Sputum Microbiota in Tuberculosis as Revealed by 16S
Man Kit Cheung1., Wai Yip Lam2., Wendy Yin Wan Fung2., Patrick Tik Wan Law3, Chun Hang Au1,
Wenyan Nong1, Kai Man Kam4, Hoi Shan Kwan1*, Stephen Kwok Wing Tsui2,5*
1School of Life Sciences, The Chinese University of Hong Kong, Hong Kong SAR, China, 2School of Biomedical Sciences, The Chinese University of Hong Kong, Hong
Kong SAR, China, 3Core Facilities Genome Sequencing Laboratory, The Chinese University of Hong Kong, Hong Kong SAR, China, 4Tuberculosis Reference Laboratory,
Department of Health, Hong Kong SAR, China, 5Hong Kong Bioinformatics Centre, The Chinese University of Hong Kong, Hong Kong SAR, China
Background: Tuberculosis (TB) remains a global threat in the 21st century. Traditional studies of the disease are focused on
the single pathogen Mycobacterium tuberculosis. Recent studies have revealed associations of some diseases with an
imbalance in the microbial community. Characterization of the TB microbiota could allow a better understanding of the
Methodology/Principal Findings: Here, the sputum microbiota in TB infection was examined by using 16S rRNA
pyrosequencing. A total of 829,873 high-quality sequencing reads were generated from 22 TB and 14 control sputum
samples. Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, and Fusobacteria were the five major bacterial phyla
recovered, which together composed over 98% of the microbial community. Proteobacteria and Bacteroidetes were more
represented in the TB samples and Firmicutes was more predominant in the controls. Sixteen major bacterial genera were
recovered. Streptococcus, Neisseria and Prevotella were the most predominant genera, which were dominated by several
operational taxonomic units grouped at a 97% similarity level. Actinomyces, Fusobacterium, Leptotrichia, Prevotella,
Streptococcus, and Veillonella were found in all TB samples, possibly representing the core genera in TB sputum microbiota.
The less represented genera Mogibacterium, Moryella and Oribacterium were enriched statistically in the TB samples, while
a genus belonging to the unclassified Lactobacillales was enriched in the controls. The diversity of microbiota was similar in
the TB and control samples.
Conclusions/Significance: The composition and diversity of sputum microbiota in TB infection was characterized for the
first time by using high-throughput pyrosequencing. It lays the framework for examination of potential roles played by the
diverse microbiota in TB pathogenesis and progression, and could ultimately facilitate advances in TB treatment.
Citation: Cheung MK, Lam WY, Fung WYW, Law PTW, Au CH, et al. (2013) Sputum Microbiota in Tuberculosis as Revealed by 16S rRNA Pyrosequencing. PLoS
ONE 8(1): e54574. doi:10.1371/journal.pone.0054574
Editor: Stefan Bereswill, Charite ´-University Medicine Berlin, Germany
Received October 24, 2012; Accepted December 12, 2012; Published January 24, 2013
Copyright: ? 2013 Cheung 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 the Research Fund for the Control of Infectious Diseases (RFCID) from the Health, Welfare and Food Bureau of the Hong
Kong SAR Government (Ref. No. 09080442). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org (HSK); email@example.com (SKWT)
. These authors contributed equally to this work.
Tuberculosis (TB) is an airborne infectious disease caused by the
bacterium Mycobacterium tuberculosis. In 2010, there were an
estimated 8.5 million cases of TB worldwide, causing 1.2 million
deaths . Global control of TB has relied on the 90-year-old and
largely ineffective Bacille Calmette-Gue ´rin (BCG) vaccine and
a few decades-old drugs . The growing epidemics of HIV/
AIDS and multidrug-resistant TB strains have further complicated
the effectiveness of disease control . As with most classical
disease investigations, TB studies have been focused on a single
causative agent. However, recent understandings suggest that
disease phenotypes are more likely to be a result of complex
microbial interactions . Characterizing the whole microbial
communities in TB infections might provide further insights into
The advent of the cloning-independent and massively parallel
454 pyrosequencing technology  has facilitated investigations of
the soil microbial populations , marine microbial communities
 and the human gut microbiota . Just recently, the
microbiota of respiratory diseases such as cystic fibrosis (CF)
[9,10], chronic obstructive pulmonary disease (COPD)  and
nosocomial pneumonia  were also characterized by using this
approach. By revealing the composition and diversity of micro-
biota associated with these diseases, these studies allow the
development of hypotheses relating the microbial communities
and the disease states. Information on the respiratory microbiota
in TB infection is currently unavailable. Here, based on 16S rRNA
pyrosequencing, the composition and diversity of sputum micro-
biota in TB infection was examined for the first time. The
PLOS ONE | www.plosone.org1January 2013 | Volume 8 | Issue 1 | e54574
microbiota of non-TB case-matched sputum samples was also
characterized for comparison.
