Cross-Platform Omics Prediction procedure: a statistical machine learning framework for wider implementation of precision medicine.

In this modern era of precision medicine, molecular signatures identified from advanced omics technologies hold great promise to better guide clinical decisions. However, current approaches are often location-specific due to the inherent differences between platforms and across multiple centres, thus limiting the transferability of molecular signatures. We present Cross-Platform Omics Prediction (CPOP), a penalised regression model that can use omics data to predict patient outcomes in a platform-independent manner and across time and experiments. CPOP improves on the traditional prediction framework of using gene-based features by selecting ratio-based features with similar estimated effect sizes. These components gave CPOP the ability to have a stable performance across datasets of similar biology, minimising the effect of technical noise often generated by omics platforms. We present a comprehensive evaluation using melanoma transcriptomics data to demonstrate its potential to be used as a critical part of a clinical screening framework for precision medicine. Additional assessment of generalisation was demonstrated with ovarian cancer and inflammatory bowel disease studies.

Whole-genome landscape of mucosal melanoma reveals diverse drivers and therapeutic targets.

Knowledge of key drivers and therapeutic targets in mucosal melanoma is limited due to the paucity of comprehensive mutation data on this rare tumor type. To better understand the genomic landscape of mucosal melanoma, here we describe whole genome sequencing analysis of 67 tumors and validation of driver gene mutations by exome sequencing of 45 tumors. Tumors have a low point mutation burden and high numbers of structural variants, including recurrent structural rearrangements targeting TERT, CDK4 and MDM2. Significantly mutated genes are NRAS, BRAF, NF1, KIT, SF3B1, TP53, SPRED1, ATRX, HLA-A and CHD8. SF3B1 mutations occur more commonly in female genital and anorectal melanomas and CTNNB1 mutations implicate a role for WNT signaling defects in the genesis of some mucosal melanomas. TERT aberrations and ATRX mutations are associated with alterations in telomere length. Mutation profiles of the majority of mucosal melanomas suggest potential susceptibility to CDK4/6 and/or MEK inhibitors.

Whole-genome landscapes of major melanoma subtypes.

Melanoma of the skin is a common cancer only in Europeans, whereas it arises in internal body surfaces (mucosal sites) and on the hands and feet (acral sites) in people throughout the world. Here we report analysis of whole-genome sequences from cutaneous, acral and mucosal subtypes of melanoma. The heavily mutated landscape of coding and non-coding mutations in cutaneous melanoma resolved novel signatures of mutagenesis attributable to ultraviolet radiation. However, acral and mucosal melanomas were dominated by structural changes and mutation signatures of unknown aetiology, not previously identified in melanoma. The number of genes affected by recurrent mutations disrupting non-coding sequences was similar to that affected by recurrent mutations to coding sequences. Significantly mutated genes included BRAF, CDKN2A, NRAS and TP53 in cutaneous melanoma, BRAF, NRAS and NF1 in acral melanoma and SF3B1 in mucosal melanoma. Mutations affecting the TERT promoter were the most frequent of all; however, neither they nor ATRX mutations, which correlate with alternative telomere lengthening, were associated with greater telomere length. Most melanomas had potentially actionable mutations, most in components of the mitogen-activated protein kinase and phosphoinositol kinase pathways. The whole-genome mutation landscape of melanoma reveals diverse carcinogenic processes across its subtypes, some unrelated to sun exposure, and extends potential involvement of the non-coding genome in its pathogenesis.

Identification, Review, and Systematic Cross-Validation of microRNA Prognostic Signatures in Metastatic Melanoma.

