- Email: [email protected]
Identifying a healthy oral microbiome through metagenomics
ORIGINAL ARTICLE
10.1111/j.1469-0691.2012.03857.x
Identifying a healthy oral microbiome through metagenomics ´ . Simo´n-Soro1, M. Pignatelli1 and A. Mira1 L. D. Alcaraz1, P. Belda-Ferre1, R. Cabrera-Rubio1, H. Romero2, A 1) Department of Genomics and Health, Center for Advanced Research in Public Health, Avda. Catalun˜a, Valencia, Spain and 2) Laboratorio de Organizacio´n y Evolucio´n del Genoma, Facultad de Ciencias/CURE, Universidad de la Repu´blica, Montevideo, Uruguay
Abstract We present the results of an exploratory study of the bacterial communities from the human oral cavity showing the advantages of pyrosequencing complex samples. Over 1.6 million reads from the metagenomes of eight dental plaque samples were taxonomically assigned through a binning procedure. We performed clustering analysis to discern if there were associations between non-caries and caries conditions in the community composition. Our results show a given bacterial consortium associated with cariogenic and noncariogenic conditions, in agreement with the existence of a healthy oral microbiome and giving support to the idea of dental caries being a polymicrobial disease. The data are coherent with those previously reported in the literature by 16S rRNA amplification, thus giving the chance to link gene functions with taxonomy in further studies involving larger sample numbers. Keywords: Binning, dental caries, metagenomics, probiotic, pyrosequencing Original Submission: 23 November 2011; Accepted: 20 March 2012 Editors: A. Moya, Rafael Canto´n, and D. Raoult Clin Microbiol Infect 2012; 18 (Suppl. 4): 54–57 Corresponding author: A. Mira, Department of Genomics and Health, Center for Advanced Research in Public Health, Avda. Catalun˜a 21, 46020 Valencia, Spain E-mail: [email protected]
potential cariogenic species to Veillonella, Propionibacterium and Atopobium, among others [4], most of which are poorly characterized species.
Introduction
Dental caries, microbiome and pyrosequencing
Unlike most infectious diseases where a single causing agent can be found responsible for the infection, oral diseases appear to be the outcome of multiple microorganisms. In periodontitis, for instance, at least three bacterial organisms have been found to be directly associated with the development of the disease [1]. Similarly, the complexity of the microbial community in the oral cavity has hampered the identification of a single aetiological agent for dental caries. It has been demonstrated that Streptococcus sobrinus and above all S. mutans are acidogenic and play an important role in caries initiation [2]. However, the use of molecular techniques like PCR amplification and cloning of the 16S rRNA gene have revealed that a high proportion of samples from cavities do not contain mutans streptococci, whereas other acid-producing bacteria are present [3]. These include Lactobacillus, Actinomyces or Bifidobacterium. Recent molecular work has confirmed these results and expanded the list of
Dental caries is probably better understood as a polymicrobial disease [5] where the interaction and synergistic effect of multiple species should be taken into account for future strategies of diagnosis, prevention and treatment. Given that a large portion of oral bacteria cannot be cultured by current laboratory techniques, the introduction of molecular approaches has provided a significant improvement in our understanding of oral microbiota. However, PCR amplification and cloning still have significant biases that do not allow microbial diversity to be fully studied, as many species or DNA segments cannot be detected. Thus, a metagenomic approach by which the total DNA from a microbial community is obtained obviating the need for culture or PCR amplification has been proposed as a promising strategy to study the full genetic pool of the human microbiome in health and disease [6]. In addition, the extraordinary increase in sequencing output and the reduction of the associated cost
ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases
CMI
Alcaraz et al. Identifying a healthy oral microbiome through metagenomics
