Hayley C. Whitaker

Analysis of the genetic phylogeny of multifocal prostate cancer identifies multiple independent clonal expansions in neoplastic and morphologically normal prostate tissue

Genome-wide DNA sequencing was used to decrypt the phylogeny of multiple samples from distinct areas of cancer and morphologically normal tissue taken from the prostates of three men. Mutations were present at high levels in morphologically normal tissue distant from the cancer, reflecting clonal expansions, and the underlying mutational processes at work in morphologically normal tissue were also at work in cancer. Our observations demonstrate the existence of ongoing abnormal mutational processes, consistent with field effects, underlying carcinogenesis. This mechanism gives rise to extensive branching evolution and cancer clone mixing, as exemplified by the coexistence of multiple cancer lineages harboring distinct ERG fusions within a single cancer nodule. Subsets of mutations were shared either by morphologically normal and malignant tissues or between different ERG lineages, indicating earlier or separate clonal cell expansions. Our observations inform on the origin of multifocal disease and have implications for prostate cancer therapy in individual cases.

Extensive transduction of nonrepetitive DNA mediated by L1 retrotransposition in cancer genomes

Introduction The human genome is peppered with mobile repetitive elements called long interspersed nuclear element–1 (L1) retrotransposons. Propagating through RNA and cDNA intermediates, these molecular parasites copy and insert themselves throughout the genome, with potentially disruptive effects on neighboring genes or regulatory sequences. In the germ line, unique sequence downstream of L1 elements can also be retrotransposed if transcription continues beyond the repeat, a process known as 3′ transduction. There has been growing interest in retrotransposition and 3′ transduction as a possible source of somatic mutations during tumorigenesis. The activity of individual L1 elements fluctuates during tumor evolution. In a lung tumor, hundreds of 3′ transductions arose from a small number of active L1 source elements (colored circles on outer rim of circle). As the tumor evolved from the preinvasive common ancestor to invasive cancer, individual elements exhibited variable activity over time. Rationale To explore whether 3′ transductions are frequent in cancer, we developed a bioinformatic algorithm for identifying somatically acquired retrotranspositions in cancer genomes. We applied our algorithm to 290 cancer samples from 244 patients across 12 tumor types. The unique downstream sequence mobilized with 3′ transductions effectively fingerprints the L1 source element, providing insights into the activity of individual L1 loci across the genome. Results Across the 290 samples, we identified 2756 somatic L1 retrotranspositions. Tumors from 53% of patients had at least one such event, with colorectal and lung cancers being most frequently affected (93% and 75% of patients, respectively). Somatic 3′ transductions comprised 24% of events, half of which represented mobilizations of unique sequence alone, without any accompanying L1 sequence. Overall, 95% of 3′ transductions identified derived from only 72 germline L1 source elements, with as few as four loci accounting for 50% of events. In a given sample, the same source element could generate 50 or more somatic transductions, scattered extensively across the genome. About 5% of somatic transductions arose from L1 source elements that were themselves somatic retrotranspositions. In three of the cases in which we sequenced more than one sample from a patient’s tumor, we were able to place 3′ transductions on the phylogenetic tree. We found that the activity of individual source elements fluctuated during tumor evolution, with different subclones exhibiting much variability in which elements were “on” and which were “off.” The ability to identify the individual L1 source elements active in a given tumor enabled us to study the promoter methylation of those elements specifically. We found that 3′ transduction activity in a patient’s tumor was always associated with hypomethylation of that element. Overall, 2.3% of transductions distributed exons or entire genes to other sites in the genome, and many more mobilized deoxyribonuclease I (DNAse-I) hypersensitive sites or transcription factor binding sites identified by the ENCODE project. Occasionally, somatic L1 insertions inserted near coding sequence and redistributed these exons elsewhere in the genome. However, we found no general effects of retrotranspositions on transcription levels of genes at the insertion points and no evidence for aberrant RNA species resulting from somatically acquired transposable elements. Indeed, as with germline retrotranspositions, somatic insertions exhibited a strong enrichment in heterochromatic, gene-poor regions of the genome. Conclusion Somatic 3′ transduction occurs frequently in human tumors, and in some cases transduction events can scatter exons, genes, and regulatory elements widely across the genome. Dissemination of these sequences appears to be due to a small number of highly active L1 elements, whose activity can wax and wane during tumor evolution. The majority of the retrotransposition events are likely to be harmless “passenger” mutations. Hitchhiking through the tumor genome Retrotransposons are DNA repeat sequences that are constantly on the move. By poaching certain cellular enzymes, they copy and insert themselves at new sites in the genome. Sometimes they carry along adjacent DNA sequences, a process called 3′ transduction. Tubio et al. found that 3′ transduction is a common event in human tumors. Because this process can scatter genes and regulatory sequences across the genome, it may represent yet another mechanism by which tumor cells acquire new mutations that help them survive and grow. Science, this issue p. 10.1126/science.1251343 Tumor genomes are peppered with mobile repeat sequences that carry along adjacent DNA when they insert into new genomic sites. Long interspersed nuclear element–1 (L1) retrotransposons are mobile repetitive elements that are abundant in the human genome. L1 elements propagate through RNA intermediates. In the germ line, neighboring, nonrepetitive sequences are occasionally mobilized by the L1 machinery, a process called 3′ transduction. Because 3′ transductions are potentially mutagenic, we explored the extent to which they occur somatically during tumorigenesis. Studying cancer genomes from 244 patients, we found that tumors from 53% of the patients had somatic retrotranspositions, of which 24% were 3′ transductions. Fingerprinting of donor L1s revealed that a handful of source L1 elements in a tumor can spawn from tens to hundreds of 3′ transductions, which can themselves seed further retrotranspositions. The activity of individual L1 elements fluctuated during tumor evolution and correlated with L1 promoter hypomethylation. The 3′ transductions disseminated genes, exons, and regulatory elements to new locations, most often to heterochromatic regions of the genome.

