Identification of genotype-selective antitumor agents using synthetic lethal chemical screening in engineered tumor cells
Sona Dolma, Stephen L. Lessnick, William C. Hahn, and Brent R. Stockwell
Cancer Cell, Vol 3, 285-296, March 2003
This report is published as a partial requirement for the graduate class in Oncogenomics, Genetics 144 taught by Dr. Charles Brenner, Dartmouth Medical School, Winter Term 2005.
The course is based on material covered in Oncogenomics: Molecular Approaches to Cancer.
PLEASE NOTE: Online subscription to Cancer Cell may be required to view some material on this page.
In a recent issue of Cancer Cell, Brent Stockwell and his colleagues at the Whitehead Institute for Biomedical Research used a novel screening technique to survey more than 23,000 chemical compounds and test for their ability to kill transformed human cells. In particular, Stockwell's group searched for compounds that were lethal to cancer cells while leaving normal healthy cells unharmed. The group identified nine compounds matching this profile, including many currently in use clinically and one previously unidentified compound which they named Erastin.
Compounds which exhibit synthetic lethalithy are lethal to tumor cells only. Specificty is determined by the presence of specific oncoproteins or by the loss of specific tumor suppressor proteins, which, these compounds modulate to confer synthetic lethality to the cell . Such agents need not target the specific proteins involved but may target proteins in the oncoprotein-linked signalling pathway. Current compounds which display a synthetic lethality include rapamycin which has been shown to be synthetically lethal to cells which are deficient fot the tumor suppressor PTEN also Imatinib Mesylate (Gleevec®) which specifically blocks the fusion protein, BCR:ABL.
To search for genotype selective antitumor agents primary fibroblast cells were transformed by introducing specific genetic elements known to be involved in oncogenesis such as hTERT, the viral oncogenes LT & ST from SV40, E6, E7 from HPV and hRASV12. By using these cell lines Dolma et al could test libraries of compounds ability to growth inhibit/kill cells of known oncogenic genotypes. Furthermore by varying the genetic elements used to transform the cells the group could assess which drug was most effective against a particular oncogenic genotype (i.e. introducing specific oncoproteins or inhibiting specific tumor suppressors) . In doing so the group could identify genotype selective antitumor agents exhibiting synthetic lethality. The ability of genotype selective compounds to serve as molecular probes is a rapidly expanding field in molecular biology/pharmacology. Such research is based on the premise of chemical genetics-that small molecules can be used to identify proteins and pathways underlying biological effects one such example is rapamycin, which lead to the discovery of target of rapamycin (TOR).
The authors made use of various primary cell lines which had being transformed through the introduction of hTERT and a various combinations of of E6, E7, SV40 (small T & large T) and hRASV12. LT blocks both p53 and Rb. E6 blocks p53 only. E7 blocks Rb only. Therefore by using different combinations of genetic elements they could delineate activation/inactivation of different oncogenic pathways. Using this approach the authors could screen a library of 23,550 compounds and elucidate whivh vompounds were most effective against which particular oncogenic phenotype.
Click here for a schematic of the screen used
Initially compounds that showed a 50% or greater inhibition of calcein staining (as a marker of cell viability) were selected for a 2 fold dilution series in the tumorigenic BJ-TERT/LT/ST/RASv12 cells and its non-transformed parent cell line, BJ. This screen was used to detect compounds with synthetic lethality i.e. compounds lethal to tumorigenic cells but not in isogenic primary cells. The authors found 9 compounds that were at least 4 times more potent against the BJ-TERT/LT/ST/RASv12 over its parent BJ cell line.
These nine compounds were subsequently tested at multiple doses in each panel of engineered cell lines. IC50 values for each compound in each cell line were calculated and used to determine whether a particular genotype was particularly sensitive to a certain compound. This was an effort to determine the genetic basis for selectivity of each compound, comparing IC50 values between cells of known genotype. They found that the 9 compounds could be categorized into 3 groups. Category one consists of four compounds sangivamycin, bouvardin, NSC146109, and echinomycin. These four compounds had no selectivity for a particular genotype. Their activity simply appeared to correlate with increased doubling tme of the cells. Therefore these compounds appear to be selective for rapidly dividing cells. Supporting this theory all agents in this category act by inhibiting DNA or inhibiting protein synthesis.
The compounds in group 2, of mitoxantrone, doxorubicin, and daunorubicin. became more potent when hTERT was introduced and even more potent when Rb is inactivated by either SV40-LT or HPV-E7. Two of the compounds in this category are anthracyclines (doxorubicin & daunorubicin). These compounds are topoisomerase II poisons, which bind to topoisomerase II and DNA and prevent the religation of double-strand DNA breaks introduced by topoisomerase II. Further investigation into the molecular events following the introduction of hTERT and inactivation of Rb showed that this results in increased expression of TOPO II. Therefore these compounds were more effective in cell lines where their putative target was expressed at a higher level.
Category three consists of two drugs, camptothecin (CPT) & erastin. Efficient CPT & erastin-induced cell death required the presence of both ST & RASv12. Although CPT & erastin have similar genetic basis of selectivity their mechanism of action differs, CPT is partially active in cells lacking Rb whereas erastin is not . The investigators found that cells expressing both RASv12 and ST upregulated the putative target of CPT, topoisomerase 1 (TOPO I). To extrapolate CPT and erastin῭s mechanism of action the investigators treated the cells with the phosphatase inhibitor okadaic acid which is known to induce TOPO I. Treatment with OA also sensitized the cells to CPT. Furthermore RNAi directed toward TOPO I signifigantly reduced CPT's effects in the ST & RASv12 cell line. Camptothecin induces apoptosis in BJ-TERT/LT/ST/RASv12 tumorigenic cells. Erastin's mechanism of action differed, it's ability to induce non apoptotic cell death is limited to BJ-TERT/LT/ST/RASv12 tumorigenic cells. Therefore it is synthetically lethal to cells harboring this genotype.
The identification of molecular genotype(s) which enhance a known or develpomental compounds clinical efficacy would be benefical to the design of treatment regimens for cancer patients. In this paper Dolma et al have provided a unique insight into the molecular mechanisms of compounds currently in clinical use (CPT and anthracyclines) and identified specific oncogenic pathways, which, if present in a tumor may increase the efficacy of these compounds. They have discovered a unique compound, erastin which exhibits synthetic lethalithy to cells with oncogenic ST & RASv12.
Although the authors limited their studies to hTERT, LT, ST, E6, E7, and RASv12 as transforming genes, future studies can make use of a wide variety of cancer-associated alleles. The authors conclude the paper by suggesting that by using such a methodology it may be possible to define the signaling networks that emanate from many oncogenes and tumor suppressors. Such studies may ultimately unravel intricate details of these and other critical signaling networks altered by oncogenic mutations and suggest drugs to target them.