FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2736544909> ?p ?o ?g. }

• W2736544909 endingPage "223.e6" @default.
• W2736544909 startingPage "210" @default.
• W2736544909 abstract "•Combined CRISPRi/a chemical-genetic screening reveals targets of therapeutic agents•Focused chemical-genetic profiling rapidly classifies agents by mechanism of action•Chemical-genetic screens with rigosertib reveal a microtubule-destabilizing signature•Targeted in vivo and in vitro approaches confirm rigosertib’s mechanism of action Chemical libraries paired with phenotypic screens can now readily identify compounds with therapeutic potential. A central limitation to exploiting these compounds, however, has been in identifying their relevant cellular targets. Here, we present a two-tiered CRISPR-mediated chemical-genetic strategy for target identification: combined genome-wide knockdown and overexpression screening as well as focused, comparative chemical-genetic profiling. Application of these strategies to rigosertib, a drug in phase 3 clinical trials for high-risk myelodysplastic syndrome whose molecular target had remained controversial, pointed singularly to microtubules as rigosertib’s target. We showed that rigosertib indeed directly binds to and destabilizes microtubules using cell biological, in vitro, and structural approaches. Finally, expression of tubulin with a structure-guided mutation in the rigosertib-binding pocket conferred resistance to rigosertib, establishing that rigosertib kills cancer cells by destabilizing microtubules. These results demonstrate the power of our chemical-genetic screening strategies for pinpointing the physiologically relevant targets of chemical agents. Chemical libraries paired with phenotypic screens can now readily identify compounds with therapeutic potential. A central limitation to exploiting these compounds, however, has been in identifying their relevant cellular targets. Here, we present a two-tiered CRISPR-mediated chemical-genetic strategy for target identification: combined genome-wide knockdown and overexpression screening as well as focused, comparative chemical-genetic profiling. Application of these strategies to rigosertib, a drug in phase 3 clinical trials for high-risk myelodysplastic syndrome whose molecular target had remained controversial, pointed singularly to microtubules as rigosertib’s target. We showed that rigosertib indeed directly binds to and destabilizes microtubules using cell biological, in vitro, and structural approaches. Finally, expression of tubulin with a structure-guided mutation in the rigosertib-binding pocket conferred resistance to rigosertib, establishing that rigosertib kills cancer cells by destabilizing microtubules. These results demonstrate the power of our chemical-genetic screening strategies for pinpointing the physiologically relevant targets of chemical agents. The ready availability of genomic sequence information, combined with conceptual advances in our understanding of the molecular etiology of diseases, is enabling precision medicine efforts, which seek to develop rational therapies that specifically address the molecular and genetic basis of a disease (Ashley, 2016Ashley E.A. Towards precision medicine.Nat. Rev. Genet. 2016; 17: 507-522Crossref PubMed Scopus (415) Google Scholar). Critical to these efforts are therapeutic agents with well-defined targets and high specificity for these targets as well as a comprehensive understanding of how the efficacy of these agents is affected by different genetic backgrounds. Identifying the targets, off-target activities, and genetic dependencies of chemical agents, however, remains one of the principal obstacles in drug development (Nijman, 2015Nijman S.M.B. Functional genomics to uncover drug mechanism of action.Nat. Chem. Biol. 2015; 11: 942-948Crossref PubMed Scopus (59) Google Scholar). This obstacle has hampered the use of natural products and small molecules identified as lead compounds in cell-based screens, creating an urgent need for methods that enable accurate and comprehensive characterization of mechanisms of action of small molecules to guide further development and treatment applications. Hypothesis-free evaluation of a molecule’s mechanism of action using systematic genetic screening provides a potential solution to these challenges (Ho et al., 2011Ho C.H. Piotrowski