In metastatic melanoma, it is vital to identify and validate biomarkers of prognosis. Previous studies have systematically evaluated protein biomarkers or mRNA-based expression signatures. No such analyses have been applied to microRNA (miRNA)-based prognostic signatures. As a first step, we identified two prognostic miRNA signatures from publicly available data sets (Gene Expression Omnibus/The Cancer Genome Atlas) of global miRNA expression profiling information. A 12-miRNA signature predicted longer survival after surgery for resection of American Joint Committee on Cancer stage III disease (>4 years, no sign of relapse) and outperformed American Joint Committee on Cancer standard-of-care prognostic markers in leave-one-out cross-validation analysis (error rates 34% and 38%, respectively). A similar 15-miRNA biomarker derived from The Cancer Genome Atlas miRNA-seq data performed slightly worse (39%) than these current biomarkers. Both signatures were then assessed for replication in two independent data sets and subjected to systematic cross-validation together with the three other miRNA-based prognostic signatures proposed in the literature to date. Five miRNAs (miR-142-5p, miR-150-5p, miR-342-3p, miR-155-5p, and miR-146b-5p) were reproducibly associated with patient outcome and have the greatest potential for application in the clinic. Our extensive validation approach highlighted among multiple independent cohorts the translational potential and limitations of miRNA signatures, and pointed to future directions in the analysis of this emerging class of markers.

Circulating tumor DNA to monitor treatment response and detect acquired resistance in patients with metastatic melanoma.

Repeat tumor biopsies to study genomic changes during therapy are difficult, invasive and data are confounded by tumoral heterogeneity. The analysis of circulating tumor DNA (ctDNA) can provide a non-invasive approach to assess prognosis and the genetic evolution of tumors in response to therapy. Mutation-specific droplet digital PCR was used to measure plasma concentrations of oncogenic BRAF and NRAS variants in 48 patients with advanced metastatic melanoma prior to treatment with targeted therapies (vemurafenib, dabrafenib or dabrafenib/trametinib combination) or immunotherapies (ipilimumab, nivolumab or pembrolizumab). Baseline ctDNA levels were evaluated relative to treatment response and progression-free survival (PFS). Tumor-associated ctDNA was detected in the plasma of 35/48 (73%) patients prior to treatment and lower ctDNA levels at this time point were significantly associated with response to treatment and prolonged PFS, irrespective of therapy type. Levels of ctDNA decreased significantly in patients treated with MAPK inhibitors (p < 0.001) in accordance with response to therapy, but this was not apparent in patients receiving immunotherapies. We show that circulating NRAS mutations, known to confer resistance to BRAF inhibitors, were detected in 3 of 7 (43%) patients progressing on kinase inhibitor therapy. Significantly, ctDNA rebound and circulating mutant NRAS preceded radiological detection of progressive disease. Our data demonstrate that ctDNA is a useful biomarker of response to kinase inhibitor therapy and can be used to monitor tumor evolution and detect the early appearance of resistance effectors.

Tumour procurement, DNA extraction, coverage analysis and optimisation of mutation-detection algorithms for human melanoma genomes.

Whole genome sequencing (WGS) of cancer patients' tumours offers the most comprehensive method of identifying both novel and known clinically-actionable genomic targets. However, the practicalities of performing WGS on clinical samples are poorly defined.This study was designed to test sample preparation, sequencing specifications and bioinformatic algorithms for their effect on accuracy and cost-efficiency in a large WGS analysis of human melanoma samples.WGS was performed on melanoma cell lines (n = 15) and melanoma fresh frozen tumours (n = 222). The appropriate level of coverage and the optimal mutation detection algorithm for the project pipeline were determined.An incremental increase in sequencing coverage from 36X to 132X in melanoma tissue samples and 30X to 103X for cell lines only resulted in a small increase (1-2%) in the number of mutations detected, and the quality scores of the additional mutations indicated a low probability that the mutations were real. The results suggest that 60X coverage for melanoma tissue and 40X for melanoma cell lines empower the detection of 98-99% of informative single nucleotide variants (SNVs), a sensitivity level at which clinical decision making or landscape research projects can be carried out with a high degree of confidence in the results. Likewise the bioinformatic mutation analysis methodology strongly influenced the number and quality of SNVs detected. Detecting mutations in the blood genomes separate to the tumour genomes generated 41% more SNVs than if the blood and melanoma tissue genomes were analysed simultaneously. Therefore, simultaneous analysis should be employed on matched melanoma tissue and blood genomes to reduce errors in mutation detection.This study provided valuable insights into the accuracy of SNV with WGS at various coverage levels in human clinical cancer specimens. Additionally, we investigated the accuracy of the publicly available mutation detection algorithms to detect cancer specific SNVs which will aid researchers and clinicians in study design and implementation of WGS for the identification of somatic mutations in other cancers.

