Selective tumor killing based on specific DNA-damage response deficiencies

Abstract

Organisms constantly undergo various stresses within their life span, which can damage their DNA. In order to maintain genomic stability and counteract the development of unwanted genomic mutations, organisms have evolved a DNA-damage response (DDR) to protect their genome. Due to the critical roles played by DDR in genomic stability, its defects can lead to cellular transformation and potentially tumorigenesis. Consequently, this also provides the opportunity to specifically target tumor cells due to a weakened ability to tolerate genotoxic stresses. In this lies a treatment strategy in which the inhibition of remaining DDR pathways can hyper-sensitize tumors to chemotherapeutic agents while minimizing deleterious effects to healthy cells. Therefore it is important to understand the genotypic background of specific tumors to determine which DDR pathways remain and can be targeted for inhibition. Tumor therapies based on the DDR are ideal not only as a means of increasing the effectiveness of current chemotherapies but also as a means to selectively target tumor cells while leaving healthy cells unharmed. Thus, targeting DDR components as a means of increasing effectiveness and discrimination of current chemotherapeutic tumor treatments is currently the focus of many studies and clinical trials.

Keywords:

• DNA-damage response
• tumor
• chemotherapy
• synergism

Overview of DNA-Damage Responses and Signaling

The genome of living organisms is constantly under the attack of exogenous and endogenous agents that damage their DNA potentially leading to harmful events such as single- or double-strand DNA breaks (SSBs or DSBs).1 It is imperative for the organisms to resolve such damage and protect their DNA to maintain genomic stability, allowing for proper replication and proliferation. To counteract the deleterious effects of DNA-damaging agents, organisms have developed DNA-damage responses (DDR) in order to maintain the stability of their genome and ensure proper replication and survival. This is achieved through the sensing of DNA-damaging events and transducing effecter kinases that regulate genes involved in various pathways that enable cells to tolerate their current state of stress.

The Ataxia telangiectasia mutated (ATM) and ATM- and Rad3-related (ATR) sensor kinases are recruited to sites of DNA damage. Upon recruitment they phosphorylate and activate a variety of target proteins that execute the DDR. The ATM sensor kinase has been implicated mainly in the activation of the DDR in response to DSBs in the genome induced by agents such as ionizing radiation.2 ATM is recruited to sites of DSB through its interaction with Nbs1 of the Mre11-Rad50-Nbs1 (MRN) complex involved in DSB repair.3–5 The Ku70/Ku80 DNA-end-binding heterodimer has been implicated in the recruitment of the MRN complex allowing for the full activation of ATM in the presence of DSB.6

The ATR sensor kinase is mainly involved in the activation of the DDR in response to single-strand DNA (ssDNA), which occurs upon stalling of a replication fork.7 A stall may result from an adduct within the DNA caused by agents such as UV radiation.7 The recruitment of ATR is mediated through its interaction with ATR-interacting protein (ATRIP) which in turn interacts with ssDNA-binding replication protein A (RPA).8,9 Complete ATR activation is achieved through interaction with the sensor Rad9Rad1-Hus1 (9-1-1) complex that also localizes to sites of stalled replication.8,9 This is mediated through independent recruitment of both the 9-1-1 complex and the ATRIP-ATR complex by RPA, as well as through interaction of the mediator protein TopBP1 with both the Rad9 subunit of the 9-1-1 complex and ATR.8,10–14

The main substrates of the ATM and ATR sensor kinases are the effecter kinases Chk1 and Chk2 (Fig. 1). Chk1 can be activated by both ATM and ATR through phosphorylation in the presence of a variety of sources of DNA damage such as ionizing radiations, UV light, hydroxyurea and topoisomerase I inhibitors.15,16 Chk1 activation plays an important role in checkpoint regulation and can arrest the cell cycle at S and G₂/M phases through its interaction with and phosphorylation of Cdc25 phosphatase, which is required for proper cell cycle progression.17,18 The phosphorylation signals degradation of Cdc25A during the G₁/S phase transition and Cdc25C during the G₂/M phase transition, effectively arresting the cell cycle though the increase in cyclin-dependent kinase phosphorylation.19,20 Furthermore, the Wee1 kinase, through Chk1 activation, aids in the G₂ checkpoint through the phosphorylation and inhibition of Cdc2, which leads to cell cycle arrest.21 Chk2 is involved in the activation of p53, an important tumor suppressor gene product, and multiple other substrates involved in cell cycle arrest and the induction of apoptosis.22,23 Chk2 may also activate the transcription factor E2F1, which can in turn lead to apoptosis in both a p53-dependent and -independent manner via positive feedback activation of ATM.24–26