J. Dixon S.J. Baryshnikova A. Costanzo M. Boone C. Combining functional genomics and chemical biology to identify targets of bioactive compounds.Curr. Opin. Chem. Biol. 2011; 15: 66-78Crossref PubMed Scopus (58) Google Scholar, Smith et al., 2010Smith A.M. Ammar R. Nislow C. Giaever G. A survey of yeast genomic assays for drug and target discovery.Pharmacol. Ther. 2010; 127: 156-164Crossref PubMed Scopus (95) Google Scholar). Extensive efforts in yeast have yielded high-throughput chemical-genetic methods for identifying the molecular targets of drugs, in which drug sensitivity is systematically profiled for a library of strains with increased or decreased gene expression levels (reviewed in Ho et al., 2011Ho C.H. Piotrowski J. Dixon S.J. Baryshnikova A. Costanzo M. Boone C. Combining functional genomics and chemical biology to identify targets of bioactive compounds.Curr. Opin. Chem. Biol. 2011; 15: 66-78Crossref PubMed Scopus (58) Google Scholar, Smith et al., 2010Smith A.M. Ammar R. Nislow C. Giaever G. A survey of yeast genomic assays for drug and target discovery.Pharmacol. Ther. 2010; 127: 156-164Crossref PubMed Scopus (95) Google Scholar). These methods are based on the principle that sensitivity to a drug is generally correlated to the expression levels of the components of the pathway it targets. Combining knockdown and overexpression profiling has provided particular utility, as both the direct target and modifiers of sensitivity can be identified with high precision (Hoon et al., 2008Hoon S. Smith A.M. Wallace I.M. Suresh S. Miranda M. Fung E. Proctor M. Shokat K.M. Zhang C. Davis R.W. et al.An integrated platform of genomic assays reveals small-molecule bioactivities.Nat. Chem. Biol. 2008; 4: 498-506Crossref PubMed Scopus (157) Google Scholar). In human cells, knockdown and knockout screens have already aided the identification of drug targets and of biomarkers predictive of responsive patient populations in selected cases (reviewed in Kampmann, 2017Kampmann M. Elucidating drug targets and mechanisms of action by genetic screens in mammalian cells.Chem. Commun. (Camb.). 2017; 53: 7162-7167Crossref PubMed Google Scholar), and the value of overexpression has been demonstrated in targeted approaches, for example in validating the AAA-ATPase p97 (VCP) as the target of CB-5083 (Anderson et al., 2015Anderson D.J. Le Moigne R. Djakovic S. Kumar B. Rice J. Wong S. Wang J. Yao B. Valle E. Kiss von Soly S. et al.Targeting the AAA ATPase p97 as an approach to treat cancer through disruption of protein homeostasis.Cancer Cell. 2015; 28: 653-665Abstract Full Text Full Text PDF PubMed Scopus (244) Google Scholar), underscoring the potential of these methods for drug target identification. With the advent of CRISPR-based screening platforms for inhibiting (CRISPRi) or activating (CRISPRa) gene expression (reviewed in Dominguez et al., 2016Dominguez A.A. Lim W.A. Qi L.S. Beyond editing: repurposing CRISPR-Cas9 for precision genome regulation and interrogation.Nat. Rev. Mol. Cell Biol. 2016; 17: 5-15Crossref PubMed Scopus (542) Google Scholar), it has now become feasible to combine genome-wide overexpression and knockdown screens, laying the groundwork for a systematic chemical-genetic strategy to identify the targets of drug candidates and other small molecules in human cells. The challenges in identifying the targets of drug candidates are clearly illustrated in the development of rigosertib (Estybon, ON 01910.Na; Figure 1A), a promising small molecule under clinical evaluation as an anti-cancer drug whose molecular target and mechanism of action remain unresolved. Rigosertib’s promise stems from its cytotoxic activity against a broad range of cancer cell lines, inducing mitotic arrest and apoptosis in these cells, as well as potent inhibition of tumor growth in mouse models (Gumireddy et al., 2005Gumireddy K. Reddy M.V.R. Cosenza S.C. Boominathan R. Baker S.J. Papathi N. Jiang J. Holland J. Reddy E.P. ON01910, a non-ATP-competitive small molecule inhibitor of Plk1, is a potent anticancer agent.Cancer Cell. 2005; 7: 275-286Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar, Reddy et al., 2011Reddy M.V.R. Venkatapuram P. Mallireddigari M.R. Pallela V.R. Cosenza S.C. Robell K.A. Akula B. Hoffman B.S. Reddy E.P. Discovery of a clinical stage multi-kinase inhibitor sodium (E)-2-2-methoxy-5-[(2′,4′,6′-trimethoxystyrylsulfonyl)methyl]phenylaminoacetate (ON 01910.Na): synthesis, structure-activity relationship, and biological activity.J. Med. Chem. 2011; 54: 6254-6276Crossref PubMed Scopus (75) Google Scholar). Rigosertib was originally identified in a screen for inhibitors of polo-like kinase 1 (PLK1) (Gumireddy et al., 2005Gumireddy K. Reddy M.V.R. Cosenza S.C. Boominathan R. Baker S.J. Papathi N. Jiang J. Holland J. Reddy E.P. ON01910, a non-ATP-competitive small molecule inhibitor of Plk1, is a potent anticancer agent.Cancer Cell. 2005; 7: 275-286Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar), but comparison of the cellular phenotypes elicited by treatment with rigosertib and the well-characterized PLK1 inhibitor BI2536 revealed marked differences (Steegmaier et al., 2007Steegmaier M. Hoffmann M. Baum A. Lénárt P. Petronczki M. Krssák M. Gürtler U. Garin-Chesa P. Lieb S. Quant J. et al.BI 2536, a potent and selective inhibitor of polo-like kinase 1, inhibits tumor growth in vivo.Curr. Biol. 2007; 17: 316-322Abstract Full Text Full Text PDF PubMed Scopus (674) Google Scholar). It was also proposed that rigosertib inhibits phosphatidylinositol 3-kinase (PI3K) signaling (Prasad et al., 2009Prasad A. Park I.W. Allen H. Zhang X. Reddy M.V. Boominathan R. Reddy E.P. Groopman J.E. Styryl sulfonyl compounds inhibit translation of cyclin D1 in mantle cell lymphoma cells.Oncogene. 2009; 28: 1518-1528Crossref PubMed Scopus (64) Google Scholar), but it is unclear whether this effect is through direct inhibition of PI3K. In addition, a microscopy-based screen classified rigosertib with microtubule-destabilizing agents (Twarog et al., 2016Twarog N.R. Low J.A. Currier D.G. Miller G. Chen T. Shelat A.A. Robust classification of small-molecule mechanism of action using a minimalist high-content microscopy screen and multidimensional phenotypic trajectory analysis.PLoS ONE. 2016; 11: e0149439Crossref PubMed Scopus (12) Google Scholar), but in vitro assays of rigosertib’s activity against microtubules have yielded conflicting results (Gumireddy et al., 2005Gumireddy K. Reddy M.V.R. Cosenza S.C. Boominathan R. Baker S.J. Papathi N. Jiang J. Holland J. Reddy E.P. ON01910, a non-ATP-competitive small molecule inhibitor of Plk1, is a potent anticancer agent.Cancer Cell. 2005; 7: 275-286Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar, Lu et al., 2015Lu T. Laughton C.A. Wang S. Bradshaw T.D. In vitro antitumor mechanism of (E)-N-(2-methoxy-5-(((2,4,6-trimethoxystyryl)sulfonyl)methyl)pyridin-3-yl)methanesulfonamide.Mol. Pharmacol. 2015; 87: 18-30Crossref PubMed Scopus (17) Google Scholar, Oussenko et al., 2011Oussenko I.A. Holland J.F. Reddy E.P. Ohnuma T. Effect of ON 01910.Na, an anticancer mitotic inhibitor, on cell-cycle progression correlates with RanGAP1 hyperphosphorylation.Cancer Res. 2011; 71: 4968-4976Crossref PubMed Scopus (30) Google Scholar). Finally, rigosertib was proposed to serve as a molecular RAS mimetic that inhibits RAS signaling (Athuluri-Divakar et al., 2016Athuluri-Divakar S.K. Vasquez-Del Carpio R. Dutta K. Baker S.J. Cosenza S.C. Basu I. Gupta Y.K. Reddy M.V.R. Ueno L. Hart J.R. et al.A small molecule RAS-mimetic disrupts RAS association with effector proteins to block signaling.Cell. 2016; 165: 643-655Abstract Full Text Full Text PDF PubMed Scopus (191) Google Scholar), but it was subsequently argued that rigosertib impacts RAS signaling indirectly through JNK-mediated inactivation of SOS1, B-Raf, and C-Raf (Ritt et al., 2016Ritt D.A. Abreu-Blanco M.T. Bindu L. Durrant D.E. Zhou M. Specht S.I. Stephen A.G. Holderfield M. Morrison D.K. Inhibition of Ras/Raf/MEK/ERK pathway signaling by a stress-induced phospho-regulatory circuit.Mol. Cell. 2016; 64: 875-887Abstract Full Text Full Text PDF PubMed Scopus (58) Google Scholar). Thus, despite a decade of studies and encouraging efficacy and tolerance in early clinical trials (Silverman et al., 2015Silverman L.R. Greenberg P. Raza A. Olnes M.J. Holland J.F. Reddy P. Maniar M. Wilhelm F. Clinical activity and safety of the dual pathway inhibitor rigosertib for higher risk myelodysplastic syndromes following DNA methyltransferase inhibitor therapy.Hematol. Oncol. 