provided by next generation sequencing has been applied to the study of gut microbiota, providing a more complete picture of human-associated bacterial communities [7]. We used a 454 GLX Titanium pyrosequencing approach to obtain over 800 Mbp of DNA sequence from supragingival dental plaque samples from eight individuals who varied in oral health status. Two of them (healthy controls, with no caries) were volunteers who had never suffered from dental caries in their lives and another four samples were from individuals with one, four, eight and 15 cavities at the moment of sampling. In addition, two samples were taken from individual cavities in order to give a first glimpse of the diversity at these diseased sites. Over 2 million pyrosequencing reads of 425 bp average length were analysed by phymmBL [8], a binning method that combines the assignment of sequences by homology and by nucleotide composition using hidden Markov models, thus allowing taxonomic binning and prediction for each single read. All the available complete whole genome sequencing as well as reference genomes for the Human Microbiome Project and the Human Oral Database
55
were used to build a local database to predict taxonomic affiliation. Filtering the reads under 200 bp, we managed to taxonomically assign over 1.6 million reads to the 1150 genomes analysed, with an estimated accuracy at the class level over 75% [8]. When a correspondence analysis was performed with the assigned reads, samples with bad oral health tended to cluster together (Fig. 1). As can be observed in the figure, the principal component separated the two healthy samples and the sample from the individual with a single cavity from the other five samples with dental plaque of individuals with more than four cavities and from the two samples within cavities. Whereas the dental plaque samples from individuals of bad oral health clustered tightly at the positive values of the main axis, the three samples from healthy individuals occupied different positions at the secondary axis. Taken together, the results show hints of a specific microbiota associated with the presence of dental caries, and there appear to be several combinations of bacteria under good oral health, a finding that should be confirmed with larger
FIG. 1. Correspondence analysis of the bacterial diversity in eight oral samples based on the taxonomic assignment of 1.6 million pyrosequencing reads by the binning PhymmBL approach. The first axis successfully separates healthy from diseased individuals. Around the healthy samples some bacterial genera are suggested to be potentially associated with absence of caries. The samples are represented with symbols according to health status: individuals that have never suffered from dental caries are marked with white teeth symbols (samples noca-01p and noca-03p); individuals with one cavity (sample ca1-01p) and four cavities (sample ca1-02p) are marked with grey teeth symbols; individuals with eight and 15 cavities (samples ca-06p and ca-04p, respectively) are marked with black teeth symbols; samples from individual cavities are marked with a black spot within a white tooth and correspond to teeth 1.6 (sample ca-06_1.6) and 4.6 (sample ca-05_4.6), following WHO nomenclature. ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases, CMI, 18 (Suppl. 4), 54–57
56
Clinical Microbiology and Infection, Volume 18 Supplement 4, July 2012
sampling of healthy volunteers. The bacterial genera which are uniquely associated with the absence of caries can be observed in the figure, and include strains with highest similarity to Neisseria, Cardiobacterium, Rothia, Kingella, Aggregatibacter or Mannheimia. Some of these bacteria are poorly known, and include for instance members of the TM7 phylum, a bacterial group which does not have a single member cultured in the laboratory but which appears to be widely present in the oral cavity [9]. Thus, we propose that efforts towards improving culturing media for oral bacteria would be highly recommended to better understand the potential beneficial role of these microbes. In the caries-associated genera we can find Dialister, Oligotropha, Basfia, Parvibaculum, Syntrophus or Treponema, among others. Interestingly the genus Streptococcus, which includes the mutans streptococci traditionally associated with caries, is not associated to a health status in the correspondence analysis, probably reflecting the preventive nature of some other species from this genus and giving new insights into the contribution of other species to the diseased status. Our results agree with those obtained by sequencing of PCR amplified 16S rRNA genes, where a number of studies have found a specific set of bacterial genera associated with non-caries oral conditions [10], including some of the healthrelated bacteria identified in our work. The finding that a given microbial community may be linked to non-caries status supports the idea of using health-associated bacteria as probiotics to prevent oral diseases [11]. The use of healthpromoting bacteria has been successfully applied in pharynx infections by inoculating with bacteriocin-producing commensal strains isolated from healthy individuals and is the basis for replacement therapies to prevent infectious disease in the gut and the oral cavity.
