A CDS2 inhibitor for use in the treatment of cancer characterised by functional suppression of CDS1

A CDS2 inhibitor is developed to target cancers with suppressed CDS1 function, exploiting the synthetically lethal interaction between CDS1 and CDS2, providing a promising therapeutic approach for improving treatment efficacy in cancers with limited response to existing therapies.

WO2024240955A1PCT designated stage expired Publication Date: 2024-11-28GENOME RES LTD +1
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
PCT/EP2024/064443
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-25
Filing Date
2024-05-24
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

Current cancer treatments, including immune checkpoint inhibition and targeted therapies, have limited efficacy for uveal melanoma and other cancers, highlighting the need for new therapeutic targets, particularly for cancers with functional suppression of CDS1, where existing therapies show modest improvements and are restricted to specific patient populations.

Method used

The development of a CDS2 inhibitor for use in treating cancers with functional suppression of CDS1, leveraging the synthetically lethal interaction between CDS1 and CDS2, where inhibition of CDS2 is selectively lethal to cancer cells with suppressed CDS1 function, allowing for targeted therapy.

Benefits of technology

The CDS2 inhibitor effectively targets cancer cells with suppressed CDS1 function, offering a potential therapeutic strategy for cancers like uveal melanoma, diffuse B cell lymphoma, and others, by inducing selective lethality in cancer cells while sparing normal cells, thus improving treatment outcomes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure EP2024064443_28112024_PF_FP_ABST
    Figure EP2024064443_28112024_PF_FP_ABST
Patent Text Reader

Abstract

Described herein is a method of treating or preventing cancer in a subject wherein the cancer has functional suppression of CDS1, the method comprising administering a therapeutically effective amount of a CDS2 inhibitor to the subject Also described are methods for predicting whether a subject having cancer, wherein the cancer has functional suppression of CDS1, is likely to respond to treatment with a CDS2 inhibitor; as well as methods of selecting, and identifying and treating a subject likely to respond to treatment with a CDS2 inhibitor. Also disclosed are methods of screening for a compound for use in treatment of cancer.
Need to check novelty before this filing date? Find Prior Art

Description

[0001] A CDS2 INHIBITOR FOR USE IN THE TREATMENT OF CANCER CHARACTERISED BY FUNCTIONAL SUPPRESSION OF CDS1

[0002] This application claims priority from GB2307864.5 filed on 25 May 2023, the contents and elements of which are herein incorporated by reference for all purposes.

[0003] Field of the Invention

[0004] The invention relates to treatments for cancer targeting Cytidine diphosphate diacylglycerol synthase 1 and 2 (CDS1 and CDS2).

[0005] Background

[0006] While the clinical benefit of immune checkpoint inhibition may have revolutionised the treatment of many cancers, including cutaneous melanoma, survival outcomes for patients remain exceptionally poor. Given its unique biology, the durable clinical efficacy reported in cutaneous melanoma from anti-cytotoxic T-lymphocyte antigen 4 (CTLA-4) and antiprogrammed death 1 receptor (PD-1) therapy is rarely observed in uveal melanoma [1 ,2], Tebantafusp, the first systemic therapy to recently receive regulatory approval, has shown modest improvements in survival benefit but is restricted to approximately only half of patients who carry the HLA-A*02:01 allele [3,4], Other systemic treatments including targeted therapies have limited, if any, tumour response [5-9, 25]. New therapeutic targets are urgently needed, and adopting a systematic approach is crucial to comprehensively identify novel tumour-intrinsic vulnerabilities.

[0007] CRISPR screening of cancer cell lines has enabled the perturbation of individual genes on a genome-wide scale, leading to the discovery of single gene essentialities. Synthetic lethal interactions have traditionally been identified by performing these large loss-of-function screens in cell lines that harbour defined genetic changes. More recently, synthetic lethal interactions have been interrogated through the development of combinatorial CRISPR technologies to simultaneously disrupt gene pairs. In cancer cells where the function of one gene is commonly lost, targeting the remaining gene within that pair could result in loss of cellular fitness or death selectively in cancer cells and not normal cells.

[0008] Of particular interest is targeting paralogue gene pairs, that are found in different genomic regions but share a common ancestral gene and hence functionality through involvement in common or parallel pathways. This functional overlap observed in paralogue gene pairs allow them to withstand stresses and loss of one gene, without ultimately affecting survival of the cell. Therefore, cells are able tolerate loss of one, but not both, of these genes. Indeed, genes that have identifiable paralogues are less likely to be essential to cell survival compared with genes that have no paralogues [26-30]. Therefore, targeting paralogue gene pairs is a valuable source of synthetic lethal interactions in many cancers.

[0009] Although there have been significant gains in the molecular understanding of many cancers, such as, for example, uveal melanoma, colon cancer, lung cancer and breast cancer, these have not translated into successful targeted therapies.

[0010] Summary of the invention

[0011] The present inventors have discovered that inhibition or genetic ablation of CDS2 is selectively lethal to those cancer cells in which CDS1 is functionally suppressed. The synthetically lethal interaction between CDS1 and CDS2 has important implications for the treatment of cancers, for example in cancers with functional suppression of CDS1 , such as cancers where CDS1 expression is low or abolished.

[0012] Accordingly, in a first aspect of the invention, there is provided a method of treating or preventing cancer in a subject, wherein the cancer has functional suppression of CDS1 , the method comprising administering a therapeutically effective amount of a CDS2 inhibitor to the subject.

[0013] In accordance with a second aspect of the invention there is provided a CDS2 inhibitor for use in a method of treating or preventing cancer characterised by functional suppression of CDS1 in a subject in need thereof, the method comprising the step of administering the CDS2 inhibitor to the subject.

[0014] In accordance with a third aspect of the invention there is provided a CDS2 inhibitor for use in the manufacture of a medicament for the treatment of cancer having functional suppression of CDS1.

[0015] In any of the first to third aspects of the invention, the method may comprise detecting presence or absence of functional suppression of CDS1 in the cancer.

[0016] The presence of functional suppression of CDS1 may have been determined in the subject prior to treatment.

[0017] In accordance with a fourth aspect of the invention there is provided a method for selecting a subject for the treatment of cancer with a cytidine diphosphate diacylglycerol synthase 2 (CDS2) inhibitor, the method comprising the steps of:

[0018] (i) providing a biological sample obtained from the subject,

[0019] (ii) detecting presence or absence of functional suppression of cytidine diphosphate diacylglycerol synthase 1 (CDS1 ) in the biological sample, and (iii) selecting the subject for treatment with the CDS2 inhibitor if the presence of functional suppression of CDS1 is detected.

[0020] In accordance with a fifth aspect of the invention there is provided a method for predicting whether a subject having cancer is likely to respond to treatment with a CDS2 inhibitor, the method comprising the steps of:

[0021] (i) providing a biological sample obtained from the subject,

[0022] (ii) detecting presence or absence of functional suppression of CDS1 in the biological sample, wherein if the presence of functional suppression of CDS1 is detected then predicting that the subject is likely to respond to treatment with the CDS2 inhibitor.

[0023] In accordance with a sixth aspect of the invention there is provided a method for identifying a subject with cancer who is likely to respond to treatment with a CDS2 inhibitor, the method comprising the steps of:

[0024] (i) providing a biological sample obtained from the subject,

[0025] (ii) detecting presence or absence of functional suppression of CDS1 in the biological sample, wherein the subject is likely to respond to treatment with the CDS2 inhibitor if the presence of functional suppression of CDS1 is detected.

[0026] The method may further comprise detecting presence or absence of functional suppression of CDS1 in a biological sample obtained from the subject.

[0027] In accordance with a seventh aspect of the invention there is provided a method of identifying and treating a subject suitable for treatment of cancer, the method comprising the steps of: i) obtaining a biological sample from the subject, ii) detecting presence or absence of functional suppression of CDS1 in the biological sample, iii) identifying the subject as suitable for treatment of cancer if the presence of functional suppression of CDS1 is detected, and iv) administering an effective amount of a CDS2 inhibitor to a subject identified as suitable for treatment. In accordance with an eighth aspect of the invention there is provided a method for detecting presence or absence of the functional suppression of CDS1 in a biological sample derived from a subject in order to select a subject for the treatment of cancer with a CDS2 inhibitor.

[0028] In accordance with a ninth aspect of the invention there is provided a method of screening for a compound for use in the treatment of cancer having functional suppression of CDS1 , the method comprising the steps of:

[0029] (i) administering the compound to cancer cells having functional suppression of CDS1 in an in vitro system or to an in vivo non-human animal model of the cancer having functional suppression of CDS1 , and

[0030] (ii) determining if there is any inhibition or suppression of cancer cell proliferation in vitro or of the cancer in the in vivo model, wherein if the compound results in inhibition or suppression of cancer cell proliferation in vitro or of the cancer in the in vivo model, then it may be useful in the treatment of cancer having functional suppression of CDS1 .

[0031] Functional suppression of CDS1 may comprise inactivating mutations, decreased transcription of a CDS1 gene resulting in low or no expression of CDS1 , decreased translation of CDS1 mRNA, heterozygous deletion of a CDS1 gene, or homozygous deletion of a CDS1 gene.

[0032] The functional suppression may be compared with a control.

[0033] The control may be the level of function of CDS1 in non-cancerous tissue in the subject or from the corresponding tissue type in a normal, healthy subject.

[0034] Inactivating mutations may completely eliminate the activity of the CDS1 protein, or it may reduce the level of activity compared to the wild-type level of activity. For example, inactivating mutations may reduce the activity of the CDS1 protein by at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or at least 99% reduced activity compared to the wild-type level of activity.

[0035] As CDS1 and CDS2 are enzymes that catalyse the conversion of PA into CDP-DAG, the level of activity can be measured directly by measuring the concentration of these substances over time using any suitable means. For example, enzyme-linked immunosorbent assay (ELISA) could be used to detect the amount of CDP-DAG at a given time, and an activity rate of production calculated from this. Metabolic profiling of the substrates of these enzymes using techniques such as nuclear magnetic resonance [NMR], liquid chromatography-mass spectrometry [LC-MS] could also be used. Decreased transcription of the CDS1 gene may result in complete elimination of CDS1 expression, or it may be a partially reduced level of transcription, i.e. low expression compared to the expression of the CDS1 gene in a control e.g. non-cancerous cells of the subject, or to the expression of the CDS1 gene in the corresponding cells of or tissue type of a normal, healthy subject. For example, expression of the CDS1 gene may be reduced by at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or at least 99% compared to the control.

[0036] Decreased translation of CDS1 mRNA may result in complete elimination of CDS1 expression, or it may be a partially reduced level of translation compared to the level of translation in a control, e.g. non-cancerous cells of the subject, or to the level of translation in the corresponding cells or tissue type of a normal, healthy subject. For example, translation of CDS1 mRNA may be reduced by at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or at least 99% compared to the control.

[0037] The CDS2 inhibitor may, for example, be an antibody, a small molecule inhibitor, or an RNA inhibitor.

[0038] The RNA inhibitor may be a short interfering RNA (siRNA), short hairpin (shRNA), guide RNA (gRNA), antisense oligonucleotide, antagomir, or microRNA.

[0039] The cancer treated may be uveal melanoma, diffuse B cell lymphoma, acute myeloid leukaemia, hepatocellular carcinoma, sarcoma, cutaneous melanoma, pancreatic cancer, skin cancer, blood cancer, brain cancer, lung cancer, colorectal cancer, breast cancer, renal cancer, cutaneous melanoma, liposarcoma, non-small cell lung cancer, or chronic myeloid leukaemia.

[0040] In accordance with any of the aspects of the invention above, the functional suppression may be of CDS2, and the treatment with a CDS1 inhibitor.

[0041] It is to be understood that cancers may have functional suppression of CDS2 rather than CDS1 . As presented herein, CDS1 and CDS2 form a synthetic lethal pair. As a result, if a cancer has functional suppression of either one of CDS1 or CDS2, then this may make it suitable for treatment with an inhibitor to the corresponding other protein or gene.

[0042] Brief description of the drawings

[0043] Embodiments and experiments will be discussed with reference to the accompanying figures, in which:

[0044] Figure 1 A shows a schematic of categories of gene pairs included in the combinatorial CRISPR library (right) and a pie chart of depicting the percentages of each category of pgRNA included in the combinatorial CRISPR library. The number of gene pairs in each group is shown, in addition to the number of single control genes and individual safe targeting guides (STG). A total of 25,449 unique paired guide RNA (pgRNA) combinations were included, targeting 568 gene pairs that were paralogous (n=293), putative synthetic lethal (n=225) or uveal-specific (n=50). Each gene within the gene pairs was combined with a safe-targeting guide (STG) to determine the effect of single gene knockout. A panel of 54 control essential and 115 control non-essential single genes were included to allow assessment of screen performance.

[0045] Figure 1 B shows the design of the dual paired gRNA construct. The first sgRNA in position 1 was placed under a hll6 promotor and the second sgRNA in position 2 was placed under a mll6 promotor. Non-identical tracrRNAs were used to minimise recombination. To assess single-guide activity, genes were paired with STGs placed in either position.

[0046] Figure 1 C shows the single gRNA pairing strategy. In position 1 , twelve gRNAs have been designed, four each against gene A, gene B and safe-targeting regions of the genome. In position 2, another twelve unique gRNAs have been designed, four each against gene A, gene B and safe-targeting regions of the genome. The 48 gRNA combinations used in the library are indicated in the shaded boxes. Safe-targeting guide, STG.

[0047] Figure 1 D shows a density plot of guide-level Iog2 fold change (FC) for each category of pgRNA across all ten human uveal melanoma cell lines. pgRNA constructs containing only STGs or STGs paired with control non-essential genes were labelled as negative controls. STGs paired with control essential genes were labelled as positive controls. STGs paired with guides against genes were labelled as single knockout. pgRNAs targeting gene pairs were labelled as double knockout.

[0048] Figure 1 E shows the calculation of the genetic interaction (Gl) score to quantify synthetic lethality. The sum of the observed Iog2 FC for sgRNAs targeting gene A and gene B was calculated to determine the predicted Iog2 FC of that pgRNA. The difference between the predicted and experimentally observed Iog2 FC was calculated. More negative Gl scores signify greater synthetic lethality.

[0049] Figure 1 F shows the number of common significantly depleted gene pairs across the uveal melanoma cell lines screened.

[0050] Figure 1 G shows the results of the combinatorial CRISPR screen in ten human uveal melanoma cell lines. Single guide analysis using BAGEL

[0050] calls CDS2 as essential in every cell line / time point. CDS1 is called as non-essential as it is not called as hit in any cell line / time point. Figure 2A shows gene expression of CDS1 across different tumour types, based upon data generated by the TCGA Research Network: https: / / www.cancer.gov / tcga. TPM; transcripts per million. Standard TCGA tumour abbreviations used.

[0051] Figure 2B shows a waterfall plot depicting CDS2 Chronos gene effect across 952 pancancer cell lines from Project Achilles, DepMap 21 Q2. Each column represents a cell line coloured by CDS1 expression.

[0052] Figure 3A shows a box plot comparison of CDS2 essentiality in CDS1 low expressers (log2(TPM+1) < 1) versus high expressers (log2(TPM+1) > 1) across 316 cell lines (left) analysed in Project Score and 937 cell lines (right) in Project Achilles, DepMap 22Q2. For each data set, a t-test (two-sided, equal variance) was performed. Box and whiskers plots indicate median and 5-95 percentile. Outliers as dots.

[0053] Left: Sanger dataset. Essentiality data is in units of quantile normalized “CRISPRcleaned” log fold change, downloaded from the Project SCORE dataset. Expression data is downloaded from Cell Model Passports, and is in units of log transcripts per million (TPM). Right: Broad dataset. Essentiality data refers to copy number-corrected Chronos scores, downloaded from the Project Achilles dataset (DepMap 22Q2). Expression data is downloaded from the DepMap portal (22Q4). 1 log TPM was used as a threshold to divide the datasets into CDS1 low- and high-expressers. For each dataset, a t-test (two-sided, equal variance) was performed to compare CDS2 essentiality in CDS1 low vs high expressers.

[0054] Figure 3B shows the correlation between CDS1 methylation and expression. Methylation fraction represents a weighted average of the methylation ratios of all CpG sites within 1000 bp from a gene’s transcriptional start site (TSS). The methylation ratio of each CpG site was determined by the number of reads where that CpG was methylated over the total number of reads covering that CpG. The weights of each CpG are calculated by the total number of reads covering that CpG overall the total number of reads covering any CpG within 1000 bp of the gene’s TSS. Data obtained from DepMap 23Q270. TPM; transcripts per million.

[0055] Figure 4A shows the number of significantly depleted genes by number of screened uveal melanoma cell lines, subdivided by genes that were members of paralogue families identified in Genome-wide single gRNA CRISPR screen.

[0056] Figure 4B shows the top five significantly depleted genes found across all ten screened uveal melanoma cell lines after known common essential genes have been removed. More essential genes are represented by a lower MAGeCK beta score and lower FDR. Figure 4C shows the validation of the findings of the screen using a 28-day competitive coculture assay. Cells were transduced with a lentivirus expressing a gRNA and a BFP marker. BFP Expression was measured by flow cytometry at baseline and at and day 14 and day 28. The proportion of surviving CDS2 gRNA-transduced cells was significantly reduced compared with non-transduced cells. Two gRNAs were tested against each gene including a safe-targeting gene (STG) control. Results were normalized to day 4 for each gRNA.

[0057] Figure 5A shows the results of a clonogenic assay showing the effects of CDS2 disruption at day 14 following the addition of doxycycline to Cas9-expressing cells that contain a doxycycline-inducible sgRNA against CDS2, compared with the DMSO control. Data representative of three independent biological replicates. The assays show reduced colony formation in cells that have been treated with doxycycline which induces CDS2 knockout, compared with the DMSO control.

[0058] Figure 5B shows the flow cytometry results from apoptosis assays at day 10 following treatment with doxycycline, compared with DMSO-treated transduced cells. Data are representative of three biological replicates performed in a single experiment. Data shown are mean and SD. Significance of combined apoptotic and dead populations shown, determined by t-test (twosided, equal variance). A, B, E and P refer to individual clonal cell lines; Cas9 refer to non-transduced control cell lines. Results show increased apoptosis in cells that have been treated with doxycycline which induces CDS2 knockout, compared with the DMSO control and Cas9 control.

[0059] Figure 6 shows disruption of CDS2 following doxycycline treatment leading to a progressive reduction in the phosphoinositides PI, PIP and PIP2 and in increase in the precursor phosphatidic acid (PA), compared with the DMSO control in cell lines OMM2.5 and MP41 . A reduction in PIP2 was also observed in MP41 with doxycycline treatment The dotted lines represent the doxycycline treated cell lines; the solid lines respresent the DMSO treated cell lines. Top row, OMM2.5; bottom row, MP41.

[0060] Figure 7A shows a schematic of the experimental procedure for the suppressor screen.

[0061] Figure 7B shows the results of the suppressor screen. Guides which are either not a hit (FDR < 0.05) or a hit in a single replicate are coloured light grey. Guides which are a hit in 2 replicates and 3 replicates are coloured in dark. Those guides which are a hit in two or more replicates and have a meanLFC > 6 are annotated with the guide’s gene symbol.

[0062] Figure 8 shows role of CDS1 and CDS2 in phosphoinositide synthesis and recycling, and the role of phosphoinositides in downstream signalling pathways in uveal melanoma. Figure 9 shows supersized Lipid Droplets (LDs) that are induced by the deletion of CDS2 in uveal melanoma cells, indicating that the cells are generating an excess of triacylglycerol by diverting the precursors not used for phosphoinositide synthesis, which accumulates in these fat stores.

[0063] Figures 10 and 11 show the results of a combinatorial gRNA CRISPR screen. CDS1 and CDS2 synthetic lethality is observed in non-small cell lung cancer cell lines and pancreatic cell lines. This effect is not observed in cutaneous melanoma cell lines which have low CDS1 expression, as CDS2 alone is found to be essential in a proportion of these cell lines.

[0064] Figures 12A shows the results of an independent combinatorial CRISPR screen in the following cell lines: colorectal, gastric, chronic myeloid leukaemia and renal. CDS2 is found to be a synthetic lethal gene pair in all 9 out of the 9 screened cell lines.

[0065] Figure 12B shows that the annotated cell lines have normal CDS1 expression.

[0066] Figure 12 C shows Identification of CDS1 / CDS2 synthetic lethality across multiple tumour types. CDS1 and CDS2 gene expression in cancers based upon data generated by the TCGA Research Network and obtained from the UCSC Xena platform. Quartiles indicated by dotted line; median indicated red line. Standard TCGA tumour abbreviations used.

[0067] LAML = Acute Myeloid Leukemia; ACC =Adrenocortical carcinoma; BLCA= Bladder Urothelial Carcinoma; LGG=Brain Lower Grade Glioma; BRCA = Breast invasive carcinoma; CESC =Cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL= Cholangiocarcinoma; COAD= Colon adenocarcinoma; ESCA= Esophageal carcinoma; GBM = Glioblastoma multiforme; HNSC= Head and Neck squamous cell carcinoma; KICH=Kidney Chromophobe; KIRC=Kidney renal clear cell carcinoma; KIRP=Kidney renal papillary cell carcinoma; LIHC=Liver hepatocellular carcinoma; LUAD= Lung adenocarcinoma; LUSC= Lung squamous cell carcinoma; DLBC= Lymphoid Neoplasm Diffuse Large B-cell Lymphoma; MESO=Mesothelioma; OV = Ovarian serous cystadenocarcinoma; PAAD = Pancreatic adenocarcinoma; PCPG = Pheochromocytoma and Paraganglioma; PRAD = Prostate adenocarcinoma; READ = Rectum adenocarcinoma; SARC= Sarcoma; SKCM = Skin Cutaneous Melanoma; STAD = Stomach adenocarcinoma; TGCT= Testicular Germ Cell Tumors; THYM Thymoma; THCA= thyroid carcinoma; UCS = Uterine Carcinosarcoma; UCEC = Uterine Corpus Endometrial Carcinoma; UVM = Uveal Melanoma.

