Comparison of stool versus rectal swab samples and storage conditions on bacterial community profiles

Abstract

Background: Sample collection for gut microbiota analysis from in-patients can be challenging. Collection method and storage conditions are potential sources of variability. In this study, we compared the bacterial microbiota from stool stored under different conditions, as well as stool and swab samples, to assess differences due to sample storage conditions and collection method.

Methods: Using bacterial 16S rRNA gene sequence analysis, we compared the microbiota profiles of stool samples stored and collected under various conditions. Stool samples (2 liquid, 1 formed) from three different patients at two hospitals were each evaluated under the following conditions: immediately frozen at -80°C, stored at 4°C for 12-48 hours before freezing at -80°C and stored at -20°C with 1-2 thaw cycles before storage at -80°C. Additionally, 8 stool and 30 rectal swab samples were collected from 8 in-patients at one hospital. Microbiota differences were assessed using the Yue and Clayton dissimilarity index (θ_YC distance) and analysis of molecular variance (AMOVA).

Results: Regardless of the storage conditions, the bacterial communities of aliquots from the same stool samples were very similar based on θ_YC distances (median intra-sample θ_YC distance: 0.035, IQR: 0.015-0.061) compared to aliquots from different stool samples (median inter-sample θ_YC distance: 0.93, IQR: 0.85-0.97) (Wilcoxon test p-value: <0.0001). For the stool and rectal swab comparison, samples from different patients, regardless of sample collection method, were significantly different (AMOVA p-values: <0.001-0.029) compared to no significant difference between all stool and swab samples (AMOVA p-value: 0.976). The θ_YC dissimilarity index between swab and stool samples was significantly lower within individuals (median 0.17, IQR: 0.10-0.27) than between individuals (median 0.93, IQR: 0.85-0.97) (Wilcoxon test p-value: <0.0001), indicating minimal differences between stool and swab samples collected from the same individual over the sampling period.

Conclusion: For gastrointestinal microbiota studies based on bacterial 16S rRNA gene sequence analysis, interim stool sample storage at 4 °C or -20 °C, rather than immediate storage at -80 °C, does not significantly alter results. Additionally, stool and rectal swab microbiotas from the same subject were highly similar, indicating that these sampling methods could be used interchangeably to assess the community structure of the distal GI tract.

Keywords: 16S rRNA gene sequences; Gastrointestinal tract; Gut microbiota; Microbial community; Microbiota; Rectal swab; Stool.

Fig. 1

Bacterial community composition of stool samples subjected to various temperature storage conditions. The relative abundances of 16S rRNA gene sequences (V4 region), classified to the genus level when possible, are shown. Labels indicate sample, storage condition (SC) as described in Table 1 and aliquot in the following format: sample_SC#_aliquot#. For example, A_2_1 is from sample A, storage condition number 2, aliquot number 1. The colors of the horizontal bars above labels correspond to sample color-coding in Fig. 2a. a Undiluted DNA for PCR. b DNA diluted 1:10 for PCR, indicated by d after aliquot number

Fig. 2

θ_YC distances between bacterial communities of stool sample aliquots subjected to various temperature storage conditions from 3 patients. a Principal coordinates analysis (PCoA) of θ_YC distances between bacterial communities of stool sample aliquots. The aliquots of each sample were represented by a different color which corresponds to the colors of the horizontal bars above labels in Fig. 1. b The θ_YC distances between aliquots was significantly lower within samples (median: 0.035, IQR: 0.015-0.061) than between samples (median: 0.93, IQR: 0.85-0.97) (Wilcoxon test p-value: <0.0001)

Fig. 3

Bacterial community composition of stool and subsequent rectal swab samples. The relative abundances of sequences classified to the genus level when possible. Labels indicate sample type (f = stool, s = swab), subject number, sample number and approximate sampling time in hours relative to stool sample collection. For example, s1_4_24 indicates a swab sample from subject 1, sample number 4, collected at approximately 24 h after the stool sample. Please note that sample s1_2_3 is distinct from other subject 1 samples, likely due to contamination from a subject 2 sample. The colors of the horizontal bars above labels correspond to subject color-coding in Fig. 4a

Fig. 4

θ_YC distances between bacterial communities of stool and swab samples. a Principal coordinates analysis (PCoA) of θ_YC distances between bacterial communities of stool and swab samples. Samples are color-coded by subject. Shape indicates sample type: triangles = stool samples and circles = swab samples. b The θ_YC distances between swab and stool samples was significantly lower within subjects (median: 0.17, IQR: 0.10-0.27) than between subjects (median: 0.93, IQR: 0.85-0.97) (Wilcoxon test p-value: <0.0001)