Spontaneous DNA Damage Propels Tumorigenicity Cell Research (2017) 27:720-721

Spontaneous DNA damage propels tumorigenicity Cell Research (2017) 27:720-721. doi:10.1038/cr.2017.43; published online 31 March 2017

High levels of endogenously gen- (TP53/p53) and the histone variant H2A as “self-inflicted” or spontaneous DSBs erated DNA damage drive oncogen- histone family member X (H2AFX/ (spDSBs), originated in viable cancer esis, sustain malignant progression H2AX). This multi-pronged cascade cells independently of ROS and RS and increase therapy resistance. In also comprises auto-activation steps, (as they occurred before S phase entry) a paper recently published in Cell feedback loops and a close interplay, through the order of events involving: Research, Liu and colleagues added and possibly redundancy, with the (1) the partial permeabilization of the additional insights into this topic by ATR serine/threonine kinase (ATR) outer mitochondrial membrane, (2) the uncovering a novel intrinsic source network [6]. release of cytochrome c1 (CYC1) from of double-strand breaks that fosters A wide range of evidence dem- mitochondria, (3) sublethal activation the aggressiveness and stemness of onstrates that unrepaired DSBs can of executioner caspases (caspase 3 malignant cells. promote tumorigenesis by inducing (CASP3), CASP6 and CASP7), and DNA double-strand breaks (DSBs) DNA mutation(s) and karytotypic (4) cleavage of chromatin DNA by the are a type of DNA damage characterized aberration(s) [7]. Moreover, established apoptotic nucleases endonuclease G by the severance of both strands of the cancer cells often show increased DSB (ENDOG) and DNA fragmentation DNA duplex. Besides being produced formation accompanied by an extensive factor β (DFFB/CAD) (Figure 1). This by exogenous genotoxins, these cyto- DDR rewiring, encompassing defects circuitry was elegantly unveiled by toxic lesions can arise from multiple in pathway(s) for DNA repair and/or CRISPR/Cas9 gene editing-mediated endogenous sources, including (1) the overactivation of the ATM and ATR knockout and epistatic experiments metabolic by-products, mostly reactive axes [4]. This holds true also for cancer revealing limited nuclease activation oxygen species (ROS); (2) prolonged stem cells (CSCs) [8], which are the in CASP3−/−, CASP3−/−CASP6−/− and stalling in DNA replication fork pro- subset of immature, self-renewing and CASP3−/−CASP6−/−CASP7−/− cancer gression as this occurs during repli- multipotent cells within the tumor mass cells, as well as a significant decrease cation stress (RS); (3) chromosome driving cancer initiation and evolution of spDSBs in tumor cells deficient for missegregation events; (4) telomere [9]. Logically, DDR-related molecules executioner caspases, ENDOG or CAD shortening; and (5) chromothripsis, a are hence regarded as candidate targets [10]. This mechanism is reminiscent to phenomenon associated with the forma- for effective anticancer therapies [4]. that observed upon exogenous perturba- tion of micronuclei [1-3]. Endogenous In a recent paper published by Cell tions [11], even though it occurred in DSBs are boosted by features associated Research, the group of Chuan-Yuan Li unperturbed conditions. with (pre)malignancy, such as increased reports an alternative, intriguing point of In subsequent experiments, Liu et proliferation rate, metabolic rewiring, view about the source of endogenously al. provided evidence in favor of the activated oncogenes, deficiencies in induced DSBs and their contribution to role of spDSBs in sustaining cancer cell DNA damage response (DDR), karyo- cancer evolution [10]. When perform- tumorigenicity and stemness potential. typic aberration(s) or chromosomal ing a comparative analysis of a panel of They demonstrated that the subpopula- instability (CIN) [4, 5]. malignant, immortalized/untransformed tion of cancer cells with elevated levels Irrespective of their origin, DSBs are and primary cell lines previously syn- of spDSBs (spDSBshigh) displayed

quickly detected by the ATM serine/ chronized in the G1 phase, these authors higher clonogenicity in soft-agar assay threonine kinase (ATM). Upon binding found high levels of endogenously in- and grew more efficiently when xeno- to DSBs, ATM activates one or more duced DSBs exclusively in tumor cells. grafted in nude mice than the spDSBslow branches of DDR by phosphorylating The presence of DSBs was assessed fraction. In line with this evidence, factors involved in DNA damage repair, by quantifying foci containing γH2AX strategies aimed at minimizing spDSBs cell cycle checkpoint and/or regulated and tumor protein p53 binding protein 1 (e.g., by knocking out effector caspases cell death, such as the checkpoint kinase (TP53BP1/53BP1) and by the COMET or knocking down pro-apoptotic BCL2 2 (CHEK2/CHK2), tumor protein p53 assay. These lesions, which were dubbed proteins) and boosting spDSBs (e.g.,

