The development of an efficient procedure for protein extraction from formalin fixed and paraffin embedded tissues and analysis of its application to reverse phase protein arrays

Abstract

Abstract Purpose: Formalin fixation coupled with paraffin embedding (FFPE) is the standard tissue fixation and storage method adopted by most health institutions in the world. However, protein extraction from FFPE tissues has been challenging due to the extensive molecular cross-linking that occurs upon formalin fixation. The reverse-phase protein array (RPPA) is a novel high-throughput technology which can detect changes in protein levels and protein functionality in numerous tissue and cell sources. It has been used to evaluate protein expression mainly in frozen preparations. Reproducibility and reliability of the technique in FFPE samples has not been demonstrated extensively so far. The purpose of this study is to develop and optimize an efficient and reproducible procedure for extraction of proteins from FFPE tissues, and evaluate its performance on RPPA. Materials and Methods: Fresh frozen and FFPE material was obtained from breast cancer cell lines, xenografts, renal and breast cancer patient tissues, as well as normal renal cortex tissue. Serial FFPE cell or tissue sections were deparaffinized and processed using six different protein extraction protocols. The yield and protein quality from these protocols were evaluated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), Coomassie blue staining and Western Blots. Protein lysates were then subjected to reverse-phase protein array profiling by using the most efficient protocol. The Pearson correlation coefficient between replicate preparations or between FFPE and matched frozen samples was calculated, and its statistical significance was determined (R software). Unsupervised hierarchical clustering and heatmaps were generated by the R package Class Discovery. Results: The protein extraction protocol that produced the highest yield and best protein quality included a boiling step in 2mM Tris buffer containing 2% sodium dodecyl sulfate at pH 9.0 for 20 minutes, followed by incubation in 60°C for 2 hours. This method was used to prepare lysates for the RPPA assay. The correlation coefficient between replicate preparations across the tested protein markers demonstrated high reproducibility of the technique, with the majority of the markers showing coefficients equal or more than 0.7. The protocol enables efficient extraction of immunoreactive protein from all four experimental settings, with the majority of protein markers (119 of the 203 tested) showing significant correlation between FFPE and frozen preparations (R>0.5). Protein markers whose expression is specifically influenced by particular cell culture conditions and pharmacological interventions also proved that results derived from FFPE specimens are biologically relevant and reliable. Importantly, the RPPA-based global protein expression profile of the FFPE preparations was able to distinguish between normal renal and tumor samples, and categorize the patient samples into distinct groups slightly more successfully than their frozen counterparts. Conclusions: Our results indicate that, with the optimized protein extraction method, FFPE tissues can be a valuable source in generating reproducible and biologically relevant proteomic profiles using reverse-phase protein arrays (RPPA).