Farm Table says:
Comparative analyses of rumen microbiomes
What is the problem?
Australia’s livestock industries are being confronted with public concerns and challenges to change with respect to resource use, environmental impact, and public health. The biology underpinning many of these challenges is microbiological in cause-and-effect, and perhaps the most pressing challenge presented to our livestock industries by society right now is to reduce methane emissions.
The purpose of this plan is to further develop and use (meta)genomics approaches to understand the greater rumen microbiota populations in livestock, using the datasets produced in Australia and abroad.
What did the research involve?
Total DNA extraction:
-DNA extractions were carried out on rumen samples collected from animals. Mixed rumen fermentations and pure microbial cultures using a modified version of the RBB+C method (Yu & Morrison, 2004).
Total RNA extraction for RNAseq:
-Total RNA was extracted from a 1ml sample by first adding 200 µl of phenol/ethanol (95:5 vol: vol) to preserve RNA transcripts.
-A metagenomic assessment of the rumen microbiome from seven animals using Hiseq Illumina was undertaken at DOE-JGI.
Assembly, gene calling, and annotation:
-The Illumina datasets were subjected to quality clipping for the removal of adapters and low quality regions using Trimmomatic (Bolger et al., 2014), by specifying all Illumina adaptor sequences (allowing 2 seed mismatches, using a 40bp palindrome clip threshold, and a 15 simple clip threshold), as well as non-default parameters of LEADING=3, TRAILING=3, SLIDING WINDOW of width 10 and minimum quality 20.
Classification and binning of assembled contigs 16S rRNA classification Scaffolds from all best idba-ud assemblies were scanned using the algorithm of Huang and colleagues (Huang et al., 2009) to identify 16S rRNA fragments.
-Non-overlapping tetramers (with one tally per reverse complement pair) were counted over 2Kb windows within the scaffolds.
– A random subset of 2.5 million reads was extracted from each sample dataset, and mapped using bowtie2 (–very-sensitive setting) against the scaffolds of the best idba-ud assembly for all the samples.
A PCA (using prcomp in R – no centering or scaling) was conducted on each of the coverage and tetramer matrices separately, as they were on different scales.
Training data for supervised vector machine
-The above 16S rRNA-classified data was used as training data for the SVM. The SVM was fit using the kernlab package in R, with the de novo OTU cluster as the classification class, and the principal components as main effects.
Analysis of microbial diversity from 16S rDNA:
-Using high throughput sequencing platforms and barcode “pyrotagging”, phylogenetic-based methods targeting the 16S rDNA gene were used to deeply characterize the microbial populations present in the rumen and faecal samples of cattle.
Microbial culturing methods:
-Anaerobic culturing media was prepared in culture tubes and batch fermentation vessels for the isolation and purification of bacterial species using the anaerobic techniques of Hungate (Hungate, 1969) as modified by Bryant (Bryant, 1972).
Identification of rRNA
-Fastq files were separated into original samples based on the barcodes used for each library. Each sample was aligned to the reference multiFASTA using Burrows-Wheeler Aligner bwa (‘mem’ algorithm) using the multithreaded option (-t 6).
Assignment to Kegg
-In order to map all reference genomes to Kyoto Encyclopaedia of Genes and Genome (KEGG) pathways, the KEGG orthology database (2011 version) was downloaded, and the genes.pep file was filtered for protein sequences that had at least one kegg pathway assignment and prepared as a blast database.
Differential expression analysis
-The binary alignment map (bam) and GFF files for the mapped transcripts to the reference database were loaded into R (version 3.1.0 (2014-04-10)) using the tracklayer and GenomicRanges libraries.
Gene alignment analysis and qPCR primer design
-For each gene of interest, a separate database was generated including all gene fragments identified from the metagenomic sequencing of low and high animals as well as the gene of interest from representative microbial genomes.
What were the key findings?
Microbial ecology of low and high methane animals:
-The initial stage of the project focused on the interrogation of the microbiome of animals identified by researchers as part of the Beef CRC as producing “Higher” than or “Lower” than predicted methane outputs.
Simulating low and high methane phenotypes through culturing
-Several combinations of bacterial species and methanogenic species were trialed before the development of consistent consortia that exhibited either high or low methane yields.
Metatranscriptomic analysis of “low” and “high” consortia fermentations.
-Extraction of microbial total RNA (all RNA types) was performed on Rhodes grass fermentations for “low” and “high” methane phenotypes at 24 and 30 hrs.
Development of qPCR assay for the monitoring of specific fermentative steps.
-Genes encoding for enzymes considered to be critical for the determination of pathways that contribute to a low methane phenotype were either identified within the literature or designed based on analysis of the metagenomics/transcriptomic data.
– Methane produced from cattle is an end by-product aimed at the removal of excess hydrogen produced from the bacterial conversion of plant polysaccharides into volatile fatty acids for the animal.
Functional genomics (metagenomics)
-Phylogenetic assignment of marker genes against curated datasets accurately identifies to the species level only a small percentage of the microbiota.
Simulating low methane phenotypes to study gene transcriptomics.
-Microbial ecology data and functional analysis of metagenomic data provide evidence of the key species that define a low methane phenotype in cattle and goats.
Metatranscriptomics of complex microbial populations, such as those within the rumen, across large numbers of animals, although feasible, would involve a large expense and generation of more relevant metagenomic and genomic scaffolds to aid in transcript mapping. Quantitative PCR primers have been designed that target multiple genes in three major fermentative pathways observed to have altered expression levels in low methane systems.
The method that advocates these species within cattle are likely to result in lower methane emissions and will revolve around increased starch and soluble sugar content in diets. In addition, these same substrates are likely to increase the abundance of the other important Tammar wallaby gut Succinivibrionaceae species in the cattle rumen.