Identifying druggable oncogenes targeted for amplification in cancer: an introduction to the ConSig-amp analysis.

Comprehensive functional analysis of the tousled-like kinase 2 frequently amplified in aggressive luminal breast cancers. Kim JA, Tan Y, Wang X, Cao X, Veeraraghavan J, Liang Y, Edwards DP, Huang S, Pan X, Li K, Schiff R. and Wang XS#. Nature Communications. 2016, In Press.

     Genomic amplifications lead to deregulations of oncogenes to which cancer cells become often addicted in specific tumors. Such events, however, usually affect a large number of genes in cancer genomes which makes it difficult to identify the primary oncogene targets of these amplifications. In our previous study, we discovered that cancer genes possess complicated yet distinctive “gene concept signature”, which include cancer-related signaling pathways, molecular interactions, transcriptional motifs, protein domains, and gene ontologies(1). Based on this observation, we developed a Concept Signature (or ConSig) analysis that prioritizes the biological importance of candidate genes underlying cancer via computing their strength of association with those cancer-related signature concepts (http://consig.cagenome.org) (1-3). In our previous study, we have applied this analysis to reveal the primary target genes of chromosome 17q amplifications in breast cancer (4). Here we postulate that the ConSig analysis may be used to effectively nominate dominantly acting cancer genes from the genomic amplifications in cancer at a genome-wide scale, which can be further translated into viable therapeutic targets by interrogating pharmacological databases. Indeed, analyses of known amplified oncogene targets (i.e. ERBB2, CCND1, MYC, PAK1, NCOA3, YWHAZ)(5-10) in breast cancer suggest that ConSig analysis can effectively point out the primary oncogenes targeted by genomic amplifications (Figure. 1). Toward this end, we have assembled a genome-wide analysis called “ConSig-Amp” to discover viable therapeutic targets in cancer from multi-dimensional genomic datasets (Figure 2a).

Figure 1. The ConSig scores, the amplification frequencies, and the correlations of expressions of the genes within the known amplified genomic regions in breast cancer. The amplification frequencies are shown in red bar chart, ConSig scores are shown in blue line chart, and the gene expression correlations based on Spearman’s statistics are shown in dot-plot. The gene names with high ConSig scores (>1.5) are shown under each chart.

    To discover new therapeutic targets in ER+ breast cancer, we analyzed the copy number (Affymetrix SNP 6.0) and RNAseq (UNC RNAseqV2) datasets available for breast tumors from The Cancer Genome Atlas Project (TCGA)(11). Normalized “level 3” data (segmented by the CBS algorithm) (14) were directly applied in the analysis. First, the copy number segments were matched with human genes based on physical coordinates to obtain gene-level copy number data. The frequency of genomic amplification of each human gene in breast cancer was assessed; breast tumors with relative copy number at the respective gene locus more than 0.7 were considered as amplification positive. Genes that are amplified in >5% of ER+ tumors were nominated, and their expressions based on RNAseq data were correlated with copy number data by Spearman’s correlation statistics. The druggability of these genes was predicted based on a drug-target database compiled from multiple sources(12-14). Then all candidates were ranked by the ConSig-amp score calculated by multiplying the Spearman’s correlation coefficient by the concept signature (ConSig) score that we have developed that prioritizes functionally important genes underlying cancer by accessing their associations with cancer-related molecular concepts(1). The detailed protocol to calculate the ConSig Score and the precomputed scores used in this study (for all human genes) are available in the website http://consig.cagenome.org (release 2).

Figure 2. ConSig-Amp identifies TLK2 as a candidate druggable target frequently amplified in breast cancer. (a) The bioinformatics workflow of ConSig-Amp to discover therapeutically relevant oncogene targets in cancer at genome-wide scale based on copy number and RNAseq datasets. The ConSig-Amp score is calculated by multiplying the ConSig score (see Methods) with the correlation between gene expression and copy number. (b) Prioritizing amplified breast cancer oncogene targets by ConSig score and Spearman’s correlation between copy number (Affymetrix SNP 6.0 array) and gene expression (RNAseq). Data shown here are from TCGA.

