16S rRNA (gene) based barcoded Illumina sequencing of the faecal microbiota of healthy human volunteers

To assess intraindividual and interindividual variability in human gut microbiota within one week, we determined the microbiota composition in the faeces of 10 healthy human volunteers. Each volunteer collected 3 faecal samples on consecutive days (moment 1, 2 or 3) within one week (30 samples). There was no (dietary) intervention involved. The faecal samples were preprocessed, followed by genomic DNA extraction. Two preprocessing methods were applied to the faecal samples. First, frozen samples were crushed with a dead blow hammer into smaller particles. During the hammering, samples were kept frozen using dry ice and liquid nitrogen. The hammered sample was stored at minus 80 degrees, and part of the hammered faeces was additionally mill homogenised in a IKA A11 Basic Analytical Mill. Genomic DNA were extracted from the faecal samples as follows: for all 30 samples, DNA was extracted from the hammered only samples. For comparison between preprocessing methods, only the first collected faecal sample (moment 1) was further additionally analysed using the faecal mill homogenised sample (+10 additional samples). Additionally, ten technical replicates were included in the dataset (the same sample subjected to sequencing twice, +10 additional samples). Therefore, the total dataset contains 50 samples.DNA was isolated from approximately 250 mg faeces using the DNeasy Powersoil Pro Kit (Qiagen, Venlo, the Netherlands) according to manufacturers instructions. Microbiota composition was determined via sequencing of the variable V4 region of the 16S rRNA gene. Triplicate PCR reactions were performed in 35 uL, containing 7 uL 5x Phusion Green HF buffer, 0.7 uL 10 mM dNTPs (Promega, Madison, USA), 0.4 uL Phusion hot start II DNA polymerase (2 U/uL), 25.5 uL nuclease free water, 0.7 uL of extracted template DNA (20 ng/uL) and 0.7 uL of each of the barcoded primers 515F (doi.org/10.1111/1462-2920.13023) and 806R (doi.org/10.3354/ame01753), in concentrations 10 uM. The primers were attached to in house barcodes (barcode strategy: doi.org/10.12688/f1000research.9227.2). Cycling conditions were as follows: 98 degrees 30 s, 25 cycles of 98 degrees 10 s, 50 degrees 10 s, 72 degrees 10 s, and 72 degrees for 7 min. Pooled PCR products were checked on a 1.3 percent agarose gel, and purified using magnetic beads (MagBio Genomics Inc., Gaithersburg, USA). PCR product concentrations were measured using Qubit dsDNA BR buffer and dye (Invitrogen, California, USA), on a DS11 FX fluorometer (DeNovix, Wilmington, USA). Afterwards, a library containing an equimolar mix (200 ng each) of purified PCR products was prepared. The amplicon length including barcodes, and primers was 307 basepairs. The PCR products were subjected to paired end P150 sequencing using the Illumina Novaseq6000 platform (Novogene, Cambridge, UK). The sequencing data was processed and demultiplexed using NGTax 2.0 using default settings (doi.org/10.12688/f1000research.9227.2). Raw sequencing data were demultiplexed by trimming sample specific barcodes and primer sequences. The SILVA reference database version 138 was used to assign taxonomy.