Abstract

Future Medicinal ChemistryVol. 10, No. 18 CommentaryFree AccessSynthetic lethality in drug development: the dawn is comingShuaishi Gao & Luhua LaiShuaishi Gao Center for Quantitative Biology, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, 100871, PR ChinaSearch for more papers by this author & Luhua Lai*Author for correspondence: E-mail Address: lhlai@pku.edu.cn Center for Quantitative Biology, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing, 100871, PR China BNLMS, State Key for Structural Chemistry of Unstable & Stable Species, College of Chemistry & Molecular Engineering, Peking University, Beijing, 100871, PR ChinaSearch for more papers by this authorPublished Online:25 Jul 2018https://doi.org/10.4155/fmc-2018-0227AboutSectionsPDF/EPUB ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinkedInReddit Keywords: CRISPRdrug combinationdrug developmentRNAisynthetic lethalitySynthetic lethality refers to the death of cells caused by concomitant perturbations of two genes (loss-of-function mutations, RNA interference, drug treatment, etc.), each of which is nonlethal alone. The concept of synthetic lethality was first proposed in 1922 by Calvin Bridges when studying the model organism Drosophila melanogaster, and synthetic lethality screening was suggested as a strategy to develop anticancer therapeutics about two decades ago [1]. With the accumulation of large-scale ‘omics’ data, much has been learned about the multilevel differences between cancer cells and normal cells, and synthetic lethality consequently offers the possibility to selectively target cancer cells, with the potential to reduce drug resistance and side effects. However, only one synthetically lethal interaction – between the poly(adenosine diphosphate [ADP]-ribose) (PARP) and breast-cancer susceptibility genes 1 and 2 (BRCA1 and BRCA2), has made the journey from discovery to clinical approval, even if the biological mechanism behind this lethality is still not completely understood [2]. Nevertheless, studies on synthetic lethality have demonstrated great potential for clinical applications. This review focuses on recent developments in synthetic lethality screening strategies, case analysis of clinical and preclinical application examples, as well as the current challenges and opportunities.Strategies for synthetic lethality screeningThe use of RNA interference (RNAi) by small interfering RNAs (siRNAs) or short hairpin RNAs (shRNAs) as a screening technology platform makes it possible to identify synthetic lethality via simultaneous knockouts. Researchers from Novartis launched the DRIVE project, using RNAi to screen about 8000 genes in nearly 400 cancer cell lines [3]. This large-scale gene depletion screening enabled the analysis of the characteristics of many genes as well as their mutual interdependence. The main drawback of RNAi is data discrepancy, which can lead to low reproducibility and poor overlap across different studies, including false discovery due to the inability to distinguish between on- and off-target effects [4]. Although increasing the screening range and depth can reduce off-target effects to some extent, this comes at the expense of increased labor intensity.The clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated protein 9 (CRISPR-Cas9) system was firstly discovered in the adaptive immune system of archaea and was utilized in gene editing in a wide range of organisms. Its function is similar to RNAi, but it offers higher reliability. This system contains a single-guide RNA (sgRNA) that allows the Cas9 endonuclease to target the specific sequence homologous to the sgRNA, which enables subsequent modification or regulation. The combination of CRISPR-Cas9-mediated perturbation and analysis of growth kinetics has become a robust method for genetic interaction screening [5]. This method can be applied in different cell lines by synthesizing a dual-gRNA library containing up to 105 defined gene pairs to analyze the changes of the cell growth rate.By comparison, RNAi utilizes siRNAs or shRNAs to directly trigger a gene knockdown, while the construction of Cas9 complex is less effective. On the other hand, an advantage of CRISPR is its lower susceptibility to systematic off-target effects. In fact, large-scale RNAi and CRISPR screenings are complementary, and increased accuracy is expected to result from their combined use. Housden et al. used a CRISPR-based method to generate stable mutations in Drosophila cell lines to mimic different disease states, combined it with RNAi to screen for synthetic interactions and discovered novel therapeutic targets for tuberous sclerosis complex therapy [6]. This combination strategy reduces the off-target effects due to the reduction of noise with dual RNAi, and, consequently, it does not require a second, time-consuming screening.Nonetheless, CRISPR and RNAi hold great promise as therapeutic agents with an extensive ability to selectively target genes that are not treatable by