HUS1 Regulates in Vivo Responses to Genotoxic Chemotherapies

SHORT COMMUNICATION HUS1 regulates in vivo responses to genotoxic chemotherapies

G Balmus1, PX Lim1, A Oswald1, KR Hume1,2, A Cassano1, J Pierre1, A Hill1, W Huang3, A August3, T Stokol4, T Southard1 and RS Weiss1

Cells are under constant attack from genotoxins and rely on a multifaceted DNA damage response (DDR) network to maintain genomic integrity. Central to the DDR are the ATM and ATR kinases, which respond primarily to double-strand DNA breaks (DSBs) and replication stress, respectively. Optimal ATR signaling requires the RAD9A-RAD1-HUS1 (9-1-1) complex, a toroidal clamp that is loaded at damage sites and scaffolds signaling and repair factors. Whereas complete ATR pathway inactivation causes embryonic lethality, partial Hus1 impairment has been accomplished in adult mice using hypomorphic (Hus1neo) and null (Hus1Δ1) Hus1 alleles, and here we use this system to define the tissue- and cell type-specific actions of the HUS1-mediated DDR in vivo. Hus1neo/Δ1 mice showed hypersensitivity to agents that cause replication stress, including the crosslinking agent mitomycin C (MMC) and the replication inhibitor hydroxyurea, but not the DSB inducer ionizing radiation. Analysis of tissue morphology, genomic instability, cell proliferation and apoptosis revealed that MMC treatment caused severe damage in highly replicating tissues of mice with partial Hus1 inactivation. The role of the 9-1-1 complex in responding to MMC was partially ATR-independent, as a HUS1 mutant that was proficient for ATR-induced checkpoint kinase 1 phosphorylation nevertheless conferred MMC hypersensitivity. To assess the interplay between the ATM and ATR pathways in responding to replication stress in vivo, we used Hus1/Atm double mutant mice. Whereas Hus1neo/neo and Atm−/− single mutant mice survived low-dose MMC similar to wild-type controls, Hus1neo/neoAtm−/− double mutants showed striking MMC hypersensitivity, consistent with a model in which MMC exposure in the context of Hus1 dysfunction results in DSBs to which the ATM pathway normally responds. This improved understanding of the inter-dependency between two major DDR mechanisms during the response to a conventional chemotherapeutic illustrates how inhibition of checkpoint factors such as HUS1 may be effective for the treatment of ATM-deficient and other cancers.

Oncogene (2016) 35, 662–669; doi:10.1038/onc.2015.118; published online 27 April 2015

INTRODUCTION The ATR pathway responds to lesions caused by genotoxins Effective cancer treatment frequently involves DNA-damaging such as ultraviolet-light, HU or MMC, which physically impede agents, such as ionizing radiation (IR), which induces double- replication fork progression, leading to accumulation of strand DNA breaks (DSBs), and various chemotherapeutics that, for RPA-coated single-stranded DNA. This represents the major example, cause interstrand crosslinks (ICLs) (mitomycin C (MMC), recruitment signal for ATR and its partner ATRIP, as well as the Cisplatin), perturb nucleotide metabolism (5-fluoruracil, hydro- RAD9A-RAD1-HUS1 (9-1-1) complex, a toroidal clamp that xyurea (HU)), inhibit topoisomerases (Topotecan), or are clastogenic structurally resembles proliferating cell nuclear antigen. The 9-1- (Bleomycin). Cells respond to such genotoxic threats by activating a 1 complex is loaded at 5′ recessed DNA ends and interacts with DNA damage response (DDR) that consists of a series of signaling TOPBP1 and RHINO, which are required to stimulate ATR kinase pathways that sense genome damage and induce effector activity.8,9 Activated ATR phosphorylates CHK1 and other down- molecules that execute protective responses, including repair, stream effectors to promote cell cycle arrest, replication fork senescence or apoptosis. The mammalian DDR includes two major 1,2 stability and DNA repair. Independently of ATR activation, the pathways, headed by the related kinases ATM and ATR. 9-1-1 complex also directly interacts with base excision repair ATM is activated primarily by DSBs, as well as oxidative stress – factors,10 12 translesion synthesis polymerases13 and members of and chromatin changes, and phosphorylates a host of down- other repair pathways, such as homologous recombination and stream effectors, including checkpoint kinase 2 (CHK2) and p53. In 14 humans, ATM mutations cause Ataxia-Telangiectasia (AT), a mismatch repair. In humans, hypomorphic loss-of-function ATR mutations are genomic instability (GIN) syndrome characterized by extreme IR 15 sensitivity, ataxia due to cerebellar degeneration, telangiectasias, associated with the rare disease Seckel Syndrome. Seckel fi 2,3 patients present with developmental defects, including dwarfism, immune de ciency and sterility. AT patients demonstrate a 16,17 significantly increased risk of hematological malignancies, reflect- microcephaly and cognitive impairment. Mouse models ing the tumor-suppressor functions of ATM. Mouse models of AT targeting the ATR pathway revealed that loss of individual ATR recapitulate many of the human phenotypes.4–6 Consistent with pathway components causes embryonic lethality, highlighting a the concept that ATM is the main responder to DSBs, Atm- continuous need for ATR pathway-mediated protection against – deficient mice succumb to low-dose IR due to gastrointestinal replication stress.18 23 Moreover, mice with partial ATR expression toxicity4 but show normal survival in response to replication stress, show severe phenotypes, including dwarfism, craniofacial defects such as that caused by low-dose MMC.7 and early lethality.24

