International Dental Journal 2011; 61 (Suppl.

1): 1122
ORIGINAL ARTICLE
doi: 10.1111/j.1875-595X.2011.00025.x

Dental caries pathogenicity: a genomic and metagenomic

perspective
Scott N. Peterson1, Erik Snesrud1, Nicholas J. Schork2 and Walter A. Bretz3
1
The J. Craig Venter Institute, Rockville MD 20886; 2The Scripps Translational Science Institute & Scripps Health, La Jolla CA 92037;
3
Department of Cariology & Comprehensive Care, New York University, College of Dentistry, New York, USA

ABSTRACT
In this review we address the subject of dental caries pathogenicity from a genomic and metagenomic perspective. The
application of genomic technologies is certain to yield novel insights into the relationship between the bacterial flora, dental
health and disease. Three primary attributes of bacterial species are thought to have direct impact on caries development,
these include: adherence on tooth surfaces (biofilm formation), acid production and acid tolerance. Attempts to define the
specific aetiological agents of dental caries have proven to be elusive, supporting the notion that caries aetiology is perhaps
complex and multi-faceted. The recently introduced Human Microbiome Project (HMP) that endeavors to characterise the
micro-organisms living in and on the human body is likely to shed new light on these questions and improve our
understanding of polymicrobial disease, microbial ecology in the oral cavity and provide new avenues for therapeutic and
molecular diagnostics developments.
Key words: Caries, biofilm, bacterial species, genomic, metagenomic, Human Microbiome Project

Dental caries is the single most common disease in
childhood with a prevalence rate five times higher than
the next most prevalent disease, asthma1. This disease
presents early in childhood where very aggressive forms
of dental decay affect the primary teeth of infants and
young children. A variety of factors, including
microbial, genetic, immunological, behavioural and
environmental, interact to contribute to dental caries
development. Over 50% of children in the United States
ages 59 years have at least one cavity or filling. This
number increases to 78% by the time children reach
adulthood (17 years of age). These adverse outcomes
are disproportionately present in disadvantaged seg-
ments of the US population including children, adults
and the elderly. Childhood caries significantly contrib-
utes to the burden of pain and is associated with a
marked decrease in quality of life and general health2,3.
Dental expenditures for treatment of childhood caries
are substantial and may involve extensive curative
(restorative) procedures under general anaesthesia. It
has been shown that dental care is the most common
unmet need among children in the United States2,3.
Unlike many disease conditions that require surrogate
samples to establish causality or that require long
periods of time to develop (e.g. periodontal diseases),
2011 FDI World Dental Federation
dental caries and the associated microbial flora can be
readily tracked over time in longitudinal studies. This is
an important feature in the context of the dental biofilm
microbiome since the changes occurring in the microb-
iome over time can be more readily associated with
caries onset and progression.
DENTAL CARIES AETIOLOGY
Micro-organisms and complex microbial communities
are responsible for the majority of biochemical trans-
formations in the oral cavity. Despite estimates that
approximately half of these micro-organisms are culti-
vable, we have a very limited understanding of the
metabolic capabilities of most species comprising the
oral microbial communities. Cariogenicity is mainly
promoted by three bacterial phenotypes including:
Adhesion, or effective competiveness in the biofilm
environment
Acid production
Acid tolerance4.
There is substantial evidence that Streptococcus
mutans is associated with the onset and development
of dental decay5. S. sobrinus has been implicated in
caries development particularly in instances where
11
Peterson et al.
caries development appears to be independent of S.
mutans. Interestingly, S. sobrinus displays higher acid
production and acid tolerance compared to S. mutans.
Its representation in dental biofilm however appears
more variable suggesting that it is less fit in the biofilm
environment compared to S. mutans. Additional acido-
genic and aciduric bacteria include Actinomyces spp.
and Bifidobacterium spp.. Traditional culture-based
methods have shown that S. mutans is the chief
pathogen for dental caries initiation1,510. Various
laboratories have reported the co-cultivation of S.
anginosus, S. constellatus, S. parasanguinis together
with S. mutans on MSB media formulations, thus
complicating the interpretation of a large body of
microbiologically derived evidence concerning
S. mutans as the primary aetiological agent of caries
development. Lactobacillus species have also been
implicated in caries progression and are often found
in advancing caries lesions1,5,10. Additional evidence
has linked the Actinomyces to root surface caries onset
and development11.
There are relatively few studies that have employed
culture-independent techniques to study dental caries-
associated flora. Surveys of microbial diversity in the
human oral microbiome based on the sequencing and
phylogenetic analysis have estimated that more than
500 microbial species inhabit the oral cavity69. Becker
and colleagues12 and Aas et al.,13 used 16S rDNA
analysis of caries-free and caries-active children and
adults, revealing species associated with dental health
such as S. sanguinis and genera associated with dental
caries in both primary and permanent dentitions such
as Streptococcus spp. including S. mutans and low-pH
non-S. mutans streptococci, Veillonela spp., Actinomy-
ces spp., Bifidobacterium spp., Lactobacillus spp.,
Propionibacterium spp., and Atopobium spp.. Munson
et al.14 and Chhour et al.15 reported that the predom-
inant microbes in a small number of adults with
advancing, deep caries lesions were S. mutans and
Lactobacillus spp. but also included the genera
Prevotella, Selenomonas, Dialister, Fusobacterium,
Bifidobacterium, and Pseudoramibacter. Corby and
colleagues conducted the largest study employing 16S
rDNA profiling of dental caries-associated flora in
infants and children who were caries-active (n = 86)
and caries-free (n = 118)16. S. mutans, Actinomyces
spp., and Lactobacillus spp. were over-abundant in
caries-active subjects whereas, beneficial species asso-
ciated with dental health included S. parasanguinis, A.
defectiva, S. mitis, S. oralis, and S. sanguinis. Taken
together, the number of species identified in these
studies ranged from 75 to 197.
Based on these previous efforts, a conceptual frame-
work has been established suggesting that dental
biofilm contains both health-associated and disease-
associated flora. While these association studies have
12
yielded significant insights, our inability to clearly relate
the causative agents of dental health and disease and
the reasons behind this must be carefully considered. As
scientists and clinicians we tend to think in binary terms
when forming cause and effect relationships. The
problem of dental decay is likely to be significantly
more complex. It may be that many species in specific
combinations will dictate the signatures associated with
dental health and disease or the transition between the
two states. Alternatively, the problem may be more
complex such that groups of related bacterial species
may be thought of as inter-changeable components in
defining caries signatures. From a microbial genomics
perspective this possibility appears reasonable since the
genomes of related bacterial species share a larger
number of genes functions than do more distantly
related species. More complicated still, and our current
working hypothesis, is that the distinction between
health and disease lies primarily at the level of the strain
selection. This topic is covered in greater detail below.
BIOFILM FORMATION
The micro-organisms that comprise the human dental
biofilm co-aggregate by a complex and dynamic process
to form a multi-species community, biofilm. On hard
surfaces (teeth), biofilms are typically several cell layers
thick and can measure up to several hundred microm-
eters7. There is much that we do not understand about
the factors governing biofilm formation, it is clear that
the process is complex and involves contributions made
by a very large number of bacterial species. Biofilm
formation on tooth surfaces is initiated by what are
known as pioneer colonisers comprised mainly of
mitis group members such as S. gordonii, S. sanguinis
and S. mitis. The establishment of these species allows
the early colonisers including S. mutans, Veillonella
spp. and bridging function provided by Fusobacteria
spp. to co-colonise the biofilm. Maturation of the
biofilm involves addition of late colonisers comprised
primarily of anaerobic, gram-negative bacteria1719.