small-molecule inhibitors or antibodies. We can design gene therapy programs based on synthetically lethal relationships between genes.Synthetic lethality can also be detected by combinational drug screening. The use of drug screen provides explicit pharmacological and clinical benefits. Wali et al. used a medium-throughput screen consisting of 768 pairwise drug combinations between six FDA-approved drugs and 128 investigational agents, to test their treatment effect on triple-negative breast cancer, which led to the discovery of excellent synergistic combinations [7]. The strongest additive effect was observed with ABT-263 and crizotinib, which targeted BCL-xL/BCL2 and ALK/MET respectively. Licciardello et al. built a combinatorial drug library for the purpose of screening, which contained 40,160 pairwise combinations of 308 small molecules [8]. Notably, they used filtering and clustering procedures based on biochemical diversity and drug mechanisms, to narrow down the large dataset of FDA-approved drugs. Using the high-throughput drug combination screen in an androgen receptor positive chronic myeloid leukemia cell line, they found the synergistic effect between flutamide and phenprocoumon, which showed promising anticancer effect on prostate cancer cells.Jerby-Arnon et al. used a combined analysis of somatic copy number alterations, somatic mutations, gene expression data and shRNA data, and developed the DAISY pipeline to identify synthetic lethal and synthetic dosage lethal interactions in cancer, which have been validated in cells using RNAi and drugs [9]. This integrative statistical analysis pipeline provides a way of combining the genetic data of cancer patients with multiple data from human cell lines.Ding et al. used somatic mutation data and a driver gene list to search for the gene's co-occurrence and mutual exclusivity [10]. The results indicated that the same somatic–somatic interactions in different cell lines can be co-occurring or mutually exclusive, again verifying that synthetic lethality is related to the cellular context.The biological significance of synthetic lethality is related to the robustness of the cell system. In silico predictions based on systems biology models can be used to simulate the complexity and alternativity of biology networks. Apaolaza et al. searched for minimal cut sets, which represent the minimum set of reactions whose deletion would cause cell death, in a genome-scale metabolic network, and integrated gene expression data to discover vulnerabilities of cancer cells [11]. They applied this framework to analyze the lethality mechanism of ribonucleotide reductase catalytic subunit M1 in multiple myeloma.Additionally, machine learning can be used to explore the sizeable synergistic space more efficiently. Preuer et al. put forward DeepSynergy, a deep learning method based on a feed-forward neural network, whose inputs contained chemical descriptors of two drugs and the genomic information of the cell line without compound treatment, which showed a higher predictive performance than other machine learning methods [12]. Specifically, the data set contained 23,062 samples covering 583 combinations derived from 38 distinct anticancer drugs, each tested against 39 cancer cell lines. In the future, we expect that machine learning methods will not only provide a high predictive performance based on existing dataset, but also predict presently unknown drug combinations.Clinical & preclinical studies of synthetic lethalityInhibitors of PARP, a single-strand damage sensor, are the first clinically approved drugs based on synthetic lethality. Tumors with mutations of the tumor suppressor gene BRCA1/2, which participates in the repair of double-stranded DNA breaks, are sensitive to PARP inhibitors, which include the FDA approved drugs olaparib (prostate/ovarian/breast cancer), rucaparib (ovarian cancer), niraparib (ovarian cancer), and the Phase-III investigative drug talazoparib (EMBRACA) [2,13].The inhibition of EGFR in BRAF-mutant colorectal cancer has entered a randomized Phase II trial [14]. BRAF inhibitors can decrease MAPK signaling, but in turn activate EGFR through a negative feedback loop, whereby the combined inhibition of EGFR and BRAF can result in a reduced growth rate of cancer cells.There are other promising synthetic lethal pairs under investigation. For example, TP53 mutations widely occur in carcinogenesis, and activating p53 alone can induce G1 arrest in cancer cells. Since inhibiting Bcl-2 can shift cancer cells from G1 arrest to apoptosis and lower their drug resistance by targeting the downstream apoptotic pathway, the combination of the p53 activator RG and the Bcl-2 inhibitor ABT (venetoclax) was shown to be an active synthetically lethal pair [15].PTEN has a high mutation rate in cancer cells, leading to the activation of the PI3K/Akt pathway. After analyzing The Cancer Genome Atlas database, Zhao et al. suggested that the chromatin remodeling gene CHD1 is a putative