Skip to main content

3' MACE RNA-sequencing allows for transcriptome profiling in human tissue samples after long-term storage.

Boneva S, Schlecht A, Böhringer D, Mittelviefhaus H, Reinhard T, Agostini H, Auw-Haedrich C, Schlunck G, Wolf J, Lange C

Publication: Lab Invest (2020) • Vol 100 • Page 1345-1355PMID: 32467590View on PubMed

Purpose

This paper compared the performance of two sequencing methods (standard RNAseq and 3’ massive analysis of cDNA ends (MACE)) in fresh and formalin-fixed, paraffin-embedded (FFPE) healthy conjunctiva specimens, and among FFPE blocks stored for between 56 d and 13.4 years.

Conclusion

In the three fresh, healthy conjunctival specimens, the percentage of reads that mapped to the human reference genome was comparable between standard RNAseq and MACE (86.2% versus 76.8%, respectively). While RNAseq analysis of fresh (unfixed) and FFPE case-matched conjunctival specimens generated a similar number of processed reads the percentage of multi-mapping reads washigher in FFPE than fresh specimens (25.8% versus 21.4%, respectively; p=0.029); and, libraries generated from FFPE specimens had lower diversity than those from fresh specimens. For MACE-RNAseq, the mean number of genes detected in FFPE and fresh specimens were similar (19,655 versus 19,717, respectively) and the mean number of normalized reads were strongly correlated (R2=0.88, p<0.001). The proportion of detected transcripts by gene type also did not differ between FFPE and fresh specimens when analyzed by MACE-RNAseq (73.2% versus 72.9%, respectively).

While the percentage of uniquely mapped reads by MACE-seq was adversely affected by the duration of FFPE block storage, with robust changes occurring within the first 1,000 days of storage, a significant correlation was not observed between the number of expressed genes detected and storage duration, nor was library diversity affected.

Studies (2)

Purpose

This study compared the performance results of traditional RNAseq and MACE sequencing in fresh (unfixed) healthy conjunctiva specimens and among case-matched, prospectively collected FFPE specimens. Fresh and healthy conjunctival samples were collected from four patients during surgery for vitreoretinal detachment and divided to yield a total of three aliquots that were immediately processed for traditional RNAseq and MACE RNAseq, or fixed in formalin for 12 h within 1 minute of excision and processed and embedded in paraffin. Total RNA was extracted from FFPE sections with Zymo’s Quick-RNA FFPE Kit, additional details were not provided. RNA quality was assessed with the RNA Pico Sensitivity Assay and RNA concentration was determined by Qubit fluorometry. Libraries for standard RNAseq were prepared using the GenXPro GmbH Kits and were sequenced on a HiSeq 2500 machine. For MACE, poly-adenylated mRNA (isolated from total RNA) was used to construct libraries with the GenXPro GmbH Kit, and sequencing was performed on a NextSeq500 machine.

Summary of Findings

RNAseq and MACE-RNAseq analysis of healthy unfixed conjunctival specimens resulted in similar mapped performances and comparable numbers of detectable expressed genes.  The majority of genes were detectable by both sequencing methods (19,659), with 2,213 and 3,251 genes identified by only MACE-RNAseq or RNAseq, respectively. While the two sequencing methods displayed similar sensitivities for the detection of genes with a low level of expression (defined as a copy number <10) and those that were mid-range (a copy number of 10-100), RNAseq detected more genes with high expression (copy number >100) than MACE-RNAseq (men 9,233 versus 8,771; p<0.001). This difference may be due in part to a preferential bias for longer transcripts for the RNAseq platform. A significant albeit weak correlation was observed between transcript length and the mean number of raw reads for RNAseq (R2=0.005, p<0.001) but not MACE-RNAseq, and a difference was observed in median transcript length between the two methods (RNAseq: 1,451 bp versus MACE-RNAseq: 643 bp; p<0.001).

RNAseq analysis of fresh (unfixed) and FFPE case-matched conjunctival specimens generated a similar number of processed reads, although fresh specimens had a modestly higher number of processed reads that uniquely mapped to the reference genome than FFPE specimens (76.7% versus 71.4%, respectively; p=0.029).  The percentage of multi-mapping reads was higher in FFPE than fresh specimens (25.8% versus 21.4%, respectively; p=0.029). Libraries generated from FFPE specimens had lower diversity than those from fresh specimens (mean of 17.5 genes that accounted for 25% of reads for FFPE versus 23.8 genes for fresh specimens). For MACE-RNAseq, the mean number of genes detected in FFPE and fresh specimens were similar (19,655 versus 19,717, respectively) and the mean number of normalized reads were strongly correlated (R2=0.88, p<0.001). With MACE-RNAseq, the proportion of detected transcripts by gene type did not differ between FFPE and fresh specimens (73.2% versus 72.9%, respectively).

  • Biospecimens

    TissueEye
  • Preservation Types

    None (Fresh)Formalin
  • Diagnoses

    Normal
    Other diagnoses
  • Platforms

    RNAAutomated electrophoresis/Bioanalyzer
    RNAFluorometry
    RNANext generation sequencing
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    Biospecimen PreservationType of fixation/preservation
    Formalin (buffered)
    None (fresh)
    Next generation sequencing SpecificTechnology platform
    MACE-RNAseq
    Traditional RNAseq