Quantitative mRNA expression analysis from formalin-fixed, paraffin-embedded tissues using 5' nuclease quantitative reverse transcription-polymerase chain reaction

Abstract

Analysis of gene expression and correlation with clinical parameters has the potential to become an important factor in therapeutic decision making. The ability to analyze gene expression in archived tissues, for which clinical followup is already available, will greatly facilitate research in this area. A major obstacle to this approach, however, has been the uncertainty about whether gene expression analyses from routinely archived tissues accurately reflect expression before fixation. In the present study we have optimized the RNA isolation and reverse transcription steps for quantitative reverse transcription-polymerase chain reaction (RT-PCR) on archival material. Using tissue taken directly from the operating room, mRNAs with half-lives from 10 minutes to >8 hours were isolated and reverse transcribed. Subsequent real-time quantitative PCR methodology (TaqMan) on these cDNAs gives a measurement of gene expression in the fixed tissues comparable to that in the fresh tissue. In addition, we simulated routine pathology handling and demonstrate that this method of mRNA quantitation is insensitive to pre-fixation times (time from excision to fixation) of up to 12 hours. Therefore, it should be feasible to analyze gene expression in archived tissues where tissue collection procedures are largely unknown.

Figure 1.

Denaturing agarose gel of RNA isolated from 3 archival prostate tissue blocks collected in 1985, 1990, and 1993, respectively. RNA isolated using the modified 3-day protocol is longer (A) and the yield is higher (B) than with the original protocol.

Figure 2.

Amplification of different sized fragments of the β-actin mRNA from fixed tissue RNA isolates. Greatly increased sensitivity of detection is obtained with fragments less than 131 basepairs. The forward primer and TaqMan probe were kept constant and used with different reverse primers to change amplicon length (inset picture). No difference is seen between the immediate fix RNA and the 12 hour pre-fixation RNA samples (see Materials and Methods for description of tissue processing). Fresh tissue RNA was used as a high quality RNA control. Standard deviations on all points were smaller than the size of the symbols and are omitted.

Figure 3.

Relative abundance of 99- and 291-bp fragments of the β-actin mRNA in fixed tissues. All points are measured relative to β-Gus and are normalized to the fresh tissue RNA.

Figure 4.

Effect of pre-fixation time on quantitation of six different mRNA species. All points are measured relative to β-Gus and are normalized to the fresh tissue RNA.

Figure 5.

Effect of fixation and pre-fixation time on quantitation of different pieces of the c-myc mRNA. Primers and probe used are shown in Table 1 . Fragments are affected differently by fixation but none of the fragments show a change in abundance over the 12-hour time series. All points are measured relative to β-Gus and are normalized to the fresh tissue RNA.