synthetic-essential gene in PTEN loss-of-function cancers [16].RAS is an oncogene product which is difficult to modulate directly using small molecules [1]. Using a CRISPR-based RNAi screening, Wang et al. profiled gene essentiality in 14 AML cell lines [17]. By comparing gene essentiality between RAS-dependent and RAS-independent cell lines, they highlighted the synthetically lethal genes assembled in the up- and downstream pathways of RAS.Limitations & opportunities of drug discovery via synthetic lethalitySynthetic lethality provides promising solutions for targeting oncogenic mutations of genes such as RTK/RAS, TP53 and MYC, which show high mutation rates in cancer patients [18]. The loss-of-function mutations in these genes make them theoretically recalcitrant to pharmaceutical therapy. Nevertheless, synthetic lethality can circumvent this barrier and decrease the proliferation and metastasis of the corresponding cancer cells.However, there are three significant hurdles in applying synthetic lethality in drug development. First, synthetic lethal interactions are rare, and it is therefore difficult to discover robust pairs due to false positive and false-negative interactions predicted by high-throughput screening technologies. Thus, compatible technologies and judgment standards should be developed. Moreover, follow-up studies including cross-validation by other strategies such as combinational drug experiments, should be performed after the initial screening.Second, the regulation of signaling pathways and the intracellular environment are heterogeneous due to the genetic and metabolic differences between different cell types, which makes synthetic interactions context-dependent. Elucidation of the mechanisms underlying the synthetic interactions and understanding of the cross-talk between pathways is of great significance to verify synthetically lethal drug targets. Along with the accumulation of personalized data, therapeutic strategies based on synthetic lethality can be incorporated with precision medicine.Third, synthetic lethality is not simply defined by loss-of-function mutation at the gene level, but rather involves complex regulation patterns encompassing gene expression, protein synthesis and phenotypes such as proliferation and metastasis. For example, in p53+/+ tumor cells, p53 is inactivated if there is a deletion of ARF, a tumor suppressor that positively regulates p53 [19]. This enlarges the conventional synthetic lethality definition and increases the number of feasible regulatory options. High-throughput screening and systematic analysis methods are needed in this area.In addition to the study of pairwise combinations, the synthetic lethality of multiple drugs should also be explored. Cocktail therapies commonly used in Traditional Chinese Medicine, as well as in AIDS treatment, provide successful examples in this direction. However, even though ongoing clinical trials are testing combinations of three, and even four different drugs, the mechanisms behind the combination regimens are still poorly understood [20]. Indeed, in addition to the availability of high-throughput technologies and big data, understanding the mechanisms of the currently known multidrug combinations is necessary to better predict the prognostic outcomes.ConclusionSynthetic lethality provides a conceptual framework for developing effective therapeutic solutions for the treatment of complex diseases. Gradually, maturing high-throughput screening technologies and computational methods facilitate the identification of synthetic lethality in specific genetic contexts and the underlying mechanisms, as well as predicting the possible synergistic pairs. Looking forward, along with the fast development of the field, more clinically applicable synthetic lethality therapeutic strategies are expected in the near future.Financial and competing interests disclosureThis work was supported in part by the Ministry of Science and Technology of China (2016YFA0502303, 2015CB910300) and the National Natural Science Foundation of China (21633001). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.No writing assistance was utilized in the production of this manuscript.References1 Hartwell LH. Integrating genetic approaches into the discovery of anticancer drugs. Science 278(5340), 1064–1068 (1997).Crossref, Medline, CAS, Google Scholar2 Lord CJ, Ashworth A. PARP inhibitors: synthetic lethality in the clinic. Science 355(6330), 1152–1158 (2017).Crossref, Medline, CAS, Google Scholar3 McDonald ER, de Weck A, Schlabach MR et al. Project DRIVE: a compendium of cancer dependencies and synthetic lethal relationships uncovered by large-scale, deep RNAi screening. Cell 170(3), 577–592 (2017).Crossref, Medline, CAS, Google Scholar4 Kaphingst KA, Persky S, Lachance C. Systematic analysis of RNAi reports identifies dismal commonality at gene-level & reveals an unprecedented enrichment in pooled shRNA screens. Comb. Chem. High Throughput Screen. 