1Department of Biomedical Sciences, Cornell University, Ithaca, NY, USA; 2Department of Clinical Sciences, Cornell University, Ithaca, NY, USA; 3Department of Microbiology and Immunology, Cornell University, Ithaca, NY, USA and 4Department of Population Medicine and Diagnostic Sciences, Cornell University, Ithaca, NY, USA. Correspondence: Dr RS Weiss, Department of Biomedical Sciences, Cornell University, T2-006C Veterinary Research Tower, Ithaca, NY 14853, USA. E-mail: [email protected] Received 1 July 2014; revised 8 March 2015; accepted 10 March 2015; published online 27 April 2015 Mitomycin C hypersensitivity in mice with partial Hus1 loss G Balmus et al 663 The role of the ATR pathway in cancer is complex and may differ in several cancer cell lines, particularly those with ATM or p53 depending on the particular stage of carcinogenesis. Some studies defects,38–41 and have shown promising results in clinical have proposed that the ATR pathway is critical for tumor trials.32,42 Thus ATR pathway dependency may represent a key suppression while others suggest an enabling role in supporting vulnerability of cancer cells.42,43 tumor growth.15,25 ATR mutations have been identified in human Because inactivation of genes encoding 9-1-1 complex subunits cancers and ATR pathway mouse mutants also have been causes embryonic lethality, we previously developed a system reported to show increased cancer predisposition in some for partial Hus1 inactivation based on a hypomorphic Hus1 instances,20,26,27 leading to the suggestion that GIN prevention allele (Hus1neo) that expresses wild-type HUS1 at reduced by the ATR pathway may suppress tumor initiation. However, levels.44,45 Hus1neo was combined with null (Hus1Δ1) or wild-type other studies indicate that ATR pathway genes are upregulated in (Hus1+) Hus1 alleles to create a Hus1 allelic series with tumors28,29 and increased expression of DDR proteins, including incrementally decreased Hus1 expression (Hus1+/+4Hus1+/neo4 HUS1 and CHK1, is observed in mouse mammary tumors.30 These Hus1+/Δ14Hus1neo/neo4Hus1neo/Δ1). Hus1+/neo and Hus1+/Δ1 mice findings are consistent with observations that in some contexts are indistinguishable from Hus1+/+ mice and in our studies are the ATR pathway is necessary for cancer cells to tolerate stresses used as normal controls. Hus1neo/Δ1 mice, which show reduced associated with neoplastic proliferation,27,31–33 a phenomenon Hus1 expression in a variety of tissues (Supplementary Figure S1), termed non-oncogene addiction that reflects the importance of are born at expected frequencies and appear grossly normal. They DDR factors in cancers and associated opportunities for ther- display no detectable increase in cancer predisposition but show apeutic intervention.34 Similar to established synthetic lethal elevated micronucleus formation in peripheral blood cells, strategies designed to exploit dependencies in DNA repair- indicating that the reduced Hus1 expression is insufficient for defective cancers,35–37 ATR pathway inhibitors have been used proper genome maintenance.44 Hus1neo/neo cells with intermediate to sensitize cancer cells to common genotoxic chemotherapeutics Hus1 expression show moderate GIN phenotypes.

Figure 1. Hypersensitivity to MMC but not IR in mice with partial Hus1 inactivation. Previously described Hus1neo, Hus1Δ1 and Atm knockout mice were utilized in accordance with federal and institutional guidelines, under a protocol approved by the Cornell University Institutional Animal Care and Use Committee. For analysis of DNA damage responses in vivo, mice were subjected to 15 Gy (a) or 9 Gy (b) γ-irradiation using a Mark I Model 68 sealed Cesium source gamma irradiator (JL Shepherd and Associates, San Fernando, CA, USA) or injected intraperitoneally with a single dose of 8 mg (c) or 4 mg (d) per kg body weight MMC (Sigma, St Louis, MO, USA). The mice were subsequently monitored daily for up to 30 days or until reaching humane end point criteria. Control Hus1+ genotypes include Hus1+/+, Hus1+/neo and Hus1+/Δ1. Kaplan–Meier survival analysis was performed using SPSS software (ver 18, Chicago, IL, USA). Hus1neo/Δ1 mice showed no signiﬁcant change in sensitivity to IR at either 15 Gy (P = 0.768, log-rank) (a) or 9 Gy (P = 0.674, log-rank) (b) doses as compared with the Hus1+ littermate controls, whereas Atm−/− mice were hypersensitive to IR (Po0.001, log-rank). By contrast, Hus1neo/Δ1 mice showed marked sensitivity to MMC at both 8 mg/kg (P = 0.009, log-rank) (c) and 4 mg/kg (Po0.001, log-rank) (d) doses.

