Dataset Information

INKA, an integrative analysis pipeline for inference of hyperactive phosphokinases in mass spectrometry-based phosphoproteomics data

ABSTRACT: As a large proportion of current targeted cancer therapies is based on specific (tyrosine) kinase inhibitors, uncovering the identity of hyperactive kinases at work in tumors is crucial for proper treatment selection. A major challenge in phosphoproteomic data analysis is to relate kinases to the extent and magnitude of their phosphorylation 'footprint', and to rank them accordingly in order to single out kinases that are crucial for tumor growth in individual patients. Previous approaches have either zeroed in on phosphorylation of kinases themselves as a read-out of kinase activity, or focused on inferring kinase activities from the phosphorylation of their (supposed) substrates. Here, we combine both kinase-centric and substrate-centric analyses. We present a computational pipeline called Integrative Inferred Kinase Activity (INKA) scoring that integrates the observed abundance of phosphopeptides derived from (i) kinases as a whole, (ii) their kinase activation loop segments, (iii) proteins deduced to be kinase substrates based on prior experimental knowledge of kinase-substrate relationships, and/or (iv) kinase substrates predicted by the sequence motif- and network-based NetworKIN algorithm. As a proof of concept, applying this pipeline to the analysis of phosphoproteomic data on seven different oncogene-driven cancer cell lines highlighted the high, top ranking inferred activity of known driver kinases in these cells. This demonstrates the potential of label-free MS-based phosphoproteomics combined with dedicated data analysis for identifying (hyper)active kinases that, in potential future applications to tumors, may be selected for targeted inhibition in a personalized treatment setting.

INSTRUMENT(S): Q Exactive

ORGANISM(S): Homo Sapiens (human)

TISSUE(S): Blood Cell, Epithelial Cell

SUBMITTER: Sander Piersma

LAB HEAD: Connie Ramona Jimenez

PROVIDER: PXD006616 | Pride | 2019-04-18

REPOSITORIES: Pride

ACCESS DATA

Dataset's files

Source:

			Action	DRS
	MaxQuant_TiOx_data.zip	Other
	MaxQuant_pTyr_HCC827_ER3.zip	Other
	MaxQuant_pTyr_HCC827parental.zip	Other
	MaxQuant_pTyr_K562_H2228_SK_Mel28.zip	Other
	QE1_140414_OPL1013_IdR_CvA_pTyrIP_A.raw	Raw

Items per page:

1 - 5 of 38

Publications

INKA, an integrative data analysis pipeline for phosphoproteomic inference of active kinases.

Beekhof Robin R van Alphen Carolien C Henneman Alex A AA Knol Jaco C JC Pham Thang V TV Rolfs Frank F Labots Mariette M Henneberry Evan E Le Large Tessa Ys TY de Haas Richard R RR Piersma Sander R SR Vurchio Valentina V Bertotti Andrea A Trusolino Livio L Verheul Henk Mw HM Jimenez Connie R CR

Molecular systems biology 20190412 4

Identifying hyperactive kinases in cancer is crucial for individualized treatment with specific inhibitors. Kinase activity can be discerned from global protein phosphorylation profiles obtained with mass spectrometry-based phosphoproteomics. A major challenge is to relate such profiles to specific hyperactive kinases fueling growth/progression of <i>individual</i> tumors. Hitherto, the focus has been on phosphorylation of either kinases or their substrates. Here, we combined label-free kinase-c ...[more]

PMID: 30979792

Similar Datasets

Project description:Renal cell carcinoma (RCC) is one of the ten most common malignancies. Even though recent paradigm changes led to inclusion of immune checkpoint inhibitor combination therapy in treatment of naive metastatic RCC, insight in the biology and management of RCC would still favour the use of a combination of kinase inhibitors to target multiple pathways, presenting the possibility of enhanced efficacy and reduced development of resistance. We have used the streamlined-feedback system control (s-FSC) technique, a phenotypic approach, to identify optimized drug combinations (ODC) in 5 different human RCC lines. This approach requires no a priori mechanistic information on drug mechanism of action, and uses statistical modelling circumventing the need to experimentally screen large numbers of combinations. Different ODC with up to 90% effectivity were obtained for the 5 cell lines, importantly, by combining drug at doses that individually only inhibit 5-25% cell viability. Cell viability was reduced and apoptosis induced, accompanied by a general inhibition of downstream survival and growth promoting effector kinase RPS6. Global phosphoproteomics analysis of the cell lines revealed that in all cases where the ODC showed >50% efficacy, the most active kinases were targets of the selected drugs in the ODC. Interestingly, we observed considerable overlap in the top 20 most active kinases in all 5 cell lines, suggesting a universal ODC might be effective in all cells. Indeed, a combination of erlotinib and dasatinib, targeting EGFR and EPHA2/SRC signaling respectively, combined with AZD4547 or axitinib (targeting VEGFR and FGFR signaling), displayed excellent activity. The effect of the 786-O cell-line specific ODC was analysed in more detail using phosphoproteomics coupled to kinase activity analysis using the INKA pipeline.approaches. As expected, the activity of the main targets of erlotinib and dasatinib was reduced, but the activity of several untargeted kinases, including AXL, PTK2 and MET, remained high or increased, having potential implications for the development of resistance. Addition of crizotinib, targeting MET and PTK2, to the 3-drug ODC but not exchange with axitinib, further enhanced drug combination efficacy. In conclusion, we show that the phenotypic s-FSC method is highly effective in selecting optimized combinations of drugs, and that phosphoproteomics analysis can help further refine the selection of drugs included in the screen or final drug mixture.