Origins and functional consequences of somatic mitochondrial DNA mutations in human cancer.

Recent sequencing studies have extensively explored the somatic alterations present in the nuclear genomes of cancers. Although mitochondria control energy metabolism and apoptosis, the origins and impact of cancer-associated mutations in mtDNA are unclear. In this study, we analyzed somatic alterations in mtDNA from 1675 tumors. We identified 1907 somatic substitutions, which exhibited dramatic replicative strand bias, predominantly C > T and A > G on the mitochondrial heavy strand. This strand-asymmetric signature differs from those found in nuclear cancer genomes but matches the inferred germline process shaping primate mtDNA sequence content. A number of mtDNA mutations showed considerable heterogeneity across tumor types. Missense mutations were selectively neutral and often gradually drifted towards homoplasmy over time. In contrast, mutations resulting in protein truncation undergo negative selection and were almost exclusively heteroplasmic. Our findings indicate that the endogenous mutational mechanism has far greater impact than any other external mutagens in mitochondria and is fundamentally linked to mtDNA replication. DOI: http://dx.doi.org/10.7554/eLife.02935.001

Engrailed-2 (EN2): A Tumor Specific Urinary Biomarker for the Early Diagnosis of Prostate Cancer

Purpose: Prostate cancer (PC) is the second most common cause of cancer related death in men. A number of key limitations with prostate specific antigen (PSA), currently the standard detection test, has justified evaluation of new biomarkers. We have assessed the diagnostic potential of Engrailed-2 (EN2) protein, a homeodomain-containing transcription factor expressed in PC cell lines and secreted into the urine by PC in men. Experimental Design: EN2 expression in PC cell lines and prostate cancer tissue was determined by semi-quantative RT-PCR and immunohistochemistry. First pass urine [without prior digital rectal examination (DRE)] was collected from men presenting with urinary symptoms (referred to exclude/confirm the presence of prostate cancer) and from controls. EN2 protein was measured by ELISA in urine from men with PC (n = 82) and controls (n = 102). Results: EN2 was expressed and secreted by PC cell lines and PC tissue but not by normal prostate tissue or stroma. The presence of EN2 in urine was highly predictive of PC, with a sensitivity of 66% and a specificity of 88.2%, without requirement for DRE. There was no correlation with PSA levels. These results were confirmed independently by a second academic center. Conclusions: Urinary EN2 is a highly specific and sensitive candidate biomarker of prostate cancer. A larger multicenter study to further evaluate the diagnostic potential of EN2 is justified. Clin Cancer Res; 17(5); 1090–8. ©2011 AACR.

Alterations in beta-catenin expression and localization in prostate cancer

Wnt signaling is thought to be important in prostate cancer, in part because proteins such as β‐catenin can also affect androgen receptor signaling. β‐Catenin forms a cell adhesion complex with E‐cadherin raising the possibility that loss of expression or a change in β‐catenin distribution in the cell could also alter downstream signaling, decreased inter‐cellular adhesion and the promotion of metastasis. A number of studies have reported the altered expression and/or localization of β‐catenin as a biomarker in prostate cancer.

