RAPID COMMUNICATION
Biological
B.J.F. Keijser1, E. Zaura2, S.M. Huse3,
J.M.B.M. van der Vossen1, F.H.J. Schuren1,
R.C. Montijn1, J.M. ten Cate2,
and W. Crielaard2*
1 TNO Quality of Life, Business Unit Food and
Biotechnology Innovations, Microbial Genomics Group,
Zeist, The Netherlands; 2 Department of Cariology
Endodontology Pedodontology, Academic Centre for
Dentistry Amsterdam (ACTA), University of Amsterdam
and VU Amsterdam, Louwesweg 1, 1066 EA Amsterdam,
The Netherlands; and 3 Josephine Bay Paul Center,
Marine Biological Laboratory, Woods Hole, MA, USA;
*corresponding author, w.crielaard@acta.nl
J Dent Res 87(11):1016-1020, 2008
ABSTRACT
A good definition of commensal microflora and
an understanding of its relation to health are
essential in preventing and combating disease. We
hypothesized that the species richness of human
oral microflora is underestimated. Saliva and
supragingival plaque were sampled from 71 and
98 healthy adults, respectively. Amplicons from
the V6 hypervariable region of the small-subunit
ribosomal RNA gene were generated by PCR,
pooled into saliva and plaque pools, and sequenced
by means of the Genome Sequencer 20 system
at 454 Life Sciences. Data were evaluated by
taxonomic and rarefaction analyses. The 197,600
sequences generated yielded about 29,000 unique
sequences, representing 22 taxonomic phyla.
Grouping the sequences in operational taxonomic
units (6%) yielded 3621 and 6888 species-level
phylotypes in saliva and plaque, respectively.
This work gives a radically new insight into the
diversity of human oral microflora, which, with
an estimated number of 19,000 phylotypes, is
considerably higher than previously reported.
Abbreviations: GS-20, Genome Sequencer
20; OTU, Operational Taxonomic Unit; PCR,
polymerase chain-reaction; 16S rDNA, small-
subunit (16S) ribosomal deoxyribonucleic acid;
UV, ultraviolet.
KEY WORDS: pyrosequencing, microflora, plaque,
saliva, diversity.
Received April 25, 2008; Last revision August 1, 2008;
Accepted August 25, 2008
A supplemental appendix to this article is published
electronically only at http://jdr.iadrjournals.org/cgi/ content/
full/87/11/1016/DC1.
1016
Pyrosequencing Analysis of the
Oral Microflora of Healthy Adults
INTRODUCTION
O
ral health and disease depend on the interplay between the host and the
oral microbial community. The impact of the microbial community on
shifting the balance from health to disease cannot be understood without
a comprehensive view of a healthy community. During the past 40 years,
a wealth of knowledge has been gathered: Over 250 oral bacterial species
have been isolated and characterized by cultivation, and over 450 species
have been identified by culture-independent molecular approaches (Paster et
al., 2006).
Recent studies where cloning and sequencing approaches of microbial
16S rDNA were applied to the oral microflora have identified still more new
species (Aas et al., 2008; Preza et al., 2008; Riggio et al., 2008). This is
not surprising, since the number of species detected in a sample is strongly
affected by the number of sequences analyzed (Schloss and Handelsman,
2005), and overall diversity is affected by the sample size (Rajilić-Stojanović
et al., 2007). This, however, limits the cloning and sequencing method
(typically at most a few thousand clones from a low number of individuals)
to identification of only the predominant micro-organisms in a sample.
Detection of low-abundance taxa requires studies on sets of sequences many
orders of magnitude larger than those currently reported.
Massively parallel pyrosequencing—a next-generation sequencing
technique—is a new molecular approach that allows for extensive
sequencing of microbial populations in a high-throughput, cost-effective
manner (Ronaghi et al., 1998; Von Bubnoff, 2008). This technique has
been successfully applied to determine bacterial diversity within various
environmental ecosystems, such as hydrothermal vents of a deep marine
biosphere (Sogin et al., 2006; Huber et al., 2007) and soil (Roesch et al.,
2007). Within the human body, vaginal microflora (Sundquist et al., 2007)
and bacteria of chronic wounds (Dowd et al., 2008) have been assessed
by this approach, but we are not aware of reports on the oral microbial
population.
We aimed to estimate the detailed species richness of oral microflora
of the healthy adult population, and, more specifically, to determine the
number of different phylotypes and their relative abundance in saliva and
supragingival plaque.
