Cross-species identification of PIP5K1-, splicing- and ubiquitin-related pathways as potential targets for RB1-deficient cells

Abstract

The RB1 tumor suppressor is recurrently mutated in a variety of cancers including retinoblastomas, small cell lung cancers, triple-negative breast cancers, prostate cancers, and osteosarcomas. Finding new synthetic lethal (SL) interactions with RB1 could lead to new approaches to treating cancers with inactivated RB1. We identified 95 SL partners of RB1 based on a Drosophila screen for genetic modifiers of the eye phenotype caused by defects in the RB1 ortholog, Rbf1. We validated 38 mammalian orthologs of Rbf1 modifiers as RB1 SL partners in human cancer cell lines with defective RB1 alleles. We further show that for many of the RB1 SL genes validated in human cancer cell lines, low activity of the SL gene in human tumors, when concurrent with low levels of RB1 was associated with improved patient survival. We investigated higher order combinatorial gene interactions by creating a novel Drosophila cancer model with co-occurring Rbf1, Pten and Ras mutations, and found that targeting RB1 SL genes in this background suppressed the dramatic tumor growth and rescued fly survival whilst having minimal effects on wild-type cells. Finally, we found that drugs targeting the identified RB1 interacting genes/pathways, such as UNC3230, PYR-41, TAK-243, isoginkgetin, madrasin, and celastrol also elicit SL in human cancer cell lines. In summary, we identified several high confidence, evolutionarily conserved, novel targets for RB1-deficient cells that may be further adapted for the treatment of human cancer.

Fig 1. Drosophila synthetic lethality eye screen in Rb1-deficient background.

Drosophila eye pictures of the following genotypes: GMR-Gal4> no RNAi (A) GMR-Gal4> Cont-i (B) GMR-Gal4> Rbf1-i (C). Scanning electron microscope images of a Drosophila eye of the following phenotypes: GMR-Gal4> no RNAi (D) GMR-Gal4> Cont-i (E) GMR-Gal4> Rbf1-i (F). Drosophila eye picture of the following phenotypes: GMR-Gal4> Ran RNAi (G), GMR-Gal4> SkpA RNAi (H), GMR-Gal4> Cdc27 RNAi (I), GMR-Gal4> Rbf1-i, Ran RNAi (J), GMR-Gal4> Rbf1-i, SkpA RNAi (K), GMR-Gal4> Rbf1-i, Cdc27 RNAi (L). GO enrichment analysis of Drosophila eye hits (M).

Fig 2. Testing human gene orthologs of the Drosophila hits in available large-scale shRNA and CRISPR human cancer cell line screens.

Data analysis workflow (A). Scatter plots illustrating Z scores in DRIVE, AVANA, COLT, DEPMAP data sets in RB1 wild-type (WT) and mutant (ALT) human cancer cell lines for candidate RB1 SL partners: SKP2 (B), EIF4A3 (C), SNRPF (D), CCNA2 (E), CDC27 (F), U2AF2 (G), RAN (H), and UBA1 (I).

Fig 3. Identifying synthetic lethal partners of RB1 in human cancer patients.

(A) We analyzed the overall survival of the TCGA cancer patients by comparing the patients with low activity of both genes (bin 1) against the low activity of RB1 and high activity of candidate partner gene (bin7) in pooled lung, breast, and prostate cancers patients. (B) Box plot comparing the relative hazard computed using log-rank test for different gene sets. Gene sets from left to right represent (i) 17155 genes not used in the Drosophila screen, (ii) the human orthologs of 1300 genes used in the Drosophila screen, (iii) human orthologs of 95 genes identified as RB1 SL partners in the Drosophila screen, and (iv) 18 genes identified as RB1 SL partners in cancer cell lines. (C) Volcano plot showing relative hazard and–log10(p-value) for human orthologs of the genes that we screened in Drosophila, with human orthologs of hits from the Drosophila eye screen and cell line hits shown with blue and red colors, respectively. The p-value was adjusted for multiple comparisons using Benjamini-Hochberg’s FDR method. Dotted horizontal line indicates the FDR of 10%, arrows indicate direction of survival and the numbers below arrows indicate the number of genes having better or worst survival in conjunction with low RB1. (D) Kaplan Meier’s survival curves along with corrected p-values for six genes which showed SL with RB1 in cancer patients and in at least two human cancer cell lines screens. Blue line–samples in bin 1, Red line–samples in bin 7.