[0068] Figure 13A shows a schematic of in vivo experiments performed in 28 female mice at 8 weeks of age (NOD-Prkdcscid-IL2rgTm1 / Rj background).

[0069] Figure 13B shows tumour growth (cm2) in mice that were xenografted with the Doxycycline- inducible CDS2 gRNA up to 50 days post-inoculation. Mice fed with the doxycycline- containing diet resulting in CDS2 disruption showed significantly reduced tumour growth compared with control mice that were fed with a normal diet not containing doxycycline. ** p<0.01.

[0070] Figure 14A shows a schematic depicting the bioinformatic pipeline for identifying cancerspecific SLIs using TCGA (left) and DepMap (right).

[0071] Figure 14B shows the correlation (Pearson’s r) between RNA, proteomic, mutation or methylation anchor gene data and gene dependency data in cancer cell lines for previously validated synthetic lethal pairs. The rank of the known target gene partners was plotted for each data type, n (cancer cell lines) = 913 (RNA), 299 (protein), 986 (damaging mutation) and 513 (methylation).

[0072] Figure 14C shows the enrichment of the previously reported synthetic lethal interactions FAM50A-FAM50B, RPP25L-RPP25, EIF1 AX-EIF1 AY, DDX3X-DDX3Y, DNAJC19-DNAJC15, TTC7A-TTC7B and MCL1 -BCL2L1 for different sample sizes. The correlation (Pearson’s r) between RNA and dependency data was calculated using all DepMap cancer cell lines or subsets corresponding to the indicated cancer types. The top 100 pairs were plotted and the data for the abovementioned synthetic lethal pairs is denoted in blue, n (cancer cell lines) = 913 (all), 122 (lung), 70 (CNS), 63 (skin) and 40 (breast).

[0073] Figure 14D shows the top predicted synthetic lethal pairs, named by their target gene and anchor gene, respectively. Y-axis presents the correlation (Pearson’s r) between the expression of the anchor gene and the dependency on the target gene (DepMap). X-axis presents the average reduction in anchor expression in patient tumour samples compared to patient healthy tissue samples (TCGA). Increases in expression are plotted as a 0% reduction. Bonferroni-corrected Pearson-associated p-values were used to determine significance cut-off. n = 913 cancer cell lines and 10.005 patient samples (9264 tumour, 741 healthy).

[0074] Figure 14E shows the average fold-change in patient tumour samples compared to patient healthy samples (TCGA) for the anchor genes of the top ten predicted synthetic lethal pairs in Fig. 14D. Note that the y-axis was cut at 2.0. n = 10.005 patient samples (9264 tumour, 741 healthy).

[0075] Figure 14F shows the RNA expression in patient tumour samples and patient healthy samples for downregulated anchor genes in panel b (TCGA). The three dotted lines in each violin represent the quartile cut-offs. Kruskal-Wallis test followed up by Dunn’s test was used, n = 10.005 patient samples (9264 tumour, 741 healthy). Figure 14G shows for each cancer lineage the percentage of ACDS2 or APARP lethal cancer cell lines (DepMap) plotted against the percentage of BRCA1 / 2-mutated or low CDS1 -expressing cancer cell lines. The cut- off for the latter was log2(TPM+1 ) of 1 .5. The cut-off for lethality was CERES<-0.5. A CERES of -1 corresponds to core essential, whereas a CERES of 0 corresponds to non-essential

[0033] . For each cell line data for the most essential PARP gene (out of 13) was included in the analysis, n = 906 cancer cell lines (40 breast).

[0076] Figure 14H shows the relative cancer mortality (SEER) plotted against the frequency of ACDS2 lethal cell lines (DepMap) for each cancer type. The fraction of ACDS2 or APARP lethal cell lines within each type is categorically depicted. The cut-off for lethality was CERES<-0.5. For each cell line, data for the most essential PARP gene (out of 13) was included in the analysis, n (cancer cell lines) = 906 (40 breast, 122 lung, 97 blood, 92 brain, 64 skin, 45 soft tissue).

[0077] Figure 141 shows the probability of survival for patients with the highest and lowest quartile of CDS1 expression of patient tumour samples (TCGA). The hazard ratio for the CDS1 -low quartile was 1.90 [1.715-2.106] compared to the CDS1 -high quartile. Log-rank Mantel-Cox test was used, n = 10163 patients

[0078] Figure 14J shows calibration gene-normalized RNA expression in healthy tissue samples (GTEx), cancer cell lines (DepMap), ACDS2 lethal cancer cell lines (DepMap, CERES<-0.5) and ACDS2 non-lethal cancer cell lines (DepMap, CERES>-0.5) in violin plots for the genes B2M, E2F1 , CDKN2B and CDS1 . The three dotted lines in each violin represent the quartile cut-offs. For significant differences the percentual change in average TPM was indicated. For controls one overall average was indicated. Kruskal-Wallis test followed up by Dunn’s test was used, n = 17382 healthy tissue samples and 913 cancer cell lines (657 ACDS2 non- lethal and 256 ACDS2 lethal).

[0079] Figure 14K shows, for the top four lineages from Figure 14H, the calibration gene normalized RNA expression in healthy tissue samples (GTEx), cancer cell lines (DepMap), ACDS2 lethal cancer cell lines (DepMap, CERES<-0.5) and ACDS2 non-lethal cancer cell lines (DepMap, CERES>-0.5) in violin plots for the CDS1 gene. The percentual change in average TPM between the healthy tissue and cancer cell lines was indicated. The three dotted lines in each violin represent the quartile cut-offs. Kruskal-Wallis test followed up by Dunn’s test was used, n (healthy tissue samples) = 578 (lung), 2642 (brain), 929 (blood), 1809 (skin) and n (cancer cell lines) = 122 (lung, 87 ACDS2 non-lethal and 35 ACDS2 lethal), 75 (brain, 50 ACDS2 non-lethal and 25 ACDS2 lethal), 48 (blood, 13 ACDS2 non- lethal and 35 ACDS2 lethal), 63 (skin, 39 ACDS2 non-lethal and 24 ACDS2 lethal). Figure 14L shows the calibration gene normalized CDS1 protein levels in matched healthy lung and lung tumour samples. The three dotted lines in each violin represent the quartile cut-offs. Mann-Whitney test was used, n = 110 tumour samples and 102 adjacent healthy lung samples.

[0080] Figure 15A shows a plot depicting synthetic lethality upon CRISPR perturbation of CDS2 (ACDS2 lethality) and CDS1 RNA expression for cancer cell lines (DepMap). Cancer cell lines were categorized by CDS1 RNA levels and cell lines used for validation are marked. Histograms depicting the distribution were added for both axes, n = 913 cancer cell lines.

[0081] Figure 15B shows a plot depicting CDS1 RNA expression as measured by RNA sequencing (DepMap) or qPCR analysis. The correlation (Pearson’s r) and the associated p-value (p) were added. Average of two independent experiments with each n = 4 replicates for 8 cancer cell lines.

[0082] Figure 15C shows a diagram depicting the method for quantifying lethality upon CRISPR perturbation using fluorescent tracker cells. The mCherry-positive percentage was quantified by flow cytometry.

[0083] Figure 15D shows a graph depicting cumulative lethality 14 days after CRISPR perturbation of CDS2 using tracker cells (see Figure 15C). CDS2 sgRNAs, positive control sgRNA (RPL19) and negative control sgRNA (sgControl) were included. Two-way ANOVA followed up by Dunnett’s test was used, n = 3 replicates for 7 cancer cell lines with 2 CDS2 sgRNAs.

[0084] Figure 15E shows representative histograms for the flow cytometry data used to calculate lethality for NCI-H2030 in Fig. 15D.

[0085] Figure 15F shows the day 8 results for the experiment shown in Fig. 15D. Two-way ANOVA followed up by Dunnett’s test was used, n = 3 replicates for 7 cancer cell lines with 2 CDS2 sgRNAs.

[0086] Figure 15H shows a diagram depicting method for quantifying synthetic lethality using ectopic CDS1 or GFP expression (left). The GFP-positive percentage was quantified by flow cytometry. The right panel shows a graph depicting quantification of synthetic lethality in melanoma. Two-way ANOVA followed up by Sidak’s test was used, n = 3 replicates for 4 cancer cell lines.

[0087] Figure 151 shows the 7 day results for the experiment figure 15H. Two-way ANOVA followed up by Sidak’s test was used. n= 3 replicates for 4 cancer cell lines.

[0088] Figure 15J shows a graph depicting cumulative lethality over 14 days after CRISPR perturbation of CDS2 using tracker cells (see Figure 15C) in indicated cell lines (A375; A549; CAKI1 ; COLO679). CDS2 sgRNA and negative control sgRNA (sgControl) were included. One sample T test was used using the cell lines as replicates, n = 3 replicates for 4 cancer cell lines.

[0089] Figure 16 shows a graph depicting a snapshot of the total protein normalized level of cleaved caspase-3 in ACDS2 or control samples of BLM and SK-MEL-2 cell lines, as determined by quantitative western blotting (Abby). Two-way ANOVA followed up by Sidak’s test was used, n = 3 replicates for 2 cancer cell lines.

[0090] Figure 17A shows a diagram depicting the in vivo variant of the method for quantifying synthetic lethality (left). A graph depicting quantification of the in vivo synthetic lethality in two melanoma cell line models is shown on the right. Two-way ANOVA followed up by Sidak’s test was used, n = 6 NOD-Scid I L2Rgnull mice per group each for 2 cancer cell lines.

[0091] Figure 17B shows representative histograms for the flow cytometry data used to calculate synthetic lethality in Figure 17A.

[0092] Figure 18A shows a diagram depicting the method used to screen for ACDS2 lethality escape mechanisms. One CDS1 -high cancer cell line and four CDS1 -negative cancer cell lines were screened.

[0093] Figure 18B shows the lethality of the screened cells between day 15 and day 25. Lethality was quantified using fluorescent tracker cells as depicted in Fig.15C. Two-way ANOVA followed up by Sidak’s test was used, n = 3 replicates for 5 cancer cell lines.

[0094] Figure 18C shows the colony formation assays with adherent cell lines used for the screen. Crystal violet was used to visualize adherent cells while scans show the results (top). Cell survival assays with the suspension screen cell line are shown at the bottom. Cell counts were used for quantification and heat maps present the result, n = 1 for 5 cancer cell lines, an independent experiment was performed for quantification (Fig. 18B).

[0095] Figure 18D shows the results of a quality control experiment. As a quality control, each control arm was compared with the original library to detect essential genes (true positives) over non-essential genes (true negatives). The resulting area under the receiver operator curve (ROC-AUC) values for each screen are presented. A ROC-AUC above 90% of maximum is considered outstanding for CRISPR screens, n = 1 for 5 cancer cell lines at 200x coverage with 4 sgRNA per gene and 1000 sgControls in the CRISPR library.

[0096] Figure 18E shows a graph depicting the screen hits identified in 3 / 4 or all 4 CDS1 -negative cancer cell lines. Effect size was calculated using the average ACDS2 lethality in the lethality assays with screen cells (panel b) and the fold-change enrichment of the second-best guides for each gene. Benjamini-Hochberg corrected p- values from MAGeCK resulted in the indicated significance, n = 1 for 4 cancer cell lines at 200x coverage with 4 sgRNA per gene and 1000 sgControls in the CRISPR library.

[0097] Figure 18F shows the average ACDS2 changes for knockouts of each other gene in the four CDS1-negative cancer cell lines plotted against the ACDS2 changes for knockouts of each other gene in the CDS1 -high cell line. The second-best sgRNA out of four guides was used to calculate effect size. The dotted line corresponds to the ACDS2 change required to compensate for the average lethality quantified in Fig. 18B. n = 1 for 5 cancer cell lines at 200x coverage with 4 sgRNA per gene and 1000 sgControls in the CRISPR library.

[0098] Figure 18G-J shows the lethality rescue plotted against the Benjamini-Hochberg corrected p- values from MAGeCK for each gene in the separate screens. The rescue was calculated using average lethality quantified in Figure 18B. The two genes significant in three cell lines (LCMT1 & EP300) and the two significant genes in two cell lines with the strongest rescue (PPP6C & PTPN11) were marked. Note that outliers above 100% were not plotted, n = 1 for 4 cancer cell lines at 200x coverage with 4 sgRNA per gene and 1000 sgControls in the CRISPR library.

[0099] Figure 19A shows an illustration of a method for phenotyping ACDS2 lethal / CDS1-low cancer cell lines and ACDS2 non-lethal / CDS1-high cancer cell lines (DepMap). The graphs on the right-hand side show genome wide calculated Pearson correlations. The five genes enriched on either side are marked. Bonferroni corrected Pearson associated p-values were used to determine significance cut-off. n = 913 cancer cell lines.

[0100] Figure 19B shows E-cadherin and vimentin expression in ACDS2 lethal and ACDS2 non- lethal cancer cell lines (DepMap). Mann-Whitney test was used, n = 913 cancer cell lines (657 ACDS2 non-lethal and 256ACDS2 lethal).

[0101] Figure 19C shows CDS1 promoter methylation in cancer cell lines and the subset of blood cancer cell lines (DepMap). Histograms were added to show the distribution of CDS1 promoter methylation. Mann-Whitney test was used, n = 792 cancer cell lines (72 blood cancer cell lines).

[0102] Figure 19D shows the gene set enrichment analysis (GSEA) results for Hallmark epithelial to mesenchymal transition. Results are from a full hallmark GSEA run on genome-wide CDS1 co-expression (Pearson’s r) utilizing DepMap and TOGA data. The Benjamini-Hochberg corrected GSEA associated p-values were used, n = 913 cancer cell lines and 11069 patient samples.

[0103] Figure 19E shows a diagram depicting the core CDS1 / 2 pathway. The lethality upon perturbation (colour) and lineage specific expression patterns (grey boxes on the left) of each enzyme in the pathway are indicated (DepMap). For abbreviations, see bottom of figure description. Gene dependency is represented by CERES, n = 913 cancer cell lines.

[0104] Figure 19F shows the correlation between ACDS2 lethality and ACDIPT lethality, APIK3CA lethality, PIK3CA inhibitor lethality (Alpelisib) or the pan-PI3K inhibitor lethality (Copanlisib) in solid cancer cell lines (DepMap). The Bonferroni corrected Pearson correlation associated p- values were used, n (cancer cell lines) =865 (CDIPT and PIK3CA), 428 (Copanlisib) and 441 (Alpelisib).

[0105] Figure 19G shows the rescue of synthetic lethality upon supplementation with phosphatidylinositol (PI) or phosphatidylcholine (PC, control). The method to quantify synthetic lethality was introduced in Figure 15H. PI and PC were complexed with lipid free bovine serum albumin. Two-way ANOVA followed up by Dunnett’s test was used, n = 3 replicates for 2 cancer cell lines.

[0106] Figure 19H shows a diagram depicting a method for quantifying the dead cell fraction using the DNA-intercalating dye DRAQ7. The plot on the right-hand side shows rescue of ACDS2 cell death upon supplementation with phosphatidylinositol (PI) in cultured BLM cells. Oneway ANOVA followed up by Sidak’s test was used, n = 3 replicates for 1 cancer cell line.

[0107] Figure 20A shows a diagram depicting the samples generated for quantifying lipids and proteins. One CDS1 -high cancer cell line and four CDS1 -negative cancer cell lines were analysed.

[0108] Figure 20B shows ACDS2 lipid changes, presented in alphabetic order. Only lipid classes detected in all samples were plotted. Stars indicate number of CDS1 -negative cancer cell lines with at least two-star significant lipid alterations compared to the CDS1 -high cancer cell line. For abbreviations see bottom of figure description. Significance was indicated by one * for each cell line with p-value<0.01 . Two-way ANOVA followed up by Sidak’s test was used, n = 4 replicates for 5 cancer cell lines.

[0109] Lipids: CE = cholesterol esters, Cerdl 8:0 = ceramide d18:0, Cerdl 8:1 = ceramide d18:1 , DG = diacylglycerol, FFA = free fatty acids, HexCER = hexocylceramides, LPC = lysophosphatidylcholine, LPE = lysophosphatidylethanolamine, PA = phosphatidic acid, PC = phosphatidylcholine, PE = phosphatidylethanolamine, PG = phosphatidylglycerol, PI = phosphatidylinositol, PS = phosphatidylserine, SM = sphingomyelin, TG = triglycerides

[0110] Figure 20C shows a graph with five ACDS2 lipid class changes presented (full results are shown in Fig. 20B). The stars indicate the number of CDS1 -negative cell lines with at least two-star significant lipid alterations compared with the CDS1-high cell line. Significance was indicated by one * for each cell line with p-value<0.01 . Two-way ANOVA followed up by Sidak’s test was used, n = 4 replicates for 5 cancer cell lines. Figure 20D shows a diagram of the CDS2 pathway with ACDS2-changed lipid classes in CDS1 -negative cancer cell lines marked with arrows. For lipid abbreviations see Figure 20B description. Included enzymes are annotated on UniProt.

[0111] Enzymes: PEMT = phosphatidylethanolamine n-methyltransferase, CEPT1 = choline / ethanolamine phosphotransferase 1 , DGATs = diacylglycerol o-acyltransferases, DGKs = diacylglycerol kinases, CDS = cytidine diphosphate diacylglycerol synthase, TAM41 = mitochondrial phosphatidate cytidylyltransferase, CDIPT = cytidine diphosphate diacylglycerol synthase inositol-3-phosphatidyltransferase, PGS1 = phosphatidylglycerophosphate synthase 1 , PTPMT1 = protein tyrosine phosphatase mitochondrial 1 .

[0112] Figure 20E shows the average ACDS2 protein changes for the four CDS1-negative cancer cell lines (y-axis) plotted against the ACDS2 protein changes in the CDS'! -high cancer cell line (x-axis). Significance was determined by students T test, n = 4 replicates for 5 cancer cell lines.

[0113] Figure 20F shows volcano plots presenting the ACDS2 -induced protein changes in five cancer cell lines. In addition to CDS2, the top and bottom five proteins significant in at least 3 / 4 CDS1 -negative cancer cell lines were marked, n = 4 replicates for 5 cancer cell lines.

[0114] Figure 20G the top ten significantly enriched GO-terms among ACDS2-changed proteins in the four CDS1 -negative cancer cell lines. Benjamini-Hochberg p-values from Stringdb were used, n = 4 replicates for 4 cancer cell lines.

[0115] Figure 20H shows cell areas staining positive for the lipid compartment dye BODIPY in live cell imaging. Two-way ANOVA followed up by Sidak’s test was used, n = 4 replicates for 2 cancer cell lines.

[0116] Figure 20I shows representative combined light and transmission electron microscopy images of a sgControl and a sgCDS2 SK-MEL-2 cell stained for lipids (Nile Red) and DNA (Hoechst). Lipid droplets are shown in white (*). DNA is marked with a #. High intensity signal areas were marked with symbols, n = 1 for 1 cancer cell line.

[0117] Figure 20J shows a cartoon depicting lipid compartments and the CDS2 pathway. The ACDS2 lipid class changes in the four CDS1 -negative cell lines are marked by arrows. For abbreviations see description of Figure 20B (lipids) or Figure 20C (enzymes).

[0118] Figure 21 A shows the gating strategy for in vitro GFP or mCherry mixing experiments.

[0119] Figure 21 B shows the gating strategy for the in vivo mixing experiment.

[0120] Figure 21 C shows the gating strategy for dead cell quantification used in the examples. Figure 22A shows the results of synthetic lethality rescue. Cell division was reduced by contact inhibition, CDK4 / 6 inhibition (Palbociclib) or both and synthetic lethality was quantified using the method described in 15H. Two-way ANOVA followed up by Tukey’s test resulted in the marked significance, n = 3 replicates for 2 cancer cell lines.

[0121] Figure 22B shows a heat map depicting the normalized enrichment score (NES) for enriched Hallmark gene sets in CDS1-low cell lines (DepMap) and their enrichment in CDS1-low cancer patient samples (TCGA). Results are from a full hallmark GSEA run on genome-wide CDS1 co-expression (Pearson’s r) utilizing DepMap and TCGA data. The Benjamini- Hochberg corrected GSEA associated p-values were used. N = 913 cancer cell lines and 11069 patient samples.

[0122] Figure 23A shows BLAST comparison between bacterial CDSA and human CDS2 with highlight of D384 (CDS2) and corresponding D249 (CDSA).

[0123] Figure 23B shows Alphafill comparison between bacterial CDSA and human CDS2 with highlight of D384 (human CDS2) and corresponding D249 (bacterial CDSA).

[0124] Figure 23C shows a graph depicting cumulative lethality over 14 days after CRISPR perturbation of CDS2 using tracker cells (see Figure 15C). CDS2 sgRNA and negative control sgRNA (sgControl) were included. Cell line variants ectopically expressing CDS1 , GFP, CDS2 fused with GFP (CDS2-GFP) or CDS2-GFP with D384A mutation were included. The CDS2-GFP’s have silent mutations making them impervious to sgCDS2. Two-way ANOVA followed up by Sidak’s test was used, n = 3 replicates for 2 cancer cell lines and for CDS2-GFP and CDS2-GFP D384A the variants were generated and included twice (1 & 2).