© 2016 Macmillan Publishers Limited Oncogene (2016) 662 – 669 Mitomycin C hypersensitivity in mice with partial Hus1 loss G Balmus et al 664 Here we used Hus1 hypomorph mice and cells to test the ATM and ATR pathways are specialized to respond to particular requirements for the 9-1-1 complex following in vivo genotoxin DNA lesions. These data raise the intriguing possibility that treatment. We show that Hus1neo/Δ1 mice are no more sensitive to naturally occurring sequence variants or cancer-associated altera- the DSB-inducing agent IR than controls but are hypersensitive to tions in Hus1 or related genes could influence chemosensitivity in the replication stress-inducers HU and MMC, through a mechanism patients.47 that is partially independent of ATR-mediated checkpoint signaling. To complement our assessment of survival following genotoxin Furthermore, we provide evidence that MMC exposure under exposure, we performed histopathological assessments and conditions of impaired Hus1 function results in DSBs to which the analyzed proliferation, apoptosis and GIN levels at various time ATM pathway responds. Consequently, Atm-deficient cancer cells points after MMC exposure, before the emergence of severe may be selectively killed by Hus1 impairment in combination with clinical signs. We initially focused our analysis on the small low-dose chemotherapy. These findings provide in vivo validation intestine because intestinal crypts are known to be sensitive to for current models of genotoxin-specific DDRs and illustrate how genotoxin exposure.48 At 24 and 48 h post-MMC treatment, the Δ modulation of checkpoint function can be used to enhance the intestines of Hus1neo/ 1 mice showed a significant increase in efficacy of conventional chemotherapeutics. damaged crypts characterized by crypt epithelial attenuation or necrosis and increased presence of neutrophils in lamina propria (Figures 2a and b, Supplementary Figures S3A and B). By 96 h RESULTS AND DISCUSSION post-MMC treatment, significant differences in the frequency of To test how an ATR pathway defect impacts in vivo sensitivity to damaged crypts were no longer observed, but at this time point Δ different DNA-damaging agents, we treated Hus1neo/ 1 mice and the small intestine from Hus1neo/Δ1 mice showed large areas where control littermates with IR, HU or MMC. Following whole-body IR crypts were completely disrupted or absent (Supplementary treatment, there were no significant differences in survival Figures S3C and S4A). We hypothesize that these severe Δ between Hus1neo/ 1 mice and Hus1+ controls. Atm-deficient mice, pathologies in Hus1neo/Δ1 mice disrupt the intestinal barrier, on the other hand, showed dramatic IR hypersensitivity as contributing to the observed lethality. Additional analyses previously reported4 (Figures 1a and b). In stark contrast, upon revealed significantly reduced proliferation at 24 h post-MMC Δ in utero HU treatment, Hus1neo/ 1 embryos showed significantly treatment in mice with partial Hus1 impairment (Figures 2c and d) increased cell death compared with littermate controls and a significant increase in apoptosis at 48 h (Figures 2e and f), (Supplementary Figure S2), highlighting a requirement for Hus1 accompanied by a significant increase in DNA damage as in replication stress responses in vivo. Proliferation, however, was indicated by staining for the DSB marker γH2AX in the crypts of reduced following HU to a similar extent in Hus1-deficient and MMC-treated Hus1neo/Δ1 mice (Figures 2g and h). In contrast, -proficient embryos. The analysis of replication stress sensitivity no Hus1 genotype-specific differences were observed for was extended to adult mice using MMC. In vitro, HUS1 down- apoptosis or proliferation in the small intestine following IR regulation previously was shown to increase cellular sensitivity to treatment, although Hus1neo/Δ1 mice did show a significant another ICL inducer, cisplatin.46 Following MMC treatment increase in damaged crypts at 24 h post-IR (Supplementary Hus1neo/Δ1 mice showed significant hypersensitivity compared Figures S4B, S5 and S6). with littermate controls (Figures 1c and d). Together, these data Because genotoxin sensitivity in the intestine relates in part to show for the first time that HUS1 dysfunction disrupts protective the proliferative nature of the tissue, we next tested whether responses to replication stress in vivo. Whereas mice with partial relatively quiescent tissues, including liver and skin, were Hus1 impairment are hypersensitive to the replication stress- sensitized to genome damage by partial Hus1 impairment but inducing agents HU and MMC, they survive exposure to the observed similar IR and MMC responses regardless of Hus1 clastogen IR as well as wild-type mice, establishing in vivo that the genotype (data not shown). We therefore assessed the

Figure 2. Decreased proliferation and increased GIN and apoptosis in the small intestine of Hus1neo/Δ1 as compared with Hus1+ mice following MMC treatment. At the indicated time points after MMC treatment, animals were humanely euthanized, and tissues were collected, formalin fixed, embedded in paraffin and serially sectioned at 5-μm thickness. Hematoxylin and eosin (H&E) staining was performed using a standard protocol. Staining for bromodeoxyuridine (BrdU) incorporation, a synthetic thymidine analog, was used to quantify cell proliferation. Two hours prior to euthanasia, mice were injected intraperitoneally with 8 mg/kg BrdU, and staining was performed with a BrdU Kit (Invitrogen, Norwalk, CT, USA) according to the manufacturer’s directions. Apoptosis quantitation was done using the Apoptag (Millipore, Billerica, MA, USA) TUNEL detection system according to the manufacturer's instructions. γ-H2AX staining was done using a mouse monoclonal antibody (Ser139, clone JBW301, Millipore) and a mouse secondary detection kit (Invitrogen). Images were obtained using an Aperio Scanscope (Aperio Technologies, Vista, CA, USA). n ⩾ 3 mice for each genotype and time point were used for quantitation. Quantitation was done on intact intestinal crypts, and at least five intestinal crypts were counted per mouse. Statistical analysis was performed using SPSS. Error bars represent s.d. (a) Representative H&E images showing increased damage in Hus1neo/Δ1 crypts when compared with Hus1+ controls after 4 mg/kg MMC. (b) Quantification of the proportion of damaged crypts at the indicated times post-MMC treatment showed a significant increase in Hus1neo/Δ1 mice at both 24 and 48 h posttreatment (*Po0.05, Mixed Model Analysis) (c) Representative images depicting BrdU-positive cells, indicative of proliferation, in the small intestinal crypts in Hus1neo/Δ1 mice and littermate controls after no treatment or 24 h after 4 mg/kg MMC. (d) Quantification of BrdU-positive cells in the small intestinal crypts at the indicated times post-MMC treatment. In mice from both groups, proliferation was similar in the absence of treatment, largely unchanged at 8 h post-MMC treatment, and reduced at 24 and 48 h after treatment, before returning to basal levels at 96 h after treatment. Significantly decreased proliferation was observed in Hus1neo/Δ1 mice as compared with littermate controls at 24 h post-MMC (*Po0.05, Mixed Model Analysis). (e) Representative images depicting TUNEL-positive apoptotic cells in small intestinal crypts from Hus1neo/Δ1 mice and Hus1+ controls after no treatment or at the indicated time points after 4 mg/kg MMC. (f) Quantification of TUNEL-positive cells in the small intestinal crypts at the indicated times post-MMC. At 8 h post-MMC treatment, mice from both genotypic groups showed significantly increased levels of apoptosis when compared with the untreated group (*Po0.01, Mixed Model Analysis), and by 96 h the level of apoptosis returned to basal levels. Significantly increased apoptosis was observed in Hus1neo/Δ1 mice as compared with littermate controls at 48 h post-MMC (*Po0.05, Mixed Model Analysis). (g) Representative images depicting γ-H2AX-positive cells, indicative of DNA damage, in small intestinal crypts from Hus1neo/Δ1 mice and littermate controls after no treatment or at the indicated time points after 4 mg/kg MMC. (h) Quantification of positive cells showed that γ-H2AX levels were increased at 8 h posttreatment and then gradually returned toward basal levels by 96 h posttreatment. There was a significant increase in γ-H2AX in Hus1neo/Δ1 mice as compared with littermate controls at 24 h posttreatment (*Po0.01, Mixed Model Analysis). Size bars correspond to 50 microns.