High cell densities serve as a signal for several species to
undergo competence development and production of
bacteriocins. Cell lysis resulting from bacteriocin pro-
duction has been conjectured to provide genomic DNAs
to the biofilm environment that may enhance DNA
exchange and recombination of foreign DNAs. The
metabolic contributions and the role of other abundant
and even low abundance species in the development
and maintenance of biofilm are relatively unknown.
The short-chain acids (butyrate, propionate, lactate,
acetate, formate) produced by the dominant species via
sugar metabolism, undoubtedly serve as substrates
(carbon sources) for other metabolically specialised
species. The nature of this metabolism and the species
responsible for it are poorly understood, however it is
2011 FDI World Dental Federation

likely that these activities contribute to the overall
regulation of pH in dental biofilm and are therefore
important to identify and characterise. We speculate
that these activities will be associated with species of
moderate and low abundance, since they are likely to be
out-competed by the primary metabolisers.
The bacteria in dental biofilms experience dramatic
fluctuations in nutrient availability, pH, oxygen tension
and osmolarity4. The catabolism of sugars creates
specific stressors to cells co-habitating dental biofilms
including: acidic pH and reactive oxygen species (ROS)
that serve to damage DNA, proteins and other macro-
molecules in the cell. Since acid production is synon-
ymous with the virulence of organisms like S. mutans,
these phenotypes have evolved to be coupled to a
variety of stress responses. A number of studies have
made links between a variety of stress responses to
biofilm formation, competence development and acid
tolerance. The regulation and co-regulation of these
complex systems appears to be accomplished by a
relatively small set of two-component systems (TCSs).
TCSs operate through the concerted action of a sensor
kinase that senses the bacterial environment and
subsequently undergoes an activation via auto-phos-
phorylation. The sensor kinase then transfers this
phosphate group to and activates the second compo-
nent of the system, the response regulator that then
modulates gene expression at one or more promoter
sites. For example in S. mutans the TCS, VicRK is
induced under conditions of oxidative stress whereas
ComDE, another TCS involved in competence induc-
tion is also activated. Deletion mutants in VicRK
display increased sensitivity to H2O220 and reduced
competence for DNA uptake. A serine threonine kinase
and phosphatase encoded by pknB and pppL respec-
tively appears to serve as a regulatory switch that links
oxidative stress and competence21. Deletion mutant
strains of pknB result in reduced competence develop-
ment, biofilm formation, cell morphology, growth rate,
oxidative stress resistance, bacteriocin expression and
acid tolerance. The specific genes displaying reduced
expression overlap substantially with those affected by
a deletion of the VicRK TCS, supporting a link between
these regulators. Yet another example linking biofilm
formation to competence and stress response comes
from characterisation of the S. mutans, SloR protein
that is known to impact biofilm formation, competence
development and resistance to oxidative stress in a
Mn2+ manner22. The long-studied stringent response
induced by nutrient (amino acid) starvation, controlled
by relRS also increases the expression of stress response
genes23. These examples illustrate a variety of regula-
tors that appear to serve as environmental sensors of
stress, whether oxidative, pH, or nutrient availability
that in turn reprogramme the gene expression patterns
of cells growing within the biofilm environment.
2011 FDI World Dental Federation
Dental caries pathogenicity
DENTAL BIOFILM METABOLISM, ACID
PRODUCTION AND ACID TOLERANCE
Sucrose is considered the most cariogenic carbohydrate
because of its correlation with the production of
organic acids and further association with the demin-
eralisation of the tooth surface11. The production of
acid impacts the resident microbial populations. The
availability of dietary carbohydrates is key to biofilm
initiation and development24. The dominance of the
Streptococci and other members of the Firmicutes,
dictate that the overall biochemical flux occurring in
the dental biofilm microbiome are dominated by the
metabolic and catabolic potentials encoded by these
genomes. Comparative analysis of the Streptococci
genomes reveals an expanded repertoire of PTS perm-
eases and ABC transporters devoted to the transport of
a wide variety of sugars. S. mutans is able to metabolise
a number of sugars and glycosides such as glucose,
fructose, sucrose, lactose, galactose, mannose, cellobi-
ose, glucosides, trehalose, maltose and a previously
unrecognised, group of sugar-alcohols25. The presence
of enzymes in the non-oxidative branch of the pentose
phosphate pathway along with a pentose PTS trans-
porter suggests an ability to metabolise various pentos-
es. S. mutans produces a number of fermentation
products (lactate, acetate, formate and ethanol) that
contribute to the acidogenic potential of the local
environment.
The adaptation to changing pH has most likely been
a significant driving force of the evolution that defines
the dental biofilm microbial community. The acid
tolerance of S. mutans and other aciduric species is
based in part to the activity of the membrane-bound,
acid-stable, proton-translocating F0F1 ATPase that
maintains intracellular pH at 7.526. Many Streptococ-
cal spp. including S. gordonii possess an arginine
deiminase system (ADS) allowing the hydrolysis of
arginine to ornithine, ammonia and C02 plus ATP27.
The genes responsible for this activity are substrate
inducible and display catabolite repression. This met-
abolic system is regulated by the TCS CiaRH and
ComDE in a reduced pH-dependent manner, defining
yet another link between acid tolerance and compe-
tence development. These deacidifying activities have
been associated with reduced caries development28,29.
The major cariogenic acid producers tend to display
increased tolerance to reduced pH and differ from acid
producers that lower pH but also act as arginolytics30.
Microbial metabolism of nitrogenous substrates can
also affect pH and have been attributed to urease
activity that serve to elevate pH30. Carbon dioxide
production represents another mechanism employed by
S. mutans to cope with acid stress and involves its
ability to convert L-malate to L-lactate plus C02 and
ATP, that contributes to deacidifying the cytosol and
13
Peterson et al.
provides energy for H+ translocation into the extracel-
lular environment via the F0F1 ATPase31. Malolactic
fermentation (MLF) converts di-carboxylic L-malate to
mono-carboxylic acid and C02. This acitivity has also
been shown to increase S. mutans acid tolerance31.
The reduced expression level of chaperones control-
ling proper protein folding, DnaK and GroEL leads to
reduced biofilm formation, acid tolerance and H202
tolerance32. In a low pH growth environment, S.
mutans increases the proportion of monounsaturated
fatty acids in their membranes. The gene responsible
for this lipid biosynthesis, fabM when deleted, results
in an extreme acid sensitivity and reduced cariogenic
potential in a rat model33. This result suggests that
maintaining higher concentrations of H+ outside the
cell is controlled in part by membrane structure. The
commonality between discrete TCS mutants of S.
mutans and their impact on a number of phenotypes
related to biofilm formation, competence development
and acid tolerance suggest functional redundancy of
these systems and a molecular complexity of their
regulation. The large number of genes involved in this
regulatory circuitry further implies the relative impor-
tance of this regulation in the survival and fitness of
the species in its dynamic and stress-filled environ-
ment. Our knowledge of the molecular details of acid
production and tolerance are largely the result of
studies focused on S. mutans. Future efforts will begin
to embrace other species present in dental biofilm. The
particular species to be studied are likely to be
dictated by findings derived from the Human Microb-
iome Project.
