The composition of the perinatal intestinal microbiota in horse

Abstract

The establishment of the intestinal microbiota is critical for the digestive and immune systems. We studied the early development of the rectal microbiota in horse, a hindgut fermenter, from birth until 7 days of age, by qPCR and 16S rRNA gene amplicon sequencing. To evaluate initial sources of the foal microbiota, we characterised dam fecal, vaginal and oral microbiotas. We utilised an amplicon sequence variant (ASV) pipeline to maximise resolution and reproducibility. Stringent ASV filtering based on prevalence and abundance in samples and controls purged contaminants while preserving intestinal taxa. Sampled within 20 minutes after birth, rectal meconium contained small amounts of diverse bacterial DNA, with a profile closer to mare feces than mouth. 24 hours after birth, rectum was colonised by Firmicutes and Proteobacteria, some foals dominated by single genera. At day 7, the rectal genera were still different from adult feces. The mare vaginal microbiota contributed to 24 h and 7 day microbiotas. It contained few lactobacilli, with Corynebacterium, Porphyromonas, Campylobacter and Helcococcus as the most abundant genera. In the oral mucosa, Gemella was extremely abundant. Our observations indicate that bacteria or bacterial components are present in the intestine immediately after birth, but the newborn microbiota changes rapidly.

Figure 1

16S rRNA gene copy numbers per sample. Blue colour indicates the negative controls and green the foal rectal samples. The boxes represent the interquartile ranges (IQR) containing 50% of samples. The horizontal line in a box indicates the median. Whiskers show maxima and minima within 1.5 × IQR. The circle indicates an outlier.

Figure 2

Accepted and rejected 16S rRNA gene sequence reads per sample. Accepted reads are indicated as blue. Deleted reads are indicated as yellow-orange, with reads classified as Ralstonia in orange. Seven 0 h samples and two 24 h samples were excluded from further analysis due to low quality (red bars). Negative control data processed with the 0 h foal data is shown.

Figure 3

Average microbiota compositions of foal rectal samples and mare fecal, vaginal and oral samples. Main colours indicate bacterial phyla. Within phyla, shades indicate individual genera. The lightest shades of each phylum show the combined abundance of the least abundant genera (<0.5% of total).

Figure 4

Shannon diversity indices of foal rectal samples and mare fecal, vaginal and oral samples. Boxplot as in Fig. 1.

Figure 5

Comparison of 0 h foal and mare microbiotas. (A) PCoA on unweighted UniFrac distances of 0 h rectal microbiota and mare fecal, oral and vaginal microbiotas from the ASV data. Colours and shapes indicate the sample types. (B) ASVs shared between 0 h foal rectal microbiota samples and various mare microbiota samples. ASVs detected in at least two animals per sample group are included.

Figure 6

Comparison of horse fecal microbiotas at various ages. (A) PCoA on unweighted UniFrac of rectal microbiota in the 0 h, 24 h and 7 d old foals and adult mares from ASV data. Colours and shapes indicate the sample types. (B) ASVs shared between rectal microbiota samples at various ages. All ASVs detected in any animal are included. Therefore, the numbers are larger than the mean ASV counts presented in Table 1.