hematopoietic compartment as another tissue with actively signiﬁcantly elevated in both the bone marrow and thymus of dividing cells that might be affected by Hus1 status. Hus1neo/Δ1 mice with partial Hus1 impairment (Supplementary Figures S7C–F). mice showed a delayed increase in myelopoiesis after MMC Peripheral blood proﬁling post-MMC revealed that, whereas white treatment, and dysplastic changes were evident in myeloid cells in blood cell and neutrophil counts were increased slightly in control the bone marrow (Supplementary Figures S7A and B). MMC- mice, these changes were reduced in mice with partial Hus1 induced DNA damage as assessed by γH2AX staining was inactivation, consistent with myelopoiesis defects (Supplementary

Figure 3. HUS1 mediates ATR-independent functions in response to MMC. (a) HUS1 mutant R4D+I152Y is properly expressed in 293T cells. A HUS1 mutant R4D+I152Y expression plasmid was generated using a Agilent QuikChange Lightning Multi Site-directed Mutagenesis Kit (Santa Clara, CA, USA) with CMV-mHUS1 plasmid as the template and 5'-CGCGTGGATCCATGAAGTTTGACGCCAAGATC GTGGACC-3' and 5'-AAGGACTTACAAGAACCCTCCTACCCAGACT GTGACGTCAGTATT-3' as mutagenic primers. The empty vector or plasmids containing WT HUS1 or HUS1 R4D+I152Y were transiently transfected into 293 T cells using polyethyleneimine with a 4:1 volume to plasmid weight ratio. The transfected cells were lysed and immunoblotted with antibodies specific for HUS169 or β-ACTIN (Sigma A5316) as a loading control. (b) Cells expressing HUS1 mutant R4D+I152Y displayed partial MMC hypersensitivity. For stable transfection in Hus1−/− p21−/− MEFs, a pGK puro plasmid was cotransfected with the Hus1 expression plasmids. Cells were then selected and maintained in media containing 1.83 μg/ml puromycin. The resulting stable cell pools were seeded in six-well dishes in triplicate for each treatment condition and allowed to proliferate for 2 days. The cells were then either left untreated or treated with 0.25, 0.5 and 1.0 μg/ml of MMC for an hour. After MMC removal, the cells were allowed to recover for 3 days and then harvested and counted using an ORFLO Moxi Z mini automated cell counter (Ketchum, ID, USA). Error bars represent s.d. Statistical comparisons were computed by two-way analysis of variance. Survival by cells expressing WT HUS1 was significantly different than that by cells expressing the R4D+I152Y mutant or empty vector; the difference between cells expressing the R4D+I152Y mutant or empty vector also was statistically significant (*Po0.05). (c) MMC- induced CHK1 phosphorylation was unperturbed in cells expressing HUS1 mutant R4D+I152Y. Hus1-deficient MEFs stably transfected with the indicated plasmids were left untreated or incubated with 10 μg/ml MMC for 3 h. Cell extracts were then prepared and immunoblotted using antibodies specific for phospho-CHK1 (S345; Cell Signaling #2341, Danvers, MA, USA), CHK1 (Santa Cruz #G-4, Table S1). Together, these data indicate that highly replicating Santa Cruz, CA, USA) or β-ACTIN (Sigma A5316) as a loading control. tissues require HUS1 to tolerate replicative genotoxic stress. Cells lacking HUS1 show significantly diminished CHK1 phosphor- To test whether the defective DDR associated with partial Hus1 ylation following MMC treatment. Both wild-type HUS1 and the impairment was related to the requirement for the 9-1-1 complex HUS1 mutant R4D+I152Y were capable of restoring MMC-induced in ATR activation, we monitored genotoxin-induced activation of pCHK1 levels, suggesting that ATR checkpoint signaling is intact in o CHK1, an ATR substrate that is robustly phosphorylated following the mutant HUS1-expressing cells. *P 0.05. ultraviolet-light treatment and to a lesser extent after IR or MMC (Supplementary Figure S8). Compared with control Hus1+/neo Δ complex not only promotes ATR activation but additionally mouse embryonic fibroblasts (MEFs), Hus1neo/ 1 cells showed mediates ATR-independent functions through HUS1, likely invol- reduced CHK1 activation. Thus one element of MMC hypersensi- ving interactions with effector proteins at DNA lesions.52,53 tivity in cells with Hus1 dysfunction may relate to impaired ATR Because the 9-1-1 complex and the ATR pathway protect activation, consistent with the well-established requirements for replication fork integrity and suppress DSB formation under ATR in replication fork stabilization and signaling through the conditions of replication stress, we hypothesized that the elevated Fanconi Anemia pathway that are critical for the response to 49 γH2AX staining we observed in MMC-treated Hus1 hypomorph crosslinking agents, such as MMC. To directly test the requirement of the 9-1-1 complex for Fanconi Anemia signaling, we mice might be explained by the conversion of MMC-induced ICLs monitored MMC-induced FANCD2 ubiquitination.50 Basal FANCD2 into DSBs that trigger an ATM-dependent DDR. This hypothesis is fi consistent with recent findings that the ATM and ATR pathways monoubiquitination was signi cantly increased in untreated Hus1- 1,54–56 deficient cells, consistent with chronic replication stress in these function as an integrated DDR network. ATR defects, for 51 instance, cause sensitivity not only to replication stress-inducing cells, but did not increase further upon MMC treatment in the 57,58 absence of Hus1 (Supplementary Figure S9). We next investigated agents but also to clastogens. Concomitant ATM and ATR pathway inhibition results in synthetic lethality, further emphasizing whether HUS1 also mediated ATR signaling-independent func- 24,51,59 tions in response to MMC, by creating a HUS1 mutant with the cooperative nature of the DDR. We previously reported alterations in two HUS1 outer surface residues, R4 and I152, synthetic lethality upon combining the most severe Hus1 neo/Δ1 located in candidate-binding pockets for HUS1-associated pro- defect in our allelic series (Hus1 ) with Atm deficiency, neo/neo teins. The R4D+I152Y double mutant, which was expressed at whereas a more moderate level of Hus1 impairment (Hus1 ) levels comparable to those of wild-type HUS1 (Figure 3a), was enabled generation of viable Hus1/Atm double mutant mice that used to complement Hus1-null MEFs. Whereas wild-type HUS1 showed dwarfism, craniofacial abnormalities and digit corrected the MMC hypersensitivity of Hus1-null MEFs, cells abnormalities.60 Interestingly, simultaneous Hus1/Atm impairment − − expressing the R4D+I152Y mutant showed significant, though in Hus1neo/neoAtm / mice did not result in greater IR sensitivity partial, MMC hypersensitivity (Figure 3b). Importantly, the R4D than Atm deficiency alone. These findings support the idea that +I152Y mutant was proficient for MMC-induced CHK1 phosphor- ATM mediates the primary DSB response and are consistent with a ylation (Figure 3c), as would be expected given that stimulation of model in which the synthetic phenotypes in Hus1/Atm double ATR by the 9-1-1 complex occurs through the RAD9A subunit.8 mutants occur because Hus1 dysfunction results in DSBs or other Together, these data suggest that, in response to MMC, the 9-1-1 lesions that feed into the ATM pathway for resolution.