THE HUMAN MICROBIOME PROJECT
The NIH recently established a Roadmap initiative
(http://nihroadmap.nih.gov/hmp) referred to as the
Human Microbiome Project (HMP). In its early phases,
this programme promises to characterise the microbial
populations that are associated with 15 to 18 body sites
of 300 healthy human subjects. The programme
initiatives include the sequencing of up to 1,000
reference genomes from both cultured and uncultured
bacteria. Approximately 250 of these genome se-
quences will be of relevance to the oral cavity with an
emphasis on species thought to be related to dental
caries and gingival disease34. While this is a relatively
new programme, oral microbiologists have appreciated
the importance of the microbiome and polymicrobial
disease for a number of years. While the human gut
harbours the largest number of bacteria, it can be
argued that the oral cavity contains the highest diversity
and complexity of microbial species. Critical to our
understanding of the role of the oral microbiome is the
knowledge that members of microbial communities
play key roles not only as putative agents responsible
14
for the onset and progression of oral diseases but also in
maintaining oral health.
THE ECOLOGY OF THE DENTAL BIOFILM
MICROBIOME
Early investigations of the human microbiome are
revisiting well-established predictions made by classical
ecological models. Among the early observations of all
human associated microbiomes is the strong selectivity
of species membership. Among the >70 known bacte-
rial divisions described to date, less than 10% are found
in association with the human environment. The
Actinobacteria, Firmicutes, Proteobacteria, Bacteroide-
tes, Fusobacteria and Spirochaetes appear to comprise
the vast majority of the bacterial divisions that live in
association with humans. The population structure (the
relative abundance and type of bacterial species making
up the microbial population) of the oral cavity is unique
compared to other microbiomes such as the skin, gut,
etc. Likewise, within the oral cavity, each domain
displays varying levels of similarities and differences.
The saliva and dental biofilm microbiomes are rela-
tively similar whereas they are quite distinct from the
microbiome associates with subgingival domains. Com-
mon to all microbiome population structures is the
representation of a relatively small number of phyla,
class, orders, families and genera. It is at the level of the
species and probably strains that we observe a massive
radiation of phylotypes or related species occupying the
same ecological niche (Figure 1). Ecological principles
dictate that natural selection has taken two primary
forms. Top down selection has generated massive
functional redundancy within microbial populations.
This prediction is consistent with the observation stated
above that many tens or hundreds of species belonging
to the same genus occupy the same microbiome. An
important consequence of top down selection and the
generation of functional redundancy in microbiomes is
the fact that presence of any single species in the
Figure 1. Taxonomic Representation of the Dental Biolm Micro-
biome.
2011 FDI World Dental Federation

population is dispensable since many other sibling
species are likely to be capable of filling the gap with
respect to community functions provided. As such it
would be anticipated that the oral microbiomes
associated with human subjects to be substantially
varied with regard to the species present and their
relative abundance in the microbiome. Indeed, this
subject-to-subject variability appears to be borne out by
early experimentation and represents a strong con-
founding effect on conducting traditional association
studies to identify the culprits of dental health and
disease. The second form of selection is bottom up
selection and has served to generate functional special-
ization. Evidence for this prediction is less obvious at
this time but is consistent with the idea that microbial
communities do not simply compete for common
energy sources but also and perhaps more importantly
cooperate to glean the most from the available
resources. Our initial examination of the dental biofilm
microbiome indeed supports the prediction of massive
gene function repertoires provided by an astonishing
number of distinct bacterial species. We remain quite
ignorant as to how these genes function in the context
of a complex ecosystem but these questions will be
actively pursued by a number of investigations in the
coming years.
STREPTOCOCCAL GENOMES AND GENOME
PLASTICITY
Over the past several years many Streptococcal ge-
nomes have been sequenced including: S. pyogenes,
S. pneumoniae, S. agalactiae, S. thermophilus, S. suis,
S. sanguinis, S. gordonii, S. equi and S. mutans (http://
www.genomesonline). Comparative genomic analyses
have made clear that streptococcal genomes have some
unique characteristics. Streptococcal genomes generally
encode approximately 2,0002,300 genes. Approxi-
mately 15% of the genome encodes phosphotranserase
systems (PTS) allowing the import of numerous sugars
and ions into the cell and the export of a variety of
compounds out of the cell. It has become clear that the
genomes of individual strains belonging to the same
species share a large set of genes commonly referred to
as the core genome. The remainder of the genome
referred to as accessory functions are generally consid-
ered non-essential and are variably present in the
genome sequence. The fraction of the genome encoding
accessory functions can be quite large, 1020% of the
total gene complement. S. mutans strain variability has
been studied in detail by Li and Caufield35. It is
conceivable that the gene pool in the form of accessory
genes for many Streptococcal spp. may exceed 20,000
genes! At this time, it is unclear what impact strain-
specific, accessory genes play in dental health and
disease, but given the large gene pools encoded by
2011 FDI World Dental Federation
Dental caries pathogenicity
species, strain-level selection must be considered a
strong possibility.
A comparison of completely sequenced genomes
derived from two S. mutans isolates, UA159 and
NN2025 showed a remarkable degree of gene variation
and conservation. Approximately 90% of the genes
present in strain NN2025 are also present in the UA159
genome36. The gene order among shared genes is
largely conserved, however a number of exceptions
involving over 30 genome rearrangements centered
symmetrically around the origin of replication were
noted. These regions tend to be flanked by phage,
transposons and Insertion Sequence elements (IS-ele-
ments). Acquisition of genes by horizontal exchange is a
major driving force in bacterial evolution, particularly
among the Streptococci, perhaps owing to their natural
competence phenotype. The ability of species present in
dental biofilms to sample genes selected over millions
of years of evolution to enable adaptation to the
dynamic dental biofilm environment is a satisfying
notion since the potential for the acquisition of genetic
traits of value in environmental fitness and stress
related resistance would appear likely and perhaps
commonplace.
S. mutans strain-specific phenotypes and genotypes
related to plasmids37,38, mutacin production39, sero-
type antigens40, competence41,42 sugar utilisation and
acid production have been noted. A comparison of nine
S. mutans isolates by comparative genomic hybridisa-
tion (CGH) using DNA microarrays revealed a total of
385 genes that were differentially present or dramati-
cally divergent in nucleotide sequence compared to
strain UA15943. Just over half of these genes encode
proteins of unknown function thus complicating pre-
dictions concerning their relationship to the caries
phenotype. These variable regions were generally
observed as blocks of co-localised genes. The functional
potential encoded by the strain-variable regions in-
cluded: galactose and sorbose utilisation, fructose and
mannose PTS transporters and DNA modifying en-
zymes. Two large gene clusters known as TnSmu1 and
TnSmu2 were also variably present among isolates.
TnSmu1 encodes a number of hypothetical genes of
unknown function, transposases and IS-elements.
