Optimized Gene Expression Analysis from FFPE Samples

Gene expression analyses on RNA isolated from formaldehyde-fixed, paraffin-embedded (FFPE) tissues are challenging due to chemical modifications of RNA, cross-links of RNA with other molecules, degradation of RNA, and the limited amounts of sample usually available. In this study, researchers at Applied Biosystems, the University of Dublin (Ireland), and the University of Bern (Switzerland) addressed the feasibility of obtaining accurate and reproducible gene expression results using FFPE samples from a wide variety of tissues and block ages. The optimized protocol presented here can be used to reliably quantify mRNA expression levels using RNA isolated from FFPE samples and real-time RT-PCR.

Introduction

When working with archival material, researchers often need to use RNA from samples that have been stored as FFPE tissue blocks. These FFPE biopsies are a valuable source of information for retrospective research, providing complete historical and clinical information on treatments, procedures, or outcomes.

Although the tissue structure of FFPE samples has been maintained for histological analysis, damage to the nucleic acid incurred through the fixation, embedding, and storage processes can impede gene expression studies. Because many of the RNA alterations cannot be reversed, researchers are faced with optimizing multiple steps of RNA gene expression analysis protocols when working with these samples. This article describes a four-step protocol and the tools you need to reproducibly and reliably study gene expression profiles from FFPE samples.

Step 1: Isolate RNA from FFPE Samples

Overview. Isolating RNA from FFPE samples provides several challenges for preserving expression patterns that most closely correlate with the RNA levels present before fixing, embedding, and storage. For example, recovering all mRNA in a quantitative manner can be difficult because some RNA will be cross-linked to other molecules in the FFPE sample. RNA fragmentation, common in FFPE samples, means that RNA isolation protocols must be able to extract all sizes of RNA. In addition, a heating step should be considered as a potential method of reversing some formaldehyde-induced modifications of nucleic acids [1–3].

Tools. As expected, the integrity of RNA isolated from FFPE samples is low, and the broad profiles are characteristic of heavily degraded RNA fragments (Figure 1). The RecoverAll™ Total Nucleic Acid Isolation Kit has been optimized to recover short RNA fragments; therefore, the maximum peak is shifted towards shorter fragment lengths. Longer fragments are also collected, as is indicated by the overall shape and the shoulder above 200 base pairs in this representative example. In comparison, two other kits (Suppliers 2 and 3) preferentially isolate longer fragments, while two kits (Suppliers 1 and 4) preferentially isolate shorter fragments (Figure 1). Compared to other kits, the RecoverAll Kit also consistently produced the highest yields of RNA (Figure 2).

Figure 1. Gel Image and Electrophoretic Traces of RNA from an FFPE Sample. RNA preparations from a single FFPE breast cancer tissue block were measured on an Agilent® 2100 bioanalyzer after utilizing different commercial extraction protocols, including the Ambion RecoverAll™ Total Nucleic Acid Isolation Kit for FFPE Tissues. Marker = RNA 6000 Ladder.

Figure 2. The Ambion RecoverAll™ Total Nucleic Acid Isolation Kit for FFPE Tissues Recovers the Highest RNA Yields Compared to Similar Protocols. RNA was isolated from duplicate samples of two FFPE tumor tissues that were six years old using different commercial extraction protocols, including the Ambion RecoverAll Total Nucleic Acid Isolation Kit.

To reverse chemical modifications of the RNA that occurred during fixation, a heating step (70ºC, 20 min) was incorporated into the RecoverAll Kit protocol after protease digestion but before nucleic acid isolation. Although the observation that cycle threshold (CT) values decrease when the sample is heated has been reproduced using a variety of tissues and cell lines (Figure 3), in some cases no improvement in sensitivity was observed (data not shown). This result might be explained by different fixation, handling, and storage methods for those samples. However, heating FFPE samples is recommended because a significant increase in the CT value has not been observed, indicating that the heating step will not be detrimental to RNA quality.

Figure 3. Heating Samples During RNA Isolation Can Improve Sensitivity of Downstream Real-Time RT-PCR Assays. Gene expression levels of four different genes (GAPDH was measured with two different TaqMan® Assays) with increasing amplicon length for the same sample are shown. RNA was extracted using the RecoverAll™ Total Nucleic Acid Isolation Kit for FFPE Tissues. For one of the FFPE samples (green), the standard protocol was modified to include a 70°C heating step for 20 min prior to the column purification. Real-time PCR was performed using the TaqMan Gene Expression Master Mix. Error bars indicate the 95% confidence intervals. Data courtesy of J Li and O Sheils, University of Dublin, Trinity College, Dublin, Ireland.

Step 2: Reverse Transcribe RNA to cDNA

Overview. Real-time PCR is a versatile method for the precise quantification of nucleic acids. Its ability to reliably quantify steady-state levels of specific RNA sequences, no matter what the species (mRNA, miRNA, snRNA, etc.), makes real-time RT-PCR (Steps 2 and 4) the method of choice for many quantitative gene expression experiments.

In two-step RT-PCR, a reverse transcription step is performed first and the resulting cDNA is then used in separate real-time PCR assays. Therefore, accurate and precise quantification of RNA targets relies upon the performance of the RT step, which means that the reverse transcriptase efficiency and fidelity should be high.

Tools. The High Capacity cDNA Reverse Transcription Kit efficiently and quantitatively produces high quality cDNA for gene expression studies. The high efficiency of the MultiScribe™ Reverse Transcriptase in this kit makes it ideal for use with FFPE samples.

Step 3: Preamplify cDNA Without Bias

Overview. Researchers strive for a high yield RNA extraction procedure for samples such as FFPE, laser capture microscopy, and needle biopsy, which are all limited in quantity. However, the amount of RNA obtained is often insufficient to perform parallel experiments on the same RNA source.