Fig 4. Protein network analysis of synthetic lethal RB1 partners from Drosophila, human cancer cell lines, and human cancer patients.

(A) PPI network of Drosophila eye hits (marked in blue color), positive controls from the literature (marked in violet) and Rbf gene (marked in orange). (B) PPI network of the full list of human orthologs of Drosophila eye hits and positive controls from literature; subnetwork 1 contains proteins that also scored as SL partners in human cancer cell lines (marked in green or red); subnetwork 2 contains proteins that also scored as SL partners in human cancer patients (marked in yellow or red nodes); the common genes from subnetwork 1 and 2 are marked in red. (C, D) Top 20 drugs from the connectivity map analysis that scored from the analysis of the full list of human orthologs of Drosophila eye hits and positive controls from literature (C); and from the analysis of subnetwork 1 that contains proteins that also scored as SL partners in human cancer cell lines (D). Columns—cMap cell lines, rows–drugs. Decreased gene expression signature is marked in blue and increased–in red. *—marks an average of gene expression signature between all tested cell lines. Scatter plots illustrating Z scores for SN-38 (E), topotecan (F), and camptothecin (G) in RB1 wild-type and mutant cell lines retrieved from CTRP.

Fig 5. Creation of a novel Rbf1/Pten/Ras Drosophila cancer model.

(A) Relative mRNA levels of Pten and Rbf1 in tubulin-Gal4 larvae expressing either control RNAi, Rbf1 RNAi, Pten RNAi or double Rbf1, Pten RNAi. Means ± SD. **p<0.01, ***p<0.001 (B) Relative luciferase levels in UAS-Luciferase,esg-Gal4,tubulinGal80ts,UAS-GFP female flies expressing either control RNAi, Rbf1 RNAi, Pten RNAi, double Rbf1,Pten RNAi, activated UAS-Ras1A, or combined Rbf1,Pten RNAi,Ras1A. Means ± SD. **p<0.01, ***p<0.001. Maximum projection of confocal images of posterior midgut (C, D) and Malpighian tubules (E) of UAS-Luciferase, esg-Gal4, tubulinGal80ts, UAS-GFP flies crossed to control RNAi (C) or combined Rbf1, Pten RNAi,Ras1A (D, E). Whole fly pictures under fluorescent microscope of UAS-Luciferase, esg-Gal4, tubulinGal80ts, UAS-GFP flies crossed to either control RNAi or combined Rbf1, Pten RNAi, Ras1A (F). Lifespan analysis of UAS-Luciferase, esg-Gal4, tubulinGal80ts, UAS-GFP flies crossed to either control RNAi, Rbf1 RNAi, Pten RNAi, double Rbf1, Pten RNAi, activated UAS-Ras1A, or combined Rbf1, Pten RNAi, Ras1A (G). Absolute (top) and relative (bottom) luciferase levels in two tester lines: UAS-Luciferase, esg-Gal4, tubulinGal80ts, UAS-GFP lines containing either control RNAi (control) or combination of Rbf1 RNAi, Pten RNAi, Ras1A overexpression and crossed to either control RNAi or RNAi against 11 genes and E2F and Skp2 as positive controls. Four biological replicates/genotype. (H).

Fig 6. RB1-deficient cancer cells are sensitive to chemical inhibitors of PIP5K1-, ubiquitin- and splicing- related pathways.

Proliferation of RB1 wild-type and RB1 mutant prostate cancer cells treated with 2.5, 1.25, or 0.625 uM of Madrasin (A) or 50, 25, or 12.5 uM of Isoginkgetin (B) for 96 hr (crystal violet staining). Data are shown as means ± SD. Proliferation of RB1 wild-type and RB1 mutant breast cancer cells treated with 50, 25, or 12.5 uM of Skp2 inhibitor (C), 10 or 5 uM of UNC3230 (D), 25 or 12.5 uM of PYR-41 (E) or 0.1, 0.05, or 0.025 uM of TAK243 (F) for 96 hr (crystal violet staining). Data are shown as means ± SD. Proliferation of RB1 wild-type and RB1 mutant osteosarcoma cells treated with 25, 12.5 or 6.25 uM of PYR-41 (G) or 0.1, 0.05, or 0.025 uM of TAK243 (H) for 96 hr (crystal violet staining). Data are shown as means ± SD.

Fig 7. Overview of the findings presented in this study.