A total of 964,556 raw 16S rRNA reads were obtained in this
study. About 578,000 reads were from the 22 TB samples and the
remaining ,387,000 reads were from the 14 controls (Table 1).
The filtering process removed about 14% of the raw sequencing
reads. There were about 499,000 and 331,000 high-quality reads
for the TB and control samples, respectively. The average read
length was about 370 bp, after removal of the primer sequences.
The average number of qualified reads for the TB group was
22,660, and that for the control group was 23,667. Rarifying the
number of reads to 13,210, there was an average of 486 and 487
OTUs in the TB and control samples, respectively. The average
Chao1 value calculated was 832 in the TB samples and 870 in the
controls. The average Shannon index was 4.85 in the TB samples
and 4.80 in the controls.
Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria, and Fusobacteria
were the major bacterial phyla recovered, which together
composed over 98% of the microbial community (Figure 1).
Firmicutes, Proteobacteria and Bacteroidetes were among the most
abundant phyla in both the TB and control samples, composing
37.6%, 31.2% and 19.2% in the TB samples and 43.6%, 27.1%
and 17.0% in the controls, respectively. Whereas the relative
abundance of Actinobacteria and Fusobacteria between the two groups
was similar, Proteobacteria and Bacteroidetes were more represented in
the TB samples and Firmicutes was more dominant in the controls.
Breakdown of the data revealed a large variation in the
community structure among the individuals (Figure 2). This was
supported by PCoA analysis, in which no obvious differential
clustering was observed between members of the TB and control
groups (Figure 3).
Sixteen major bacterial genera were recovered in the sputum
samples here (Figure 4). The most abundant genera in the TB
samples were Neisseria (28.0%), Streptococcus (27.8%) and Prevotella
(16.8%), whereas those for the controls were Streptococcus (31.8%),
Neisseria (22.0%) and Prevotella (14.4%). Neisseria and Prevotella were
more represented in the TB samples and Streptococcus was more
predominant in the controls. Lactococcus, Pseudomonas and un-
classified Enterobacteriaceae were less dominant and prevalent in the
TB samples (Figures 4 and 5). Within the 16 major genera, only
Actinomyces, Fusobacterium, Leptotrichia, Prevotella, Streptococcus, and
Veillonella were recovered in all the TB samples, and all these
genera besides Leptotrichia were found in all the 36 samples
(Figure 5). Breakdown of the data again revealed inter-member
variations in the community structure at the genus level (Figure 6).
Two OTUs together composed more than 90% of the
dominant genus Neisseria in both sample groups (Table 2). OTU
6783 was most similar to the uncultured clone NSV3Q1b18
sampled from a showerhead swab , and contributed 57.2%
and 88.7% of the genus in the TB and control samples,
respectively. On the other hand, OTU 7988 was most similar to
the uncultured clone 7H59 from the subgingival plaque of
a periodontitis-free patient , and it was more represented in
Neisseria in the TB samples (37.4%) than in the controls (4.1%).
About one-third of another major genus Streptococcus was repre-
sented by three OTUs in both sample groups (Table 2). OTU
1734 was most similar to a S. mitis clone sampled from the oral
cavity , and was more represented in Streptococcus in the TB
samples (12.0%) than in the controls (1.6%). On the other hand,
OTU 6370 was most similar to a S. parasanguinis clone sampled
from the oral cavity, and was more represented in the control
samples (16.3%) than in the TB samples (5.6%). OTU 7204 was
most similar to the S. salivarius strain ATCC 7073, and its
contribution in the genus was relatively comparable in both groups
(12.4% in TB and 14.6% in control). Two OTUs together
composed more than one-third of the genus Prevotella in both the
TB and control samples (Table 2). They were most similar to
uncultured bacterial clones sampled from the human oral cavity.
