Cell death biomarker

Abstract

The invention relates to cell death of cancer cells, and in particular to biomarkers that may be used to identify cancer cells that are sensitive to death receptor ligand (DRL)-induced cell death. The invention also extends to prognostic methods and kits for identifying cancer cells that are sensitive to DRL-induced cell death. The invention further extends to novel compositions and therapeutic methods using such compositions for treating cancer.

Claims

1. A method of treating an individual suffering from cancer, wherein the cancer is selected from mesothelioma, renal cell carcinoma or cholangiocarcinoma, and wherein the cancer is sensitive to death receptor ligand (DRL)-induced cell death, the method comprising: (i) detecting for the presence of a mutant BAP1 gene or mutant BAP1 protein, or for a reduced level of expression of a wild-type BAP1 gene or a lower wild-type BAP1 protein concentration compared to the level of expression or protein concentration in a reference cell that is a BAP1 wild-type cell that is resistant to DRL-induced cell death, or for reduced or non-binding of an ASXL protein to a wild-type BAP1 protein compared to the level of binding in a reference cell that is a BAP1 wild-type cell, which is resistant to DRL-induced cell death; and (ii) administering, or having administered, a therapeutically effective amount of a death receptor ligand to the individual.

2. The method according to claim 1, wherein the mutant BAP1 gene is a gene that encodes a non-functional or enzymatically inactive BAP1 protein, or a BAP1 protein that exhibits reduced binding to an ASXL protein compared to the level of binding in a reference cell, which is resistant to DRL-induced cell death.

3. The method according to claim 1, wherein the reduced level of expression is at least a 10%, 15%, 25%, 35%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% or a 100% reduction compared to the reference cell.

4. The method according to claim 1, wherein the nucleotide sequence of the mutant BAP1 gene is SEQ ID NO. 5 or a fragment or variant thereof.

5. A method according to claim 1, wherein the death receptor ligand is selected from a group consisting of: TRAIL, TNF alpha; FAS ligand (FASL); recombinant TRAIL comprising dulanermin; antibody to a death receptor; mapatumuab; drozitumumab; conatumumab; lexatumumab; tigatuzumab; Medi-3038; Medi-3039; and LBY-135; or a combination thereof.

Description

(1) For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying diagrammatic drawings, in which:—

(2) FIG. 1A shows various types of mutations in different malignant pleural mesothelioma (MPM) cell lines. FIG. 1B is an analysis of TCGA exomes for enrichment of BAP1 loss of function mutations. FIG. 1s is a schematic of BAP1 gene exons with Mutations annotated from 5180 TCGA exomes.

(3) FIG. 2A is a volcano plot showing drug-genome interactions in MPM. The volcano plot displays the mean delta AUC by genotype for 92 library compounds. The Y-axis denotes adjusted p-Value, X-axis denotes effect size. Size of circle indicates number of mutant lines in cohort tested. FIG. 2B is a histogram which shows the results of a 6-day viability assay for multiple mesothelioma cell lines (n=19). rTRAIL (40 ng/ml). Met 5a is a mesothelial normal control line. FIG. 2C is a scatter plot showing AUC following 6 day viability assay in response to 40 ng/ml rTRAIL (normalised to DMSO treated control) in cell lines discretized by BAP1 mutation status. BAP1 mutation status significantly correlates with response to rTRAIL. Two-tailed t-test, p=0.015.

(4) FIG. 3A shows the results of a viability assay in which various MPM cells lines are treated with rTRAIL. MPM cell lines were treated with a dose range from 0.5 ng/ml to 100 ng/ml and cell viability was measured using Syto-60 assay. Based on their cell viability, the cell lines were classified into resistant (red), partially sensitive (orange) and sensitive (green). Cell lines were western blotted to probe the expression of BAP1 protein expression. FIG. 3B shows the results of a viability assay in which various MPM cells lines are treated with rTRAIL. Three BAP1 wild-type cell lines (MPP-89, H2869 & H2818) and four BAP1 mutant cell lines (H2722, H2461, H28 and H2731) were treated with TRAIL (0-1000 ng/ml) for 24 hours and cell death quantified using an Annexin V/DAPI cell death assay.

(5) FIG. 4 is a graph (and Western blot), which shows that knocking down BAP1 in a BAP1 wild-type mesothelioma cell line H2818 confers increased cell death response to rTRAIL.

(6) FIG. 5A is an immunoblot for BAP1 protein in BAP1 null mesothelioma lines following transfection with empty vector and BAP1 expression vector. FIG. 5B is a dose-response curve for an Annexin V/DAPI cell death assay performed with rTRAIL on the BAP1 null H226 parental line, a BAP1 wt overexpressing stable line, and BAP1 c91 hydrolase inactive stable cell line. FIG. 5C is a graph showing the effect of rTRAIL on cell viability of an untransduced H226 cell line, a BAP1 (transduced) cell line, and a H226 cell line with the NLS deleted.

(7) FIG. 6A shows that cell death is significantly dysregulated with the loss of the BAP1 catalytic ubiquitin hydrolation domain. Comparing GEX profile of C91 variant (catalytically inactive) BAP1 transduced H226 with H226 BAP1 wild-type transduced H226 Kegg pathway analysis on significantly dysregulated genes analysis with adj p <0.05 and FDR <20%. FIG. 6B is a graph showing RMA normalised gene centered mRNA expression of IAP genes BIRC2 and BIRC3 in c91 mutant BAP1 vs wild-type BAP1 expressing H226 cell line. The Western blot shows dysregulation of IAP family proteins in H226 cell line expressing catalytically inactivated C91 mutant BAP1.