Figure 4. Increased dependency on the ATM pathway following MMC treatment of mice and cells with partial Hus1 impairment. (a) Kaplan– Meier survival analysis of control mice, including Hus1+/neoAtm+/+ (n = 6), Hus1neo/neoAtm+/+ (n = 11) and Hus1+/neoAtm−/− (n = 4) mice, as compared with Hus1neo/neoAtm−/− (n = 7) double mutant mice following intraperitoneal injection of 4 mg/kg MMC. No lethality was observed following MMC treatment of control and single mutant mice, whereas MMC sensitivity was significantly increased in Hus1/Atm double mutant mice (Po0.001; log-rank). (b) Flow cytometric analysis of cultured lymphoma cells from Hus1+/neoAtm−/− and Hus1neo/neoAtm−/− mice. For thymic lymphoma culture, tumors were isolated from mice of the selected genotypes, disaggregated and grown in RPMI 1640 with 25 mM HEPES, 200 mM L-glutamine (Lonza, Basel, Switzerland) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Hyclone, Logan, UT, USA), 1% penicillin and streptomycin (Mediatech, Manassas, VA, USA), 1% nonessential amino acids (Mediatech), 55 μM 2β-mercaptoethanol (Sigma) and stimulated with phytohaemagglutinin (PHA-P; Sigma). Live cells were stained using phycoerythrin-conjugated anti-CD4 (BD Biosciences, Franklin Lakes, NJ, USA; clone GK1.5) and fluorescein-conjugated anti-CD8a (BD Biosciences, clone 53-6.7) antibodies, in the presence of Fc block (eBioscience, San Diego, CA, USA), in 2% FBS/phosphate-buffered saline. Flow data were acquired on a LSRII flow cytometry system (BD Biosciences), and analyzed using FlowJo software (FlowJo LLC., Ashland, OR, USA). (c) Primary Hus1+/neoAtm−/− and Hus1neo/neoAtm−/− lymphoma cultures were treated with the indicated doses of MMC, and viability was assessed 24 h later. MMC sensitivity was significantly greater in Hus1/Atm double mutant (Hus1neo/neoAtm−/−) lymphoma cells as compared with Atm single mutant (Hus1+/neoAtm−/−) controls (*Po0.01, ANOVA). (d) Hypothetical model showing the interaction between the ATM and ATR pathways in response to genotoxic stress caused by IR or MMC. Whereas Atm-null mice are hypersensitive to IR independent of Hus1 status, sub-optimal Hus1 expression causes increased sensitivity to replication stress-inducing agents, such as MMC. MMC treatment of mice with partial Hus1 impairment results in DSB accumulation, creating a greater requirement for ATM-mediated genome protection and leading to synthetic lethality in Atm/Hus1 double mutant mice upon MMC treatment. We propose that pharmacological impairment of HUS1 or possibly other elements of the ATR pathway could be used to sensitize ATM-deficient tumor cells to genotoxic chemotherapies.

To test this hypothesis in the context of replication stress, intact ATM function, a moderate Hus1 defect as in Hus1neo/neo mice we treated Hus1/Atm double mutant mice with MMC. can be overcome through the ATM pathway, resulting in limited Although Hus1neo/neo or Atm−/− single mutant mice showed no MMC hypersensitivity (Figure 4a), whereas a more severe detectable MMC hypersensitivity compared with wild-type Hus1 defect in Hus1neo/Δ1 mice is incompatible with survival − − controls, Hus1neo/neoAtm / double mutant mice were highly following MMC exposure (Figure 1d), even in the presence of sensitive to low-dose MMC (Figure 4a). These data reinforce the functional ATM. idea of a highly integrated DDR, such that mice with partial Hus1 The striking MMC hypersensitivity in Hus1/Atm double mutant impairment cannot properly respond to ICLs, leading to DSBs that mice led us to hypothesize that Hus1 impairment would sensitize require the ATM pathway for survival (Figure 4d). In mice with Atm mutant cancers to chemotherapy. Our previous research