[0125] P-values in the figures are indicated by * (<0.05), ** (<0.01 ) and *** (<0.001 ). Deviations from this method of indicating p-values are indicated in the description of a figure.

[0126] Detailed description of the invention

[0127] Aspects and embodiments of the invention will now be discussed with reference to the accompanying figures. Further aspects and embodiments will be apparent to those skilled in the art.

[0128] The present disclosure relates to the finding that CDS1 and CDS2 have a synthetically lethal relationship across different cancer types, especially in cancers exhibiting functional suppression of CDS1.

[0129] Accordingly, provided is a method of treating or preventing cancer in a subject, wherein the cancer has functional suppression of CDS1 , the method comprising the step of administering a therapeutically effective amount of a CDS2 inhibitor to the subject. Diphosphate-diacylglycerol synthases (CDS) are critical enzymes that catalyse the formation of cytidine diphosphate-diacylglycerol (CDP-DAG) from phosphatidic acid (PA). Two isoforms of human CDS, CDS1 and CDS2 have been identified. These enzymes are believed to be localised to the endoplasmic reticulum. CDS2 is encoded by CDS2 (gene ID: 8760). Human CDS2 is available under Uniprot Identifier: 095674 and ENSEMBL ID: ENSG00000101290. The amino acid sequence has the database reference NP 003809.1 and is encoded by the nucleotide sequence available under the database reference NM 003818.4. CDS2 has a broad expression pattern and is found in virtually every tissue. CDS1 is encoded CDS1 (gene ID: 1040) located on 4q21. Human CDS1 is available under Uniport identifier: Q92903 and ENSEMBL ID: NSG00000163624. The amino acid sequence has the database reference NP 001254.2 and is encoded by the nucleotide sequence of the database reference NM 001263.4. CDS1 is strongly expressed in epithelial tissues, for example in the photoreceptor layer of adult retinas, suggesting a role in phototransduction. The two isoforms are discussed in detail in D’Souza (2014)

[0031] .

[0130] In the present invention, tumour, malignant tumour, cancer, malignant neoplasm, carcinoma, sarcoma, and the like are collectively referred to as "tumour" or "cancer".

[0131] In the present invention, the "subject" means a human or a non-human animal. The subject means, for example, a human or a non-human mammal affected by a disease on which a CDS2 inhibitor is expected to have therapeutic effects. Preferred examples of the disease on which a CDS2 inhibitor is expected to have therapeutic effects can include cancer. A subject suitable for treatment as described herein, may be a non-human mammal, such as a rodent (e.g. a guinea pig, a hamster, a rat, a mouse), murine (e.g. a mouse), canine (e.g. a dog), feline (e.g. a cat), equine (e.g. a horse), cattle, a primate, simian (e.g. a monkey or ape), a monkey (e.g. marmoset, baboon), an ape (e.g. gorilla, chimpanzee, orang-utan, gibbon), or a human. In some preferred embodiments, the individual is a human. In other preferred embodiments, non-human mammals, especially mammals that are conventionally used as models for demonstrating therapeutic efficacy in humans e.g. murine, primate, porcine, canine, or rabbit animals) may be employed.

[0132] Preferred examples of the "subject" according to the present invention can include humans and non-human mammals suspected of having cancer, and humans and non-human mammals diagnosed or previously with cancer.

[0133] In the present invention, the "biological sample" refers to a tissue, a liquid, or a cell isolated from an individual, or a mixture thereof. Examples thereof can include, but are not particularly limited to, biopsied tumour, spinal fluid, pleural fluid, intra-abdominal fluid, lymph, skin sections, blood, urine, faeces, sputum, the respiratory organs, the intestinal tract, the genitourinary tract, saliva, milk, the digestive organs, and cells collected therefrom.

[0134] Preferred examples of the "biological sample" can include a portion of resected tissue obtained during surgery performed for the purpose or treating cancer, a portion of tissue collected by biopsy or the like from a subject suspected of having cancer, and cells derived from blood, pleural fluid, or intra-abdominal fluid. The biological sample may be obtained from cancerous tissue or may contain cancer cells.

[0135] The biological sample may be protein extracts or nucleic acid extracts prepared from a tissue, a liquid, or a cell isolated from an individual, or a mixture thereof, etc. The preparation of the protein extracts or the nucleic acid extracts can be carried out by use of a protein preparation method or a nucleic acid preparation method known per se in the art.

[0136] The biological sample is preferably a biological sample collected before treatment with a CDS2 inhibitor. Use of such a biological sample enables sensitivity for the CDS2 inhibitor to be predicted before carrying out treatment with the CDS2 inhibitor, as described below. As a result, whether or not to apply treatment involving the CDS2 inhibitor to the subject can be determined, i.e., the subject to whom treatment involving the CDS2 inhibitor is applied can be selected based on the presence or absence of functional suppression of CDS1 in the cancer.

[0137] The method of treatment may thus comprise detecting the presence or absence of functional suppression of CDS1 in the cancer.

[0138] The "functional suppression of cytidine diphosphate diacylglycerol synthase 1 (CDS1)" according to the present invention includes deletion, decreased expression, inactivation, and the like, resulting in reduced or abolished activity of CDS1, or increased CDS1 protein degradation. The deletion of CDS1 includes homozygous deletion and heterozygous deletion. The decreased expression of CDS1 includes both decreased expression at the transcriptional level and decreased expression at the translational level.

[0139] In some cases, CDS1 may be functionally suppressed by means of an inactivating or loss- of-function (complete or partial) mutation (e.g. point mutations), chromosomal deletions, truncations, or breaks, in CDS1, polymorphisms in the encoding nucleic acid, mutations or polymorphisms in a gene encoding a regulatory factor, decreased transcription of a CDS1 gene resulting in low or no expression of CDS1 , decreased translation of CDS1 mRNA, heterozygous deletion of a CDS1 gene, or homozygous deletion of a CDS1 gene.

[0140] In some cases, the decreased expression may be the result of transcriptional repression of CDS1 by the zinc finger transcription factor zinc finger E-box binding homeobox 1 (ZEB1). In some cases, the decreased expression of CDS1 may be the result of repressed gene transcription due to methylation of the CDS1 promoter. In some cases, the decreased expression may be the result of clustered regularly interspaced short repeats (CRISPR)- meditated DNA modification as described herein.

[0141] In some cases, the functional suppression of CDS1 in a cancer may be due to the administration of a CDS1 inhibitor to the subject sequentially or simultaneously with the CDS2 inhibitor. In some instances, administration of the CDS1 inhibitor may intra-tumoral, e.g. by intra-tumoral injection. The CDS1 inhibitor may be, but is not particularly limited to, synthetic low molecular compound libraries, expression products of gene libraries, peptide libraries, siRNA, antibodies, substances released by bacteria, extracts and culture supernatants of cells (microbes, plant cells, and animal cells), purified or partially purified polypeptides, including antibodies and fragments or derivatives thereof, e.g. single domain antibodies, marine organisms, plant- or animal-derived extracts, and random phage peptide display libraries. CDS1 inhibitors are also commercially available and the skilled person will be familiar with methods for validating these for the present purposes.

[0142] In the present specification, the term "decreased expression" or “reduced expression” means that the expression level is lower or abolished when compared with a control (e.g., an average expression level of normal healthy subjects, such as in the corresponding tissue type, or an expression level in noncancer tissue of the same patient).

[0143] Examples of approaches for "detecting the functional suppression of CDS1" according to the present invention include, but are not particularly limited to, the detection methods set out below. Methods for extracting genomic DNA or RNA from a biological sample derived from a subject are not particularly limited, and an approach known in the art can be appropriately selected for use. Examples of the method for extracting genomic DNA include the SDS phenol method (a method which involves denaturing proteins in tissues preserved in a urea- containing solution or ethanol, with a proteolytic enzyme (proteinase K), a surfactant (SDS), and phenol, and extracting DNA from the tissues by precipitation with ethanol), and DNA extraction methods using Clean Columns (registered trademark, manufactured by NexTtec Biotechnologie GmbH), AquaPure (registered trademark, manufactured by Bio-Rad Laboratories, Inc.), ZR Plant / Seed DNA Kit (manufactured by Zymo Research Corp.), AquaGenomic Solution (registered trademark, manufactured by MoBiTec GmbH), prepGEM (registered trademark, manufactured by ZyGEM NZ Ltd.), Buccal Quick (registered trademark, manufactured by TrimGen Corp.), or the like. Examples of the method for extracting RNA include extraction methods using phenol and chaotropic salt (more specifically, extraction methods using a commercially available kit such as TRIzol (manufactured by Invitrogen Corp.) or ISOGEN (manufactured by Wako Pure Chemical Industries, Ltd.)), and methods using other commercially available kits (RNA Prep Total RNA Extraction Kit (manufactured by Beckman Coulter Inc.), RNeasy Mini (manufactured by Qiagen N.V.), RNA Extraction Kit (manufactured by Pharmacia Biotech), etc.). Examples of reverse transcriptases for use in further preparing cDNA from the extracted RNA include, but are not particularly limited to, reverse transcriptase derived from retrovirus such as RAV (Rous associated virus) or AMV (avian myeloblastosis virus), and reverse transcriptase derived from mouse retrovirus such as MMLV (Moloney murine leukemia virus).

[0144] The deletion of CDS1 can be confirmed by analysing the genotype of the CDS1 gene. The analysis of the genotype of the CDS1 gene can employ any method known in the art. For example, the analysis of the genotype can be carried out by detecting the copy number variation (CNV) of the CDS1 gene DNA. CNV refers to a phenomenon in which 1 kb or more of genomic DNA on the chromosomes is of 1 copy or less or 3 copies or more, though being of 2 copies in a normal human somatic cell, i.e., the diploid genome. It can be considered that: when the gene copy number is 1 copy or less, the genomic DNA is deleted; and when the gene copy number is 3 copies or more, the genomic DNA is duplicated. The detection of CNV can be carried out by use of a method known in the art. Specific examples of such a method can include an array CGH method, a single base polymorphism (SNP) array method, quantitative real time polymerase chain reaction (qPCR), a multiplex ligationdependent probe amplification method, next-generation sequencing, and digital PCR.

[0145] In the present specification, the term "decreased expression" or “reduced expression” means that the expression level is lower when compared with a control (e.g., an average expression level of normal subjects, an expression level in the corresponding tissue type of a normal, healthy subject, an expression level in noncancer tissue of the same patient, or an expression level in other cancerous tissue).

[0146] The level of expression of CDS1 may be reduced by at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90%, at least 95%, or at least 99% compared to the level of expression of CDS1 in the control.

[0147] As CDS1 and CDS2 are enzymes that catalyse the conversion of PA into CDP-DAG, the level of activity can be measured directly by measuring the concentration of these substances over time using any suitable means. For example, enzyme-linked immunosorbent assay (ELISA) may be used to detect the amount of CDP-DAG at a given time, and an activity rate of production calculated from this. Metabolic profiling of the substrates of these enzymes using techniques such as nuclear magnetic resonance [NMR], liquid chromatography-mass spectrometry [LC-MS] could also be used. A CDS activity assay using cell lysates with fluorescent PA substrate and running thin layer chromatography for detecting the product CDP-DAG as described in

[0061] may also be used.

[0148] Specifically, the analysis of the expression of the CDS1 gene in a biological sample can be carried out, for example, by measuring the amount of mRNA which is a transcript of the gene to be assayed, or measuring the amount of a protein which is a product of the gene to be assayed.

[0149] A gene expression detection method known in the art can be used as a method for measuring the amount of mRNA. For example, the measurement of the amount of mRNA can be carried out by use of a number of molecular biological approaches such as Northern blotting, dot blot, polymerase chain reaction (PCR), reverse transcription PCR (RT-PCR), quantitative RT-PCR, hybridization, DNA array method, next-generation sequencing, nanopore sequencing and digital PCR. Also, the measurement of the amount of mRNA can be carried out by a method known in the art using, as a probe, a polynucleotide having a DNA sequence that hybridizes under stringent conditions to the gene to be assayed. For example, for the preparation of a probe, the probe is appropriately bound to a label such as a fluorescent label. This labelled probe is hybridized to mRNA isolated and purified from a biological sample or cDNA synthesized from the mRNA. Then, fluorescence intensity derived from the hybridized probe can be measured to thereby detect the amount of mRNA of the gene to be assayed. The probe may be immobilized, for use, onto a support such as glass beads or a glass substrate. Specifically, the probe can also be used in the form of a DNA array or a DNA chip in which the probe prepared for the gene to be assayed is immobilized on a support. The support is not particularly limited as long as the support permits immobilization of polynucleotides. The support may have any shape or may be made of any material. General examples of the support can include: inorganic materials such as glass plates, silicon wafers, and resins; nitrocellulose as a natural polymer material; and nylon as a synthetic polymer material.

[0150] A commercially available DNA chip or DNA array can be used. The polynucleotide to be immobilized onto a support may be a synthetic oligonucleotide. A nucleic acid derivative, which is capable of being fluorescently labeled, may be introduced into the sequence of the synthetic oligonucleotide. Any of the Affymetrix DNA chip technique, which can synthesize the oligonucleotide of interest on a support, and the Stanford DNA chip technique, which immobilizes a synthesized DNA fragment as a DNA probe by spotting, can be used. Alternatively, the desired polynucleotide may be immobilized by spotting onto the surface of the upper end of columns of the 3D-Gene product (manufactured by Toray Industries, Inc.) which acts as a support having a three-dimensional structure. The phrase "hybridize under stringent conditions" means that the hybridization is maintained even after washing treatment at 42°C with, for example, a buffer solution containing 1 x SSC (0.15 M NaCI and 0.015 M sodium citrate) and 0.1% sodium dodecyl sulfate (SDS). In addition to the temperature conditions described above, various other factors influence the stringency of hybridization. Those skilled in the art can achieve stringency, which is equivalent to the stringency of hybridization shown above as an example, by combining various factors.

[0151] A probe and a primer set for quantitatively detecting mRNA or cDNA derived from the CDS1 gene are not particularly limited as long as the mRNA or the cDNA can be specifically detected. An oligonucleotide consisting of 12 to 26 bases is preferred. Such a probe and a primer set can be appropriately designed on the basis of information on the nucleotide sequence of the gene to be assayed. Then, the oligonucleotide having the determined sequence can be synthesized according to routine methods using, for example, a DNA synthesizer. Alternatively, the desired primer or probe may be selected, for use, from commercially available primers or probes for gene detection.

[0152] A protein assay method known in the art can be used as a method for quantitatively measuring the CDS1 protein. For example, various methods using an antibody against the CDS1 protein can be applied thereto. Specific examples thereof can include Western blot, enzyme-linked immunosorbent assay (ELISA), and radioimmunoassay (RIA).

[0153] The antibody against the CDS1 protein is directed to the CDS1 protein as an antigen. A human antibody, a mouse antibody, a rat antibody, a rabbit antibody, a sheep antibody, or the like can be appropriately used as long as the antibody specifically binds to the antigen. The antibody may be a polyclonal antibody or may be a monoclonal antibody, or a fragment or derivative thereof such as Fab and F(ab’)2 fragments, or a single domain antibody. A monoclonal antibody is preferred because homogeneous antibodies can be stably produced. The polyclonal antibody and the monoclonal antibody can be prepared by methods well known to those skilled in the art. Alternatively, the desired antibody may be selected, for use, from commercially available antibodies, for example those described in Blunsom et al

[0065] .

[0154] Monoclonal antibody-producing hybridomas can be prepared, generally, by use of a technique known in the art as follows: the antigen of interest or cells expressing the antigen of interest is used as a sensitizing antigen. The desired animal is immunized with this sensitizing antigen according to an ordinary immunization method, and the resulting immunocytes can be fused with known parent cells by an ordinary cell fusion method, followed by the selection of cells (hybridoma cells) producing the desired monoclonal antibody by an ordinary screening method to prepare the hybridomas. The preparation of the hybridomas can be carried out according to, for example, the method of Milstein et al. ("Methods of Enzymology", 1981 , Vol. 73, p. 3-46 ). In this context, the preparation of the monoclonal antibody can employ the CDS1 protein or a fragment thereof as an antigen. Those skilled in the art can easily obtain the CDS1 protein or the fragment thereof according to a method described in a textbook, for example, Sambrook et al. ed., "Molecular Cloning A Laboratory Manual", 2nd edition, Vol. 1-3, Cold Spring Harbor Laboratory Press, N.Y., 1989 .

[0155] For the quantification of the CDS1 protein, the protein, the fragment thereof, or the antibody may be immobilized, for use, on a support. The support is not limited as long as the support permits immobilization of proteins. General examples thereof can include: inorganic materials such as glass plates, silicon wafers, and resins; nitrocellulose as a natural polymer material; and nylon and polystyrene as synthetic polymer materials.

[0156] The CDS1 protein can also be detected by use of mass spectrometry (MS). Particularly, liquid chromatography-mass spectrometry (LC / MS) is sensitive and therefore advantageous. The measurement by mass spectrometry can be performed, for example, by preparing proteins from a biological sample, labelling the proteins, fractionating the proteins, subjecting the fractionated proteins to mass spectrometry, and identifying the CDS1 protein from the mass spectrometry value. An isotope labelling reagent known in the art can be used as the label. An appropriate labelling reagent can be obtained as a commercially available product. Also, the fractionation can be performed by a method known in the art and can be performed using, for example, a commercially available strongly cationic column.

[0157] Meanwhile, it is known in the art that promoter hypermethylation is partly responsible for the decreased expression of a gene. Thus, it is also possible to detect the presence or absence of the functional suppression of CDS1 by using the methylation of the CDS1 gene promoter as an index. The detection of promoter methylation can employ a method known in the art, for example, a method of directly detecting, by nucleotide sequence determination, changes in nucleotide sequence after treatment with bisulfite having the activity of converting methylated cytosine to uracil, or a method of indirectly detecting this change using a restriction endonuclease that can recognize (or cleave) a nucleotide sequence before the bisulfite treatment, but cannot recognize (or cleave) the nucleotide sequence after the bisulfite treatment.

[0158] In the present invention, the phrase "detect mutation" means detecting mutation of genomic DNA. If the mutation of genomic DNA is reflected in a base change in a transcript or an amino acid change in a translation product, the phrase "detect mutation" is also meant to include detecting the change in the transcript or the translation product (i.e., indirect detection). In some embodiments, the method of the present invention is a method for detecting mutation by directly determining the nucleotide sequence of the CDS1 gene region in a biological sample derived from a subject. In the present specification, the " CDS1 gene region" means a given region of genomic DNA comprising the CDS1 gene. The region comprises expression control regions (e.g., a promoter region and an enhancer region) of the CDS1 gene, a 3'-terminal noncoding region of the CDS1 gene, and the like. Mutation in these regions may influence, for example, the transcriptional activity of the CDS1 gene.

[0159] The determination of the nucleotide sequence can be performed by a method generally known to those skilled in the art, such as the Maxam-Gilbert method or the Sanger method.

[0160] The determined nucleotide sequence of DNA or cDNA can be compared with a control (e.g., the nucleotide sequence of DNA or cDNA derived from a noncancer tissue of the same patient) thereby to determine the presence or absence of the mutation of the CDS1 gene region in the test tissue.

[0161] The method for detecting mutation of the CDS1 gene region can be performed by various methods capable of detecting mutation, in addition to the method which involves directly determining the nucleotide sequence of DNA or cDNA.

[0162] For example, the mutation detection according to the present invention can be performed by a method as mentioned below. First, a DNA or cDNA sample is prepared from a biological sample. Subsequently, an oligonucleotide probe having a nucleotide sequence complementary to a nucleotide sequence containing the mutation of the CDS1 gene region is prepared by labelling with a reporter fluorescent dye and a quencher fluorescent dye. Then, the oligonucleotide probe is hybridized to the DNA sample. Further, the nucleotide sequence containing the mutation of the CDS1 gene region is amplified by using the DNA sample hybridized with the oligonucleotide probe as a template. Then, fluorescence emitted by the reporter fluorescent dye, due to the decomposition of the oligonucleotide probe resulting from amplification, is detected. Subsequently, the detected fluorescence is compared with a control. Examples of such a method include the double dye probe method, i.e. the so-called TaqMan(R) probe method.

[0163] In another method, a DNA or cDNA sample is prepared from a biological sample. Subsequently, in a reaction system involving an intercalator that emits fluorescence when inserted between the strands of a DNA duplex, the nucleotide sequence containing the mutation of the CDS1 gene region is amplified by using the DNA sample as a template. Then, the temperature of the reaction system is changed, and variation in the intensity of fluorescence emitted by the intercalator is detected. The detected variation in the intensity of fluorescence in conjunction with the change in the temperature is compared with a control. Examples of such a method include HRM (high resolution melting).

[0164] In a further alternative method a DNA or cDNA sample is firstly prepared from a biological sample. Subsequently, DNA containing the mutation site of the CDS1 gene region is amplified. Further, the amplified DNA is cleaved with restriction enzymes. Subsequently, the DNA fragment is separated according to its size. Subsequently, the size of the detected DNA fragment is compared with a control. Examples of such a method include a method based on restriction fragment length polymorphism (RFLP), and PCR-RFLP.