Materials & Methods
Samples
The study was approved by the Medical Ethical Committee of the Free University
Amsterdam. Participants (healthy adults who had not used antibiotics in the
preceding 3 mos) were informed verbally and participated on the basis of informed
consent. Sampling was performed in the morning before the participants ate
breakfast. Saliva was collected from a group of 71 individuals by mouthrinse
with 10 mL sterile saline for 30 sec (Boutaga et al., 2007). The saline solution
had been UV-irradiated to avoid DNA contamination. Samples were placed on
ice immediately and stored at -80°C until further use. We collected supragingival
plaque from a group of 98 individuals by sampling buccal dental surfaces using a
J Dent Res 87(11) 2008 Pyrosequencing of Commensal Oral Microflora 1017

sterile, DNA-free wooden toothpick. The toothpick tip was cut off, to multiple V6 reference sequences, and/or where identical V6
placed in an Eppendorf tube, and stored at -80°C. sequences were derived from longer sequences mapping to different
taxa, tags were assigned to the lowest common taxon. We created
Molecular Techniques OTUs and OTU rarefaction curves by aligning unique tag sequences
Individual saliva samples were vortexed, and a 0.3-mL quantity was and calculating distance matrices as previously described (Sogin et
transferred to a sterile screw-cap Eppendorf tube with 0.3 g zirconia- al., 2006), and using DOTUR (Schloss and Handelsman, 2005) to
silica beads (diameter, 0.1 mm; Biospec Products, Bartlesville, OK, create clusters at the unique, 0.03, 0.06, and 0.10 levels. Additional
USA). For plaque samples, the toothpick tip was placed in a sterile estimators were calculated with EstimateS (Version 7.5, R. K.
screw-cap Eppendorf tube with 0.3 g zirconia-silica beads, and a Colwell, http://purl.oclc.org/estimates).
100-mL quantity of sterile UV-irradiated water was added. Then, 0.2
mL phenol was added, and the samples were homogenized with a Results
Mini-beadbeater (Biospec Products) for 2 min. DNA was extracted
More than 273,000 PCR amplicons were sequenced, of which
with the AGOWA mag Mini DNA Isolation Kit (AGOWA, Berlin,
about 197,600 reads passed quality control (Table 1). The
Germany), quantified (Nanodrop ND-1000; NanoDrop Technologies,
individual sequences could be clustered into 28,937 unique V6
Montchanin, DE, USA), and analyzed with the Agilent 2100
tag sequences, representing 22 known phyla or candidate divisions
Bioanalyser (Agilent Technologies, Santa Clara, CA, USA).
(see also the Appendix Table). The vast majority of sequences
PCR amplification of the 16S rDNA hypervariable
(99.6%) belonged to one of the seven major phyla: Actinobacteria,
V6-region was performed with 2 forward primers—(909Fw)
Bacteroides, Firmicutes, Fusobacteria, Proteobacteria,
gcctccctcgcgccatcag-AAACTYAAARRAATTGACGG and
Spirochetes, or candidate division TM7 (Fig. 1). Of the major
(917Fw) gcctccctcgcgccatcag-GAATTGACGGGGRCCCGCA—
phyla (Fig. 1), Actinobacteria, Fusobacteria, and Spirochetes were
and 1 reverse primer—(1061Rv) gccttgccagcccgctcag-
overrepresented in plaque, while Bacteroides, Firmicutes, and
TCACGRCACGAGCTGACGAC. These primers included the
Proteobacteria sequences were more abundant in saliva.
454 Life Sciences (Branford, CT, USA) Adapter A (for forward
At the genus level, sequences from saliva and plaque
primers) and B (for reverse primers) fused to the 59 end of the
represented 318 different genera. From these, saliva comprised
SSU-rRNA bacterial primer sequence (upper case). Primers
185 genera, plaque, 267. About 11% of all sequences could
were designed to provide the best combination between a high
not be identified at the genus level and were classified at the
phylogenetic recovery and limited distance to the V6 hypervariable
next highest possible resolution level—family, order, class, or
region (Baker et al., 2003; Horz et al., 2005). The PCR amplicon
phylum. Of all sequences, 97% corresponded to 31 genera and
library was created for each individual DNA sample (71 saliva
14 other taxa (Table 2, marked with *) in plaque and saliva.
and 98 plaque samples). The amplification mix contained 2 units
Three genera (Prevotella, Streptococcus, and Veillonella)
of Goldstar DNA polymerase (Eurogentec, Liège, Belgium) and
constituted about 50% of the salivary microflora. In plaque,
Goldstar polymerase (Eurogentec) with 2.5 mM MgCl2.