Conclusions The data presented here suggest that a more holistic view of the probiotic approach against dental caries may be needed, as the microbial contribution to good or bad oral health is probably related to bacterial consortia rather than to individual species. It is well established that dental caries is a multifactorial disease where diet, teeth shape and composition, saliva pH or the immune system among others play a role in the tendency to develop the disease. The development of metagenomics and next generation sequencing techniques now allows the contribution of different microbial consortia to oral diseases to be investigated. This is a challenging period in which experimental work should be designed to determine whether a combination of microbial species could be
CMI
successfully transplanted to the oral cavity and form a stable biofilm that prevents cavities and contributes to oral health. Additionally, the microbial consortia associated with cariogenic conditions may also have important consequences for designing preventive strategies against dental caries. Promising results in passive and active immunization against antigens from specific oral pathogens have been obtained in the last decade [12]. However, if oral diseases are polymicrobial, single-species-based immunization strategies may be limited. We have previously proposed the use of surface antigens shared among several oral pathogens as more efficient targets for the design of vaccines against oral diseases [13]. Metagenomic approaches like the one presented here will help to determine the species against which these efforts should be directed.
Acknowledgements This work was funded by the following projects from the Spanish MICINN: SAF2009-13032-C02-02 from the I+D program, BIO2008-03419-E from the EXPLORA program and MICROGEN CSD2009-00006 from the Consolider-Ingenio program, and a CONACyT fellowship to LDA.
Transparency Declaration The authors do not have any potential conflicts of interest to declare.
References 1. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent RL. Microbial complexes in subgingival plaque. J Clin Periodontol 1998; 25: 134–144. 2. Loesche WJ. Role of Streptococcus mutans in human dental decay. Microbiol Rev 1986; 50: 353–380. 3. Corby PM, Lyons-Weiler J, Bretz WA et al. Microbial risk indicators of early childhood caries. J Clin Microbiol 2005; 43: 5753–5759. 4. Aas JA, Griffen AL, Dardis SR et al. Bacteria of dental caries in primary and permanent teeth in children and young adults. J Clin Microbiol 2008; 46: 1407–1417. 5. Kleinberg I. A mixed-bacteria ecological approach to understanding 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: 108–125. 6. Mullany P, Hunter S, Allan E. Metagenomics of dental biofilms. Adv Appl Microbiol 2008; 64: 125–136. 7. Qin J, Li R, Raes J et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010; 464: 59–65. 8. Brady A, Salzberg SL. Phymm , Phymm BL. metagenomic phylogenetic classification with interpolated Markov models. Nat Methods 2009; 6: 673–676.
ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases, CMI, 18 (Suppl. 4), 54–57
CMI
Alcaraz et al. Identifying a healthy oral microbiome through metagenomics
9. Paster BJ, Boches SK, Galvin JL et al. Bacterial diversity in human subgingival plaque. J Bacteriol 2001; 183: 3770–3783. 10. Zaura E, Keijser BJF, Huse SM, Crielaard W. Defining the healthy ‘core microbiome’ of oral microbial communities. BMC Microbiol 2009; 9: 259. 11. Belda-Ferre P, Alcaraz LD, Cabrera-Rubio R et al. The oral metagenome in health and disease. ISME J 2012; 6: 46–56.
57
12. Abiko Y. Passive immunization against dental caries and periodontal disease: development of recombinant and human monoclonal antibodies. Crit Rev Oral Biol Med 2000; 11: 140–158. 13. Mira A. Horizontal gene transfer in oral bacteria. In: Rogers AH, ed. Oral molecular microbiology. Caister Academic Press: Norfolk, UK, 2007; 65–85.
ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases, CMI, 18 (Suppl. 4), 54–57
10.1111/j.1469-0691.2012.03857.x
Identifying a healthy oral microbiome through metagenomics ´ . Simo´n-Soro1, M. Pignatelli1 and A. Mira1 L. D. Alcaraz1, P. Belda-Ferre1, R. Cabrera-Rubio1, H. Romero2, A 1) Department of Genomics and Health, Center for Advanced Research in Public Health, Avda. Catalun˜a, Valencia, Spain and 2) Laboratorio de Organizacio´n y Evolucio´n del Genoma, Facultad de Ciencias/CURE, Universidad de la Repu´blica, Montevideo, Uruguay
Abstract We present the results of an exploratory study of the bacterial communities from the human oral cavity showing the advantages of pyrosequencing complex samples. Over 1.6 million reads from the metagenomes of eight dental plaque samples were taxonomically assigned through a binning procedure. We performed clustering analysis to discern if there were associations between non-caries and caries conditions in the community composition. Our results show a given bacterial consortium associated with cariogenic and noncariogenic conditions, in agreement with the existence of a healthy oral microbiome and giving support to the idea of dental caries being a polymicrobial disease. The data are coherent with those previously reported in the literature by 16S rRNA amplification, thus giving the chance to link gene functions with taxonomy in further studies involving larger sample numbers. Keywords: Binning, dental caries, metagenomics, probiotic, pyrosequencing Original Submission: 23 November 2011; Accepted: 20 March 2012 Editors: A. Moya, Rafael Canto´n, and D. Raoult Clin Microbiol Infect 2012; 18 (Suppl. 4): 54–57 Corresponding author: A. Mira, Department of Genomics and Health, Center for Advanced Research in Public Health, Avda. Catalun˜a 21, 46020 Valencia, Spain E-mail: [email protected]
potential cariogenic species to Veillonella, Propionibacterium and Atopobium, among others [4], most of which are poorly characterized species.
Introduction
Dental caries, microbiome and pyrosequencing
Unlike most infectious diseases where a single causing agent can be found responsible for the infection, oral diseases appear to be the outcome of multiple microorganisms. In periodontitis, for instance, at least three bacterial organisms have been found to be directly associated with the development of the disease [1]. Similarly, the complexity of the microbial community in the oral cavity has hampered the identification of a single aetiological agent for dental caries. It has been demonstrated that Streptococcus sobrinus and above all S. mutans are acidogenic and play an important role in caries initiation [2]. However, the use of molecular techniques like PCR amplification and cloning of the 16S rRNA gene have revealed that a high proportion of samples from cavities do not contain mutans streptococci, whereas other acid-producing bacteria are present [3]. These include Lactobacillus, Actinomyces or Bifidobacterium. Recent molecular work has confirmed these results and expanded the list of
Dental caries is probably better understood as a polymicrobial disease [5] where the interaction and synergistic effect of multiple species should be taken into account for future strategies of diagnosis, prevention and treatment. Given that a large portion of oral bacteria cannot be cultured by current laboratory techniques, the introduction of molecular approaches has provided a significant improvement in our understanding of oral microbiota. However, PCR amplification and cloning still have significant biases that do not allow microbial diversity to be fully studied, as many species or DNA segments cannot be detected. Thus, a metagenomic approach by which the total DNA from a microbial community is obtained obviating the need for culture or PCR amplification has been proposed as a promising strategy to study the full genetic pool of the human microbiome in health and disease [6]. In addition, the extraordinary increase in sequencing output and the reduction of the associated cost
ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases
CMI
Alcaraz et al. Identifying a healthy oral microbiome through metagenomics