TnSmu2 is a 50Kb genomic island (2% of the genome)
that appears to be acquired from a foreign source as
evidenced by its altered G+C content relative to the
remainder of the genome, encodes a cluster of nonrib-
osomal peptides (NRPs) and polyketide synthases
(PKSs) required for mutacin production. The presence
of this island is variable across S. mutans strains but
notably so too are the specific genes residing within the
cluster. The mechanism that creates these hybrid gene
clusters remains unknown, but suggests that S. mutans
may generate a large number of bacteriocin variants
related to mutacin. The significance of this variation
15
Peterson et al.
remains to be characterized. Deletion of this cluster
results in reduced growth in aerobic conditions and
sensitivity to H202. A similar comparative genome
hybridisation study performed with isolates represent-
ing a limited geographical distribution revealed highly
similar conclusions44. Nakano and colleagues per-
formed MultiLocus Sequence Typing (MLST) of 102
isolates that resulted in an array of 92 distinct sequence
types. This result implies a very large degree of genetic
variability among S. mutans strains within the world-
wide population. We may infer from these studies as
well as others related to non oral species that microbial
genomes have an enormous repertoire of genes encoded
in strain-specific genomic regions. Therefore an impor-
tant question to address is whether strain-variable genes
provide any selective advantage to cells in dental health
or disease.
ANTAGONISM OR MUTUALISM?
Several studies have reported the effect that the growth
or activity of one species has on another. While these
studies generally refer to these activities as examples of
antagonism, it is not clear whether this designation is
appropriate. Tong and colleagues reported that S.
oligofermentans supresses the growth of S. mutans
through the production of H202 via a lactate oxidase
activity that involves L-amino acid oxidase. This
enzyme is able to use several L-amino acids as
substrates45. The authors referred to this activity as
the competitors main weapon to gain a competitive
advantage over S. mutans and as an example of a
counter-offensive strategy used during chemical war-
fare within the oral microbial community. While at
surface level this description may appear reasonable, it
is perhaps incorrect. Other examples of mutual-
ism antagonism include observations that S. salivarius
can strongly inhibit S. mutans biofilm formation via
inactivation of the competence stimulating peptide
(CSP). This inhibition is dependent on the S. mutans
girA gene product that encodes a bacitracin exporter46.
Yet another example involves S. mutans production of
mutacin resulting in the killing of S. mitis group
members. S. mitis group members are known to
produce H202 as a means of inhibiting S. mutans
growth. The complexity of the dental biofilm necessi-
tates the cooperative behaviour of its membership.
Complex mutualistic relationships may contribute to an
overall ecological homeostasis that, in some instances,
may be disrupted by a small number of cheaters whose
change in abundance and functional activities may
directly or indirectly impact other species in a non-
mutualistic manner. S. mutans may indeed be a cheater,
but the species that exhibit growth inhibitory activities
toward S. mutans may be more appropriately thought
of as maintaining biofilm ecological homeostasis. The
16
increased abundance and rampant, uncontrolled
growth of acidogenic species would no doubt do
serious harm to the biofilm population, reducing its
complexity to a few aciduric species that may display
significantly reduced overall fitness. The fact that this
outcome is not observed suggests that apparent antag-
onism may instead reflect bacterial mechanisms for
maintaining a stable microbiome in the face of a
dynamic and sometimes hostile environment.
MICROBIOME POPULATION STRUCTURE AND ITS
CONSEQUENCES FOR COMPREHENSIVE
CHARACTERISATION
The population structure of the oral microbiome or any
of its sub-domains represents a strong barrier to its
comprehensive characterisation. Most, if not all human
microbiomes appear to have similar characteristics in
that commonly one or a few microbial species are
dominant in the population, representing 510% of the
total microbial population. The next most prevalent
species in the population are also innumerous (510
species) representing <1% of the microbial population.
Within this diversity, many hundreds of species are
present at an abundance of 0.1% and many more
hundreds of species are present at an abundance of
0.01% of the total microbial population (Figure 2). The
biological significance of the microbiome population
structure is not yet clear but suggests that microbiomes
possess an inherent flexibility by virtue of harbouring a
very large and diverse spectrum of species in the form of
a low abundance reservoir. It may be safely assumed
that this reservoir contributes strongly to the subject-to-
subject variability observed in the dental biofilm
microbiome. The practical consequence of the microb-
iome population structure is that the application of
brute force DNA sequencing methods will have limited
power to characterise moderate and low abundance
species. Based on our knowledge of the dental biofilm
microbiome we have modelled the DNA sequence
Figure 2. Population Structure of the Dental Biolm Microbiome.
The large dynamic range of the dental biolm microbiome spans over
4 logs. The inset depicts a zoom in on the most abundant species.
2011 FDI World Dental Federation

coverage expected through the theoretical generation of
10Gb of DNA sequence data. The most abundant DNA
sequences present in dental biofilm samples are human
DNAs. Nearly 80% of the sequences generated from
dental biofilm are derived from the human genome,
although this may vary substantially depending on the
sampling practices used. Just two species make up
nearly 25% of the dental biofilm microbiota (group 1),
five additional species display an average abundance
between 14% of the total bacterial population (group
2). Approximately 100 species display abundance in the
range of 0.1% (group 3). Finally many hundreds of
species are present at an abundance of 0.01% or less
(group 4). The application of deep sequencing of this
microbiome (10Gb of data) yields >100X coverage of
the two genomes in group 1, 8X coverage of the five
species present in group 2, 0.8X coverage of the
genomes in group 3 and finally genomes in group 4
result in less than 0.01X coverage. Therefore only seven
genomes achieve reasonable sequence coverage. The
genomes of the vast majority of species genomes
achieve very low to virtually no sequence coverage at
all.
In order to contend with the difficulties posed by the
population structure of the oral microbiome, we have
developed an experimental approach to study microbi-
omes using a strategy we call Re-Association Kinetic
Enrichment (RAKE) to enrich or normalise 16S rDNA
and metagenomic DNA samples through the removal of
the most abundant sequences present in the nucleic acid
population (Cot analysis). This capability is based on
the long-standing observation that the rate that dena-
tured DNA strands re-associate to form double strand
molecules is directly proportional to the relative
abundance of the sequences in the population47.
Allowing denatured DNAs to re-associate to achieve a
ratio of [dsDNA] [ssDNA] = 10, followed by recovery
of the single-strand fraction on hydroxyapatite columns
allows us to achieve a 10-fold enrichment normalisa-
tion of the genomic DNAs. The use of this classical
method in the context of human microbiome charac-
terisation is likely to facilitate a more comprehensive
characterisation of complex microbial populations and
allow more in depth comparisons of microbiomes in
association with both health and disease.
16S rDNA SEQUENCE PROFILING OF THE DENTAL
BIOFILM MICROBIOME
The sequencing of the 16S rDNA gene has become a
mainstay for bacterial phylogenetic profiling. The 16S
rDNA gene has evolutionary characteristics that make it
well suited for this purpose. The 16S rDNA gene evolves
very slowly relative to other genes encoded in bacterial
genomes. Generally, two strains belonging to the same
species have identical 16S rDNA sequences or may differ
2011 FDI World Dental Federation
Dental caries pathogenicity
at a small number of base positions. The more closely
related two microbial species are (the time since diverg-
ing from their last common ancestor) the more similar are
their 16S rDNA sequences. A large database of annotated
16S rDNA sequences is available and widely used for
profiling microbial species. Across the 16S rDNA genes
length are a series of highly conserved sequences
interspersed with variable sequences. While the con-
served sequence blocks may not differ among bacteria
belonging to the same genus, the variable regions evolve
more freely and provide a means of distinguishing closely
related species from one another. The maximum dis-
criminatory capacity is obtained when the entire 16S
rDNA gene sequence is obtained since it allows compar-
ison of all of the variable regions in the gene allowing
higher resolution assignments to be made. The value of
this approach is that it is culture independent. It has been
estimated that more than 50% of the bacterial species
present in the oral cavity are currently uncultivable. Our
ability to amplify the 16S rDNA gene by PCR directly
from dental biofilms does not require the intermediate
step of cultivation.