50% of all sequences was occupied by 6 genera: Streptococcus,
After denaturation (94°C; 2 min), 30 cycles were performed
Veillonella, Corynebacterium, Actinomyces, Fusobacterium,
that consisted of denaturation (94°C; 30 sec), annealing (50°C; 40
and Rothia.
sec), and extension (72°C; 80 sec). DNA was isolated by means of
At the species level, defined as OTUs at 3% difference, about
the MinElute kit (Qiagen, Hilden, Germany), visualized with the
5600 and 10,000 phylotypes were found in saliva and plaque,
Agilent 2100 Bioanalyser, and quantified (Nanodrop ND-1000).
respectively, while a more conservative approach—6% difference
Individual amplicon libraries were pooled in equimolar amounts and
among OTUs—yielded 3600 and 6800 phylotypes (Table 1). The
sequenced unidirectionally in the reverse direction (B-adaptor) by
most diverse genus in saliva was the genus Prevotella, while in
means of the Genome Sequencer 20 (GS-20) system (Roche, Basel,
plaque it was the genus Streptococcus (Table 2).
Switzerland) at 454 Life Sciences. Sequences are available in the
GenBank Short Read Archive, accession number
SRA001159 (ftp://ftp.ncbi.nih.gov/pub/TraceDB/ Table1. Sequencing Information and Diversity Estimates for the Bacteria in Saliva and Plaque
ShortRead/SRA001159).
Saliva Plaque
Data Analysis
GS-20 sequencing data were processed as Total number of sequences generated 100,537 172,659
previously described (Sogin et al., 2006). In Total number of V6 sequences passeda 73,485 124,188
brief, we trimmed sequences by removing primer
Total number of unique V6 tag sequences 10,754 18,224
sequences and low-quality data, sequences that did
Total OTUs at 3% difference (phylotypes) 5669 10,052
not have an exact match to the reverse primer that
Total OTUs at 6% difference (phylotypes) 3621 6888
had an ambiguous base call (N) in the sequence,
Chao1 estimator of richness at 3% difference 12,611 26,204
or that were shorter than 50 nt after trimming. We
(95% CI) (11,940 to 13,355) (25,023 to 27,477)
then calculated the percent difference between
each unique tag sequence and the closest match ACE estimator of richness at 3% difference 12,650 27,266
in a database of 41,460 unique eubacterial and (95% CI) (12,365 to 12,946) (26,720 to 27,830)
1911 unique archeal V6 sequences, representing Chao1 estimator of richness at 6% difference 8480 18,922
141,014 SSU rRNA sequences from the Silva (95% CI) (7881 to 9163) (17,850 to 20,100)
database (Pruesse et al., 2007). Taxa were assigned ACE estimator of richness at 6% difference 8315 19,875
to each full-length reference sequence according to (95% CI) (8073 to 8569) (19,361 to 20,410)
the Ribosomal Database Project Classifier (Cole
a Trimmed reads that passed quality control as described in Sogin et al. (2006).
et al., 2005). In cases where tags were equidistant
1018 Keijser et al. J Dent Res 87(11) 2008
Figure 1. Relative abundance of the main phyla identified in saliva
(dark bars) and plaque (light bars). Only phyla with a relative
abundance greater than 0.4% are shown. These 7 predominant phyla
together account for > 99% of sequences identified. Total numbers of
sequences were N = 73,485 and N = 124,188 in saliva and in plaque,
respectively.
From all sequences, from 0.7 to 2.6% (depending on a
similarity distance used) can be considered new, previously
unidentified, phylotypes: 5305 sequences differed more than
10% from the reference sequence in the database, while 1328
sequences differed by more than 30% from the nearest reference.
The richness of total bacterial communities of saliva and
plaque was estimated by rarefaction analysis. The shapes of
the rarefaction curves (Figs. 2A, 2B) indicate that bacterial
richness of the sampled saliva and plaque is not yet complete.
