RNAscope Multiplex FISH Signal Assessment in FFPE and Fresh Frozen Tissues: The Effect of Archival Duration on RNA Expression.
Author(s): Sari AIP, Copeland K, Nuwongsri P, Pipatsakulroj W, Jinawath A, Israsena N, Lertsittichai P, Chirappapha P, Shiao MS, Jinawath N
Publication: J Histochem Cytochem, 2025, Vol. , Page 221554241311971
PubMed ID: 39844680 PubMed Review Paper? No
Purpose of Paper
This paper compared the levels of four housekeeping mRNAs in unmatched (22-24 specimens each) and eight case-matched archival formalin-fixed, paraffin-embedded (FFPE) and frozen optimal cutting temperature (OCT)-embedded breast tumor specimens using RNAscope and investigated the effects of storage duration of the FFPE and frozen specimens on the RNAscope results. Additionally, effects of section thickness on RNAscope signal and RNA integrity number (RIN) on transcript levels were investigated using frozen tissue specimens.
Conclusion of Paper
All four house-keeping genes were detected using RNAscope in 26 of the 30 (86.7%) FFPE and 32 of the 32 (100%) frozen tissue specimens. FISH signal was more distinct and/or the signal was strongest in frozen OCT-embedded sections that were 7 µM rather than 10 µM thick, fixed in 4% paraformaldehyde (PFA) rather than 10% neutral buffered formalin (NBF), at room temperature rather than 4°C, and for 20 min rather than 30 min. Cell counts were comparable between 4 µM-thick FFPE and 7 µM-thick frozen sections, but the signals were sharper in the 4 µM FFPE sections, with the authors attributing the improvement to the thinner sections. For each of the four housekeeping genes, frozen sections had higher transcript counts per cell, a lower percentage of sections with no detectable transcripts per cell, a higher percentage of sections with 10-15 or >15 transcripts per cell, and higher H-Scores than FFPE sections. Principal component analysis (PCA) based on the expression of the housekeeping genes showed some separation of FFPE and frozen specimens, but there was overlap between the groups. Based on NormFinder results, HPRT1 was the most stable of the four housekeeping genes evaluated.
When only the eight matched pairs were considered, the signal was generally stronger in frozen than FFPE specimens, but the significance of the difference was influenced by the duration of specimen storage, with a significantly stronger signal observed in specimens stored for 3 or more years than those stored for 2 years. Further, the strongest correlation in signal between matched frozen and FFPE specimens was observed in two of the specimens stored for 2 years. PCA based on housekeeping gene expression in the entire sample set showed some separation of FFPE specimens that were archived for 0-3, 4-7 and ≥8 y, but there was overlap between the 0-3 and 4-7 y groups. No separation based on the storage duration of frozen specimens was observed in the PCA of the entire frozen specimen sample set. The number of transcripts per 1000 cells was significantly lower in FFPE specimens stored for ≥8 years than those stored for 4-7 or 1-3 years and for FFPE specimens stored 4-7 versus 0-3 years, but differences in frozen specimens were limited to fewer POLR2A transcripts per 1000 cells in specimens stored 0-3 years versus 4-7 years. The number of transcripts per cell and H-scores were correlated with storage duration of FFPE specimens, but no significant correlations between the number of transcripts per cell or H-scores and storage duration were observed for frozen specimens. While the number of transcripts per cell was very weakly correlated with RIN for frozen specimens, there was no correlation between RIN and storage duration.
Studies
-
Study Purpose
This study compared the levels of four housekeeping mRNAs in unmatched (22-24 specimens each) and eight case-matched archival formalin-fixed, paraffin-embedded (FFPE) and frozen optimal cutting temperature (OCT)-embedded breast tumor specimens using RNAscope and investigated the effects of storage duration of the FFPE and frozen specimens on the RNAscope results. Thirty FFPE and 32 frozen breast cancer tumor specimens, including 8 that were case-matched, were collected between 2014 and 2021. FFPE specimens were fixed in 10% neutral buffered formalin for the duration specified by the College of American Pathologists, paraffin embedded and stored at room temperature for 0-3 years (14 specimens), 4-7 years (20 specimens) or ≥ 8 years (4 specimens) before analysis (no further details provided). Frozen specimens were frozen in OCT cryo-gel (no further details provided) and stored for 0-3 (12 specimens) or 4-7 years (20 specimens) at an unspecified temperature before analysis. All specimens were confirmed to contain >50% cancer cells. Sections (4 µM for FFPE and 7 µM for frozen) were mounted on Superfrost Plus slides. Levels of UBC, PPIB, POLR2A, and HPRT1 were assayed using the RNAscope Multiplex Fluorescent v2 Kit with the optional antigen retrieval step for FFPE specimens and following fixation of frozen specimens in 4% paraformaldehyde at room temperature for 20 min. RNAscope signals were imaged within 2 weeks using the Vectra Polaris Automated Quantitative Pathology Imaging System, and 3-5 regions of interest were captured at 40X magnification for analysis. Images were analyzed using the CellProfiler software. H-scores were calculated based on the transcript count per cell for each gene. For RNA integrity analysis, RNA was extracted from frozen sections using the RNeasy Mini Kit, and RIN was assessed using a TapeStation 2000.
