Ratio-Dependent Synergism of a Doxorubicin and Olaparib

Dec 28, 2017 - Inhibitors of the poly(ADP-ribose) polymerase (PARP) enzymes are a relatively new class of molecularly targeted small molecule drugs th...
7 downloads 11 Views 2MB Size
Subscriber access provided by the University of Exeter

Article

Ratio-Dependent Synergism of a Doxorubicin and Olaparib Combination in 2D and Spheroid Models of Ovarian Cancer Sina Eetezadi, James C Evans, Yen Ting Shen, Raquel De Souza, Micheline Piquette-Miller, and Christine Allen Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.7b00843 • Publication Date (Web): 28 Dec 2017 Downloaded from http://pubs.acs.org on December 29, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

2

Ratio-Dependent Synergism of a Doxorubicin and Olaparib Combination in 2D and Spheroid Models of Ovarian Cancer

3 4

Sina Eetezadi*, James C. Evans*, Yen-Ting Shen, Raquel De Souza, Micheline Piquette-Miller and Christine Allen#

5 6

Leslie Dan Faculty of Pharmacy, University of Toronto, 144 College Street, Toronto, Ontario M5S 3M2, Canada

7

*Authors made equal contribution

1

8 9 10 11 12 13 14

# Correspondence to: Prof. Christine Allen Leslie Dan Faculty of Pharmacy University of Toronto 144 College Street Toronto, Ontario M5S 3M2 Canada

15 16

E-mail: [email protected]

17 18

The authors declare no potential conflicts of interest.

1 ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Abstract Ovarian cancer is the fourth leading cause of death in women in developed countries. Even though patients with the most lethal form of the disease (HGSOC; high grade serous ovarian cancer) respond well to initial treatment, they often relapse with progressively resistant disease. Inhibitors of the poly(ADP-ribose) polymerase (PARP) enzymes are a relatively new class of molecularly targeted small molecule drugs that show promise in overcoming resistance. The present study explores the combination of a DNA damaging agent, doxorubicin (DOX), with the PARP inhibitor, olaparib (OLP) in order to achieve optimal synergy of both drugs in serous ovarian cancer. This drug combination was evaluated and optimized in 2D monolayers and 3D multicellular tumor spheroids (MCTS) using a genetically and histologically characterized panel of nine OC cell lines with or without BRCA1 or BRCA2 mutations. Combination index (CI) values of DOX and OLP were determined using the Chou and Talalay method. The potency of this drug combination was found to rely heavily on the molar ratios at which the two drugs are combined. In general, MCTS growth inhibition was reflective of the patterns predicted by the CI values obtained in monolayers. Promising combination ratios identified in this study warrant further preclinical and clinical investigation.

36 37 38 39

Keywords: Drug Combination, PARP Inhibitor, Serous Ovarian Cancer, Synthetic Lethality, Doxorubicin, Olaparib, Multicellular Tumor Spheroids, BRCA, Combination Index

40

2 ACS Paragon Plus Environment

Page 2 of 35

Page 3 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Introduction

41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Ovarian cancer (OC) is the fourth leading cause of death in women of developed countries, with dismal survival improvements achieved in the past three decades 1, 2. While the median survival rate for OC as a whole is 40 to 50 % at 10 years post-diagnosis, it falls significantly to 21 % and less than 6 % when the disease is diagnosed at stages III and IV, respectively, as occurs in the majority of cases1. About 90 % of OCs are epithelial ovarian cancer (EOC), and between 60 and 70 % of EOCs are high-grade serous ovarian cancer (HGSOC), the most common and deadliest form of the disease that is widely associated with p53 gene mutations 3-5. As a consequence of the initially asymptomatic or ambiguously symptomatic nature of EOC, most patients present with advanced, metastatic disease at diagnosis, which is currently treated by tumor cytoreductive surgery prior to carboplatin and paclitaxel adjuvant chemotherapy 3, 6, 7. HGSOC patients initially respond well to this regimen (response rate (RR) >80 %), but as many as 75 % relapse, often within 18 months, with progressively resistant disease 4, 8. Once platinum resistance occurs, selection of the most appropriate treatment for further management of the disease remains a challenge 1, 9. Pegylated liposomal doxorubicin (PLD, Doxil® / Caelyx®) is predominant in the treatment of OC recurrence, however it is associated with low RRs of 9 - 16 % 10, 11.

