736496

research-article2017
ADRXXX10.1177/0022034517736496Advances in Dental ResearchThe Caries Microbiome

Advances
Advances in Dental Research
2018, Vol. 29(1) 78­–85
The Caries Microbiome: Implications © International & American Associations
for Dental Research 2018

for Reversing Dysbiosis Reprints and permissions:

sagepub.com/journalsPermissions.nav
DOI: 10.1177/0022034517736496
https://doi.org/10.1177/0022034517736496
journals.sagepub.com/home/adr

A.C.R. Tanner1,2, C.A. Kressirer1,2, S. Rothmiller3, I. Johansson4,

and N.I. Chalmers5

Abstract
The oral microbiome plays a critical role in maintaining oral health. Frequent dietary carbohydrate intake can lead to dysbiosis of the
microbial community from overproduction of acid with selection for increases in acidogenic, acid-tolerant bacteria. Knowledge of the
caries-associated microbiome is key in planning approaches to reverse the dysbiosis to achieve health. For risk assessment and treatment
studies, it would be valuable to establish whether microbial monitoring requires assay of multiple species or whether selected key
species would suffice. Early investigations of the oral microbiota relied on culture-based methods to determine the major bacteria in
health and disease. Microbial monitoring using gene probes facilitated study of larger populations. DNA probe methods confirmed and
expanded the importance of transmission of bacteria from mother to infant and association of preselected species, including mutans
streptococci and lactobacilli with caries in larger populations. 16S ribosomal RNA (rRNA) probes confirmed the wide diversity of
species in oral and caries microbiomes. Open-ended techniques provide tools for discovery of new species, particularly when strain/
clone identification includes gene sequence data. Anaerobic culture highlighted the caries association of Actinomyces and related species.
Scardovia wiggsiae, in the Actinomyces/Bifidobacterium family, and several Actinomyces species have the cariogenic traits of acid production
and acid tolerance. Next-generation sequencing combined with polymerase chain reaction methods revealed a strong association with
mutans streptococci in a high caries population with poor oral hygiene and limited access to care. A population with a lower caries
experience generally had lower or no Streptococcus mutans detection but harbored other acidogenic taxa in the microbiome. Study of
the microbiome suggests a role for the assay of selected putative cariogenic species in more aggressive diseases. For many populations
with caries progression, however, assay of multiple species will likely be warranted to determine the caries profile of the population
and/or individuals under study.

Keywords: microbiology, pediatric dentistry, biofilms, Streptococcus mutans, Actinomyces, risk assessment

Introduction biofilm as a whole, including species in the core community

and species that can balance the acid produced after normal
The caries microbiome plays a critical role and is the primary dietary intakes of carbohydrates. Major putative caries patho-
etiology in dental caries. Use of the term microbiome was gens include the mutans streptococci, Streptococcus mutans
ascribed to Joshua Lederberg (2003) with the following and Streptococcus sobrinus, and various Lactobacillus species.
description: “the ecological community of commensal, symbi- An extended ecological plaque hypothesis (Takahashi and
otic, and pathogenic microorganisms that literally share our Nyvad 2011) incorporates roles in caries initiation and progres-
body space and have been all but ignored as determinants of sion for other acid-tolerant Streptococcus, Actinomyces,
health and disease.” The microbiome that naturally colonizes Bifidobacterium, and Lactobacillus species. There is a critical
teeth in health is a biofilm community that can counterbalance role for knowledge of the caries microbiome for caries risk
acid production from dietary intakes of carbohydrates to main- assessment, prevention, and treatment and for the development
tain an intact tooth surface, for example, by ammonia produc-
tion from arginine or urea (Burne and Marquis 2000). Dysbiosis 1
of the biofilm, with change of the bacterial composition, occurs The Forsyth Institute, Cambridge, MA, USA
2
Harvard School of Dental Medicine, Boston, MA, USA
when excessive and frequent intakes of carbohydrates result in 3
Bundeswehr Institute of Pharmacology and Toxicology, Munich,
acid production that exceeds the buffering capacity of the Germany
healthy microbiome. This increases risk of demineralization of 4
Department of Odontology/section of Cariology, Umeå University,
the tooth surface and can lead to dental caries. Associated risk Umeå, Sweden
5
factors for caries include the mineralization status of teeth and DentaQuest Institute, Columbia, MD, USA
quantity and composition of saliva in the oral cavity. Corresponding Author:
Understanding of the dysbiosis of the biofilm has included A.C.R. Tanner, The Forsyth Institute, 245 First Street, Cambridge, MA
focus on selected acidogenic and aciduric caries-associated 02142, USA.
species, considered caries pathogens, and modulation of the Email: annetanner@forsyth.org
The Caries Microbiome 79