THE DENTAL BIOFILM MICROBIOME
To profile the species present in the dental biofilm
microbiome the authors generated near full-length
rDNA amplicons comprised of non-enriched rDNA
and rDNAs that had undergone a 64-fold, 1,000-fold
and 10,000-fold RAKE enrichment for subsequent 16S
rDNA sequencing. The amplicons were derived from
two pools: four caries-free (C-F) and four caries-active
(C-A) twin pairs. These subjects (57-years-old) were
medically healthy individuals. The clinical examina-
tions included diagnostic imaging systems (digital fibre-
optic trans-illumination and quantitative light fluores-
cence). C-A twins had an average of 12 (range = 617)
decayed tooth surfaces where as C-F twins presented
with a decay component = 0. We performed a deep
survey of the composition and complexity of the dental
biofilm microbiome to establish its population structure
and to enable a comprehensive comparison of the
species OTUs present in C-F and C-A microbiomes. An
OTU (Operational Taxonomic Unit) is intended to
approximate the definition of a species and is defined
here as a 16S rDNA sequence with >97% sequence
identity across its entire length. Sequences with less
identity are considered distinct OTUs. In total we
sequenced 47,659 rDNAs that yielded 36,773 high
quality clone sequences (C-F = 17,099 and C-
A = 19,674). In our work detailing the species present
in the dental biofilm microbiome we have identified
nearly 1,500 species, a number significantly higher than
any previous report. It is difficult to establish how close
this number is to reaching saturation but is likely to be
far from the actual species diversity present.
17
Peterson et al.
The depth of our sequencing efforts limits the extent
that we can evaluate less abundant species except for
the fact that the number of OTUs related to singleton
16S rDNA sequence appeared to accumulate in a
relatively linear fashion without signs of decline. We
speculate that the number of species present in the
dental biofilm microbiome may easily exceed several
thousands. This point must be considered in the proper
context however. We do not know much about the
significance of very low abundance species in microbi-
omes. While we can be relatively certain that their
overall contribution to metabolic flux is minimal, they
may produce compounds and or toxins that play a
significant role in altering the population structure that
defines healthy and disease microbiomes. Another
consideration is whether the low abundance species
represent transient entities rather than true residents of
the oral cavity. The bacteria ingested in our diets could
indeed become pseudo members of the dental biofilm
microbiome. Yet another possibility is that strains
belonging to abundant species may exist as low
abundance reservoir members. Under appropriate envi-
ronmental selection the abundance of these strains may
be radically altered. These are early days for the
microbiome efforts and presumably additional efforts
will start to shed light on some of these topics.
Among the nearly 1,500 OTUs observed in our in-
depth analysis of the dental biofilm microbiome were
135 named species. This is by far the largest number of
OTUs identified using full-length 16S rDNA sequence
data. Our data indicate that the species present span at
least four orders of magnitude in abundance. Only four
bacterial divisions are present in the dental biofilm
microbiome at a significant abundance. The dominance
of the Firmicutes is striking and comprises more than
80% of the dental biofilm microbiome. The next most
abundant Phylum is Proteobacteria (11%) followed by
Bacteriodetes (3%) and Fusobacteria (2%). Nearly
66% of the Firmicutes represent Streptococcal spp.
With respect to the Streptococcus, the species repre-
sentation may represent something close to a species
continuum within the oral cavity48. The notion of the
species continuum is consistent with the ecological
prediction of top down selection. Other significant
genera represented among the Firmicutes are Abiotro-
phia spp. and Granulicatella spp. although their
abundance is approximately 20-fold less than the
Streptococci. The most prevalent Proteobacteria, were
Neisseria spp. and Campylobacter spp.
BIOCHEMICAL FLUX IN THE DENTAL BIOFILM
MICROBIOME
Two dominant species present in the dental biofilm
microbiome, S. sanguinis and S. mitis comprise approx-
imately 25% of the total microbial biomass. There are
18
only six additional species present in the dental biofilm
microbiome that are present at frequencies of more
than 1% and together, these eight species make up
nearly 40% of the bacterial biomass of dental biofilm.
The five most abundant species in our survey are all
Streptococci: S. sanguinis, S. mitis, S. oralis, S. gordonii
and S. mutans. The remaining three species are
Abiotrophia defectiva, Granulicatella adjacens and
finally Campylobacter gracilis (Figure 3). It is of
interest to note that the genus Abiotrophia, was
originally comprised of four member species was re-
classified to form a second group referred to as
Granulicatella. Both of these were previously referred
to as nutritionally variant Streptococci. These organ-
isms have been classified as fastidious bacteria that
require vitamin B6 analogs for growth. Despite the
naming of these species they are indeed phylogeneti-
cally distinct from the Streptococci. Unlike the Strep-
tococci, Abiotrophia defectiva and Granulicatella adj-
acens have limited sugar metabolic potential, although
both are known to produce acid from sucrose. Inter-
estingly, Campylobacter gracilis is an asaccharolytic
organism, incapable of metabolising sugars and there-
fore most likely not a direct contributor to caries
development via acid production.
DIFFERENTIALLY ABUNDANT GENERA SPECIES IN
C-F AND C-A MICROBIOMES
Our profiling of the C-F and C-A dental biofilm
microbiome revealed several genera that were differen-
tially abundant. The Streptococcus were in slightly
higher abundance (>50%) overall in C-A compared to
C-F samples. The next most abundant genera Cam-
pylobacter, Neiseria, Gemella and Veillonella were
mixed in their representation. The first three were over-
abundant in the C-F microbiome (>50%), whereas the
Veillonella were slightly over-represented in the C-A
microbiome (>50%). The relationship beyween the
Figure 3. The most abundant species present in the dental biolm
microbiome.
2011 FDI World Dental Federation

Streptococcus and Veillonella is of interest in that the
Veillonella are known to actively metabolise
the products of primary fermentation (lactate). In this
regard as the acid production increases in the C-A tooth
surface environment, so too may be the secondary
metabolism of the Veillonella.
We compared the species profile of the C-F (n = 4)
and C-A (n = 4) pools to identify differentially abun-
dant species. In total we identified 48 species displaying
differential abundance (3-fold or greater). We identified
29 species genera that were over-represented in C-F
and 19 that were over-represented in C-A pools. S.
mutans displayed the largest abundance difference
when comparing C-F and C-A dental biofilm microbi-
omes (65-fold).
SPECIES PROFILING IN THE ORAL CAVITY:
PATHOGEN DETECTION AND DIAGNOSTICS
Despite its widespread use in microbial taxonomy,
the 16S rDNA sequence data has specific limitations,
particularly in the context of the human microbiome.
The 16S rDNA sequence evolves very slowly and
therefore, the comparison of sequences between any
two strains belonging to the same species or any two
closely related species possess zero or a small number
of base differences. It is highly likely that the
distinction between microbiomes in health and dis-
ease lies primarily in the species and strains present.