2015; 33: 57-66Crossref PubMed Scopus (33) Google Scholar), rigosertib’s molecular target(s) remains undefined, impeding its clinical development. Results from a recent phase 3 clinical trial indicated that rigosertib has no overall benefit in treatment of myelodysplastic syndrome, but detailed analysis suggested that rigosertib could benefit a subgroup of patients, which has motivated an ongoing phase 3 clinical trial (Garcia-Manero et al., 2016Garcia-Manero G. Fenaux P. Al-Kali A. Baer M.R. Sekeres M.A. Roboz G.J. Gaidano G. Scott B.L. Greenberg P. Platzbecker U. et al.ONTIME study investigatorsRigosertib versus best supportive care for patients with high-risk myelodysplastic syndromes after failure of hypomethylating drugs (ONTIME): a randomised, controlled, phase 3 trial.Lancet Oncol. 2016; 17: 496-508Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar). Clearly, efforts to identify a cohort of patients who could benefit from rigosertib would be facilitated by knowledge of its targets and genetic dependencies. More broadly, the example of rigosertib underscores the difficulty in definitively identifying the molecular target and mechanism of action of a small molecule even when it has robust cell biological effects, a challenge that has plagued both clinical pharmacology and chemical-genetic efforts (Arrowsmith et al., 2015Arrowsmith C.H. Audia J.E. Austin C. Baell J. Bennett J. Blagg J. Bountra C. Brennan P.E. Brown P.J. Bunnage M.E. et al.The promise and peril of chemical probes.Nat. Chem. Biol. 2015; 11: 536-541Crossref PubMed Scopus (541) Google Scholar, Munoz, 2017Munoz L. Non-kinase targets of protein kinase inhibitors.Nat. Rev. Drug Discov. 2017; 16: 424-440Crossref PubMed Scopus (76) Google Scholar). Here, we present a generalizable chemical-genetic strategy that combines CRISPRi and CRISPRa screens to define the mechanism of action of chemical agents, and we apply it to identify the target of rigosertib. A challenge facing any chemical-genetic approach is that modulating expression of many genes can impact sensitivity to a drug through indirect effects (e.g., by slowing cell growth), impeding identification of the direct target. We show that our integrated CRISPRi/a approach overcomes this challenge by providing a filter for removing genes that impact drug sensitivity through such indirect mechanisms, enabling identification of genes whose protein levels directly dictate drug sensitivity. Indeed, genome-wide profiling of rigosertib sensitivity in this manner strongly implicated microtubule destabilization as rigosertib’s main mechanism of action. Comparison to other drugs using focused chemical-genetic profiling similarly suggested that rigosertib acts as a microtubule-destabilizing agent, which we then confirmed using targeted approaches. Our results reveal the mechanism of cancer cell killing by rigosertib as well as genetic dependencies of rigosertib sensitivity, possibly informing patient selection and providing routes for rational engineering toward improved clinical utility. More broadly, this work serves as a general blueprint for the use of CRISPR-based chemical-genetic screens in drug target identification. We first sought to identify genetic modulators of rigosertib sensitivity in a genome-wide and unbiased manner. For this purpose, we leveraged our CRISPRi/CRISPRa functional genomics platform to screen for genes whose knockdown or overexpression affects sensitivity to rigosertib (Figure 1B). Briefly, we infected chronic myeloid leukemia (K562) cells expressing either dCas9-KRAB (CRISPRi) (Gilbert et al., 2013Gilbert L.A. Larson M.H. Morsut L. Liu Z. Brar G.A. Torres S.E. Stern-Ginossar N. Brandman O. Whitehead E.H. Doudna J.A. et al.CRISPR-mediated modular RNA-guided regulation of transcription in eukaryotes.Cell. 2013; 154: 442-451Abstract Full Text Full Text PDF PubMed Scopus (2254) Google Scholar) or dCas9 fused to the SunTag (SunCas9) and a SunTag-binding single-chain antibody fused to VP64 (CRISPRa) (Tanenbaum et al., 2014Tanenbaum M.E. Gilbert L.A. Qi L.S. Weissman J.S. Vale R.D. A protein-tagging system for signal amplification in gene expression and fluorescence imaging.Cell. 2014; 159: 635-646Abstract Full Text Full Text PDF PubMed Scopus (904) Google Scholar) with our first-generation genome-scale CRISPRi and CRISPRa sgRNA (single-guide RNA) libraries (targeting 15,977 genes) (Gilbert et al., 2014Gilbert L.A. Horlbeck M.A. Adamson B. Villalta J.E. Chen Y. Whitehead E.H. Guimaraes C. Panning B. Ploegh H.L. Bassik M.C. et al.Genome-scale CRISPR-mediated control of