of new anticaries therapies. A major consideration is what on samples from young children with aggressive caries in
components of the microbiome are most efficient, effective, Saipan, Micronesia (Milgrom et al. 2000). Background and
and realistic to monitor to evaluate reversal of the microbiome diet information was obtained by survey and caries status and
dysbiosis to that compatible with health. In this short review, 2 oral microbiota evaluated in 6- to 36-mo infants and that of
hypotheses are considered: 1) monitoring target putative caries their primary caregivers (Tanner et al. 2002). S. mutans was the
pathogens (e.g., S. mutans and Lactobacillus species) or 2) primary caries-associated species, and its detection in the
monitoring the community seeking upset of balance away from infants was strongly associated with its detection in the child’s
health. A more detailed review of the caries microbiome was primary caregiver (Tanner et al. 2002). Indeed, all the 38
previously presented (Tanner et al. 2016). assayed species that included caries, gingival, and periodontal
species were detected more frequently in children when also
detected in the caregiver. The age at which species were
Exploring the Caries Microbiome detected was more strongly associated with species detection
The bacteria on teeth related to dental caries have been studied in the caregiver than the infant age. Implications for therapy
for over 100 y using microbial methods contemporary to the from these observations include that any treatments designed
time of study. An early director of The Forsyth Institute, Percy to prevent or reduce colonization of S. mutans would need to
Howe, published a paper on the microbiology of dental caries in start very early. Furthermore, the high concordance between
1917 that noted the key role of bacteria currently recognized as the microbiota detected in caregivers and infants indicates the
Lactobacillus and Bifidobacterium species. Howe also pio- value of treating caregivers/mothers to reduce the likelihood of
neered chemotherapeutic treatments incorporating silver (Brown transmission of pathogenic species.
1952), approaches that have had a resurgence of interest. In another population in Boston, Massachusetts, young chil-
Microbial studies into dental caries continued at The Forsyth dren were monitored for the presence of caries during an
Institute. Harold Jordan and coworkers showed microbial trans- extended pediatrician well-child visit (Nunn et al. 2009).
mission between mothers to children using bacteriocin typing Demographic, dietary, and diet history data were obtained from
of S. mutans (Berkowitz and Jordan 1975). Ronald Gibbons the child’s caregiver in addition to oral microbial samples from
considered S. mutans as a major pathogen of dental caries and, children. Bacteria were assayed by DNA probes and by poly-
with van Houte, demonstrated increased adherence of S. mutans merase chain reaction (PCR) for S. mutans (Kanasi, Johansson,
for tooth surfaces over other oral mucosal sites (Gibbons and et al. 2010). There was a higher detection frequency of species
van Houte 1971). With Caufield, Gibbons studied efficacy of including S. mutans and other caries-associated species from
topical iodine for suppressing S. mutans (Caufield and Gibbons tooth than tongue samples, suggesting that sampling the tongue
1979). van Houte also published seminal microbiology findings could underestimate detection of S. mutans and other caries-
in caries etiology that continue to provide background to studies associated species. Comparing microbial methods, a species-
today. van Houte was a strong advocate of a caries microbiota specific PCR assay for S. mutans yielded a higher detection
that did not rely on S. mutans solely for cariogenicity, and he frequency in caries children than the DNA probe assay, suggest-
described the acidogenicity and caries association of several ing improved detection sensitivity by the more targeted PCR
bacterial groups, including nonmutans streptococci and bifido- method. In modeling of the microbial data by caries status by
bacteria in addition to lactobacilli and S. mutans (van Houte et partial least squares regression, there was a higher specificity
al. 1996; van Houte et al. 1991). Several studies demonstrated for S. mutans with caries detection with the PCR assay than by
that anaerobic approaches improved viable recovery of oral probes. Further different associations with caries with several
bacteria, including those associated with dental caries by Lactobacillus and Actinomyces species were observed. These
Loesche and coworkers (Loesche and Syed 1973). Findings observations suggest specificity in the microbiome of dental
from these studies were replicated by other investigators caries at the species level for several genera and that an all
(Hardie et al. 1977) to provide a background for understanding Lactobacillus species assay, as from a selective medium or
the composition of the caries-associated biofilm to include genus PCR test, may not differentiate the caries-associated lac-
specific highly aciduric and acidogenic species, particularly tobacilli from probiotic health–associated species.
S. mutans, and other acidogenic species in Actinomyces, In both of these populations of young children, S. mutans
Bifidobacterium, and nonmutans streptococci. was the principal caries-associated species observed, although
the association was microbiological method dependent: tar-
geted assay by PCR was more sensitive than a multispecies
Preselected Species Assay probe approach.
Using DNA Probes
Molecular methods using DNA and 16S ribosomal RNA Open-Ended Microbiology Methods,
(rRNA)–based probes facilitated analysis of oral bacteria from 16S rRNA-Based, Anaerobic Culture,
more individuals than with culture-based methods. In one
Next-Generation Sequencing
approach, DNA probes made from cultured bacterial species
and bacterial samples were applied to membranes in a checker- The whole genomic probes were limited to detecting a
board-type array (Socransky et al. 1994). This assay was used preselected set of cultured species and limits enquiry into
80 Advances in Dental Research 29(1)