© 2016 Macmillan Publishers Limited Oncogene (2016) 662 – 669 Mitomycin C hypersensitivity in mice with partial Hus1 loss G Balmus et al 668 showed that partial Hus1 impairment does not significantly alter 11 Wang W, Lindsey-Boltz LA, Sancar A, Bambara RA. Mechanism of stimulation of the kinetics or histopathological features of thymic lymphomas in human DNA ligase I by the Rad9-rad1-Hus1 checkpoint complex. J Biol Chem Atm-deficient mice.60 We therefore cultured lymphomas from Atm 2006; 281: 20865–20872. single and Hus1/Atm double mutant mice and compared their 12 Friedrich-Heineken E, Toueille M, Tännler B, Bürki C, Ferrari E, Hottiger MO et al. MMC sensitivity. The tumors were composed mainly of CD4 and The two DNA clamps Rad9/Rad1/Hus1 complex and proliferating cell nuclear 353 CD8-positive cells in mice of either genotype (Figure 4b) as antigen differentially regulate Flap endonuclease 1 activity. J Mol Biol 2005; : 980–989. previously reported for Atm−/− mice.4 Notably, lymphomas from fi 13 Kai M, Wang TSF. Checkpoint activation regulates mutagenic translesion synth- Hus1/Atm double mutant mice were signi cantly more sensitive to esis. Genes Dev 2003; 17:64–76. −/− low-dose MMC than those from Atm mutants (Figure 4c; 14 Bai H, Madabushi A, Guan X, Lu A-L. Interaction between human mismatch repair Po0.01 Mixed Model Analysis). Together, these data argue that recognition proteins and checkpoint sensor Rad9-Rad1-Hus1. DNA Repair 2010; 9: targeted inhibition of the 9-1-1 complex may be an effective 478–487. strategy for tumors with ATM pathway defects, which represent a 15 O'Driscoll M. Diseases associated with defective responses to DNA damage. Cold significant proportion of all cancers.61 Importantly, loss of either Spring Harb Perspect Biol 2012; 4: a012773. HUS1 or ATR in mice synergizes with p53 loss to promote cell 16 Goodship J, Gill H, Carter J, Jackson A, Splitt M, Wright M. Autozygosity mapping 62,63 of a Seckel Syndrome locus to chromosome 3q22.1-q24. Am J Hum Genet 2000; death and not tumorigenesis. ATR pathway inhibitors are 67 – currently in clinical trials for cancer therapy,64 and our data :498 503. 17 O'Driscoll M, Ruiz-Perez VL, Woods CG, Jeggo PA, Goodship JA. A splicing support their use in conjunction with replication stress-inducing mutation affecting expression of ataxia-telangiectasia and Rad3-related protein agents, such as HU or MMC. Given that the 9-1-1 complex also has (ATR) results in Seckel syndrome. Nature Genet 2003; 33: 497–501. ATR-independent functions, our findings back the development of 18 Han L, Hu Z, Liu Y, Wang X, Hopkins KM, Lieberman HB et al. Mouse Rad1 deletion inhibitors of the 9-1-1 complex itself, in addition to existing drugs enhances susceptibility for skin tumor development. Mol Cancer 2010; 9:67. that target ATR or CHK1. Critical for predicting the success of such 19 Jeon Y, Ko E, Lee KY, Ko MJ, Park SY, Kang J et al. TopBP1 deficiency causes an therapies will be knowledge of ATM pathway status and the early embryonic lethality and induces cellular senescence in primary cells. J Biol replicative potential of the tumors. Together with growing Chem 2011; 286: 5414–5422. evidence that ATR, CHK1 and the 9-1-1 complex are important 20 Brown EJ, Baltimore D. ATR disruption leads to chromosomal fragmentation and 14 – for tumor progression,38,39,42,65–68 these data illustrate how early embryonic lethality. Genes Dev 2000; :397 402. fi fi 21 Liu Q, Guntuku S, Cui XS, Matsuoka S, Cortez D, Tamai K et al. Chk1 is an essential pathway-speci c DDR inhibitors could enhance the ef cacy and kinase that is regulated by Atr and required for the G(2)/M DNA damage minimize the toxicity of conventional chemotherapies. checkpoint. Genes Dev 2000; 14: 1448–1459. 22 Weiss RS, Enoch T, Leder P. Inactivation of mouse Hus1 results in genomic 14 CONFLICT OF INTEREST instability and impaired responses to genotoxic stress. Genes Dev 2000; : 1886–1898. The authors declare no conflict of interest. 