Tools. cDNA preamplification with the TaqMan® PreAmp Master Mix Kit can address these issues and increase data quality. During the preamplification step, cDNA is amplified using methods that do not introduce biases in RNA representation. The amplified product can then be diluted and used in subsequent real-time PCR experiments (Figure 4).

Figure 4. Increase Specificity by Performing RNA Preamplification. One tenth of a 10-cycle preamplification was used in a TaqMan® real-time PCR assay. These conditions are predicted to produce a gain in CT value (DCT) of ~5.7, relative to the original sample (i.e., 10 preamplification cycles increases template 1024-fold; one tenth of that is 102.4 times the original sample, which is equivalent to a CT value of ~5.7). For most protocols, there was not a large difference between the observed and predicted change in CT value; therefore, no preamplification bias was introduced. (TaqMan Assay for CDKN1B gene, amplicon length = 71 bp; average CT ± range of two independent extractions). FFPE = laser capture microdissection FFPE tissue; Fresh = matched fresh frozen sample; PreAmp = preamplification step used; Data courtesy of J Li and O Sheils, University of Dublin, Trinity College, Dublin, Ireland.

Step 4: Analyze cDNA Using Real-Time PCR

Overview. Real-time PCR can be used to quantify mRNA levels from small RNA input amounts and over a broad dynamic range (>7 logs) compared to commonly used techniques (Northern blot analysis and RNase protection assay). Novel chemistries, such as the TaqMan 5' nuclease, and instrumentation platforms both enable detection of very low copy numbers.

FFPE treatment and storage of tissue samples leads to RNA that is randomly damaged across its length. For amplifications from these samples, there is a direct correlation between target amplicon size and the number of intact target templates present. Therefore, shorter amplicons will perform better (Figure 5) [4–6], and we recommend using primers that generate amplicons of less than 150 bp.

Figure 5. Effects of Amplicon Size on Real-Time RT-PCR Using Compromised Input RNA. Higher CT values are observed for artificially degraded (alkaline treated) RNA isolated from a ZR75 breast cancer cell line (reverse transcription was performed using the High Capacity cDNA Reverse Transcription Kit, Real-time PCR was performed using TaqMan® Universal PCR Master Mix and custom TaqMan MGB probes). Data courtesy of A Oberli, A Baltzer, and R Jaggi, University of Bern, Switzerland.

Tools. TaqMan® Gene Expression Assays are primer/probe sets for real-time PCR analysis of specific genes. The FAM™-dye labeled Minor Groove Binding (MGB) probes (for increased sequence specificity), and the small amplicon sizes (often well below 100 base pairs) make these assays an ideal solution for real-time RT-PCR using degraded RNA from FFPE samples. (The amplicon length for each TaqMan Gene Expression Assay can be found at www.allgenes.com.) Target detection increased with decreasing amplicon length, as observed by lower CT values for shorter amplicons.

Applied Biosystems scientists have optimized several master mixes for various applications, and we recommend the TaqMan Gene Expression Master Mix for analysis of FFPE samples. This master mix contains all components necessary for PCR assays, except cDNA and primer/probes. It is optimized for use with Applied Biosystems real-time PCR systems, which have been acknowledged as the gold standard in real-time PCR whether you need high-throughput capability (7900HT Fast Real-Time System) or are just getting started (StepOne™ Real-Time PCR System).

Scientific Contributors
Jinghuan Li and Orla Sheils • University of Dublin, Trinity College, Dublin, Ireland
Andrea Oberle, Anna Baltzer, and Rolf Jaggi • University of Bern, Switzerland
Marco Pirotta, Rick Conrad, Emily Zeringer, Danny Lee, John Pfeifer, and Kathy Lee • Applied Biosystems

Real-Time PCR Decision TreeFor additional information about selecting reagents for other sample types (e.g., microRNA or blood) or applications (e.g., genotyping), use the web-based, interactive Real-Time PCR Decision Tree

References1. Li J and Sheils O (2007) in preparation.

2. Hamatani K, Eguchi H, Takahashi K, Koyama K, Mukai M, Ito R, Taga M, Yasui W, Nakachi K (2006) Improved RT-PCR amplification for molecular analyses with long-term preserved formalin-fixed, paraffin-embedded tissue specimens. J Histochem Cytochem 54(7):773–780.

3. Kuramochi H, Hayashi K, Uchida K, Miyakura S, Shimizu D, Vallböhmer D, Park S, Danenberg KD, Takasaki K, Danenberg PV (2006) Vascular endothelial growth factor messenger RNA expression level is preserved in liver metastases compared with corresponding primary colorectal cancer. Clin Cancer Res 12(1):29–33.

4. Gloghini A, Canal B, Klein U, Dal Maso L, Perin T, Dalla-Favera R, Carbone A (2004) RT-PCR analysis of RNA extracted from Bouin-fixed and paraffin-embedded lymphoid tissues. J Mol Diagn 6(4):290–296.

5. Antonov J, Goldstein DR, Oberli A, Baltzer A, Pirotta M, Fleischmann A, Altermatt HJ, Jaggi R (2005) Reliable gene expression measurements from degraded RNA by quantitative real-time PCR depend on short amplicons and a proper normalization. Lab Invest 85(8):1040–1050.

6. Godfrey TE, Kim SH, Chavira M, Ruff DW, Warren RS, Gray JW, Jensen RH (2000) Quantitative mRNA expression analysis from formalin-fixed, paraffin-embedded tissues using 5' nuclease quantitative reverse transcription polymerase chain reaction. J Mol Diagn 2(2):84–91.