1. A Preliminary Analysis of the Microbiota of Canine Dental Plaque Introduction Dental plaque is a complex multi species biofilm which has been extensively studied in humans owing to its importance in the aetiology of oral diseases. Relatively little work has been carried out to advance our understanding of the canine oral microbiota and its relationship with canine oral health. Intensive research attention on human dental plaque has led to numerous suggestions and theories about the causes of oral diseases such as periodontal disease. Most of these ideas should be equally applicable to the dental plaque of dogs and other animals, but to make the most of these advances in understanding, the oral microbiota and ecology of the target animal must be suitably characterised. There are many published research papers relating to the oral microbiota of dogs, but frequently they focus on bite wounds, or animal models of human periodontitis, while neglecting to consider its importance to the health of the dog. The aim of this study was to investigate the bacterial composition of canine dental plaque, motivated by the desire to improve oral healthcare for dogs. A culture based approach was used for isolation of bacteria, coupled with 16S rRNA gene sequencing for the identification of recovered bacterial isolates. Materials and Methods Plaque sampling was performed on dogs during routine dental treatment, after scoring the disease status using standard indices adapted for canines. Healthy dogs were sourced from the Waltham Centre for Pet Nutrition, and dogs with periodontal disease were sourced from a local veterinary clinic. Plaque samples were taken from the buccal surface of upper third premolars at the gingival margin (healthy sites) or at the base of the periodontal pocket (diseased sites) and immediately transferred into reduced transport fluid before being transported to the laboratory. The samples were serially diluted and plated out onto Columbia agar base, and anaerobe agar both containing 5 % horse blood before being incubated at 37 o C aerobically with 5 % CO 2, and anaerobically respectively. After 3 - 10 days each distinguishable colony morphotype was isolated and subjected to comparative sequence analysis of the 16S rRNA gene. The 16S rRNA gene was amplified by PCR using the primers 27F and 1492R. DNA sequencing was performed with an ABI 310 Genetic Analyser using the 357F primer for all isolates, and also using the 27F and 530F primers for extended sequencing. Sequences were identified by searching public databases, and grouped according to bacterial genus. Phylogenetic trees were then generated for each group using ClustalX to align sequences, and the PHYLIP DNADist and Neighbour programs to generate trees. Sequences of related bacteria from public databases were also included in the trees, which were used to group the isolated bacteria into phylotypes. The distance separating known species on each tree was used as a guide for the assignment of phylotypes so that they should approximately reflect distinct species. University College London Results (continued) From Table 1 it can be seen that the bacterial composition of dental plaque was different between samples from healthy and diseased sites. In particular, Actinomyces, Corynebacterium, and Neisseria species were abundant at healthy sites (9.2 %, 14.2 %, 11.0 %), while they were rare or not detected at diseased sites. The disease associated microbiota was, however, dominated by Porphyromonas, Prevotella, and Fusobacterium species, which together comprised 87.2 % of the biota. It is also evident that within genera, the predominant species was often different depending upon the disease status of the sample; for example Porphyromonas gulae and Porphyromonas macacae were isolated at high frequency from disease associated samples and low frequency from health associated samples, while the inverse was true of Porphyromonas canoris which was only detected in healthy sites. Discussion Grouping bacteria using phylogenetic trees based on 16S rRNA gene sequences allowed isolates from different samples to be compared and the relatedness of new species and clades to be assessed. A broad range of bacterial species belonging to at least 17 genera were isolated from the dental plaque of 5 dogs. These genera are typical of those found in human dental plaque, including Actinomyces, Porphyromonas, Fusobacterium, Neisseria, and Streptococcus. Identification of the bacteria to species level, however, revealed that the vast majority were not species normally found in the human oral cavity. In addition the proportions of certain genera in the canine plaque did not match those typically found in human plaque; for example Streptococcus species, which are common in human plaque (e.g. S. sanguis), were detected in only one sample, comprising less than 2 % of the total healthy microbiota, and again the species was one not normally found in the human biota (S. suis). The greatest difference between health and disease samples was detected in the Bacteroides, Porphyromonas, and Prevotella species, which were observed to more than double in frequency from 32 % of the biota in health to 78 % in disease. This result is in agreement with the literature which has repeatedly demonstrated an association of these genera with periodontal disease in both humans and dogs. However, Porphyromonas gingivalis, which is widely considered the most important periodontal pathogen in humans, was not detected at all, though