Chk1 and Chk2 also share many redundant roles and substrates involved in the activation of the DDR and effector targets, such as the BRCA homologous recombination repair (HRR) pathway.27 The Chk1/Chk2-dependent role in BRCA2-mediated HRR is achieved through its interaction with Rad51, a key component of the strand-invasion step of HRR.28–30 Brit1, another direct substrate of both ATM and ATR, is also a key regulator of the BRCA HRR pathway, both indirectly through Chk1 activation and through direct activation of BRCA1.31

Due to important roles of DDR in maintaining genome stability and ensuring proper replication and proliferation, defects in this pathway are often associated with inherited diseases, particularly tumorigenesis. In order for tumors to develop, cell cycle checkpoint must be disarmed, which is often accompanied by defects in other DNA-repair mechanisms.32 Thus many genes in the DDR pathway mediated by ATM and ATR are found to be defective to allow for tumor formation and progression.

The tight connection between DDR defects and tumors has been the focus of many anti-cancer therapies. Defects that have allowed the formation of the tumor have also left the deficient cells more vulnerable to genotoxic agents. Thus many therapies involve the selective killing of tumor cells due to their increased chemo- and radio-sensitivities in comparison to normal proliferating cells. By inhibiting the remaining repair pathway(s) tumor cells have, they would become extremely sensitive to the specific damaging agents, whereas normal cells contain multiple pathways to deal with genotoxic stresses and can still rely on their redundant repair pathways to tolerate the damage. Furthermore, the increased resistance normal cells display over tumor cells in the presence of specific inhibitors inducing a synthetic sensitivity may eventually lead to tumor therapies that leave normal cells entirely unharmed. These therapies based on DDR defects illustrate the importance of understanding the genotype of a specific tumor in order to best treat and remove cancer from infected organisms.

ATM and ATR Sensor Kinases as a Target for Inhibition and Cancer Therapy

The sensor kinases ATM and ATR have been well-characterized in protecting cells from exogenous sources of DNA damage, through activation of both checkpoint and other DNA-damage responses. As a result of the protective nature of the ATM and ATR pathways in genomic stability and inducing apoptosis, ATM and ATR themselves as well as other members of their respective signaling pathways have been chosen as targets for cancer-related therapies. The positive effect of the inhibition of ATM and ATR was first visualized through the radio- and chemo-sensitizing effects of caffeine.33 It has also been shown that caffeine has the ability to selectively kill p53 mutant cells.34 These studies laid the basis for the studies centered on the inhibition of the ATM and ATR signaling pathways as a means of cancer therapy.

As a result of caffeine's ability to inhibit phosphoinositide-3-kinase (PI3K)-related kinases such as ATM and ATR, and thus enhance DNA-damage sensitivity, a search for more potent inhibitors of the PI3K kinases (PIKK) began. The compounds LY294002 and wortmannin were also found to inhibit PIKK to varying degrees; however, like caffeine, they lacked specificity to ATM and/or ATR.35,36 Using these drugs as a template for more specific inhibitors resulted in the identification of 2-morpholin-4-yl-6-thianthren-1-pyran-4-one (KU-55933) as a highly specific inhibitor for ATM.37 The KU-55933 compound was found to induce sensitivity to IR and other DSB-inducing chemicals such as topoisomerase II inhibitors.37 Recently, KU-55933 inhibition of ATM has been shown to prevent the complete activation of Akt,38 a signal transducer overexpressed in some cancers, which promotes increased cell proliferation and survival.39 It was found that tumor cells treated in combination with KU-55933 and rapamycin, which targets mTOR, a downstream target of Akt, not only induced apoptosis, but also decreased cell proliferation more than either drug alone.38