In terms of the less represented taxa, genera Moryella,
Mogibacterium and Oribacterium were found statistically enriched in
the TB samples (p,0.05) whereas a genus belonging to the
unclassified Lactobacillales was enriched in the control samples
(p,0.05) (Table 3). At the species level, eight OTUs, including
those showing 99% identity to Prevotella melaninogenica, Lactobacillus
crispatus, Streptococcus anginosus, and Parvimonas micra, were found
enriched in the TB samples (p,0.05). Two other OTUs, including
one that is 99% identical to Aggregatibacter aphrophilus, were enriched
in the control samples (p,0.05) (Table 3).
Table 1. Patient and sequencing run characteristics.
Number of patients2214
Patient age (median, range)41, 20–66*59, 22–82
Patient gender (male/female)13/8* 6/8
Raw sequencing reads577,999386,557
Average reads per sample22,660 23,667
*Information on age and sex are only available for 21 TB patients.
Figure 1. Relative abundance of dominant bacterial phyla in TB
and control samples. Only phyla .1% in either group were included.
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org2January 2013 | Volume 8 | Issue 1 | e54574
TB remains a global threat in the 21st century. Lead by the
World Health Organization, the Stop TB Partnership has planned
to halve the global TB prevalence and mortality by 2015 and to
eliminate the disease as a public health problem by 2050 . A
better understanding of the disease from an angle different from
the traditional single-pathogen-based approach could help to meet
these targets. Indeed, instead of the presence of a single causative
pathogen, it was shown that some diseases are associated with and
might result from an imbalance in the microbial community,
a condition known as dysbiosis. Inflammatory bowel disease 
and obesity  are two of the well-known examples. The advent
of next-generation sequencing technologies, such as 454 pyrose-
quencing, has greatly advanced studies of the human microbiota
. However, unlike the widely studied gut microbiota, the
investigation of the lung-associated microbiota is relatively new
. In this study, the sputum microbiota in TB infection was
characterized by using 16S rRNA pyrosequencing for the first
time, aiming to advance understanding of the disease in the
microbial community level.
Figure 2. Relative abundance of dominant bacterial phyla in all individuals under study. Only phyla .1% in either group were included.
The 22 samples on the left were from the TB patients, and the 14 samples on the right were from the control group.
Figure 3. PCoA plots of TB and control samples. The plots were based on a) weighted and b) unweighted UniFrac distances. The two principal
coordinates combined explained 66.44% and 24.31% of the variations in the weighted and unweighted cases, respectively. Red dots represent TB
samples; blue dots represent control samples.
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org3January 2013 | Volume 8 | Issue 1 | e54574
About one million raw sequencing reads were generated for
a total of 36 sputum samples in this study. However, the
throughput is still insufficient to capture every member in the
microbial community. An example particularly worth mentioning
is M. tuberculosis. Although being the causative agent of the disease,
M. tuberculosis represents only a very small portion of the whole
microbiota in the TB sputum samples (Figure 4). To obtain
positive results in the culture and the smear tests, there should be
at least 100 M. tuberculosis cells/ml and ,104acid-fast bacilli (AFB)
cells/ml, respectively . Because the average total bacterial load
in sputum samples could reach 109cells/ml , 105reads may be
needed to recover a single read of M. tuberculosis in a particular TB
sample. Since the average sequencing throughput was ,27,000
reads per sample here, it is not unexpected that no reads of M.
tuberculosis were recovered in some of the TB sputum samples.
Nevertheless, the sequencing throughput is sufficient to reveal the
major phyla and genera of the sputum microbiota, and it allows
meaningful comparisons between the TB and control groups.
Five major bacterial phyla were recovered in the sputum
samples collected in this study; they were the Firmicutes,
Proteobacteria, Bacteroidetes, Actinobacteria, and Fusobacteria. They are
also the predominant phyla in other human body sites, including
Figure 4. Relative abundance of dominant bacterial genera in TB and control samples. Only genera .1% in either group were included.
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org4January 2013 | Volume 8 | Issue 1 | e54574
the oral cavity, skin and colon . In addition, they dominate the
microbiota in the CF sputum  and the bronchial tract of
COPD patients . The three most dominant phyla retrieved in
this study (Firmicutes, Proteobacteria and Bacteroidetes) are also the most
commonly identified phyla in the normal lung microbiota .