(8) FIG. 7A is a volcano plot showing drug-genome interactions when rTRAIL was used as an anchor drug in combination with the library of 94 single agent compounds. Synergy was described using delta AUC metric. FIGS. 7B & 7C are graphs showing the effect of rTRAIL on cell viability of MPM cells in the presence of LCL161. TRAIL resistant MPM cells were treated with either 0-1000 ng/ml of TRAIL alone or a combination of 5 μM LCL161 and 0-1000 ng/ml of TRAIL for 24 hours and cell death was quantified by Annexin V/DAPI assay. FIG. 7D is a graph showing the effect of rTRAIL on cell viability of cells in the presence of the IAP inhibitor, LCL161. BAP1-transduced or BAP1 C91A-transduced H226 cells were treated either with 0-1000 ng/ml of TRAIL alone or combination of 5 μM LCL161 and 0-1000 ng/ml of TRAIL for 24 hours and cell death was quantified by Annexin V/DAPI assay.

(9) FIG. 8A is a graph showing the effect of rTRAIL on the viability of various cancer cell lines. Bladder (RT4) and Breast (HCC1187) cancer cell lines with nonsense mutations in BAP1 show sensitivity to rTRAIL while renal cell cancer cell lines (769P & RCC10RGB) with missense mutation and wild-type renal (BB65RCC) and bladder cancer (SW1710) cell lines are resistant to TRAIL. FIG. 8B shows that knockdown of BAP1 in Breast cancer cell line MDA MB-231 increases sensitivity to rTRAIL.

(10) FIG. 9A is a protocol of an in vivo experiment. FIG. 9B is a box plot showing the weight of tumours extracted from mice injected with mutant or wild-type BAP1-expressing cells after treatment with rTRAIL. Tumour weights of mutated BAP1 xenografts are significantly smaller than wild-type BAP1 xenografts after TRAIL treatment. FIG. 9C is a graph showing that TRAIL treatment reduces the tumour burden (measured by bioluminescence) of mutated BAP1 xenografts when compared to TRAIL treated wild-type BAP1 xenografts or untreated BAP1 mutated and BAP1 wild-type xenografts.

(11) FIG. 10 is a graph, which shows that BAP1 the that catalytic domain of BAP1 also regulates the sensitivity of H226 cells to the cell death-inducing ligands, TRAIL, FASL and TNFα. Untransduced BAP1-negative H226 cells, BAP1-expressing and catalytically dead BAP1-expressing H226 cells were treated with 100 ng/ml of FASL, TRAIL and TNF-alpha for 24 hours and cell death was quantified by Annexin V/DAPI assay. *p<0.05 indicating significant difference between untransduced H226 cells H226 BAP1 expressing cells. NS no significant difference between untransduced H226 cells and BAP1 C91A transduced cells. #p<0.05 indicating significant difference between untransduced H226 cells and BAP1 C91A transduced cells.

(12) FIG. 11 is a graph, which shows that BAP1 that catalytic domain of BAP1 also regulates the sensitivity of H226 cells to the cell death inducing ligands, TRAIL, FASL and TNFα. Untransduced BAP1-negative H226 cells, BAP1-expressing, ASXL binding site deleted BAP1-expressing H226 cells and catalytically dead BAP1-expressing H226 cells were treated with 100 ng/ml of FASL, TRAIL and TNF-alpha for 24 hours and cell death was quantified by Annexin V/DAPI assay.

EXAMPLES

(13) The inventors have discovered that mutation of the BAP1 tumour suppressor gene confers sensitivity to therapeutic modulation of the apoptotic pathway in human cancers. They have explored and validated this association in malignant pleural mesothelioma, bladder carcinoma and breast carcinoma and have evidence that it can be extended to between 1-36% human cancers, including renal cell carcinoma, and cervical cancer and uveal melanoma. Although the data described herein focuses on rTRAIL, a recombinant protein that activates the TRAIL pathway by binding to TRAIL receptor 1 (TRAIL-R1, also known as death receptor 4; DR 4) and TRAIL-R2 (also known as DR 5), BAP1 is also found to modulate other pro-apoptotic pathways, such as the FAS ligand pathway or the TNF pathway or intrinsic apoptotic pathway (see Example 6).

(14) Materials and Methods

(15) Whole Exome Sequencing

(16) DNA was extracted using the column extraction technique as per manufacturer's instructions (QIAGEN). Genomic libraries were prepared using the Illumina paired end sample prep kit following the manufacturer's instructions. Exome enrichment was performed using the Agilent SureSelect Human All Exon 50 Mb kit following the manufacturer's recommended protocol. Each exome was sequenced using the 75-bp paired end protocol on an Illumina HiSeq 2000 DNA Analyser to produce approximately 5-10 Gb of sequence per exome. Sequencing reads were aligned to the human genome (NCBI build GrCh 37) using the Burrows-Wheeler aligner (BWA) algorithm with default settings (17). Unmapped reads and PCR duplicates were excluded from the analysis. Average coverage of the cell line exomes at lox or higher was 80%.

(17) Copy Number Annotation

(18) DNA was extracted as above. DNA was outsourced to AROS for SNP 6.0 array. Copy number annotation was derived from the PICNIC algorithm [18].

(19) Variant Detection

(20) The CaVEMan algorithm was used to call single nucleotide substitutions [19]. The algorithm uses a naïve Bayesian classifier to estimate the posterior probability of each possible genotype (wild-type, germline or somatic mutation) at each base. To call insertions and deletions, split read mapping was implemented as a modification of the Pindel Algorithm [19]. Pindel searches for one read anchored on the genome with the other read mapped in two portions, spanning a putative insertion/deletion. For both algorithms, an identical putative normal from the CGP panel of tumours was nominated that has been used in all cell lines studied without available matched normal tissue. Significant post processing filtering against various panels of normal was subsequently undertaken to eliminate as many germline single nucleotide polymorphisms as possible. These include the 1000 genomes database, DB SNP, and an internal panel of CGP normal. Following these steps missense variants were annotated using the FATHM algorithm (Cancer Genome Project) as to potential functional consequence of the variant.

