE (2005) Recovery of potential pathogens and interfering
bacteria in the nasopharynx of otitis media-prone children and their smoking
and nonsmoking parents. Arch Otolaryngol Head Neck Surg 131: 509–12.
9. White AN, Ng V, Spain CV, Johnson CC, Kinlin LM, et al. (2009) Let the sun
shine in: effects of ultraviolet radiation on invasive pneumococcal disease risk in
Philadelphia, Pennsylvania. BMC Infect Dis 9: 196.
10. Kinlin LM, Spain CV, Ng V, Johnson CC, White AN, et al. (2009)
Environmental exposures and invasive meningococcal disease: an evaluation
of effects on varying time scales. Am J Epidemiol 169: 588–95.
11. Watson K, Carville K, Bowman J, Jacoby P, Riley TV, et al. (2006) Upper
respiratory tract bacterial carriage in Aboriginal and non-Aboriginal children in
a semi-arid area of Western Australia. Pediatr Infect Dis J 25: 782–90.
12. Trotter CL, Greenwood BM (2007) Meningococcal carriage in the African
meningitis belt. Lancet Infect Dis 7: 797–803.
13. Kunin V, Engelbrektson A, Ochman H, Hugenholtz P (2009) Wrinkles in the
rare biosphere: pyrosequencing errors can lead to artificial inflation of diversity
estimates. Environ Microbiol 12: 118–23.
14. Quince C, Lanzen A, Curtis TP, Davenport RJ, Hall N, et al. (2009) Accurate
determination of microbial diversity from 454 pyrosequencing data. Nat
Methods 6: 639–41.
15. Costello EK, Lauber CL, Hamady M, Fierer N, Gordon JI, et al. (2009)
Bacterial Community Variation in Human Body Habitats Across Space and
Time. Science 2009 326: 1694–7.
16. Brook I (2010) Effects of exposure to smoking on the microbial flora of children
and their parents. Int J Pediatr Otorhinolaryngol 74: 447–50.
17. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, et al.
(2009) A core gut microbiome in obese and lean twins. Nature 457: 480–4.
18. 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: 13780–5.
19. Senok AC, Verstraelen H, Temmerman M, Botta GA (2009) Probiotics for the
treatment of bacterial vaginosis. Cochrane Database Syst Rev. pp CD006289.
20. Oztuna F, Ozlu T, Bulbul Y, Buruk K, Topbas M (2006) Does cold environment
affect Streptococcus pneumoniae adherence to rat buccal epithelium? Respira-
tion 73: 546–51.
21. van Gils EJ, Veenhoven RH, Hak E, Rodenburg GD, Bogaert D, et al. (2009)
Effect of reduced-dose schedules with 7-valent pneumococcal conjugate vaccine
on nasopharyngeal pneumococcal carriage in children: a randomized controlled
trial. Jama 302: 159–67.
22. van de Pol AC, Wolfs TF, Jansen NJ, Kimpen JL, van Loon AM, Rossen JW
(2009) Human bocavirus and KI/WU polyomaviruses in pediatric intensive care
patients. Emerg Infect Dis 15: 454–7.
23. Sogin ML, Morrison HG, Huber JA, Mark Welch D, Huse SM, et al. (2006)
Microbial diversity in the deep sea and the underexplored "rare biosphere". Proc
Natl Acad Sci U S A 103: 12115–20.
24. Huse SM, Dethlefsen L, Huber JA, Mark Welch D, Relman DA, et al. (2008)
Exploring microbial diversity and taxonomy using SSU rRNA hypervariable tag
sequencing. PLoS Genet 4: e1000255.
25. Pruesse E, Quast C, Knittel K, Fuch BM, Ludwig W, et al. (2007) SILVA: a
comprehensive online resource for quality checked and aligned ribosomal RNA
sequence data compatible with ARB. Nucleic Acids Res 35: 7188–96.
26. Cole JR, Chai B, Farris RJ, Wang Q, Kulam SA, et al. (2005) The Ribosomal
Database Project (RDP-II): sequences and tools for high-throughput rRNA
analysis. Nucleic Acids Res 33: D294–6.
27. Schloss PD, Handelsman J (2005) Introducing DOTUR, a computer program
for defining operational taxonomic units and estimating species richness. Appl
Environ Microbiol 71: 1501–6.
28. Saeed AI, Bhagabati NK, Braisted JC, Liang W, Sharov V, et al. (2006) TM4
microarray software suite. Methods Enzymol 411: 134–93.
29. Tusher VG, Tibshirani R, Chu G (2001) Significance analysis of microarrays
applied to the ionizing radiation response. Proc Natl Acad Sci U S A 98:
Dynamics of Nasopharyngeal Microbiota in Infants
PLoS ONE | www.plosone.org8 February 2011 | Volume 6 | Issue 2 | e17035