These collectively suggest the prevalence of these major phyla in
normal and diseased lung microbiota, as well as the human
microbiota in general. Within the 16 major genera recovered in
this study, only Actinomyces, Fusobacterium, Leptotrichia, Prevotella,
Streptococcus, and Veillonella were found in all the TB samples, and
they may represent the core genera in the TB sputum microbiota.
Similarly, except Leptotrichia, the remaining five genera were found
Figure 5. Prevalence of dominant bacterial genera in TB (n=22) and control (n=14) samples. Only genera .1% in either group were
Figure 6. Relative abundance of dominant bacterial genera in all individuals under study. Only genera .1% in either group were
included. The 22 samples on the left were from the TB patients, and the 14 samples on the right were from the control group.
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org5 January 2013 | Volume 8 | Issue 1 | e54574
in all 36 samples, and they may represent the core genera in the
sputum microbiota in general. However, additional studies are
needed to confirm this.
Streptococcus, Neisseria and Prevotella were the three most pre-
dominant genera recovered in the sputum samples collected in the
current study. They are also the major genera retrieved in the
sputum of patients with COPD , nosocomial pneumonia 
and CF [9,10]. This is in contrast to the normal lung microbiota,
in which the genus Pseudomonas dominates . The genus
Streptococcus encompasses both commensal and pathogenic Gram-
positive bacteria inhabiting various human body sites, including
the oral cavity and the upper respiratory tract. Streptococcus
pneumoniae, for instance, is a well-known pathogen associated with
pneumonia . Neisseria is a group of Gram-negative bacteria
that colonizes human mucosal epithelia. Commensal species
include N. lactamica and N. mucosa, while pathogenic species such
as N. meningitidis could cause pneumonia . Prevotella contains
obligate anaerobes commonly isolated from the human oral and
intestinal tracts, and it can cause lower respiratory tract infections.
Members within this genus are believed to cause pulmonary
exacerbations through polymicrobial interactions that enhance the
virulence of the established pathogens .
Each of the three most predominant genera was dominated by
two to three major OTUs that together composed one-third to
90% of the groups. OTU 1734 of the genus Streptococcus is 100%
identical to S. mitis and was more represented in Streptococcus in the
TB samples than in the controls. Although usually commensal, S.
mitis could become a significant pathogen in immunocompromised
patients as well as individuals implicated in a wide range of
diseases, including pneumonia . Indeed, genome study of the
S. mitis B6 strain has revealed homologues of many previously
identified S. pneumoniae virulence factors such as autolysins .
Also, incubation of S. mitis in the gingival epithelial cells was
reported to induce the expression of human b-defensin 2 that can
kill other oral pathogens . It is thus reasonable to speculate
that the increased relative abundance of the opportunistic
Table 2. List of abundant OTUs composing the three major bacterial genera.
TB (%)Control (%)BLAST result (Identity; isolation source; GenBank accession no.)*
BacteroidetesPrevotella2225 19.726.6Uncultured clone 071030_015; oral cavity; JQ471460
BacteroidetesPrevotella799914.410.7Uncultured clone 071070_272; oral cavity; JQ476683
FirmicutesStreptococcus173412.01.6 Streptococcus mitis clone; oral cavity; GU422750
FirmicutesStreptococcus6370 5.616.3Streptococcus parasanguinis clone; oral cavity; GU412056
Firmicutes Streptococcus720412.414.6Streptococcus salivarius strain ATCC 7073; NA; NR_042776
ProteobacteriaNeisseria678357.2 88.7Uncultured clone NSV3Q1b18; showerhead swab; EU629781
ProteobacteriaNeisseria7988 37.4 4.1Uncultured clone 7H59; subgingival plaque; JX010905
*All with 100% sequence similarity.
Table 3. List of genera and species significantly different between TB and control groups.
abundance (%)Enriched in
Phylum GenusSpecies (% identity)*
Bacteroidetes PrevotellaPrevotella melaninogenica (99%)0.00500.00030.0292v
BacteroidetesPrevotellaPrevotella histicola (97%)0.00560.00040.0481v
ProteobacteriaNeisseria Neisseria flavescens (92%)0.002200.0413v
Firmicutes StreptococcusStreptococcus parasanguinis (92%)0.002900.0342 v
FirmicutesLactobacillus Lactobacillus crispatus (99%) 0.02270.00040.0363 v
FirmicutesStreptococcusStreptococcus anginosus (99%) 0.00630.00030.0380v
FirmicutesStreptococcusStreptococcus cristatus (97%)0.01380.00120.0430v
FirmicutesParvimonasParvimonas micra (99%)0.12770.01180.0492v
FirmicutesGranulicatellaGranulicatella adiacens (92%) 0 0.0011 0.0435v
ProteobacteriaAggregatibacterAggregatibacter aphrophilus (99%)0.00040.00450.0228v
‘OTUs clustered at 97% similarity.