(21) Combination (Genome-Drug) Therapeutic Screen Approaches

(22) Manual “single dose” combination screening was undertaken using 96 well formats. Cells were plated on day 1 in previously optimized seeding densities in 180 μl if media. On day 2 20 μl of a 10× concentration of media from a stock of drugs was added. Cells were then allowed to grow for 72 hrs or 6 days and fixed at the end of the assay. Drug wells were compared to DMSO treated control wells.

(23) Single agent high throughput 5 point viability screening was undertaken in 384 well formats [18] using robotic liquid handling with fixing with 4% paraformaldehyde and staining for viability with Syto60 nucleic acid dye (Invitrogen) (see below). Single agent dose response curves were derived for each library of 85-95 drugs according to the experiment used, and log IC50 or area under the curve (AUC) metrics were derived for each library compound in each cell line according to a previously derived formula [18]. Using this data various 2-drug synergy was measured with a Delta AUC metric.

(24) A binary event matrix was compiled for the cell lines in the mesothelioma screen by aggregating copy number and exome data and this was used as input classifiers for genomic correlation. Data from this therapeutic screen was then analysed using a Multivariate Analysis of Variance (MANOVA) [18] to annotate the sensitizing effect of genotype on dose response. The results are presented as a volcano plot demonstrating significance of the interaction (above a Benjamin Hochberg false discovery threshold) and magnitude of effect size.

(25) Analyses of TCGA Data

(26) Frequency of BAP1 truncating mutations in various cancer types is based upon data generated by The Cancer Genome Atlas (TCGA) Research Network.

(27) Cell Culture

(28) 293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), nonessential amino acids, 50 U/mL penicillin, 50 μg/mL streptomycin, and 1% sodium pyruvate. Human mesothelioma cell lines were cultured in RPMI-1640 medium supplemented with 10% FBS, 1% penicillin-streptomycin and 1% sodium pyruvate (H2369, H2373, H2461, H2591, H2595, H2722, H2731, H2795, H2803, H2804, H2869, H290, H513, IST-MES1, MPP-89, MSTO-211H, NCI-H2052, NCI-H2452, NCI-H226, NCI-H28) or Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12) supplemented with 10% FBS, nonessential amino acids, 50 U/mL penicillin, 50 μg/mL streptomycin, and 1% sodium pyruvate (H2818, H2810). Cells were maintained at 37° C. at 5% CO.sub.2.

(29) Western Blotting

(30) Cell monolayers were washed in phosphate buffered saline (PBS) and lysed in radioimmunoprecipitation assay (RIPA) buffer (Sigma-Aldrich) and protease inhibitors (Complete-mini; Roche) on ice. Lysates were centrifuged at 14000 rpm for ten minutes and the supernatant aspirated. Protein concentration was calculated from a standard curve of bovine serum albumin using the BCA assay (Calbiotech) according to the manufacturer's instructions. Lysates were prepared to the appropriate concentration and 4× Laemelli buffer and 10× reducing agent added prior to the sample being heated at 70° C. for ten minutes. Lysates were subjected to SDS-PAGE on pre-cast 4-12% Bis-Tris gels (Invitrogen) at 200V for 1 hr. Protein was transferred onto a nitrocellulose membrane using an iBlot gel transfer device (Invitrogen) as per manufacturer's instructions. Membranes were blocked in 5% milk in tris-buffered saline with Tween 20 (TBS-T) before the addition of primary antibody (at 1:1000 in TBS-T unless otherwise stated) overnight at 4° C. The following day the membrane was washed three times in TBS-T and secondary antibody added (at 1:2500 in TBS-T) Antibodies used include BAP1 (C-4; Santa Cruz sc-28383), alpha tubulin (Cell Signalling #2125), c-IAP1 (Cell signaling #7065), cIAP2 (Cell signaling, #3130), Livin (Cell Signalling #5471), Survivin (Cell signaling, #2803), Alexa Fluor® 488 (Invitrogen A-21202). Immunoblots were imaged using an ImageQuant™ LAS 4000 biomolecular imager.

(31) Cell Viability Assays

(32) Adherent cell lines were seeded 24 hours before drugging. Cells were trypsinised and counted before seeding at the optimal density for the size of well (96 or 384) and duration of assay. 72 hr after drug treatment cells were fixed with 4% paraformaldehyde for 30 minutes. Following two washes of dH.sub.2O 100 μl Syto60 nucleic acid stain (Invitrogen) was added at a final concentration of 1 μM and plates fixed for 1 hr at room temperature. Quantification of fluorescent signal was achieved using excitation/emission wavelength of 630/695 nM.

(33) Cell Death Assays

(34) Adherent mesothelioma cell lines were plated in a 96 well plate at approximately 10000 cells per well. Cells were plated and given 1 day to adhere at which time drug was added. After 48 hrs media, including floating cells, was collected from each well. The remaining adherent cells were washed with PBS and mobilised with 0.05% trypsin in EDTA. All cells were collected into tubes containing the previously removed media and pelleted by centrifugation (300 g, 5 minutes). Cells were then re-suspended in 1× Annexin V binding buffer with 10 μl/1 ml concentration of Annexin V-647 antibody (Invitrogen) and incubated for 15 minutes at room temperature. DAPI (41 g/ml) was then added to each sample before flow cytometry analysis as below. Annexin V−/DAPI− cells were judged to be viable, AnnexinV+/DAPI− cells were considered to be undergoing apoptosis (early apoptotic phase), and Annexin V+/DAPI+ cells were considered late apoptotic or necrotic, and recorded as dead.

(35) Flow Cytometry Analyses

(36) Cells were washed with phosphate buffered saline and fixed by incubation in 4% paraformaldehyde for 20 minutes at room temperature. For intracellular BAP1 staining, fixed cells were permeabilised in 0.1% triton X-100 in PBS for 20 minutes on ice, washed twice with PBS, incubated with primary antibody (C-4; Santa Cruz sc-28383, 1:100) for 20 minutes on ice, washed twice again and incubated with a fluorescent secondary antibody (Alexa Fluor® 488, 1:200 (Invitrogen A-21202)). Cells were washed twice with PBS and suspended in PBS for flow cytometry analysis. Cells analysed as part of the cell death assays were prepared as above.