58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84

Inhibitors of the poly(ADP-ribose) polymerase (PARP) enzymes are a class of molecularly targeted small molecule drugs that show promise in the setting of resistant recurrence. The first PARP inhibitor to be clinically assessed is olaparib (OLP, Lynparza®), which has been approved by the FDA for refractory, advanced OC associated with germline breast cancer susceptibility gene (BRCA) mutations12. In August 2017, Lynparza® was additionally approved by the FDA as a maintenance therapy for women with recurrent epithelial ovarian, fallopian tube or primary peritoneal cancer who have achieved complete or partial response to platinium-based chemotherapy, irrespective of their BRCA status 13. Also in 2017, another PARP inhibitor, niraparib (Zejula™), was approved by the FDA for maintenance therapy for adults with recurrent EOC that is responsive (either complete or partial) to platinum based chemotherapy 14. PARP inhibitors capitalize on deficiencies or aberrations in DNA damage repair pathways that are present in about half of all HGSOC.1, 6, 15, 16 . About 51 % of HGSOC display deficiencies in homologous recombination repair (HRR), the high-fidelity DNA double-strand break (DSB) repair pathway, 15 – 20 % of which display germline or somatic mutations in the BRCA1 or BRCA2 genes 15, 17, 18. Through active recruitment of repair proteins, the PARP1 enzyme plays a critical role in base excision repair, which is the predominant mechanism in the repair of DNA single strand breaks (SSBs) 19-21. Upon catalytic inhibition by a PARP inhibitor, PARP1 becomes trapped at SSB sites, which leads to stalling of the replication fork and subsequent formation of SSBs 22. Additionally, PARP inhibitors hinder the role of PARP1 in DSB repair given that PARP1 is reported to aid in the recruitment of HRR proteins 19. On its own, PARP inhibition is relatively non-toxic to cells with intact HRR capabilities. However, synthetic lethality occurs upon PARP inhibition in cells with deficiencies in HRR due to BRCA mutations 20, 21 or the “BRCAness” phenotype, which renders cells more susceptible to DNA damage 15, 19. PARP inhibitors have shown promise clinically in both BRCA-deficient and proficient HGSOCs 23-25. In a phase II trial involving recurrent, platinum-sensitive HGSOC patients, it was found that the inclusion of the PARP inhibitor 3 ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

85 86

olaparib into the paclitaxel-carboplatin standard regimen significantly improved progression-free survival, especially in the patient subpopulation with BRCA mutations 26.

87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107

Due to heightened susceptibility to SSB and compromised repair capabilities, the combination of PARP inhibitors with DNA-damaging agents has the potential to synergistically affect HGSOC cells. The combination of a PARP inhibitor with doxorubicin (DOX) is particularly promising in HGSOC. DOX inhibits DNA polymerase activity and topoisomerase II, and induces free radical formation, causing DSBs and SSBs 11. Further, DOX shows enhanced efficacy in HRR-deficient OC, as BRCA mutation carriers treated with PLD show greater RR and longer overall survival (OS) compared to sporadic cases 10, 11 . Moreover, PARP inhibitors have been shown to enhance DOX antitumor activity in p53deficient cancers 27, 28, which is promising for HGSOC, as 90-100% of cases show p53 mutations 15, 16, 29. Further, PARP inhibitors increase topoisomerase II expression, thereby specifically potentiating the action of DOX 30. Yet, in clinical practice, dose limiting toxicities have been a major hurdle for such combinations, and have resulted in the need to halt a number of clinical trials and/or to reduce drug doses mid-trial 6, 31. Generally, combination therapies have been designed with the assumption that maximal therapeutic effect is achieved when the maximum tolerated dose (MTD) of each drug is employed. However, it has been shown that the effect of drug combinations does not only depend on the nature of each drug’s distinct mechanism of action, but also on the molar ratios at which they have been combined 32, 33. For example, a given drug may induce a strong cell defense response that in turn reduces the therapeutic effect of a second drug when given in combination. Hence, the choice of ratio is important; while certain ratios are synergistic, others can be antagonistic 34.