The BARD1 BRCT domain contributes to p53 binding, cytoplasmic and mitochondrial localization, and apoptotic function.

BARD1 is a breast cancer tumor suppressor with multiple domains and functions. BARD1 comprises a tandem BRCT domain at the C-terminus, and this sequence has been reported to target BARD1 to distinct subcellular locations such as nuclear DNA breakage sites and the centrosome through binding to regulatory proteins such as HP1 and OLA1, respectively. We now identify the BRCT domain as a binding site for p53. We first confirmed previous reports that endogenous BARD1 binds to p53 by immunoprecipitation assay, and further show that BARD1/p53 complexes locate at mitochondria suggesting a cellular location for p53 regulation of BARD1 apoptotic activity. We used a proximity ligation assay to map three distinct p53 binding sequences in human BARD1, ranging from weak (425-525) and modest (525-567) to strong (551-777 comprising BRCT domains). Deletion of the BRCT sequence caused major defects in the ability of BARD1 to (1) bind p53, (2) localize to the cytoplasm and mitochondria, and (3) induce Bax oligomerization and apoptosis. Our data suggest that BARD1 can move to mitochondria independent of p53, but subsequently associates with p53 to induce Bax clustering in part by decreasing mitochondrial Bcl-2 levels. We therefore identify a role for the BRCT domain in stimulating BARD1 nuclear export and mitochondrial localization, and in assembling mitochondrial BARD1/p53 complexes to regulate specific activities such as apoptotic function.

MicroRNA and mRNA expression profiling in metastatic melanoma reveal associations with BRAF mutation and patient prognosis.

The role of microRNAs (miRNAs) in melanoma is unclear. We examined global miRNA expression profiles in fresh-frozen metastatic melanomas in relation to clinical outcome and BRAF mutation, with validation in independent cohorts of tumours and sera. We integrated miRNA and mRNA information from the same samples and elucidated networks associated with outcome and mutation. Associations with prognosis were replicated for miR-150-5p, miR-142-3p and miR-142-5p. Co-analysis of miRNA and mRNA uncovered a network associated with poor prognosis (PP) that paradoxically favoured expression of miRNAs opposing tumorigenesis. These miRNAs are likely part of an autoregulatory response to oncogenic drivers, rather than drivers themselves. Robust association of miR-150-5p and the miR-142 duplex with good prognosis and earlier stage metastatic melanoma supports their potential as biomarkers. miRNAs overexpressed in association with PP in an autoregulatory fashion will not be suitable therapeutic targets.

Increased MAPK reactivation in early resistance to dabrafenib/trametinib combination therapy of BRAF-mutant metastatic melanoma.

One-third of BRAF-mutant metastatic melanoma patients treated with combined BRAF and MEK inhibition progress within 6 months. Treatment options for these patients remain limited. Here we analyse 20 BRAF(V600)-mutant melanoma metastases derived from 10 patients treated with the combination of dabrafenib and trametinib for resistance mechanisms and genetic correlates of response. Resistance mechanisms are identified in 9/11 progressing tumours and MAPK reactivation occurred in 9/10 tumours, commonly via BRAF amplification and mutations activating NRAS and MEK2. Our data confirming that MEK2(C125S), but not the synonymous MEK1(C121S) protein, confers resistance to combination therapy highlight the functional differences between these kinases and the preponderance of MEK2 mutations in combination therapy-resistant melanomas. Exome sequencing did not identify additional progression-specific resistance candidates. Nevertheless, most melanomas carried additional oncogenic mutations at baseline (for example, RAC1 and AKT3) that activate the MAPK and PI3K pathways and are thus predicted to diminish response to MAPK inhibitors.

Melanomas of unknown primary have a mutation profile consistent with cutaneous sun-exposed melanoma.