(37) Flow cytometry analysis was conducted on an LSRFortessa cell analyser (BD Biosciences) and data analysed with FlowJo software.

(38) mRNA Microarray

(39) The mRNA from catalytically inactive BAP1 expressing H226 cells (H226 C91A) and WT BAP1 expressing H226 cells (H226 BAP1) was extracted and run on an Illumina HT12 array.

(40) Pathway Analysis

(41) The significantly differentially expressed genes identified from the mRNA microarray were analysed using KEGG pathway analysis.

(42) Plasmids

(43) The cDNA full-length clone of human BAP1 was obtained in a pCMV6-AC backbone (Origene, SC117256), which was cloned into a PCCL.CMV lentiviral backbone for all further experiments. BAP1 mutant constructs were generated using site directed mutagenesis kits (NEB) and confirmed by full length DNA sequencing. Short hairpin RNAs were obtained through UCL RNAi library in a GIPZ shRNAmir lentiviral vector (Dharmacon V2LHS41473). The sequence (SEQ ID NO. 44) for the short hairpin is as follows:

(44) TABLE-US-00044 [SEQ ID NO. 44] TAAAGGTGCAGATGAACTC
Lentivirus Production and Concentration

(45) Lentiviruses were generated by transfection of 293T cells with the lentivirus vector plasmids together with the packaging plasmid pCMVdR8.2 and envelope plasmid pMDG.2 using jetPEI (Polyplus Transfection) as the transfection reagent. The 293T cells were incubated at 37° C. and the medium containing the lentiviruses harvested at 24 and 48 hrs. The lentivirus was concentrated by ultracentrifugation at 18000 rpm for 2 hours at 4° C. (SW28 rotor, Optima LE80K Ultracentrifuge, Beckman) and stored at −80° C. before use.

(46) Lentivirus titration was performed by transducing 293T cells with serial dilutions of virus in the presence of 4 μg/ml polybrene. After 4 days cells were analyzed for the percentage of BAP1 positive cells using flow cytometry. Viral was calculated as follows:
Titre (transduction units (TU)/ml)=Proportion of BAP1 positive cells×number of seeded cells/volume of virus (ml)

(47) MPM cells were then transduced with a range of multiplicity of infections (MOIs) in the presence of 4 μg/ml polybrene and transduction efficacy assessed by flow cytometry analysis. The optimal population (lowest MOI at which >90% transduction achieved) was selected for further experiments.

(48) shRNA Experiments

(49) Lentivirus encoding shRNA targeting BAP1 was generated as per the lentivirus production protocol above. MPM cells (H2818) were transduced and treated with puromycin 200 μg/mL until a pure population was achieved. Immunoblotting was performed to assess efficacy of the shRNA knockdown.

(50) Animals

(51) All animal studies were approved by the University College London Biological Services Ethical Review Committee and licensed under the UK Home Office regulations and the Evidence for the Operation of Animals (Scientific Procedures) Act 1986 (Home Office, London, UK). Mice were purchased from Charles River, kept in individually ventilated cages under specific pathogen-free conditions and had access to sterile-irradiated food and autoclaved water ad libitum.

(52) Xenograft Mouse Models

(53) Groups of 8 week old NOD.CB17-Prkdcscid/NcrCrl (NOD SCID) mice (Charles River) were injected on each flank with 1 million cells of luciferase transduced mesothelioma cell lines (H226 BAP and H226 C91A) in a 1:1 mixture of matrigel and media. When tumours were established, as assessed by bioluminescent imaging (IVIS), at 14 days following injection of tumour cells, treatment was began with either vehicle or isoleucine zipper TRAIL (izTRAIL) [20]. Either vehicle or izTRAIL were given intraperitoneally once daily at a dose of 600 mcg for the duration of the experiment. Tumour size was assessed at days 0, 13, 19, 26 and 41 using bioluminescent imaging (IVIS). Mice were culled at 42 days and tumours removed and weighed. Six mice per group were treated. Researchers were not blinded in these experiments.

(54) Statistical Analyses

(55) Statistical analysis was performed using GraphPad Prism V. 4 (GraphPad Software). In vivo experiments with multiple groups were analysed using repeated measures ANOVA, and single-group data were assessed using Student t test. All in vitro experiments were performed in triplicate unless specified otherwise.

Example 1—TRAIL Targets BAP1 Mutant Mesothelioma Cells

(56) The inventors carried out a combinatorial chemical screen in 15 mesothelioma cell lines (together with the Met 5a mesothelial normal control line) using 94 small molecule inhibitors and chemotherapy agents (see Table 1) either alone or in combination with the ligand tumour necrosis factor (TNF)-related cell death inducing ligand (TRAIL). To detect examples of extreme drug sensitivity, the inventors analysed for statistical associations between response and the mutational status of these cell lines based on a set of 8 genes recently identified as being candidate cancer genes in mesothelioma (see FIG. 1A). Of note BAP1 mutations are well recognised across cancer types (see FIGS. 1B and C). The largest effect of a mutation on drug response was that of mesothelioma cells harbouring a mutation in the deubiquitinase BAP1 and treated with TRAIL (see FIG. 2A). There was no significant effect on cell viability observed in the control normal mesothelial cell line MET-5A included in the screen (see FIG. 2B). BAP1 mutant cells were significantly more sensitive to TRAIL than their wild-type counterparts (see FIGS. 2B & 2C). Furthermore, the BAP1 mutations detected in these cell lines would be predicted to be truncating (see Table 2). The inventors confirmed by immunoblot that BAP1 mutations were usually associated with loss of protein expression and the mutant cell lines are generally sensitive to TRAIL (see FIG. 3A).