childhood caries (ECC) was an unnamed

“Bifidobacterium CX010” species, in
addition to S. mutans (Becker et al.
2002).
To better understand the caries micro-
biome a comparison of anaerobic culture
(Tanner, Kressirer, et al. 2011) and 16S
rRNA cloning analysis (Kanasi, Dewhirst,
et al. 2010), methods revealed increased
detection of Actinomyces by culture meth-
ods (Tanner 2015). Furthermore, since
species detected by molecular cloning
methods were also cultured, this sug-
gested that most of microbiota associated
with dental caries is cultivable as previ-
ously described (Munson et al. 2004).
Highest counts and diversity were from
the blood agar and included several
unnamed taxa recognized previously from
noncultural methods. These isolates were
submitted to the Human Oral Microbiome
Database (HOMD) pipeline for full
genome sequencing. Some “unculture-
ability” can thus be explained by
Figure 1.  Severe early childhood caries (ECC) and caries-free (CF) microbiomes from blood sequence-based versus biochemical tests
pH 7 and acid pH 5 agar isolation. The anaerobic microbiota was characterized and compared
for characterization. These findings sug-
from 40 caries-free and 42 severe ECC children (Tanner, Kressirer, et al. 2011). Microbiomes
of children were compared using principal component analysis. The microbiota from ECC from gest that it is unlikely there are major
acidic agar isolation was more tightly clustered than that from caries-free children, with some groups of “hidden” species for consider-
separation of the ECC and CF communities. The microbiota cultured on blood agar (pH 7) was ation in caries treatment regimens as long
more widely distributed than that from acid agar. There was considerable overlap in the ECC and
CF communities. as methods can encompass the microbiota
present.
Different segments of the microbiota
new bacterial associations with disease. Sequence-based anal- were cultured on a rich blood–containing agar compared with
yses, initially using PCR of the 16S rRNA gene, cloning ampli- the low pH acid isolation medium (Figure 1). The greatest sep-
cons, and their sequencing, however, showed a much larger aration between microbial populations from caries-free and
diversity of species, including many previously unrecognized/ ECC children was from the acid isolation medium. This obser-
uncultured in oral samples (Dewhirst et al. 2010). Sequence- vation suggests that the major differentiation between health
based data were used to design bacterial probes that were ini- and caries is in the acid-tolerant population selected during the
tially assayed in a “reverse-checkerboard” assay (Socransky et disease process, rather than the predominant microbiota that
al. 1994) and then miniaturized to arrays, for example, the includes the core microbiome. Acid isolation suppressed total
Forsyth-based Human Oral Microbe Identification Microarray counts about 10-fold (Hughes et al. 2012), particularly for
(HOMIM) microarray (Preza et al. 2009). Together, these 16S most Gram-negative genera, except for Veillonella species
rRNA-based approaches allowed a relatively rapid and/or in- (Tanner, Kressirer, et al. 2011). Acid-tolerant species included
depth assay of the wide diversity of species from multiple sites. selected Streptococcus and Actinomyces species, including the
In particular, there was emergence of a wider spectrum of spe- mutans streptococci, Actinomyces odontolyticus and several
cies in “Gram-positive rod” genera as caries associated. These unnamed Actinomyces species, Bifidobacterium/Scardovia,
included Atopobium, Propionibacterium, and Bifidobacterium and Lactobacillus species. The 2 major caries-associated spe-
in addition to Actinomyces and Lactobacillus (Aas et al. 2008). cies, however, S. mutans and Scardovia wiggsiae, were signifi-
Furthermore, many Gram-negative genera and species were cantly different between health and disease from both acid and
detected in carious lesions, including Selenomonas, Fusobacterium, blood agar isolation. Nevertheless, the proportion of S. mutans
Porphyromonas, and Haemophilus (Corby et al. 2005; Gross and, by extension, S. wiggsiae was at low overall proportions
et al. 2010). At the same time, there were increasing reports of of the microbiome since S. mutans from selective isolation
a lack of detection of S. mutans in carious lesions and recogni- comprised at most 10% of the total blood and 30% of the acid
tion of nonmutans caries-associated microbiotas, consistent agar counts (Hughes et al. 2012). Together, findings from com-
with previous observations from culture-based analyses paring detection of S. mutans by different methods have impli-
(Hardie et al. 1977; Boyar et al. 1989; van Houte et al. 1994). cations in designing microbial monitoring. To explore the
One taxon that emerged as associated with severe early microbiota, open-ended cultural or molecular methods would
The Caries Microbiome 81