23 Hopkins KM, Auerbach W, Wang XY, Hande MP, Hang H, Wolgemuth DJ et al. Deletion of mouse rad9 causes abnormal cellular responses to DNA damage, genomic instability, and embryonic lethality. Mol Cell Biol 2004; 24: 7235–7248. ACKNOWLEDGEMENTS 24 Murga M, Bunting S, Montaña MF, Soria R, Mulero F, Cañamero M et al. A mouse We thank Amy Lyndaker, Steve Jackson and Yaron Galanti for helpful discussions and model of ATR-Seckel shows embryonic replicative stress and accelerated aging. comments on the manuscript; Natasha Karp for help with statistical analysis and the Nature Genet 2009; 41:891–898. staff of the Cornell Lab Animal Services and CARE programs for excellent animal care. 25 Smith J, Tho LM, Xu N, Gillespie DA. The ATM-Chk2 and ATR-Chk1 pathways in This work was supported by a National Institutes of Health grant R01 CA108773 to DNA damage signaling and cancer. Adv Cancer Res 2010; 108:73–112. RSW; a Cornell University College of Veterinary Medicine Graduate Research 26 Lam MH, Liu Q, Elledge SJ, Rosen JM. Chk1 is haploinsufficient for multiple Assistantship to GB; a Cornell University College of Veterinary Medicine Clinical functions critical to tumor suppression. Cancer Cell 2004; 6:45–59. Fellowship to KRH; and a National Center for Research Resources grant 27 Gilad O, Nabet BY, Ragland RL, Schoppy DW, Smith KD, Durham AC et al. (S10RR023781) for instrumentation. Combining ATR suppression with oncogenic Ras synergistically increases genomic instability, causing synthetic lethality or tumorigenesis in a dosage- dependent manner. Cancer Res 2010; 70: 9693–9702. REFERENCES 28 Maniwa Y, Yoshimura M, Bermudez VP, Yuki T, Okada K, Kanomata N et al. Accumulation of hRad9 protein in the nuclei of nonsmall cell lung carcinoma cells. 1 Cimprich KA, Cortez D. ATR: an essential regulator of genome integrity. Nat Rev Cancer 2005; 103:126–132. Mol Cell Biol 2008; 9:616–627. 2 McKinnon PJ. ATM and the molecular pathogenesis of ataxia telangiectasia. Annu 29 Zhu A, Zhang CX, Lieberman HB. Rad9 has a functional role in human prostate 68 – Rev Pathol 2012; 7: 303–321. carcinogenesis. Cancer Res 2008; : 1267 1274. 3 McKinnon PJ, Caldecott KW. DNA strand break repair and human genetic disease. 30 Zhang M, Behbod F, Atkinson RL, Landis MD, Kittrell F, Edwards D et al. fi Annu Rev Genomics Hum Genet 2007; 8:37–55. Identi cation of tumor-initiating cells in a p53-null mouse model of breast cancer. 68 – 4 Barlow C, Hirotsune S, Paylor R, Liyanage M, Eckhaus M, Collins F et al. Cancer Res 2008; : 4674 4682. Atm-deficient mice: a paradigm of ataxia telangiectasia. Cell 1996; 86:159–171. 31 Murga M, Campaner S, Lopez-Contreras AJ, Toledo LI, Soria R, Montaña MF et al. 5 Elson A, Wang Y, Daugherty CJ, Morton CC, Zhou F, Campos-Torres J et al. Exploiting oncogene-induced replicative stress for the selective killing of Myc- 18 – Pleiotropic defects in ataxia-telangiectasia protein-deficient mice. Proc Natl Acad driven tumors. Nat Struct Mol Biol 2011; : 1331 1335. Sci USA 1996; 93: 13084–13089. 32 Schoppy DW, Ragland RL, Gilad O, Shastri N, Peters AA, Murga M et al. Oncogenic 6 Xu Y, Ashley T, Brainerd EE, Bronson RT, Meyn MS, Baltimore D. Targeted dis- stress sensitizes murine cancers to hypomorphic suppression of ATR. J Clin Invest ruption of ATM leads to growth retardation, chromosomal fragmentation during 2012; 122:241–252. meiosis, immune defects, and thymic lymphoma. Genes Dev 1996; 10: 2411–2422. 33 Kawasumi M, Lemos B, Bradner JE, Thibodeau R, Kim Y-S, Schmidt M et al. 7 Kass EM, Helgadottir HR, Chen C-C, Barbera M, Wang R, Westermark UK et al. Protection from UV-induced skin carcinogenesis by genetic inhibition of the Double-strand break repair by homologous recombination in primary mouse ataxia telangiectasia and Rad3-related (ATR) kinase. Proc Natl Acad Sci USA 2011; somatic cells requires BRCA1 but not the ATM kinase. Proc Natl Acad Sci USA 2013; 108: 13716–13721. 110: 5564–5569. 34 Luo J, Solimini NL, Elledge SJ. Principles of cancer therapy: oncogene and non- 8 Delacroix S, Wagner JM, Kobayashi M, Yamamoto K-I, Karnitz LM. The Rad9-Hus1- oncogene addiction. Cell 2009; 136: 823–837. Rad1 (9-1-1) clamp activates checkpoint signaling via TopBP1. Genes Dev 2007; 21: 35 Jackson SP, Bartek J. The DNA-damage response in human biology and disease. 1472–1477. Nature 2009; 461: 1071–1078. 9 Cotta-Ramusino C, McDonald ER, Hurov K, Sowa ME, Harper JW, Elledge SJ. A DNA 36 Farmer H, McCabe N, Lord CJ, Tutt ANJ, Johnson DA, Richardson TB et al. damage response screen identifies RHINO, a 9-1-1 and TopBP1 interacting protein Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. required for ATR signaling. Science 2011; 332: 1313–1317. Nature 2005; 434:917–921. 10 Helt CE, Wang W, Keng PC, Bambara RA. Evidence that DNA damage detection 37 Ellisen LW. PARP inhibitors in cancer therapy: promise, progress, and puzzles. machinery participates in DNA repair. Cell Cycle 2005; 4: 529–532. Cancer Cell 2011; 19:165–167.