(57) TABLE-US-00045 TABLE 1 Compounds used in combinatorial chemical screen with 15 mesothelioma cell lines together with the Met5a mesothelial normal control line. min max Targeted conc conc compound target process/pathway (uM) (uM) AICAR AMPK agonist metabolism 7.81 2000 Camptothecin DNA topoisomerase I DNA replication 0.0004 0.1 Vinblastine Microtubules cytoskeleton 0.0004 0.1 Cisplatin DNA crosslinking DNA replication 0.0234 6 Docetaxel Microtubules cytoskeleton 0.0000 0.0125 Gefitinib EGFR EGFR signalling 0.0020 0.5 ABT-263 Bcl-2, Bcl-xL, and Bcl-w apoptosis regulation 0.0078 2 Vorinostat HDAC inhibitor Class I, IIa, chromain histone 0.0391 10 IIb, IV acetylation Nilotinib Bcr-Abl ABL signalling 0.0078 2 AZD-2281 PARP1/2 Genome integrity 0.0195 5 Bosutinib SRC, ABL, TEC ABL signalling 0.0078 2 Lenalidomide TNF alpha other 0.0195 5 Axitinib PDGFR, KIT, VEGFR RTK signalling 0.0078 2 AZD7762 Chk 1/2 Genome integrity 0.0078 2 GW 441756 Trk A RTK signalling 0.0078 2 CEP-701 FLT3, JAK2, NTRK1, RET RTK signalling 0.0078 2 SB 216763 GSKa/b WNT signalling 0.0391 10 17-AAG Hsp90 other 0.0039 1 AMG-706 VEGFR, RET, c-KIT, PDGFR RTK signalling 0.0078 2 KU-55933 ATM Genome integrity 0.0391 10 BIBW2992 EGFR, HER2 EGFR signalling 0.0020 0.5 GDC-0449 SMO other 0.0391 10 PLX4720 RAF ERK MAPK 0.0391 10 signalling BX-795 TBK1, PDK1, IKK, AURKB/C other 0.0195 5 NU-7441 DNAPK Genome integrity 0.0078 2 SL 0101-1 RSK, AURKB, PIM3 ERK MAPK 0.0391 10 signalling BI-D1870 RSK1/2/3/5, PLK1, AURKB cell cycle 0.0195 5 BIRB 0796 p38, JNK2 JNK and p38 0.0391 10 signalling JNK Inhibitor VIII JNK JNK and p38 0.0391 10 signalling 681640 Wee1, Chk1 cell cycle 0.0078 2 Nutlin-3a p53-MDM2 interaction p53 pathway 0.0313 8 mirin MRE11-Rad50-Nbs1 complex cell cycle 0.3906 100 PD-173074 FGFR1, FGFR3 RTK signalling 0.0078 2 ZM-447439 Aurora B mitosis 0.0156 4 RO-3306 Cdk1 cell cycle 0.0195 5 MK-2206 AKT1/2 PI3K signalling 0.0156 4 PD-0332991 Cdk 4/6 cell cycle 0.0156 4 PF477736 Chk 1 (Chk2) cell cycle 0.0039 1 GW843682X (AN-13) Plk1 mitosis 0.0195 5 NVP-BEZ235 PI3K Class 1 and mTORC1/2 PI3K signalling 0.0010 0.25 GDC0941 PI3K (class 1) PI3K signalling 0.0156 4 AZD8055 mTORC1/2 TOR signalling 0.0078 2 PD-0325901 MEK 1/2 ERK MAPK 0.0010 0.25 signalling AZD6482 PI3K beta PI3K signalling 0.0195 5 Obatoclax Mesylate Bcl-2, Bxl-xl, Mcl-1 apoptosis regulation 0.0391 10 EHT 1864 Rac GTPases cytoskeleton 0.0391 10 BMS-708163 gamma-secretase complex other 0.0195 5 5-Fluorouracil antimetabolite mitosis 0.0781 20 Paclitaxel Beta subunit of tubulin cytoskeleton 0.0000 0.01 PF-02341066 MET, ALK RTK signalling 0.0039 1 Sorafenib PDGFR, KIT, VEGFR RTK signalling 0.0156 4 BI-2536 Plk1, 2, 3 mitosis 0.0020 0.5 BMS-536924 IGF-1R IGFR signalling 0.0156 4 GSK1904529A IGF-IR and IR IGFR signalling 0.0195 5 AKT inhibitor VIII AKT1/2/3 PI3K signalling 0.0195 5 PF-4708671 p70 S6KA TOR signalling 0.0391 10 JNJ-26854165 MDM2 p53 pathway 0.0391 10 LY317615 PKC beta other 0.0391 10 BMS-754807 IGF-1R/IR IGFR signalling 0.0391 10 TW 37 BCL-2, BCL-XL apoptosis regulation 0.0195 5 Embelin XIAP apoptosis regulation 0.0391 10 Erlotinib EGFR EGFR signalling 0.0078 2 AZ628 BRAF ERK MAPK 0.0078 2 signalling AG-014699 PARP1/2 Genome integrity 0.0195 5 Gemcitibine nucleoside analog DNA replication 0.0391 10 GSK269962A ROCK1/2 cytoskeleton 0.0195 5 SB-505124 TGFbetaR-I (ALK5) other 0.0391 10 Tamoxifen ER other 0.0195 5 Fulvestrant ER other 0.0039 1 Anastrozole ER other 0.0391 10 JQ1 BRD2, BRD3, BRD4 chromatin other 0.0039 1 YK 4-279 RNA helicase A other 0.0391 10 CHIR-99021 GSK3B WNT signalling 0.0391 10 (5Z)-7-Oxozeaenol TAK1 other 0.0391 10 FK866 NAMPT inhibitor metabolism 0.0039 1 BMS-345541 IKK-beta other 0.0391 10 AZ960 JAK2 other 0.0391 10 BMN-673 PARP Genome integrity 0.0391 10 XAV 939 Tankyrase (PARP5a) WNT signalling 0.0195 5 GSK1120212 MEK1, MEK2 ERK MAPK 0.0039 1 signalling GSK2118436 BRAF ERK MAPK 0.0391 10 signalling Temozolomide DNA akylating agent DNA replication 0.1172 30 Olaparib + DNA damage response Genome integrity 0.0391 10 Temozolomide AZD2281 PARP Genome integrity 0.0391 10 Bicalutamide Androgen receptor other 0.0391 10 PF-562271 FAK cytoskeleton 0.0391 10 PAC-1 Caspase 3 activator apoptosis regulation 0.0391 10 INCB-18424 JAK1, JAK2, TYK2 other 0.0391 10 OSI-906 IGFR-1 IGFR signalling 0.0098 2.5 Epirubicin DNA damage DNA replication 0.0391 10 Cyclophosphamide DNA akylating agent DNA replication 0.0391 10 Carboplatin DNA damage DNA replication 0.0391 10 Everolimus mTOR TOR signalling 0.0195 5 LCL161 SMAC mimetic apoptosis regulation 0.0391 10 rTRAIL Death receptor ligand apoptosis regulation 0.39 ng/ml 100 ng/ml DMSO CONTROL NA