be appropriate. If, however, the study goal is

to detect or suppress target acidogenic caries- A B
associated taxa, then methods that include
acid isolation or focus on species specific
identification, including PCR and quantitative
PCR (qPCR) methods, would have greater
sensitivity.
A major issue posttherapy is caries pro-
gression to new sites and recurrent caries
around restorations, particularly for severe
ECC following restorative therapy under gen-
eral anesthesia (Berkowitz et al. 2011).
Furthermore, S. mutans levels have not been Streptococcus mutans, Scardovia S. mutans, Scardovia wiggsiae,
reliable in predicting children with treatment wiggsiae, Streptococcus sobrinus, Streptococcus sobrinus and Diet (> 4
Veillonella parvula, Acnomyces cariogenic drinks a day)
relapse (Berkowitz et al. 2011; Hughes et al.
gerencseriae, Streptococcus cristatus ROC = 81.1
2012). Pre- and posttreatment monitoring ROC = 82.2
indicated that successful treatment was
accompanied by a changed microbial profile Figure 2.  Receiver operator curves (ROCs) from discriminant analysis for early childhood
of several species detected using a multispe- caries (ECC) risk assessment. Sensitivity, the true positive rate, is plotted against the
cies microarray, although not significantly for false-positive rate, 1 – specificity. Values above the diagonal lines indicate good fit of the
model, whereas values below the diagonal lines indicate poor classification. The diagonal is a
S. mutans (Tanner, Kent, et al. 2011). The nondiscrimination line where sensitivity is equal to 1 – specificity. The true- and false-positive
changed profile was not observed in the rates were quite similar using either 6 caries-associated species or 3 species and diet. (A)
microbiota of children with caries progression ROC using 6 caries-associated species. (B) ROC using Streptococcus mutans, Streptococcus
and new lesions. This finding suggests that sobrinus, and Scardovia wiggsiae with >4 liquid cariogenic foods/d from Palmer et al. (2010).
Red line area under curve (AOC) for caries-free and blue line AOC for ECC. This example
microbial shifts are possible but would require highlights the value of incorporating data from different types of risk factors, in this example,
monitoring for several taxa to evaluate the microbiology and diet.
change.
Risk assessment for dental caries frequently incorporates species, designated by HOT (HOT = Human Oral Taxon [www.
diet and bacteria, and the interrelatedness was shown by an homd.org]) numbers, were cultured that were acid tolerant;
association between S. mutans detection and intake frequency thus, the acidogenic potential at low pHs was of interest to
of putative cariogenic foods (Palmer et al. 2010). In the severe explore cariogenic potential (van Houte 1994). S. wiggsiae and
ECC population studies, risk assessment using 8 species, several Actinomyces isolates were acid tolerant when tested on
including S. mutans, S. wiggsiae, and S. sobrinus, gave a a low pH agar (Table). Using a microtiter plate format, strains
receiver operator curve (ROC) 82.2, whereas risk assessment were acidogenic with similar terminal pH values, at initial neu-
with only 3 species, S. mutans, S. wiggsiae, and S. sobrinus, tral and low pHs. Final pH readings and acid-tolerant growth of
and addition of diet (over 4 cariogenic beverages, including ice some Actinomyces were similar to those of S. mutans and S.
cream and sweetened yogurt a day) yielded a very similar sobrinus reference strains tested. Other Actinomyces strains
value of ROC 81.1 (Figure 2). Thus, considering diet as well as were less acid tolerant and acidogenic, suggesting considerable
the microbiota added to risk assessment as previously advo- heterogeneity of potential cariogenicity within Actinomyces.
cated (Bratthall and Hansel Petersson 2005). In clinical prac- Actinomyces have long been recognized as components of the
tice, the tests selected would need to be tailored to the caries microbiome and are a key group in caries initiation and
population under study as caries is not a homogeneous disease progression under the extended ecological plaque hypothesis
and likely include factors other than diet and bacteria. Modeling (Takahashi and Nyvad 2011). Acidogenicity under acid condi-
on microbiota, diet, and clinical characteristics showed 2 tions by members of Actinomyces as well as Scardovia affirms
groups of ECC children, highlighting heterogeneity in caries the importance of these species in the caries microbiome and
populations (Palmer et al. 2010). Together, these observations very likely as components in caries risk assessment.
highlight the value of pretreatment to know what disease/ In addition to severe ECC, S. wiggsiae was associated with
microbiota and dietary risk factors the individual patient has caries initiation in 2 different populations of adolescents with
before designing therapy and addressing the caries dysbiosis. fixed orthodontic appliances (Tanner et al. 2012; Torlakovic
The new species, S. wiggsiae (Downes et al. 2011), which et al. 2012). The microbiota associated with fixed orthodontic
was identified as caries associated in severe ECC (Tanner, appliances, however, has been studied mainly relative to gingi-
Kressirer, et al. 2011), was previously recognized as the vitis, and both S. mutans and S. wiggsiae were also associated
unnamed “Bifidobacterium” clone, CX010 (Becker et al. 2002). with gingivitis, although the primary association was with the
S. wiggsiae was detected in 40% of caries children without white spot lesions (Tanner et al. 2012). Both caries and gingi-
S. mutans (Tanner, Kressirer, et al. 2011), suggesting further val inflammation are plaque associated, and consideration of
study was indicated on the potential cariogenicity of the new both would be needed to differentiate the primary disease asso-
species (Kressirer et al. 2017). Several new Actinomyces ciation and importance in treatment effects, if any.
82 Advances in Dental Research 29(1)

Table.  Acid Tolerance (Measured as Growth on Agar pH 5.5) and Acidogenicity (Measured as Final pH in Growth Medium) of Mutans Streptococci,
Scardovia, and Actinomyces Strains.

Final pH Readings
Growth
  Agar Initial pH 7.0 Initial pH 5.5
a b b b b
Bacterial Strains pH 5.5 Glucose Sucrose Fructose Glucose Sucroseb Fructoseb