The effects of PIKK inhibitors were analyzed to ensure that their use would only affect cells being challenged under DNA-damaging conditions. This led to the characterization of ATM inhibition synthetic lethal phenotype, via KU-55933, with genes involved in the Fanconi anemia pathway.40 The ATM and Fanconi anemia pathways appear to play parallel and compensatory roles in maintaining genomic stability and cell viability.40 DNA replication-fork stalling can be resolved through activation of the Fanconi anemia pathway, which can stabilize and coordinate repair at the fork through the promotion of HRR and translesion synthesis.41 Currently, the role of ATM in the tolerance of DNA replication-fork stalling is unclear. ATM may provide time for the resolution of the stalled fork through activation of DDR checkpoints or potentially play a direct role in repair through signaling non-homologous end joining42 or HRR.43 The potential for targeting ATM, as a means of inducing synthetic lethality, is very promising as many Fanconi anemia complementation (FANC) group genes are often mutated in a variety of cancers, such as head, neck, lung, ovarian and cervical cancers.40 The reverse synthetic lethality was demonstrated utilizing ATM-deficient tumor cells treated with EF24, a potent inhibitor for FANCD2.44 This along with other synthetic lethal phenotypes has potential to be an effective means of combating specific tumors without harming normal cells.

Due to their radio- and chemo-sensitization phenotypes, the inhibition of both ATM and ATR has been pursued. The compound schisandrin B (SchB) was identified during a screen for potent inhibitors of ATR.45 SchB, a dibenzocyclooctadiene, is derived from Fructus schisandrae, and has been utilized as a traditional Chinese remedy for the treatment of hepatitis and myocardial disorders.46,47 It has been shown to act as a specific inhibitor for ATR, causing decreased phosphorylation of downstream targets BRCA1 and Chk1, leading to the abrogation of the G₂/M checkpoint following UV treatment.45 The specific nature of ATR inhibition allows for a novel targeting of cancer cells to undergo apoptosis that have lost p53-dependent apoptotic function due to p53 mutations. This scenario is supported by the ability of SchB to enhance doxorubicin-induced apoptosis of tumor cells, while leaving normal cells unaffected.48 Hence, SchB becomes an excellent candidate for cancer therapy, although further studies are required to determine its clinical feasibility.

Through studies showing the tendency of Atr^+/− Mlh1^−/− (a component of mismatch repair, MMR) mice to develop tumors,49 it has been suggested that the ATR pathway may act in concert with MMR to maintain genome stability. This has been attributed to the accumulation of chromosome instability (CIN), as suppression of both ATR and MMR by siRNA resulted in an increased susceptibility to CIN.50 The increase in CIN was also observed when either ATR or Chk1 was depleted in MMR-defective cell lines.50 The clinical significance of this phenotype was addressed by showing that tumor cell lines deficient in both ATR and MMR pathways displayed an increased sensitivity over cells deficient in either of the individual pathways alone to the chemotherapeutic agent 5-fluorouracil (5-FU).50 The mechanism of this synthetic relationship in increasing susceptibility to chemotherapeutic agents is not fully understood, but may be attributed to an increased incidence of DSB when both pathways are compromised.50 The connection between ATR and MMR defects has therapeutic potential in treating nonpolyposis colon carcinomas that are characteristically resistant to 5-FU,51 as well as the wide variety of other common MMR-deficient tumors such as colorectal, gastric, endometrial and bladder cancers.52 Thus, the harnessing of this synthetic phenotype may lead to the ability to lower doses of chemotherapeutic drugs reducing damage to healthy cells.

Recently identified compounds, NU6027 and VE-821, have been shown to exhibit potent and specific inhibition of ATR while not affecting other PIKK. NU6027 has been shown to enhance hydroxyurea and cisplatin cytotoxicity in an ATR-dependent manner as well as display a synthetic lethal phenotype with defects in SSB repair, such as PARP-1 inhibition.53 Although, NU6027 studies reinforce targeting ATR inhibition as a potential cancer therapy it shows insufficient aqueous solubility for in vivo evaluation and is thus unlikely to be clinically pursued.53 The inhibition of ATR by VE-821 has been shown to be lethal to HCT116 cancer cells alone and increase sensitivity to various genotoxic agents and antimetabolites.54 VE-821 also induces a strong conditional synthetic lethal phenotype between the ATR and ATM-p53 tumor suppressor pathway in combination with cisplatin.54 These studies support the pursuit of ATR inhibition and involve synthetic lethal phenotypes as a novel means of improving current tumor therapies.