(58) TABLE-US-00046 TABLE 2 BAP1 mutation status in selected cell lines. BAPI rTRAIL mRNA SAMPLE_NAME COSMIC_ID DESCRIPTION ZYGOSITY response expression H226 Deletion Homozygous Sensitive H2461 1290810 Deletion - heterozygous Sensitive 0.32 Frameshift H2722 1290812 HomDel homozygous Resistant −2.52 H2731 1240134 Essential Splice heterozygous Sensitive −0.41 H2795 1290813 Essential Splice heterozygous Sensitive* −0.45 H2804 1240136 Essential Splice heterozygous Sensitive 0.51 IST-MES1 907173 Essential Splice homozygous Unknown −0.33 NCI-H2452 908462 Substitution - homozygous Resistant −0.04 Missense NCI-H28 908470 Essential Splice heterozygous Sensitive 1.27 NCI-H226 905941 HomDel homozygous Sensitive** ND H2595 1240132 Wild-type heterozygous Unknown −2.05 H2369 1290808 Wild-type heterozygous Resistant −0.64 H2373 1290809 Wild-type heterozygous Resistant 1.65 H2591 1240131 Wild-type heterozygous Resistant 0.93 H2803 1240135 Wild-type heterozygous Resistant 0.62 H2810 1240137 Wild-type heterozygous Resistant 1.06 H2818 1290814 Wild-type heterozygous Resistant 0.41 H2869 1240138 Wild-type heterozygous Resistant −0.67 H290 1240139 Wild-type heterozygous Unknown −0.15 H513 1240141 Wild-type heterozygous Resistant 0.11 MPP-89 908150 Wild-type heterozygous Resistant 0.54 MSTO-211H 908152 Wild-type heterozygous Sensitive** −0.28 NCI-H2052 688058 Wild-type heterozygous Resistant 0.92

Example 2—Modulation of BAP1 Expression Determines TRAIL Sensitivity Through Activation of Cell Death

(59) TRAIL binds via two active transmembrane death receptors, DR 4 and DR 5, triggering a caspase cascade and subsequently cell death. The viability effect of TRAIL observed in BAP1 mutant cells was indeed associated with an increased fraction of cells stained with the apoptotic marker Annexin V (see FIG. 3B).

(60) The inventors therefore next examined whether modulation of BAP1 expression in mesothelioma cells resulted in changes in TRAIL sensitivity. The ablation of BAP1 protein with the use of a lentiviral shRNA in the BAP1 wild-type cell line H2818 promoted a shift towards increased sensitivity in the BAP1 null compared to the BAP1 competent parental line (see FIG. 4). The BAP1 null cell line NCI-H226, which possesses a homozygous deletion of BAP1, was transduced with a BAP1 expression vector to restore expression of wild-type full length BAP1 (see FIG. 5A) or the catalytically dead C91A mutant. Treatment of the null NCI-H226 cell line with a dose range of TRAIL resulted in increased cell death which was significantly reduced in the BAP1 expressing H226 cell line (see FIG. 5b). The C91A variant however phenocopied the response of the BAP1 null parental cell line indicating a functional Ubiquitin hydrolase catalytic domain is critical for sensitivity to TRAIL. The nuclear localization signal (NLS) also plays a key role in imparting TRAIL resistance as deletion of NLS results in significant reduction in BAP1 induced TRAIL resistance (see FIG. 5c).

Example 3—Loss of BAP1 Expression and Function Modulates Components of the Apoptotic Machinery

(61) The H226 mesothelioma cell line harbours a homozygous deletion of BAP1, resulting in complete loss of BAP1 expression. The inventors further examined the effect of this catalytically inactive BAP1 on differential mRNA gene expression as well as carrying out a signalling pathway impact analysis (SPIA), as previously described (PMID 18990722). Among those pathways significantly altered when comparing wild-type versus c91a mt BAP1, was that of cell death pathways (see FIG. 6A). In particular, there was decreased expression of members of the IAP family in H226 cells stably transduced with the catalytically dead C91A mutant (see FIG. 6B). The largest effects were seen in CIAP2, and this was confirmed by western blot (see FIG. 6C).

