You are here:
Lack of concordance in microarray gene expression responses to Phenobarbital in companion aged FFPE and Frozen liver samples
Citation:
Hester, S., G. Carswell, B. Vallanat, A. Deangelo, AND C. Wood. Lack of concordance in microarray gene expression responses to Phenobarbital in companion aged FFPE and Frozen liver samples. Presented at Society of Toxicology, March 10 - 14, 2013.
Impact/Purpose:
This abstract will be presented at the Society of Toxicology meeting March 10-14, 2013, San Antonio, TX.
Description:
Despite the immense potential value of public and private biorepositories, direct utilization of archival tissues for molecular profiling has been limited. A major reason for this limited use is the difficulty in obtaining reliable transcriptomic profiles from formalin-fixed paraffin-embedded (FFPE) tissue samples with highly fragmented nucleic acid. The goal of this study was to evaluate transcriptional responses in 16 year-old FFPE and companion frozen (FROZ) liver samples from F344 rats treated for 2 yrs with control water or water with 0.06% Phenobarbital (PB), a known CAR/PXR inducer and rodent liver mitogen. RNA isolation and gene expression was performed using the Rat Illumina Bead Array® for 16 samples, paired FFPE and FROZ. The RNA integrity numbers (RINs) ranged from 2-3 for FFPE and 3-5 for FROZ, and RNA yield for both FFPE (2 10µm sections) and FROZ (20 mg) samples was 2-4 micrograms. Nugen Ovation® and Encore® reagents optimized for FFPE tissues were used to amplify and label the RNA followed by hybridization to arrays. We selected 12 of 16 samples based on successful hybridization performance for analysis. Differential gene expression was assessed using Rank Products followed by pathway mapping. Results showed only 10 significantly altered PB-responsive genes in common between paired FFPE (n=1373) and FROZ (n=721) samples whereas pathway analysis identified only a 10% overlap in pathways significantly altered by PB between FFPE and FROZ samples. Pathway profiles for both FFPE and FROZ samples were unique to their respective sample type. The lack of concordance in genomic responses to PB suggests that traditional microarray platforms are inadequate for genomic profiling of aged FFPE samples, despite improved RNA isolation and labeling methods. Further work will evaluate more recent next-generation sequencing (NGS) technologies for transcriptomic analysis of archival specimens. This abstract does not reflect EPA policy