gene repression and activation.Cell. 2014; 159: 647-661Abstract Full Text Full Text PDF PubMed Scopus (1544) Google Scholar). After harvesting a subpopulation at the outset of the experiment (t0), we cultured the remaining cells at a coverage of >1,000 cells/sgRNA without treatment or with rigosertib treatment. We then measured the relative abundance of each sgRNA in each population by next-generation sequencing to reveal how each sgRNA affects growth in the absence of rigosertib (γ) and sensitivity to rigosertib (ρ). In particular, ρ represents the normalized difference in abundance between the treated and untreated populations for each sgRNA (Figure 1B; STAR Methods) (Kampmann et al., 2013Kampmann M. Bassik M.C. Weissman J.S. Integrated platform for genome-wide screening and construction of high-density genetic interaction maps in mammalian cells.Proc. Natl. Acad. Sci. USA. 2013; 110: E2317-E2326Crossref PubMed Scopus (81) Google Scholar), with ρ > 0 indicating that expression of the sgRNA confers protection against treatment and ρ < 0 indicating sensitization. Rigosertib sensitivity phenotypes of targeting sgRNAs were well correlated in biological replicates (for sgRNAs with |ρ| > 0.1 in either replicate, Pearson r2 [CRISPRi] = 0.70, Pearson r2 [CRISPRa] = 0.61), whereas those of non-targeting control sgRNAs were clustered around zero (Figure 1C). The resulting gene-level phenotypes revealed 1,102 genes for which knockdown or overexpression strongly affects rigosertib sensitivity (|ρ| > 0.15; Figures 1D and S1A; Data S1). The majority of genes with strong effects on rigosertib sensitivity have protective phenotypes only in either the CRISPRi or the CRISPRa screens (hits along the axes in Figure 1D). Most of these protective phenotypes, however, are correlated with substantial negative effects on growth in the absence of rigosertib (Figure S1B), suggesting that the protective phenotypes could be caused by indirect effects on rigosertib sensitivity mediated through broader changes in cell physiology. For example, knockdown of carbamoyl phosphate synthetase 2 (CAD) strongly protects cells against rigosertib, but it also reduces untreated cell growth dramatically (γ = −0.51) and leads to cell-cycle arrest in S phase (Adamson et al., 2016Adamson B. Norman T.M. Jost M. Cho M.Y. Nuñez J.K. Chen Y. Villalta J.E. Gilbert L.A. Horlbeck M.A. Hein M.Y. et al.A multiplexed single-cell CRISPR screening platform enables systematic dissection of the unfolded protein response.Cell. 2016; 167: 1867-1882Abstract Full Text Full Text PDF PubMed Scopus (488) Google Scholar). Thus, CAD knockdown could protect against rigosertib simply by preventing cells from reaching mitosis, the cell-cycle phase that is likely affected by rigosertib (Gumireddy et al., 2005Gumireddy K. Reddy M.V.R. Cosenza S.C. Boominathan R. Baker S.J. Papathi N. Jiang J. Holland J. Reddy E.P. ON01910, a non-ATP-competitive small molecule inhibitor of Plk1, is a potent anticancer agent.Cancer Cell. 2005; 7: 275-286Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar). Other essential genes, such as those encoding subunits of the mitochondrial ribosome, similarly have protective phenotypes in CRISPRi but no phenotypes in CRISPRa. Conversely, genes with growth phenotypes in CRISPRa have protective phenotypes in CRISPRa but generally no phenotypes in CRISPRi. Gene enrichment analyses of hits from either screen alone show the strongest enrichments for categories comprising these essential genes (Figures S1C and S1D). To prioritize hits, we therefore compared the CRISPRi and CRISPRa phenotypes for all genes (Figures 1D and 1E), reasoning that genes directly involved in the process(es) targeted by rigosertib might have strong and oppositely signed phenotypes. In this comparison, two genes stood out: KIF2C and TACC3. KIF2C is the most sensitizing hit in the CRISPRa screen and a strongly protective hit in the CRISPRi screen, whereas KIF2C knockdown has no effect on untreated growth. Conversely, TACC3 knockdown sensitizes cells to rigosertib, whereas overexpression is protective. Notably, both genes are involved in regulating microtubule dynamics (Figure 1F): KIF2C encodes the microtubule depolymerase MCAK (Tanenbaum et al., 2011Tanenbaum M.E. Medema R.H. Akhmanova A. Regulation of localization and activity of the microtubule depolymerase MCAK.BioArchitecture. 