[0165] In a further alternative method, first, a DNA or cDNA sample is prepared from a biological sample. Subsequently, DNA containing the mutation site of the CDS1 gene region is amplified. Further, the amplified DNA is dissociated into single-stranded DNA. Subsequently, the single-stranded DNA thus obtained by dissociation is separated on a nondenaturing gel. The mobility of the separated single-stranded DNA on the gel is compared with a control. Examples of such a method include PCR-SSCP (single-strand conformation polymorphism).

[0166] In a further alternative method, first, a DNA or cDNA sample is prepared from a biological sample. Subsequently, DNA containing the mutation site of the CDS1 gene region is amplified. Further, the amplified DNA is separated on a gel where the concentration of a DNA denaturant is gradually elevated. Subsequently, the mobility of the separated DNA on the gel is compared with a control. Examples of such a method include denaturant gradient gel electrophoresis (DGGE).

[0167] A further alternative method employs DNA containing the mutation site of the CDS1 gene region prepared from a biological sample, and a substrate on which an oligonucleotide probe that hybridizes to the DNA is immobilized. Examples of such a method include DNA array methods.

[0168] In a further alternative method a DNA or cDNA sample is firstly prepared from a biological sample. Also, an "oligonucleotide primer having a nucleotide sequence complementary to one base immediately downstream from the base of the mutation site of the CDS1 gene region, and its downstream nucleotide sequence" is prepared. Subsequently, ddNTP primer extension reaction is performed by using the DNA as a template and by using the primer. Subsequently, the primer extension reaction product is applied to a mass spectrometer to conduct mass measurement. Subsequently, the genotype is determined from the mass measurement results. Subsequently, the determined genotype is compared with a control. Examples of such a method include MALDI-TOF / MS.

[0169] In a further alternative method a DNA or cDNA sample is firstly prepared from a biological sample. Subsequently, an oligonucleotide probe consisting of 5'-"a nucleotide sequence complementary to the base of the mutation site of the CDS1 gene region, and its upstream nucleotide sequence"-"a nucleotide sequence that hybridizes to neither one base immediately downstream from the mutation site of the CDS1 gene region nor its downstream nucleotide sequencers' (flap) is prepared. Also, an "oligonucleotide probe having a nucleotide sequence complementary to the base of the mutation site of the CDS1 gene region and its downstream nucleotide sequence" is prepared. Subsequently, these two types of oligonucleotide probes are hybridized to the prepared DNA. Subsequently, the hybridized DNA is cleaved with a single-stranded DNA cleaving enzyme to liberate the flap. In the present invention, examples of the single-stranded DNA cleaving enzyme include, but are not particularly limited to, cleavase. In this method, subsequently, a reporter fluorescence- and quencher fluorescence-labeled oligonucleotide probe having a sequence complementary to the flap is hybridized to the flap. Subsequently, the intensity of the generated fluorescence is measured. Subsequently, the measured fluorescence intensity is compared with a control. Examples of such a method include the Invader method.

[0170] In a further alternative method a DNA or cDNA sample is firstly prepared from a biological sample. Subsequently, DNA containing the mutation site of the CDS1 gene region is amplified. Then, the amplified DNA is dissociated into single strands. Only one of the single strands of the dissociated DNA is then isolated. Subsequently, a base extension reaction is performed, base by base, from near the base of the mutation site of the CDS1 gene region. Pyrophosphoric acid formed during the reaction brings about enzymatic emission of light, and the intensity of the light is measured. Then, the measured fluorescence intensity is compared with a control. Examples of such a method include pyrosequencing.

[0171] In a further alternative method a DNA or cDNA sample is firstly prepared from a biological sample. Subsequently, DNA containing the mutation site of the CDS1 gene region is amplified. Subsequently, an "oligonucleotide primer having a nucleotide sequence complementary to one base immediately downstream from the base of the mutation site of the CDS1 gene region, and its downstream nucleotide sequence" is prepared.

[0172] Subsequently, a single base extension reaction is performed by using the amplified DNA as a template and the prepared primer in the presence of fluorescently labelled nucleotides. Then, the degree of polarization of fluorescence is measured. Subsequently, the measured degree of polarization of fluorescence is compared with a control. Examples of such a method include AcycloPrime.

[0173] In a further alternative method a DNA or cDNA sample is firstly prepared from a biological sample. Subsequently, DNA containing the mutation site of the CDS1 gene region is amplified. Subsequently, an "oligonucleotide primer having a nucleotide sequence complementary to one base immediately downstream from the base of the mutation site of the CDS1 gene region, and its downstream nucleotide sequence" is prepared. Subsequently, a single base extension reaction is performed by using the amplified DNA as a template and the prepared primer in the presence of fluorescently labelled nucleotides. Subsequently, the base species used in the single base extension reaction is determined. Subsequently, the determined base species is compared with a control. Examples of such a method include SNuPE. If the mutation results in an amino acid change (e.g., substitution, deletion, or insertion) in the CDS1 protein, the sample prepared from the biological sample may be a protein. In such a case, the mutation detection can employ a method using a molecule (e.g., an antibody) specifically binding to a site with the amino acid change caused by the mutation.

[0174] The "CDS2 inhibitor" according to the present invention is not particularly limited and may be a compound known in the art or may be a compound identified by the screening described below. A CDS2 inhibitor may be any compound that reduces or abolishes CDS2 activity.

[0175] CDS2 inhibitors may be, but are not particularly limited to, targeted protein degraders such as Proteolysis-targeting chimera (PROTAC), synthetic low molecular compound libraries, expression products of gene libraries, peptide libraries, siRNA, antibodies, substances released by bacteria, extracts and culture supernatants of cells (microbes, plant cells, and animal cells), purified or partially purified polypeptides, including antibodies and fragments or derivatives thereof, such as Fab, scFv, single domain antibodies, marine organisms, planter animal-derived extracts, and random phage peptide display libraries, CRISPR-based inhibitors as described herein.

[0176] The CDS2 inhibitor may inhibit the enzymatic activity of CDS2. Inhibition of the enzymatic activity may be tested as described above.

[0177] The CDS2 inhibitor for use in the present invention may be formulated as pharmaceutical compositions for clinical use and may comprise a pharmaceutically acceptable carrier, diluent or adjuvant.

[0178] The CDS2 inhibitor can be orally administered or parenterally administered (e.g., intravenously administered, intraarterially administered, or locally administered) in various forms such as tablets, powders, granules, capsules, and solutions according to its characteristics to a cancer patient. The dose is not particularly limited as long as the dose is an amount effective for treating cancer by inhibiting CDS2. The dose can be appropriately selected according to the properties of the compound as well as the age, body weight, symptoms, and health conditions of the cancer patient, the degree of progression of cancer, etc. The dosing frequency is not particularly limited and can be appropriately selected according to purpose. For example, a daily dose may be administered once a day or may be administered in several divided portions per day. In the case of administering the compound inhibiting CDS2 to a human, the dose range is approximately 0.01 mg / kg body weight to approximately 500 mg / kg body weight, preferably approximately 0.1 mg / kg body weight to approximately 100 mg / kg body weight, per day. For administration to a human the daily dose is, preferably, administered once a day or in two to four divided portions per day, and this administration is preferably repeated at an appropriate interval. Also, the daily dose may exceed the amount described above, if necessary, at a physician's discretion.

[0179] The CDS2 inhibitor may be mixed with a pharmaceutically acceptable carrier, and the composition can be formulated by a pharmaceutical method known in the art to prepare a pharmaceutical product. Examples of the pharmaceutically acceptable carrier include, but are not limited to, sterile water, saline, plant oils, solvents, bases, emulsifiers, suspending agents, surfactants, stabilizers, flavors, aromatic substances, excipients, vehicles, antiseptics, binders, diluents, tonicity agents, soothing agents, bulking agents, disintegrants, buffers, coating agents, lubricants, colorants, sweeteners, thickeners, corrigents, solubilizers, and other additives.

[0180] The composition may be formulated for topical, parenteral, systemic, intravenous, intraarterial, intramuscular, intrathecal, intraocular, intratumoural, subcutaneous, oral or transdermal routes of administration which may include injection. Injectable formulations may comprise the selected compound in a sterile or isotonic medium.

[0181] Administration is preferably in a "therapeutically effective amount", this being sufficient to show benefit to the subject. The actual amount administered, and rate and time-course of administration, will depend on the nature and severity of the disease being treated.

[0182] Prescription of treatment, e.g. decisions on dosage etc, is within the responsibility of general practitioners and other medical doctors, and typically takes account of the disorder to be treated, the condition of the individual patient, the site of delivery, the method of administration and other factors known to practitioners. Examples of the techniques and protocols mentioned above can be found in Remington’s Pharmaceutical Sciences, 20th Edition, 2000, pub. Lippincott, Williams & Wilkins.

[0183] Alternatively, targeting therapies may be used to deliver the CDS2 inhibitor more specifically to certain types of cell, by the use of targeting systems such as antibody or cell specific ligands. Targeting may be desirable for a variety of reasons; for example if the agent is unacceptably toxic, or if it would otherwise require too high a dosage, or if it would not otherwise be able to enter the target cells. A composition may be administered alone or in combination with other treatments, either simultaneously or sequentially dependent upon the condition to be treated by any convenient route.

[0184] A CDS2 inhibitor, or composition comprising a CDS2 inhibitor, as described herein may be administered in combination with an anti-cancer agent. Examples of suitable anti-cancer agents include chemoactive agents, for example alkylating agents such as platinum complexes including cisplatin, mono(platinum), bis(platinum), tri-nuclear platinum complexes and carboplatin, thiotepa and cyclosphosphamide (CYTOXAN); alkyl sulfonates such as busulfan, improsulfan and piposulfan; aziridines such as benzodopa, carboquone, meturedopa, and uredopa; ethylenimines and methylamelamines including altretamine, triethylenemelamine, trietylenephosphoramide, triethylenethiophosphaoramide and trimethylolomelamime; nitrogen mustards such as chlorambucil, chlornaphazine, cholophosphamide, estramustine, ifosfamide, mechlorethamine, mechlorethamine oxide hydrochloride, melphalan, novembichin, phenesterine, prednimustine, trofosfamide, uracil mustard; nitrosureas such as carmustine, chlorozotocin, fotemustine, lomustine, nimustine, ranimustine; antibiotics such as aclacinomysins, actinomycin, authramycin, azaserine, bleomycins, cactinomycin, calicheamicin, carabicin, caminomycin, carzinophilin, chromomycins, dactinomycin, daunorubicin, detorubicin, 6-diazo-5-oxo-L-norleucine, doxorubicin, epirubicin, esorubicin, idarubicin, marcellomycin, mitomycins, mycophenolic acid, nogalamycin, olivomycins, peplomycin, potfiromycin, puromycin, quelamycin, rodorubicin, streptonigrin, streptozocin, tubercidin, ubenimex, zinostatin, zorubicin; antimetabolites such as methotrexate and 5-fluorouracil (5-FU) gemcitabine, fluorouracil, capecitabine, methotrexate sodium, ralitrexed, pemetrexed, tegafur, cytosine arabinoside, thioguanine, 5-azacytidine, 6-mercaptopurine, azathioprine, 6-thioguanine, pentostatin, pitavastatin, fludarabine phosphate, and cladribine; folic acid analogues such as denopterin, methotrexate, pteropterin, trimetrexate; purine analogs such as fludarabine, 6- mercaptopurine, thiamiprine, thioguanine; pyrimidine analogs such as ancitabine, azacitidine, 6-azauridine, carmofur, cytosine arabinoside, dideoxyuridine, doxifluridine, enocitabine, floxuridine, 5-FU; androgens such as calusterone, dromostanolone propionate, epitiostanol, mepitiostane, testolactone; anti-adrenals such as aminoglutethimide, mitotane, trilostane; folic acid replenishers such as folinic acid; aceglatone; aldophosphamide glycoside; aminolevulinic acid; amsacrine; bestrabucil; bisantrene; edatraxate; defofamine; demecolcine; diaziquone; elformithine; elliptinium acetate; etoglucid; gallium nitrate; hydroxyurea; lentinan; lonidamine; mitoguazone; mitoxantrone; mopidamol; nitracrine; pentostatin; phenamet; pirarubicin; podophyllinic acid; 2-ethylhydrazide; procarbazine; PSK; razoxane; sizofuran; spirogermanium; tenuazonic acid; triaziquone; 2, 2', 2"- trichlorotriethylamine; urethan; vindesine; dacarbazine; mannomustine; mitobronitol; mitolactol; pipobroman; gacytosine; arabinoside (Ara-C); taxoids, e.g. paclitaxel (TAXOL) and docetaxel (TAXOTERE); chlorambucil; gemcitabine; 6-thioguanine; mercaptopurine; platinum analogs such as cisplatin and carboplatin; vinblastine; platinum; etoposide (VP-16); ifosfamide; mitomycin C; mitoxantrone; vincristine; vinorelbine; binbiastine; vindesine; navelbine; novantrone; teniposide; daunomycin; aminopterin; ibandronate; CPT11 ; topoisomerase inhibitor RFS 2000; difluoromethylornithine (DMFO); retinoic acid; esperamicins; capecitabine (XELODA); Topoisomerase inhibitors such as doxorubicin HCI, daunorubicin citrate, mitoxantrone HCI, actinomycin D, etoposide, topotecan HCI, teniposide (VM-26), and irinotecan and pharmaceutically acceptable salts, acids or derivatives of any of the above.

[0185] In the present invention, it has been found that CDS1 and CDS2 are in a synthetic lethal relationship in cancer cells, and the inhibition of CDS2 in cancer cells having the functional suppression of CDS1 is capable of suppressing the growth of the cancer cells. On the basis of this finding, responsiveness to the treatment of cancer with a compound inhibiting CDS2 can be evaluated by using the functional suppression of CDS1 as a therapeutic index. Thus, the present invention provides a method for predicting responsiveness to the treatment of cancer with a CDS2 inhibitor, comprising using a biological sample derived from a subject, detecting the presence or absence of the functional suppression of CDS1 in the biological sample, and thereby determining that the subject for which the functional suppression of CDS1 has been detected has responsiveness to the treatment of cancer with a CDS2 inhibitor.

[0186] The subject for whom the functional suppression of CDS1 has been detected in this way is also suitable for the treatment of cancer with a CDS2 inhibitor. A patient who is to receive the treatment of cancer with a CDS2 inhibitor, or otherwise can be selected by using the functional suppression of CDS1 as reference, and efficiently treated. Thus, the present invention provides a method for selecting a subject for the treatment of cancer with a CDS2 inhibitor, comprising using a biological sample derived from the subject, detecting the presence or absence of the functional suppression of CDS1 in the biological sample, and selecting the subject for whom the functional suppression of CDS2 has been detected as a subject for the treatment of cancer with a CDS2 inhibitor.

[0187] The present invention also provides a method for identifying a subject with cancer who is likely to respond to treatment with a CDS2 inhibitor, the method comprising the steps of providing a biological sample obtained from the subject, detecting presence or absence of functional suppression of CDS1 in the biological sample, wherein the subject is likely to respond to treatment with the CDS2 inhibitor if the presence of functional suppression of CDS1 is detected.

[0188] Also provided is a method of identifying and treating a subject suitable for treatment of cancer, the method comprising the steps of obtaining a biological sample from the subject, detecting presence or absence of functional suppression of CDS1 in the biological sample, identifying the subject as suitable for treatment of cancer if the presence of functional suppression of CDS1 is detected, and administering an effective amount of a CDS2 inhibitor to a subject identified as suitable for treatment.

[0189] The CDS2 inhibitor may be as described above or as identified by the method screening described herein.

[0190] The present invention provides a method for screening for a compound for use in the treatment of cancer, i.e. for a CDS2 inhibitor, having the functional suppression of CDS1 described herein, the method comprising the step of selecting the compound by determining whether or not it inhibits CDS2 as an index. Since CDS1 and CDS2 are orthologous enzymes, the screening system of the present invention can be established by using a test to confirm the activity of these enzymes is suppressed when the test compound is screened.

[0191] Thus, provided herein is a method of screening for a compound for use in the treatment of cancer having functional suppression of CDS1 , the method comprising the steps of:

[0192] (i) administering the compound to cancer cells having functional suppression of CDS1 in an in vitro system or to an in vivo non-human animal model of the cancer having functional suppression of CDS1 , and

[0193] (ii) determining if there is any inhibition or suppression of cancer cell proliferation in vitro or of the cancer in the in vivo model, wherein if the compound results in inhibition or suppression of cancer cell proliferation in vitro or of the cancer in the in vivo model, then it may be useful in the treatment of cancer having functional suppression of CDS1 .

[0194] In vitro models as used herein may include tumour cell lines. Examples of tumour cell lines include, but are not limited to melanoma cell lines, such as SK-MEL-2 or 0MM2, SK-MEL- 147, BLM, D10, A375, or other cancers such as NCI-H2030, NCI-H1975, LPS141 , K562, Jurkat, CAKI1 COLO679 and A549.

[0195] In vivo models as used herein may include, but are not particularly limited to, mouse models. For example, the mouse models may be immunodeficient NOD-Scid I L2Rgnull mice xeno transplanted with a cancer cell line as described above. The cancer cell line may have functional suppression or deletion of CDS1 due to naturally occurring inactivating or loss-of-function (complete or partial) mutation (e.g. point mutations), chromosomal deletions, truncations, or breaks, in CDS1, polymorphisms in the encoding nucleic acid, mutations or polymorphisms in a gene encoding a regulatory factor, decreased transcription of a CDS1 gene resulting in low or no expression of CDS1 , decreased translation of CDS1 mRNA, heterozygous deletion of a CDS1 gene, or homozygous deletion of a CDS1 gene. Alternatively, CDS1 expression may be low or abolished due to perturbation of CDS1 , for example, by CRISPR using a single guide (sg)RNA, or siRNA.

[0196] The present inventors show that the synthetic lethality between CDS1 and CDS2 is caused at least partially by increased apoptosis in cancer cells. Accordingly, inhibition or suppression of cancer cell proliferation may be tested by measuring the levels of apoptosis. Increased apoptosis following treatment with the compound relative to untreated cancer cell, is indicative of suppression of cancer cell proliferation. The skilled person will be familiar with assays that measure apoptosis, for example by measuring the levels of caspase 3. A relative increase in caspase 3 in the cancer cell is indicative of increased apoptosis (and thus reduced proliferation). Other ways of quantifying cell proliferation include cell viability assays, e.g., Brdu detection by flow cytometry, ELISA, or immune histochemistry, detecting the presence of and proliferation assays, e.g. measuring the presence of cell proliferation markers such as Ki-67 by flow cytometry. The skilled person will be familiar with other methods for assessing inhibition or suppression of cancer cell proliferation.

[0197] Examples of the test compound to be applied to the screening system of the present invention include, but are not particularly limited to, synthetic low molecular compound libraries, expression products of gene libraries, peptide libraries, siRNA, antibodies, substances released by bacteria, extracts and culture supernatants of cells (microbes, plant cells, and animal cells), purified or partially purified polypeptides, marine organisms, plant- or animal-derived extracts, and random phage peptide display libraries.

[0198] A subject with cancer may display at least one identifiable sign, symptom, or laboratory finding that is sufficient to make a diagnosis of cancer in accordance with clinical standards known in the art. Examples of such clinical standards can be found in textbooks of medicine such as Harrison’s Principles of Internal Medicine, 15th Ed., Fauci AS et al., eds., McGraw- Hill, New York, 2001 . In some instances, a diagnosis of a cancer in an individual may include identification of a particular cell type (e.g. a cancer cell) in a sample of a body fluid or tissue obtained from the individual.

[0199] The subject may have minimal residual disease (MRD) after an initial cancer treatment. The term “treatment”, as used herein in the context of treating a condition, pertains generally to treatment and therapy in which some desired therapeutic effect is achieved, for example, the inhibition of the progress of the condition, and includes a reduction in the rate of progress, a halt in the rate of progress and amelioration of the condition, and cure of the condition.

[0200] Treatment may be any treatment and therapy, whether of a human or an animal (e.g. in veterinary applications), in which some desired therapeutic effect is achieved, for example, the inhibition or delay of the progress of the condition, and includes a reduction in the rate of progress, a halt in the rate of progress, amelioration of the condition, cure or remission (whether partial or total) of the condition, preventing, delaying, abating or arresting one or more symptoms and / or signs of the condition or prolonging survival of a subject or patient beyond that expected in the absence of treatment.

[0201] Treatment as a prophylactic measure (i.e. prophylaxis) is also included. For example, an individual susceptible to or at risk of the occurrence or re-occurrence of cancer may be treated as described herein. Such treatment may prevent or delay the occurrence or reoccurrence of cancer in the individual.