Depending on the cut-off used in sequence differences
between OTUs (6% or 3%), the estimates of the richness of
total bacterial communities ranged between 8480 and 12,650
phylotypes in saliva and from 18,922 to 26,202 phylotypes in
plaque (Table 1, Fig. 2).
The relative abundance of the phylotypes indicates that
the majority of OTUs is present at low abundance (Figs. 2C,
2D). The 1000 most abundant OTUs account for 90-95% of all
sequences.
Discussion
An understanding of the composition of the oral microbial
ecosystem in relation to oral health is essential for the prevention
and treatment of oral diseases. The introduction of high-throughput
pyrosequencing has dramatically increased the resolution at which
microbial communities can be analyzed. The main drawback
of pyrosequencing is the current length restriction of sequences
obtained: 100nt for the GS-20 and 250nt for the current GS-FLX.
As a consequence, the method cannot be used to produce full-
length 16s rDNA sequences, traditionally used for microbial
taxonomic studies. The method does allow for the analysis of small
hypervariable regions of the 16s rDNA gene, as has been used in
several studies (Sogin et al., 2006; Huber et al., 2007; Sundquist
et al., 2007). This approach, however, provides a challenge to
the accurate assignment of the bacterial taxonomic groups and
to estimation of the microbial richness on the basis of these short
sequence tags. First studies have demonstrated the feasibility
of this approach (Liu et al., 2007; Sundquist et al., 2007). Tag
sequencing has been shown to provide sufficient resolution to
explore microbial communities, and conclusions obtained from
full-length sequences could be recaptured. However, the taxonomy
resolution of the sequence tags obtained may vary between/among
the different taxonomic groups.
We have used 454 GS-20 pyrosequencing to explore the
composition of the oral microbial flora by targeting the 16S
rDNA hypervariable V6-region. This region has been used in
several similar studies (Huber et al., 2007; Sogin et al., 2006)
and has the advantage that the hypervariable part is flanked
directly by well-conserved regions that can be used for PCR
amplification (Baker et al., 2003). To define the membership
of the tags found in species-level phylotypes (OTU), we used a
cut-off of 3% difference, facilitating a direct comparison with
other studies, as well as a more conservative cut-off of 6%.
At a 6% cut-off, our studies revealed 3621 and 6888 species-
level phylotypes in saliva and plaque samples, respectively.
This number is far greater than the current estimates of oral
microbial diversity. Analysis of our data further indicates that
the total microbial species richness is in the order of 19,000-
26,000, depending on the cut-off chosen (6% or 3%).
So far, there are no other metagenomic studies on human
microbial flora using pyrosequencing or species-richness
estimates available for comparison. Our results on human
commensal oral microbial flora showed lower diversity than in
the deep marine biosphere, where about 37,000 bacterial and
about 3000 archaeal phylotypes have been estimated at 3%
difference (Huber et al., 2007), while soil samples showed a
somewhat lower diversity (from 5000 to 20,000 OTUs at 3%
difference; Roesch et al., 2007) compared with the oral flora.
The vast majority of the unexpectedly large number of
species-level phylotypes is present at very low levels. This
is illustrated by the fact that of the 3600 OTUs identified
in saliva, or the 6800 OTUs identified in plaque, the 1000
most common OTUs account for 95% of all sequences. Five
percent of sequences thus represented the majority of OTUs
identified in saliva or plaque, indicative of a low abundance. In
addition, estimations of species richness in the oral microbiota
suggest that our data covered approximately 50% of the full
microbiota. Additional species are likely to be present at even
lower abundance. At this stage, it is hard to predict which
role bacteria, present in low numbers, play in oral ecology.
However, it is an oversimplification to neglect their presence.
The taxonomic distribution of our metagenomic data
at the phylum level is in general agreement with previous
findings (Munson et al., 2002; Aas et al., 2005; Preza et
al., 2008). Firmicutes (genus Streptococcus and Veillonella)
and Bacteroidetes (genus Prevotella) were the predominant
phyla in saliva, while Firmicutes and Actinobacteria (genus
Corynebacterium and Actinomyces) dominated supragingival
plaque. The short-read length of 454 pyrosequencing data
and the incomplete nature of 16S rDNA databases limit the
phylogenetic resolution (Sundquist et al., 2007). In our data,
89% of the sequences could be identified at the genus level.
