Tissues from routine pathology archives are suitable for microRNA analyses by quantitative PCR

Abstract

Background: MicroRNAs have recently taken centre stage as short non-coding RNAs that regulate mRNA expression.

Aim/methods: To assess the feasibility of using microRNA techniques on routinely processed tissues, the accessibility of two representative microRNAs was examined by real-time quantitative PCR in 86 human formalin-fixed paraffin-embedded (FFPE) samples from liver, breast, bone marrow, lymphatic tissues and colon. Murine liver was used to analyse the influence of fixation time and different fixatives.

Results: High-quality microRNA was successfully extracted from routinely processed formalin-fixed tissues, resembling PCR amplification results from snap-frozen material analysed in parallel. While fixation time did not affect microRNA accessibility, non-buffered formalin or fixative supplements such as glutaraldehyde influenced PCR results. Storage of human tissues for up to 7 years did not cause a significant deterioration of microRNA. However, microRNA quality in human archival material following routine processing 10-20 years ago was decreased. Oxidation by ambient air during storage and fixation in non-buffered formalin is a possible reason for loss of microRNA quality.

Conclusion: The assessment of microRNAs in readily obtained formalin-fixed paraffin-embedded samples is a highly promising tool in molecular pathology when similarly treated samples are analysed. Therefore, microRNA analyses will gain wider acceptance as an adjunct to morphological tissue assessment in routine pathology and retrospective studies.

Figure 1. Formalin-fixed paraffin-embedded (FFPE) versus snap-frozen samples of liver tissue. (A) Level of miR-122a microRNA normalised against miR-16 microRNA in snap-frozen (n = 5) and in FFPE mouse liver tissues (n = 5). The mean value of the snap-frozen samples served as calibrator. Error bars indicate SD. (B) miR-122a detection in snap-frozen and FFPE mouse liver tissues, and (C) miR-16 in matched samples of human snap-frozen and FFPE tissues of liver and colon (see also table 1)

Figure 2. Level of miR-16 microRNA in different formalin-fixed paraffin-embedded tissues. Median of miR-16 level in different organs determined by real-time PCR of 10 ng RNA for each sample. Outlying points are displayed as circles and an extreme outlying point is displayed as an asterisk.

Figure 3. Quantitative real-time PCR expression of miR-122a microRNA in correlation to different length of formalin fixation and different fixatives. Quantitative real-time PCR analysis of miR-122a. The mean values of normalised miR-122a levels of the 24 h snap-frozen samples served as calibrator. Two asterisks indicate a high level of statistical significance (p<0.01).

Figure 4. Level of miR-16 microRNA in archived formalin-fixed paraffin-embedded samples from three decades. miR-16 levels were determined in formalin-fixed paraffin-embedded human lymph nodes by real-time PCR. A 10 ng quantity of total RNA for each sample was used and miR-16 levels were taken from a standard curve. Error bars indicate standard deviation and asterisks indicate a significant decrease of miR-16 level after 17 and 27 years of storage (p<0.01).