Synthetic lethal screening identifies compounds activating iron-dependent, nonapoptotic cell death in oncogenic-RAS-harboring cancer cells

Abstract

We screened small molecules to identify two compounds, which we named RSL3 and RSL5, that have increased lethality in the presence of oncogenic RAS. Counter screening with biologically active compounds defined aspects of the mechanism of action for RSL3 and RSL5, such as a nonapoptotic, MEK-dependent, and iron-dependent oxidative cell death. Erastin, a previously reported compound with RAS-selective lethality, showed similar properties. RNA interference experiments targeting voltage-dependent anion channel 3 (VDAC3), a target of erastin, demonstrated that RSL5 is a scaffold that acts through VDACs to activate the observed pathway. RSL3 activated a similar death mechanism but in a VDAC-independent manner. We found that cells transformed with oncogenic RAS have increased iron content relative to their normal cell counterparts through upregulation of transferrin receptor 1 and downregulation of ferritin heavy chain 1 and ferritin light chain.

Figure 1. Overview of Our Synthetic Lethal Screening Strategy

(A) Two genes (in this case, RAS and Protein X, a target of a small molecule) are synthetic lethal if perturbations in either alone has little consequence to the cell, but simultaneous perturbation of both causes death. Perturbation of protein X can be produced by either genetic mutation or small-molecule treatment. (B) Characteristics of isogenic cell lines used for synthetic lethal screening. Sequential introduction of oncogenic elements converts human primary fibroblast cells into tumorigenic cells. Phenotypic changes associated with expression of each genetic element are indicated along with cell doubling time. Both RAS^V12 harboring cells (BJ-TERT/LT/ST/RAS^V12 and BJ-TERT/p53DD/CDK4/cyclinD1/ST/RAS^V12 cells) can form tumors in nude mice but differ in cell doubling time, which allows filtering out proliferation-rate-dependent lethal compounds. (C) Flow chart of our screening strategy. (D) Structure of RSL5 and RSL3.

Figure 2. Effect of RSL5 on Four BJ-Derived Cell Lines and Results of Counter Screening with Bioactive Molecules

(A) BJ-TERT, BJ-TERT/LT/ST, BJ-TERT/LT/ST/RAS^V12, and DRD cells were treated with RSL5 in 384-well plates for 48 hr. Error bars indicate one standard deviation of triplicate data. (B and C) BJ-TERT/LT/ST/RAS^V12 cells were treated with RSL5 in the presence of the indicated bioactive molecules in 384-well plates for 24 hr. The concentration of each bioactive compound is listed in Table S1. Percent growth inhibition was measured with alamar blue. The graph is a representative outcome of multiple experiments.

Figure 3. Effect of RSL3 on Four BJ-Derived Cell Lines and Results of Counter Screening with Bioactive Molecules

(A) BJ-TERT, BJ-TERT/LT/ST, BJ-TERT/LT/ST/RAS^V12, and DRD cells were treated with RSL3 in 384-well plates for 48 hr. Data represents the mean ± SD of triplicate samples. (B and C) BJ-TERT/LT/ST/RAS^V12 cells were treated with RSL3 in the presence of indicated bioactive compounds in 384-well plates for 24 hr. The concentration of each bioactive compound was listed in Table S1. Percent growth inhibition was measured with alamar blue. The graph is a representative outcome of multiple experiments.

Figure 4. RSL5 and Erastin Utilize VDAC3 to Create Synthetic Lethality with RAS

HT1080 cells were infected with lentivirus-encoding shRNAs targeting VDAC3 and treated with erastin (A), RSL5 (B), or RSL3 (C). Expression of shRNAs targeting VDAC3 prevented cell death caused by RSL5 and erastin, whereas the sensitivity of RSL3 was not changed. Percent growth inhibition was determined with alamar blue. Error bars indicate one standard deviation of triplicate data.

Figure 5. RSL3 Induces Rapid and Nonapoptotic Cell Death in Oncogenic RAS Containing Tumorigenic Cells