Being aware of the low taxonomic resolution below the genus
level, we did not attempt to identify the individual species,
and thus report here only the abundance of OTUs at 3% and
6% differences. Based on full-length 16S rDNA sequences,
a resolution of 3% OTUs is considered to be the microbial
J Dent Res 87(11) 2008 Pyrosequencing of Commensal Oral Microflora 1019
species level (Acinas et al., 2004),
but is also believed to underestimate
the true number of species (Pedrós-
Alió, 2006). One percent cut-off levels
have also been used (Munson et al.,
2004). It has been demonstrated that
short pyrosequencing reads suffice
for accurate microbial community
analysis (Liu et al., 2007), but how
sequence diversity of full-length 16S
rDNA sequences relates to that of the
hypervariable regions and how this
would reflect taxonomic boundaries
remains to be determined.
Interpretation of the relative
abundance of different taxa identified
by any molecular method must be done
with care. The steps of DNA extraction,
PCR amplification (choice of primers),
and intrinsic differences in 16S rDNA
copy number can/may result in a skewed
taxonomic interpretation of the microbial
community, and probably result in an
underestimation of microbial diversity
(Horz et al., 2005). Measures used
in this study to avoid technical errors
included UV irradiation of solutions and
an efficient DNA extraction protocol
(Keijser et al., 2007). Implementation
of a stringent quality control of the
sequences further excluded potential
sources for artifacts (Huse et al., 2007).
With the increasing read length of the
454 pyrosequencing technology, other
hypervariable regions, such as the
V3 and V4, become amenable to the
technique. These improvements are
likely to provide an even more complete
view of the composition of microbial
communities.
By applying this novel approach
in oral microbial sample analysis, we
created a radically new insight into the
richness of the diversity of commensal
human oral microflora of both saliva
and plaque. We detected diversity at
least one order of magnitude higher
than previously described and estimated
that more than 19,000 species-level
phylotypes contribute to the ultimate
oral species diversity. This new insight
provides a challenge for taxonomists
and ecologists to determine the impact
of this enormous microbial richness
in relation to the physiology of the
oral microbiota of individuals, and
in relation to the oral health-disease
equilibrium.
ACKNOWLEDGMENTS
We thank Mieke Havekes and Louise Nederhoff for helpful
Table 2. Relative Abundance of Top 45 Genera, Covering 97% of Total Sequences, and the Number
of Species-level Phylotypes per Genus in Saliva and Plaque Pools
Relative Amount (%) Number of Phylotypesb
of Total Saliva Plaque
Phylum Genusa  Saliva Plaque  3% 6%  3% 6%
Firmicutes Streptococcus 19.68 14.69 842 540 1038 705
Firmicutes Veillonella 12.60 9.24 376 194 432 250
Actinobacteria Corynebacterium 0.36 8.74 33 26 443 296
Actinobacteria Actinomyces 2.48 7.57 138 88 700 445
Fusobacteria Fusobacterium 1.78 6.60 113 58 577 334
Actinobacteria Rothia 2.88 6.43 141 78 366 223
Bacteroidetes Prevotella 19.70 5.92 956 531 637 355
Bacteroidetes Capnocytophaga 0.52 5.52 54 33 629 416
Proteobacteria Neisseria 8.18 3.87 425 272 348 224
Proteobacteria Haemophilus 3.73 2.76 122 91 109 84
Fusobacteria Leptotrichia 1.09 2.28 87 57 274 201
Proteobacteria Campylobacter 0.78 2.15 28 13 114 62
TM7 TM7_genera_incertae_sedis 1.88 1.86 149 89 308 195
Firmicutes Selenomonas 0.63 1.79 53 36 197 109