BEZ-235, a well-characterized inhibitor of PI3K and mTOR, has recently been shown to inhibit additional PIKK, ATM and ATR.55 The sensitizing effects of BEZ-235 on tumor cells displayed in multiple pre-clinical, phase I and II trails may in fact be a result of the inhibition of the DDR kinases, ATM and ATR as their effect cannot be explained through PI3K and mTOR inhibition alone.55–57 Additional studies and clinical trials should be designed accounting for the inhibition of ATM and ATR by BEZ-235 to harness the full potential this compound has as a therapeutic agent.

Chk1 and Chk2 Effecter Kinases as a Target for Inhibition and Cancer Therapy

The effecter kinases Chk1 and Chk2 are important targets for cancer therapy, as they act as direct substrates for the sensor kinases ATM and ATR and mediate the transduction of damage signals to many effecter proteins. The manipulation of this signal transduction is not only a therapeutic target, but also a source of tumorigenesis. Thus the dual nature of manipulating the DNA-damage signal transduction pathways creates difficulties in determining whether a specific effecter is an adequate therapeutic target.

Chk1 inhibition has been widely studied due to its ability to induce p53-independent apoptosis in response to cytotoxic drugs or radiation.58,59 Chk1 inhibitors, such as UCN-01, have the ability to induce apoptosis in p53^−/− cells following IR, thus showing the promise of Chk1 inhibitors as therapeutic agents, since they can selectively target p53^−/− tumors leaving cells with a functional checkpoint unharmed.60 Lidamycin (LDM), an inducer of DSBs currently undergoing phase II clinical trials, can induce p53-independent apoptosis which selectively targets p53^−/− cells due to their lack of a functional Chk1-mediated G₂/M checkpoint.61,62 Another Chk1 inhibitor, AZD7762, has been shown to sensitize multiple tumor xenographs by interfering with HRR and the abrogation of the G₂/M checkpoint.63

UCN-01, the first Chk1 inhibitor tested in human trials has experienced problems during clinical trials due to non-selective inhibition and poor in vivo activity, which raises questions regarding its ability to be utilized as an effective anti-cancer agent.64,65 Additional Chk1 inhibitors that show more promise, AZD7762, SCH900776, LY2603618 and LY2606368, are currently undergoing clinical trials. AZD7762 and SCH900776 are in phase I clinical trials in combination with gemcitabine, a nucleoside analog, and irinotecan, a topoisomerase I inhibitor.66–68 Further studies involving AZD7762 have supported its use in combination with gemcitabine and radiation as a therapy for advanced pancreatic cancers.69 LY2606368 is undergoing a sole phase I trial, unaccompanied by any genotoxic agents, and LY2603618 is also being assessed in multiple phase I and II clinical trails in combination with a variety of chemotherapeutic agents. These clinical studies further support the use of Chk1 inhibition as a means to sensitize tumors to current therapies.

Chk2 inhibition has shown significantly less potential and more controversy as a target for cancer therapy in comparison to Chk1. Many inhibitors of Chk2 show very low specificity and display inhibition of Chk1 as well.70 In addition, inhibition of Chk2 via complete knockdown in HCT116 cells expressing various levels of p53 showed no inhibition of G₂/M checkpoint or LDM-induced apoptosis as well as no increase in sensitivity when combined with a Chk1 knockdown.62 Multiple lines of evidence show that Chk2 inhibition by interference RNA offers no further therapeutic value over the inhibition of Chk1 alone,71 or when combined with antimetabolites.63

However, there are also studies that support the use of Chk2 inhibition as a tumor therapy as it leads to mitotic catastrophe72 and interference RNA or expression of a dominant negative CHK2 leads to enhanced apoptosis following IR or doxorubicin treatment.73 Chk2 has recently been implicated in mitosis as well as genomic stability, as the loss of Chk2 is sufficient to induce CIN,74 providing new phenotypes to potentially lead to novel therapeutic targets. The potential role(s) of Chk2 in genomic stability is further supported in murine studies showing inhibition of both p53 and Chk2 leads to an additive genomic instability following irradiation.75

Chk2 inhibition shows potential in cancer therapies as a means of not only reducing tumor cells but also protecting normal proliferating cells. Chk2-deficient mice exhibit radio-resistance indicating a radio-protective role of Chk2 inhibitors for somatic cells.76 Chk2 inhibition-induced radioprotection has been further validated through analysis of radio-resistance of normal proliferating cells upon treatment with Chk2 inhibitors, 2-arylbenzimidazole77 and VRX0466617.78 However the radio-protective aspect of Chk2 inhibition must be extensively studied before being clinically considered due to the potential of increasing resistance to radiation therapy of tumor cells as well as healthy cells.