Summary of Findings:
All four house-keeping genes were detected by RNAscope in 26 of the 30 (86.7%) FFPE and 32 of the 32 (100%) frozen specimens evaluated. Cell counts were comparable between 4 µM-thick FFPE and 7 µM-thick frozen sections, but the signals were sharper in the FFPE sections, which the authors attributed to thinner sections. The cell number adjusted transcript count for each of the four mRNAs was higher for frozen than FFPE sections (P<0.001); and when binned based on the number of transcripts per cell, FFPE specimens had a higher percentage of sections with no detectable transcripts per cell (P<0.0001, all four transcripts) and a higher percentage of sections with 10-15 or >15 transcripts per cell (P<0.0001, all) than frozen sections. The H-score for all four transcripts was correspondingly higher in frozen than FFPE sections (P<0.001, all). Principal component analysis (PCA) based on housekeeping gene expression of the entire sample set showed some separation of FFPE and frozen specimens, but overlap between the groups was observed. As expected, expression of UBC and PPIB were higher than POLR2A and HPRT1 in both FFPE and frozen specimens (P<0.001, all). Based on NormFinder results, HPRT1 was the most stable of the four housekeeping genes evaluated. When only the eight matched pairs were considered, RNAscope signals were generally stronger in frozen than FFPE specimens, but the significance of the effect was influenced by specimen storage, as FFPE specimens stored for 3 or more years were more likely to have significantly stronger RNAscope signal than those stored for 2 years. Further, the strongest correlation in signal between matched frozen and FFPE specimens was observed in two of the specimens stored for 2 years (0.824, 0 0.857 versus 0.119-0.722 for pairs stored 3-6 years and 0.280 in the remaining pair stored 2 years). PCA based on housekeeping gene expression in the entire sample set showed some separation of FFPE specimens that were archived for 0-3, 4-7 and ≥8 y, but there was also some overlap between the 0-3 and 4-7 y groups. Frozen specimens did not separate based on storage duration in the PCA of the entire sample set of frozen specimens. The number of transcripts per 1000 cells was significantly lower in the FFPE specimens stored for ≥8 years, than those stored for 4-7 (P<0.0001, all transcripts) or 1-3 years (P<0.0001, all transcripts), and for UBC, PPIB and POLR2A in FFPE specimens stored 4-7 versus 0-3 years (P<0.0001, P<0.001 and P<0.05, respectively). In frozen specimens, differences between specimens stored for 0-3 versus 4-7 years were limited to slightly fewer POLR2A transcripts per 1000 cells in specimens stored 0-3 years (P<0.05). While the number of transcripts per cell was very weakly correlated with the RIN of frozen specimens (R2<0.2, P<0.05; all), there was no correlation between RIN and storage duration. The number of transcripts per cell was correlated with the storage duration of FFPE specimens (R2<0.35, P<0.0001 for UBC and PPIB; R2<0.17, P<0.0001 for POLR2A; and R2<0.048, P<0.01 for HPRT1), but no significant correlations between the number of transcripts per cell and storage duration were observed for frozen specimens. Similarly, the H-score was correlated with the storage duration of FFPE specimens (R2<0.33, P<0.0001 for UBC and PPIB; R2<0.15, P<0.0001 for POLR2A; and R2<0.047, P<0.01 for HPRT1), but no significant correlations between the H-score and storage duration were observed for frozen specimens.
Biospecimens
Preservative Types
- Frozen
- OCT
- Formalin
Diagnoses:
- Neoplastic - Carcinoma
Platform:
Analyte Technology Platform RNA In situ hybridization RNA Automated electrophoresis/Bioanalyzer Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Biospecimen Preservation Type of fixation/preservation Formalin (buffered)
OCT
Storage Storage duration 0-3 years
4-7 years
8 years
-
Study Purpose
The purpose of this study was to investigate the effects of section thickness and fixation duration on RNAscope signals in frozen breast tumor specimens embedded in OCT. Thirty-two archival (stored for 0-7 years) breast cancer tumor specimens frozen in OCT cryo-gel were evaluated (no further details provided). All specimens were confirmed to contain >50% cancer cells. Sections that were 7 µM and 10 µM thick were mounted on Superfrost Plus slides for comparison. To optimize fixation, 7 µM sections were fixed in: 1) 4% paraformaldehyde (PFA) versus 10% neutral buffered formalin (NBF) (temperature and duration not specified, 2) 4% PFA at room temperature versus 4°C (duration not specified), and 3) 4% PFA for 20 min versus 30 min (temperature not specified, likely room temperature). Levels of UBC, PPIB, POLR2A, and HPRT1 were assayed using the RNAscope Multiplex Fluorescent v2 Kit. RNAscope signals were imaged within 2 weeks of the assay using the Vectra Polaris Automated Quantitative Pathology Imaging System, and 3-5 regions of interest were captured at 40X magnification for analysis. Images were analyzed using CellProfiler software. H-scores were calculated based on transcript count per cell for each gene.
Summary of Findings:
FISH signal was more distinct and the number of dots were higher for PPIB (P=0.032) due to better signal separation when 7 µM thick sections rather than 10 µM thick sections of frozen OCT-embedded tumor specimens were used, but there was no difference in the number of cells counted or UBC, POLR2A or HPRT1 dots. Nuclear DAPI staining was stronger when tumor specimens were fixed in 4% PFA than 10% NBF. Fixation of OCT-embedded frozen specimens in 4% PFA at room temperature led to much higher signal than observed in specimens fixed at 4°C. Delineations between cells were more distinct when fixed in 4% PFA for 20 min compared to those fixed for 30 min.
Biospecimens
Preservative Types
- Formalin
- Frozen
Diagnoses:
- Neoplastic - Carcinoma
Platform:
Analyte Technology Platform RNA In situ hybridization Pre-analytical Factors:
Classification Pre-analytical Factor Value(s) Biospecimen Aliquots and Components Aliquot size/volume 7 µM
10 µM
Biospecimen Preservation Type of fixation/preservation Paraformaldehyde
Formalin (buffered)
Biospecimen Preservation Time in fixative 20 min
30 min
Biospecimen Preservation Temperature of fixation/preservation Room temperature
4°C