Bacteroidetes Porphyromonas 3.37 1.32 164 85 197 105
Proteobacteria fam. Pasteurellaceae* 1.79 1.30 101 71 117 83
Proteobacteria Kingella 0.08 1.27 14 9 78 45
Firmicutes order Lactobacillales* 1.97 1.12 69 37 89 56
Proteobacteria Cardiobacterium 0.04 1.08 3 2 83 57
Firmicutes Gemella 1.49 1.03 86 45 101 65
Spirochaetes Treponema 0.21 0.86 36 26 118 92
Proteobacteria class Gammaproteobacteria* 4.01 0.84 156 123 125 100
Actinobacteria fam. Propionibacteriaceae* 0.02 0.73 3 2 72 38
Proteobacteria order Burkholderiales* 0.26 0.66 15 13 71 53
Proteobacteria fam. Neisseriaceae* 0.22 0.61 52 40 140 112
Firmicutes Parvimonas 0.15 0.59 8 4 47 26
Firmicutes fam. Lachnospiraceae* 0.11 0.54 25 16 93 61
Actinobacteria Actinobaculum 0.01 0.54 1 1 26 17
Firmicutes order Clostridiales* 0.61 0.52 104 86 180 155
Bacteroidetes Paludibacter 0.04 0.51 3 2 38 19
Firmicutes Dialister 0.09 0.47 14 12 47 29
Bacteroidetes order Bacteroidales* 2.63 0.42 164 84 74 57
Bacteroidetes fam. Flavobacteriaceae* 0.19 0.39 14 9 38 27
Firmicutes Eubacterium 0.15 0.39 25 18 62 41
Proteobacteria Derxia 0.06 0.38 8 7 46 31
Bacteroidetes Tannerella 0.09 0.36 13 9 46 25
Bacteroidetes fam. Prevotellaceae* 0.10 0.28 17 11 50 33
Firmicutes Abiotrophia 0.15 0.22 5 4 25 17
Firmicutes fam. Acidaminococcaceae* 0.24 0.20 27 20 26 22
Actinobacteria fam. Coriobacteriaceae* 0.03 0.20 3 3 21 14
Actinobacteria fam. Bifidobacteriaceae* 0.05 0.19 7 4 24 17
Proteobacteria Aggregatibacter 0.18 0.16 17 12 24 20
Proteobacteria Actinobacillus 0.52 0.16 24 16 20 15
Actinobacteria Propionibacterium 0.02 0.15 5 2 44 29
Firmicutes Oribacterium 0.65 0.13 54 30 34 25
a Taxa marked with asterisk could not be assigned to any of the genera and are shown at the
lowest common taxon: family (fam.), order, class, or phylum.
b The species-level phylotypes were defined as OTUs at 3% difference from other OTUs (Acinas et
al., 2004) and by a more conservative approach: 6% difference between OTUs was needed to
belong to different species.
discussions and excellent technical assistance. Sue Huse was
supported on a subcontract to Mitchell L. Sogin from the Woods
Hole Center for Oceans and Human Health, funded by the National
1020 Keijser et al. J Dent Res 87(11) 2008

Microbiol 8:43.
Horz HP, Vianna ME, Gomes BP,
Conrads G (2005). Evaluation of universal
probes and primer sets for assessing
total bacterial load in clinical samples:
general implications and practical use in
endodontic antimicrobial therapy. J Clin
Microbiol 43:5332-5337.
Huber JA, Mark Welch DB, Morrison HG,
Huse SM, Neal PR, Butterfield DA, et al.
(2007). Microbial population structures in
the deep marine biosphere. Science 318:97-
100.
Huse SM, Huber JA, Morrison HG, Sogin
ML, Mark Welch D (2007). Accuracy
and quality of massively parallel DNA
pyrosequencing. Genome Biol 8: R143.
Keijser BJ, ter Beek A, Rauwerda H,
Schuren F, Montijn R, van der Spek H,
et al. (2007). Analysis of temporal gene
expression during Bacillus subtilis spore
germination and outgrowth. J Bacteriol
189:3624-3634.
Liu Z, Lozupone C, Hamady M,
Bushman FD, Knight R (2007). Short
pyrosequencing reads suffice for accurate
microbial community analysis. Nucleic
Acids Res 35:e120.
Munson MA, Pitt-Ford T, Chong B,
Weightman A, Wade WG (2002).
Molecular and cultural analysis of the
microflora associated with endodontic
infections. J Dent Res 81:761-766; errata
in J Dent Res 82:69, 2003, and J Dent Res
Figure 2. Plots of the numbers of different operational taxonomic units (OTUs) in saliva (A) and plaque 82:247, 2003.