The 16S rDNA sequence has essentially no power at
this taxonomic level. We are developing a cost-
effective alternative to 16S rDNA sequencing. An
ideal signature sequence is one that cleanly discrim-
inates a species strains from its nearest neighbours,
while at the same time being inclusive of all members
of the species (strains). The use of conserved, protein-
coding sequences as surrogate markers for species
profiling is advantageous since the DNA sequence
identities shared by conserved orthologous proteins
within a species are typically in the 95100% range,
compared to much reduced identities (<85%) across
species. It is not cost-effective, to perform deep 16S
rDNA sequence profiling, or hybridisation-based
(PhyloChip) methods on large human cohorts. Among
the culture-free strategies currently available for
profiling microbial species, present in complex mic-
robiota, one that has been overlooked and thereby
under-utilised is the use of qPCR. This method is
widely accepted as being highly specific and is
considered the gold standard for quantitative accu-
racy. The method has a linear dynamic range of
detection that spans over seven orders of magnitude.
The human microbiome project continues to generate
many reference genome sequences relevant to the oral
cavity. We are using capture array technology to
allow the targeted DNA sequencing of seven con-
2011 FDI World Dental Federation
Dental caries pathogenicity
served, orthologous genes encoded in all eubacterial
species in an iterative fashion to establish a compre-
hensive sequence database. This experimental ap-
proach seeks to mimic the utility provided by
bacterial MLST sequence data but applied in this
case to a microbiome49.
Recently, a number of scientists have been applying
genome sequence capture methodologies to selectively
sequence exons of interest encoded in the human
genome50,51. This method relies on the ability to
represent short DNA sequences (60-mers) on glass
slide microarrays. Hybridisation of denatured geno-
mic DNA on these chips allows selective capture of
specific sequences of interest. These captured se-
quences are then subjected to high throughput DNA
sequencing. This same methodology may be effec-
tively applied to the dental biofilm microbiome. The
seven target genes were selected based on the
following three criteria:
Conservation throughout all Eubacteria and Ar-
chaebacteria
Low probability for horizontal gene exchange
Appropriate rates of evolution to allow species and
strain level sequence changes to be exploited.
Sequences were mined from databases from genera
found in the dental biofilm microbiome. The DNA
sequences corresponding to the seven selected genes
are represented on DNA microarrays as a series of 60-
mers tiling the length of each of the seven genes and
are referred to as seed sequences. Each 60-mer is
capable of capturing relatively large DNAs, 500
1,000 bp with 020% nucleotide sequence divergence,
(80100% identity with seed sequences). These
sequences are assembled and unique orthologous
sequences (>2% divergence) are identified and used
to design a second capture array. If one considers a
phylogenetic tree representing a microbiome, the seed
sequences define several branch tips. The captured
orthologues represent an expansion of sequence
clouds surrounding the seed sequence branches.
Given the large phylogenetic breadth of the dental
biofilm microbiome samples this procedure is carried
out iteratively as multiple capture cycles. We antici-
pate that in each iteration we will expand the size and
number of clouds originating from the initial seed
sequences. It is also anticipated that in early cycles, the
complexity of captured sequences will expand,
perhaps exponentially. As one approaches saturation
of coverage, the number of novel alleles sequenced will
begin to sharply decrease. Using this procedure we will
be in a position to design specific primer pairs for
target genes allowing reliable measurements of species
and strain abundance in dental biofilm microbiomes.
This alternative strategy presents a powerful potential
means for performing diagnostic assays for oral
diseases.
19
Peterson et al.
METAGENOMIC DNA SEQUENCING
DNA sequencing technologies continue to make rapid
advances allowing ever-greater amount of DNA se-
quence data generation at lower cost. These advances
and the application of sequencing technology to the
human microbiome are therefore not coincidental. Our
ability to generate Gigabase quantities of DNA sequence
of genomic DNAs isolated from microbiome samples is
now routine. It is reasonable to expect these boundaries
to shift substantially in the next decade. Comparison of
metagenomic DNA sequence data provides a systematic
and objective way to determine which gene functions are
over or under-represented in C-F and C-A dental biofilm
microbiomes. In the context of the dental biofilm
microbiome, sugar metabolism and its relationship to
acid production are of direct relevance to dental caries.
The reduced pH in C-A dental biofilms provides a
discrete selective pressure with the potential to cause
shifts in species and sub-species present. The positive and
negative selection necessarily alters the relative balance
of genes and gene functions present in the C-F and C-A
dental biofilm microbiome. In this regard, a direct
comparison of metagenomic DNA sequence data derived
from C-F and C-A subjects will reveal those genes and
functional pathways that allow species adaptation and
strain-level niche exploitation in a reduced pH environ-
ment. The genes driving these adaptations will be
revealed as over- and under-represented sequences in
C-F and C-A microbiomes. The very large quantity of
DNA sequence data generated in metagenomic studies
provide statistical power to these inferences. There is
indeed coherence to over- and under-represented genes
encoded in metagenomic data since the over under-
represented genes frequently cluster into over under-
represented biochemical pathways.
Metabolic reconstruction is the process for predicting
an organism or microbial communitys complement of
biochemical reactions and the participating metabolites
from a combination of gene sequence analysis, experi-
mental data and reasoning based on network structure.
Metabolic reconstruction uses annotated gene functions
of proteins to place activities into the context of
pathways, thus providing a way to visualise physiolog-
ically relevant processes. The comparison of functionally
annotated genes and metabolic pathways of any two
microbiomes yields differences in the collective reper-
toire of metabolic and protein functions present. The
over and under-represented functions or pathways reflect
the specific selective pressure(s) or exploitable energetic
opportunities associated with a particular microbial
ecosystem. Examples of such comparisons include the
large expansion of photorhodopsins in microbial popu-
lations found in the Sargasso Sea48 and from the Global
Ocean Sampling (GOS)53. Other examples include the
limited metabolic diversity encountered in the acid mine
20
drainage system54 and the over-abundance of non-host
encoded functions identified in the microbiome of the
human gastrointestinal tract55.
GENE EXPRESSION IN DENTAL BIOFILMS
A number of gene expression studies using DNA
microarray technology have been conducted in S. mutans
to date. These studies have revealed much concerning the
mechanisms and gene products involved in biofilm
formation and acid tolerance. Studies examining amino
acid starvation56, oxygen tension57, sugar transport58,
Mn2+ depletion59 biofilm formation60 and a variety of
responses to stress tolerance have been conducted. The
power of many of these studies exploited the use of
strains that deleted one or more TCS regulators thereby
allowing insights to the mechanistic aspects and the
participating genes involved in complex biological
phenomena such as acid stress tolerance to be achieved.
For example, deletion of the S. mutans vicK gene known
to be involved in H202 resistance, alters the expression of
89 transcripts primarily involved in transport and cell
membrane integrity61. Under low pH growth conditions,
14% of the S. mutans genome displays altered gene
expression. One hundred and sixty-nine genes are up-
regulated and 108 genes are down-regulated. Among the
up-regulated genes were numerous TCS including:
CiaHR, LeuSR, LiaSR, ScnKR, HK Rr1037 1038 and
ComDE62. Systematic deletions of each of the 14 TCS led
to a common set of phenotypes including: reduced acid
tolerance and reduced growth and fitness63. The com-
plement of genes involved in acid tolerance suggest that
S. mutans has evolved a complex system for coping with
acid stress. The TCS encoding HK11 RR11 has been
implicated in biofilm formation. A strain deleted in
RR11 displayed deficiencies in biofilm formation and a
20-fold reduction in competence development. Biofilm
formation was affected in the presence of oxidative stress
once again highlighting the connection between biofilm
formation and stress response. In total 5% (174 genes) of
the genome was differentially expressed in planktonic
compared to biofilm grown cultures. Among the differ-
entially regulated genes were a number of additional
TCS60. Gene expression studies to define the competence
stimulating peptide (ComC) regulon have been reported,
identifying 277 genes that are differentially expressed in
a comC knockout strain64 including a bacteriocin
involved in cell lysis. The comC gene was shown to be
induced in a low pH (pH = 5.0) dependent manner. The
cell lysis activity was shown to affect S. mutans itself, a
process that has been referred to a fratricide.