2011; 1: 80-87Crossref PubMed Scopus (27) Google Scholar) and TACC3 is a microtubule-binding protein that, directly or indirectly, promotes microtubule stability, especially during mitosis (Hood and Royle, 2011Hood F.E. Royle S.J. Pulling it together: the mitotic function of TACC3.BioArchitecture. 2011; 1: 105-109Crossref PubMed Scopus (42) Google Scholar). Similarly, rigosertib sensitivity is affected by modulation of several tubulin isoform-encoding genes as well as other microtubule-associated genes, including TACC3- and KIF2C-interacting proteins such as CKAP5 and KIF18B (Figure 1D). Thus, genetic manipulations that destabilize microtubules sensitize cells to rigosertib, whereas stabilization of microtubules protects cells against rigosertib, suggesting that rigosertib’s cytotoxicity, directly or indirectly, arises from a perturbation of the microtubule network (Figure 1F). To validate the screen results, we measured the effects of KIF2C and TACC3 knockdown or overexpression on rigosertib sensitivity in individual re-tests. We infected K562 CRISPRi and CRISPRa cells with constructs expressing KIF2C- or TACC3-targeting sgRNAs or a non-targeting control sgRNA and used flow cytometry to monitor how the fraction of sgRNA-expressing cells changed after treatment with rigosertib (Kampmann et al., 2013Kampmann M. Bassik M.C. Weissman J.S. Integrated platform for genome-wide screening and construction of high-density genetic interaction maps in mammalian cells.Proc. Natl. Acad. Sci. USA. 2013; 110: E2317-E2326Crossref PubMed Scopus (81) Google Scholar, Gilbert et al., 2014Gilbert L.A. Horlbeck M.A. Adamson B. Villalta J.E. Chen Y. Whitehead E.H. Guimaraes C. Panning B. Ploegh H.L. Bassik M.C. et al.Genome-scale CRISPR-mediated control of gene repression and activation.Cell. 2014; 159: 647-661Abstract Full Text Full Text PDF PubMed Scopus (1544) Google Scholar). In this internally controlled growth assay, an increase or decrease in the fraction of sgRNA-expressing cells indicates that sgRNA-expressing cells grow faster or slower than untransduced cells, respectively. Thus, relative enrichment >1 indicates that expression of the sgRNA confers protection against treatment, whereas enrichment <1 indicates sensitization. Consistent with the screen phenotypes, K562 CRISPRi cells expressing a KIF2C-targeting sgRNA were enriched after rigosertib treatment, whereas K562 CRISPRi cells expressing a TACC3-targeting sgRNA were depleted after rigosertib treatment (Figure 2A). K562 CRISPRa cells exhibited the opposite behavior (Figure S2A). We confirmed knockdown or overexpression of KIF2C and TACC3 by qRT-PCR (Figure S2B). Thus, the expression levels of KIF2C and TACC3 reproducibly dictate rigosertib sensitivity and resistance. Knockdown of KIF2C also protected both HeLa (cervical carcinoma) and H358 cells (non-small-cell lung cancer) against rigosertib, as indicated by enrichment of sgRNA-expressing cells, and knockdown of TACC3 sensitized both cell lines to rigosertib (Figures 2B, 2C, S2C, and S2D). By contrast, knockdown of KIF2C or TACC3 did not alter the sensitivity of the H358 cells, which are driven by an activating G12C mutation in KRAS, to the specific K-RAS(G12C) inhibitor ARS-853 (Patricelli et al., 2016Patricelli M.P. Janes M.R. Li L.S. Hansen R. Peters U. Kessler L.V. Chen Y. Kucharski J.M. Feng J. Ely T. et al.Selective inhibition of oncogenic KRAS output with small molecules targeting the inactive state.Cancer Discov. 2016; 6: 316-329Crossref PubMed Scopus (446) Google Scholar), indicating that rigosertib does not act on the RAS pathway, as had been suggested recently (Athuluri-Divakar et al., 2016Athuluri-Divakar S.K. Vasquez-Del Carpio R. Dutta K. Baker S.J. Cosenza S.C. Basu I. Gupta Y.K. Reddy M.V.R. Ueno L. Hart J.R. et al.A small molecule RAS-mimetic disrupts RAS association with effector proteins to block signaling.Cell. 2016; 165: 643-655Abstract Full Text Full Text PDF PubMed Scopus (191) Google Scholar). Together, these results suggest that rigosertib’s genetic interactions and consequently its mechanism of action are conserved across different cell types. The genome-wide screen indicated that genetic destabilization of microtubules potentiates rigosertib’s cytotoxicity, which would be consistent with direct or indirect microtubule destabilization by rigosertib or with other modes of inhibition of mitosis. To