3.1. Tissue Storage
Prostatectomy material was obtained from a patient treated for localized prostate cancer at the University Hospital of Tuebingen, Germany under an Institutional Review Board approved protocol (395/2008BO1). The cancer tissue was cut into two equal parts. Subsequently, one portion was fixed in 10% neutral buffered formalin and embedded in paraffin and the other portion was fixed using the cryo-conservation method. For the same patient, matched normal prostatic tissue was also fixed using the FFPE protocol.
3.2. DNA Extraction and Preparation
The FFPE normal prostatic, FFPE tumor and fresh frozen specimen were cut into 3 μm thick sections and stained with hematoxylin and eosin. The sections were assessed by a pathologist (S.P.) to identify the tumor region (tissue containing > 80% tumor cells) or the absence of tumor (i.e., normal prostatic tissue). A 3 mm biopsy needle was then used to punch three cores from each sample for DNA extraction. Core punches, restricted to the tumor region, were performed rather than tissue sections to maintain the homogeneity of the tumor sample. The three cores for each sample were pooled and DNA isolation was performed using the RecoverAll™ Total Nucleic Acid Isolation Kit (Ambion) for the FFPE samples and the PureLink™ Genomic DNA Kit for the fresh frozen specimen.
3.3. Library Construction
Three micrograms of each DNA sample was treated to obtain the SOLiD pre-capture library according to the manufacturer’s protocol (Applied Biosystems, Inc.). DNA was sheared using the Covaris S2 to produce fragments with a base pair target range of 150–180 (Covaris, Inc.). The fragments were end repaired and purified using the SOLiD Library Column Purification Kit (Applied Biosystems). The resulting blunt-ended fragments were then ligated to P1 and P2 adaptors. The desired fragment lengths were obtained by running the samples through precast gels (E-Gel SizeSelect 2%, Invitrogen by Life Technologies) for size selection. The purified adapter ligated fragments were subjected to nick translation and library amplification using Platinum PCR Amplification Mix (Invitrogen by Life Technologies) and P1 and P2 amplification primers with 12 amplification cycles, to obtain the genomic pre-captured library.
The libraries were quantified using the Agilent Bioanalyzer DNA 1000 chip and the High-Sensitivity DNA chip.
3.4. Targeted Capture and Exome Sequencing
Targeted enrichment was performed with an ABI SOLiD optimized SureSelect human whole exome kit (Agilent SureSelect All Exon G3362, version 1, ELID 027495). The kit is designed to enrich for 165,637 exons (~18,000 genes) covering a total of 37-Mb genomic sequences. Capture libraries were hybridized in solution according to the SureSelect Target Enrichment protocol for the Applied Biosystems SOLiD System (SureSelect Human All Exon Kit and Custom Designs; Version 1.5.1, April 2010). The prepared exome library was further used for emulsion PCRs following the manufacturer’s instructions (Applied Biosystems SOLiD™ 4 System Templated Bead Preparation Guide by life technologies, March 2010) based on a library concentration of 0.5 pM. For each sample, one quad of a SOLiD sequencing slide (Life Technologies) was processed for sequencing as single-end 50-bp reads. Approximately 6 to 8 Gbs of sequence was generated per capture library on the SOLiD4 system.
3.5. Bioinformatic Analyses
The sequencing color space reads were mapped to the reference human genome (UCSC hg19) using the BLAT-like fast accurate search tool (BFAST, v0.6.5a) [17
]. SAM-files have been filtered using SAMTools (v0.1.15) [18
] with the following criteria: PHRED-like consensus of ≥ 30, removal of PCR duplicates. Using the GATKs (v1.1-37-ge63d9d8) [19
] default pipeline, a realignment was performed. SNVs were subsequently called using GATK, using default parameters [20
]. Also a SNV had to be present in ≥ 25% in the reads at the position, to be declared as an SNV.
3.6. Determination of Copy Number Variations
Comparing the FFPE tumor sample against FFPE normal prostatic sample and fresh frozen tumor against FFPE normal prostatic sample, the copy number variation analysis was performed. For this a window of length 350 has been shifted over each position of the samples. For each window the ratio between normal and tumor tissue has been calculated. Furthermore, for normalization purposes, the log2 has been applied to these ratios.
For prioritization and annotation snpEff (v2.0.4) [21
] has been used, using default parameters and including ENSEMBLE release 64.