A genomic method that is being applied recently due to
advances in short read DNA sequence technology is
RNASeq. This method allows gene expression to be
characterised by the sequencing of millions of cDNAs
that are mapped to genome sequences to allow the
2011 FDI World Dental Federation

identification of differentially expressed genes. Applying
this method to the dental biofilm microbiome is currently
underway in our laboratory and will provide a first
glimpse into the gene expression of a large number of
species present in dental biofilms. The complication that
arises is the reduced ability to cleanly map cDNA
sequences to reference genome sequences (due to
sequence divergence) or to attribute cDNA sequences
to appropriate species in the absence of reference genome
sequence. Despite these challenges, the generation of
RNASeq data for microbiomes and comparing profiles
generated from C-F and C-A subjects will shed light on
some of the common mechanisms used by species beyond
S. mutans enabling fitness in the biofilm environment and
how these species cope with nutrient and pH dynamics.
We anticipate a great variety of such mechanisms to be
revealed by such analyses.
CONCLUSIONS
The accumulated data pertaining to biofilm formation,
acid production and acid tolerance mechanisms has
shed light on a number of fascinating evolutionary
accomplishments of the oral microbiome. Our current
understanding of the regulatory networks operating
with regard to biofilm formation, competence develop-
ment and stress responses come largely from S. mutans
but nevertheless illustrate the relative dependence that
this species places with respect to its regulatory and
functional capacity to maintain fitness in the face of a
generally difficult lifestyle. The co-regulation of these
complex systems by overlapping sets of regulators
provides a potential means for the development of
effective therapeutics against S. mutans. The creation of
the human microbiome project is timely and promises
to bring a new set of approaches to long-standing
problems. Many of these approaches are brute force in
nature and will have limited capacity to contend with
the very large dynamic range of species abundance
present in microbiomes. Nevertheless, necessity is the
mother of all invention and we may look forward to
methodological advances to progress in concert with
the inevitable technological advances that will be
employed over the next decade and beyond in order
to better understand dental health and disease.
Conflicts of interest
Dr Bretz is a consultant for Procter & Gamble. The
remaining authors declare no conflicts of interest.
REFERENCES
1. Loesche WJ, Grenier E. Detection of Streptococcus mutans in
plaque samples by the direct fluorescent antibody test. J Dent Res
1976 55: A8793.
2011 FDI World Dental Federation
Dental caries pathogenicity
2. Acs G, Shulman R, Ng MW, Chussid S. The effect of dental
rehabilitation on the body weight of children with early child-
hood caries. Pediatr Dent 1999 21: 109113.
3. Low W, Tan S, Schwartz S. The effect of severe caries on the
quality of life in young children. Pediatr Dent 1999 21: 325326.
4. Lemos JA, Abranches J, Burne RA. Responses of cariogenic
streptococci to environmental stresses. Curr Issues Mol Biol 2005
7: 95107.
5. Loesche WJ. Role of Streptococcus mutans in human dental de-
cay. Microbiol Rev 1986 50: 353380.
6. Darveau RP, Tanner A, Page RC. The microbial challenge in
periodontitis. Periodontol 2000 1997 14: 1232.
7. Kolenbrander PE. Oral microbial communities: biofilms, interac-
tions, and genetic systems. Annu Rev Microbiol 2000 54: 413437.
8. Moore WE, Moore LV. The bacteria of periodontal diseases.
Periodontol 2000 1994 5: 6677.
9. Hutter G, Schlagenhauf U, Valenza G, et al. Molecular analysis of
bacteria in periodontitis: evaluation of clone libraries, novel phyl-
otypes and putative pathogens. Microbiology 2003 149: 6775.
10. Loesche WJ, Straffon LH. Longitudinal investigation of the role
of Streptococcus mutans in human fissure decay. Infecti Immun
1979 26: 498507.
11. van Houte J. Role of micro-organisms in caries etiology. J Dent
Res 1994 73: 672681.
12. Becker MR, Paster BJ, Leys EJ, et al. Molecular analysis of bac-
terial species associated with childhood caries. J Clin Microbiol
2002 40: 10011009.
13. Aas JA, Griffen AL, Dardis SR, Lee AM, Olsen I, Dewhirst FE, et
al. Bacteria of dental caries in primary and permanent teeth in
children and young adults. J Clin Microbiol 2008 46: 14071417.
14. Munson MA, Banerjee A, Watson TF, et al. Molecular analysis of
the microflora associated with dental caries. J Clin Microbiol
2004 42: 30233029.
15. Chhour KL, Nadkarni MA, Byun R, et al. Molecular analysis of
microbial diversity in advanced caries. J Clin Microbiol 2005 43:
843849.
16. Corby PM, Lyons-Weiler J, Bretz WA, et al. Microbial risk
indicators of early childhood caries. J Clin Microbiol 2005 43:
57535759.
17. Kolenbrander PE, Palmer RJ Jr, Rickard AH, et al. Bacterial
interactions and successions during plaque development. Peri-
odontol 2000 2006 42: 4779.
18. Rosan B, Lamont RJ. Dental plaque formation. Microbes Infect
2000 2: 15991607.
19. Zijnge V, van Leeuwen MB, Degener JE, et al. Oral biofilm
architecture on natural teeth. PLoS One 2010 5: e9321.
20. Deng DM, Liu MJ, ten Cate JM, et al. The VicRK system of
Streptococcus mutans responds to oxidative stress. J Dent Res
2007 86: 60610.
21. Banu LD, Conrads G, Rehrauer H, et al. The Streptococcus
mutans serine threonine kinase, PknB, regulates competence
development, bacteriocin production, and cell wall metabolism.
Infect Immun 2010 78: 22092220.
22. Rolerson E, Swick A, Newlon L, et al. The SloR Dlg metall-
oregulator modulates Streptococcus mutans virulence gene
expression. J Bacteriol 2006 188: 50335044.
23. Lemos JA, Lin VK, Nascimento MM, et al. Three gene products
govern (p)ppGpp production by Streptococcus mutans. Mol
Microbiol 2007 65: 15681581.
24. Paes Leme AF, Koo H, Bellato CM, et al. The role of sucrose in
cariogenic dental biofilm formationnew insight. J Dent Res 2006
85: 878887.
25. Ajdic D, McShan WM, McLaughlin RE, et al. Genome sequence
of Streptococcus mutans UA159, a cariogenic dental pathogen.
Proc Natl Acad Sci USA 2002 99: 1443414439.
21
Peterson et al.
26. Bender GR, Sutton SV, Marquis RE. Acid tolerance, proton
permeabilities, and membrane ATPases of oral streptococci. In-
fect Immun 1986 53: 331338.
27. Burne RA, Marquis RE. Alkali production by oral bacteria and pro-
tection against dental caries. FEMS Microbiol Lett 2000 193: 16.
28. Liu Y, Burne RA. Multiple two-component systems of Strepto-
coccus mutans regulate agmatine deiminase gene expression and
stress tolerance. J Bacteriol 2009 191: 73637366.
