Chapter 2.4.2 : Microbial Community Analysis Using High-Throughput Amplicon Sequencing

PCR and sequencing workflow for a one-step PCR amplification protocol to generate sequence-ready amplicon libraries. Shown are the Earth Microbiome Project (EMP) primers for amplification and sequencing of the V4 variable region of microbial 16S rRNA genes on the Illumina sequencing platform. doi:10.1128/9781555818821.ch2.4.2.f1

PCR and sequencing workflow for a two-step PCR amplification protocol to generate sequence-ready amplicon libraries. Shown are the Earth Microbiome Project (EMP) primers with Fluidigm common sequence linkers for amplification and sequencing of the V4 variable region of microbial 16S rRNA genes on the Illumina sequencing platform. Two separate PCR stages are required to generate template-specific amplicons, and to attach sample-specific barcodes and sequencing adapters. doi:10.1128/9781555818821.ch2.4.2.f2

Comparison of 16S rRNA gene of Staphylococcus aureus (top strand; accession number L37597) and S. epidermidis strain DSM20044 T (bottom strand; accession number LN681574). A basic local alignment search tool (BLAST) analysis was performed to compare the two gene sequences. Mismatches are highlighted in red. The region of comparison begins at the 5′ end of the molecule, directly after the commonly used domain-level primer 27F, and ends just before the commonly used domain-level primer 1492R. Common primer sequences, highlighted in green, include 341F, 534R, 806R, 907R, 1114R, and 1392R (from left to right). Approximate locations of the nine microbial variable regions are indicated. Note that no differences exist between the two sequences in the V4 region, commonly targeted by the 515F/806R primer set. doi:10.1128/9781555818821.ch2.4.2.f3

Relationship between total PCR cycles and presence of putative chimeric sequences in the output amplicon sequence data. A single gDNA sample from mammalian feces (chinchilla) was PCR amplified using a two-step PCR protocol, utilizing the EMP primers and Fluidigm common sequence linkers. A systematic variation of first-stage and second-stage PCR cycles was performed, with a range of 20–36 cycles total. After sequencing on an Ion Torrent PGM, sequence data were demultiplexed and chimeras were identified using usearch61. A linear relationship between total cycle number and chimera formation (R 2 = 0.65) was observed. doi:10.1128/9781555818821.ch2.4.2.f4

Relationship between OTU creation and similarity threshold. The number of identified OTUs generated at each similarity threshold is standardized by the number of OTUs identified at 100% similarity (solid line). The percent change in number of OTU at each threshold was also calculated (dashed line) and shows that the greatest change occurs from a 98% to 97% threshold cutoff. The data are derived from six amplicon data sets generated on a Roche 454 pyrosequencer. doi:10.1128/9781555818821.ch2.4.2.f5

Effect of single-sample versus complete study clustering on observed microbial diversity. A consistent decrease in the number of measured OTUs was observed when sample diversity was estimated based on total data set clustering as opposed to one-by-one sample clustering and diversity calculations. The four data sets, from freshwater (DB1), seawater (DB2), marine sediments (DB3), and microbial mats (DB4), consisted of 10–40 samples each, with amplicons generated on a Roche 454 pyrosequencer. doi:10.1128/9781555818821.ch2.4.2.f6