A novel method for RNA extraction from FFPE samples reveals significant differences in biomarker expression between orthotopic and subcutaneous pancreatic cancer patient-derived xenografts.

Malachia Hoover,Jason Fleming,Jonathan A Kelber,Yvess Adamian,Matthew H Katz,Robert Doebler,Michael Bouvet,Jacqueline Lee,Ali Maawy,Mark Brown,Robert M Hoffman,Alexander Chang,Armen Gharibi

doi:10.18632/oncotarget.11809

Abstract

Next-generation sequencing (NGS) can identify and validate new biomarkers of cancer onset, progression and therapy resistance. Substantial archives of formalin-fixed, paraffin-embedded (FFPE) cancer samples from patients represent a rich resource for linking molecular signatures to clinical data. However, performing NGS on FFPE samples is limited by poor RNA purification methods. To address this hurdle, we developed an improved methodology for extracting high-quality RNA from FFPE samples. By briefly integrating a newly-designed micro-homogenizing (mH) tool with commercially available FFPE RNA extraction protocols, RNA recovery is increased by approximately 3-fold while maintaining standard A260/A280 ratios and RNA quality index (RQI) values. Furthermore, we demonstrate that the mH-purified FFPE RNAs are longer and of higher integrity. Previous studies have suggested that pancreatic ductal adenocarcinoma (PDAC) gene expression signatures vary significantly under in vitro versus in vivo and in vivo subcutaneous versus orthotopic conditions. By using our improved mH-based method, we were able to preserve established expression patterns of KRas-dependency genes within these three unique microenvironments. Finally, expression analysis of novel biomarkers in KRas mutant PDAC samples revealed that PEAK1 decreases and MST1R increases by over 100-fold in orthotopic versus subcutaneous microenvironments. Interestingly, however, only PEAK1 levels remain elevated in orthotopically grown KRas wild-type PDAC cells. These results demonstrate the critical nature of the orthotopic tumor microenvironment when evaluating the clinical relevance of new biomarkers in cells or patient-derived samples. Furthermore, this new mH-based FFPE RNA extraction method has the potential to enhance and expand future FFPE-RNA-NGS cancer biomarker studies.

Highlights

Next-Generation Sequencing (NGS), such as RNA-seq or whole-genome analysis, of patientderived tumor tissue in combination with state-of-the art bioinformatics holds great potential to improve disease outcomes by identifying novel biomarkers of cancer onset, progression and therapy resistance www.impactjournals.com/oncotarget [1,2,3]
The combination of our mH technology with the Qiagen Formalin-fixed paraffinembedded (FFPE) RNA extraction kit has significantly improved RNA quality and yield from pancreatic ductal adenocarcinoma (PDAC) patient and cell line derived xenograft FFPE samples
This method has been validated using qPCR to compare gene expression of important PDAC biomarkers previously published from fresh-frozen samples with those generated from our FFPE samples for the same cells and same microenvironments (Figure 3)

Summary

Introduction

Next-Generation Sequencing (NGS), such as RNA-seq or whole-genome analysis, of patientderived tumor tissue in combination with state-of-the art bioinformatics holds great potential to improve disease outcomes by identifying novel biomarkers of cancer onset, progression and therapy resistance www.impactjournals.com/oncotarget [1,2,3]. Among the advantages of NGS over other technologies is the ability to perform high-resolution transcriptome monitoring to identify isoform variations, outlier expression patterns and expression signatures of low expressing genes or long noncoding RNAs [4]. These data can help to differentiate between unique molecular subtypes of various malignancies for the purpose of guiding diagnosis, prognosis and therapeutic interventions [5, 6]. A significant need exists for the development of more efficient methods to recover high-quality nucleic acid material from FFPE samples without sacrificing large amounts of these valuable samples [8]

Methods

Results

Conclusion