provided by next generation sequencing has been applied to the study of gut microbiota, providing a more complete picture of human-associated bacterial communities [7]. We used a 454 GLX Titanium pyrosequencing approach to obtain over 800 Mbp of DNA sequence from supragingival dental plaque samples from eight individuals who varied in oral health status. Two of them (healthy controls, with no caries) were volunteers who had never suffered from dental caries in their lives and another four samples were from individuals with one, four, eight and 15 cavities at the moment of sampling. In addition, two samples were taken from individual cavities in order to give a first glimpse of the diversity at these diseased sites. Over 2 million pyrosequencing reads of 425 bp average length were analysed by phymmBL [8], a binning method that combines the assignment of sequences by homology and by nucleotide composition using hidden Markov models, thus allowing taxonomic binning and prediction for each single read. All the available complete whole genome sequencing as well as reference genomes for the Human Microbiome Project and the Human Oral Database
55
were used to build a local database to predict taxonomic affiliation. Filtering the reads under 200 bp, we managed to taxonomically assign over 1.6 million reads to the 1150 genomes analysed, with an estimated accuracy at the class level over 75% [8]. When a correspondence analysis was performed with the assigned reads, samples with bad oral health tended to cluster together (Fig. 1). As can be observed in the figure, the principal component separated the two healthy samples and the sample from the individual with a single cavity from the other five samples with dental plaque of individuals with more than four cavities and from the two samples within cavities. Whereas the dental plaque samples from individuals of bad oral health clustered tightly at the positive values of the main axis, the three samples from healthy individuals occupied different positions at the secondary axis. Taken together, the results show hints of a specific microbiota associated with the presence of dental caries, and there appear to be several combinations of bacteria under good oral health, a finding that should be confirmed with larger
FIG. 1. Correspondence analysis of the bacterial diversity in eight oral samples based on the taxonomic assignment of 1.6 million pyrosequencing reads by the binning PhymmBL approach. The first axis successfully separates healthy from diseased individuals. Around the healthy samples some bacterial genera are suggested to be potentially associated with absence of caries. The samples are represented with symbols according to health status: individuals that have never suffered from dental caries are marked with white teeth symbols (samples noca-01p and noca-03p); individuals with one cavity (sample ca1-01p) and four cavities (sample ca1-02p) are marked with grey teeth symbols; individuals with eight and 15 cavities (samples ca-06p and ca-04p, respectively) are marked with black teeth symbols; samples from individual cavities are marked with a black spot within a white tooth and correspond to teeth 1.6 (sample ca-06_1.6) and 4.6 (sample ca-05_4.6), following WHO nomenclature. ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases, CMI, 18 (Suppl. 4), 54–57
56
Clinical Microbiology and Infection, Volume 18 Supplement 4, July 2012
sampling of healthy volunteers. The bacterial genera which are uniquely associated with the absence of caries can be observed in the figure, and include strains with highest similarity to Neisseria, Cardiobacterium, Rothia, Kingella, Aggregatibacter or Mannheimia. Some of these bacteria are poorly known, and include for instance members of the TM7 phylum, a bacterial group which does not have a single member cultured in the laboratory but which appears to be widely present in the oral cavity [9]. Thus, we propose that efforts towards improving culturing media for oral bacteria would be highly recommended to better understand the potential beneficial role of these microbes. In the caries-associated genera we can find Dialister, Oligotropha, Basfia, Parvibaculum, Syntrophus or Treponema, among others. Interestingly the genus Streptococcus, which includes the mutans streptococci traditionally associated with caries, is not associated to a health status in the correspondence analysis, probably reflecting the preventive nature of some other species from this genus and giving new insights into the contribution of other species to the diseased status. Our results agree with those obtained by sequencing of PCR amplified 16S rRNA genes, where a number of studies have found a specific set of bacterial genera associated with non-caries oral conditions [10], including some of the healthrelated bacteria identified in our work. The finding that a given microbial community may be linked to non-caries status supports the idea of using health-associated bacteria as probiotics to prevent oral diseases [11]. The use of healthpromoting bacteria has been successfully applied in pharynx infections by inoculating with bacteriocin-producing commensal strains isolated from healthy individuals and is the basis for replacement therapies to prevent infectious disease in the gut and the oral cavity.