Melanoma of unknown primary (MUP) is an uncommon phenomenon whereby patients present with metastatic disease without an evident primary site. To determine their likely site of origin, we combined exome sequencing from 33 MUPs to assess the total rate of somatic mutations and degree of UV mutagenesis. An independent cohort of 91 archival MUPs was also screened for 46 hot spot mutations highly prevalent in melanoma including BRAF, NRAS, KIT, GNAQ, and GNA11. Results showed that the majority of MUPs exhibited high somatic mutation rates, high ratios of C>T/G>A transitions, and a high rate of BRAF (45 of 101, 45%) and NRAS (32 of 101, 32%) mutations, collectively indicating a mutation profile consistent with cutaneous sun-exposed melanomas. These data suggest that a significant proportion of MUPs arise from regressed or unrecognized primary cutaneous melanomas or arise de novo in lymph nodes from nevus cells that have migrated from the skin.

Effects of BRAF inhibitors on human melanoma tissue before treatment, early during treatment, and on progression.

Selective BRAF inhibitors (BRAFi) are a standard of care for the treatment of BRAF(V) (600) -mutant metastatic melanoma. We analyzed a unique set of serial triplicate human metastatic melanoma tumor biopsies to identify biomarkers of BRAFi response and resistance. Morphologic features and immunohistochemical biomarkers were analyzed in 37 metastatic melanoma biopsies at pretreatment (PRE), early during treatment (EDT), and on progression (PROG) from 15 patients treated with a BRAFi and correlated with response and outcome. At EDT, proliferative markers decreased regardless of response, whereas markers of cell death increased in responders. High expression of nuclear p27 at baseline was the strongest predictor of a poorer OS and predicted worse response. The results show that BRAFi are universally antiproliferative, regardless of clinical response, whereas markers of cell death increased only in responders. The addition of therapies targeting the cell cycle machinery may improve the response and duration of BRAFi, and investigation of the mechanisms of apoptosis may provide additional therapeutic targets.

Intratumoral molecular heterogeneity in a BRAF-mutant, BRAF inhibitor-resistant melanoma: a case illustrating the challenges for personalized medicine.

Targeted therapies are increasingly being used to treat a variety of cancers. Their efficacy depends upon the accurate detection and targeting of a specific mutation or aberration in the tumor. All cancers, such as melanoma, are molecularly heterogeneous, with drug-resistant subclones present before the treatment or emerging as a result of targeted therapies. Here, we show intralesional molecular heterogeneity in a progressing V600E BRAF-mutant melanoma metastasis from a patient treated for 7 months with the BRAF inhibitor vemurafenib. In the single metastasis, two distinct subclones were observed, both V600E BRAF-mutant and only one with an additional G13R NRAS mutation. Molecular heterogeneity even at the intralesional level shows that personalizing or adjusting therapies based on genotyping of a portion of a single lesion may not accurately depict the molecular profile or drivers of oncogenesis across the entire patient's melanoma.

Mutational analysis of cutaneous squamous cell carcinomas and verrucal keratosis in patients taking BRAF inhibitors.

B-RAF inhibitors (BRAFi) have been shown to improve rates of overall and progression-free survival in patients with stage IV metastatic melanoma positive for the BRAF V600E mutation. However, the main drawback is the development of verrucal keratosis (hyperkeratotic papules with verruca-like characteristics with benign histological findings) and cutaneous squamous cell carcinomas (cuSCC). We have found upstream mutations in RAS as well as PIK3CA in both verrucal keratosis and cuSCC. This suggests that verrucal keratosis is an early clinical presentation of cuSCC in patients on BRAFi.

A high-throughput panel for identifying clinically relevant mutation profiles in melanoma.