16(9), 665–681 (2013).Crossref, Medline, Google Scholar5 Shen JP, Zhao D, Sasik R et al. Combinatorial CRISPR-Cas9 screens for de novo mapping of genetic interactions. Nat. Methods 14(6), 573–576 (2017).Crossref, Medline, CAS, Google Scholar6 Housden BE, Valvezan AJ, Kelley C et al. Identification of potential drug targets for tuberous sclerosis complex by synthetic screens combining CRISPR-based knockouts with RNAi. Sci. Signal. 8(393), 1–22 (2015).Crossref, CAS, Google Scholar7 Wali VB, Langdon CG, Held MA et al. Systematic drug screening identifies tractable targeted combination therapies in triple-negative breast cancer. Cancer Res. 77(2), 566–578 (2017).Crossref, Medline, CAS, Google Scholar8 Licciardello MP, Ringler A, Markt P et al. A combinatorial screen of the CLOUD uncovers a synergy targeting the androgen receptor. Nat. Chem. Biol. 13(7), 771–778 (2017).Crossref, Medline, CAS, Google Scholar9 Jerby-Arnon L, Pfetzer N, Waldman YY et al. Predicting cancer-specific vulnerability via data-driven detection of synthetic lethality. Cell 158(5), 1199–1209 (2014).Crossref, Medline, CAS, Google Scholar10 Ding L, Bailey MH, Porta-Pardo E et al. Perspective on oncogenic processes at the end of the beginning of cancer genomics. Cell 173(2), 305–320 (2018).Crossref, Medline, CAS, Google Scholar11 Apaolaza I, San José-Eneriz E, Tobalina L et al. An in-silico approach to predict and exploit synthetic lethality in cancer metabolism. Nat. Commun. 8(1), 459 (2017).Crossref, Medline, Google Scholar12 Preuer K, Lewis RPI, Hochreiter S, Bender A, Bulusu KC, Klambauer G. DeepSynergy: predicting anti-cancer drug synergy with Deep Learning. Bioinformatics 34(9), 1538–1546 (2017).Crossref, Google Scholar13 Zimmer AS, Gillard M, Lipkowitz S, Lee J-M. Update on PARP inhibitors in breast cancer. Curr. Treat. Options Oncol. 19(5), 21 (2018).Crossref, Medline, Google Scholar14 Kopetz S, McDonough SL, Morris VK et al. Randomized trial of irinotecan and cetuximab with or without vemurafenib in BRAF -mutant metastatic colorectal cancer (SWOG 1406). J. Clin. Oncol. 35(4 Suppl.), 520–520 (2017).Crossref, Google Scholar15 Pan R, Ruvolo V, Mu H et al. Synthetic lethality of combined Bcl-2 inhibition and p53 activation in AML: mechanisms and superior antileukemic efficacy. Cancer Cell 32(6), 748–760 (2017).Crossref, Medline, CAS, Google Scholar16 Zhao D, Lu X, Wang G et al. Synthetic essentiality of chromatin remodelling factor CHD1 in PTEN-deficient cancer. Nature 542(7642), 484–488 (2017).Crossref, Medline, CAS, Google Scholar17 Wang T, Yu H, Hughes NW et al. Gene essentiality profiling reveals gene networks and synthetic lethal interactions with oncogenic ras. Cell 168(5), 890–903 (2017).Crossref, Medline, CAS, Google Scholar18 Sanchez-Vega F, Mina M, Armenia J et al. Oncogenic signaling pathways in the cancer genome atlas. Cell 173(2), 321–337 (2018).Crossref, Medline, CAS, Google Scholar19 Sherr CJ, Bertwistle D, Den Besten W et al. p53-dependent and -independent functions of the arf tumor suppressor. Cold Spring Harb. Symp. Quant. Biol. 70, 129–137 (2005).Crossref, Medline, CAS, Google Scholar20 Kapoor P, Rajkumar SV. Multiple myeloma in 2016: fresh perspectives on treatment and moments of clarity. Nat. Rev. Clin. Oncol. 14(2), 73–74 (2017).Crossref, Medline, Google ScholarFiguresReferencesRelatedDetailsCited Byahctf1 and kras mutations combine to amplify oncogenic stress and restrict liver overgrowth in a zebrafish model of hepatocellular carcinoma17 January 2023 | eLife, Vol. 12Pinpointing Cancer Sub-Type Specific Metabolic Tasks Facilitates Identification of Anti-cancer Targets23 March 2022 | Frontiers in Medicine, Vol. 9Discovery of a New CaMKII-Targeted Synthetic Lethal Therapy against Glioblastoma Stem-like Cells4 March 2022 | Cancers, Vol. 14, No. 5Genetic landscape of T cells identifies synthetic lethality for T-ALL20 October 2021 | Communications Biology, Vol. 4, No. 1Integration of genome-level data to allow identification of subtype-specific vulnerability genes as novel therapeutic targets6 July 2021 | Oncogene, Vol. 40, No. 33Synthetic Lethality through the Lens of Medicinal Chemistry2 November 2020 | Journal of Medicinal Chemistry, Vol. 63, No. 23Synthetic lethality on drug discovery: an update on cancer therapy31 March 2020 | Expert Opinion on Drug Discovery, Vol. 15, No. 7 Vol. 10, No. 18 Follow us on social media for the latest updates Metrics History Received 7 June 2018 Accepted 7 June 2018 Published online 25 July 2018 Published in print September 2018 Information© 2018 Newlands PressKeywordsCRISPRdrug combinationdrug developmentRNAisynthetic lethalityFinancial and competing interests disclosureThis work was supported in part by the Ministry of Science and Technology of China (2016YFA0502303, 2015CB910300) and the National Natural Science Foundation of China (21633001). The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.No writing assistance was utilized in the production of this manuscript.PDF download

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call