Skip to main content

Evaluation of the Use of Formalin-Fixed and Paraffin-Embedded Archive Gastric Tissues for Microbiota Characterization Using Next-Generation Sequencing.

Pinto-Ribeiro I, Ferreira RM, Pereira-Marques J, Figueiredo C, Macedo G, Carneiro F, Pinto V

Publication: Int J Mol Sci (2020) • Vol 21PMID: 32046034View on PubMed

Purpose

This paper compared the microbiome of case-matched frozen and formalin-fixed, paraffin-embedded (FFPE) non-neoplastic gastric tissue blocks that were sectioned after 6, 12 and 18 months of storage. To assess potential contamination introduced during paraffin embedding, the authors compared the microbiome of FFPE gastric specimens and paraffin blocks containing no tissue specimen.

Conclusion

FFPE non-neoplastic gastric tissue specimens generated a mean of 62,920 reads representing 56 operational taxonomic units (OTUs), but the paraffin control block that lacked any tissue produced a mean of 62,095 reads and 58 OTUs. Microbial alpha-diversity (Shannon Index) was lower in the paraffin control than for any of the FFPE specimens evaluated, and the microbial community of the FFPE specimens was distant from that in the paraffin control as measured by weighted Unifrac distances of >0.5 for 14 of the 15 FFPE specimens. Proteobacteria and Actinobacteria were the most prevalent phyla in the paraffin control (77.4% and 19.1%, respectively) and the FFPE specimens (33.4% and 43.6%, respectively), but the FFPE specimens also had a high prevalence of Firmicutes (14.5%) and other phyla.  While 15 OTUs were only found in the paraffin control, 43 were found in both the paraffin control and FFPE specimens, of which 4 represented 93% of the bacterial abundance in the paraffin control. When the 58 OTUs found in the paraffin control were subtracted from microbiome data from FFPE specimens, the relative abundance of Proteobacteria and Bacteroidetes increased (54.1% increase, P<0.001 and 9.9% increase, P=0.021, respectively) and the relative abundance of Actinobacteria decreased (11.9% decrease, P<0.0001).

The median number of OTUs was higher in frozen specimens than case-matched FFPE specimens (75 versus 31, P<-0.0001), regardless of whether the FFPE blocks were stored for 6 (50 OTUs, P=0.011), 12 (34 OTUs, P=0.0008) or 18 (25 OTUs, P=0.0002) months (storage temperature not specified) before sectioning. However, alpha and beta diversity were comparable among FFPE and frozen specimens. Interestingly, while the microbial community of FFPE blocks stored for 12 months was considered distant from the communities of FFPE blocks stored for 6 (P=0.0021) or 18 (P<0.001) months and from the case-matched frozen specimens (P<0.0001), no other differences were observed. At the phyla level, FFPE specimens had more Proteobacteria (P= 0.042) and fewer Actinobacteria (P=0.033) than case-matched frozen specimens. When FFPE specimens were separated by storage duration, those stored for 12 months before sectioning had fewer Bacteroidetes than the frozen control (P=0.039), and those stored for 18 months had fewer Actinobacteria than frozen controls (P=0.045). At the phyla level, the taxonomic profiles were mostly consistent for case-matched frozen and FFPE specimens, regardless of storage duration, while differences were much more pronounced at the order level.

Studies (1)

