Background
Despite an incidence rate of 6 % of all childhood cancers [
1], neuroblastoma is responsible for 15 % of all childhood cancer deaths [
2]. Tumours originate from tissues derived from primordial neural crest cells and subsequently can arise anywhere in the sympathetic nervous system [
3]. Fifty percent of all primary tumours manifest in the adrenal medulla [
2]. Patients with high risk disease undergo multimodal treatment, involving intensive chemo- and radiotherapy following surgical resection. However, despite rigorous treatment, there is only a 40 % overall survival rate [
2]. This could possibly be improved with immunotherapy, which has proven an effective treatment for high-risk neuroblastoma patients in remission [
4], but further improvements are necessary to limit adverse cytotoxic effects.
Ninety percent of neuroblastoma tumours express the noradrenaline transporter (NAT) [
5], allowing the active uptake of catecholamine neurotransmitters. Targeted radiotherapy using radioiodinated meta-iodobenzylguanidine (
131I-MIBG) exploits this characteristic of neuroblastoma cells. The radiopharmaceutical
131I-MIBG is a structural analogue of noradrenaline, facilitating its selective accumulation by neuroblastoma tumour cells.
131I-MIBG has demonstrated efficacy as a single agent [
6,
7]. However, the optimal use of
131I-MIBG has yet to be defined [
8], and increasingly it is administered in combination with cytotoxic drug therapy [
9‐
11]. Indeed, a Clinical Oncology Group pilot study (NCT01175356/ANBL09P1) is currently investigating the efficacy of
131I-MIBG in combination with intensive induction chemotherapy in high-risk neuroblastoma patients.
Poly(ADP-ribose) polymerases (PARPs) mediate the post-translational modification of target proteins following hydrolysis of the PARP substrate, nicotinamide adenine dinucleotide (NAD
+) [
12,
13]. The first discovered PARP enzyme, and hence the most comprehensively studied, is PARP-1 [
14,
15]. Upon detection of DNA strand breaks, PARP-1 catalytic activity is increased 500-fold [
13], resulting in the ADP-ribosylation of target proteins including histones, components of DNA repair pathways and PARP-1 auto-modification [
16]. PARP-1 inhibition was shown to exhibit synthetic lethality in cells lacking
BRCA-1 and
BRCA-2 [
17,
18], two important components of homologous recombination repair of DNA double strand breaks [
19]. Inhibition of PARP-1 function in BRCA-deficient cell lines, either by genetic silencing of
PARP-1 [
18] or pharmacologically using a PARP-1 inhibitor [
17], prompted the accumulation of DNA lesions that were not repaired by homologous recombination.
PARP-1 inhibitors have shown great promise when used in combination with treatments that cause substantial DNA damage, including ionising radiation [
20‐
23], DNA alkylating agents [
20,
24] and the topoisomerase-1 poisons topotecan or irinotecan [
25,
26]. Indeed, we have shown previously that the second generation PARP-1 inhibitor PJ34 enhanced the efficacy of 3-way modality treatment involving
131I-MIBG and topotecan [
22]. However, it has been suggested that PJ34 may be toxic to normal cells [
27,
28]. Innovative PARP-1 inhibitors, such as olaparib and rucaparib, have greater specificity, enhanced target affinity, and have now progressed to clinical evaluation [
12,
16,
29]. Rucaparib was the first PARP-1 inhibitor to enter clinical trials [
30] and olaparib was the first PARP-1 inhibitor to gain FDA approval for the treatment of germline
BRCA-deficient ovarian cancer. Both rucaparib and olaparib have shown promise in phase II/III clinical trials, both as monotherapies in
BRCA-mutated breast cancer [
31], ovarian cancer [
32] and prostate cancer [
33], and in combination with cytotoxic drug therapy [
34‐
36].
Therefore, PARP-1 inhibition is a promising approach not only to the targeting of BRCA-deficient cancers which are deficient in DNA repair capacity, but also to the enhancement of the efficacy of DNA damaging chemo- and radiotherapies. Indeed, increased PARP-1 expression has previously been associated with greater neuroblastoma cell genomic instability, higher neuroblastoma stage and poor overall survival [
37], suggesting these tumours will be susceptible to PARP-1 inhibition. PARP-1 inhibitors are also being evaluated clinically for the treatment of children with refractory or recurrent malignancies, such as solid neoplasms, acute lymphoblastic leukaemia, central nervous system neoplasms and neuroectodermal tumours (NCT02116777/ADVL1411). In the present study, we determined the radiosensitising potential of rucaparib and olaparib, two PARP-1 inhibitors currently undergoing phase II/III clinical investigation, in combination with external beam X-radiation or the neuroblastoma-targeting radiopharmaceutical
131I-MIBG. We also examined the effect of combination treatment on cell cycle progression and the persistence of DNA damage.