Example 4—Combination Drug Screen Demonstrates Synergy Between SMAC Mimetic LCL161 and rTRAIL in BAP1 Competent Cell Lines

(62) rTRAIL was used as an anchor drug in combination with the library of 94 single agent compounds described above. Synergy was described using delta AUC metric (ref Wessles et al) and this was correlated with the previously described genomic subgroups. The inventors have shown that drugs such as SMAC mimetic LCL161, DNA helicase inhibitor YK-4279 and the tyrosine kinase inhibitor sorafenib to increase the efficacy of DRL-induced apoptosis in otherwise resistant cells. One of the most synergistic findings of this screen was the association of sensitivity to the SMAC mimetic LCL161 and rTRAIL in BAP1 wild-type MPM (see FIG. 7 A). This was validated by treating TRAIL resistant wild-type BAP1 expressing cells with combination of LCL161 and TRAIL. Further validation was also performed in the H226 cell lines stably expressing wild-type BAP1 that was previously demonstrated to be resistant to TRAIL alone (see FIGS. 5 and 7B-D). The combination of the IAP inhibitor, LCL161, and TRAIL showed a synergistic increase in cell death in the both mutant and wild-type line indicating that DRL induced cell death in BAP1 mutant and wild-type cells can be enhanced by combining with other agents (see FIG. 7B-D). FIG. 7D in particular shows that both BAP1 mutant and wild-type cells undergo cell death in response to treatment with the combination of TRAIL and LCL161. Endogenous SMAC/Diablo is a specific natural inhibitor of IAP's (14). This data suggests that in the BAP1 competent state, BAP1 loss can be phenocopied by specifically mimicking this inhibitory effect on IAPs resulting in a net inactivation of IAP's and sensitivity to rTRAIL. This would lend further support to the idea that the BAP1/extrinsic apoptotic pathway perturbation seen is related to a specific dysregulation of net activity of IAP's.

Example 5—Extension of BAP1/TRAIL Effect to Other Tissues Harboring BAP1 Loss of Function Mutations

(63) The deubiquitinase BAP1 is frequently mutated in pleural mesothelioma (36%), uveal melanoma (47%) and intrahepatic cholangiocarcinomas (25%) as previously noted. To determine whether additional BAP1 mutant tumours occur that may also be amenable to this therapeutic approach, the inventors extended this analysis to a cohort of 5180 tumour samples in 20 cancer types using variant data from The Cancer Genome Atlas (TCGA). Truncating BAP1 mutations were also observed in a diverse range of cancer types, with frequencies of up to 6% (see FIGS. 1b, and c) Carbone, M. et al, Nature Reviews Cancer 13, 153-159 (March 2013). When the inventors extended their analysis to a panel of 1001 cancer cell lines that had previously been submitted for whole exome and copy number analysis, they identified 17 cell lines harbouring truncating mutations in BAP1. These included clear cell kidney cancer, bladder cancer and breast cancer cell lines. Treatment of these cell lines with TRAIL resulted in a marked viability effect compared to BAP1 wild-type cell lines from the same cancer type (see FIG. 8A). The inventors also inactivated BAP1 using a lentiviral shRNA in the breast cancer cell line, MDA-MB231, and observed an exaggerated apoptotic response to rTRAIL (see FIG. 8B). This suggests that TRAIL therapy may be efficacious in other forms of cancer (in addition to mesothelioma), and that inhibition of BAP1 or by targeting the pathway which BAP1 induces TRAIL resistance it is possible to sensitise cancer cells to TRAIL.

(64) The inventors demonstrated the efficacy of targeting BAP1 mutant cells with TRAIL in vivo by mice xenograft models. Mutant and wild-type BAP1 cells were injected subcutaneously into left and right flanks of mice. The mice were treated with either rTRAIL or vehicle (see FIG. 9A). The tumours were weighed at the end of the experiment and the weights of mutated BAP1 tumours in mice that received TRAIL was significantly less than wild-type BAP1 tumours with TRAIL treatment or mutant and wild-type tumours of mice with vehicle treatment (see FIG. 9B). The tumour burden was tracked in real time over a period of 4 weeks. The tumour burden of mutated BAP1 xenografts in mice that received TRAIL was significantly less than wild-type BAP1 xenografts with TRAIL treatment or mutant and wild-type xenografts of mice with vehicle treatment (see FIG. 9C).

Example 6—Role of ASXL Binding Site on of BAP1 Function

(65) The inventors have also demonstrated that a mutation in the ASXL protein binding site of the BAP1 gene impairs BAP1-induced TRAIL resistance (see FIG. 1i). BAP1 has been shown to form a complex with proteins ASXL1, ASXL2 or ASXL3. Mutation of the binding site for ASXL protein inhibits formation of BAP1-ASXL complexes. The BAP1-ASXL complex has been shown to deubiquitinate Histone 2A, and other substrates, and both BAP1 and ASXL1, ASXL2, or ASXL3 are required for this deubiquitination function. This complex is an important regulator of the Polycomb Respressor Complex and gene transcription. The inventors have shown that the BAP1 wild-type and ASXL3 mutant (truncating mutation) cell line H513 is TRAIL sensitive. Hence loss of function of ASXL1, ASXL2 or ASXL3 increases the sensitivity of cells to DRL induced cell death. Mutations of ASXL1, ASXL2 or ASXL3 also predict sensitivity to DLR and hence can be used as a biomarker for cell death independent of BAP1 mutational status.

Example 7—Extension of BAP1/TRAIL Effect to Other Extrinsic Death Pathways

(66) Although the data in this application focus on rTRAIL, a recombinant protein that activates the TRAIL pathway by binding to DR 4 and DR 5 receptors, the observed BAP1 mutation-sensitisation extends to other extrinsic apoptotic pathways including the FAS ligand pathway and the TNFalpha pathway (see FIG. 10).

SUMMARY

(67) The inventors have found that BAP1 is an important regulator of whether a cell will undergo cell death in response to the activation of a death receptor by a death receptor ligand, such as TRAIL, TNF alpha (TNFα) and FAS ligand (FASL). Specifically, non-functional or low expression of wild-type BAP1 causes cells to become sensitive to death receptor ligand-induced cell death. Consequently, it has been discovered that a mutant BAP1 gene or a mutant BAP1 protein, or a cancer cell with low expression of a wild-type BAP1 protein may be used as a biomarker of sensitivity to DRL-induced cell death.