Purpose

This study compared the microbiome of case-matched frozen and formalin-fixed, paraffin-embedded (FFPE) non-neoplastic gastric tissue blocks that were sectioned after 6, 12 and 18 months of storage as FFPE blocks. To assess potential contamination introduced during paraffin embedding, the authors compared the microbiome of FFPE gastric specimens and paraffin blocks containing no tissue specimen. Five non-neoplastic gastric specimens (diagnosis not specified) were cut into two fragments. One fragment was snap-frozen in liquid nitrogen and the other was fixed in 10% neutral buffered formalin (NBF) and embedded in paraffin. The FFPE block was trimmed to expose the tissue and 10 µm sections were cut after 6, 12 and 18 months of storage (temperature not specified). FFPE specimens were deparaffinized in xylene and frozen specimens were homogenized in ATL buffer, then all specimens were lysed in proteinase K at 56°C (duration not specified). DNA was then extracted with the QIAamp DNA FFPE Tissue Kit. The total bacterial load was quantified using the 16S rRNA gene real-time PCR assay. The V5-V6 hypervariable regions of the 16S rRNA gene were PCR amplified using universal primers and used to prepare sequencing libraries with the Illumina 16S Metagenomic Sequencing Library Kit. Libraries were pair-end sequenced on a MiSeq instrument. Reads were quality-filtered by requiring the number of expected errors to be <1.0 and trimmed to a fixed length of 29,000 reads before being dereplicated and clustered into OTUs with the UPARSE algorithm and alignment to two databases. Alpha-diversity and beta-diversity were assessed by the Shannon index and weighted UniFrac distances, respectively.

Summary of Findings

FFPE non-neoplastic gastric tissue specimens generated a mean of 62,920 reads representing 56 operational taxonomic units (OTUs), but the paraffin control block that lacked any tissue produced a mean of 62,095 reads and 58 OTUs. Microbial alpha-diversity (Shannon Index) was lower in the paraffin control than for any of the FFPE specimens evaluated, and the microbial community of the FFPE specimens was distant from that in the paraffin control as measured by weighted Unifrac distances of >0.5 for 14 of the 15 FFPE specimens. Proteobacteria and Actinobacteria were the most prevalent phyla in the paraffin control (77.4% and 19.1%, respectively) and the FFPE specimens (33.4% and 43.6%, respectively), but the FFPE specimens also had a high prevalence of Firmicutes (14.5%) and other phyla.  While 15 OTUs were only found in the paraffin control, 43 were found in both the paraffin control and FFPE specimens, of which 4 represented 93% of the bacterial abundance in the paraffin control. When the 58 OTUs found in the paraffin control were subtracted from microbiome data from FFPE specimens, the relative abundance of Proteobacteria and Bacteroidetes increased (54.1% increase, P<0.001 and 9.9% increase, P=0.021, respectively) and the relative abundance of Actinobacteria decreased (11.9% decrease, P<0.0001).

The median number of OTUs was higher in frozen specimens than case-matched FFPE specimens (75 versus 31, P<-0.0001), regardless of whether the FFPE blocks were stored for 6 (50 OTUs, P=0.011), 12 (34 OTUs, P=0.0008) or 18 (25 OTUs, P=0.0002) months (storage temperature not specified) before sectioning. However, alpha and beta diversity were comparable among FFPE and frozen specimens. Interestingly, while the microbial community of FFPE blocks stored for 12 months was considered distant from the communities of FFPE blocks stored for 6 (P=0.0021) or 18 (P<0.001) months and from the case-matched frozen specimens (P<0.0001), no other differences were observed. At the phyla level, FFPE specimens had more Proteobacteria (P= 0.042) and fewer Actinobacteria (P=0.033) than case-matched frozen specimens. When FFPE specimens were separated by storage duration, those stored for 12 months before sectioning had fewer Bacteroidetes than the frozen control (P=0.039), and those stored for 18 months had fewer Actinobacteria than frozen controls (P=0.045). At the phyla level, the taxonomic profiles were mostly consistent for case-matched frozen and FFPE specimens, regardless of storage duration, while differences were much more pronounced at the order level.

  • Biospecimens

    TissueStomach
  • Preservation Types

    FrozenFormalin
  • Diagnoses

    Not specified
  • Platforms

    DNANext generation sequencing
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen PreservationType of fixation/preservation
    Formalin (buffered)
    Snap frozen
    StorageStorage duration
    12 months
    18 months
    6 months