Oncogene (2016) 662 – 669 © 2016 Macmillan Publishers Limited Mitomycin C hypersensitivity in mice with partial Hus1 loss G Balmus et al 669 38 Ma CX, Cai S, Li S, Ryan CE, Guo Z, Schaff TW et al. Targeting Chk1 in p53-deficient 56 Matsuoka S, Ballif BA, Smogorzewska A, McDonald ER, Hurov KE, Luo J et al. ATM triple-negative breast cancer is therapeutically beneficial in human-in-mouse and ATR substrate analysis reveals extensive protein networks responsive to tumor models. J Clin Invest 2012; 122: 1541–1552. DNA damage. Science 2007; 316:1160–1166. 39 Origanti S, Cai S-R, Munir AZ, White LS, Piwnica-Worms H. Synthetic lethality of 57 Adams KE, Medhurst AL, Dart DA, Lakin ND. Recruitment of ATR to sites of ionising Chk1 inhibition combined with p53 and/or p21 loss during a DNA damage radiation-induced DNA damage requires ATM and components of the MRN response in normal and tumor cells. Oncogene 2013; 32:577–588. protein complex. Oncogene 2006; 25: 3894–3904. 40 Reaper PM, Griffiths MR, Long JM, Charrier J-D, Maccormick S, Charlton PA et al. 58 Jazayeri A, Falck J, Lukas C, Bartek J, Smith GCM, Lukas J et al. ATM- and cell cycle- Selective killing of ATM- or p53-deficient cancer cells through inhibition of ATR. dependent regulation of ATR in response to DNA double-strand breaks. Nat Cell Nat Chem Biol 2011; 7: 428–430. Biol 2006; 8:37–45. 41 Hall AB, Newsome D, Wang Y, Boucher DM, Eustace B, Gu Y et al. Potentiation of 59 Weiss RS, Leder P, Vaziri C. Critical role for mouse Hus1 in an S-phase DNA tumor responses to DNA damaging therapy by the selective ATR inhibitor VX-970. damage cell cycle checkpoint. Mol Cell Biol 2003; 23:791–803. Oncotarget 2014; 5: 5674–5685. 60 Balmus G, Zhu M, Mukherjee S, Lyndaker AM, Hume KR, Lee J et al. 42 Bartek J, Mistrik M, Bartkova J. Thresholds of replication stress signaling in cancer Disease severity in a mouse model of ataxia telangiectasia is modulated development and treatment. Nat Struct Mol Biol 2012; 19:5–7. by the DNA damage checkpoint gene Hus1. Hum Mol Genet 2012; 21: 43 Broustas CG, Lieberman HB. DNA damage response genes and the development 3408–3420. of cancer metastasis. Radiat Res 2014; 181: 111–130. 61 Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA, Kinzler KW. Cancer 44 Levitt PS, Liu H, Manning C, Weiss RS. Conditional inactivation of the mouse Hus1 genome landscapes. Science 2013; 339: 1546–1558. cell cycle checkpoint gene. Genomics 2005; 86:212–224. 62 Ruzankina Y, Pinzon-Guzman C, Asare A, Ong T, Pontano L, Cotsarelis G et al. 45 Levitt PS, Zhu M, Cassano A, Yazinski SA, Liu H, Darfler J et al. Genome main- Deletion of the developmentally essential gene ATR in adult mice leads to age- tenance defects in cultured cells and mice following partial inactivation of the related phenotypes and stem cell loss. Cell Stem Cell 2007; 1: 113–126. essential cell cycle checkpoint gene Hus1. Mol Cell Biol 2007; 27: 2189–2201. 63 Yazinski SA, Westcott PMK, Ong K, Pinkas J, Peters RM, Weiss RS. Dual inactivation 46 Kinzel B, Hall J, Natt F, Weiler J, Cohen D. Downregulation of Hus1 by antisense of Hus1 and p53 in the mouse mammary gland results in accumulation of oligonucleotides enhances the sensitivity of human lung carcinoma cells to damaged cells and impaired tissue regeneration. Proc Natl Acad Sci USA 2009; cisplatin. Cancer 2005; 94: 1808–1814. 106: 21282–21287. 47 Church D, Kerr R, Domingo E, Rosmarin D, Palles C, Maskell K et al. “Toxgnostics”: 64 Fokas E, Prevo R, Hammond EM, Brunner TB, McKenna WG, Muschel RJ. Targeting an unmet need in cancer medicine. Nat Rev Cancer 2014; 14: 440–445. ATR in DNA damage response and cancer therapeutics. Cancer Treat Rev 2014; 40: 48 Potten CS. A comprehensive study of the radiobiological response of the murine 109–117. (BDF1) small intestine. Int J Radiat Biol 1990; 58:925–973. 65 Brooks K, Oakes V, Edwards B, Ranall M, Leo P, Pavey S et al. A potent Chk1 49 Flynn RL, Zou L. ATR: a master conductor of cellular responses to DNA inhibitor is selectively cytotoxic in melanomas with high levels of replication stress. Trends Biochem Sci 2011; 36: 133–140. replicative stress. Oncogene 2013; 32:788–796. 50 Kottemann MC, Smogorzewska A. Fanconi anaemia and the repair of Watson and 66 Tang Y, Dai Y, Grant S, Dent P. Enhancing CHK1 inhibitor lethality in glioblastoma. Crick DNA crosslinks. Nature 2013; 493:356–363. Cancer Biol Ther 2012; 13: 379–8867. 51 Weiss RS,, Matsuoka S, Elledge SJ, Leder P. Hus1 acts upstream of Chk1 in a 67 la Torre de J, Gil-Moreno A, García A, Rojo F, Xercavins J, Salido E et al. Expression mammalian DNA damage response pathway. Current Biol 2002; 12:73–77. of DNA damage checkpoint protein Hus1 in epithelial ovarian tumors correlates 52 Xu X, Guardiani C, Yan C, Ivanov I. Opening pathways of the DNA clamps with prognostic markers. Int J Gynecol Pathol 2008; 27:24–32. proliferating cell nuclear antigen and Rad9-Rad1-Hus1. Nucleic Acids Res 2013; 41: 68 Broustas CG, Zhu A, Lieberman HB. Rad9 protein contributes to prostate tumor 10020–10031. progression by promoting cell migration and anoikis resistance. J Biol Chem 2012; 53 Jansen JG, Fousteri MI, de Wind N. Send in the clamps: control of DNA translesion 287: 41324–41333. synthesis in eukaryotes. Mol Cell 2007; 28:522–529. 69 Lyndaker AM, Lim PX, Mleczko JM, Diggins CE, Holloway JK, Holmes RJ et al. 54 Shiotani B, Zou L. ATR signaling at a glance. J Cell Sci 2009; 122: 301–304. Conditional inactivation of the DNA damage response gene Hus1 in mouse testis 55 Maréchal A, Zou L. DNA damage sensing by the ATM and ATR kinases. Cold Spring reveals separable roles for components of the RAD9-RAD1-HUS1 complex in Harb Perspect Biol 2013; 5: a012716–a012716. meiotic chromosome maintenance. PLoS Genet 2013; 9: e1003320.

Supplementary Information accompanies this paper on the Oncogene website (http://www.nature.com/onc)