Methods
Reagents
Rucaparib and olaparib were purchased from Selleckchem (Suffolk, UK) and were reconstituted using phosphate buffered saline (PBS) and dimethyl sulfoxide (DMSO), respectively. Drugs were then diluted in culture medium, maximum DMSO concentration was 0.2 % (v/v). Unless otherwise stated, all other cell culture reagents were purchased from Life Technologies (Paisley, UK) and all chemicals were purchased from Sigma-Aldrich (Poole, UK).
Cell culture
Human neuroblastoma SK-N-BE(2c) cells were purchased from the American Type Culture Collection. SK-N-BE(2c) cells were maintained in high glucose Dulbecco’s Modified Eagle Medium (DMEM) containing 15 % (v/v) foetal calf serum, 2 mM L-glutamine and 1 % (v/v) non-essential amino acids. Human glioblastoma UVW cells [
38] were transfected with a plasmid containing the bovine noradrenaline transporter (NAT) gene [
39]. UVW/NAT cells were maintained in Minimum Essential Medium (MEM) containing 10 % (v/v) foetal calf serum, 2 mM L-glutamine, 1 % (v/v) non-essential amino acids and 1 mg/ml geneticin. Cells were incubated at 37 °C, 5 % CO
2 in a humidified incubator, and were passaged every 3-4 days. Cell lines were cultured in this study for less than 6 months after resuscitation.
Clonogenic assay
Monolayers were cultured at a density of 105 cells in 25 cm2 flasks. Cells in the exponential growth phase were treated with fresh culture medium containing rucaparib or olaparib and were simultaneously irradiated using an Xstrahl RS225 X-Ray irradiator (Xstrahl Limited, Surrey, UK) at a dose rate of 0.93 Gy/min. After 24 h incubation at 37 °C, cells were seeded in 21.5 cm2 petri dishes at a density of 500 (SK-N-BE(2c)) or 250 (UVW/NAT) cells per dish in triplicate. After 8 days (UVW/NAT) or 14 days (SK-N-BE(2c)), colonies containing ≥50 cells were fixed with 50 % (v/v) methanol in PBS and stained with crystal violet. Stained colonies were counted and expressed as a fraction of the untreated, unirradiated control. Radiation survival curves were fitted assuming a linear-quadratic relationship between survival and radiation dose using GraphPad Prism 5.01 (GraphPad Software, San Diego, USA). The data were used to calculate the dose required to sterilise 50 % of clonogens (IC50), as well as the dose-enhancement factor at IC50 (DEF50).
PARP-1 activity assay
Cells were seeded at a density of 1x105 (SK-N-BE(2c)) or 0.5x105 (UVW/NAT) cells on to glass coverslips in 6-well plates. After 48 h, fresh medium was added containing rucaparib or olaparib, before incubating for 1.5 h at 37 °C. PARP-1 activity was stimulated by treatment with 20 mM hydrogen peroxide for 20 min at room temperature in the dark. PBS or DMSO treatment of 0.09 % (v/v) in medium constituted negative controls. Cells were fixed with ice cold methanol/acetone (1:1) on ice for 15 min, before blocking with 2 % (w/v) bovine serum albumin (BSA) in PBS for 30 min at room temperature. Fixed cells were incubated for 1 h at room temperature with a 1:200 dilution of mouse anti-PADPR monoclonal antibody (Abcam, Cambridge, UK; Cat# ab14459) in antibody buffer (10 mM Tris–HCl pH 7.5, 150 mM NaCl, 0.1 % (w/v) BSA in distilled water). Bound anti-PADPR primary antibody was visualised after 1 h incubation at room temperature using goat anti-mouse Alexa Fluor 488-conjugated secondary antibody (Life Technologies, Paisley, UK; Cat# A11029), at a dilution of 1:500 in antibody buffer. Cells were fixed by treatment with 4 % (w/v) paraformaldehyde for 30 min at room temperature in the dark, before mounting on to microscope slides using Vectashield mounting medium containing DAPI nuclear stain (Vector Laboratories, Peterborough, UK). Fluorescence was visualised by means of a Zeiss Axio Observer LSM 780 confocal microscope, using identical laser power and gain settings for all images.
131I-MIBG synthesis and treatment
No-carrier-added (n.c.a.)
131I-MIBG was prepared using a solid phase system wherein the precursor of
131I-MIBG was attached to an insoluble polymer via the tin-aryl bond [
40,
41]. The reaction conditions, HPLC purification procedure, and radiochemical yield were as described previously [
42]. Cells were treated with
131I-MIBG for 2 h, by which time
131I-MIBG uptake was maximal [
43].
