The microbiome associated with equine periodontitis and oral health

Abstract

Equine periodontal disease is a common and painful condition and its severe form, periodontitis, can lead to tooth loss. Its aetiopathogenesis remains poorly understood despite recent increased awareness of this disorder amongst the veterinary profession. Bacteria have been found to be causative agents of the disease in other species, but current understanding of their role in equine periodontitis is extremely limited. The aim of this study was to use high-throughput sequencing to identify the microbiome associated with equine periodontitis and oral health. Subgingival plaque samples from 24 horses with periodontitis and gingival swabs from 24 orally healthy horses were collected. DNA was extracted from samples, the V3-V4 region of the bacterial 16S rRNA gene amplified by PCR and amplicons sequenced using Illumina MiSeq. Data processing was conducted using USEARCH and QIIME. Diversity analyses were performed with PAST v3.02. Linear discriminant analysis effect size (LEfSe) was used to determine differences between the groups. In total, 1308 OTUs were identified and classified into 356 genera or higher taxa. Microbial profiles at health differed significantly from periodontitis, both in their composition (p < 0.0001, F = 12.24; PERMANOVA) and in microbial diversity (p < 0.001; Mann-Whitney test). Samples from healthy horses were less diverse (1.78, SD 0.74; Shannon diversity index) and were dominated by the genera Gemella and Actinobacillus, while the periodontitis group samples showed higher diversity (3.16, SD 0.98) and were dominated by the genera Prevotella and Veillonella. It is concluded that the microbiomes associated with equine oral health and periodontitis are distinct, with the latter displaying greater microbial diversity.

Figure 1

Two-dimensional ordination of equine microbial profiles at oral health and periodontal disease by principal component analysis (PCA). The OTU data was subsampled at 16 000 reads/sample and log2-transformed before the PCA.

Figure 2

Diversity analysis on equine microbial profiles at health and periodontal disease. A Observed species richness or number of OTUs/sample; B estimated species richness or Chao-1 and C Shannon diversity index. Healthy microbiomes were statistically significantly less diverse than microbiomes with periodontal disease (p < 0.001, Mann–Whitney test).

Figure 3

Visualisation of most significant taxa (genus or higher level) that differentiate between health and periodontal disease in equine microbiomes. In total 107 taxa were significantly different. Only 69 taxa that had LDA scores above 3 are shown. The taxa are ranked by the effect size in LEfSe.

Figure 4

The relative abundance of most discriminative genera between health and disease. A Gemella and B Actinobacillus in health; C Prevotella and D Veillonella in periodontitis, based on LDA scores in LEfSe. Values are expressed as a percentage.

Figure 5

Taxonomic representation of statistically significant differences between the healthy and periodontitis samples at family or higher taxonomic level. All 51 significantly different taxa had LDA scores above 3. Differences are represented in colour (shades of red periodontal disease, green health).