Streptococcus mutans ATCC 25175 2 3.83 ± 0.00 3.76 ± 0.00 3.79 ± 0.00 3.95 ± 0.02 3.87 ± 0.00 3.89 ± 0.00
Streptococcus mutans SJ 2 3.86 ± 0.00 3.82 ± 0.00 3.80 ± 0.00 4.11 ± 0.01 4.05 ± 0.00 4.06 ± 0.00
Streptococcus sobrinus ATCC 33478 2 2.98 ± 0.01 2.97 ± 0.02 2.91 ± 0.01 3.78 ± 0.05 3.71 ± 0.00 3.68 ± 0.00
Streptococcus sobrinus ATCC 27352/6715 2 3.72 ± 0.03 3.73 ± 0.00 3.74 ± 0.00 3.99 ± 0.01 4.01 ± 0.01 4.03 ± 0.00
Scardovia wiggsiae DSM 22547/C155A 2 3.30 ± 0.01 3.36 ± 0.01 3.12 ± 0.01 3.47 ± 0.03 3.65 ± 0.03 3.52 ± 0.03
Scardovia wiggsiae H42A2/F0424 2 3.51 ± 0.02 3.26 ± 0.00 3.23 ± 0.05 4.02 ± 0.03 3.85 ± 0.02 3.75 ± 0.01
Actinomyces johnsonii (H122B8) 2 3.43 ± 0.01 3.48 ± 0.01 3.49 ± 0.01 4.09 ± 0.01 4.12 ± 0.01 4.10 ± 0.01
Actinomyces sp. HOT 170 (A84A17) 2 3.52 ± 0.04 3.44 ± 0.00 3.40 ± 0.00 3.75 ± 0.07 3.82 ± 0.01 3.89 ± 0.00
Actinomyces oris (H49B49) 1 3.77 ± 0.00 3.70 ± 0.01 3.83 ± 0.01 4.29 ± 0.00 4.27 ± 0.01 4.29 ± 0.00
Actinomyces naeslundii I (H101A18) 1 3.80 ± 0.02 3.82 ± 0.01 3.87 ± 0.02 4.21 ± 0.01 4.21 ± 0.01 4.12 ± 0.04
Actinomyces sp. HOT 179 (A62A40) 2 3.82 ± 0.00 3.80 ± 0.00 3.84 ± 0.00 4.28 ± 0.02 4.25 ± 0.00 4.25 ± 0.00
Actinomyces sp. HOT 448 (T39B26) 2 3.88 ± 0.00 3.91 ± 0.02 3.98 ± 0.03 4.26 ± 0.01 4.24 ± 0.00 4.19 ± 0.04
Actinomyces graevenitzii (A62A23) 2 3.95 ± 0.01 4.01 ± 0.00 4.01 ± 0.00 3.27 ± 0.02 3.34 ± 0.02 3.26 ± 0.00
Actinomyces dentalis (H22B8) 1 4.11 ± 0.00 4.07 ± 0.01 4.08 ± 0.06 4.24 ± 0.01 4.25 ± 0.01 4.21 ± 0.03
Actinomyces sp. HOT 180 (T17B14) 2 4.12 ± 0.13 3.97 ± 0.00 3.88 ± 0.02 4.17 ± 0.00 4.16 ± 0.00 4.15 ± 0.02
Actinomyces sp. HOT 177 (H49B36) 2 4.18 ± 0.00 4.17 ± 0.00 4.29 ± 0.00 3.69 ± 0.01 3.76 ± 0.01 3.80 ± 0.00
Actinomyces sp. HOT 171 (H49B26) 2 4.18 ± 0.01 4.22 ± 0.04 4.32 ± 0.11 4.31 ± 0.01 4.35 ± 0.00 4.32 ± 0.01
Actinomyces sp. HOT 172 (H304PB12) 2 4.26 ± 0.03 4.27 ± 0.01 4.25 ± 0.01 4.14 ± 0.05 4.06 ± 0.02 4.01 ± 0.02
Actinomyces sp. HOT 169 (T1B18) 1 4.28 ± 0.05 4.21 ± 0.01 4.19 ± 0.01 4.36 ± 0.04 4.26 ± 0.00 4.28 ± 0.04
Actinomyces sp. HOT 178 (H11B25) 1 4.36 ± 0.01 4.34 ± 0.01 4.22 ± 0.00 4.43 ± 0.01 4.43 ± 0.00 4.30 ± 0.00
Actinomyces sp. HOT 877 (T1B25) 1 4.38 ± 0.03 4.43 ± 0.00 4.23 ± 0.00 4.43 ± 0.02 4.52 ± 0.00 4.31 ± 0.00
Actinomyces gerencseriae (H74B34) 1 4.52 ± 0.02 4.57 ± 0.00 4.51 ± 0.01 3.85 ± 0.01 3.92 ± 0.03 3.80 ± 0.04
Actinomyces odontolyticus (H106B16) 1 4.63 ± 0.05 4.69 ± 0.01 4.57 ± 0.00 4.48 ± 0.04 4.54 ± 0.00 4.45 ± 0.00
Actinomyces israelii (A87A37) 1 4.65 ± 0.03 4.66 ± 0.13 4.66 ± 0.10 3.85 ± 0.00 4.09 ± 0.01 4.02 ± 0.12
Actinomyces naeslundii II (H403B5) 1 4.69 ± 0.03 4.56 ± 0.00 4.68 ± 0.05 4.57 ± 0.01 4.56 ± 0.00 4.59 ± 0.00
Actinomyces sp. HOT 175 (T20B12) 1 4.75 ± 0.02 4.71 ± 0.02 4.85 ± 0.01 4.55 ± 0.01 4.54 ± 0.01 4.61 ± 0.02
Actinomyces meyeri (H101A17) 2 4.79 ± 0.06 4.45 ± 0.01 4.48 ± 0.00 4.34 ± 0.46 3.75 ± 0.02 3.75 ± 0.04
Actinomyces sp. HOT 181 (H304PA1) 2 4.94 ± 0.01 4.22 ± 0.01 4.21 ± 0.00 4.55 ± 0.05 4.11 ± 0.01 4.12 ± 0.02
Actinomyces massiliensis (T20B27) 2 5.13 ± 0.02 5.15 ± 0.04 5.18 ± 0.03 4.83 ± 0.03 4.92 ± 0.05 4.93 ± 0.05
Actinomyces georgiae (T41B19) 1 5.42 ± 0.10 5.16 ± 0.13 5.15 ± 0.03 4.50 ± 0.05 4.28 ± 0.06 4.34 ± 0.04

