The genomic complexity of primary human prostate cancer

Abstract

Prostate cancer is the second most common cause of male cancer deaths in the United States. However, the full range of prostate cancer genomic alterations is incompletely characterized. Here we present the complete sequence of seven primary human prostate cancers and their paired normal counterparts. Several tumours contained complex chains of balanced (that is, 'copy-neutral') rearrangements that occurred within or adjacent to known cancer genes. Rearrangement breakpoints were enriched near open chromatin, androgen receptor and ERG DNA binding sites in the setting of the ETS gene fusion TMPRSS2-ERG, but inversely correlated with these regions in tumours lacking ETS fusions. This observation suggests a link between chromatin or transcriptional regulation and the genesis of genomic aberrations. Three tumours contained rearrangements that disrupted CADM2, and four harboured events disrupting either PTEN (unbalanced events), a prostate tumour suppressor, or MAGI2 (balanced events), a PTEN interacting protein not previously implicated in prostate tumorigenesis. Thus, genomic rearrangements may arise from transcriptional or chromatin aberrancies and engage prostate tumorigenic mechanisms.

Figure 1

Graphical representation of 7 prostate cancer genomes. Each Circos plot depicts the genomic location in the outer ring and chromosomal copy number in the inner ring (red = copy gain; blue = copy loss). Interchromosomal translocations and intrachromosomal rearrangements are shown in purple and green, respectively. Genomes are organized according to the presence (top row) or absence (bottom row) of the TMPRSS2-ERG gene fusion.

Figure 2

Complex structural rearrangements in prostate cancer. (a) Schematic of “closed chain” pattern of chromosomal breakage and rejoining. Breaks are induced in a set of loci (left), followed by an exchange of free ends without loss of chromosomal material (middle), leading to the observed pattern of balanced (copy neutral) translocations involving a closed set of breakpoints (right). (b) Complex rearrangement in prostate PR-1701. TMPRSS2-ERG is produced by a closed quartet of balanced rearrangements involving 4 loci on chromosomes 1 and 21. Top: Each rearrangement is supported by the presence of discordant read pairs in the tumor genome but not the normal genome (colored bars connected by blue lines). Thin bars represent sequence reads; directionality represents mapping orientation on the reference genome. Figures are based on the Integrative Genomics Viewer (http://www.broadinstitute.org/igv). Bottom: Schematic of breakpoints and balanced translocations. Hatched lines indicate sequences that are duplicated in the derived chromosomes at the resulting fusion junctions. (c) Complex rearrangement in prostate PR-2832 involving breakpoints and fusions at 9 distinct genomic loci. Hatched lines indicate sequences that are duplicated or deleted in the derived chromosomes at the resulting fusion junctions. For breakpoints in intergenic regions, the nearest gene in each direction is shown. In addition to the sheer number of regions involved, this complex rearrangement is notable for the abundance of breakpoints in or near cancer related genes, such as TBK1, MAP2K4, TP53, and ABL1.

Figure 3

Association between rearrangement breakpoints and genome-wide transcriptional/histone marks in prostate cancer. ChIP-Seq binding peaks were defined previously for the TMPRSS2-ERG positive (ERG+) prostate cancer cell line VCaP. For each genome, enrichment of breakpoints within 50 kb of each set of binding peaks was determined relative to a coverage-matched simulated background (see Methods). ERG+ prostate tumors are in black; ETS-negative prostate tumors are in white. Enrichment is displayed as the ratio of the observed breakpoint rate to the background rate near each indicated set of ChIP-Seq peaks. Rearrangements in ETS-negative tumors are depleted near marks of active transcription (AR, ERG, H3K4me3, H3K36me3, Pol II, and H3ace) and enriched near marks of closed chromatin (H3K27me3). P-values were calculated according to the binomial distribution and are displayed in Supplementary Figure S5 and Supplementary Table 6. Significant associations passing a false discovery rate cutoff of 5% are marked with an asterisk.

Figure 4

Disruption of CADM2 and the PTEN pathway by rearrangements. (a) Location of intragenic breakpoints in CADM2. (b) CADM2 break-apart demonstrated by FISH in an independent prostate tumor. (c) Location of intragenic breakpoints in PTEN (top) and MAGI2 (bottom). (d) MAGI2 inversion demonstrated by FISH in an independent prostate tumor, using probes flanking MAGI2 (red and green) and an external reference probe also on chromosome 7q (green). The probes and strategy for detecting novel rearrangements by FISH are diagrammed in Supplementary Figure S8.