The NCI-60 Methylome and Its Integration into CellMiner

Abstract

A unique resource for systems pharmacology and genomic studies is the NCI-60 cancer cell line panel, which provides data for the largest publicly available library of compounds with cytotoxic activity (~21,000 compounds), including 108 FDA-approved and 70 clinical trial drugs as well as genomic data, including whole-exome sequencing, gene and microRNA transcripts, DNA copy number, and protein levels. Here we provide the first readily usable genome-wide DNA methylation database for the NCI-60, including 485,577 probes from the Infinium HumanMethylation450k BeadChip array, which yielded DNA methylation signatures for 17,559 genes integrated into our open access CellMiner version 2.0 (https://discover.nci.nih.gov/cellminer). Among new insights, transcript versus DNA methylation correlations revealed the epithelial/mesenchymal gene functional category as being influenced most heavily by methylation. DNA methylation and copy number integration with transcript levels yielded an assessment of their relative influence for 15,798 genes, including tumor suppressor, mitochondrial, and mismatch repair genes. Four forms of molecular data were combined, providing rationale for microsatellite instability for 8 out of the 9 cell lines in which it occurred. Individual cell line analyses showed global methylome patterns with overall methylation levels ranging from 17 to 84%. A six-gene model including PARP1, EP300, KDM5C, SMARCB1 and UHRF1 matched this pattern. Additionally, promoter methylation of two translationally relevant genes Schlafen 11 (SLFN11) and methylguanine methyltransferase (MGMT) served as indicators of therapeutic resistance or susceptibility, respectively. Overall, our database provides a resource of pharmacological data that can reinforce known therapeutic strategies and identify novel drugs and drug targets across multiple cancer types.

Integrative analysis of pharmacogenomics in major cancer cell line databases using CellMinerCDB

“…Data source comparisons. CellMinerCDB integrates four prominent cancer cell line data sources (CellMiner NCI-60 (2,(8)(9)(10), Sanger/Massachusetts General Hospital GDSC (3), the Broad/Novartis CCLE, and the Broad CTRP) (5,6) and a tissue-specific dataset encompassing 66 small cell lung cancer lines (NCI SCLC) (11) (Figure 1). Collectively, these sources provide drug activity and molecular profiling data for approximately 1,400 distinct cancer cell lines (Figure 1b, Supplementary Data).…”

“…Despite ongoing data acquisition and processing efforts, gaps exist with respect to genomic profiling data in sources other than the CellMiner NCI-60 ( Figure 1b, dark grey table entries). For the GDSC gene mutation and methylation data, we took advantage of processing pipelines developed for the NCI-60 (10,12) to compute gene-level summary data from publicly available raw data. Remaining source-specific molecular profiling data gaps can be filled within CellMinerCDB by effectively extending data provided by one source to another.…”

“…The latest versions of these data can also be downloaded from https://discover.nci.nih.gov/cellminer/loadDownload.do. Detailed information is provided in (1,9,10,36).…”

“…DNA methylation. Data were obtained using the Illumina Infinium Human Methylation 450 platform as described (10).Values lie between 0 (lack of methylation) and 1 (complete methylation). microRNA expression.…”

Schlafen11 Expression Is Associated With the Antitumor Activity of Trabectedin in Human Sarcoma Cell Lines

“…In addition, a previously reported analysis of DNA methylation using the cancer cell line panel NCI60 has identified SLFN11 CpG promoter island hypermethylation as a predictive biomarker of platinum resistance (18,26). Some studies have revealed that the DNA methyltransferase (22).…”