Downstream Effecters as Targets of Synthetic Lethality-Based Therapies

In order for tumor cells to develop, DDR defects must accumulate to allow for uncontrolled proliferation. These characteristics may be exploited in a number of ways such as radio- or chemosensitization or through synthetic lethality by simultaneous loss of two separate but nevertheless redundant DNA-repair pathways. Tumor treatments which harness synthetic lethal phenotypes are not only beneficial as they increase the effectiveness of the chemotherapeutic agent but also as a means of protecting healthy cells by selectively sensitizing the defective tumors. A well-characterized synthetic lethal phenotype that is currently utilized as a means of radio-sensitizing cancer cells consists of creating defects in both HRR and base excision repair (BER) pathways.

Poly (ADP-ribose) polymerase 1 (PARP-1) has been investigated as a target for cancer therapy and for inducing synthetic lethality due to its key regulatory role in the BER pathway through interaction with SSBs.79,80 PARP-1 inhibitors have exhibited the ability to sensitize tumor cells to cytotoxic agents such as temozolomide, a topoisomerase I inhibitor, platinum and radiations.81–83 PARP-1 inhibition has been evaluated in phase II clinical trials and tumors showed specific sensitization to temozolomide.83–86 Furthermore, PARP-1 inhibition caused selective killing of tumors deficient in BRCA1 or BRCA2 with minimal effect on normal cells,87,88 which is largely attributed to the defect in HRR.89 However, PARP-1 inhibition is not limited to BRCA1 and BRCA2 mutations.90–93 In an HRR-deficient background, PARP-1 inhibition induces genomic instability and lethality probably through aberrant activation of the error-prone process of non-homologous end joining.94 The requirement of non-homologous end joining and/or HRR for the synthetic lethal interactions with PARP-1 inhibition reinforces the need to determine the genetic background of tumor cell lines to determine optimal treatment. Additional phase I and II clinical trials have confirmed the ability of PARP-1 inhibition to sensitize BRCA1- and BRCA2-deficient tumors to cytotoxic agents as a result of the synthetic lethality due to the loss of both HRR and BER.86,95–97 As a result of the success of PARP-1 inhibition in targeting HRR-deficient tumors, multiple phase I and II clinical trials are being conducted to examine other potential synthetic lethal phenotypes PARP-1 exhibits with DSB repair and ATM pathways.93,98 The PARP-1 inhibitor, BSI-201, also known as iniparib, is currently undergoing phase II clinical trials for treatment of BRCA-deficient tumors as well as phase III clinical trials administering iniparib in conjunction with chemotherapy drugs gemcitabine and carboplatin.99 ABT-888, also known as Veliparib, is currently undergoing a wide variety of phase I and II clinical trials in combination with various genotoxic and radiation therapies.100–112

Cell cycle checkpoint inhibition is a common target for cancer therapy utilized to induce p53-independent apoptosis. Thus, Wee1, a tyrosine kinase involved in the G₂/M checkpoint,113 has been a focus of interest. Wee1 inhibition, just as Chk1 described previously, was able to preferentially sensitize p53^−/− tumor cells in the presence of DNA damage due to the loss of both p53-dependent and independent cell cycle checkpoints.114 The synthetic lethal phenotype was further confirmed through small molecule inhibition of Wee1 in p53-deficient tumor cells.115,116 Wee1 inhibition shows promising results as a therapeutic agent in preclinical and its first clinical trials utilizing MK-1775, a small molecule inhibitor of Wee1.117 Recent studies have confirmed Wee1 inhibition by MK-1775 to selectively sensitize p53-deficient tumor-derived cells to the chemotherapeutic drug gemcitabine.118 Other Wee1 inhibitors have been shown to increase the effect of selective radiation killing in specific tumors.119

