Although the reverse PreGen4 primer had sequence mismatches with all the Bacteroides sequences, six sequences had 9–11 consecutive matching
sequences at the 3′ end (data not shown). Thus, the PreGen4 primers may potentially result in the nonspecific amplification of Bacteroides sequences described above. CH5424802 chemical structure Therefore, from the in silico analysis it was concluded that g-Prevo primers are more specific to ruminal Prevotella than PreGen4 primers. Based on their valid coverage and high specificity to ruminal Prevotella, the g-Prevo primers were selected to be used in this study. Real-time PCR quantification of Prevotella revealed that the relative abundance of this genus in the total rumen bacteria of sheep was as high as 19.7% (Table 3). On the other hand, the commonly cultivated ruminal Prevotella species, P. bryantii and P. ruminicola,
accounted for only 0.6% and 3.8%, respectively (Table 3). The relative abundance of Prevotella tended to increase when the animals were switched to a concentrate diet, although one animal showed no difference in the proportion of Prevotella in either diet (data not shown). In order to demonstrate the proportion of classical ruminal bacterial species, the relative abundance of individual species was aggregated (Table 3). The sum of the relative abundance this website of 12 representative rumen bacterial species in the two dietary conditions accounted for 2.4–4.9% of the total rumen bacteria. The relative abundance of the rumen fibrolytic species (F. succinogenes, R. albus and R. flavefaciens) tended to decrease in concentrate-fed sheep. In particular, the abundance of F. succinogenes decreased significantly (P<0.001)
when the sheep were fed a high-concentrate diet. The DGGE fingerprints of rumen Prevotella from the same diet showed a similar banding pattern and tended to cluster according to the diet, although a certain degree of animal-to-animal variation was observed (Fig. 1). The DGGE fingerprints revealed unique bands for either the hay or the concentrate diet, although there were common banding positions for the two dietary conditions. Comparative analysis of the ADP ribosylation factor DGGE profile across diet showed consistently more bands in samples from hay-fed animals (Fig. 1). A total of 139 16S rRNA gene sequences, 60 from sheep on a hay diet and 79 from sheep on a concentrate diet, were subjected to sequence analysis after discarding those suspected to be chimeras. Good’s coverages of the hay and concentrate libraries were 43.3% and 65.8%, respectively. Although the libraries were not comprehensive, we obtained diverse sequences of Prevotella, and diet specificity was supported by both DGGE and library analysis. Based on a 97% sequence similarity criterion (Stackebrandt & Goebel, 1994), only 17 clones (12.2%) from the two libraries were considered to represent the characterized rumen Prevotella species (P. ruminicola or P. bryantii) and the remaining 122 clones (87.