Final pH glucose broth: bold = pH <4.0, italic = pH 4 to <4.5, bold italic = pH ≥4.5 Actinomyces strains showed a range of final pH values in media at
initial pH 7 and pH 5.5. Differences were significant between strains at initial pH 7 for all groups (P < 0.01) but not for media at initial pH 5.5 (P < 0.05)
when tested by analysis of variance with Bonferroni adjustment. Strains tested were either reference strains or clinical isolates (Tanner, Kressirer, et al.
2011). Strains for testing were cultured anaerobically on blood agar. For low pH agar growth, medium containing Bacto yeast agar (10 g/L), brain heart
infusion (26 g/L), and tryptic soy agar (20 g/L) was prepared and pH adjusted using 0.2 N HCl to pH 7.0 and pH 5.5. After test strains were inoculated
and incubated anaerobically for 3 to 4 d, growth on agar at pH 5.5 was compared with that at pH 7.0. For acid production, neutralized soya peptone
(5 g/L), trypicase peptone pancreatic digest of casein (3 g/L), and Bacto yeast extract (10 g/L) was prepared without added carbohydrate and filter
sterilized sugars added to final concentrations of 1% glucose, 1% sucrose, 1% fructose, or no added sugar for controls and distributed in flat-bottomed
96-well microtiter plates. Broths were inoculated with suspensions of test strains and incubated anaerobically for 48 h. Final pH values were read with
a microelectrode. Tests were run in triplicate and mean values and standard deviation of the mean calculated.
a
2 = equivalent to growth on agar at pH 7. 1 = less than growth on agar at pH 7.
b
Mean of triplicate readings ± standard deviation of mean.

Next-generation sequencing is a molecular open-ended S. mutans in the Romanian children: 85.7% compared to
approach to investigate the microbiome and was used in paral- 50% in the caries-active Swedish children. Detection of the
lel with PCR to evaluate microbiotas of Swedish and Romanian Streptococcus sobrinus was significantly different between
adolescent populations that had very different access to dental populations at 50% of the Romanian children but not detected
care and caries experience (Johansson et al. 2016). Adolescents in the Swedish population examined. Thus, clear differences in
from Romania, mean age 14.4 y and minimal previous expo- association of mutans streptococci with caries was observed,
sure to dental care, on examination had abundant plaque and a which appeared associated with extent of caries, access to den-
mean decayed missing filling surface (DMFS) score of 20.1. tal care, and lower socioeconomic status in general.
This contrasted from 17-y-old Swedish caries-free and caries- The microbiome assayed by next-generation sequencing
active (mean DMFS 7.5) children who had experienced routine revealed the microbial diversity and microbiomes that charac-
dental care and had minimal detectable dental plaque. In addi- terized caries-active Romanian, caries-active Swedish, and
tion to a higher caries experience in the Romanian than caries-free children beyond S. mutans (Johansson et al. 2016)
Swedish adolescents, there was a higher detection frequency of (Fig. 3). In addition to S. mutans and S. sobrinus, the more
The Caries Microbiome 83

Figure 3.  Microbiome of Romanian (high caries, blue), Swedish (low caries, yellow), and Swedish caries-free adolescents (red). Partial least squares
(PLS) regression modeling using caries grouping, microbiota from pyrosequencing, and sweet snack intake (Johansson et al. 2016). (A) PLS scatterplot
illustrating separate clustering of the low-, high-, and no-caries groups based on the microbiota and diet. (B) PLS loading plot of the microbial variables
separating the clinical groups highlighting the Streptococcus and Actinomyces taxa, the major-acid producing taxa, in each clinical condition and numbers
of taxa detected in each group. Figure modified from Figure 3 in Johansson et al. (2016).

acidogenic microbiota of the higher caries Romanian children was characterized by 8 Actinomyces species, not characteristic
included several nonmutans Streptococcus species. The micro- of the Swedish caries children, and nonmutans streptococci.
biota that characterized the Swedish population with caries Comparison of the caries microbiota of these 2 very different
included, among the more fermentative species, unnamed populations highlights differences in the association of S. mutans
Actinomyces taxa and Streptococcus constellatus. In contrast, in relation to caries, particularly clear since both populations
the acidogenic microbiota of the Swedish caries-free children were monitored at the same time by the same microbiological
84 Advances in Dental Research 29(1)