Synthetic lethal phenotypes are of great interest in the field of cancer therapy, as they allow for selective killing of tumor cells deficient in various DDR pathways. As a result of this potential benefit, many screens are being undertaken to find novel targets of inhibition to induce synthetic lethality in tumors. Brit1, a positive regulator of homologous recombination repair which is activated through both ATM and ATR, is found to be misregulated in many murine tumors31 and is thus the current subject of highthroughput screening to identify inhibitor compounds that may induce a novel synthetic lethal phenotype to be potentially used as a therapeutic agent.120

Conclusion

Conditional synthetic lethality as a means of treating tumors is becoming an increasingly powerful tool in the fight against cancer. This strategy has two appealing aspects that make it superior to classical treatment of tumors that rely solely on chemo- and radiotherapies. The synthetic lethality induced by targeting the DDR sensitizes tumor cell lines to classical treatments while normal cells are affected to a far lesser extent. This permits selective killing of tumor cells and may potentially even allow for chemo- and radiotherapies that completely abolish negative effects to normal healthy cells.

The current state of knowledge with regard to DDR has allowed for the development of many genetically inspired therapies specifically targeting tumor cells based on their unique phenotypes. Some selected DDR inhibitors undergoing clinical trials are listed in Table 1. However, the understanding of DDR and other differentially regulated pathways within tumor cells is still incomplete. Furthermore, the known components of DDR have yet to be exhaustively tested as targets to allow for the selective killing of tumor cells. Thus, continued research is required to hopefully identify novel components of DDR that may serve as targets for tumor therapies. The search for chemicals and drugs acting on known targets within DDR must also be expanded in the hope of identifying novel selective therapies.

Figures and Tables

Figure 1 ATM/ATR DNA-damage response pathway. Open boxes indicate damage signals; gray boxes indicate the genetic components; filled boxes represent the downstream consequences of the response pathway. Arrows indicate the ability of one component of the pathway to lead to the activation of a subsequent component.

Table 1 Inhibitors of the DDR undergoing clinical trials

Inhibitor	Target	Trial(s)*	Phase*	Condition*
BEZ-235	PI3K mTOR ATM ATR DNA-PKcs	8	Phase I/II	Breast cancer, endometrial cancer, various solid tumors
AZD7762	Chk1 Chk2	1	Phase I	Various solid tumors
LY2603618	Chk1	7	Phase I/II	Non-small cell lung cancer, pancreatic neoplasms, solid tumors
LY2606368	Chk1 Chk2	1	Phase I	Advanced cancer, colorectal cancer, con-small cell lung cancer, ovarian cancer
SCH900776	Chk1	2	Phase I	Leukemia, lymphoma, neoplasms, various solid tumors
UCN-01	Chk1 PDK-1	1	Phase II	Lymphoma
ABT-888 (Veliparib)	PARP-1	40	Phase I/II	Breast cancer, colorectal cancer, leukemia, lymphoma, primary peritoneal cancer, neoplasms, ovarian cancer, solid tumors, skin cancer
AZD2281 (KU-0059436) (Olaparib)	PARP-1	25	Phase I/II	Breast cancer, colorectal cancer, ovarian cancer, neoplasms, triple negative breast cancer, solid tumors
BSI-201 (Iniparib)	PARP-1	17	Phase I/II/III	Advanced solid tumors, advanced uterine carcinosarcoma, breast cancer, glioblastoma, lung cancers, ovarian cancers, primary peritoneal cancer, triple negative breast cancer
AG014699	PARP-1	2	Phase I/II	Advanced solid tumors, breast cancer, ovarian cancer
CEP-8983	PARP-1	1	Phase I/II	Solid tumors
CEP-9722	PARP-1	3	Phase I/II	Solid tumors
MK-4827	PARP-1	4	Phase I	Glioblastoma multiforme, leukemia, melanoma, neoplasms, ovarian cancer, various solid tumors
MK-1775	Wee1	1	Phase II	Epithilial ovarian cancer

* Clinical trial information was obtained from www.clinicaltrials.gov.

Acknowledgments

We wish to thank Michelle Hanna for proofreading the manuscript and other laboratory members for helpful discussion. This work was supported by Natural Sciences and Engineering Research Council of Canada Discovery Grant No. 138338-2009 to W.X., M.B. is a recipient of the Graduate Scholarship from the College of Medicine, University of Saskatchewan.