(68) Thus, the invention also encompasses an advantageous: method of determining if an individual's cancer cell is sensitive to death receptor ligand (DRL)-induced cell death; kit for determining if an individual's cancer cell is sensitive to DRL-induced cell death; method of selectively inducing death receptor ligand induced cell death in an individual suffering from a cancer that is insensitive to death receptor ligand induced cell death; a method of sensitising to DRL-induced cell death, an individual suffering from a cancer that is insensitive to DRL-induced cell death; a composition comprising a BAP1 inhibitor and a death receptor ligand; and a method of treating, preventing or ameliorating an individual suffering from a cancer, which is insensitive to DRL-induced cell death.

REFERENCES

(69) 1. Lanphear B P, Buncher C R. Latent period for malignant mesothelioma of occupational origin. Journal of occupational medicine: official publication of the Industrial Medical Association. 1992; 34(7):718-21. Epub 1992/07/01. 2. Lemen R A, Dement J M, Wagoner J K. Epidemiology of asbestos-related diseases. Environmental health perspectives. 1980; 34:1-11. Epub 1980/02/01. 3. Algranti E, Saito C A, Carneiro A P, Moreira B, Mendonca E M, Bussacos M A. The next mesothelioma wave: Mortality trends and forecast to 2030 in Brazil. Cancer epidemiology. 2015. Epub 2015/09/01. 4. Treasure T, Waller D, Tan C, Entwisle J, O'Brien M, O'Byrne K, et al. The Mesothelioma and Radical surgery randomized controlled trial: the Mars feasibility study. Journal of thoracic oncology: official publication of the International Association for the Study of Lung Cancer. 2009; 4(10):1254-8. Epub 2009/08/08. 5. Vogelzang N J, Rusthoven J J, Symanowski J, Denham C, Kaukel E, Ruffle P, et al. Phase III study of pemetrexed in combination with cisplatin versus cisplatin alone in patients with malignant pleural mesothelioma. Journal of clinical oncology: official journal of the American Society of Clinical Oncology. 2003; 21(14):2636-44. Epub 2003/07/16. 6. Guo G, Chmielecki J, Goparaju C, Heguy A, Dolgalev I, Carbone M, et al. Whole-exome sequencing reveals frequent genetic alterations in BAP1, NF2, CDKN2A, and CUL1 in malignant pleural mesothelioma. Cancer research. 2015; 75(2):264-9. Epub 2014/12/10. 7. Bott M, Brevet M, Taylor B S, Shimizu S, Ito T, Wang L, et al. The nuclear deubiquitinase BAP1 is commonly inactivated by somatic mutations and 3p21.1 losses in malignant pleural mesothelioma. Nature genetics. 2011; 43(7):668-72. Epub 2011/06/07. 8. Martin M, Masshofer L, Temming P, Rahmann S, Metz C, Bornfeld N, et al. Exome sequencing identifies recurrent somatic mutations in EIF1AX and SF 3B1 in uveal melanoma with disomy 3. Nature genetics. 2013; 45(8):933-6. Epub 2013/06/26. 9. Gerlinger M, Horswell S, Larkin J, Rowan A J, Salm M P, Varela I, et al. Genomic architecture and evolution of clear cell renal cell carcinomas defined by multiregion sequencing. Nature genetics. 2014; 46(3):225-33. Epub 2014/02/04. 10. Chan-On W, Nairismagi M L, Ong C K, Lim W K, Dima S, Pairojkul C, et al. Exome sequencing identifies distinct mutational patterns in liver fluke-related and non-infection-related bile duct cancers. Nature genetics. 2013; 45(12):1474-8. Epub 2013/11/05. 11. Narasimhan S R, Yang L, Gerwin B I, Broaddus V C. Resistance of pleural mesothelioma cell lines to apoptosis: relation to expression of Bcl-2 and Bax. The American journal of physiology. 1998; 275(1 Pt 1):L165-71. Epub 1998/08/05. 12. Hinds M G, Norton R S, Vaux D L, Day C L. Solution structure of a baculoviral inhibitor of apoptosis (IAP) repeat. Nature structural biology. 1999; 6(7):648-51. Epub 1999/07/15. 13. de Almagro M C, Vucic D. The inhibitor of apoptosis (IAP) proteins are critical regulators of signaling pathways and targets for anti-cancer therapy. Experimental oncology. 2012; 34(3):200-11. Epub 2012/10/17. 14. Yu J, Wang P, Ming L, Wood M A, Zhang L. SMAC/Diablo mediates the proapoptotic function of PUMA by regulating PUMA-induced mitochondrial events. Oncogene. 2007; 26(29):4189-98. Epub 2007/01/24. 15. Behjati S et al. Recurrent PTPRB and PLCG1 mutations in angiosarcoma. Nat Genet, 2014 April; 46(4):376-9. doi: 10.1038/ng.2921. 16. Tarpey P S et al Frequent mutation of the major cartilage collagen gene COL2A1 in Chondrosarcoma. Nat Genet. 2013, August; 45(8): 923-926. 17. Li, H and Durbin, R Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics, 2009. 25(14): p 1754-60. 18. Garnett, M. J., et al., Systemic identification of genomic markers of drug sensitivity in cancer cells. Nature, 2012. 483 (7391): p. 473-81. 19. Behjati, S. et al., Mobile DNA in cancer. Extensive transduction of nonrepetitive DNA mediated by Li retrotransposition in cancer genome. Science, 2014. 345(6196): p. 1251343. 20. Ganten T, Koschny R, Sykora J, Schulze-Bergkamen H, Buchler P, Haas T, et al. Preclinical differentiation between apparently safe and potentially hepatotoxic applications of TRAIL either alone or in combination with chemotherapeutic drugs. Clin Cancer Res. 2006; 12:2640-6