    This analysis revealed several known kinase targets in breast cancer such as ERBB2, PAK1, RPS6KB1, and PTK2(15, 16), together with a new candidate kinase target, TLK2 (Figure 2b). ERBB2, RPS6KB1, and TLK2 all locate at the peaks of both frequent amplifications and high ConSig scores in Chr17q (Figure 3). Such coincidence of the two parameters at high levels provided integrated evidence about their role as the primary targets of these amplified genomic regions.

Figure 3. Frequent gene amplifications in Chr17q with significantly correlated gene expressions. Chr17q genes amplified in >2% of breast cancers as well as having Spearman’s correlation coefficient R>0.5 are shown in the chart. The concept signature scores for these genes are shown in the blue line chart. The three lead amplified kinase targets (ERBB2, RPS6KB1, and TLK2) nominated by ConSig-Amp analysis are shown in the chart. All three targets locate at the peaks of both genomic amplifications and ConSig scores. This coincidence provided integrated evidence about their functional importance in breast cancer. TLK2 locates in a small peak region of genomic amplifications close to the RPSKB1 amplicon. This figure is based on the copy number data and RNAseq expression data from TCGA.

    Here we demonstrated the implementation of the Concept Signature analysis, to facilitate cancer target discovery from genomic datasets. This analysis automatically recognizes the complex molecular fingerprints in cancer genes and enables high-throughput assessment of the function of candidate targets underlying cancer. Interestingly, analyses of known amplified oncogenes suggest that ConSig scores provide independent evidence to identify oncogenes targeted by genomic amplifications other than the correlation of copy number with gene expression. This observation has led to our development of the genome-wide ConSig-Amp analysis to assess the functional importance of genes within the amplified regions of cancer genome. By interrogating different types of genomic and pharmacological data, our integrative ConSig-Amp analysis enables effective navigation of the complex cancer signaling network to reveal key oncogene targets that are directly druggable. In addition, this analysis can also be integrated with other genomic alternations revealed by genomic or transcriptomic sequencing, such as somatic point mutations or recurrent gene fusions, to nominate cancer genes that are targeted by multiple types of genetic alternations.

    Applying ConSig-amp to the genomic data from TCGA has led to our discovery of a novel cell cycle kinase target TLK2 that are upregulated by genomic amplifications in more aggressive and lethal form of ER+ breast cancers. This discovery suggested the application of ConSig-Amp in discovering previously uncharacterized cancer genes targeted for amplifications in the tumor genomes. The criteria to determine a true gene amplification has been suggested previously: physical mapping of the amplifications in multiple tumors, correlation of gene expression with copy number increase, association with clinical outcome, and its biological function in cancer(15). TLK2 gene amplification fulfills all these criteria: a) TLK2 locates within a consensus region of chr17q23.2 amplifications in breast cancers; b) TLK2 overexpression in breast cancer is primarily driven by increased copy number; c) TLK2 overexpression correlates with poor clinical outcome of ER+ breast cancer patients irrespective of endocrine treatment; d) TLK2 inhibition potently and selectively inhibits the growth of the breast cancer cells with TLK2 amplification and overexpression; e) Our biological studies strongly support the role of TLK2 in cell cycle regulation, anti-apoptosis, and enhanced aggressiveness of ER+ breast cancers.

Figure 4. Identification of TLK2 as an amplified kinase target in aggressive luminal breast cancer. (a) The bioinformatics workflow of ConSig-Amp to discover therapeutically relevant oncogene targets in cancer at genome-wide scale based on TCGA copy number and RNAseq datasets. (b) Kaplan-Meier plots based on multiple gene expression datasets showing correlation of TLK2 overexpression with the outcome of systemically untreated or endocrine-treated ER+ breast cancer patients. (c) A schematic of normal G1/S cell cycle signaling and their alternations following TLK2 inhibition (black arrows). (d) The effect of TLK2 inhibition in the MCF7 xenograft tumors inducibly expressing a TLK2 shRNA, in the presence or absence of concomitant tamoxifen treatment. Figure shows the Kaplan–Meier survival plot comparing the progression-free survival of different treatment groups.

    Consistent with our findings, the latest phosphoproteomic study of TCGA breast tumors by The Clinical Proteomic Tumor Analysis Consortium (CPTAC) independently identified TLK2 as an amplicon-associated highly phosphorylated kinases in luminal breast cancer(17), which further support the significance of TLK2 amplification and its preferential association with luminal tumors. Our study is the first comprehensive analysis of TLK2 function in aggressive luminal breast cancers, which will timely complement the CPTAC paper.

