Cancer Diagnosis Program | Biorepositories and Biospecimen Research Branch

Search BRD Papers

Comparison of Methods To Collect Fecal Samples for Microbiome Studies Using Whole-Genome Shotgun Metagenomic Sequencing.

Author(s): Byrd DA, Sinha R, Hoffman KL, Chen J, Hua X, Shi J, Chia N, Petrosino J, Vogtmann E

Publication: mSphere, 2020, Vol. 5, Page

PubMed ID: 32250964 PubMed Review Paper? No

Purpose of Paper

Conclusion of Paper

Studies

1. Study Purpose

   This study compared the microbiome of frozen fecal specimens with aliquots of the same 15 specimens that were preserved in RNAlater, 95% ethanol, FIT tubes and FOBT cards, then stored at room temperature for 0 or 4 days, and frozen at -80°C before analysis. Feces were collected from 15 volunteers into an Exakt Pak canister and delivered to the laboratory. Specimens were manually mixed using a spatula and aliquoted. Aliquots were then immediately frozen at -80°C (control), mixed with RNAlater, mixed with 95% ethanol, smeared onto FOBT cards, or mixed with the fluid in FIT tubes. Matched aliquots in each preservative were frozen immediately and after 4 days at room temperature.  DNA was extracted from specimens using the PowerMag Soil DNA Isolation Kit and quantified using PicoGreen. Specimens were sequenced using TruSeq SBS Kits on a HiSeq 2000 instrument. Data was processed using a pipeline developed at the Baylor College of Medicine. Statistical analysis was conducted using R and SAS.

   Summary of Findings:

   Fecal specimen aliquots that were directly frozen without preservation had the highest number of observed species (580 species) and Shannon index, with significantly fewer species detected in specimens preserved with FIT tubes (549 species, P=0.02), FOBT cards (539 species, P=0.002), RNAlater (537 species, P=0.001) and 95% ethanol (534 species, P<0.001).  When limited to genes that had >70% identity and coverage to known KEGG orthologs (k-genes), the highest number of observed species and Shannon index were for FOBT-preserved specimens (30,894 genes), but when compared to specimens frozen immediately, differences were limited to fewer k-genes in RNAlater-preserved specimens (28,369 genes versus 30,290 genes in control, P=0.009).  For all preservatives, except RNAlater, the observed number of k-genes and Shannon index were higher on day 0 than on day 4 of post-preservation room temperature storage. Variability in species and k-gene coverage was primarily explained by interindividual differences (71% and 68%, respectively), with only 9-10% of the variability explained by collection method/solution and <1% of the variability explained by storage duration. The relative abundance of the detected species was dependent on the preservative used, but specimens that were frozen without a preservative solution had more species with an abundance that was greater than 2% than in any of the preserved specimens. Interclass correlation coefficients (ICCs) relative to  fecal specimens frozen on day 0 were lowest for the abundance of Actinobacteria, Bacteroidetes, and Firmicutes in specimens preserved with 95% ethanol (0.46, 0.23 and 0.33, respectively) or FOBT cards (0.30, 0.36, and 0.40, respectively) but were >0.50 in specimens preserved in RNAlater or FIT tubes. ICCs for alpha diversity metrics (Shannon index and observed species/genes) relative to frozen specimens were higher for specimens preserved with RNAlater (0.68-0.91) or FOBT cards (0.46-0.87) than specimens preserved with FIT tubes (0.44-0.84) or 95% ethanol (0.40-0.81). ICCs with frozen specimens for axis 1 of Bray-Curtis and Jaccard distances were highest for fecal specimens preserved with RNAlater (0.94 and 0.91, respectively) followed by specimens preserved with FIT tubes (0.89 and 0.87, respectively) or FOBT cards (0.86-0.85) and lowest for 95% ethanol-preserved specimens (0.67 and 0.65, respectively), with much lower ICCs for axis 2 (0.33-0.50). ICCs with frozen specimens for the most abundant species were low (<40%) for 3, 5, 6, and 7 of the 20 most abundant species in fecal specimens preserved with RNAlater, FOBT cards, FIT tubes and 95% ethanol. ICCs for k-genes, modules, and pathways with frozen specimens were highly variable, but tended to be lowest in specimens preserved in 95% ethanol. ICC’s for observed species, observed gene, Shannon index, Bray-Curtis distance and Jaccard distance matrices between fecal specimens preserved by the same method on day 0 and day 4 were high for specimens preserved with FIT tubes (0.86-0.99), FOBT cards (0.93-0.99), and RNAlater (0.97-0.99), but were lower for specimens preserved in 95% ethanol (0.57-0.77 most, 0.93 for observed species). Further, in 95% ethanol -specimens, ICCs between day 0 and day 4 post-preservation room temperature storage were poor (<40%) for 4 of 20 species, 3 of 20 k-genes, 2 of 10 modules, and 3 of 10 pathways.

   Biospecimens

   • Bodily Fluid - Feces

   Preservative Types

   • Ethanol
   • RNAlater
   • Other Preservative
   • Frozen

   Diagnoses:

   • Not specified

   Platform:

   Analyte	Technology Platform
   DNA	Fluorometry
   DNA	Next generation sequencing

   Pre-analytical Factors:

   Classification	Pre-analytical Factor	Value(s)
   Biospecimen Preservation	Type of fixation/preservation	Fecal immunochemical test (FIT) FOBT Card Ethanol RNAlater Frozen
   Storage	Time at room temperature	0 days 4 days

   View more