Prostate cancer is one of the most common forms of cancer diagnosed each year and has a high mortality rate. The current clinical standard for diagnosis of prostate cancer is based on histological analysis and by using the Gleason score. These methods of diagnosis can be fairly subjective depending on the situation. The researcher proposed that, because cancer is caused by multiple genetic aberrations, next-generation sequencing technologies can be an ideal tool for identifying and examining these aberrations simultaneously.

The researcher used two different sample types, snap frozen and paraffin preserved to increase the sample size and to establish that both sample types were suitable for sequencing and gave similar quality of results. This has an important consequence in increasing the use of paraffin samples for resequencing and the use of preserved clinical tissue samples. The sequencing samples were used in two projects—SureSelect 3.5 Mb Enrichment using the Illumina Genome Analyzer platform and using SureSelect Human All Exon Enrichment with Applied Biosystems™ SOLiD Sequencing technologies—to identify non-silent small nuclear variations that are present in the prostate cancer samples.

In the first approach using SureSelect, the researchers used a small target region of 3.5 Mb and the enrichment regions were limited to exons. After the fragmentation, the regions chosen were only those fragments of DNA that contained exons. The library thus generated was used with SureSelect probes to select all relevant cancer genes and susceptibility loci that were found by previous genome-wide association studies. Three different sample sizes were used (0.5 μg, 1.5 μg and 3.0 μg), and all three sample sizes were able to identify > 98% of targets in the extended targeted regions (Table 1). This result indicates that the SureSelect platform is capable of very high readouts even with small sample sizes and is a suitable clinical diagnostic tool for identifying genetic variation in prostate cancer samples.

Table 1. Enrichment of 13,307 prostate cancer specific regions: Effect of sample size. Three different sample sizes were used (0.5 μg, 1.5 μg and 3.0 μg).


The cumulative normalized coverage plot shows that about 93% of the targeted regions were enriched about 5-fold (Figure 1). This means that more than 93% of the bases within baits received more than one-fifth of the mean coverage. This shows that the coverage achieved using the SureSelect platform was uniform.  Using the small targeting regions of 3.9 Mb, the study identified an average of 5000 SNVs out of which 600 were unknown and 100 were nonsilent tumor- specific SNVs (Table 2). The DNA enrichment method worked for both cryo and FFPE samples. The percent enrichment in the target regions obtained from both sample types was similar, suggesting that both methods of preservation can be used for producing high-quality results



Figure 1. Cumulative coverage plot showing the fraction of bases covered against the normalized coverage.


Table 2. Summary of individual patient results.

The other approach was to use the Applied Biosystems™ SOLiD platform in a whole exome enrichment to identify the genes that are behind the cause of cancer. In the SOLiD sequencing approach, after the ligase reaction, the di-base encoding called each nucleotide position twice, therefore, the error rate was very low. The results were highly repetitive and both normal and tumor samples gave a 75-fold enrichment (maximum enrichment is 80-fold) (Table 3). The unique mappable reads were in the order of 53.5x106 in normal tissues and 45.0x106 in tumor samples.


Table 3. Whole exome Enrichment SOLiD sequencing.


The cumulative normalized coverage plot shows that 90% of all target regions had a 5x coverage for normal tissue, and that tumor tissue had about 86% of all target regions, with a 5x coverage (Figure 2). This indicates that, for both normal and tumor tissue, the uniformity of coverage is excellent on the SOLiD platform.



Figure 2. Cumulative normalized coverage plots for normal (left) and tumor (right) tissue samples.

Using the whole exome enrichment sequencing data, the number of SNVs in the targeted regions was approximately 16,000 SNVs in both normal and tumor tissues. In the extended regions, this number was closer to 22,000.

The researchers have established that the next-gen sequencing technologies for Agilent’s SureSelect Target Enrichment System and SOLiD platforms are powerful tools for identifying genetic variations in cancer samples, using prostate cancer samples as an example. They have further established that both cryo-preserved and paraffin-preserved samples give the same high-quality results, thus increasing the possible number of samples that can be used in these studies. Researchers are currently sequencing more than 100 clinical samples. The sequencing data will then be analyzed using a combination of computational and experimental techniques to identify the genetic makeup of prostate cancer.

To watch this webinar, click here.