[0202] In particular, treatment may include inhibiting cancer growth, including complete cancer remission, and / or inhibiting cancer metastasis. Cancer growth generally refers to any one of a number of indices that indicate change within the cancer to a more developed form. Thus, indices for measuring an inhibition of cancer growth include a decrease in cancer cell survival, a decrease in tumor volume or morphology (for example, as determined using computed tomographic (CT), sonography, or other imaging method), a delayed tumor growth, a destruction of tumor vasculature, or improved performance in delayed hypersensitivity skin test. In the present invention, “cancer” is not particularly limited, and every cancer can be targeted. Examples of cancer include, but are not limited to, acute myeloid leukaemia, adrenocortical carcinoma, bladder cancer, e.g. bladder urothelial carcinoma, brain cancer, such as brain lower grade glioma, breast cancer, such as breast invasive carcinoma, cervical cancer such as cervical squamous cell carcinoma and endocervical adenocarcinoma, lung cancer, prostate cancer, colorectal (colon) cancer, colon adenocarcinoma, esophageal carcinoma, glioblastoma multiforme, skin cancer, melanoma, skin cutaneous melanoma; pancreatic cancer, e.g. pancreatic adenocarcinoma, ovarian cancer, e.g. ovarian serous cystadenocarcinoma, leukemia, lymphoma, liver cancer, e.g liver hepatocellular carcinoma, lung cancer, e.g. lung adenocarcinoma or lung squamous cell carcinoma; Lymphoid Neoplasm Diffuse Large B-cell Lymphoma, kidney (renal) cancer, e.g. kidney Chromophobe, kidney renal papillary cell carcinoma, or kidney renal clear cell carcinoma, mesothelioma; stomach (gastric) cancer, e.g. stomach adenocarcinoma, thyroid cancer, e.g. thyroid carcinoma, testicular cancer, bone cancer, sarcoma, multiple myeloma, gallbladder cancer, e.g. Testicular Germ Cell Tumors, head and neck cancers (e.g., oral, throat, laryngeal cancers, such as head and neck squamous cell carcinoma,), glioblastoma, neuroblastoma, uterine (endometrial) cancer, e.g. uterine Carcinosarcoma or uterine corpus endometrial carcinoma, anal cancer, rectum adenocarcinoma, prostate cancer, e.g. prostate adenocarcinoma, sarcoma, thymoma, vulvar cancer, penile cancer, intestinal cancer, blood cancer, eye (ocular) cancer, retinoblastoma, and uveal melanoma.

[0203] ***

[0204] The features disclosed in the foregoing description, or in the following claims, or in the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, or a method or process for obtaining the disclosed results, as appropriate, may, separately, or in any combination of such features, be utilised for realising the invention in diverse forms thereof.

[0205] While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

[0206] For the avoidance of any doubt, any theoretical explanations provided herein are provided for the purposes of improving the understanding of a reader. The inventors do not wish to be bound by any of these theoretical explanations.

[0207] Any section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

[0208] Throughout this specification, including the claims which follow, unless the context requires otherwise, the word “comprise” and “include”, and variations such as “comprises”, “comprising”, and “including” will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

[0209] It must be noted that, as used in the specification and the appended claims, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Ranges may be expressed herein as from “about” one particular value, and / or to “about” another particular value. When such a range is expressed, another embodiment includes from the one particular value and / or to the other particular value. Similarly, when values are expressed as approximations, by the use of the antecedent “about,” it will be understood that the particular value forms another embodiment. The term “about” in relation to a numerical value is optional and means for example + / - 10%.

[0210] Examples

[0211] An approach was used that agnostically interrogates uveal melanoma genomes to identify genetic vulnerabilities for therapeutic targets. A comprehensive CRISPR KO screening of 10 human uveal cell lines was performed, using genome-wide and combinatorial libraries to map monogenetic and digenetic dependencies. Synthetic lethal interactions were identified by interrogating 568 gene pairs and applying separate genome-wide CRISPR KO libraries on the same cell lines to identify single gene essentialities. This identified CDS2 as a novel genetic dependency in the context of low CDS1 expression in vitro and in vivo (Example 1 ). Example 2 shows the results of an unbiased computation analysis to identify cancer-specific SLI, using public genomic and gene expression resources, further supporting the strong SLI between CDS1 and CDS2 in a variety of cancers with low CDS1 expression.

[0212] Detailed method information is provided below these results.

[0213] EXAMPLE 1 - Synthetic lethal interaction between CDS1 and CDS2 in uveal melanoma

[0214] The present inventors collected, validated, whole genome sequenced and proteome profiled a collection of 10 uveal cell line models

[0215] Combinatorial CRISPR library design and screening.

[0216] A total of 25,499 pgRNAs targeting 568 gene pairs, 169 single control genes and safe targeting regions were included in a combinatorial CRISPR library. Three main categories of gene pairs were selected for inclusion in the CRISPR library, namely uveal melanoma gene pairs, paralogues, and putative synthetically lethal genes(Fig. 1 A). The library consisted primarily of paralogous gene pairs (293) and putative synthetic lethal gene pair targets (225). Paralogue pairs were selected such that they had a minimum protein sequence identity of 45% (Ensembl v92). Those with single fly (Drosophila melanogaster) and worm (Caenorhabditis elegans) orthologues, each with at least one lethal phenotype (FlyMine, FB2015_15; WormMine, WS251 ) were included in the library.

[0217] Additional paralogous / non-paralogous putative synthetic lethal gene pairs were identified by linear modelling and data integration with MASH-up

[0045] using pan-cancer TCGA (The Cancer Genome Atlas) expression

[0046] and copy number data together with data from loss- of-function RNAi-based screens and copy number profiles from the CCLE cancer cell line datase

[0020] . To further systematically identify associations between gene expression level and essentiality to prioritize candidate pairs where low expression of one gene predicted essentiality for another, CRISPR-cas9 and RNA-Seq data sets from Project Score [47; 48] were used. These two approaches identified 125 and 100 gene pairs of the abovementioned 225, respectively. In addition to these collections, we also included a subset of gene pairs relevant to uveal melanoma (50), with one member of the pair known to have a gene expression profile predictive of uveal melanoma outcome and where low expression of this gene has been observed in some tumours, while the other gene in the pair was either a paralogous gene or a candidate synthetic lethal gene based upon the database of synthetic lethal interactions (SynLethDB) [21 ,22, 23], a database of synthetic lethality based collated from computational predictions and text mining, or identified from mutual exclusivity analysis of TCGA uveal melanoma genome data. Of note, this group of genes included drivers for uveal melanoma such as KIT and BAP1 paired with other genes. Ten safe-targeting guides (STGs), which had previously been designed against regions of the genome with no annotated function, were also used in the library (data not shown). These STGs were paired with control essential and non-essential single genes to allow technical assessment of screen performance, and also with sgRNAs targeting a gene in a pair to allow the single gene knockout effect to be computed in comparison to co-disruption of both members of a pair.

[0218] Figure 1 B shows the design of the paired guide (pg)RNA. The pgRNA construct was designed with each sgRNA placed under an independent promoter (human U6 or mouse U6) and a modified tracrRNA scaffold for one of the sgRNAs to reduce recombination (Figure 1 b). To compensate for any possible promoter bias, an equal number of guides for each gene were positioned behind each promoter, selecting 8 individual sgRNAs per gene. This resulted in a total of 32 unique sgRNA pairings for every gene pair as shown in Figure 1 C. Genes with fewer than 6 available unique sgRNAs were excluded from the library.

[0219] Ten human uveal melanoma cell lines were engineered to express Cas9 and had >85% Cas9 activity prior to CRISPR screening using a reporter assay. Screening was performed in technical triplicate and a minimum library representation of 3000x was maintained throughout the screen. Cells were harvested at 28 days post-transduction and at 20 cell doublings. Non-Cas9 cell lines were infected with the CRISPR library and harvested at day 7 and used as a baseline to compare gRNA abundances with. As part of the assessment of screen performance, the correlation between replicates within each cell line with the median being 0.838 was computed (data not shown). When the dropout of control essential and non- essential single genes paired with STGs was compared, a strong depletion of essential genes was observed (median Iog2 fold change (FC) -0.31 ) with little dependency on non- essential genes (median log2FC -0.07) across all cell lines and area under the curve (AUC) values in excess of 0.98 (see Figure 1 D). It was possible to recall the essential genes for all cell lines using BAGEL, a gene essentiality classifier

[0050] . The log2FC of all double knockout pgRNAs targeting gene pairs were analysed and compared these values with the log2FC of single gene knockouts where vectors contained a gRNA targeting a gene paired with a STG. This analysis revealed significantly lower log2FCs within the double knockout gene pair group (p < 2.2e-16, one-tailed K-S test), which indicates the generally more deleterious effect of disrupting two genes in these cells and strongly suggest a synthetic lethal effect on cell fitness for interacting gene pairs.

[0220] Identification of synthetic lethal interactions

[0221] To identify the abovementioned synergistic gene pairs, the Bliss independence model

[0049] was used, which predicts the dual knockout FC to equal the sum of the individual knockout FCs.

[0222] This allowed for the quantification of the difference between the predicted and observed FC as the genetic interaction (Gl) score (See Figure 1 E). For each cell line, a gene pair was considered a ‘hit’ if the mean normalised Gl score was less than -0.5 and the false discovery rate (FDR) below 0.01 across all guide pairs and replicates (data not shown). At this threshold, we identified 105 unique gene pairs (18.5% of the library gene pairs) that were significantly depleted in at least one of the 10 screened cell lines (Data not shown). The majority of hits were reproduced in at least two cell lines (62.9%), indicating that there were relatively few hits that were cell line-specific in the screen. The analysis then focussed on significant digenic interactions that were shared amongst at least 6 of our 10 screened cell lines, which allowed us to consider the top 22.0% (23 / 105) of the total gene pair hits across all cell lines (See Figure 1 F)

[0223] BAGEL called CDS2 as essential in every cell line and time point, and CDS1 as non- essential i.e. not called as a hit in any cell line and timepoint with no greater than additive effect on survival (Figure 1G).

[0224] CDS1 / CDS2 synthetic lethal interaction

[0225] CDS1 gene expression was low in uveal melanomas

[0051] (median: 6.1 , range: 2.9-9.1 ; log2(TPM+1 )) when compared with CDS2 gene expression (median: 10.8, range: 10.0-12.4; log2(TPM+1 )). To determine whether this expression pattern could be observed in other cancers, TCGA pan-cancer gene expression obtained from the UCSC Xena platform was analysed

[0052] , Whilst CDS2 was widely expressed amongst all cancer subtypes analysed, CDS1 expression was low or absent in cases of cutaneous melanoma, glioblastoma, haematological malignancies, hepatocellular carcinoma and sarcoma (see Figures 2A and B), suggesting that targeting CDS2 in these cancers might be a viable therapeutic strategy.

[0226] To determine whether the CDS1 / CDS2 synthetic lethal interaction was observed in other tumour types, the analysis of CDS2 gene essentiality was extended to include 316 cell lines from Project Score

[0053] and 937 cell lines from Project Achilles (DepMap 22Q2)

[0032] . Any uveal melanoma cell lines present within the data sets were excluded from the analysis to ensure that any observed effect was not driven by uveal melanoma alone. Importantly, CDS2 was significantly more essential in CDS1 low-expressers (log2(TPM+1) < 1) in both Project Score (p=2.52e-11 ) and Project Achilles (p=8.95e-109), consistent with a synthetic lethal relationship between CDS1 and CDS2 (Figure 3A). Analysis of methylation data from these cell lines (DepMap 23Q2) revealed promoter methylation in cell lines that had low CDS1 expression (Figure 3B), suggesting a probable mechanism of gene silencing.

[0227] Genome-wide CRISPR KO screen to identify single gene essentialities

[0228] Genome-wide sgRNA CRISPR KO screens were performed in ten uveal melanoma cell lines (92.1 , Mel202, Mel270, Mel285, MP38, MP41 , MP46, 0MM2.1 , OMM2.3 and OMM2.5) that had been engineered to express Cas9 to (1 ) to identify additional uveal melanoma-specific dependencies of single genes that were not assessed in the combinatorial CRISPR library;

[0229] (2) to further examine digenic paralogue CRISPR dependencies by perturbing each member within a paralogue family and integrating these data with of other paralogue family members;

[0230] (3) as orthogonal / complementary validation of the combinatorial CRISPR screen results. The screens were performed in technical triplicate per cell line, where each was transduced with the Human Improved Genome-wide Knockout CRISPR Library v1 .1 , cultured for 14 days and harvested to for gRNA quantification. We excluded known essential genes and filtered for genes that had a MAGeCK MLE beta score of < -0.5 and FDR < 0.05. This analysis identified 171 previously undefined essential genes across 10 cell lines, including 28 genes that were essential in 7 or more cell lines (Figures 4A-B).

[0231] There were a total of 1447 significantly depleted genes that were identified from the analysis, of which 834 were known Achilles essential genes. When we focused on genes that were significantly depleted in at least 9 out of the 10 screened cell lines, 77 genes that were not control essential genes were identified. After excluding DepMap 23Q2 CRISPR Inferred Common essentials that were essential in nearly all cell lines, we identified CDS2, POLR2A, H2AC14, MITFand SOX10 as essential genes in uveal melanoma (Figure 4B).

[0232] The results of the screen were genetically validated by selecting 3 dropout genes and a control gene (STG2) that did not drop out. Two gRNAs were selected against each gene, one from the original library used in the screen and another from Human MinLibCas9. The results of the genetic validation at day 14 were concordant with the dropout screen results. When we continued the validation experiment to day 28, we found that ITGB5 and ZEB2 became significantly depleted in the tested cell lines. This may indicate that some genes require longer before dropout. However, it was evident that the effect of CDS2 dropout was visible from day 14 for all cell lines (See Figure 4C). For further validation experiments we engineered doxycycline-inducible CDS2 KO clones. Using these clones clonogenic and apoptosis assays were performed and demonstrated a reduced colony formation ability and increased apoptosis when CDS1 and CD2 are lost in these cell lines.

[0233] Mechanistic insights into the effects of CDS2 loss

[0234] Multiple clonal cell lines containing a doxycycline-inducible sgRNA against CDS2 to test the effect of CDS2 loss and to model the dynamics of the synthetic lethal interaction with CDS1 were generated. To do this, a CDS2 gRNA was placed under the control of a doxycycline- inducible promoter, in cell lines that constitutively expressed Cas9. In colony-forming assays, doxycycline-treatment of MP41 and OMM2.5 cell lines which express low CDS1 , resulted in fewer and smaller colonies than isogenic DMSO- treated control populations (Figure 5A). Of note, in apoptosis assays we observed a significant increase in apoptotic (Annexin V+ / DAPI- ) and dead cells (Annexin V+ / DAPI+) following the induction of CDS2 sgRNA expression, suggesting that CDS1 / 2 loss results in the induction of cell death rather than cytostasis (Figure 5B).

[0235] Figure 8 shows the role of CDS1 and CDS2 in phosphoinositide synthesis and recycling, and the role of phosphoinositides in downstream signalling pathways in uveal melanoma. LC-MS was used to quantify phosphoinositides (PIPn) and their metabolic precursor PA following treatment with either doxycycline or DMSO control of a 7-day time course. We observed that following disruption of CDS2 there was a progressive reduction in the phosphoinositides PI, PIP and PIP2. These PIPn which are normally involved in downstream pathways including PI3K-Akt and MAPK no longer drive these pathways (Figure 6).

[0236] In particular, uveal melanomas that have chronic PLC signalling become affected which could result in a loss of cellular fitness of the cancer cells. The corresponding increase in PA, which serves as the substrate for CDS enzymes, is redirected to a different pathway resulting in TAG production and the formation of lipid droplets. We observed supersized Lipid Droplets (LDs) that are induced by the deletion of CDS2 (Figure 9). This effect has been previously reported (often revealed by overloading the cells with oleic acid), but in the uveal cells it seems particularly impressive, probably indicating that the cells are generating an excess of TAG (by diverting the precursors not used for PI synthesis), which keeps accumulating in these fat stores.

[0237] Together, these results indicate that CDS2 loss limits phosphoinositide availability and has a profound effect on cancer cell fitness.

[0238] To gain additional mechanistic insights we next attempted to identify genes whose disruption could rescue the lethality observed following CDS2 loss, reasoning that any such genes would represent possible paths to drug resistance in patients treated with an agent that inhibited / depleted CDS2. To do this we conducted a genetic suppressor screen using the MinLibCas9 library in an MP41 -derived clonal cell line carrying a doxycycline-inducibleCDS2 gRNA allele (See Figure 7; see Methods). Importantly, we were unable to identify any genes under significant positive selection at an FDR below 0.05, suggesting no escape to loss of CDS2 in uveal melanoma (data not shown). This is notable as it suggests that an inhibitor of CDS2 may provide effective and sustained anti-tumour activity with minimal bypass / resistance mechanisms.

[0239] Synthetic lethal interaction of CDS1 / CDS2 in vivo

[0240] \Ne next assessed whether the effect of CDS2 loss on cell proliferation extends into in vivo setting grafting the OMM2.5 clonal cell line containing the doxycycline-inducible CDS2 gRNA construct into mice (see Figure 13A). CDS2 disruption resulted in in a significant reduction in uveal melanoma cell expansion and prolonged survival. Mice that were xenografted with the Dox-inducible line and fed a Dox-containing diet (625 mg / kg; ENVIGO) had significantly reduced tumour growth compared with controls were fed a normal diet not containing doxycycline. The results of this experiment are shown in Figure 13B. Of note, compared with control mice fed a normal diet, we saw no evidence of resistance / rebound tumour growth even up to 48 days after the initiation of the experiment suggesting that within this timeframe resistance was not acquired.

[0241] Together, these data indicate that CDS2 loss affects cancer cell fitness.

[0242] Genetic suppressor screen to identify genes involved in rescue from CDS2 loss

[0243] To identify genes whose disruption could rescue the lethality following CDS2 loss, and thus represent a path to drug resistance in patients treated with an agent that inhibited or depleted CDS2 a genetic suppressor screen was conducted using the minimal genome-wide human CRISPR-Cas9 library (MinLibCas9) in an MP41 -derived clonal cell line carrying a doxycycline-inducible CDS2 gRNA. The genes identified by enriched sgRNA representation within the CDS2 knockout cells are shown in Figure 7B. This is notable as it suggests that an inhibitor of CDS2 may provide effective and sustained anti-tumour activity with minimal bypass / resistance mechanisms, if any.

[0244] Testing in other cancer types

[0245] We performed a dual guide CRISPR screen in cell lines from NSCLC, pancreatic cancer or cutaneous melanoma cell lineage.

[0246] Data were analysed using Bassik analysis

[0063] , the mean normalised Gl score here refers to the residual i.e. the difference expected when summing the median of the single gene knockout for CDS1 and CDS2, compared with what is actually observed when using a dual guide targeting both simultaneously.

[0247] A mean normalised Gl <-0.5, FDR <0.01 and neither gene is single essential (calculated by BAGEL and MAGeCK analysis) was considered to indicate a synthetic lethal.

[0248] Figure 10 shows published cell line expression for CDS1 against the mean normalised Gl. Dark symbols correspond to non-synthetic lethal cell lines, while light corresponds to those which were considered to have CDS1 / 2 synthetic lethal. The shape of the symbol represents the cancer types cell background. We found CDS1 and CDS2 to be synthetic lethal in cells from lung or pancreatic cancer backgrounds.

[0249] Figure 11 shows the same information but dark and light symbols indicate whether CDS2 was essential or non-essential respectively.

[0250] Further analysis of other cell lines to test for CDS1 / 2 synthetic lethality was performed, similar to above. The residual on the x axis is the additional detrimental effect on cellular fitness not accounted for by the sum of the two genes individually, and this is significant in 9 out of the 9 screened cell lines. The following cell lines showed that they would be suitable for CDS1 / 2 synthetic lethal treatment:

[0251] C2BBe1 - Colorectal adenocarcinoma

[0252] COLO678 - Colon carcinoma

[0253] SW837 - Colorectal adenocarcinoma

[0254] NUGC3 - Gastric adenocarcinoma

[0255] AGS - Gastric adenocarcinoma

[0256] GCIY - Gastric adenocarcinoma

[0257] 769P - Renal cell carcinoma

[0258] RCCFG2 - Clear cell renal cell carcinoma (kidney cancer) HAP1 - chronic myelogenous leukemia

[0259] This is shown in Figure 12A and 12B.

[0260] Figure 12C shows that CDS1 / CDS2 synthetic lethality is observed across many different tumour cell types, including kidney cancer, ovarian cancer, and prostate cancer. Together, these data show that CDS1 and CDS2 have a synthetically lethal relationship in a variety of cancers.

[0261] Detailed method procedures

[0262] Cell Culture

[0263] Human uveal melanoma cell lines were authenticated by short tandem repeat profiling and screened negative for mycoplasma. Cell lines were cultured in RPMI 1640 or DMEM and supplemented with fetal bovine serum. All cell lines underwent lentiviral transduction with pKLV2-EF1 aBsd2ACas9-W (Addgene, #68343) to generate Cas9-expressing cell lines. Cas9 activity was confirmed to be >85% for all cell lines prior to screening using a BFP / GFP reporter assay1. All Cas9 expressing cell lines were regularly blasticidin-selected to maintain Cas9 activity

[0264] Combinatorial CRISPR library design and construction

[0265] The combinatorial CRISPR gRNA library was constructed using to the protocol available at Protocols. io (dx.doi.org / 10.17504 / protocols.io.bpqhmmt6) based on the method established by Vidigal and Ventura [2], gRNAs from published genome-wide CRISPR knockout libraries were used to select eight gRNAs per gene, comprising four gRNAs from the Brunello library [3] and four gRNAs from the Toronto KnockOut library [4], If there were fewer than eight original gRNAs available from these two libraries, the remaining gRNAs were selected from the Sabatini / Lander library5or designed using the Genetic Perturbation Platform (Broad Institute). Each gene pair had a total of 48 unique gRNA combinations (Fig 1 C), of which 16 comprised pairing a gRNA with a safe-targeting guide. Eight previously published safetargeting guides (STGs) designed against genomic regions with no previously defined function across 127 cell lines [6] were selected and validated by performing a surveyor assay using the Surveyor Mutation Detection Kit (IDT), and served as negative controls.

