You are here:
Enhancement of RNA from Formalin-Fixed Paraffin-Embedded (FFPE) Samples
Citation:
Hester, S., L. Wehmas, C. Yauk, M. Roy, M. Gosink, D. D. Wilk, T. Hill, AND C. Wood. Enhancement of RNA from Formalin-Fixed Paraffin-Embedded (FFPE) Samples. SOT 56th Annual Meeting 2016, Baltimore, Maryland, March 12 - 16, 2017.
Impact/Purpose:
Impact statement The major issue in FFPE genomics is RNA quality. The standard fixative used for preserving tissue samples is formaldehyde, which may alter nucleic acids through fragmentation, protein cross-links, adducts, and other chemical modifications. Use of FFPE archival resources has been limited to date RNA-sequencing offers a promising way to address this problem. Most samples in tissue banks are stored as formalin-fixed paraffin-embedded (FFPE) blocks. It has been estimated that over 1 billion FFPE samples exist globally across government, academic, industry, medical and veterinary school, and hospital archives and that more than 20 million FFPE samples are archived annually just in the U.S. To date, however, use of these resources for genomic research has been limited by a lack of methods for generating reliable quantitative data and by inconsistent methods for genomic profiling of degraded RNA from formalin-fixed paraffin-embedded (FFPE) samples. The goal of this work is to inform new ways to reliably access quantitative genomic information from archival samples. Major objectives of these studies are (1) to evaluate pre-analytical variables affecting RNA-Seq results across different sample types (e.g., age in block, formalin fixation time) and using different methods to demodify poor-quality RNA (resulting from formalin exposure) for subsequent sequencing, and (2) to evaluate genomic responses using RNA-Seq in FFPE samples following exposure to reference carcinogens (e.g., PB, DCA, DEHP, A4, and FUR) and determine concordance with corresponding fresh-frozen (FROZ) samples. Methods to rescue and improve poor-quality RNA for use in sequencing will have positive implications in research and clinical settings.
Description:
Enhancement of RNA from Formalin-Fixed Paraffin-Embedded (FFPE) Samples Susan Hester1, Leah Wehmas1, Carole Yauk2, Marc Roy3, Mark M. Gosink3, Deidre D. Wilk4, Thomas Hill III5, Charles E. Wood11Office of Research and Development, US EPA, RTP, NC 27709, USA, 2Environmental Health Science and Research Bureau, Health Canada, Ottawa K1A 0K9, Canada, 3Pfizer Drug Safety R&D, Groton, CT, 4GlaxoSmithKline, King of Prussia, PA 19406, 5ORISE.Low RNA yield and quality are important issues when using FFPE samples for genomic analyses. In this this study we evaluated methods to “demodify” FFPE RNA highly fragmented and cross-linked by fixation. Primary endpoints were RNA recovery, RNA-Sequencing quality metrics, and target gene counts in response to a reference chemical (phenobarbital, PB). Frozen (FROZ) mouse liver samples from control and PB groups (n=6/group) were divided and preserved as follows: (1) FROZ only; (2) 70% ethanol (ETOH) for 3 months; (3) 10% buffered formalin for 18 hours followed by ETOH for 3 months, and (4) 10% buffered formalin for 3 months. Samples from Groups 2-4 were then processed to FFPE blocks and sectioned for RNA isolation. Group 4 samples received no demodification treatment (4A), 0.5hr 70°C incubation with TAE (4B), overnight 65°C incubation with an organocatalyst using Qiagen (4C) or Purelink (4D) isolation kits, or an overnight 65°C incubation without an organocatalyst (4E). RNA was ribodepleted and sequenced using the Illumina HiSeq platform to at least 25 million reads/sample. RNA yields in FFPE Groups 2, 3, and 4A were 18-37% of those in FROZ samples. Use of overnight incubation +/- organocatalyst improved RNA yields by >3-fold compared to no incubation (Group 4A). No group differences were observed for RNA-Seq reads (28-39 million per sample) or read alignment (76-92%). Genome coverage was higher in Group 1 (12%) compared to Group 4A (3%) but returned to 13-16% with overnight incubation (Groups 4C-4E). Across PB-treated FFPE samples, average fold-change for the marker gene Cyp2b10 were compressed 2.3 fold compared to FROZ and DEMOD groups (4B-4E) rescued up to 18% of this signal loss. These findings highlight methods to improve RNA yield and sequencing results when using FFPE samples. This abstract does not reflect EPA policy.