Fluorescence Activated Cell Sorting (FACS) analysis
Cells were seeded at a density of 7x105 (SK-N-BE(2c)) or 4x105 (UVW/NAT) cells into 75 cm2 flasks. After 48 h, fresh medium was added containing rucaparib or olaparib and cells were simultaneously irradiated before incubating for 1.5 h at 37 °C. Cells were trypsinised and washed with PBS, before fixing with 70 % (v/v) ethanol in water at -20 °C. Ethanol was removed by washing with PBS. Cells were permeabilised by treatment with 0.05 % (v/v) Triton X-100 in PBS containing a 1:50 dilution of rabbit anti-phospho-Histone H2AX(Ser139)-Alexa Fluor 647-conjugated monoclonal antibody. After 40 min incubation at room temperature, excess antibody was removed by washing with PBST buffer (0.1 % (v/v) Tween 20 in PBS). Finally, cell pellets were resuspended in PBS containing propidium iodide (10 μg/ml) and RNase A (200 μg/ml), before analysis using a BD FACSVerse flow cytometer (BD BioSciences, Oxford, UK). FACS data were quantified using FlowJo 7.6.5 software. For cell cycle analysis, cells were treated separately, and were incubated with propidium iodide and RNase A only as detailed above.
γH2AX immunofluorescent microscopy
Cells were seeded as for PARP-1 activity assay. Fresh medium was added containing rucaparib or olaparib, and cells were simultaneously irradiated, before incubating for 1.5 h at 37 °C. After treatment, cells were fixed with 4 % (w/v) paraformaldehyde for 30 min at room temperature before blocking with 2 % (w/v) BSA (in PBS) for 30 min at room temperature. Fixed cells were then incubated overnight at 4 °C with a 1:50 dilution of rabbit anti-phospho-Histone H2AX(Ser139)-Alexa Fluor 647-conjugated monoclonal antibody (Cell Signalling Technology, supplied by New England Biolabs, Hitchin, UK, Cat# 9720), followed by overnight incubation with a 1:250 dilution of mouse anti-β-tubulin (Life Technologies, Paisley, UK) in antibody buffer (10 mM Tris–HCl pH 7.5, 150 mM NaCl, 0.1 % (w/v) BSA in distilled water). Bound anti-β-tubulin primary antibody was visualised after 1 h incubation at room temperature using goat anti-mouse Alexa Fluor 488-conjugated secondary antibody (Life Technologies, Paisley, UK; Cat# A11029), at a dilution of 1:500 in antibody buffer. Cells were mounted and fluorescence visualised as for PARP-1 activity assay.
Statistical analysis
All statistical analyses were performed using GraphPad Prism version 5.01 (GraphPad Software, California, USA). The number of experimental repeats is provided in figure legends. Data are presented as means ± standard error of the mean (SEM). Statistical significance was determined by either the unpaired Student’s two-tailed t test, or the one-way ANOVA followed by post-hoc testing using Bonferroni correction for multiple comparisons. A probability (p) value < 0.05 was considered statistically significant and < 0.01 highly significant.
Discussion
Patients with high-risk neuroblastoma have an overall survival rate of 40 % despite multi-modal treatment [
2]. Therefore, they present a significant challenge to paediatric oncologists. Single agent treatment with
131I-MIBG is effective in the clinical management of high-risk neuroblastoma. However, recent studies indicate that maximal benefit will be achieved through its administration in combination with radiosensitising drugs [
46‐
49]. In this study, we observed, in pre-clinical models of neuroblastoma, that the third generation PARP-1 inhibitors, rucaparib and olaparib, significantly enhanced the efficacy of ionising radiation, in the form of external beam X-rays or
131I-MIBG. Our results indicate that the mechanism of radiosensitisation entails prolonged DNA damage and accumulation of cells in G
2/M phase of the cell cycle. PARP-1 inhibitors rucaparib and olaparib were comparable with respect to their potentiation of the lethality of X-irradiation or
131I-MIBG. Accordingly, both PARP-1 inhibitors may be considered of benefit to high-risk neuroblastoma patients undergoing targeted radiotherapy.
Since the discovery of the synthetic lethality of PARP-1 inhibition in cells deficient in homologous recombination (HR) [
17,
18], there has been much interest in the therapeutic application of PARP-1 inhibitors. PARP-1 inhibitors have proven an effective monotherapy in
BRCA-mutated breast cancer [
31], ovarian cancer [
32] and prostate cancer [
33]. However, tumours proficient in HR repair may also be susceptible to treatment with PARP-1 inhibitors if administered in combination with cytotoxic drug therapy [
34‐
36] and radiotherapy [
50]. Here, we provide pre-clinical evidence supporting the use of PARP-1 inhibition in combination with external beam X-radiation or
131I-MIBG. The current study focused on rucaparib and olaparib, the first PARP-1 inhibitors to enter clinical trial [
30‐
33,
36] and gain FDA approval, respectively.
