The NCI-60 cell lines, which represent cancers of the lung, colon, brain, ovary, breast, prostate, and kidney, as well as leukemia and melanoma, are the most frequently studied human tumor cell lines in cancer research and have generated the most extensive cancer pharmacology database worldwide. As recently reported by Ogan D. Abaan, PhD, postdoctoral fellow at the National Cancer Institute, and colleagues in Cancer Research, NCI investigators have performed a comprehensive analysis of coding variants in the NCI-60 panel identified by whole-exome sequencing.1
This work provides a list of possible cancer-specific variants for use by the research community and identifies pharmacogenomic correlations between specific variants in cancer-related genes and anticancer agents as an illustration of how such data can be used to generate and validate hypotheses for therapeutic strategies. Some of the basic and representative findings in this research effort are described below.
The whole-exome sequencing variants identified in the NCI-60 lines were divided into two groups: type I variants (> 1.2 million) corresponding to common (and possibly germline) variants and type II variants (60,005) enriched for acquired cancer-specific variants. The overall pattern of mutation was divergent among cell lines, ranging from 172 to 9,205 type II variants.
Cell lines with known microsatellite instability had very high type II variant counts, although the colon cancer cell line HCC2998, which is not known to have microsatellite instability, had the highest number of type II variants. In contrast to known microsatellite instability cell lines, nearly all of these variants were single-nucleotide variants, suggesting that the hypermutation arises from a non–microsatellite instability mechanism.
HCC2998 harbors a POLE exonuclease domain missense variant coding for a P286R mutation in POLε. Other data indicate that impaired POLε proofreading results in a high rate of single-nucleotide substitutions and increased tumor formation, and POLE mutations have been reported in colorectal cancer. HCC2998 thus seems to exemplify this phenomenon—providing a tool for further research and exemplifying the utility of the whole-exome sequencing data.
Analysis of the NCI-60 whole-exome sequencing variants according to gene mutations already present in the Catalogue of Somatic Mutations in Cancer (COSMIC) database showed that most of the variants in the 10 most frequently mutated genes are already annotated in COSMIC, but numerous novel variants in the 10 genes were also observed. Absence of normal tissue makes it impossible to validate these variants as somatic, but they are also not found in other established databases.
In addition to mutational data, the study provides drug sensitivity data for tens of thousands of compounds. It profiles the influence of specific variants for TP53, BRAF, KRAS, NRAS, PIK3CA, PTEN, and ERBBs on response to clinically relevant targeted agents (nutlin, vemurafenib [Zelboraf], selumetinib, hypothemycin, rapamycin, wortmannin, perifosine, erlotinib [Tarceva], afatinib [Gilotrif], lapatinib [Tykerb], and neratinib) and identifies aspects of the findings that merit further investigation.
For example, in one set of studies, activity of compounds was assessed for each gene with at least five cell lines containing a type II variant. TP53, the most frequently mutated gene in the NCI-60 panel, showed strong correlation with drug response. MDM2 inhibitors were effective in cell lines with wild-type p53, with two clinically relevant MDM2 inhibitors (including nutlin-3) showing the greatest negative correlation with mutant p53.
In contrast, another compound (NSC-670177) showed significant selectivity for p53 mutant cells. However, the compound known as RITA (NSC-652287), initially identified as a DNA cross-linking agent, showed little evidence of selective activity in p53 wild-type cell lines status and limited correlation with nutlin-3 activity. These findings raise doubt about claims that RITA acts specifically as a p53-reactivating compound and that it could potentially be useful for induction of tumor cell apoptosis.
As an example of how the entire NCI-60 database can be used for querying drug response parameters, the investigators cited analysis of the EGFR inhibitor erlotinib, which has activity that is highly correlated with gefitinib (Iressa) and lapatinib in the NCI-60 panel. Overall, high expression of EGFR and ERBB2 were determinants of NCI-60 cell line response to erlotinib.
However, the colon and central nervous system cell lines were generally insensitive to erlotinib despite high EGFR and ERBB2 expressions. This can be explained by taking into account mutations in the MAPK or PI3K pathways, a common mechanism of resistance, which are present in all seven colon and four of six central nervous system NCI-60 cell lines.
Potential Effects of Common Variants on Drug Response
As noted by the investigators, the power of whole-exome sequencing (ie, instead of focused sequencing of preselected genes) was also shown by the coincidental finding of a significant correlation between a germline in-frame deletion (delCAATGT) in ATAD5 in some cell lines and the increased sensitivity of these lines to the DNA-damaging agent bleomycin and bleomycin analogs—a finding showing that common variants in the human population may have a significant effect on drug response. ATAD5 is essential for maintaining genome stability and is known to be mutated in endometrial cancer.
Genotype calls identified 10 cell line heterozygous (5) or homozygous (5) for the delCAATGT deletion, 3 of which are renal cell lines (ACHN, CAKI-1, RXF-393) in which early studies suggest activity of dimethane sulfonate analogs. It was also observed that additional germline variants in ATAD5 are present exclusively in the same 10 cell lines.
In searching possible haplotypes in the HapMap database, the investigators discovered a region of linkage-disequilibrium spanning more than 300 kb. They noted, “[T]his particular haplotype could be a response modifier during chemotherapy with DNA-damaging agents. These results illustrate the discovery potential of exonic variant data when integrated with previously available NCI-60 databases.”
The investigators concluded: “As new cancer genes are identified through large-scale sequencing studies, the data presented here for the NCI-60 will be an invaluable resource for identifying cell lines with mutations in such genes for hypothesis-driven research. To enhance the utility of the data for the greater research community, the genomic variants are freely available in different formats and from multiple sources including the CellMiner [discover.nci.nih.gov/cellminer] and Ingenuity [www.ingenuity.com] websites.” ■
Disclosure: For full disclosure of the study authors, visit cancerres.aacrjournals.org.
1. Abaan OD, Polley EC, Davis SR, et al: The exomes of the NCI-60 panel. Cancer Res 73:4372-4382, 2013.