#Using RDP classifier, down to the genus level.
*Against the NCBI 16S rRNA database.
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org6January 2013 | Volume 8 | Issue 1 | e54574
pathogen in TB infections could alter the microbial community in
TB lung and affect TB pathogenesis and progression. Unlike OTU
1734, OTU 6370, which is 100% identical to S. parasanguinis, was
more represented in the control samples. S. parasanguinis is
a member of the viridans streptococci that constitute the major
population of the human oral microbiota. It is also one of the early
colonizers of the tooth surface . OTU 7204 is most similar to
S. salivarius and was of relatively comparable abundance in
Streptococcus in both groups. Although being another member of
the viridans streptococci, S. salivarius commonly causes bacteremia
. The two dominant Neisseria OTUs and the two dominant
Prevotella OTUs are most similar to uncultured clones with limited
information. Additional research works on these organisms are
urged to better understand the potential effects of these abundant
members in TB or sputum microbiota.
The less abundant taxa may also affect the dynamics of the
microbial community and clinical outcomes . Mogibacterium,
Moryella and Oribacterium were the genera enriched statistically in
the TB samples collected in this study. Mogibacterium is a group of
Gram-positive anaerobic bacteria, in which members such as M.
timidum have been identified in cases of acute lung infections .
On the other hand, Moryella and Oribacterium are genera of
anaerobic bacteria that were identified in the recent decade
[33,34]. Little information is available since their discovery. At the
species level, eight less represented OTUs were found significantly
enriched in the TB samples, each with an over 10-fold difference
in relative abundance between the sample groups. Four of them
are with high similarity (99%) to known bacterial species, including
Prevotella melaninogenica, Lactobacillus crispatus, Streptococcus anginosus,
and Parvimonas micro. P. melaninogenica is a Gram-negative anaerobic
bacterium that represents a ubiquitous member of the oral
commensal microbiota. However, putative virulence factors such
as hemagglutination were reported in some strains .
Lactobacilli are Gram-positive bacteria that are usually regarded
as beneficial to human health. However, L. crispatus strain
M206119 was found to exacerbate dextran sulfate sodium
(DSS)-induced colitis in mice by interfering with their murine
inflammatory responses . It was reported that S. anginosus
infections might simulate tuberculosis . Indeed, three cases of
co-infection by S. anginosus and M. tuberculosis were reported
previously . P. micra is a Gram-positive anaerobe usually
isolated from the human oral cavity. However, association of the
bacterium with respiratory and gastrointestinal tract infections was
reported previously . The enrichment of these opportunistic
pathogens in the TB sputum revealed here suggests their
association with TB pathogenesis or progression.
Bacterial community diversity in CF sputum was found to be
positively correlated with the patient stability [9,36]. It was also
reported that a decrease in microbial diversity in the gastrointes-
tinal tract is associated with an increased incidence of in-
flammatory bowel disease . However, bacterial diversity was
found to be significantly higher among asthmatic patients than in
controls . In our study, the microbial diversity was found to be
similar in the TB and the control sputum samples. Thus, there
does not seem to be a general relationship between microbial
diversity and the disease state. Indeed, this is not unexpected
because different diseases may involve different sets of microbiota,
within which interactions could be complex, including both
synergistic and antagonistic effects among members .
The current study represents the first characterization of the
sputum microbiota in TB infection based on 16S rRNA
pyrosequencing. The work allows future examination of potential
roles played by the diverse microbiota in TB pathogenesis and
progression, and it could ultimately facilitate advances in TB
treatment. However, additional studies on the uncultured taxa are
urged. Further investigations based on metatranscriptomics,
metaproteomics and metabolomics on the RNA, protein and
metabolite levels, respectively, are also expected to provide further
insights into understanding the disease.
Materials and Methods
The study was approved by the Joint CUHK-NTEC Clinical
Research Ethical Committee at the Prince of Wales Hospital in
Hong Kong. Written informed consent was obtained from all the
studied subjects for sample collection and subsequent analysis.