[0266] Lentivirus production was carried out using 7.2 x 107293T cells transfected with 24 pg VSV- G, 40 pg pMDLg / pRRE, 20 pg pRSV-Rev, 67.2 pg pAdvantage and 69.6 ug of library plasmid. Media was replaced 16 hours after transduction and viral supernatant harvested after 48 hours, filtered through a 0.45 pM membrane (Merck, SLHP033RS), and stored at - 80°C.

[0267] Combinatorial gRNA CRISPR screening

[0268] For each cell line, 9.0 x107cells were transduced with the combinatorial CRISPR library at the required volume to achieve a transduction efficiency of 30% (1000x library coverage). Puromycin selection (2-4 pg / mL) was performed at day 3 post-transduction for a total of 7 days. Screening was performed in technical triplicate and a minimum library representation of 3000x was maintained throughout the screen. Cells were harvested at two timepoints to account for the wide range of doubling times between cell lines. Cells were first harvested at 28 days post-transduction and additionally at 20 cell doublings. Cas9-negative cell lines were transduced under the same conditions and harvested at day 7 to allow calculation of gRNA fold change. Genomic DNA was extracted for each replicate from cell pellets using a Blood & Cell Culture DNA Maxi Kit (Qiagen, 13362). PCR amplification and sequencing was performed as previously described [7],

[0269] Analysis of the combinatorial gRNA CRISPR screen

[0270] We used the Bliss independence model to determine whether the observed effect of both genes is greater than the sum effect of both genes individually using the following formula:

[0271] Expected log2FC (glg2) = Observed Iog2FC(gl] + Observed Iog2FC(g2]

[0272] Assessment of CDS1 expression and CDS2 essentiality

[0273] To test the association between CDS1 expression and CDS2 essentiality across all cancer types, gene essentiality data was obtained from Project Score (Sanger Institute) and Project Achilles (Broad Institute). Gene expression data obtained from Cell Model Passports and the DepMap download portal (22Q4). For the Sanger (quantitative) essentiality scores, we used quantile normalized log fold change values corrected using CRISPRcleanR [8]. Binary essentiality calls were available for the Sanger dataset (https: / / binarydepscores.tsv.zip / ). Quantitative essentiality data (copy number corrected Chronos (Dempster et al, 2021) scores) for the Broad Dataset was downloaded from the DepMap portal (22Q2 release). Since no binary essentiality calls were available, a cut-off of -0.6 was used to annotate CDS2 essentiality. Cell lines with CDS1 expression log2(TPM+1) < 1 were annotated as low- expressers, and the rest were annotated as high-expressers. Single group data were assessed using Student's t test.

[0274] Whole-genome single gRNA CRISPR knockout screening

[0275] Genome wide single gRNA CRISPR knockout screening was performed using the Human Improved Genome-wide Knockout CRISPR Library v1 .1 . This library was based upon the previously published Human Improved Genome-wide Knockout CRISPR Library v1.0 (Addgene, 67989). It contains 5 sgRNAs targeting 1 ,876 additional genes that encode predefined fitness genes and genes that encode epigenetic-related proteins and kinases, as well as an additional 1 ,004 non-targeting sgRNAs. For each cell line, 3.3 x 107cells were transduced with the required volume of Human CRISPR Library v1 .1 [9] to achieve a transduction efficiency of 30% (100x library coverage). Puromycin selection (2-4 ug / mL) was performed at day 3 post-transduction for a total of 7 days. Screening was performed in technical triplicate and cells were cultured throughout the screen at a minimum library representation of 500x. Cells were harvested 14 days post-transduction, pelleted and stored at -80°C. Genomic DNA was extracted from cell pellets using a Blood & Cell Culture DNA Maxi Kit (Qiagen, 13362). PCR amplification and sequencing was performed as previously described[1].

[0276] Single-guide RNA count preprocessing and bias correction

[0277] All cell lines were processed using CRISPRcleanR8v3.0.1 (https: / / github.com / francescojm / CRISPRcleanR). Only sgRNAs in common with the Human CRISPR Library v1 .0 were analysed to prevent unbalanced targeting of essential genes with additional guides in version 1.1. sgRNAs with less than 30 read counts in the plasmid were removed. The remaining sgRNA raw counts were normalised by their total number within their replicate. Depletion and enrichment log fold changes (logFCs) for individual sgRNAs were quantified at the individual replicate level between library plasmid read counts and post-library-transduction read counts. Gene-independent responses to CRISPR-Cas9 targeting were corrected using the default parameters. MAGeCK MLE vO.5.9.5 ( h ttps : / / so u reef o rg e . n et / p roj ects / m ag eck / ) was run on the normalised corrected sgRNAs’ treatment counts (after applying the inverse transformation on the corrected logFCs previously described[8]).

[0278] Candidate essential genes identified as those with a MAGeCK beta score <1 in at least 8 of the 10 screened cell lines and a FDR <0.05. Previously defined essential genes downloaded from DepMap Public 22Q4 were removed from the candidate essential genes list. These included the intersection of Blomen et al. and Hart et al. essentials. To identify novel essential genes, genes determined as dependencies across all screened cell lines from DepMap 22Q4 (CRISPR Inferred Common Essentials) were also removed.

[0279] Validation by gRNA competitive proliferation assay gRNA competitive proliferation assays were performed using pKLV2-U6gRNA5(Bbsl)- PGKpuro2ABFP-W. To validate individual target genes, one gRNA from the original CRISPR library used in the screens, and a new gRNA derived from MinLibCas9

[0010] was used. Viral supernatants were collected 48 h after transfection and transductions were performed in biological triplicate. Flow cytometry analysis was performed on 6-well plates using a LSRII Flow cyometer (BD) to determine the proportion of BFP+ cells at days 4, 14 and 28.

[0280] Construction of inducible CDS2 knockout isogenic cell line

[0281] A gRNA against CDS2 was cloned into a tet-inducible vector pRSGTEBIeo-U6Tet-sg-EF1 - TetRep-2A-Bleo (Cellecta) according to the manufacturer’s instructions and sent for Sanger sequencing to confirm correct assembly. The vector was transduced into MP41-Cas9 and OMM2.5-Cas9 cells. Each transduced cell line was sorted at single cell density into 96 well plates containing media and zeocin selected for 4 weeks. To confirm editing of the CDS2 gene, genomic DNA was extracted from colonies using the Gentra Puregene Blood Kit (QIAGEN) and sent for Sanger sequencing. Western blot analysis was performed on CDS2 knockout clones using an anti-CDS2 antibody (Abnova H00008760-M01 2B9) and an anti- vinculin antibody (ThermoFisher MA5-11690) as a loading control.

[0282] In vitro CDS2 knockout functional assessment

[0283] Apoptosis assessment

[0284] Apoptosis and cell death were assessed by flow cytometry using Annexin V staining. Dox- inducible CDS2 KO cells were plated in 12-well plates in technical and biological triplicate. A total of 7.5 x 104cells were plated in each well with 1 ml of media. The following day, media was replaced with media containing 0.1 ug / ml doxycycline. At day 7 (check 4,7,10) media and floating and adherent cells were collected from each well. Cells were washed with 500ul of Annexin V Binding Buffer (BD Biosciences) and resuspended in 10Oul Annexin V with Sul PE Annexin V (BioLegend) for 15 minutes at room temperature. DAPI (1 :5000) was added to each sample for 5 minutes and then pelleted and resuspended in 500ul Annexin V Binding Buffer for flow cytometry analysis. Annexin V- / DAPI- cells were deemed to be viable and Annexin V+ / DAPI- cells were considered to be early apoptotic. DAPI+ cells that were considered to be late apoptotic or necrotic were recorded as dead.

[0285] Colony forming ability

[0286] In each 6-well plate, 200 dox-inducible cells were seeded per well. The following day, media was replaced with media containing 0.1 ug / ml doxycycline or the equivalent volume of dimethyl sulfoxide (DMSO). After 14 days colonies were washed with ice cold PBS and 100% ethanol and stained with 1% crystal violet solution.

[0287] Quantification of PA and PIPn

[0288] For PA and PI analysis cells were cultured in 6-well plates at appropriate densities for each time point and treated with 0.1 ug / ml doxycycline or the equivalent volume of DMSO. At days 3, 5 and 7, culture medium was aspirated from each well and the cells were washed in ice- cold PBS before their quenching in 1 ml of 1 M ice-cold HCI. The cells from parallel wells were disaggregated in 1 ml of TriplE and counted to determine the cell number per well at each time point. The cells in the HCI wells were scraped and resuspended, and a volume of the cell suspension containing 7x105 cells was pelleted and frozen. A modified Folch method was used to extract their lipids and PA and PI were analysed using LC-MS as previously described

[0011] . Technical duplicates were averaged for each clone. Data from three MP41 clones were pooled together. Data from OMM2.5 clone P collected in three independent experiments were pooled. PA, PI, PIP and PIP2 ISDs were used for ISD correction. No ISDs were used for PC or PE. PC and PE were analysed in 1 / 1 Oth aliquot of the pre-derivatization organic phase (Folch extraction). To compensate for the heterogeneity between clones and experiments normalization by the area of 34:1 PC and normalization by the average values for the DMSO treatment at times 3, 5 and 7 was performed.

[0289] Genome-wide CRISPR suppressor screen

[0290] To identify genes whose deletion rescues lethality following CDS2 loss, a genetic suppressor screen was performed using the minimal genome-wide human CRISPR-Cas9 library (MinLibCas9) in a clonal population of doxycycline-inducible CDS2 KO in MP41 . The genes identified by enriched sgRNA representation within the CDS2 knockout cells were subsequently validated. Infections were carried out as described in. Media containing 8 ug / ml polybrene and the pooled CRISPR library were added to MP41 clone at a representation of 1000 cells per sgRNA with an approximate MOI of 0.3 as determined by FACS. Following puromycin selection, cells were split and treated with either 0 or 0.1 ug / ml of doxycycline, replenished every 3 days. Approximately 500 cells per sgRNA was maintained at each passage for the duration of the screen. Following 21 days, genomic DNA was extracted using and PCR amplified and sequenced. sgRNAs with less than 30 read counts across all samples were removed. To account for between sample sequencing depth biases, total normalisation was performed on raw sample counts by adding a pseudocount of 5 and normalising to 10 million reads. Log2FCs were calculated from filtered and noramlised counts, between DMSO-treated read counts and dox-treated read counts. MAGeCK vO.5.9.3 was run to identify genes significantly under positive selection at a FDR threshold of < 0.05. Genes that had guides which were a hit in two or more replicates and have a meanLFC > 6 were selected for further analysis.

[0291] In vivo assessment of CDS2 essentiality in OMM2.5 xenografts

[0292] A total of 35 female 8 week old NXG mice (SCID, Cat: SM-NXG-F) were subcutaneously injected with 2 x 106OMM2.5 clone P cells (in 50 pl PBS and 50ul Matrigel) into the right flank. All mice were fed with regular laboratory SAFE 105 chow diet (Safe Diets, Augy, France) until tumours reached 30mm2. Mice were randomly assigned to receive SAFE 105 or a Doxycycline diet (Dox; 625 mg / kg; Envigo, TD.01306) for the duration of the study. Tumours were measured every 24 hours. EXAMPLE 2 - Synthetic lethal interaction between CDS1 and CDS2 in mesenchymal cancers

[0293] CDS1 and CDS2 constitute a synthetic lethal gene pair across cancer types

[0294] In addition to genomic alterations, cancers have also distinct gene expression profiles relative to healthy tissue. Publicly available datasets provide extensive transcriptional information as well as genetic dependencies in cancer. For example, the cancer dependency map (DepMap) is a comprehensive research initiative providing matched CRISPR screening and gene expression data for 913 cell lines across 30 cancer lineages (2021 Q2) [32; 33].

[0295] Additionally, the Cancer Genome Atlas (TCGA) provides genome-wide gene expression data for 9264 tumour and 741 healthy samples

[0062] , Together, these data sets provide approximately 1 e9 datapoints on gene expression and genetic dependencies in cancer and can be mined to predict SLIs.

[0296] To query for SLIs based on differential gene expression between cancer and normal tissue, a bioinformatic pipeline was set up using public datasets on gene dependencies and RNA expression (Figure 14A). This pipeline identifies anchor-target gene pairs, in which the anchor gene shows a relatively reduced expression level in cancer compared to normal tissue, and in which disruption of the target gene results in cancer lethality.

[0297] The cancer dependency map includes additional data that were incorporated into the pipeline to score SLIs, including promotor methylation data, protein expression data, damaging mutation data and cancer type data [32; 33; 34; 35]. RNA expression data provided good power to detect previously established synthetic lethal interactions, including BRCA1-PARP1 and WRN-MLH1 (see Figure 14B, C, D). The inclusion of all cancer types simultaneously resulted in a high-resolution detection of established synthetic lethal interactions compared to analysing specific cancer types separately (Figure 14C). Therefore, the SLI scores for all gene pairs with RNA expression data for all cancer types were determined simultaneously.

[0298] The CDS1 -CDS2 and NAA10-NAA11 gene pairs received high SLI scores (Figures 14D-F). However, the NAA10-NAA11 SLI failed to show cancer specificity and was therefore not pursued. In contrast, the CDS1 -CDS2 SLI demonstrated significant cancer specificity.

[0299] The analysis confirmed FAM50A-B, DDX3X-DDX3Y and EIF1 AX-EIF1 AY as cancer-specific SLIs, in agreement with previous reports (Figure 14D, F). Similarly, PARP-associated synthetic lethality was demonstrated for breast cancer cell lines (Figure 14G, left panel). For several other SLIs, cancer specificity was previously inferred from genomic data but did not score here as such, for example WRN-MLH1 . This is conceivably due to the relatively low frequency of genomic alterations in MLH1 coupled to our focus on transcriptomic rather than genomic data.

[0300] In contrast, CDS2 loss was associated with lethality in a large number of cancer types (Figure 14G, right panel). Several of these concerned common cancers, including lung, blood, brain and skin cancer (Figure 14H). From here onwards, lethality as a function of CDS2 loss will be referred to as ACDS2 lethality. Patients with CDS'! -low cancers showed significantly worse survival compared to CDS1 -high cancer patients (Figure 141).

[0301] To confirm and quantify cancer-specific loss of CDS1 expression, pan-cancer DepMap expression data and pan-tissue GTEx expression data was compared (Figure 14J). The data was calibrated on housekeeping genes, while control analyses on reference genes were performed to determine reliability. In line with the TCGA data (Figure 14D), we observed common absence, or low expression, of CDS1 in cancer cell lines compared to healthy tissue (64% average reduction, p-value<0.0001 ). As expected, given the SLI, low or no CDS1 expression strongly correlated with ACDS2 lethality in cancer cell lines (90% average reduction, p-value<0.0001). We also observed that lung, blood, brain and skin cancers have reduced CDS1 levels compared to their lineages of origin (Figure 14K)Asuggesting that these cancers may suppress transcription of CDS1 during cancer development. For lung cancer, differential CDS1 levels were confirmed using a cohort comprising patient-matched proteomics data from healthy and tumour tissue (Figure 14L). These in silico analyses predict that the CDS1-CDS2gene pair constitutes a human cancer-associated SLI, specifically in common CDS1-\ow cancers.

[0302] CDS1 and CDS2 constitute a synthetic lethal gene pair across cancer types

[0303] A panel of cancer cell lines which are also in the DepMap were used for wet lab validation of computer predictions (Figure 15A). Their CDS1 RNA levels were confirmed by qPCR analysis (Figure 15B). Low throughput CRISPR perturbations were used to quantify the lethality inferred from genome-wide CRISPR knockout screens. By including fluorescent tracker cells serving as an internal control for each experimental condition, we quantified lethality over extended timeframes, showing minimal variation (Figure 15C).

[0304] Upon perturbation of CDS2 with one of two separate sgRNAs, we were able to confirm ACDS2 lethality in CDS1-low or CDS1 -negative cell lines across several cancer types (Figure 15D-F). Furthermore, as predicted, a CDS1-high cell line was ACDS2 non-lethal. Compared to the DepMap data, we observed remarkably strong lethality. Furthermore, ACDS2 lethality was still evident in cell lines with low levels of CDS1 RNA (Figure 15D). To validate the CDS1 dependency for the ACDS2 lethality, CDS1 was either ectopically expressed in CDS1 -negative cancer cell lines or perturbed CDS1 in CDS1 -proficient cancer cells. When CDS1 was introduced into two CDS1-negative cell lines, the ACDS2 lethality was largely negated. Conversely, when CDS1 was perturbed in a CDS1 -high cell line it exhibited increased ACDS2 lethality (Figure 15G). To validate these observations, control GFP-expressing cancer cell lines were admixed with CDS1 -restored cell lines in an additional panel of four human melanoma models. The results confirmed strong synthetic lethality upon perturbation of CDS2 in this panel (Figure 15H and Figure 151). ACDS2 lethality was confirmed in an additional cell panel (Figure 15J). Together, these functional experiments confirm that CDS1 and CDS2 constitute a synthetic lethal pair across a panel of cancer cell lines in vitro.

[0305] ACDS2 lethality is associated with apoptosis

[0306] To investigate whether ACDS2 results in apoptotic cell death in vitro tumour samples were collected and cleaved caspase-3 was measured by quantitative western blotting as a measure of apoptosis (Figure 16). The cleaved caspase-3 signal was quantified by comparison to the total protein signal in the same capillary. As a positive control, cleaved caspase 3 levels were analyzed upon induction of apoptosis by TPCA-1 +TNF (BLM) or staurosporine (SK-MEL-2). ACDS2 cancer cells showed a significant increase in cleaved caspase-3 in cancer cells, indicating that ACDS2 lethality is associated with tumor cell apoptosis.

[0307] ACDS2 lethality in vivo

[0308] To investigate whether the SLI between CDS1 and CDS2 observed in silico and in vitro can be recapitulated in vivo, control GFP-expressing cancer cell lines were admixed with CDS1- restored cell lines. Tumour cells were inoculated into immunodeficient NOD-Scid IL2Rgnull mice and analysed by flow cytometry of the tumours 17 days later. A striking inability of CDS2-perturbed cells to contribute to tumour formation in vivo was observed (Figure 17A-B), indicating that inhibition of CDS2 reduces tumour formation. Together, these results demonstrate that CDS1 and CDS2 form a robust synthetic lethal pair in cancer, in silico, in vitro and in vivo.

[0309] No common escape mechanism for ACDS2 lethality

[0310] The computational and in vitro and in vivo functional validation data above demonstrate the broad cancer range and reproducibility of the CDS1-2 SLI. To investigate whether any cells can rewire their signalling network such that they can escape from this SLI, unbiased, genome-wide CRISPR knockout rescue screens were carried out in a panel of four CDS1 - negative human cancer cell lines and, as a control, one CDS1-high cancer cell line (Fig. 18A). In parallel, cells from the screens were used to track ACDS2 lethality during the screen. These analyses confirmed ACDS2 lethality in four CDS1-negative cancer cell lines and extended our data on the lack of ACDS2 lethality in CDS'! -high cancer cell lines to an additional cancer cell line (Figure 18B). In addition, colony formation assays were performed to visualize the lethal effect during the screen (Figure 18C - note that for K562 a different readout was used because it is a suspension cell line).

[0311] Analysis of the dropout of essential genes confirmed the high quality of the screens (Figure 18D). Potential escape mechanisms were determined using ACDS2 lethality quantified with tracker cells. The results of the screens (Figure 18E-F) indicated no common escape mechanism of ACDS2 lethality. For example, the screen performed in SK-MEL-2 cells yielded no significant enrichment even after an additional 14 days (32 days in total; Figure 18G), while the other screens yielded only some cell line-specific rescue (Figures 18G-J). These findings suggest that no common escape to the combined loss of CDS1 and CDS2 is possible, which is in agreement with the idea that CDS1 and CDS2 together serve as a bottleneck for PI synthesis.

[0312] Together these data indicate that no common escape mechanism exists for ACDS2 lethality.

[0313] Mesenchymal cancers depend on CDS2 for PI synthesis

[0314] To understand which cancer types show reduced CDS1 expression and, hence, ACDS2 lethality, CDS1 -low cancers were characterised using publicly available data. CDS1 -high cancer cells express high levels of the epithelial marker gene E-cadherin, whereas CDS1- low cancer cell lines instead express mesenchymal markers like ZEB1 , ZEB2 and vimentin (Figure 19A-B). Mechanistically, this is in agreement with the notion that CDS1 expression is suppressed in mesenchymal cancers by the transcription factor ZEB1 , previously reported to bind the CDS1 locus and to suppress CDS1 expression [36,37], Besides this major regulatory mechanism, we also observed that a rare subset of the blood lineage cancer cell lines exhibit methylation of the CDS1 promoter (Figure 19C). Overall, these findings indicate that the suppression of CDS1 expression in mesenchymal-like cancers by ZEB1 results in their strong dependency on CDS2.

[0315] Gene-set enrichment analysis indicated that CDS1-low cancers are enriched for the Hallmark EMT gene-set (p = 0.0001 DepMap, p = 0.009 TCGA, Figure 19D). Mesenchymal- transitioned cancers are common, highly metastatic and therapy-refractory [54-58]. These findings are in agreement with our observations that low CDS1 expression is common, more frequent in cancers compared to their healthy tissue of origin and associated with worse survival (Figure 14G-H; Figure 141, K). Together, these results suggest that suppression of CDS1 expression is an integral element of EMT in cancer. Next, the pathway in which CDS1 and CDS2 are involved was further explored (Figure 19E). We incorporated expression and dependency data from the DepMap of the enzymes involved. In line with the requirement for CDS1 / 2, the next enzyme in the pathway, cytidine diphosphate diacylglycerol synthase inositol-3-phosphatidyltransferase (CDIPT), is highly essential for survival. In addition, CDS2 and CDIPT are significantly co-dependent in DepMap cancer cell lines, which is indicative of a similar mechanism of dependency (Figure 19F). Accordingly, when supplementing the cell culture media with exogenous PI, a significant rescue of cell death was observed, indicating that ACDS2 lethality is due, at least in part, to insufficient availability of PI (Figure 19G and Figure 19H).

