Metagenomic Analysis of 16S Clostridium perfringens Amplicons Corroborates C. perfringens Counts on Select Agar and C. perfringens PCR Analyses of Bacteria in Broiler Farm Litter

摘要

The purpose of this study was twofold-first, to determine whether analysis of bacterial 16S ribosomal RNA (rRNA) in poultry litter corroborated standard Clostridium perfringens counts and PCR assay, and second, to find whether a correlation between 16S rRNA analysis and netB or Tpel toxin PCR intensity with chick mortality existed. At three time points of growout (0, 2, and 4 wk) litter samples were collected from 23 broiler houses representing eight farms during a coccidiosis vaccine control program. DNA extracted from these samples was used for microbiota determination by sequencing the hypervariable V3V4 region of bacterial 16s rRNA. Obtained sequences were analyzed by QIIME 2 and the Greengenes database for taxonomic composition and relative abundance of C. perfringens in the litter bacterial population. Clostridium perfringens counts on select agar and semiquantitative PCR for C. perfringens were compared with 16S analysis for equivalence testing. Relative abundance of C. perfringens estimated by 16S analysis and semiquantitative PCR for netB and Tpel toxin DNA were analyzed by Pearson linear correlation and statistical equivalence analyses with cumulative chick mortality at 4 and 9 wk growout. When data from all time points were combined, abundance estimates by C. perfringens 16S were statistically equivalent (a = 0.10) to both C. perfringens PCR and C. perfringens counts. Yet, no correlations were observed between any estimate of C. perfringens abundance and cumulative percent chick mortality at 4 or 9 wk growout. However, correlation analyses revealed a significant linear relationship between netB signal at 0 wk (r = 0.55) and 4 wk (r = 0.46) and cumulative mortality at 9 wk growout (P, 0.05). Similarly, abundance of Tpel at 0 and 2 wk showed a linear relationship with cumulative percent mortality at both 4 and 9 wk growout (0.44 <= r <= 0.54, P < 0.05). No correlations were observed between any other genera or species determined by 16S and cumulative percent chick mortality.