Although the radiosensitising capacity of PARP-1 inhibitors has previously been demonstrated in vitro [
22,
51‐
55], this is the first study to show synergism between rucaparib or olaparib with
131I-MIBG. Simultaneous treatment with 10 μM rucaparib or olaparib effectively halved the external beam X-radiation dose or the
131I-MIBG activity concentration required to kill 50 % of clonogens (IC
50) derived from human neuroblastoma SK-N-BE(2c) cells, and human glioma UVW cells genetically engineered to express the noradrenaline transporter (NAT). Rucaparib or olaparib displayed similar radiosensitising potency. Furthermore, combination treatment produced greater than additive cell kill, indicating the potential for enhanced therapeutic benefit.
The present study demonstrated that rucaparib, olaparib and X-irradiation monotherapies significantly increased the proportion of cells in the G
2/M phase of the cell cycle, which would also include a small proportion of cells in late S phase, and has been reported by others [
56]. This is associated with increased radiosensitivity [
57], due to the doubling of the amount of DNA susceptible to radiation trajectory following DNA synthesis in the preceding S phase. This indicates the importance of determining the optimal scheduling of the components of combination treatment to maximise therapeutic benefit. For example, we previously reported that simultaneous delivery of PJ34 (a second generation PARP-1 inhibitor), the topoisomerase inhibitor topotecan and
131I-MIBG maximised the efficacy of this 3-way combination [
22]. Notably, olaparib-induced radiosensitisation was shown to be replication dependent [
52], suggesting that the effects of PARP-1 inhibition would have greater effect in rapidly proliferating tumour cells [
58].
The toxicity of PARP-1 inhibition is hypothesised to involve the accumulation of single strand breaks in irradiated cells, which are subsequently converted to double strand breaks upon collision with the advancing replication fork [
52,
59]. Double strand breaks are quantified following analysis of γH2AX foci [
45]. In response to genotoxic agents such as irradiation, the histone variant protein H2AX becomes phosphorylated at serine residue 139 at the site of double strand breaks [
45]. We demonstrate here that, at cytotoxic concentrations, both PARP-1 inhibitors increased the accumulation of radiation-induced DNA damage and prevented the restitution of this damage 24 h after irradiation. Our results are supportive of others who also show that radiation-induced DNA damage remains unrepaired up to 22 h after irradiation following exposure to olaparib [
55] or rucaparib [
51]. Interestingly, Senra et al. hypothesised that olaparib-induced radiosensitisation was not only the result of impaired DNA repair, but also the result of olaparib-induced vasodilation [
55]. The widening of tumour-associated blood vessels could be of therapeutic benefit, resulting in increased efficiency of drug delivery as well as well as re-oxygenation of hypoxic radioresistant regions of tumours [
60,
61]. Significantly, rucaparib also causes vasodilation [
60,
62].
The anti-tumour effect of PARP-1 inhibitors is not only due to inhibition of PARP-1 catalytic activity. PARP-1 inhibitor toxicity has also been attributed to ‘PARP trapping’, whereby PARP-1 is confined at the site of DNA damage, thus preventing DNA repair, replication and transcription, culminating in cell death [
63]. Indeed, clinical stage PARP-1 inhibitors display a range of capacities to trap PARP-1 at the site of DNA damage, which could influence the selection of PARP-1 inhibitors for clinical use [
64].
PARP-1 inhibitors are increasingly being considered for the treatment of neuroblastoma. Rucaparib has been shown to improve the efficacy of the alkylating agent temozolomide in neuroblastoma pre-clinical models [
24]. The alternative PARP-1 inhibitor niraparib (formerly MK-4827) effectively sensitised a panel of neuroblastoma cells to external beam radiation. The degree of radiosensitisation was shown to be independent of
MYCN amplification [
65].
MYCN amplification occurs in 25 % of all primary neuroblastomas and is used for neuroblastoma risk stratification [
2]. However, to our knowledge, this is the first study to demonstrate the radiosensitising potential of rucaparib and olaparib in combination with
131I-MIBG. Abnormalities in the non-homologous end joining repair pathway, such as increased PARP-1 and DNA Ligase protein expression, have been implicated in neuroblastoma cell survival and pathogenicity [
37]. Indeed, increased PARP-1 expression was shown to correlate with increased genomic instability in neuroblastoma cell lines, including SK-N-BE(2c), and was also associated with higher neuroblastoma stage and poor overall survival [
37], suggesting these tumours will be particularly susceptible to PARP-1 inhibition.
Acknowledgements
The authors wish to thank Dr. Sally Pimlott and Dr. Sue Champion for radiopharmaceutical synthesis; Dr. Mathias Tesson for assistance with combination analysis; Dr. Shafiq Ahmed for assistance with FACS analysis.