Success with molecular-based targeted drugs in the treatment of cancer has ignited extensive research efforts within the field of personalized therapeutics. However, successful application of such therapies is dependent on the presence or absence of mutations within the patient's tumor that can confer clinical efficacy or drug resistance. Building on these findings, we developed a high-throughput mutation panel for the identification of frequently occurring and clinically relevant mutations in melanoma. An extensive literature search and interrogation of the Catalogue of Somatic Mutations in Cancer database identified more than 1,000 melanoma mutations. Applying a filtering strategy to focus on mutations amenable to the development of targeted drugs, we initially screened 120 known mutations in 271 samples using the Sequenom MassARRAY system. A total of 252 mutations were detected in 17 genes, the highest frequency occurred in BRAF (n = 154, 57%), NRAS (n = 55, 20%), CDK4 (n = 8, 3%), PTK2B (n = 7, 2.5%), and ERBB4 (n = 5, 2%). Based on this initial discovery screen, a total of 46 assays interrogating 39 mutations in 20 genes were designed to develop a melanoma-specific panel. These assays were distributed in multiplexes over 8 wells using strict assay design parameters optimized for sensitive mutation detection. The final melanoma-specific mutation panel is a cost effective, sensitive, high-throughput approach for identifying mutations of clinical relevance to molecular-based therapeutics for the treatment of melanoma. When used in a clinical research setting, the panel may rapidly and accurately identify potentially effective treatment strategies using novel or existing molecularly targeted drugs.

Identification of TFG (TRK-fused gene) as a putative metastatic melanoma tumor suppressor gene.

High density SNP arrays can be used to identify DNA copy number changes in tumors such as homozygous deletions of tumor suppressor genes and focal amplifications of oncogenes. Illumina Human CNV370 Bead chip arrays were used to assess the genome for unbalanced chromosomal events occurring in 39 cell lines derived from stage III metastatic melanomas. A number of genes previously recognized to have an important role in the development and progression of melanoma were identified including homozygous deletions of CDKN2A (13 of 39 samples), CDKN2B (10 of 39), PTEN (3 of 39), PTPRD (3 of 39), TP53 (1 of 39), and amplifications of CCND1 (2 of 39), MITF (2 of 39), MDM2 (1 of 39), and NRAS (1 of 39). In addition, a number of focal homozygous deletions potentially targeting novel melanoma tumor suppressor genes were identified. Because of their likely functional significance for melanoma progression, FAS, CH25H, BMPR1A, ACTA2, and TFG were investigated in a larger cohort of melanomas through sequencing. Nonsynonymous mutations were identified in BMPR1A (1 of 43), ACTA2 (3 of 43), and TFG (5 of 103). A number of potentially important mutation events occurred in TFG including the identification of a mini mutation "hotspot" at amino acid residue 380 (P380S and P380L) and the presence of multiple mutations in two melanomas. Mutations in TFG may have important clinical relevance for current therapeutic strategies to treat metastatic melanoma.

BARD1 translocation to mitochondria correlates with Bax oligomerization, loss of mitochondrial membrane potential, and apoptosis.

The breast cancer regulatory protein-1 (BRCA1)-associated RING domain 1 (BARD1) gene is mutated in a subset of breast/ovarian cancers. BARD1 functions as a heterodimer with BRCA1 in nuclear DNA repair. BARD1 also has a BRCA1-independent apoptotic activity. Here we investigated the link between cytoplasmic localization and apoptotic function of BARD1. We used immunofluorescence microscopy and deconvolution analysis to resolve BARD1 cytoplasmic staining patterns and detected endogenous BARD1 at mitochondria. BARD1 was also detected in mitochondrial cell fractions by immunoblotting. The targeting of BARD1 to mitochondria was modestly stimulated by DNA damage and did not require BRCA1 as indicated by RNA interference and peptide-competition experiments. Transiently expressed yellow fluorescence protein-BARD1 localized to mitochondria, and the targeting sequences were mapped to both the N and C terminus of BARD1. Ectopic yellow fluorescence protein-BARD1 induced apoptosis and loss of mitochondrial membrane potential in MCF-7 breast tumor cells. BARD1 apoptotic function was associated with stimulation of Bax oligomerization at mitochondria. This distinguishes it from BRCA1, which is pro-apoptotic but did not induce Bax oligomerization. The cancer-associated BARD1 splice-variant DeltaRIN (lacks the BRCA1 binding domain and ankyrin repeats) was recruited to mitochondria but did not stimulate apoptosis or alter membrane permeability. We propose that BARD1 has two main sites of action in its cellular response to DNA damage, the nucleus, where it promotes cell survival through DNA repair, and the mitochondria, where BARD1 regulates apoptosis.