[0316] CDS2 acts upstream of the PI3K signaling module regulating growth and survival. Unexpectedly, DepMap analysis revealed that ACDS2 lethality is not accompanied by lethality with either genetic loss of PIK3CA or pharmacologic PIK3CA inhibition (Alpelisib). Instead, a strong negative correlation between ACDS2 lethality and API3KCA lethality was observed (Figure 19F). This negative correlation was independent of PI3K isoforms, as judged by pan-PI3K inhibitor responses. These results demonstrate an essential role of CDS for PI synthesis in mesenchymal cancers. Furthermore, they unexpectedly point to a contribution of CDS1 and CDS2 to survival signalling independent of the classical PI3K pathway.

[0317] Expression of either CDS1 or CDS2 is required for lipid homeostasis

[0318] To biochemically define the effects of CDS2 perturbation in CDS1 -negative cancer cells, multi-omics was performed in a panel of four CDS1 -negative cancer cell lines and a control CDS1 -high cancer cell line (Figure 20A). Lipidomics was performed to quantify buildup or depletion of lipids, specifically those in the CDS pathway. In addition, proteomics was performed to identify deregulated cellular processes upon CDS2 loss.

[0319] Lipidomic analysis allowed for quantification of the major lipid classes in the cells, with phosphorylated PI (PIP) serving as the major signalling molecule (Figure 20B). Major changes as a function of CDS1 expression were detected upon CDS2 loss (Figures 20B-D). For example, cholesterol esters and triglycerides were significantly upregulated.

[0320] Furthermore, a strong buildup of CDS2 substrates and depletion of the downstream product PI was observed.

[0321] The proteomics analysis allowed for quantification of -7000 proteins in each cell line. As expected, the levels of CDS2 protein had dropped 10-fold in all five cell lines upon CDS2 CRISPR perturbation (Figure 20E). In addition, large groups of proteins were commonly and significantly down- or up-regulated in CDS1 -low cancer cell lines, suggesting an orchestrated response (Figure 20E-F). GO-term enrichment analysis of these proteins revealed major induction of cholesterol import and production upon ACDS2 lethality (Figure 20G).

[0322] In line with the build-up of cholesterol esters and triglycerides, CDS2-perturbed cells formed large lipid droplets, which were visible by both light and electron microscopy, staining positively for the lipid dyes BODIPY and Nile Red. Quantification of BODIPY-stained live cells revealed that on average 3% of the imaged cells comprised lipid (Figure 20H). Combined light and electron microscopy of fixed cells showed these lipid droplets in more detail (Figure 201). Together, these findings again corroborate the CDS1 / 2 biochemical bottleneck to support lipid homeostasis, demonstrating that lethality upon CDS2 perturbation in CDS1-low cell lines is accompanied by major changes in lipid metabolism and expression of the cholesterol pathway (Figure 20J).

[0323] CDS dependency is linked to proliferation

[0324] Gene set enrichment analysis of co-expressed genes indicated that ACDS2 lethal cell lines express gene sets related to cell proliferation including E2F targets and Myc target genes (Fig. 22A-B). Accordingly, when we reduced cell division using either the CDK4 / 6 inhibitor Palbociclib or contact inhibition we observed substantial rescue of synthetic lethality in two tumour cell lines Fig. 22A). This further strengthens the association between the CDS1 / 2 SLI and cancer, since increased proliferation is a hallmark of cancer.

[0325] CDS2 enzymatic activity is involved in CDS1 / 2 SLI

[0326] To investigate whether CDS2’s enzymatic activity is involved in the synthetic lethality we rescued with enzymatically inactive CDS2. For this, D384A mutant was generated based on a previous study showing that D249A removes enzymatic activity in bacterial CDSA

[0064] , The D384 / D249 region is conserved between the two proteins in terms of amino acids and ion binding (see below BLAST & AlphaFill) (Figure 23A-B).

[0327] Figure 23C shows that synthetic lethality could not be rescued by cell line variants CDS2- GFP with D384A mutation compared to controls, indicating that enzymatic activity is involved in CDS2 synthetic lethality in CDS1 Io cancer cells.

[0328] Conclusion

[0329] Taken together, these data show that CDS2 serves as a synthetic lethal target in Many cancers, and in particular mesenchymal cancers. The bioinformatic pipeline described above revealed a potential therapeutic index due to differential CDS1 expression in cancer cells (low) and normal cells (high). CDS1 and CDS2 were confirmed to constitute a functional SLI across many cancer types, particularly those of mesenchymal origin, representing approximately 40% of all cancers. The SLI is mechanistically explained by the fact that the combined loss of CDS1 and CDS2 is incompatible with cell survival, in combination with the low or absent expression of CDS1 in mesenchymal cancer cells resulting in ACDS2 lethality. The series of CRISPR knockout rescue screens confirmed not only the synthetic lethal relationship between CDS1 and CDS2, but also that they form a bottleneck in a pathway required for PI synthesis that apparently is essential for cancer cell viability. Mechanistically, ACDS2 lethality causes cancer cell apoptosis, which can be rescued at least in part by supplementing exogenous PI. The latter is consistent with a requirement for CDS2 to produce PI. These data unexpectedly indicate that PI synthesis is required for cell survival independent of classical PI3K-dependent survival signalling, because cell lines that depend on CDS1 / 2 for survival did not require PI3K. These data indicate that CDS2 inhibition would be beneficial to eliminate mesenchymal-like cancers, which are common cancers that are highly metastatic and therapy-refractory.

[0330] Detailed methods

[0331] Cancer-specific synthetic lethality prediction

[0332] Briefly, to rank potential synthetic lethal candidate pairs, the Pearson correlation between fitness data (CERES) and mRNA expression, proteomics, mutation or methylation data was calculated with 21 Q2 DepMap data. Each pair includes the anchor gene (expression gene) and the target gene (fitness effect gene), respectively. To validate this approach, several positive control synthetic lethal pairs were selected. Positive control analyses were run using all pairs with the anchor genes. To identify the most promising new synthetic lethal interactions, the anchored approach was replaced with a genome-wide analysis encompassing 3.4e8 gene pairs. From the results, the top pair for each anchor was ranked to form a hit list. Then, all duplicate target genes were removed to prevent false positives from cross-correlated anchor genes. Reanalyzed TCGA RNA data and healthy or tumor annotation was downloaded from accession GSE6294462. The fold change in expression was calculated for each anchor gene in the synthetic lethality prediction.

[0333] Patient mortality and survival analysis

[0334] TCGA pan cancer atlas RNA and outcome data was downloaded on Januari 12th 2023 from: https: / / gdc.cancer.gov / about-data / publications / pancanatlas

[0335] United States patient mortality data by cancer type was collected on Januari 18th 2023 from: https: / / seer.cancer.Qov / archive / csr / 1975 2017 / results sinqle / sect 01 table.O4 2pqs.pdf 21 Q2 DepMap data for cell lines with available RNA and dependency data was used for ACDS2 lethality. Lineages of patient mortality data and DepMap data were matched manually.

[0336] Comparing CDS1 expression between healthy tissue and cancer cell lines

[0337] GTEx v8 TPM was downloaded from the portal on January 1st 508 2022. 21 Q2 DepMap data was downloaded from the DepMap portal. Calibration genes with available TPM in DepMap and GTEx were used for normalization. Up- and down- regulation controls were selected from the literature. Briefly, transcripts per million for each calibration gene were normalized to their own average expression. Then the TPM for B2M, CDS1 , CDKN2B and E2F1 were normalized to the average of the calibration gene expression. Calibrated results are similar to uncalibrated results due to intrinsic total RNA calibration of the GTEx and DepMap dataset TPM data. The analysis was also performed for blood, skin, lung and brain lineage data with the original labelling from the DepMap and GTEx, also including soft tissue ATRT in the DepMap brain category.

[0338] Healthy and tumor proteome data

[0339] Tumor and normal adjacent lung proteome data from Satpathy et al. was downloaded from entry PDC000234 and analyzed. Specifically, Iog2-transformed ratios to internal controls were collected and sample names were mapped to their healthy or tumor tissue origin. Before plotting and statistical testing Iog2-transformation was undone resulting in ratios to internal control. The internal control was reported to be a mix of all samples. Still, calibration genes with available ratios in all samples were used for normalization. Each calibration gene ratio was normalized to its own average ratio. Then the ratio for CDS1 was normalized to the geometric mean of the normalized calibration gene ratios for each sample.

[0340] Quantitative polymerase chain reaction to quantify RNA

[0341] Three housekeeping genes for calibration were selected from a previously described list of options based on stable pan-cancer cell line expression in the DepMap

[0038] . qPCR was performed as previously described

[0039] . The average of quadruplicate CT values was calculated for each gene. To normalize to calibration genes the geometric mean of the three calibration gene CT values was subtracted from the CDS1 CT value. To normalize to A431 , the CT value of A431 was subtracted from the other cell line CT values. By exp2 CT values were transformed to fold changes.

[0342] CHMP2A: CGCGAGCGACAGAAACTAGAG (Forward primed); CCCGCATCAATACAAACTTGC (Reverse primer) VPS29: TGCAACAGTTTGCCAGCTAAA (Forward primed); CCTCTGCAACAGGGCTAAGC (Reverse primer)

[0343] PSMB2 ATCCTCGACCGATACTACACAC (Forward primer) GAACACTGAAGGTTGGCAGAT (Reverse primer)

[0344] CDS1 CCTTGTCATCCAAAATCTGT (Forward primer) GAATCCTTCCCAAGTCTTTT (Reverse primer)

[0345] Lentiviral transduction

[0346] Cell lines were lentivirally transduced as previously described

[0039] . Transduced cells were always kept on selection during cell culture in vitro. The complete selection was verified by the loss of the control untransduced cells on selection. Low throughput CRISPR perturbation using pLentiCRISPR v2 Puro or Blast and ectopic expression using pLX304 Blast (CDS1 , GFP, CDS2-GFP and CDS2-GFP D384A.) or pCDH Puro (mCherry) was performed as previously described

[0050] . For Fig. 15G Blast resistance was introduced to mCherry- transduced cells using pLentiCRISPR v2 Blast. In Figure 15D sgCDS2 #1 and #2 were used and for all other experiments sgCDS2 #2 was used. In Fig.15G sgCDSI was introduced using pLentiCRISPR v2 Blast. sgCDSI : 5’ GTGTCGCCGCCACACCGCGA 3’ sgCDS2 #1 : 5’ ACCAGGGCAGATCATATGAG 3’ sgCDS2 #2: 5’ AGTAAAGGAAATGAACCGG 3’ sgControl: 5’ AGCTTTCGAAATTGAGTGTC 3’

[0347] Determining (synthetic) lethality by tracking fluorescent cells

[0348] Briefly, after lentiviral transduction (1 day) and puromycin selection of sgRNA-transduced cells (2 days) the indicated cell mixes were prepared. For the in vivo experiments cells were selected one day longer. To determine lethality sgRNA-transduced cells were mixed with non-sgRNA transduced mCherry-positive cells. On the day of mixing (day 0) and at the indicated later time points samples from each cell mix were analyzed by flow cytometry to determine the ratio between negative and positive cells. Final ratios were normalized to day 0 ratios and to sgControl ratios. For lethality, normalized ratios were transformed into percentages to get survival percentages and subtracted from 100 to get lethality percentages For the experiments in Figure 15J and 23C the data was corrected for background signal from mCherry negative cells derived from unmixed mCherry positive control cells. To determine synthetic lethality ectopic CDS1 - or GFP-expressing cells, both transduced with the same sgRNA, were mixed on day 0. Then, lethality was calculated for the GFP-positive cells in an identical manner as for the mCherry-negative cells.

[0349] Flow cytometry

[0350] Flow cytometry data was collected on a BD Fortessa flow cytometer. One of several live / dead dyes was always included: DAPI (D9542-50MG, Sigma), PI (537059-100MG, Merck), DRAQ7 (D15106, Thermofisher) and fixable near IR (L34976, Thermofisher). 0.1% BSA in PBS was used as a washing and staining solution. Samples were washed three times (including resuspension as one wash) in 0.1%BSA in PBS, once for dead cell experiments to preserve dead cells and six times for in vivo samples to reduce postdigestion debris. The gating strategies used are presented in Figure 21.

[0351] Cell line culture

[0352] Cell lines were cultured in DMEM (41966052, Gibco) 10% FCS (3101120, Sigma) or RPMI (21875034, Thermofisher) 10% FCS for lung and blood cancer cell lines. Cells were passaged at subculturing ratios corresponding to 80% confluence at the next passage. During experiments medium was supplemented with penicillin & streptomycin (P / S, 15140122, Invitrogen).

[0353] Immunoblotting

[0354] Quantitative westerns (Abby) for cleaved caspase-3 (primary 9664S; CST) with total protein were prepared and ran according to the manufacturer’s instructions with the manufacturer’s reagents. CRISPR-perturbed samples for cell death analysis were split on day 11 after transduction and harvested on day 12 after transduction. Samples for immunoblotting by simple western were collected as previously described with the additional inclusion of the floating cell fraction. As a control, cleaved caspase-3 was measured upon induction of apoptosis with 1 pM staurosporine (9953S, CST) in SK-MEL-2 or 200 ng / mL TNF (11343017, Immunotools) + 10 pM TPCA-1 (S2824, Selleckchem) in BLM.

[0355] Animal studies

[0356] All animal studies were approved by the animal ethics committee of the Netherlands Cancer Institute (NKI) and performed under approved NKI CCD (Centrale Commissie Dierproeven) projects according to the ethical and procedural guidelines established by the NKI and Dutch legislation. Mice were housed in single-use standard cages at controlled filtered air humidity (55%), temperature (21 °C) and light cycle. All housing material, food and water were autoclaved or irradiated before use. Synthetic lethality in vivo was quantified as it was in vitro with the following adaptations. Puromycin selection was not continued in vivo. 2e6 melanoma cell mixes prepared on day 0 were injected subcutaneously into the flank of weight-randomized 13-15 week-old NOD-Scid IL2Rgnull mice (Jax). Mice were numbered and researchers were blinded from this step onwards. Tumor sizes were measured by caliper. Tumors were harvested, digested and processed into homogeneous single cell solutions. Then samples were stained for human B2M (316320 Biolegend) to exclude mouse cells and analyzed by flow cytometry. Results were normalized to day 0 and sgControl. For each group one sample was unblinded during flow cytometry acquisition to set voltages and gates for all groups.

[0357] Genome-wide CRISPR perturbation

[0358] Screening was performed as published previously [40-42], The one vector Brunello library was amplified in liquid culture at 20.000x coverage and maintenance of the original sgRNA distribution was confirmed. Screen cells were transduced at an MOI of 0.3-0.7 with Polybrene (TR-1003-G, Sigma) at 8 pg / mL. Screens were performed at 200x coverage. Novaseq was used to sequence the screen samples and the amplified Brunello library. The following sequencing primers with unique dual indexes per sample were used:

[0359] Forward sequencing primer (N = unique index for each sample)

[0360] 5’AATGATACGGCGACCACCGAGATCTACACNNNNNNNNACACTCTTTCCCTACACGACG CTCTTCCGATCTTCTTGTGGAAAGGACGAAACACCG 3’

[0361] Reverse sequencing primer (N = unique index for each sample)

[0362] 5’ CAAGCAGAAGACGGCATACGAGATNNNNNNNNGTGACTGGAGTTCAGACGTGTGCT

[0363] CTTCCGATCTCGACTCGGTGCCACTTTTTCAAGTT 3’

[0364] To allow drop-out of essential sgRNAs, cells were cultured for 10 days after Brunello library transduction. After single sgControl and sgCDS2 lentivirus transduction cells were cultured for 18 days. For SK-MEL-2 comparatively little cells were available due to lethality and no hits were detected. To ensure detection of any escape mechanism, a cryopreserved sample from day 7 was cultured until day 32 after single sgControl and sgCDS2 transduction, and those results were used for the presented analyses. The colony formation assays were started with screen cells on day 8. For K562, MCF7, NCI-654 H2030, LPS141 and SK-MEL- 2 cell lines, 125.000, 125.000, 62.500, 31.250 and 125.000 cells were seeded, respectively. The crystal violet stainings were done when control cells had reached confluence, which was after 4, 14, 12, 7 and 12 days, respectively. A seeding density corresponding to the longest culture time for MCF7 was chosen to ensure maximum sensitivity for detecting a potential difference. For K562 suspension cells, the cells were instead counted with Trypan Blue using an automatic cell counter (Biorad TC20). In addition, lethality was tracked from day 8 until day 19 using tracker cells and used to calculate rescue percentages. The quality control ROC-AUC curves were generated as described previously

[0042] ,

[0365] Lethality and synthetic lethality rescue experiments

[0366] For the synthetic lethality rescue experiments the method to track synthetic lethality by tracking GFP-positivity was used. The indicated cell mixes were generated 10 days after transduction with sgControl or sgCDS2 lentivirus. After overnight attachment, treatment was started and samples were harvested to measure GFP positivity at starting point. Cells were split and refreshed on day 3 and measured on day 7 after start of treatment. Phospholipids were complexed with BSA. 500 pM stock solutions were generated by incubating phospholipid with 6702.5% fatty-acid free BSA from Sigma (#A8806) in DMEM at 37 °C for 1 hour. Liver PC 671 (840055P-25MG) or PI (840042P-25MG) from Avanti Polar Lipids was used. Final GFP- / GFP+ratios were normalized to input ratios and sgControl ratios to obtain the synthetic lethal fraction. For rescue, the fold change in the synthetic lethal fraction was calculated by dividing by the untreated synthetic lethal fractions and transformed into rescue percentages by subtraction from 1 and multiplication by 100. The resulting rescue percentages correspond to the part of the total synthetic lethality that was negated in the treated conditions. For the overnight lethality rescue, CRISPR-perturbed samples for cell death analysis were split on day 11 after transduction and harvested on day 12 after transduction. PI (840042P-100MG) from Avanti Polar Lipids was dissolved in medium and added to the cells at 2 mM. For contact inhibition 10x the normal amount of cells were seeded and cells were not split, resulting in maximum confluence from start to end. Palbociclib (Selleckchem, S1116) was used.

[0367] Gene set enrichment analyses

[0368] Hallmark gene set enrichment analyses were performed using GSEA 4.1.0 (Java tool) with the settings h.all.v2022.1 .Hs. symbols. gmt, 25000 permutations and No_Collapse on the Pearson correlations generated as described from DepMap or TOGA data [59-60]. Gene set enrichment for go term processes was performed using Stringdb “multiple proteins” function on the list of 37 genes significant only in the four CDS 1 -negative cell lines.

[0369] Lipidomics

[0370] Quantitative targeted lipidomics was carried out using a flow-injection assay employing lipid class separation by differential mobility spectroscopy and selective multiple reaction monitoring (MRM) per lipid species (Lipidyzer platform). Briefly, after the addition of >60 deuterated internal standards, lipids were extracted using methyl tert-butyl ether. Organic extracts were subsequently dried under a gentle stream of nitrogen and reconstituted in running buffer. Lipids were subsequently analyzed using flow- injection in MRM mode employing a Shimadzu Nexera series HPLC and a Sciex QTrap 5500 mass spectrometer. For further analysis, lipidomic data normalized to cell number was used to calculate the quadruplicate fold changes for plotting and statistical analysis.

[0371] Proteomics

[0372] Frozen cell pellets were heated for 7 min. at 950700 C in 1x S-Trap Lysis buffer (5% SDS, 50 mM TEAB pH 8.5), after which DNA was sheared by probe sonication. Aliquots comprising 50 pg of protein were reduced with 20 mM DTT (20 min. at 550702 C) and alkylated with 40 mM iodoacetamide (30 min. at room temperature in the dark), after which proteins were digested overnight with trypsin (Sigma-Aldrich; enzyme / substrate ratio 1 :10) on S-Trap Micro spin columns according to the manufacturer’s instructions (ProtiFi, NY, USA). Peptides were eluted, vacuum-dried and stored at -80°C until LC-MS / MS analysis. LC-MS / MS was performed by nanoLC-MS / MS on an Orbitrap Exploris 480 mass spectrometer (Thermo Scientific) connected to a Proxeon nLC1200 system. Peptides were directly loaded onto the analytical column (ReproSil-Pur 120 C18-AQ, 2.4pm, 75 pm x 500 mm, packed in-house) and eluted at 250 nL / min in a 90-minutes gradient containing a nonlinear increase from 6% to 30% solvent B (solvent A was 0.1% formic acid / water and solvent B was 0.1% formic acid / 80% acetonitrile). The Exploris 480 was run in data-independent acquisition (DIA) mode, with full MS resolution set to 120,000 at m / z 200, MS1 mass range was set from 350-1400, normalized AGC target was 300% and maximum IT was 45ms. DIA was performed on precursors from 400-1000 in 48 windows of 13.5 Da with an overlap of 1 Da. Resolution was set to 30,000 and normalized CE was 27.