108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

Based on the potential of combining PARP inhibitors with DNA damaging agents, it was hypothesized that by evaluating a range of specific molar ratios of the PARP inhibitor OLP with the DNA-damaging agent DOX, it would be possible to maximize the therapeutic index of this drug combination by identifying synergistic ratios of the two drugs. To assess this strategy in a setting representative of clinical disease presentation, we sought to employ in vitro models encompassing the heterogeneity of EOC. About 100 OC cell lines are currently available through public cell banks 35; however, the histopathological origin and genomic characteristics are unknown for the majority of these lines, limiting their utility in accurately representing aspects of interest of the disease 35, 36. For this study, multiple cell lines were carefully selected, based on data from three recent large-scale cell line characterization studies, to histologically and molecularly represent an array of serous and HGSOCs, including HRR-deficient variations 35-37. The potential of each cell line to form multicellular tumor spheroids (MCTS) using an established liquid-overlay technique was assessed 38. MCTS are spherical, 500 µm structures comprised of an outer region of proliferative cells surrounding intermediate regions of quiescent cells 39, 40. MCTS are clinically representative 3D models of metastatic cell spheroids that form in advanced disease states as well as microscopic residual disease that remains after cytoreductive surgery in HGSOC 41, 42. In particular, to the best of our knowledge, MCTS of BRCA mutated cell lines have not been reported in the literature thus far. In this study, combination effects of DOX and OLP at ten different molar ratios were evaluated in 2D

4 ACS Paragon Plus Environment

Page 4 of 35

Page 5 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

128 129 130

Molecular Pharmaceutics

and 3D cell models of the selected ovarian cancer cell lines using the well-established Chou and Talalay method to evaluate treatment synergy 43. In order to ascertain the mechanism of action of this drug combination, the γH2AX assay was used to quantify DNA DSBs.

5 ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

131

Materials and Methods

132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147

Materials DOX was purchased from Polymed (Houston, TX, USA). OLP was obtained from Tongchuang Pharma, Suzhou Co. Ltd (China). Acetic acid, P-nitrophenyl phosphate, TritonTM-X-100, G-418 disulfate salt, sodium pyruvate and sodium acetate were purchased from Sigma-Aldrich (Oakville, ON, CA). MCDB 105 and DMEM cell media were obtained from Sigma-Aldrich, while all other cell media, sterile phosphate buffered saline (PBS, pH = 7.4), CellMaskTM Green plasma membrane stain and fetal bovine serum (FBS) were obtained from Life Technologies (Burlington, ON, CA). MEGM™ Mammary Epithelial Cell Growth Medium Kit was purchased from Lonza (Mississauga, ON, CA). UWB1.289, UWB1.289+BRCA1, OV-90 and SKOV3 cell lines were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). PEO1, PEO4 and COV362, originally from the European Collection of Cell Cultures (ECACC, Public Health England; Salisbury, UK), were purchased through Sigma-Aldrich. HEYA8 was obtained from the M. D. Anderson Cancer Center (Houston, TX, USA). OVCAR8 was obtained from the Biological Testing Branch of the National Cancer Institute (NCI; Frederick, MD, USA).

148 149 150 151 152 153 154 155 156 157 158 159 160 161

Cell Culture All cell lines were cultured as monolayers in the media recommended by the supplier, and maintained in a dedicated cell culture incubator at 37 °C, 5 % CO2 atmosphere and 90 % relative humidity. UWB1.289 and UWB1.289+BRCA1 were maintained in a 1:1 mixture of RPMI-1640 and MEGMTM including the five components of the MEGM™ Mammary Epithelial Cell Growth Medium Kit and 3 % FBS. UWB1.289+BRCA1 growth media was additionally supplemented with 200 µl/ml G-418 to maintain BRCA1 expression. HEYA8 and OVCAR8 were grown in RPMI-1640 media with 10 % FBS. PEO1 and PEO4 were maintained in RPMI-1640 media with 10 % FBS and additionally supplemented with 2 mM sodium pyruvate. COV362 was cultured in DMEM and SKOV3 in McCoy's 5a medium, both supplemented with 10 % FBS, and OV-90 was grown in a 1:1 mix of MCDB 105 and Media 199 supplemented with 15 % FBS. Penicillin/streptomycin was added to all media at a concentration of 100 units/mL penicillin and 100 µg/mL streptomycin. Cell passage number was maintained below 25.

162 163 164 165 166 167 168 169 170 171 172

Cell Line Authentication Short tandem repeat (STR) analysis was performed by The Centre for Applied Genomics at The Hospital for Sick Children (Toronto, ON, CA) to verify the authenticity of cell lines employed here. To this end, the GenePrint®10 System (Promega Corporation; Madison, WI, USA) was used, following manufacturer’s instructions. Briefly, the kit employs an STR multiplex assay that amplifies nine loci (the ANSI Standard (ASN-0002) recommends eight) and the Amelogenin gender-determining marker in a single PCR amplification. To prepare each reaction, 10 ng of genomic DNA was incubated with a master mix (5 µL GenePrint® 10 5X Master Mix, 5 µL GenePrint® 10 5X Primer Pair Mix, water for a 25 µL total volume). The cycling conditions were 96 °C for 1 min; 30 cycles of: 94 °C for 10 s, 59 °C for 1 min, 72 °C for 30 s; 1 cycle of 60 °C for 10 min and 4 °C maintenance.

