Ecological characterization of the colonic microbiota of normal and diarrheic dogs

Abstract

We used terminal restriction fragment polymorphism (T-RFLP) analysis to assess (1) stability of the fecal microbiota in dogs living in environments characterized by varying degrees of exposure to factors that might alter the microbiota and (2) changes in the microbiota associated with acute episodes of diarrhea. Results showed that the healthy canine GI tract harbors potential enteric pathogens. Dogs living in an environment providing minimal exposure to factors that might alter the microbiota had similar microbiotas; the microbiotas of dogs kept in more variable environments were more variable. Substantial changes in the microbiota occurred during diarrheic episodes, including increased levels of Clostridium perfringens, Enterococcus faecalis, and Enterococcus faecium. When diet and medications of a dog having a previously stable microbiota were changed repeatedly, the microbiota also changed repeatedly. Temporal trend analysis showed directional changes in the microbiota after perturbation, a return to the starting condition, and then fluctuating changes over time.

Figure 1

Scheme for sample processing.

Figure 2

T-RFLP analysis of fecal samples from research colony dogs. Panel (a) community composition expressed as % area of peaks under the T-RFLP electropherogram; bacterial groups were combined at the taxonomic level of order. Panel (b) dendrogram based on Bray-Curtis similarities of community composition among the five dogs; numbers at nodes indicate percentage of trees having an equivalent node in jackknife analysis. Panel (c) community richness parameter, Rr. Panel (d) community functional organization, Fo.

Figure 3

T-RFLP analysis of fecal samples from two pets in a single household. Panel (a) community composition expressed as % area of peaks under the T-RFLP electropherogram; bacterial groups were combined at the taxonomic level of order. Panel (b) dendrogram based on Bray-Curtis similarities of community composition among the samples from the two dogs; numbers at nodes indicate percentage of trees having an equivalent node in jackknife analysis. Panel (c) community richness parameter, Rr. Panel (d) community functional organization, Fo. Panel (e) % change parameter for community dynamics.

Figure 4

T-RFLP analysis of fecal samples from a single pet with two diarrheic episodes. Panel (a) community composition expressed as % area of peaks under the T-RFLP electropherogram; bacterial groups were combined at the taxonomic level of order. Panel (b) dendrogram based on Bray-Curtis similarities of community composition among the samples from this dog; numbers at nodes indicate percentage of trees having an equivalent node in jackknife analysis. Panel (c) community richness parameter, Rr. However, Rr is not reported when there was only a single peak ≥1% of the area under the electropherogram (days 168 and 198); the value of Rr was 0.12 on day 172 and 1.5 on day 221. Panel (d) community functional organization, Fo. However, Fo is not reported when there was only a single peak ≥1% of the area under the electropherogram (days 168 and 198). Panel (e) % change parameter for community dynamics. Panel (f) time lag analysis. Arrows below the X-axis indicate the day of onset of an episode of diarrhea.

Figure 5

Q-PCR analysis of bacterial groups in fecal samples from a single pet with two diarrheic episodes. Panel (a) Clostridium group I; Panel (b) Clostridium group XIVa; Panel (c) Bacteroides spp.; Panel (d) E. coli; Panel (e) E. faecium; Panel (f) E. faecalis. R ² values for the standard curves were 0.995, 0.986, 0.956, 0.958, 0.980, and 0.998, respectively. Error bars indicate the standard deviation of triplicate samples. Lack of a bar indicates that no product was detected in 40 cycles. The sample for day 252 was exhausted before Q-PCR for E. faecium, E. faecalis, and E. coli was performed. Population variability (PV) values are shown at the upper right of each panel.