distinguish between these possibilities, we devised a focused chemical-genetic profiling strategy to compare rigosertib’s pattern of drug-gene interactions (the set of drug-sensitivity phenotypes) to those of drugs that directly target the microtubule network or various other steps of mitosis, with the expectation that drugs with the same target would have similar patterns of drug-gene interactions (Figure 3A) (Jiang et al., 2011Jiang H. Pritchard J.R. Williams R.T. Lauffenburger D.A. Hemann M.T. A mammalian functional-genetic approach to characterizing cancer therapeutics.Nat. Chem. Biol. 2011; 7: 92-100Crossref PubMed Scopus (71) Google Scholar). We selected 514 genes with strong rigosertib-sensitivity phenotypes in our genome-wide CRISPRi screen and generated a sublibrary of 5,390 sgRNAs targeting these genes (STAR Methods). Using a pooled screening strategy, we evaluated how knockdown of each of these genes affects sensitivity to rigosertib as well as to a panel of mitosis-targeting drugs: ABT-751 and vinblastine (two microtubule-destabilizing agents that bind to different sites on tubulin), BI2536 (a PLK1 inhibitor), blebbistatin (a myosin II inhibitor that inhibits cytokinesis), S-trityl-l-cysteine (STLC; an inhibitor of the mitotic kinesin Eg5), and alisertib (an aurora A kinase inhibitor). Sensitivity phenotypes (ρ) derived from replicate drug treatments were highly correlated (r2 ≥ 0.84 for all drugs tested; Figures 3B and S3A; Data S1), and the rigosertib-sensitivity phenotypes were well correlated with those from the genome-wide screen (Figure S3B). Comparing the phenotypes across the different drug treatments revealed a high overall correlation of phenotypes between rigosertib and ABT-751 (r2 ≥ 0.86 for each pairwise comparison of replicate phenotype sets) as well as, to a slightly lower extent, between rigosertib and vinblastine (r2 ≥ 0.79) (Figures 3B and S3C). In contrast, rigosertib’s sensitivity phenotypes were less correlated with those of the other drugs. Indeed, in hierarchical clustering, rigosertib, ABT-751, and vinblastine formed a cluster separate from the other drugs (Figure 3B). The similarity between the sensitivity phenotypes of rigosertib and ABT-751 also extended to analogous CRISPRa sublibrary screens (257 genes; Figure 3C; Data S1). Thus, rigosertib’s pattern of drug-gene interactions closely resembles those of microtubule-destabilizing agents, particularly that of ABT-751, which binds to the colchicine site on tubulin (Dorléans et al., 2009Dorléans A. Gigant B. Ravelli R.B.G. Mailliet P. Mikol V. Knossow M. Variations in the colchicine-binding domain provide insight into the structural switch of tubulin.Proc. Natl. Acad. Sci. USA. 2009; 106: 13775-13779Crossref PubMed Scopus (208) Google Scholar). In analyzing these data, we observed that knockdown of several genes required for cell growth in the absence of drugs appeared to confer resistance to all tested drugs (Figures S3C and S3D), a pattern that was also observed in the genome-wide screen and may reflect indirect effects (Figure S1B). To test if the sensitivity phenotypes of these essential genes affected the comparison of the genetic profiles of the different drugs, we segregated the genes in our dataset into clusters (Figure S3E), removed the cluster of genes with strong negative effects on untreated growth and uniformly protective effects against all drugs, and repeated the comparison between the different drugs (Figure S3F). With these genes excluded, the sensitivity phenotypes for all drugs were correlated slightly less well, but rigosertib clustered even more closely with the microtubule-destabilizing agents ABT-751 and vinblastine compared to the drugs targeting other essential mitotic proteins. As KI" @default.
• W2736544909 created "2017-07-31" @default.
• W2736544909 creator A5002692347 @default.
• W2736544909 creator A5011792616 @default.
• W2736544909 creator A5024109873 @default.
• W2736544909 creator A5026663934 @default.
• W2736544909 creator A5033614969 @default.
• W2736544909 creator A5035219354 @default.
• W2736544909 creator A5041503324 @default.
• W2736544909 creator A5050072827 @default.
• W2736544909 creator A5060001686 @default.
• W2736544909 creator A5068209077 @default.
• W2736544909 creator A5070304364 @default.
• W2736544909 creator A5073245145 @default.
• W2736544909 creator A5073914369 @default.