29. Liu Y, Zeng L, Burne RA. AguR is required for induction of the
Streptococcus mutans agmatine deiminase system by low pH and
agmatine. Appl Environ Microbiol 2009 75: 26292637.
30. Kleinberg I. A mixed-bacteria ecological approach to under-
standing the role of the oral bacteria in dental caries causation: an
alternative to Streptococcus mutans and the specific-plaque
hypothesis. Crit Rev Oral Biol Med 2002 13: 108125.
31. Sheng J, Marquis RE. Malolactic fermentation by Streptococcus
mutans. FEMS Microbiol Lett 2007 272: 196201.
32. Lemos JA, Luzardo Y, Burne RA. Physiologic effects of forced
down-regulation of dnaK and groEL expression in Streptococcus
mutans. J Bacteriol 2007 189: 15821588.
33. Fozo EM, Scott-Anne K, Koo H, et al. Role of unsaturated fatty
acid biosynthesis in virulence of Streptococcus mutans. Infect
Immun 2007 75: 15371539.
34. Nelson KE, Weinstock GM, Highlander SK, et al. Human Mi-
crobiome Jumpstart Reference Strains C. A catalog of reference
genomes from the human microbiome. Science 2010 328: 994999.
35. Li Y, Caufield PW. Arbitrarily primed polymerase chain reaction
fingerprinting for the genotypic identification of mutans strepto-
cocci from humans. Oral Microbiol Immunol 1998 13: 1722.
36. Maruyama F, Kobata M, Kurokawa K, et al. Comparative
genomic analyses of Streptococcus mutans provide insights into
chromosomal shuffling and species-specific content. BMC Ge-
nomics 2009 10: 358.
37. Caufield PW, Ratanapridakul K, Allen DN, et al. Plasmid-con-
taining strains of Streptococcus mutans cluster within family and
racial cohorts: implications for natural transmission. Infect Im-
mun 1988 56: 32163220.
38. Caufield PW, Wannemuehler YM, Hansen JB. Familial clustering
of the Streptococcus mutans cryptic plasmid strain in a dental
clinic population. Infect Immun 1982 38: 785787.
39. Kamiya RU, Napimoga MH, Hofling JF, et al. Frequency of four
different mutacin genes in Streptococcus mutans genotypes iso-
lated from caries-free and caries-active individuals.
J Med Microbiol 2005 54: 599604.
40. Shibata Y, Ozaki K, Seki M, et al. Analysis of loci required for
determination of serotype antigenicity in Streptococcus mutans
and its clinical utilization. J Clin Microbiol 2003 41: 41074112.
41. Li YH, Lau PC, Lee JH, et al. Natural genetic transformation of
Streptococcus mutans growing in biofilms. J Bacteriol 2001 183:
897908.
42. Klein MI, Bang S, Florio FM, et al. Genetic diversity of compe-
tence gene loci in clinical genotypes of Streptococcus mutans.
J Clin Microbiol 2006 44: 30153020.
43. Waterhouse JC, Swan DC, Russell RR. Comparative genome
hybridization of Streptococcus mutans strains. Oral Microbiol
Immunol 2007 22: 103110.
44. Zhang L, Foxman B, Drake DR, et al. Comparative whole-gen-
ome analysis of Streptococcus mutans isolates within and among
individuals of different caries status. Oral Microbiol Immunol
2009 24: 197203.
45. Tong H, Chen W, Shi W, et al. SO-LAAO, a novel L-amino acid
oxidase that enables Streptococcus oligofermentans to outcom-
pete Streptococcus mutans by generating H2O2 from peptone.
J Bacteriol 2008 190: 47164721.
46. Tamura S, Yonezawa H, Motegi M, et al. Inhibiting effects of
Streptococcus salivarius on competence-stimulating peptide-
22
dependent biofilm formation by Streptococcus mutans. Oral
Microbiol Immunol 2009 24: 152161.
47. Peterson DG, Schulze SR, Sciara EB, et al. Integration of Cot
analysis, DNA cloning, and high-throughput sequencing facili-
tates genome characterization and gene discovery. Genome Res
2002 12: 795807.
48. Venter JC, Remington K, Heidelberg JF, et al. Environmental
genome shotgun sequencing of the Sargasso Sea. Science 2004
304: 6674.
49. Bishop CJ, Aanensen DM, Jordan GE, et al. Assigning strains to
bacterial species via the internet. BMC Biol 2009 7: 3.
50. Ng SB, Turner EH, Robertson PD, et al. Targeted capture and
massively parallel sequencing of 12 human exomes. Nature 2009
461: 272276.
51. Porreca GJ, Zhang K, Li JB, et al. Multiplex amplification of large
sets of human exons. Nat Methods 2007 4: 931936.
52. Turnbaugh PJ, Hamady M, Yatsunenko T, et al. A core gut mi-
crobiome in obese and lean twins. Nature 2009 457: 480484.
53. Rusch DB, Halpern AL, Sutton G, et al. The Sorcerer II Global
Ocean Sampling expedition: northwest Atlantic through eastern
tropical Pacific. PLoS Biol 2007 5: e77.
54. Schloss PD, Handelsman J. A statistical toolbox for metage-
nomics: assessing functional diversity in microbial communities.
BMC Bioinformatics 2008 9: 34.
55. Turnbaugh PJ, Ley RE, Mahowald MA, et al. An obesity-asso-
ciated gut microbiome with increased capacity for energy harvest.
Nature 2006 444: 10271031.
56. Nascimento MM, Lemos JA, Abranches J, et al. Role of RelA of
Streptococcus mutans in global control of gene expression.
J Bacteriol 2008 190: 2836.
57. Ahn SJ, Wen ZT, Burne RA. Effects of oxygen on virulence traits
of Streptococcus mutans. J Bacteriol 2007 189: 85198527.
58. Ajdic D, Pham VT. Global transcriptional analysis of Strepto-
coccus mutans sugar transporters using microarrays. J Bacteriol
2007 189: 50495059.
59. Arirachakaran P, Benjavongkulchai E, Luengpailin S, et al.
Manganese affects Streptococcus mutans virulence gene expres-
sion. Caries Res 2007 41: 503511.
60. Perry JA, Levesque CM, Suntharaligam P, et al. Involvement of
Streptococcus mutans regulator RR11 in oxidative stress response
during biofilm growth and in the development of genetic com-
petence. Lett Appl Microbiol 2008 47: 439444.
61. Senadheera D, Krastel K, Mair R, et al. Inactivation of VicK
affects acid production and acid survival of Streptococcus mu-
tans. J Bacteriol 2009 191: 64156424.
62. Gong Y, Tian XL, Sutherland T, et al. Global transcriptional
analysis of acid-inducible genes in Streptococcus mutans: multiple
two-component systems involved in acid adaptation. Microbiol-
ogy 2009 155: 33223332.
63. Levesque CM, Mair RW, Perry JA, et al. Systemic inactivation
and phenotypic characterization of two-component systems in
expression of Streptococcus mutans virulence properties. Lett
Appl Microbiol 2007 45: 398404.
64. Perry JA, Jones MB, Peterson SN, et al. Peptide alarmone sig-
nalling triggers an auto-active bacteriocin necessary for genetic
competence. Mol Microbiol 2009 72: 905917.
Correspondence to:
Dr Scott N. Peterson,
The J. Craig Venter Institute,
9704 Medical Center Drive,
Rockville MD 20886,USA.
Email: scottp@jcvi.org
2011 FDI World Dental Federation