The rs10993994 risk allele for prostate cancer results in clinically relevant changes in microseminoprotein-beta expression in tissue and urine.

Background Microseminoprotein-beta (MSMB) regulates apoptosis and using genome-wide association studies the rs10993994 single nucleotide polymorphism in the MSMB promoter has been linked to an increased risk of developing prostate cancer. The promoter location of the risk allele, and its ability to reduce promoter activity, suggested that the rs10993994 risk allele could result in lowered MSMB in benign tissue leading to increased prostate cancer risk. Methodology/Principal Findings MSMB expression in benign and malignant prostate tissue was examined using immunohistochemistry and compared with the rs10993994 genotype. Urinary MSMB concentrations were determined by ELISA and correlated with urinary PSA, the presence or absence of cancer, rs10993994 genotype and age of onset. MSMB levels in prostate tissue and urine were greatly reduced with tumourigenesis. Urinary MSMB was better than urinary PSA at differentiating men with prostate cancer at all Gleason grades. The high risk allele was associated with heterogeneity of MSMB staining and loss of MSMB in both tissue and urine in benign prostate. Conclusions These data show that some high risk alleles discovered using genome-wide association studies produce phenotypic effects with potential clinical utility. We provide the first link between a low penetrance polymorphism for prostate cancer and a potential test in human tissue and bodily fluids. There is potential to develop tissue and urinary MSMB for a biomarker of prostate cancer risk, diagnosis and disease monitoring.

LYRIC/AEG-1 Is Targeted to Different Subcellular Compartments by Ubiquitinylation and Intrinsic Nuclear Localization Signals

Purpose: LYRIC/AEG-1 has been reported to influence breast cancer survival and metastases, and its altered expression has been found in a number of cancers. The cellular function of LYRIC/AEG-1 has previously been related to its subcellular distribution in cell lines. LYRIC/AEG-1 contains three uncharacterized nuclear localization signals (NLS), which may regulate its distribution and, ultimately, function in cells. Experimental Design: Immunohistochemistry of a human prostate tissue microarray composed of 179 prostate cancer and 24 benign samples was used to assess LYRIC/AEG-1 distribution. Green fluorescent protein-NLS fusion proteins and deletion constructs were used to show the ability of LYRIC/AEG-1 NLS to target green fluorescent protein from the cytoplasm to the nucleus. Immunoprecipitation and Western blotting were used to show posttranslational modification of LYRIC/AEG-1 NLS regions. Results: Using a prostate tissue microarray, significant changes in the distribution of LYRIC/AEG-1 were observed in prostate cancer as an increased cytoplasmic distribution in tumors compared with benign tissue. These differences were most marked in high grade and aggressive prostate cancers and were associated with decreased survival. The COOH-terminal extended NLS-3 (amino acids 546-582) is the predominant regulator of nuclear localization, whereas extended NLS-1 (amino acids 78-130) regulates its nucleolar localization. Within the extended NLS-2 region (amino acids 415-486), LYRIC/AEG-1 can be modified by ubiquitin almost exclusively within the cytoplasm. Conclusions: Changes in LYRIC/AEG-1 subcellular distribution can predict Gleason grade and survival. Two lysine-rich regions (NLS-1 and NLS-3) can target LYRIC/AEG-1 to subcellular compartments whereas NLS-2 is modified by ubiquitin in the cytoplasm.

Nuclear LYRIC/AEG-1 interacts with PLZF and relieves PLZF-mediated repression.

LYRIC/AEG-1 and its altered expression have been linked to carcinogenesis in prostate, brain and melanoma as well as promoting chemoresistance and metastasis in breast cancer. LYRIC/AEG-1 function remains unclear, although LYRIC/AEG-1 is activated by oncogenic HA-RAS, through binding of c-myc to its promoter, which in turn regulates the key components of the PI3-kinase and nuclear factor-κB pathways. We have identified the transcriptional repressor PLZF as an interacting protein of LYRIC/AEG through a yeast two-hybrid screen. PLZF regulates the expression of genes involved in cell growth and apoptosis including c-myc. Coexpression of LYRIC/AEG-1 with PLZF leads to a reduction in PLZF-mediated repression by reducing PLZF binding to promoters. We have confirmed that nuclear LYRIC/AEG-1 and PLZF interact in mammalian cells via the N- and C termini of LYRIC/AEG-1 and a region C terminal to the RD2 domain of PLZF. Both proteins colocalize to nuclear bodies containing histone deacetylases, which are known to promote PLZF-mediated repression. Our data suggest one mechanism for cells with altered LYRIC/AEG-1 expression to evade apoptosis and increase cell growth during tumourigenesis through the regulation of PLZF repression.