Sample Collection and DNA Extraction
Sputum samples were collected from the Tuberculosis Refer-
ence Laboratory of the Hong Kong government in a routine TB
screening exercise. All samples in this study were from patients
who are ethnically Hong Kong Chinese. All the patients were free
of HIV. No anti-TB or other antibiotic medication was given to
the patients within four weeks of sputum collection. Smear-positive
and culture-positive TB sputum samples were collected from 22
patients (13 males and eight females, aged between 20 and 66;
data missing for one patient) (Table 1, Dataset S1). Control
sputum samples were also collected from 14 individuals (six males
and eight females, aged between 22 and 82) with TB-resembling
coughing symptoms but being culture-negative for comparison.
The sputum samples were treated with a 3% NaOH solution,
neutralized with a phosphate buffer and then centrifuged. About
200 ml supernatant was subjected to genomic DNA extraction
using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA, USA).
PCR and Pyrosequencing
PCR was performed using composite primers flanking the
hypervariable V1–V2 region of the 16S rRNA gene: A-8F (59-
CTGCTGCCTYCCGTA-39) . Sequencing adaptors and
sample-specific multiplex identifiers (MIDs) were also added to
the 59 ends of the primers according to the manufacturer’s
protocol. PCR reactions were carried out using the PlatinumH
PCR SuperMix High Fidelity Kit (Invitrogen, Carlsbad, CA, USA).
The PCR thermal regime consisted of an initial denaturation of
30 s at 94uC, followed by 28 cycles of 30 s at 94uC, 30 s at 55uC
and 40 s at 68uC. PCR products were purified using the MinElute
PCR Purification Kit (Qiagen, Valencia, CA, USA) and the quality
of the purified products was assessed using Agilent Bioanalyzer
2100. Pyrosequencing of the purified PCR products was
performed uni-directionally from the A-8F primer end and on
a GS FLX-Titanium platform in a single full-plate run. The
sequencing data were submitted to the NCBI short read archive
(accession number SRA058505).
and B-357R (59-
Raw sequencing reads were demultiplexed, quality-filtered and
analyzed using QIIME 1.4.0 . Briefly, reads shorter than
200 bp or longer than 1000 bp, with an average quality score
lower than 25 in a sliding window of 50 bp, with mismatching
primer sequences, or with ambiguous bases (Ns) were removed
from downstream analyses. Quality-filtered reads were clustered
into operational taxonomic units (OTUs) at 97% similarity.
Taxonomic assignment of representative OTUs was performed
using the RDP Classifier  at a 0.8 confidence threshold against
the Greengenes core set . The dataset was rarified before
alpha diversity calculations. Principal coordinate analysis (PCoA)
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org7January 2013 | Volume 8 | Issue 1 | e54574
was performed using the weighted and unweighted UniFrac Download full-text
distances . Statistical analyses were performed using SAS.
Information of patients and samples exam-
Conceived and designed the experiments: WYWF HSK SKWT.
Performed the experiments: WYWF PTWL. Analyzed the data: MKC
WYL CHA WN. Contributed reagents/materials/analysis tools: KMK.
Wrote the paper: MKC WYL.
1. WHO (2011) Global tuberculosis control: WHO report 2011. Available: http://
www.who.int/tb/publications/global_report/en/index.html. Accessed 2012
2. Young DB, Perkins MD, Duncan K, Barry CE 3rd(2008) Confronting the
scientific obstacles to global control of tuberculosis. J Clin Invest 118: 1255–
3. Comas I, Gagneux S (2009) The past and future of tuberculosis research. PLoS
Pathog 5: e1000600.
4. Clemente JC, Ursell LK, Parfrey LW, Knight R (2012) The impact of the gut
microbiota on human health: an integrative view. Cell 148: 1258–1270.
5. Margulies M, Egholm M, Altman WE, Attiya S, Bader JS, et al. (2005) Genome
sequencing in microfabricated high-density picolitre reactors. Nature 437: 376–
6. Palmer K, Biasi C, Horn MA (2012) Contrasting denitrifier communities relate
to contrasting N2O emission patterns from acidic peat soils in arctic tundra.
ISME J 6: 1058–1077.
7. Cheung MK, Au CH, Chu KH, Kwan HS, Wong CK (2010) Composition and
genetic diversity of picoeukaryotes in subtropical coastal waters as revealed by
454 pyrosequencing. ISME J 4: 1053–1059.