6 ACS Paragon Plus Environment

Page 6 of 35

Page 7 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

173 174 175 176 177

Samples were run on an AB 3130 Genetic Analyzer using POP7, Promega’s internal lane standard 600, dye set F and analyzed in GeneMapper v3.7. Since allelic nomenclature for the 10 loci is standardized worldwide, the obtained STR profiles were compared to those made available by each cell line’s corresponding repository or previous publications (S1 Table).

178 179 180 181 182 183 184 185 186 187 188 189 190 191

BRCA1 and BRCA2 Mutation Sequencing Regions pertaining to BRCA1 and BRCA2 gene variants of interest were amplified for Sanger sequencing by The Centre for Applied Genomics at The Hospital for Sick Children (Toronto, ON, CA) to confirm BRCA mutation status of relevant cell lines. Primer sequences are outlined in S2 Table. Briefly, each 25 µL PCR reaction was prepared with 2mM dNTPs (2.5µL), 10X PCR buffer II (2.5µL; Life Technologies), 25mM MgCl2 (1.3 µL), AmpliTaq DNA polymerase (0.25µL; Life Technologies), 5M betaine (5 µL), water (10.45 µL), 10uM primers (1 µL each) and genomic DNA (1 µL; normalized to 50-100 ng/µL). Samples were amplified by PCR using the following cycling conditions: 30 cycles of 94 ºC for 10 s, 60 ºC for 30 s and 72 ºC for 1 min. Agarose gel was used to visualize the PCR products, which were purified using the AxyPrep PCR Clean-up Kit (Axygen Biosciences). A total of 50 ng of purified product was Sanger sequenced from the forward primer using BigDye3.1 chemistry on a ABI3730XL capillary DNA analyzer (Life Technologies).

192 193 194 195 196 197 198 199

Doubling Time For evaluation of cell proliferation rates, subconfluent cells were harvested, seeded onto 6well plates at a seeding density of 100,000 cells per well and allowed to adhere overnight. Over a 6-day period, cells were detached at selected timepoints with TrypLETM Express cell dissociation solution and counted using a Countess II FL automated cell counter (Life Technologies). Cell doubling time was determined by nonlinear regression based on an exponential growth equation that was fit to the cell counts at each timepoint using Graphpad® Prism software.

200 201 202 203 204 205 206 207 208 209 210 211 212 213 214

Monolayer IC50 Evaluation Monolayer cell cytotoxicity for DOX, OLP and all drug combinations was determined using the acid phosphatase (APH) assay based on a method previously published by our group 38. Briefly, subconfluent cells were harvested, seeded onto 96-well plates and allowed to adhere overnight prior to incubation for 72 h using ten 1:4 serial dilutions of single drug or drug combinations (n = 6 wells per dilution). Seeding density was determined separately for each cell line so that the control was maintained subconfluent through to the end of the experiment (5000 cells/well for OV-90 and COV362; 4000 cells/well for PEO1 and PEO4; 3000 cells/well for UWB1.289, UWB1.289+BRCA1 and SKOV3; 1000 cells/well for OVCAR8 and HEYA8 cell lines). Following treatment, cells were washed with PBS prior to assessment of viability using the APH assay. In brief, cells were incubated with 100 µL of freshly prepared reaction buffer (sodium acetate buffer at pH 5.5 containing 1 % Triton-X supplemented with 2 mg / ml p-nitrophenyl phosphate) for 2 h at 37 °C. Cell viability was determined by measuring the UV absorbance at 485 nm using an automated 96-well plate reader (Synergy 2 from BioTek; Winooski, VT, USA)

7 ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

215 216 217

following addition of 10 µL 1 M sodium hydroxide to each well. Average absorbance (A) for each drug concentration of three independent experiments performed on different days was expressed relative to controls as follows:  % =

 –  

 –  

218 219 220

The concentration that kills 50 % (IC50, or fraction affected (Fa) = 0.5) and 75 % (Fa = 0.75) of cells was determined by fitting the data to the Hill equation using Graphpad® Prism software.

