Optimizing microbiome sequencing for small intestinal aspirates: validation of novel techniques through the REIMAGINE study

Abstract

Background: The human small intestine plays a central role in the processes of digestion and nutrient absorption. However, characterizations of the human gut microbiome have largely relied on stool samples, and the associated methodologies are ill-suited for the viscosity and low microbial biomass of small intestine samples. As part of the REIMAGINE study to examine the specific roles of the small bowel microbiome in human health and disease, this study aimed to develop and validate methodologies to optimize microbial analysis of the small intestine.

Results: Subjects undergoing esophagogastroduodenoscopy without colon preparation for standard of care were prospectively recruited, and ~ 2 ml samples of luminal fluid were obtained from the duodenum using a custom sterile aspiration catheter. Samples of duodenal aspirates were either untreated (DA-U, N = 127) or pretreated with dithiothreitol (DA-DTT, N = 101), then cultured on MacConkey agar for quantitation of aerobic gram-negative bacteria, typically from the class Gammaproteobacteria, and on blood agar for quantitation of anaerobic microorganisms. DA-DTT exhibited 2.86-fold greater anaerobic bacterial counts compared to DA-U (P = 0.0101), but were not statistically different on MacConkey agar. DNA isolation from DA-U (N = 112) and DA-DTT (N = 43) samples and library preparation for 16S rRNA gene sequencing were also performed using modified protocols. DA-DTT samples exhibited 3.81-fold higher DNA concentrations (P = 0.0014) and 4.18-fold higher 16S library concentrations (P < 0.0001) then DA-U samples. 16S rRNA gene sequencing revealed increases in the detected relative abundances of obligate and facultative anaerobes in DA-DTT samples, including increases in the genera Clostridium (false discovery rate (FDR) P = 4.38E-6), Enterococcus (FDR P = 2.57E-8), Fusobacterium (FDR P = 0.02) and Bacteroides (FDR P = 5.43E-9). Detected levels of Gram-negative enteropathogens from the phylum Proteobacteria, such as Klebsiella (FDR P = 2.73E-6) and Providencia (FDR P < 0.0001) (family Enterobacteriaceae) and Pseudomonas (family Pseudomonadaceae) (FDR P = 0.04), were also increased in DA-DTT samples.

Conclusions: This study validates novel DTT-based methodology which optimizes microbial culture and 16S rRNA gene sequencing for the study of the small bowel microbiome. The microbial analyses indicate increased isolation of facultative and obligate anaerobes from the mucus layer using these novel techniques.

Keywords: 16S rRNA gene sequencing; Methodology optimization; Microbial culture; Microbiome; Mucus layer; Small intestine.

Conflict of interest statement

The authors declare that they have no competing interests.

Figures

Fig. 1

Workflow for pretreatment and microbial culture, including the number of subjects in each group

Fig. 2

Workflow for DNA extraction and 16S rRNA gene sequencing of duodenal aspirate (DA) samples, including the number of subjects in each group

Fig. 3

Final quantification of 16S libraries from DA-U (N = 112) and DA-DTT (N = 43) samples after 35 PCR cycles. The Mann-Whitney test was used to compare the median value of groups

Fig. 4

Sunburst representation of the overall distribution of the small intestinal microbiome as determined by 16S rRNA sequencing. On the left: Relative microbial abundance detected in DA-U (no pretreatment, N = 112). On the right: Relative microbial abundance detected in DA-DTT (pretreatment with DTT, N = 43)

Fig. 5

Alpha diversity rarefaction curves for DA-DTT (N = 112) and DA-U (N = 43) samples. Samples were rarefied to the least numbers of sequences obtained

Fig. 6

Alpha diversity indices of DA samples pre-treated with DTT (DA-DTT, N = 112) and untreated DA (DA-U, N = 43). Left: Shannon entropy diversity for DA-DTT and DA-U samples. Right: Simpson’s index diversity for DA-DTT and DA-U samples

Fig. 7

Beta diversity of DA-U and DA-DTT based on the weighted UniFrac metric. Principal Coordinates Analysis plot of binary and abundance-weighted Unifrac distances between DA-DTT (shown in orange, N = 43) and DA-U (shown in blue, N = 112)