8. Saulnier DM, Riehle K, Mistretta TA, Diaz MA, Mandal D, et al. (2011)
Gastrointestinal microbiome signatures of pediatric patients with irritable bowel
syndrome. Gastroenterology 141: 1782–1791.
9. Filkins LM, Hampton TH, Gifford AH, Gross MJ, Hogan DA, et al. (2012)
Prevalence of streptococci and increased polymicrobial diversity associated with
cystic fibrosis patient stability. J Bacteriol 194: 4709–4717.
10. Guss AM, Roeselers G, Newton IL, Young CR, Klepac-Ceraj V, et al. (2011)
Phylogenetic and metabolic diversity of bacteria associated with cystic fibrosis.
ISME J 5: 20–29.
11. Cabrera-Rubio R, Garcia-Nu ´n ˜ez M, Seto ´ L, Anto ´ JM, Moya A, et al. (2012)
Microbiome diversity in the bronchial tract of patients with chronic obstructive
pulmonary disease. J Clin Microbiol 50: 3562–3568.
12. Zhou Y, Lin P, Li Q, Han L, Zheng H, et al. (2010) Analysis of the microbiota of
sputum samples from patients with lower respiratory tract infections. Acta
Biochim Biophys Sin (Shanghai) 42: 754–76.
13. Feazel LM, Baumgartner LK, Peterson KL, Frank DN, Harris JK, et al. (2009)
Opportunistic pathogens enriched in showerhead biofilms. Proc Natl Acad
Sci U S A 106: 16393–16399.
14. You M, Mo S, Watt RM, Leung WK Prevalence and diversity of Synergistetes taxa
in periodontal health and disease. J Periodontal Res. In press.
15. Dewhirst FE, Chen T, Izard J, Paster BJ, Tanner AC, et al. (2010) The human
oral microbiome. J Bacteriol 192: 5002–5017.
16. Stop TB Partnership (2011) The global plan to stop TB 2011–2015. Available:
http://www.stoptb.org/global/plan/. Accessed 2012 Oct 24.
17. Spor A, Koren O, Ley R (2011) Unravelling the effects of the environment and
host genotype on the gut microbiome. Nat Rev Microbiol 9: 279–290.
18. Turnbaugh PJ, Ridaura VK, Faith JJ, Rey FE, Knight R, et al. (2009) The effect
of diet on the human gut microbiome: a metagenomic analysis in humanized
gnotobiotic mice. Sci Transl Med 1: 6ra14.
19. Cho I, Blaser MJ (2012) The human microbiome: at the interface of health and
disease. Nat Rev Genet 13: 260–270.
20. Beck JM, Young VB, Huffnagle GB (2012) The microbiome of the lung. Transl
Res 160: 258–266.
21. Vignesh R, Balakrishnan P, Shankar EM, Murugavel KG, Hanas S, et al. (2007)
Value of single acid-fast bacilli sputum smears in the diagnosis of tuberculosis in
HIV-positive subjects. J Med Microbiol 56: 1709–1710.
22. Zhao J, Schloss PD, Kalikin LM, Carmody LA, Foster BK, et al. (2012) Decade-
long bacterial community dynamics in cystic fibrosis airways. Proc Natl Acad
Sci U S A 109: 5809–5814.
23. Nobbs AH, Lamont RJ, Jenkinson HF (2009) Streptococcus adherence and
colonization. Microbiol Mol Biol Rev 73: 407–450.
24. Brigham KS, Sandora TJ (2009) Neisseria meningitidis: epidemiology, treatment
and prevention in adolescents. Curr Opin Pediatr 21: 437–443.
25. Field TR, Sibley CD, Parkins MD, Rabin HR, Surette MG (2010) The genus
Prevotella in cystic fibrosis airways. Anaerobe 16: 337–344.
26. Mitchell J (2011) Streptococcus mitis: walking the line between commensalism and
pathogenesis. Mol Oral Microbiol 26: 89–98.
27. Denapaite D, Bru ¨ckner R, Nuhn M, Reichmann P, Henrich B, et al. (2010) The
genome of Streptococcus mitis B6 - what is a commensal? PLoS One 5: e9426.
28. Eberhard J, Pietschmann R, Falk W, Jepsen S, Dommisch H (2009) The
immune response of oral epithelial cells induced by single-species and complex
naturally formed biofilms. Oral Microbiol Immunol 24: 325–330.