References:

1. Wang XS, Prensner JR, Chen G, Cao Q, Han B, Dhanasekaran SM, et al. An integrative approach to reveal driver gene fusions from paired-end sequencing data in cancer. Nat Biotechnol. 2009;27:1005-11.
2. Wang X-S, Shankar S, Dhanasekaran SM, Ateeq B, Prensner JR, Yocum AK, et al. Characterization of KRAS Rearrangements in Metastatic Prostate Cancer. Cancer Discovery. 2011;doi: 10.1158/2159-8274.CD-10-0022.
3. Veeraraghavan J, Tan Y, Cao XX, Kim JA, Wang X, Chamness GC, et al. Recurrent ESR1-CCDC170 rearrangements in an aggressive subset of oestrogen receptor-positive breast cancers. Nat Commun. 2014;5:4577.
4. Fan Y, Ge N, Wang X, Sun W, Mao R, Bu W, et al. Amplification and over-expression of MAP3K3 gene in human breast cancer promotes formation and survival of breast cancer cells. J Pathol. 2014;232:75-86.
5. Borg A, Baldetorp B, Ferno M, Killander D, Olsson H, Sigurdsson H. ERBB2 amplification in breast cancer with a high rate of proliferation. Oncogene. 1991;6:137-43.
6. Lundgren K, Brown M, Pineda S, Cuzick J, Salter J, Zabaglo L, et al. Effects of cyclin D1 gene amplification and protein expression on time to recurrence in postmenopausal breast cancer patients treated with anastrozole or tamoxifen: a TransATAC study. Breast Cancer Res. 2012;14:R57.
7. Bonilla M, Ramirez M, Lopez-Cueto J, Gariglio P. In vivo amplification and rearrangement of c-myc oncogene in human breast tumors. J Natl Cancer Inst. 1988;80:665-71.
8. Bostner J, Ahnstrom Waltersson M, Fornander T, Skoog L, Nordenskjold B, Stal O. Amplification of CCND1 and PAK1 as predictors of recurrence and tamoxifen resistance in postmenopausal breast cancer. Oncogene. 2007;26:6997-7005.
9. Osborne CK, Bardou V, Hopp TA, Chamness GC, Hilsenbeck SG, Fuqua SA, et al. Role of the estrogen receptor coactivator AIB1 (SRC-3) and HER-2/neu in tamoxifen resistance in breast cancer. J Natl Cancer Inst. 2003;95:353-61.
10. Li Y, Zou L, Li Q, Haibe-Kains B, Tian R, Desmedt C, et al. Amplification of LAPTM4B and YWHAZ contributes to chemotherapy resistance and recurrence of breast cancer. Nat Med. 2010;16:214-8.
11. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490:61-70.
12. Wishart DS, Knox C, Guo AC, Shrivastava S, Hassanali M, Stothard P, et al. DrugBank: a comprehensive resource for in silico drug discovery and exploration. Nucleic Acids Res. 2006;34:D668-72.
13. Chen X, Ji ZL, Chen YZ. TTD: Therapeutic Target Database. Nucleic Acids Res. 2002;30:412-5.
14. Anastassiadis T, Deacon SW, Devarajan K, Ma H, Peterson JR. Comprehensive assay of kinase catalytic activity reveals features of kinase inhibitor selectivity. Nature biotechnology. 2011;29:1039-45.
15. Santarius T, Shipley J, Brewer D, Stratton MR, Cooper CS. A census of amplified and overexpressed human cancer genes. Nature reviews Cancer. 2010;10:59-64.
16. Glenisson M, Vacher S, Callens C, Susini A, Cizeron-Clairac G, Le Scodan R, et al. Identification of new candidate therapeutic target genes in triple-negative breast cancer. Genes Cancer. 2012;3:63-70.
17. Philipp Mertins, D. R. Mani, Kelly V. Ruggles, Michael A. Gillette, Karl R. Clauser, Pei Wang, et al. Proteogenomics connects somatic mutations to signalling in breast cancer. Nature. 2016;doi:10.1038/nature18003.