Conclusions The data presented here suggest that a more holistic view of the probiotic approach against dental caries may be needed, as the microbial contribution to good or bad oral health is probably related to bacterial consortia rather than to individual species. It is well established that dental caries is a multifactorial disease where diet, teeth shape and composition, saliva pH or the immune system among others play a role in the tendency to develop the disease. The development of metagenomics and next generation sequencing techniques now allows the contribution of different microbial consortia to oral diseases to be investigated. This is a challenging period in which experimental work should be designed to determine whether a combination of microbial species could be
CMI
successfully transplanted to the oral cavity and form a stable biofilm that prevents cavities and contributes to oral health. Additionally, the microbial consortia associated with cariogenic conditions may also have important consequences for designing preventive strategies against dental caries. Promising results in passive and active immunization against antigens from specific oral pathogens have been obtained in the last decade [12]. However, if oral diseases are polymicrobial, single-species-based immunization strategies may be limited. We have previously proposed the use of surface antigens shared among several oral pathogens as more efficient targets for the design of vaccines against oral diseases [13]. Metagenomic approaches like the one presented here will help to determine the species against which these efforts should be directed.
Acknowledgements This work was funded by the following projects from the Spanish MICINN: SAF2009-13032-C02-02 from the I+D program, BIO2008-03419-E from the EXPLORA program and MICROGEN CSD2009-00006 from the Consolider-Ingenio program, and a CONACyT fellowship to LDA.
Transparency Declaration The authors do not have any potential conflicts of interest to declare.
References 1. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent RL. Microbial complexes in subgingival plaque. J Clin Periodontol 1998; 25: 134–144. 2. Loesche WJ. Role of Streptococcus mutans in human dental decay. Microbiol Rev 1986; 50: 353–380. 3. Corby PM, Lyons-Weiler J, Bretz WA et al. Microbial risk indicators of early childhood caries. J Clin Microbiol 2005; 43: 5753–5759. 4. Aas JA, Griffen AL, Dardis SR et al. Bacteria of dental caries in primary and permanent teeth in children and young adults. J Clin Microbiol 2008; 46: 1407–1417. 5. Kleinberg I. A mixed-bacteria ecological approach to understanding 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: 108–125. 6. Mullany P, Hunter S, Allan E. Metagenomics of dental biofilms. Adv Appl Microbiol 2008; 64: 125–136. 7. Qin J, Li R, Raes J et al. A human gut microbial gene catalogue established by metagenomic sequencing. Nature 2010; 464: 59–65. 8. Brady A, Salzberg SL. Phymm , Phymm BL. metagenomic phylogenetic classification with interpolated Markov models. Nat Methods 2009; 6: 673–676.
ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases, CMI, 18 (Suppl. 4), 54–57
CMI
Alcaraz et al. Identifying a healthy oral microbiome through metagenomics
9. Paster BJ, Boches SK, Galvin JL et al. Bacterial diversity in human subgingival plaque. J Bacteriol 2001; 183: 3770–3783. 10. Zaura E, Keijser BJF, Huse SM, Crielaard W. Defining the healthy ‘core microbiome’ of oral microbial communities. BMC Microbiol 2009; 9: 259. 11. Belda-Ferre P, Alcaraz LD, Cabrera-Rubio R et al. The oral metagenome in health and disease. ISME J 2012; 6: 46–56.
57
12. Abiko Y. Passive immunization against dental caries and periodontal disease: development of recombinant and human monoclonal antibodies. Crit Rev Oral Biol Med 2000; 11: 140–158. 13. Mira A. Horizontal gene transfer in oral bacteria. In: Rogers AH, ed. Oral molecular microbiology. Caister Academic Press: Norfolk, UK, 2007; 65–85.
ª2012 The Authors Clinical Microbiology and Infection ª2012 European Society of Clinical Microbiology and Infectious Diseases, CMI, 18 (Suppl. 4), 54–57