methods. Clinical and treatment studies suggest the need to References

evaluate the disease-associated microbiota of the population as Aas JA, Griffen AL, Dardis SR, Lee AM, Olsen I, Dewhirst FE, Leys EJ, Paster
well as individuals before testing for risk assessment and treat- BJ. 2008. Bacteria of dental caries in primary and permanent teeth in chil-
ment effects. For caries in the absence of mutans streptococci, dren and young adults. J Clin Microbiol. 46(4):1407–1417.
Becker MR, Paster BJ, Leys EJ, Moeschberger ML, Kenyon SG, Galvin JL,
monitoring for a wide array of species, including Actinomyces Boches SK, Dewhirst FE, Griffen AL. 2002. Molecular analysis of bac-
species, would be warranted. terial species associated with childhood caries. J Clin Microbiol. 40(3):
1001–1009.
Berkowitz RJ, Amante A, Kopycka-Kedzierawski DT, Billings RJ, Feng C.
2011. Dental caries recurrence following clinical treatment for severe early
Conclusion childhood caries. Pediatr Dent. 33(7):510–514.
Berkowitz RJ, Jordan HV. 1975. Similarity of bacteriocins of Streptococcus
The oral microbiota is a complex community that, under the mutans from mother and infant. Arch Oral Biol. 20(11):725–730.
influence of a carbohydrate-frequent diet, produces an acidic Boyar RM, Thylstrup A, Holmen L, Bowden GH. 1989. The microflora associ-
environment on the tooth that can lead to demineralization and ated with the development of initial enamel decalcification below orthodon-
tic bands in vivo in children living in a fluoridated-water area. J Dent Res.
dental caries. Several species in the dysbiosis community of 68(12):1734–1738.
aggressive caries are strongly caries associated, acidogenic, Bratthall D, Hansel Petersson G. 2005. Cariogram—a multifactorial risk assess-
and acid tolerant and are more likely to contribute to the caries ment model for a multifactorial disease. Community Dent Oral Epidemiol.
33(4):256–264.
process than other microbiome residents. Aggressive caries Brown RW. 1952. Dr. Howe and The Forsyth Infirmary. Cambridge (MA):
would likely be amenable to monitoring for selected putative Harvard University Press.
Burne RA, Marquis RE. 2000. Alkali production by oral bacteria and protection
caries pathogens as proposed under the first hypothesis for against dental caries. FEMS Microbiol Lett. 193(1):1–6.
microbial monitoring. These diseases include severe ECC and Caufield PW, Gibbons RJ. 1979. Suppression of Streptococcus mutans in the
the progressing disease characterized by the previously mouths of humans by a dental prophylaxis and topically-applied iodine.
J Dent Res. 58(4):1317–1326.
untreated Romanian adolescents with high caries experience. Corby PM, Lyons-Weiler J, Bretz WA, Hart TC, Aas JA, Boumenna T, Goss J,
Not all caries-microbiotas are similar, however, and the Corby AL, Junior HM, Weyant RJ et al. 2005. Microbial risk indicators of
acidogenic role can be carried out by different species in differ- early childhood caries. J Clin Microbiol. 43(11):5753–5759.
Dewhirst FE, Chen T, Izard J, Paster BJ, Tanner AC, Yu WH, Lakshmanan
ent individuals and populations. Key caries-associated species A, Wade WG. 2010. The human oral microbiome. J Bacteriol. 192(19):
may not be identified easily, and different approaches for 5002–5017.
microbial monitoring would be needed. For this less-marked Downes J, Mantzourani M, Beighton D, Hooper S, Wilson MJ, Nicholson A,
Wade WG. 2011. Scardovia wiggsiae sp. Nov., isolated from the human
dysbiosis, the development of rapid multitaxon assays has the oral cavity and clinical material, and emended descriptions of the genus
potential to provide tools to evaluate the caries-associated Scardovia and Scardovia inopinata. Int J Syst Evol Microbiol. 61(Pt 1):
community and detect posttreatment changes consistent with a 25–29.
Gibbons RJ, van Houte J. 1971. Selective bacterial adherence to oral epithe-
more health-compatible microbiota. Monitoring under these lial surfaces and its role as an ecological determinant. Infect Immun. 3(4):
conditions would be more in line with the second hypothesis 567–573.
with microbial community-based monitoring. Gross EL, Leys EJ, Gasparovich SR, Firestone ND, Schwartzbaum JA, Janies
DA, Asnani K, Griffen AL. 2010. Bacterial 16s sequence analysis of severe
In clinical practice, microbial monitoring would have the caries in young permanent teeth. J Clin Microbiol. 48(11):4121–4128.
highest value in progressing disease to determine appropriate Hardie JM, Thomson PL, South RJ, Marsh PD, Bowden GH, McKee AS,
Fillery ED, Slack GL. 1977. A longitudinal epidemiological study on den-
end points of therapy and provide patient feedback by knowing tal plaque and the development of dental caries—interim results after two
lowered detection of disease-associated taxa. For clinical years. J Dent Res. 56 Spec No:C90–C98.
research, updating knowledge of the caries-associated microbi- Hughes CV, Dahlan M, Papadopolou E, Loo CY, Pradhan NS, Lu SC, Mathney
JM, Bravoco A, Kent RL Jr, Tanner AC. 2012. Aciduric microbiota and
ome is critical prior to developing new anticaries therapies. mutans streptococci in severe and recurrent severe early childhood caries.
Pediatr Dent. 34(2):16–23.
Author Contributions Johansson I, Witkowska E, Kaveh B, Lif Holgerson P, Tanner AC. 2016. The
microbiome in populations with a low and high prevalence of caries. J Dent
A.C.R. Tanner, contributed to conception, design, data acquisi- Res. 95(1):80–86.
Kanasi E, Dewhirst FE, Chalmers NI, Kent R, Jr., Moore A, Hughes CV,
tion, analysis, and interpretation, drafted and critically revised the
Pradhan N, Loo CY, Tanner AC. 2010. Clonal analysis of the microbiota of
manuscript; C.A. Kressirer, S. Rothmiller, contributed to design severe early childhood caries. Caries Res. 44(5):485–497.
and data acquisition, critically revised the manuscript; I. Johansson, Kanasi E, Johansson I, Lu SC, Kressin NR, Nunn ME, Kent R Jr, Tanner AC.
contributed to conception, design, data acquisition, and interpreta- 2010. Microbial risk markers for childhood caries in pediatricians’ offices.
J Dent Res. 89(4):378–383.
tion, critically revised the manuscript; N.I. Chalmers, contributed Kressirer CA, Smith DJ, King WF, Dobeck JM, Starr JK, Tanner AC. 2017.
to design, data acquisition, and analysis, critically revised the Scardovia wiggsiae and its potential role as a caries pathogen. J Oral Biosci.
manuscript. All authors gave final approval and agree to be 59(3):135–141.
Lederberg J. 2003. Of men and microbes. N Perspect Q. 20(3):52–55.
accountable for all aspects of the work. Loesche WJ, Syed SA. 1973. The predominant cultivable flora of carious
plaque and carious dentine. Caries Res. 7(3):201–216.
Acknowledgments Milgrom P, Riedy CA, Weinstein P, Tanner AC, Manibusan L, Bruss J. 2000.
Dental caries and its relationship to bacterial infection, hypoplasia, diet,
This work was conducted with support from the Colgate-Palmolive and oral hygiene in 6- to 36-month-old children. Community Dent Oral
Company and from the William Bingham 2nd Trust. The authors Epidemiol. 28(4):295–306.
Munson MA, Banerjee A, Watson TF, Wade WG. 2004. Molecular analysis
declare no potential conflicts of interest with respect to the author- of the microflora associated with dental caries. J Clin Microbiol. 42(7):
ship and/or publication of this article. 3023–3029.
The Caries Microbiome 85