Protein trafficking in response to DNA damage.

Human cells are prone to a range of natural environmental stresses and administered agents that damage or modify DNA, resulting in a cellular response typified by either cell death, or a cell cycle arrest, to permit repair of the genomic damage. DNA damage often elicits movement of proteins from one subcellular location to another, and the redistribution of proteins involved in genomic maintenance into distinct nuclear DNA repair foci is well documented. In this review, we discuss the DNA damage-induced trafficking of proteins to and from other distinct subcellular organelles including the nucleolus, mitochondria, Golgi complex and centrosome. The extent of intracellular transport suggests a dynamic and possibly co-ordinated role for protein trafficking in the DNA damage response.

Nuclear targeting and cell cycle regulatory function of human BARD1.

The BARD1 gene is mutated in a subset of breast and ovarian cancers, implicating BARD1 as a potential tumor suppressor. BARD1 gains a ubiquitin E3 ligase activity when heterodimerized with BRCA1, but the only known BRCA1-independent BARD1 function is a p53-dependent proapoptotic activity stimulated by nuclear export to the cytoplasm. We described previously the nuclear-cytoplasmic shuttling of BARD1, and in this study, we identify the transport sequences that target BARD1 to the nucleus and show that they are essential for BARD1 regulation of the cell cycle. We used deletion mapping and mutagenesis to define two active nuclear localization signals (NLSs) present in human BARD1 that are not conserved in rodent BARD1. Site-directed mutagenesis of the primary bipartite NLS abolished BARD1 nuclear import and caused its cytoplasmic accumulation. Using flow cytometry and 5-bromo-2-deoxyuridine incorporation assays, we discovered that transiently expressed BARD1 can elicit a p53-independent cell cycle arrest in G1 phase, and that this was abrogated by mutation of the BARD1 NLS but not by mutation of the nuclear export signal. Thus, BARD1 regulation of the cell cycle is a nuclear event and may be linked to its induced expression during mitosis and its possible involvement in the DNA damage checkpoint.

Issues arising from the prenatal diagnosis of some rare trisomy mosaics--the importance of cryptic fetal mosaicism.

OBJECTIVES: To add to the knowledge of fetal mosaicism, confined placental mosaicism (CPM), and uniparental disomy (UPD), in rare trisomies detected at prenatal diagnosis. METHODS: The origin of rare trisomy mosaics, mostly (8/11) seen in amniocytes, was examined in 11 cases by follow-up karyotyping and the study of microsatellite inheritance. RESULTS: Of the rare trisomies presented, three were mosaic trisomy 16 (two of which were CPM), and the remainder comprised single cases of mosaic trisomies of 8, 9, 10, 11, 12, 14, 5 and 15--the last two being CPM. Cases varied in parental derivation and meiotic versus post-zygotic origin but no case involved UPD. There was evidence for cryptic fetal mosaicism in three cases (5, 7, 11)--involving chromosomes 11, 14 and 16. CONCLUSIONS: These cases contribute further data to phenotypes associated with rare trisomies and the relative influences on the phenotype of CPM, UPD and fetal mosaicism. From sparse published data, we estimate that approximately 10% of apparent CPM cases for a rare trisomy (i.e. aneuploid CVS, normal amniocytes) may actually be cryptic fetal mosaics undetected in cultured amniocytes. In many cases, this cryptic mosaicism may be of limited clinical significance, but in others, the associated phenotypic effects may be obvious. There is no general approach to resolve this issue; the finding of even a few similar aneuploid cells in different amniocyte culture vessels may be clinically significant. It may be useful to study such an amniocyte culture with FISH with the relevant centromeric probe. Careful follow-up is recommended, particularly for infants where apparent correction of autosomal trisomy has occurred.