Optimizing RNA extraction methods for high-throughput transcriptome sequencing of formalin-fixed paraffin-embedded cardiac tissue specimens

Abstract

Archived FFPE cardiac tissue specimens are valuable for molecular studies aimed at identifying biomarkers linked to mortality in cardiovascular disease. Establishing a reliable and reproducible RNA extraction method is critical for generating high-quality transcriptome sequences for molecular assays. Here, the efficiency of four RNA extraction methods: Qiagen AllPrep DNA/RNA method (Method QP); Qiagen AllPrep DNA/RNA method with protocol modification on the ethanol wash step after deparaffinization (Method QE); CELLDATA RNA extraction (Method BP) and CELLDATA RNA extraction with protocol modifications on the lysis step (Method BL) was compared on 23 matching FFPE cardiac tissue specimens (n = 92).In comparing RNA quality metrics across FFPE RNA extract, nucleic acids extracted deploying Method QE and QP produced the highest RNA yield. However, Method QE outperformed Method QP as more extract from Method QE had DV 200 values above 30%. Both method BL and BP produced similar range of RNA purity and yield but more extract from Method BL had DV 200 values above 30% compared to Method BP. When accessing distribution value, Method BL outperformed Methods BP, QE, and QP as more extracts from Method BL had DV 200 values above 30% compared to other methods (PDV200<0.001; Kruskal-Wallis). Method QE outperformed other methods in terms of RNA yield. RNA extracts from Method QE, characterized by high RNA yield, achieved sequencing results comparable to those from Method BL, characterized by high DV200 values. Our findings reveal that optimizing protocols can yield higher-quality RNA, facilitating the exploration of more disease conditions with high-resolution transcriptome profiling.

Fig 1. Overview of experiment design.

Fig 2. Assessment of RNA quality metrics in cardiac tissue FFPE derived RNA extract.

(A) Box and whisker plots (min to max) of RNA concentration and RNA purity assessment of 260/230 and A260/280 absorbance ratio based on absorption spectra in RNA samples extracted using Method QP versus Method QE (B) Box and whisker plots (min to max) of RNA concentration and RNA purity assessment of 260/230 and A260/280 absorbance ratio based on absorption spectra in RNA samples extracted using Method BL versus Method BP (C) Comparison across four methods of extraction, Box and whisker plots (min to max) of RNA concentration, RNA purity assessment of 260/230 and A260/280 absorbance ratio across method BL and BP vs Method QE and Method QP (D) Electrophoregram/RNA fragment length across the four methods. For all groups, the asterisk (*) denotes a significant difference between groups, Paired Wilcoxon Signed Rank Test (*p<0.05, ***p<0.005, ****p<0.001).

Fig 3. Sequence data comparison between Method BL and Method QE.

(A) Correlation plot of normalized gene and transcript count. Data was normalized by log2(TPM+1). Each dot constitutes a gene or transcript. (B) Distribution of Exon, introns, and rRNA mapped reads (C) Heat map of top 25 genes and transcript across extracts from Method QE and Method BL.

Fig 4. RT-qPCR analysis comparing Method QE and the optimized extraction method (Method BL).

(A) Box plots of individual Ct values for genes MYL2, MYH7, and MYH6 depict significant differences in gene expression detection. For all groups, an asterisk (*) denotes a significant difference between groups, Paired Wilcoxon Signed Rank Test (*p<0.05, ***p<0.005, ****p<0.001).