Nunn ME, Dietrich T, Singh HK, Henshaw MM, Kressin NR. 2009. Prevalence anaerobic microbiota of severe early childhood caries. J Clin Microbiol.
of early childhood caries among very young urban Boston children com- 49(4):1464–1474.
pared with us children. J Public Health Dent. 69(3):156–162. Tanner AC, Milgrom PM, Kent R Jr, Mokeem SA, Page RC, Liao SI, Riedy
Palmer CA, Kent R, Jr., Loo CY, Hughes CV, Stutius E, Pradhan N, Dahlan CA, Bruss JB. 2002. Similarity of the oral microbiota of pre-school children
M, Kanasi E, Arevalo Vasquez SS, Tanner AC. 2010. Diet and caries- with that of their caregivers in a population-based study. Oral Microbiol
associated bacteria in severe early childhood caries. J Dent Res. 89(11): Immunol. 17(6):379–387.
1224–1229. Tanner AC, Sonis AL, Lif Holgerson P, Starr JR, Nunez Y, Kressirer CA,
Preza D, Olsen I, Willumsen T, Boches SK, Cotton SL, Grinde B, Paster BJ. Paster BJ, Johansson I. 2012. White-spot lesions and gingivitis microbiotas
2009. Microarray analysis of the microflora of root caries in elderly. Eur J in orthodontic patients. J Dent Res. 91(9):853–858.
Clin Microbiol Infect Dis. 28(5):509–517. Torlakovic L, Klepac-Ceraj V, Ogaard B, Cotton SL, Paster BJ, Olsen I. 2012.
Socransky SS, Smith C, Martin L, Paster BJ, Dewhirst FE, Levin AE. 1994. Microbial community succession on developing lesions on human enamel.
“Checkerboard” DNA-DNA hybridization. Biotechniques. 17(4):788–792. J Oral Microbiol. 4(1):e16125.
Takahashi N, Nyvad B. 2011. The role of bacteria in the caries process: ecologi- van Houte J. 1994. Role of micro-organisms in caries etiology. J Dent Res.
cal perspectives. J Dent Res. 90(3):294–303. 73(3):672–681.
Tanner AC. 2015. Anaerobic culture to detect periodontal and caries pathogens. van Houte J, Lopman J, Kent R. 1994. The predominant cultivable flora of
J Oral Biosci. 57(1):18–26. sound and carious human root surfaces. J Dent Res. 73(11):1727–1734.
Tanner AC, Kent RL, Jr, Holgerson PL, Hughes CV, Loo CY, Kanasi E, van Houte J, Lopman J, Kent R. 1996. The final pH of bacteria comprising the
Chalmers NI, Johansson I. 2011. Microbiota of severe early childhood car- predominant flora on sound and carious human root and enamel surfaces.
ies before and after therapy. J Dent Res. 90(11):1298–1305. J Dent Res. 75(4):1008–1014.
Tanner AC, Kressirer CA, Faller LL. 2016. Understanding caries from the oral van Houte J, Sansone C, Joshipura K, Kent R. 1991. Mutans streptococci and
microbiome perspective. J Calif Dent Assoc. 44(7):437–446. non-mutans streptococci acidogenic at low pH, and in vitro acidogenic
Tanner AC, Mathney JM, Kent RL, Chalmers NI, Hughes CV, Loo CY, potential of dental plaque in two different areas of the human dentition.
Pradhan N, Kanasi E, Hwang J, Dahlan MA, et al. 2011. Cultivable J Dent Res. 70(12):1503–1507.