29. Geng J, Chiu CH, Tang P, Chen Y, Shieh HR, et al. (2012) Complete genome
and transcriptomes of Streptococcus parasanguinis FW213: phylogenic relations and
potential virulence mechanisms. PLoS One 7: e34769.
30. Han XY, Kamana M, Rolston KV (2006) Viridans streptococci isolated by
culture from blood of cancer patients: clinical and microbiologic analysis of 50
cases. J Clin Microbiol 44: 160–165.
31. Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, et al. (2010) A human gut
microbial gene catalogue established by metagenomic sequencing. Nature 464:
32. Hill GB, Ayers OM, Kohan AP (1987) Characteristics and sites of infection of
Eubacterium nodatum, Eubacterium timidum, Eubacterium brachy, and other asacchar-
olytic eubacteria. J Clin Microbiol 25: 1540–1545.
33. Carlier JP, K’ouas G, Bonne I, Lozniewski A, Mory F (2004) Oribacterium sinus
gen. nov., sp. nov., within the family ‘Lachnospiraceae’ (phylum Firmicutes). Int J Syst
Evol Microbiol 54: 1611–1615.
34. Carlier JP, K’ouas G, Han XY (2007) Moryella indoligenes gen. nov., sp. nov., an
anaerobic bacterium isolated from clinical specimens. Int J Syst Evol Microbiol
35. Haraldsson G, Meurman JH, Ko ¨no ¨nen E, Holbrook WP (2005) Properties of
hemagglutination by Prevotella melaninogenica. Anaerobe 11: 285–289.
36. Zhou FX, Chen L, Liu XW, Ouyang CH, Wu XP, et al. (2012) Lactobacillus
crispatus M206119 exacerbates murine DSS-colitis by interfering with in-
flammatory responses. World J Gastroenterol 18: 2344–2356.
37. Jacobs JA, Pietersen HG, Walenkamp GH, Stobberingh EE, Soeters PB (1994)
Intervertebral infection caused by Streptococcus milleri. A case report. Clin Orthop
Relat Res 302: 183–188.
38. Rabun ˜al R, Corredoira J, Monte R, Coira A (2009) Co-infection by Streptococcus
anginosus and Mycobacterium tuberculosis: three case reports. J Med Case Rep 3: 37.
39. Murdoch DA, Shah HN (1999) Reclassification of Peptostreptococcus magnus (Prevot
1933) Holdeman and Moore 1972 as Finegoldia magna comb. nov. and
Peptostreptococcus micros (Prevot 1933) Smith 1957 as Micromonas micros comb.
nov. Anaerobe 5: 555–559.
40. Ott SJ, Musfeldt M, Wenderoth DF, Hampe J, Brant O, et al. (2004) Reduction
in diversity of the colonic mucosa associated bacterial microflora in patients with
active inflammatory bowel disease. Gut 53: 685–93.
41. Huang YJ, Nelson CE, Brodie EL, Desantis TZ, Baek MS, et al. (2011) Airway
microbiota and bronchial hyperresponsiveness in patients with suboptimally
controlled asthma. J Allergy Clin Immunol 127: 372–381.
42. Pe ´dron T, Sansonetti P (2008) Commensals, bacterial pathogens and intestinal
inflammation: an intriguing me ´nage a ` trois. Cell Host Microbe 3: 344–347.
43. McKenna P, Hoffmann C, Minkah N, Aye PP, Lackner A, et al. (2008) The
macaque gut microbiome in health, lentiviral infection, and chronic enteroco-
litis. PLoS Pathog 4: e20.
44. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, et al. (2010)
QIIME allows analysis of high-throughput community sequencing data. Nat
Methods 7: 335–336.
45. Wang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naive Bayesian classifier for
rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl
Environ Microbiol 73: 5261–5267.
46. DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, et al. (2006)
Greengenes, a chimera-checked 16S rRNA gene database and workbench
compatible with ARB. Appl Environ Microbiol 72: 5069–5072.
47. Lozupone C, Knight R (2005) UniFrac: a new phylogenetic method for
comparing microbial communities. Appl Environ Microbiol 71: 8228–8235.
Pyrosequencing Sputum Microbiota in Tuberculosis
PLOS ONE | www.plosone.org8January 2013 | Volume 8 | Issue 1 | e54574