Location: Forage-animal Production ResearchTitle: Evaluation of PCR-DGGE as a method to recapitulate host phylogeny by fecal microbial community fingerprint Author
Submitted to: Congress on Gastrointestinal Function
Publication Type: Abstract Only
Publication Acceptance Date: 3/19/2013
Publication Date: 4/16/2013
Citation: Flythe, M.D., Lawrence, L.M., Strasinger, L.A., Fisher, T.M., Gellin, G.L., Harlow, B.E. 2013. Evaluation of PCR-DGGE as a method to recapitulate host phylogeny by fecal microbial community fingerprint. Congress on Gastrointestinal Function. 67-68. Interpretive Summary:
Technical Abstract: Background: Recent studies indicate that host animal could be the primary factor determining the composition of the gastrointestinal microbiome. If host phenotype dictates microbiome composition, then composition should recapitulate host phylogeny. Polymerase chain reaction-denaturing gradient gel electrophoresis (PCR-DGGE) was evaluated as a means to compare the fecal microbiomes of horses to those of other herbivores. The hypothesis was that the PCR-DGGE banding patterns from horses would be more similar to each other than to those of other animals. Methods: Fecal samples were collected from horses (Equus ferus caballus; n=30), Hartmann’s mountain zebras (Equus zebra hartmannae; n=2), white rhinocerous (Ceratotherium simum; n=2), Malayan tapirs (Tapirus indicus; n=2), domestic goats (Capra aegarus hircus; n=2), African elephant (Loxodonta africana; n=1) and Asian elephant (Elephas maximus; n=1). PCR was performed on extracted DNA to amplify the 16S rRNA gene (primers: 341F-GC, 901R). Amplicons were separated on an acrylamide gel (urea-formamide 40-60%, 56° C, 69 V, 17 h). Gel images were normalized with external standards included in every gel and analyzed with BioNumerics software (AppliedMaths). The similarities of lanes were compared as both bands and densitometric curves. Results: Band-based analysis (Dice’s algorithm, UPGMA) indicated little similarity between the two zebra samples (31.9%) and between the two rhinocerous samples (47.1%), which resulted in mixed-species groups. However, in curve-based analysis (Pearson’s r, UPGMA) each sample was most similar to a sample from the same species. One exception was a horse sample that grouped with zebra. The highest similarity between any two samples was between two horses (94.8%). Overall similarity between all samples was 7.5%. Statistical resampling (Jackknife, mean similarity) indicated confidence values of 76.7% and 93.3% for associating horse samples with the correct species and Order, respectively. Conclusions: These results indicate that comparisons based on densitometric curves are more useful than binary data from called bands when evaluating fecal microbial communities with PCR-DGGE. Note that the error in the curve-based analysis was a horse sample that grouped with zebras, the two most closely related host species in the study.