• W2736544909 creator A5077031542 @default.
• W2736544909 creator A5077510821 @default.
• W2736544909 date "2017-10-01" @default.
• W2736544909 modified "2023-10-15" @default.
• W2736544909 title "Combined CRISPRi/a-Based Chemical Genetic Screens Reveal that Rigosertib Is a Microtubule-Destabilizing Agent" @default.
• W2736544909 cites W1516646538 @default.
• W2736544909 cites W1540299613 @default.
• W2736544909 cites W1926612287 @default.
• W2736544909 cites W1963878289 @default.
• W2736544909 cites W1968327387 @default.
• W2736544909 cites W1976023782 @default.
• W2736544909 cites W1982846895 @default.
• W2736544909 cites W1987788840 @default.
• W2736544909 cites W1990511694 @default.
• W2736544909 cites W2005308681 @default.
• W2736544909 cites W2014031830 @default.
• W2736544909 cites W2014209528 @default.
• W2736544909 cites W2023518810 @default.
• W2736544909 cites W2026682681 @default.
• W2736544909 cites W2029108033 @default.
• W2736544909 cites W2036218826 @default.
• W2736544909 cites W2038161360 @default.
• W2736544909 cites W2041636661 @default.
• W2736544909 cites W2044195791 @default.
• W2736544909 cites W2048454522 @default.
• W2736544909 cites W2063013559 @default.
• W2736544909 cites W2064571598 @default.
• W2736544909 cites W2083713217 @default.
• W2736544909 cites W2093855538 @default.
• W2736544909 cites W2095072642 @default.
• W2736544909 cites W2101749257 @default.
• W2736544909 cites W2108959691 @default.
• W2736544909 cites W2113144687 @default.
• W2736544909 cites W2120318990 @default.
• W2736544909 cites W2127857566 @default.
• W2736544909 cites W2131321938 @default.
• W2736544909 cites W2132356916 @default.
• W2736544909 cites W2133465414 @default.
• W2736544909 cites W2137753785 @default.
• W2736544909 cites W2152505144 @default.
• W2736544909 cites W2156739666 @default.
• W2736544909 cites W2169373177 @default.
• W2736544909 cites W2173348076 @default.
• W2736544909 cites W2204875887 @default.
• W2736544909 cites W2237971187 @default.
• W2736544909 cites W2276949837 @default.
• W2736544909 cites W2294083716 @default.
• W2736544909 cites W2300615341 @default.
• W2736544909 cites W2339828620 @default.
• W2736544909 cites W2420436228 @default.
• W2736544909 cites W2513408092 @default.
• W2736544909 cites W2523099503 @default.
• W2736544909 cites W2553383160 @default.
• W2736544909 cites W2562947047 @default.
• W2736544909 cites W2592378969 @default.
• W2736544909 cites W2610333492 @default.
• W2736544909 doi "https://doi.org/10.1016/j.molcel.2017.09.012" @default.
• W2736544909 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/5640507" @default.
• W2736544909 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/28985505" @default.
• W2736544909 hasPublicationYear "2017" @default.
• W2736544909 type Work @default.
• W2736544909 sameAs 2736544909 @default.
• W2736544909 citedByCount "186" @default.
• W2736544909 countsByYear W27365449092017 @default.
• W2736544909 countsByYear W27365449092018 @default.
• W2736544909 countsByYear W27365449092019 @default.
• W2736544909 countsByYear W27365449092020 @default.
• W2736544909 countsByYear W27365449092021 @default.
• W2736544909 countsByYear W27365449092022 @default.
• W2736544909 countsByYear W27365449092023 @default.
• W2736544909 crossrefType "journal-article" @default.
• W2736544909 hasAuthorship W2736544909A5002692347 @default.
• W2736544909 hasAuthorship W2736544909A5011792616 @default.
• W2736544909 hasAuthorship W2736544909A5024109873 @default.
• W2736544909 hasAuthorship W2736544909A5026663934 @default.
• W2736544909 hasAuthorship W2736544909A5033614969 @default.
• W2736544909 hasAuthorship W2736544909A5035219354 @default.
• W2736544909 hasAuthorship W2736544909A5041503324 @default.
• W2736544909 hasAuthorship W2736544909A5050072827 @default.
• W2736544909 hasAuthorship W2736544909A5060001686 @default.
• W2736544909 hasAuthorship W2736544909A5068209077 @default.
• W2736544909 hasAuthorship W2736544909A5070304364 @default.
• W2736544909 hasAuthorship W2736544909A5073245145 @default.

• next