it is frequently reported in canine studies. This result suggests that P. gingivalis may not be a normal member of the canine oral microbiota as has been previously reported, but may have been reported in the literature either due to mis-identification, or because the reports were made before species now recognised as distinct were separated. Conclusions The dental plaque biofilm of dogs has a distinct composition of bacterial species compared to human dental plaque, though it shares close similarities at the genus level. Continuing with this approach, and also using molecular methods to detect culture resistant bacteria could help to accurately identify the bacteria associated with oral health and disease in dogs and other animals, thus allowing oral healthcare measures to be more accurately targeted. Comparison of the periodontal pathogens found in different animals could provide a valuable insight into the aetiology of periodontal diseases. Acknowledgements This work was funded by the Biotechnology and Biological Sciences Research Council, and the Waltham Centre for Pet Nutrition. For Oral Health Care Sciences www.eastman.ucl.ac.uk Eastman Dental Institute Isolate identity healthdisease Actinomyces species9.2%<0.1% A. canis - like. Possible new species*6.1%<0.1% A. turicensis - like0.7%<0.1% A. naeslundii / A. bowdenii*0.9%ND A. hordeovulneris0.5%ND A. hordeovulneris0.4%ND A. europaeusND<0.1% A. hyovaginalis - like*0.5%ND Bacteroides, Porphyromonas, and Prevotella species31.9%77.6% Prevotella heparinolytica3.6%ND Prevotella heparinolyticaND18.2% Bacteroides sp.1.9%ND Porphyromonas gulae*4.8%27.7% Porphyromonas canoris*21.1%ND Porphyromonas cansulciND1.3% Porphyromonas endodontalisND0.9% Porphyromonas macacaeND29.6% Por. catoniae - like. Likely new species*0.5%ND Porphyromonas cangingivalisND Campylobacter species2.1%3.0% C. rectus - like. Possible new species*1.8%3.0% C. Curvus0.3%ND Corynebacterium species14.2%ND Corynebacterium sp. Likely new genus*3.9%ND C. felinum*2.4%ND C. jeikium - like. Likely new species*7.5%ND Corynebacterium sp.0.5%ND Fusobacterium and Filifactor species0.7%9.6% Fusobacterium alocisND1.1% Fusobacterium nucleatum - like. Possible new species*0.1%8.5% Filifactor villosum0.6%ND Actinobacillus and Haemophilus species0.6%<0.1% Haemophilus / Actinobacillus - like. Likely new species*0.6%ND H. haemoglobinophilusND<0.1% Haemophilus sp.ND<0.1% not determined7.3%4.3% no sequence available4.0%2.3% not determined - ambiguous or few rel's3.3%1.9% Neisseria species11.0%1.7% N. canis*0.9%ND N. Canis – like, group 1ND1.7% N. canis – like, group 2*9.7%ND N. Weaveri0.5%<0.1% Others23.0%3.6% C. gingivalis - like*0.6%ND Weeksella zoohelcum*1.3%ND G. palaticanis*5.8%ND Leptotrichia - like. Likely new species.*0.7%ND Peptostreptococcus sp.0.0%3.6% M. caviae<0.1%ND P. stomatis / P. dagmatis*4.6%0.0% Cardiobacterium sp.3.2%ND Ultramicrobacterium sp.* S. suis 4.9% 1.9% ND 0.0% Total viable count per ml7.91E+042.12E+07 Table 1. Summary of bacterial species isolated from dental plaque of healthy dogs (n=3) and dogs with periodontal disease (n=2). Species identities are based on closest clustering sequences and BLAST searches on public databases. 0.01 N. denitrificans N. animalis N. elongata cp04.01 cp05.07 N. canis N.macaca N. flavescens N.gonorrhoea N.polysacc Figure 2. Example neighbour joining phylogenetic tree (832 alignment positions) showing two Neisseria species which were selected for extra sequencing. Some isolates were selected for additional sequencing to provide a more accurate phylogenetic classification for the group to which they belong. 0.1 N. dentiae cp62.05 cps01.28 cp06.07 cp01.10 cps01.33 cp08.26 N. weaveri cp08.01 cp09.16 cp17.05 N. flavescens N. polysaccharea N. elongata cp07.09 N. canis cp05.07 cp06.31 cp08.11 cp08.12 cps01.16 cp01.09 cp07.07 cp03.11 cp03.09 cp09.01 cp62.01 cp62.02 cp09.08 cp02.03 cp02.07 cps01.20 cp04.01 cp06.15 cp01.11 cp07.21 cp07.10 cp17.04 cp09.13 cp08.09 cp08.04 cp03.06 cp08.10 cp01.07 cp09.02 cp05.02 Neisseria weaveri group Neisseria elongata group Neisseria canis group Neisseria canis - like Group 1 Neisseria canis - like Group 2 Figure 1. Example neighbour joining phylogenetic tree for partial 16S rRNA gene sequences (250 alignment positions) of Neisseria species isolated from the canine oral cavity of 9 dogs. Key for Figures 1 & 2 Isolate numbers from this study shown in black Isolates from this study chosen for additional sequencing shown in blue Sequences obtained from public databases for comparison shown in red D.R. Elliott 1, D.A. Spratt 1, C. Buckley 2, M-L. Baillon 2, M. Wilson 1 1 Division of Infection and Immunity, Eastman Dental Institute, University College London, United Kingdom 2 Waltham Centre for Pet Nutrition, Waltham-on-the-Wolds, LE14 4RT, United Kingdom. Results Figure 1 shows an example phylogenetic tree as used to group the isolated bacteria for each genus. Longer sequences were obtained for selected sequences and these were used to further characterise some groups by database searches and additional phylogenetic trees as shown in Figure 2. These data were then combined with viable counting results from primary isolations to determine the frequency at which each group was isolated for each sample category (health or disease) as shown in Table 1. ND = Not Detected * Indicates species identities marked which were determined using longer sequences of approximately 900 bases. Other species identities determined using sequences of approximately 300 bases.