Prognostic and Therapeutic Impact of Argininosuccinate Synthetase 1 Control in Bladder Cancer as Monitored Longitudinally by PET Imaging

Targeted therapies have yet to have significant impact on the survival of patients with bladder cancer. In this study, we focused on the urea cycle enzyme argininosuccinate synthetase 1 (ASS1) as a therapeutic target in bladder cancer, based on our discovery of the prognostic and functional import of ASS1 in this setting. ASS1 expression status in bladder tumors from 183 Caucasian and 295 Asian patients was analyzed, along with its hypothesized prognostic impact and association with clinicopathologic features, including tumor size and invasion. Furthermore, the genetics, biology, and therapeutic implications of ASS1 loss were investigated in urothelial cancer cells. We detected ASS1 negativity in 40% of bladder cancers, in which multivariate analysis indicated worse disease-specific and metastasis-free survival. ASS1 loss secondary to epigenetic silencing was accompanied by increased tumor cell proliferation and invasion, consistent with a tumor-suppressor role for ASS1. In developing a treatment approach, we identified a novel targeted antimetabolite strategy to exploit arginine deprivation with pegylated arginine deiminase (ADI-PEG20) as a therapeutic. ADI-PEG20 was synthetically lethal in ASS1-methylated bladder cells and its exposure was associated with a marked reduction in intracellular levels of thymidine, due to suppression of both uptake and de novo synthesis. We found that thymidine uptake correlated with thymidine kinase-1 protein levels and that thymidine levels were imageable with [(18)F]-fluoro-L-thymidine (FLT)-positron emission tomography (PET). In contrast, inhibition of de novo synthesis was linked to decreased expression of thymidylate synthase and dihydrofolate reductase. Notably, inhibition of de novo synthesis was associated with potentiation of ADI-PEG20 activity by the antifolate drug pemetrexed. Taken together, our findings argue that arginine deprivation combined with antifolates warrants clinical investigation in ASS1-negative urothelial and related cancers, using FLT-PET as an early surrogate marker of response.

Integration of copy number and transcriptomics provides risk stratification in prostate cancer: A discovery and validation cohort study

Background Understanding the heterogeneous genotypes and phenotypes of prostate cancer is fundamental to improving the way we treat this disease. As yet, there are no validated descriptions of prostate cancer subgroups derived from integrated genomics linked with clinical outcome. Methods In a study of 482 tumour, benign and germline samples from 259 men with primary prostate cancer, we used integrative analysis of copy number alterations (CNA) and array transcriptomics to identify genomic loci that affect expression levels of mRNA in an expression quantitative trait loci (eQTL) approach, to stratify patients into subgroups that we then associated with future clinical behaviour, and compared with either CNA or transcriptomics alone. Findings We identified five separate patient subgroups with distinct genomic alterations and expression profiles based on 100 discriminating genes in our separate discovery and validation sets of 125 and 103 men. These subgroups were able to consistently predict biochemical relapse (p = 0.0017 and p = 0.016 respectively) and were further validated in a third cohort with long-term follow-up (p = 0.027). We show the relative contributions of gene expression and copy number data on phenotype, and demonstrate the improved power gained from integrative analyses. We confirm alterations in six genes previously associated with prostate cancer (MAP3K7, MELK, RCBTB2, ELAC2, TPD52, ZBTB4), and also identify 94 genes not previously linked to prostate cancer progression that would not have been detected using either transcript or copy number data alone. We confirm a number of previously published molecular changes associated with high risk disease, including MYC amplification, and NKX3-1, RB1 and PTEN deletions, as well as over-expression of PCA3 and AMACR, and loss of MSMB in tumour tissue. A subset of the 100 genes outperforms established clinical predictors of poor prognosis (PSA, Gleason score), as well as previously published gene signatures (p = 0.0001). We further show how our molecular profiles can be used for the early detection of aggressive cases in a clinical setting, and inform treatment decisions. Interpretation For the first time in prostate cancer this study demonstrates the importance of integrated genomic analyses incorporating both benign and tumour tissue data in identifying molecular alterations leading to the generation of robust gene sets that are predictive of clinical outcome in independent patient cohorts.