(A) Kinetic analysis of cell death induced by erastin, RSL5, and RSL3. RSL3 induced growth inhibition as early as 8 hr after drug treatment. Cells were seeded in 6-well plates in the presence of 5 µg/ml of erastin, 20 µg/ml of RSL5, or 0.5 µg/ml of RSL3. Cell viability was determined by Trypan Blue exclusion after 4, 8, 11, 12, and 13 hr of drug treatment. The graph is a representative outcome of twice experiments. (B) Time-dependent effect of RSL3 on BJ-TERT, BJ-TERT/LT/ST, and BJ-TERT/LT/ST/RAS^V12 cells. Cells were seeded in 6-well plates in the presence of the increasing concentration of RSL3. After 24, 48, and 72 hr, the number of viable cells was determined by using a Vi-Cell, and percent growth inhibition was calculated. Bar graph is a representative outcome of multiple independent experiments. (C) Photograph of BJ-TERT and BJ-TERT/LT/ST/RAS^V12 cells treated with 5 µg/ml of erastin, 20 µg/ml of RSL5, 0.5 µg/ml of RSL3, or 1 µM of staurosporine for 24 hr. (D) Staurosporine-treated but not erastin-treated, RSL5-treated, or RSL3-treated BJ-TERT/LT/ST/RAS^V12 cells contain the PARP-1 fragment produced by caspase-3/7 cleavage. Lysates of each treated samples were prepared. Ten micrograms of each lysate was analyzed by western blot with an antibody directed against human PARP-1.

Figure 6. The Selective Lethality of RSL3 Is Not Confined to Mutant HRAS-Containing Cancer Cell Lines

(A) RSL3 was active in killing four KRAS mutation harboring cancer cells. A549, Calu-1, PaCa-2, and HCT116 cells were treated with RSL3 in 384-well plates for 24 hr. Percent growth inhibition was calculated with alamar blue. Error bars indicate one standard deviation of triplicate data. (B) The activity of RSL3 was tested in 60 different human cancer cell lines. The left chart shows the result of 18 cell lines consisting of representative sensitive and resistant groups. Zero viability indicates a cytostatic effect of a compound, whereas negative viability indicates a cytotoxic effect. The right table shows results of the COMPARE analysis for paclitaxel and RSL3. The sensitivity profile of paclitaxel across 60 cancer cell lines has a strong correlation with those of other microtubule modulating agents. However, the top scored PCC value for RSL3 was 0.503, suggesting that RSL3 has a distinct sensitivity profile compared to other anticancer agents. (C) Attenuating RAS signaling in Calu-1 cells by infecting with lentivirus containing shRNAs targeting H, N, or KRAS confers resistance to RSL3-induced lethality. Virus-infected cells were seeded in 384-well plates and treated with RSL3 for 24 hr. Percent growth inhibition was determined with alamar blue. Error bars indicate one standard deviation of triplicate data.

Figure 7. Oncogenic RAS Signaling Increases the Cellular Iron Content by Modulating Expression of TfR1, FTH, and FTL

(A) Iron chelators with different structures were able to suppress erastin-induced, RSL5-induced, and RSL3-induced cell death. BJ-TERT/LT/ST/RAS^V12 cells were treated with erastin (left), RSL5 (center), or RSL3 (right) in the presence or absence of iron chelators. Value represents the mean ± SD of triplicate data. (B) BJ-TERT/LT/ST/RAS^V12 cells have a greater basal iron pool than their isogenic counterparts. BJ-TERT, BJ-TERT/LT/ST, BJ-TERT/LT/ST/RAS^V12 cells were stained with PGSK, and the fluorescence profile of stained cells was analyzed by flow cytometry. (C) BJ-TERT/LT/ST/RAS^V12 cells have more TfR1 protein than that in their isogenic counterparts. Lysates of BJ-TERT, BJ-TERT/LT/ST, and BJ-TERT/LT/ST/RAS^V12 cells were analyzed by western blot with an antibody directed against TfR1. We included an antibody directed against actin for assessing protein loading amount. (D–E) Knocking down TfR1 in BJ-TERT/LT/ST/RAS^V12 cells partially inhibits cell death induced by erastin. Whole-cell lysates from BJ-TERT/LT/ST/RAS^V12 cells infected with either nontargeting shRNA (N.T.) or three different TfR1-targeting shRNAs (658, 659, 660) were prepared, and the level of TfR1 in each cell line was determined by western blot with an antibody against TfR1 (D). The same set of cell lines was treated with indicated concentration of erastin and percent growth inhibition was determined with alamar blue (E). Error bars indicate one standard deviation of triplicate data. (F) Oncogenic RAS upregulates TfR1 but downregulates FTH1 and FTL. Cellular RNAs were prepared from each cell line, and real-time PCR was performed with each gene-specific primer set. The expression levels of TfR1, FTH1, and FTL were first normalized to the level of endogenous control (RPLPO). The relative expression level of each gene among the three cell lines was expressed as a ratio of transcripts in a cell line to those in BJ-TERT. Value represents the mean ± SD of three measurements of each sample.