[0373] RAW files were analyzed with DIA-NN (version 1.8) using standard settings. Fragment spectra were searched against the Swissprot human database (version 2022 02; 20,375 entries) by selecting ‘FASTA digest for library-free search’; Trypsin / P was specified as protease specificity allowing a maximum of 1 miscleavage; N-terminal excision (M) and carbamidomethylation (C) were selected as fixed modifications and match between runs was applied. Protein group abundances were extracted from the DIA-NN result files, imported into Perseus (1 .6.15.0) and Log2-transformed. For each cell line, values were filtered for presence in at least 4 out of 4 replicates in at least the WT or sgCDS2 sample group. Missing values were replaced by an imputation-based normal distribution using a width of 0.3 and a downshift of 2.4. Differentially expressed proteins were determined using a two- sided t-test (thresholds: p-value<0.05 and Log abundance ratio [sgCDS2 / sgControl] < -0.4A> 0.4). Light microscopy

[0374] Live cells stained with BODIPY (2 pM, D3922) were imaged on a regularly maintained Zeiss 732 Axio Observer Z1 microscope at the NKI imaging facility using PhaseContrast- and Fluorescence-equipment. A home-built temperature-controlled live cell chamber and 5% CO2 incubation system were used. Data was analyzed using a custom Imaged code. Briefly, for each condition low complexity area on phase contrast was excluded to calculate cell area and BODIPY fluorescence was used to calculate lipid area. Cell lines were imaged 14 days (SK-MEL-2) or 20 days (NCI-H2030) days after transduction.

[0375] Electron microscopy

[0376] Correlative light electron microscopy (CLEM) was performed as described before [43,44], Briefly, for CLEM 150-300 nm cryosections of cells were placed onto an EM-grid, stained with Nile red (Sigma, 72485) and Hoechst 33342 (Thermo Fisher, H3570) and imaged with a Wide-field Fluorescence Microscope (Leica DM6, 100x oil objective). After washing and contrasting for EM purposes, the identical grid was imaged using a transmission electron microscope (Talos L120c Thermo Fisher, with Ceta 16M camera).

[0377] Statistics and replication

[0378] Significance was tested and depicted as indicated. To ensure reproducibility, we addressed experimental variation by including multiple replicates, we addressed biological variation by including multiple cell lines and we employed multiple methods whenever possible.

[0379] References

[0380] 1 . Tzelepis, K. et al. A CRISPR Dropout Screen Identifies Genetic Vulnerabilities and Therapeutic Targets in Acute Myeloid Leukemia. Cell Rep 17, 1193-1205 (2016).

[0381] 2. Vidigal, J. A. & Ventura, A. Rapid and efficient one-step generation of paired gRNA CRISPR-Cas9 libraries. Nat Common 6, 8083 (2015).

[0382] 3. Doench, J. G. et al. Optimized sgRNA design to maximize activity and minimize off- target effects of CRISPR-Cas9. Nat Biotechnol 34, 184-191 (2016).

[0383] 4. Hart, T. et al. High-Resolution CRISPR Screens Reveal Fitness Genes and Genotype-Specific Cancer Liabilities. Cell 163, 1515-1526 (2015).

[0384] 5. Wang, T. et al. Identification and characterization of essential genes in the human genome. Science (1979) 350, 1096-1101 (2015).

[0385] 6. Morgens, D. W. et al. Genome-scale measurement of off-target activity using Cas9 toxicity in high-throughput screens. Nat Common 8, 15178 (2017).

[0386] 7. Thompson, N. A. et al. Combinatorial CRISPR screen identifies fitness effects of gene paralogues. Nat Common 12, 1302 (2021).

[0387] 8. Iorio, F. et al. Unsupervised correction of gene-independent cell responses to CRISPR-Cas9 targeting. BMC Genomics 19, 604 (2018).

[0388] 9. Behan, F. M. et al. Prioritization of cancer therapeutic targets using CRISPR-Cas9 screens. Nature 568, 511-516 (2019).

[0389] 10. Gongalves, E. et al. Minimal genome-wide human CRISPR-Cas9 library. Genome Biol 22, 40 (2021 ).

[0390] 11. Pelster MS, Gruschkus SK, Bassett R, Gombos DS, Shephard M, Posada L, et al. Nivolumab and Ipilimumab in Metastatic Uveal Melanoma: Results From a Single-Arm Phase II Study. J Clin Oncol 2021 ; 39 (6):599-607.

[0391] 12. Piulats JM, Espinosa E, de la Cruz Merino L, Varela M, Alonso Carrion L, Martin- Algarra S, et al. Nivolumab Plus Ipilimumab for Treatment-Naive Metastatic Uveal Melanoma: An Open-Label, Multicenter, Phase II Trial by the Spanish Multidisciplinary Melanoma Group (GEM-1402). J Clin Oncol 2021 ; 39 (6):586-598.

[0392] 13. Nathan P, Hassel JC, Rutkowski P, Baurain JF, Butler MO, Schlaak M, et al. Overall Survival Benefit with Tebentafusp in Metastatic Uveal Melanoma. N Engl J Med 2021 ; 385 (13):1196-1206. 14. Carvajal RD, Butler MO, Shoushtari AN, Hassel JC, Ikeguchi A, Hernandez-Aya L, et al. Clinical and molecular response to tebentafusp in previously treated patients with metastatic uveal melanoma: a phase 2 trial. Nat Med 2022; 28 (11):2364-2373.

[0393] 15. Shoushtari AN, Khan S, Komatsubara K, Feun L, Acquavella N, Singh-Kandah S, et al. A Phase lb Study of Sotrastaurin, a PKC Inhibitor, and Alpelisib, a PI3Kalpha Inhibitor, in Patients with Metastatic Uveal Melanoma. Cancers (Basel) 2021 ; 13 (21).

[0394] 16. Shoushtari AN, Ong LT, Schoder H, Singh-Kandah S, Abbate KT, Postow MA, et al. A phase 2 trial of everolimus and pasireotide long-acting release in patients with metastatic uveal melanoma. Melanoma Res 2016; 26 (3):272-277.

[0395] 17. Shoushtari AN KR, Panageas K, Murthy RK, Jung M, Shah R, O'Donnell B, Khawaja TT, Shames Y, Prempeh-Keteku NA, Ambrosini G. . A randomized phase 2 study of trametinib with or without GSK2141795 in patients with advanced uveal melanoma. J Clin Oncol 2016; 34:9511.

[0396] 18. Joseph J Sacco PDN, Sarah Danson , Paul Lorigan , Steve Nicholson , Christian OttensmeierPhilippa Corrie , Neil Steven , Andrew Goodman , James M. G. Larkin , T. R. Jeffry Evans , Satish Kumar , Sarah E Coupland , Paul Silcocks , Ernie Marshall Sunitinib versus dacarbazine as first-line treatment in patients with metastatic uveal melanoma. J Clin Oncol 2013; 31 :9031.

[0397] 19. Luke JJ, Olson DJ, Allred JB, Strand CA, Bao R, Zha Y, et al. Randomized Phase II Trial and Tumor Mutational Spectrum Analysis from Cabozantinib versus Chemotherapy in Metastatic Uveal Melanoma (Alliance A091201). Clin Cancer Res 2020; 26 (4):804-811 .

[0398] 20. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, et al. The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature 2012; 483 (7391):603-607.

[0399] 21 . Onken MD, Worley LA, Ehlers JP, Harbour JW. Gene expression profiling in uveal melanoma reveals two molecular classes and predicts metastatic death. Cancer Res 2004; 64 (20):7205-7209.

[0400] 22. Onken MD, Worley LA, Tuscan MD, Harbour JW. An accurate, clinically feasible multi-gene expression assay for predicting metastasis in uveal melanoma. J Mol Diagn 2010; 12 (4):461-468.

[0401] 23. Guo J, Liu H, Zheng J. SynLethDB: synthetic lethality database toward discovery of selective and sensitive anticancer drug targets. Nucleic Acids Res 2016; 44 (D1 ):D1011- 1017. 24. Thompson NA, Ranzani M, van der Weyden L, Iyer V, Offord V, Droop A, et al. Combinatorial CRISPR screen identifies fitness effects of gene paralogues. Nat Common 2021 ; 12 (1):1302.

[0402] 25. Patel, S. P. et al. A phase II study of gefitinib in patients with metastatic melanoma. Melanoma Res. 21 , 357-363 (2011).

[0403] 26. Blomen VA, Majek P, Jae LT, Bigenzahn JW, Nieuwenhuis J, Staring J, Sacco R, van Diemen FR, Oik N, Stukalov A, Marceau C. Gene essentiality and synthetic lethality in haploid human cells. Science. 2015 Nov 27;350(6264):1092-6.

[0404] 27. Dandage R, Landry CR. Paralog dependency indirectly affects the robustness of human cells. Molecular systems biology. 2019 Sep;15(9):e8871 .

[0405] 28. De Kegel B, Ryan CJ. Paralog buffering contributes to the variable essentiality of genes in cancer cell lines. PLoS genetics. 2019 Oct 25; 15(10):e1008466.

[0406] 29. Gu Z, Steinmetz LM, Gu X, Scharfe C, Davis RW, Li WH. Role of duplicate genes in genetic robustness against null mutations. Nature. 2003 Jan 2;421 (6918):63-6.

[0407] 30. Wang T, Birsoy K, Hughes NW, Krupczak KM, Post Y, Wei JJ, Lander ES, Sabatini DM. Identification and characterization of essential genes in the human genome. Science. 2015 Nov 27;350(6264):1096-101.

[0408] 31 . D’Souza, Kim, Balia, Epand. Distinct properties of the two isoforms of CDP- diacylglycerol synthase. Biochemistry 2014 Dec 2;53(47):7358-67

[0409] 32. DepMap 21 Q2 Public. Available at: https: / / figshare.com / articles / dataset / DepMap_21 Q2_Public / 14541774 / 2. (Accessed: 20th October 2021)

[0410] 33. Meyers, R. M. et al. Computational correction of copy number effect improves specificity of CRISPR-Cas9 essentiality screens in cancer cells. Nat. Genet. 49, 1779-1784 (2017).

[0411] 34. Nusinow, D. P. et al. Quantitative Proteomics of the Cancer Cell Line Encyclopedia.

[0412] Cell 180, 387-402.e16 (2020).

[0413] 35. Ghandi, M. et al. Next-generation characterization of the Cancer Cell Line Encyclopedia. Nature 569, 503-508 (2019).

[0414] 36. Gemmill, R. M. et al. ZEB1 -responsive genes in non-small cell lung cancer. Cancer Lett. 300, 66-78 (2011). 37. Feingold, E. A. et al. The ENCODE (ENCyclopedia of DNA Elements) Project. Science 306, 636-640 (2004)

[0415] 38. Eisenberg, E. & Levanon, E. Y. Human housekeeping genes, revisited. Trends in Genetics 29, 569-574 (2013)

[0416] 39. Apriamashvili, G. etal. Ubiquitin ligase STUB1 destabilizes I FNy-receptor complex to suppress tumor IFNy signaling. Nat. Commun. 13, 1-16 (2022).

[0417] 40. Vredevoogd, D. W. et al. Augmenting Immunotherapy Impact by Lowering Tumor TNF Cytotoxicity Threshold. Cell 178, 585-599.e15 (2019).

[0418] 41. Sanson, K. R. etal. Optimized libraries for CRISPR-Cas9 genetic screens with multiple modalities. Nat. Commun. 9, 5416 (2018).

[0419] 42. Joung, J. et al. Genome-scale CRISPR-Cas9 knockout and transcriptional activation screening. Nat. Protoc. 12, 828-863 (2017).

[0420] 43. Grootemaat, A. E. et al. Lipid and Nucleocapsid N-Protein Accumulation in COVID-19

[0421] Patient Lung and Infected Cells. Microbiol. Spectr. 10, e01271-21 (2022).

[0422] 44. Bedard, M. etal. A terpene nucleoside from M. tuberculosis induces lysosomal lipid storage in foamy macrophages. J. Clin. Invest. 133, e161944 (2023)

[0423] 45. Cho, H., Berger, B. & Peng, J. Compact Integration of Multi-Network Topology for Functional Analysis of Genes. Cell Syst 3, 540-548 e5 (2016).

[0424] 46. Cancer Genome Atlas Research, N. et al. The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet 45, 1113-20 (2013).

[0425] 47. Dwane, L. et al. Project Score database: a resource for investigating cancer cell dependencies and prioritizing therapeutic targets. Nucleic Acids Res 49, D1365-D1372 (2021).

[0426] 48. Pacini, C. et al. Integrated cross-study datasets of genetic dependencies in cancer.

[0427] Nat Commun 12, 1661 (2021 ).

[0428] 49. Geary, N. Understanding synergy. Am J Physiol Endocrinol Metab 304, E237-53 (2013).

[0429] 50. Kim, E. & Hart, T. Improved analysis of CRISPR fitness screens and reduced off- target effects with the BAGEL2 gene essentiality classifier. Genome Med 13, 2 (2021).

[0430] 51 . Robertson, A.G. et al. Integrative Analysis Identifies Four Molecular and Clinical

[0431] Subsets in Uveal Melanoma. Cancer Cell 32, 204-220 e15 (2017). 52. Goldman, M.J. et al. Visualizing and interpreting cancer genomics data via the Xena platform. Nat Biotechnol 38, 675-678 (2020).

[0432] 53. Behan, F.M. et al. Prioritization of cancer therapeutic targets using CRISPR-Cas9 screens. Nature 568, 511-516 (2019).

[0433] 54. Yang, J. & Weinberg, R. A. Epithelial-Mesenchymal Transition: At the Crossroads of Development and Tumor Metastasis. Developmental Cell 14, 818-829 (2008).

[0434] 55. Mani, S. A. et al. The Epithelial-Mesenchymal Transition Generates Cells with

[0435] Properties of Stem Cells. Cell 133, 704-715 (2008).

[0436] 56. Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell 144, 646-74 (2011)

[0437] 57. Fischer, K. R. et al. Epithelial-to-mesenchymal transition is not required for lung metastasis but contributes to chemoresistance. Nature 527, 472-476 (2015).

[0438] 58. Van Espen, B. et al. RNF185 Control of COL3A1 Expression Limits Prostate Cancer

[0439] Migration and Metastatic Potential. Mol. Cancer Res. OF1-OF14 (2023). doi:10.1158 / 1541- 7786.MCR-23-0512

[0440] 59. Mootha, V. K. et al. PGC-1a-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat. Genet. 34, 267-273 (2003).

[0441] 60. Subramanian, A. et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. II. S. A. 102, 15545- 15550 (2005).

[0442] 61 . Blunsom, Nicholas J et al. “Mitochondrial CDP-diacylglycerol synthase activity is due to the peripheral protein, TAMM41 and not due to the integral membrane protein, CDP- diacylglycerol synthase 1.” Biochimica et biophysica acta. Molecular and cell biology of lipids vol. 1863,3 (2018): 284-298. doi:10.1016Zj.bbalip.2017.12.005

[0443] 62. Weinstein, J. N. et al. The cancer genome atlas pan-cancer analysis project. Nat.

[0444] Genet. 45, 1113-1120 (2013).

[0445] 63. Han, K., et al., Synergistic drug combinations for cancer identified in a CRISPR screen for pairwise genetic interactions. Nat Biotechnol, 2017. 35(5): p. 463-474 64. Liu X, Yin Y, Wu J, Liu Z. Structure and mechanism of an intramembrane liponucleotide synthetase central for phospholipid biosynthesis. Nature Communications. 2014 Jun;5:4244.

[0446] 65. Blunsom NJ, Gomez-Espinosa E, Ashlin TG, Cockcroft S. Mitochondrial CDP- diacylglycerol synthase activity is due to the peripheral protein, TAMM41 and not due to the integral membrane protein, CDP-diacylglycerol synthase 1. Biochim Biophys Acta Mol Cell Biol Lipids. 2018;1863(3):284-298.

Claims

Claims1 . A method of treating or preventing cancer in a subject, wherein the cancer has functional suppression of CDS1 , the method comprising administering a therapeutically effective amount of a CDS2 inhibitor to the subject.

2. A CDS2 inhibitor for use in a method of treating or preventing cancer characterised by functional suppression of CDS1 in a subject in need thereof, the method comprising the step of administering the CDS2 inhibitor to the subject.

3. A CDS2 inhibitor for use in the manufacture of a medicament for the treatment of cancer having functional suppression of CDS1 in a in a biological sample obtained from the subject.

4. The method according to claim 1 , or CDS2 inhibitor for use according to claim 2 or claim 3, wherein the method comprises detecting presence or absence of functional suppression of CDS1 in the cancer.

5. The method or the CDS2 inhibitor for use according to claim 4, wherein the presence of functional suppression of CDS1 has been determined in the subject prior to treatment.

6. A method for selecting a subject for the treatment of cancer with a cytidine diphosphate diacylglycerol synthase 2 (CDS2) inhibitor, the method comprising the steps of:(i) providing a biological sample obtained from the subject,(ii) detecting presence or absence of functional suppression of cytidine diphosphate diacylglycerol synthase 1 (CDS1 ) in the biological sample, and(iii) selecting the subject for treatment with the CDS2 inhibitor if the presence of functional suppression of CDS1 is detected.

7. A method for predicting whether a subject having cancer is likely to respond to treatment with a CDS2 inhibitor, the method comprising the steps of:(i) providing a biological sample obtained from the subject,(ii) detecting presence or absence of functional suppression of CDS1 in the biological sample, wherein if the presence of functional suppression of CDS1 is detected then predicting that the subject is likely to respond to treatment with the CDS2 inhibitor.

8. A method for identifying a subject with cancer who is likely to respond to treatment with a CDS2 inhibitor, the method comprising the steps of:(i) providing a biological sample obtained from the subject,(ii) detecting presence or absence of functional suppression of CDS1 in the biological sample, wherein the subject is likely to respond to treatment with the CDS2 inhibitor if the presence of functional suppression of CDS1 is detected.

9. A method of identifying and treating a subject suitable for treatment of cancer, the method comprising the steps of: i) obtaining a biological sample from the subject, ii) detecting presence or absence of functional suppression of CDS1 in the biological sample, iii) identifying the subject as suitable for treatment of cancer if the presence of functional suppression of CDS1 is detected, and iv) administering an effective amount of a CDS2 inhibitor to a subject identified as suitable for treatment.

10. A method for detecting presence or absence of the functional suppression of CDS1 in a biological sample derived from a subject in order to select a subject for the treatment of cancer with a CDS2 inhibitor.

11. A method of screening for a compound for use in the treatment of cancer having functional suppression of CDS1 , the method comprising the steps of:(i) administering the compound to cancer cells having functional suppression of CDS1 in an in vitro system or to an in vivo non-human animal model of the cancer having functional suppression of CDS1 , and(ii) determining if there is any inhibition or suppression of cancer cell proliferation in vitro or of the cancer in the in vivo model, wherein if the compound results in inhibition or suppression of cancer cell proliferation in vitro or of the cancer in the in vivo model, then it may be useful in the treatment of cancer having functional suppression of CDS1 .

12. The method of claim 11 , wherein inhibition or suppression of cancer cell proliferation may be determined by measuring tumour regression, quantifying circulating tumour cells, quantifying tumour markers, or by cell survival and proliferation assay, optionally wherein the cell survival and proliferation assay comprises quantifying the number of dead cancer cells by flow cytometry, optionally using a live / dead cell dye.

13. The method according to any one of claims 1 or 4-12 or the CDS2 inhibitor for use according to any one of claims 2-5, wherein functional suppression of CDS1 comprises inactivating mutations, decreased transcription of a CDS1 gene, decreased translation of CDS mRNA, decreased expression of CDS1 , no expression of CDS1 , heterozygous deletion of a CDS1 gene, or homozygous deletion of a CDS1 gene.

14. The method or CDS2 inhibitor for use according to claim 13, wherein the functional suppression is compared with a control.

15. The method or CDS2 inhibitor for use according to claim 14, wherein the control is the level of function of CDS1 in non-cancerous tissue in the subject or from the corresponding tissue type in a normal, healthy subject.

16. The method according to any one of claims 1 or 4 to 15, or the CDS2 inhibitor for use according to any one of claims 2 to 5, wherein the CDS2 inhibitor is an antibody, a small molecule inhibitor or an RNA inhibitor.

17. The method according to any one of claims 1 or 4 to 16, or the CDS2 inhibitor for use according to any one of claims 2 to 5, wherein the RNA inhibitor is a short interfering RNA (siRNA), short hairpin (shRNA), guide RNA (gRNA), antisense oligonucleotide, antagomir, or microRNA.

18. The method according to any one of claims 1 , 4 to 17, or the CDS2 inhibitor for use according to any one of claims 2 to 5, wherein the cancer is uveal melanoma, diffuse B cell lymphoma, acute myeloid leukaemia, hepatocellular carcinoma, sarcoma, cutaneous melanoma, pancreatic cancer, skin cancer, blood cancer, brain cancer, lung cancer, colorectal cancer, breast cancer, renal cancer, cutaneous melanoma, liposarcoma, non-small cell lung cancer, or chronic myeloid leukaemia.

19. The method according to any one of claims 1 , 4 to 18, or the CDS2 inhibitor for use according to any one of claims 2 to 5, wherein the cancer is uveal melanoma, skin cancer, blood cancer, brain cancer, breast cancer, or lung cancer.

Citation Information

Patent Citations

  • Synthetic lethal treatments

    GB202307864D0

  • Method for efficiently inducing blood vessel degeneration and application

    CN109136357A

  • Methods for the treatment of neurological disorders

    WO2022104011A1