The effects of freezing on faecal microbiota as determined using MiSeq sequencing and culture-based investigations

Abstract

Background: High-throughput sequencing has enabled detailed insights into complex microbial environments, including the human gut microbiota. The accuracy of the sequencing data however, is reliant upon appropriate storage of the samples prior to DNA extraction. The aim of this study was to conduct the first MiSeq sequencing investigation into the effects of faecal storage on the microbiota, compared to fresh samples. Culture-based analysis was also completed.

Methods: Seven faecal samples were collected from healthy adults. Samples were separated into fresh (DNA extracted immediately), snap frozen on dry ice and frozen for 7 days at -80°C prior to DNA extraction or samples frozen at -80°C for 7 days before DNA extraction. Sequencing was completed on the Illumina MiSeq platform. Culturing of total aerobes, anaerobes and bifidobacteria was also completed.

Results: No significant differences at phylum or family levels between the treatment groups occurred. At genus level only Faecalibacterium and Leuconostoc were significantly different in the fresh samples compared to the snap frozen group (p = 0.0298; p = 0.0330 respectively). Diversity analysis indicated that samples clustered based on the individual donor, rather than by storage group. No significant differences occurred in the culture-based analysis between the fresh, snap or -80°C frozen samples.

Conclusions: Using the MiSeq platform coupled with culture-based analysis, this study highlighted that limited significant changes in microbiota occur following rapid freezing of faecal samples prior to DNA extraction. Thus, rapid freezing of samples prior to DNA extraction and culturing, preserves the integrity of the microbiota.

Fig 1. Microbiological enumeration of total aerobes (A), total anaerobes (B) and bifidobacteria (C) in log CFUg-1 faeces from fresh (blue), -80°C frozen (green) and snap samples (red).

RM one-way ANOVA statistical testing showed no significant difference between storage conditions fresh, frozen at -80°C or snap frozen for the three groups tested.

Fig 2. Relative abundances of bacterial phyla in the fresh (blue), snap (red) and -80°C frozen samples (green).

Other contains phyla present at < 1% of assignable sequences at phylum level.

Fig 3. Relative abundances of bacterial genera in the fresh, snap and frozen samples.

Statistically significant differences in genera are indicated with an asterisk (*) (p<0.05). The Other category contains all other genera present at <0.01% of assignable reads at genus level. No significant differences in any of these genera between the 3 groups was found using non-parametric Kruskall-Wallis analysis, where statistical significance was accepted as p<0.05, adjusted for ties.

Fig 4. Visualisation of the PCoA analysis based on unweighted UniFrac distance matrices.

Results indicate samples separate by subject ID rather than storage method. Section A: samples by subject ID. Section B: samples by storage method.