Methodology article

Optimization of laser capture microdissection and RNA amplification for gene expression profiling of prostate cancer

Dagmar M Kube^1,5 , Cemile D Savci-Heijink¹ , Anne-Françoise Lamblin⁴ , Farhad Kosari^1,2 , George Vasmatzis^1,2 , John C Cheville^1,2 , Donald P Connelly^3,4 and George G Klee^1,2

¹Department of Laboratory Medicine and Pathology, Mayo Clinic College of Medicine, 200 First St. S.W., Rochester, Minnesota 55905, USA

²Comprehensive Cancer Center, Mayo Clinic College of Medicine, 200 First St. S.W., Rochester, Minnesota 55905, USA

³Department of Laboratory Medicine and Pathology, University of Minnesota, 425 Delaware St. S.E., Minneapolis, Minnesota 55455, USA

⁴Cancer Center Informatics Shared Resource, University of Minnesota, 425 Delaware St. S.E., Minneapolis, Minnesota 55455, USA

⁵Fish & Richardson P.C., 60 South Sixth Street, Minneapolis, Minnesota 55402, USA

author email corresponding author email

BMC Molecular Biology 2007, 8:25doi:10.1186/1471-2199-8-25

Published:	21 March 2007

Abstract

Background

To discover prostate cancer biomarkers, we profiled gene expression in benign and malignant cells laser capture microdissected (LCM) from prostate tissues and metastatic prostatic adenocarcinomas. Here we present methods developed, optimized, and validated to obtain high quality gene expression data.

Results

RNase inhibitor was included in solutions used to stain frozen tissue sections for LCM, which improved RNA quality significantly. Quantitative PCR assays, requiring minimal amounts of LCM RNA, were developed to determine RNA quality and concentration. SuperScript II™ reverse transcriptase was replaced with SuperScript III™, and SpeedVac concentration was eliminated to optimize linear amplification. The GeneChip^® IVT labeling kit was used rather than the Enzo BioArray™ HighYield™ RNA transcript labeling kit since side-by-side comparisons indicated high-end signal saturation with the latter. We obtained 72 μg of labeled complementary RNA on average after linear amplification of about 2 ng of total RNA.

Conclusion

Unsupervised clustering placed 5/5 normal and 2/2 benign prostatic hyperplasia cases in one group, 5/7 Gleason pattern 3 cases in another group, and the remaining 2/7 pattern 3 cases in a third group with 8/8 Gleason pattern 5 cases and 3/3 metastatic prostatic adenocarcinomas. Differential expression of alpha-methylacyl coenzyme A racemase (AMACR) and hepsin was confirmed using quantitative PCR.