(B) as a function of the number of sequences sampled, also known as rarefaction curves and the relative Munson MA, Banerjee A, Watson TF,
abundance of the OTUs in saliva (C) and plaque (D). The 0% curve is based on all unique sequences (N Wade WG (2004). Molecular analysis of
= 10,754 in saliva, N = 18,244 in plaque); the 3%, 6%, and 10% curves contain OTUs with differences the microflora associated with dental caries.
that do not exceed 3%, 6%, or 10%, respectively.  The steepness of the curves (A,B) indicates that a J Clin Microbiol 42:3023-3029.
large fraction of the species diversity  has not yet been sampled. (C,D) The relative abundance of the Paster BJ, Olsen I, Aas JA, Dewhirst FE
OTUs in saliva (C) and plaque (D). The x-axis indicates the individual OTUs, ranked according to their (2006). The breadth of bacterial diversity
relative abundance (high to low). The y-axis indicates the cumulative abundance of the OTUs. As can be in the human periodontal pocket and other
observed, for plaque as well as saliva, the 1000 most abundant OTUs represent 90-95% of all sequences. oral sites. Periodontol 2000 42:80-87.
Pedrós-Alió C (2006). Marine microbial
diversity: can it be determined? Trends
Microbiol 14:257-263.
Preza D, Olsen I, Aas JA, Willumsen T,
Grinde B, Paster BJ (2008). Bacterial profiles of root caries in elderly
Institutes of Health and National Science Foundation (NIH/NIEHS patients. J Clin Microbiol 46:2015-2021.
1 P50 ES012742-01 and NSF/OCE 0430724). We also thank the Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig W, Peplies J, et al. (2007).
ACTA Research Institute for financial support. SILVA: a comprehensive online resource for quality checked and aligned
ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res
35:7188-7196.
References Rajilić-Stojanović M, Smidt H, de Vos WM (2007). Diversity of the human
Aas JA, Paster BJ, Stokes LN, Olsen I, Dewhirst FE (2005). Defining the gastrointestinal tract microbiota revisited. Environ Microbiol 9:2125-2136.
normal bacterial flora of the oral cavity. J Clin Microbiol 43:5721-5732. Riggio MP, Lennon A, Rolph HJ, Hodge PJ, Donaldson A, Maxwell AJ, et
Aas JA, Griffen AL, Dardis SR, Lee AM, Olsen I, Dewhirst FE, et al. (2008). al. (2008). Molecular identification of bacteria on the tongue dorsum of
Bacteria of dental caries in primary and permanent teeth in children and subjects with and without halitosis. Oral Dis 14:251-258.
young adults. J Clin Microbiol 46:1407-1417. Roesch LF, Fulthorpe RR, Riva A, Casella G, Hadwin AK, Kent AD, et al.
Acinas SG, Klepac-Ceraj V, Hunt DE, Pharino C, Ceraj I, Distel DL, et al. (2007). Pyrosequencing enumerates and contrasts soil microbial diversity.
(2004). Fine-scale phylogenetic architecture of a complex bacterial ISME J 1:283-290.
community. Nature 430:551-554. Ronaghi M, Uhlén M, Nyrén P (1998). A sequencing method based on real-
Baker GC, Smith JJ, Cowan DA (2003). Review and re-analysis of domain- time pyrophosphate. Science 281:363, 365.
specific 16S primers. J Microbiol Methods 55:541-555. Schloss PD, Handelsman J (2005). Introducing DOTUR, a computer program
Boutaga K, Savelkoul PH, Winkel EG, van Winkelhoff AJ (2007). for defining operational taxonomic units and estimating species richness.
Comparison of subgingival bacterial sampling with oral lavage for Appl Environ Microbiol 71:1501-1506.
detection and quantification of periodontal pathogens by real-time poly­ Sogin ML, Morrison HG, Huber JA, Mark Welch D, Huse SM, Neal PR, et al.
mer­ase chain reaction. J Periodontol 78:79-86. (2006). Microbial diversity in the deep sea and the underexplored “rare
Cole JR, Chai B, Farris RJ, Wang Q, Kulam SA, McGarrell DM, et al. biosphere”. Proc Natl Acad Sci USA 103:12115-12120.
(2005). The Ribosomal Database Project (RDP-II): sequences and tools Sundquist A, Bigdeli S, Jalili R, Druzin ML, Waller S, Pullen KM, et al.
for high-throughput rRNA analysis. Nucleic Acids Res 33:D294-D296. (2007). Bacterial flora-typing with targeted, chip-based pyrosequencing.
Dowd SE, Sun Y, Secor PR, Rhoads DD, Wolcott BM, James GA, BMC Microbiol 7:108.
et al. (2008). Survey of bacterial diversity in chronic wounds using Von Bubnoff A (2008). Next-generation sequencing: the race is on. Cell
pyrosequencing, DGGE, and full ribosome shotgun sequencing. BMC 132:721-723.