SPRINGER NATURE | Cell Research | www.nature.com/cr 721 by irradiating spDSBslow cells with low Figure 1 Mechanism and consequences doses of x-rays) dwindled and enhanced of spDSB formation in tumors. spDSBs in cancer cells are generated via a mechanism the in vitro clonogenicity and in vivo involving the release from mitochondria of tumorigenicity of malignant cells, re- sublethal amounts of ENDOG and CYC1, spectively [10]. which in turn unleashes CAD via the ac- The increase in tumor aggressiveness tivation of CASP3. This is followed by the translocation of ENDOG and CAD from the was associated with the constitutive cytosol to the nucleus, where these nucle- phosphorylation/activation of ATM. ases catalyze DNA cleavage. Upon their Activated ATM ultimately ignited the generation, spDSBs ignite an ATM/NF-κB/ pro-tumorigenic nuclear factor-kappaB STAT3 cascade that promotes malignant tumorigenicity and stemness. (NF-κB) and signal transducer and activator of transcription 3 (STAT3) pathways instead of triggering DDR. Thus, knockout of executioner caspases, Ilio Vitale1, 2, ENDOG or CAD, reduced the level Guido Kroemer3, 4, 5, 6, 7, 8, 9 of phosphorylated ATM (pATM) and 1Department of Biology, University of Rome “Tor STAT3 (pSTAT3), and abrogated NF-κB Vergata”, Rome 00133, Italy; 2Regina Elena Na- activation. In addition, cells deficient tional Cancer Institute, Rome 00144, Italy; 3Uni- for ATM did not constitutively activate versité Paris Descartes/Paris V, Sorbonne Paris 4 the NF-κB/STAT3 axis, and displayed Cité, Paris 75006, France; Université Pierre et Marie Curie/Paris VI, Paris 75006, France; reduced soft-agar colony-forming abil- 5Equipe 11 labellisée Ligue contre le Cancer, ity and in vivo tumor growth [10]. In Centre de Recherche des Cordeliers, Paris 75006, these experimental settings, knockout France; 6INSERM, U1138, Paris 75006, France; 7 of ATR abolished the clonogenic and Metabolomics and Cell Biology Platforms, −/− Gustave Roussy Comprehensive Cancer Institute, tumorigenic potential of ATM cancer Villejuif 94805, France; 8Pôle de Biologie, Hôpi- cells, confirming the functional inter- tal Européen Georges Pompidou, AP-HP, Paris relationship between these DDR kinases colorectal cancer patients display ongo- 75015, France; 9Karolinska Institute, Department that has previously been reported [6]. ing RS response, at baseline, including of Women’s and Children’s Health, Karolinska University Hospital, Stockholm 17176, Sweden Finally, taking advantage of patient- the overactivation of ATM, associated Correspondence: Ilio Vitalea, Guido Kroemerb derived glioma cells, Liu and colleagues with TP53 mutations and increased aE-mail: [email protected] demonstrated that the subpopulation chromosomal content [12]. It will be bE-mail: [email protected] expressing the CSC marker prominin interesting to analyze (1) whether the 1 (PROM1/CD133) displayed higher TP53 status and/or aberrant karyotype References levels of spDSBs, pATM and pSTAT3 play a role in spDSB generation, (2) what is the mechanism preventing DDR 1 Crasta K, Ganem NJ, Dagher R, et al. Na- than the CD133-negative subpopula- ture 2012; 482:53-58. tion. By employing CRISPR/Cas9 execution upon spDSB-mediated ATM 2 Janssen A, van der Burg M, Szuhai K, et al. technology on these patient-derived activation, and (3) whether a threshold Science 2011; 333:1895-1898. glioma cells, Liu et al. demonstrated level of spDSB tolerability exists and, 3 Zeman MK, Cimprich KA. Nat Cell Biol if so, might be exploited for therapeutic 2014; 16:2-9. that ATM knockout decreased the CSC 4 O’Connor MJ. Mol Cell 2015; 60:547-560. fraction. This was accompanied by the purposes. 5 Vitale I, Manic G, Senovilla L, et al. Trends downregulation of STAT3 and resulted Importantly, the study of Liu and Cancer 2015; 1:124-135. in diminished sphere-growing ability, coauthors support the development of 6 Marechal A, Zou L. Cold Spring Harb Per- biomarker-driven anti-cancer strategies spect Biol 2013; 5:a012716. clonogenicity and in vivo tumor growth 7 Halazonetis TD, Gorgoulis VG, Bartek J. [10]. In these experimental settings, the based on the inhibition of the ATM or Science 2008; 319:1352-1355. forced expression of constitutively ac- ATR axis to deplete aggressive cancer 8 Maugeri-Sacca M, Bartucci M, De Maria R. tive STAT3 rescued the stemness and (stem) cells bearing high levels of Mol Cancer Ther 2012; 11:1627-1636. spDSBs and overactivated ATM/ATR. 9 Kreso A, Dick JE. Cell Stem Cell 2014; tumorigenic potential of patient-derived 14:275-291. glioma cells. To fully translate this evidence into the 10 Liu X, Li F, Huang Q, et al. Cell Res 2017; Reportedly, CSCs of distinct tissue clinics, it appears urgent to identify spe- 27:764-783. origins constitutively activate DDR to cific markers of spDSBs and to elucidate 11 Ichim G, Lopez J, Ahmed SU, et al. Mol the precise origin of spDSBs, and the Cell 2015; 57:860-872. resist genotoxic perturbations [8]. In this 12 Manic G, Signore M, Sistigu A, et al. Gut context, our own observations indicate mechanisms and consequences of their 2017; pii:gutjnl-2016-312623. that a subset of CSCs derived from formation during tumorigenesis. www.cell-research.com | Cell Research | SPRINGER NATURE