221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236

Determination of Combination Index (CI) Values CI values were determined according to a widely used method established by Chou and Talalay 44, 45. Briefly, to determine each CI value, the following monolayer cytotoxicity studies were conducted: 1. DOX as a single drug; 2. OLP as a single drug; and 3. DOXOLP combinations at a series of specific molar ratios. IC50 values that resulted from the drugs when used in combination were between 50 µM to 5 nM for both drugs, depending on the cell line and drug ratio used. For each experiment, starting concentrations were chosen such that the IC50 value was the intermediate concentration employed in the ten serial dilutions chosen, as outlined above. For all experiments, Fa = 0.5 and Fa = 0.75 molar drug concentrations were determined, as outlined above. CI values were then calculated for each combination and for both effect levels based on the formula below for mutually non-exclusive drugs as established by Chou and Talalay, where DSD is the concentration of DOX that kills 50% of cells (for Fa = 0.5) or 75% of cells (for Fa = 0.75), DSO is the concentration of OLP at Fa = 0.5 or Fa = 0.75, DCD is the concentration of DOX in combination with OLP at Fa = 0.5 or Fa= 0.75, and DCO is the concentration of OLP in combination with DOX at Fa = 0.5 or Fa = 0.75:  =

 #  # + + ! !# ! !#

237 238 239 240 241

Based on this method, CI values are indicative of strong synergism (< 0.7), synergism (0.7 – 0.9), additive effect (0.9 – 1.1), antagonism (1.1 – 3.3) or strong antagonism (> 3.3) 44. Microsoft Excel was used to generate a tricolor system based on these values, where strong synergism is represented by green, additive effect by yellow and strong antagonism by red. The software interpolates the color of each value in between these constraints accordingly.

242 243 244 245 246 247 248 249 250

γ-H2AX Determination In order to quantify the level of DNA DSBs, 1 x 10⁵ UWB1.289 and UWB1.289+BRCA1 cells were seeded on 18 x 18 mm glass coverslips in six well plates. Cells were treated for 72 hours with either; a) DOX, b) OLP c) a synergistic ratio of DOX:OLP (i.e. 50:1) or an d) antagonistic ratio of DOX:OLP (1:2). As DOX is the more potent drug in the combination (Figure 1), the respective IC50 for DOX was used as the DOX concentration in the DOX, synergistic and antagonistic ratios. For the combination groups, the OLP concentration was determined relative to the DOX. Finally, the concentration of OLP in the OLP group was determined from the highest 8 ACS Paragon Plus Environment

Page 8 of 35

Page 9 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

251 252 253 254 255 256 257 258 259 260 261

concentration of OLP used in the combination groups. Cells were incubated with 10 µM of 5-ethynyl-2’-deoxyuridine (EdU) for 30 mins prior to fixation in order to exclude those cells in the S phase that harbor endogenous DSBs. Cells were fixed at room temperature for 20 minutes in 4 % paraformaldehyde/0.2 % Triton-X-100 (pH 8.2). Click-iTt EdU Alexa Fluort 647 kit was used to stain S phase cells that had incorporated EdU. γ-H2AX foci were stained using Anti-phospho Histone H2AX clone JBW301 (1:100) (Millipore, Billerica, MA) overnight at 4°C. 4’6-diamidino-2-phenylindole (DAPI; Invitrogen) was used for nuclear staining. Images were acquired using the Zeiss LSM 710 confocal/twophoton microscope using a 633/1.4 oil immersion objective lens. For each replicate, at least 20 nuclei were acquired using ZEN black software (Oberkochen, Germany). Samples were analysed using ImageJ software (NIH, Bethesda, MD).

262 263 264 265 266 267 268 269 270 271 272 273 274 275 276

MCTS Growth Studies MCTS establishment was attempted for all cell lines using a method previously reported by our group 46. Subconfluent cells were seeded onto non-adherent 96-well round-bottomed Sumilon PrimeSurfaceTM spheroid plates (MS-9096U; Sumitomo Bakelite, Tokyo, Japan) containing the recommended complete growth media for each cell line and incubated for 10 days at 37°C, such that each well contained a single MCTS. Images were taken using a light microscope with a 10x objective lens (VistaVisionTM; VWR, Mississauga, ON, CA) connected to a digital camera (DV-2B; VWR). The volume of each MCTS was determined according to a published method 46. Briefly, the 2D cross-sectional area was measured using an automated image analysis macro developed for use with the ImageJ software package (Version 1.48V) and volumes were calculated assuming spherical shape. Growth curves were fit to the Gompertz growth equation of tumor growth 46. For morphology studies, MCTS were incubated for 1 h with 1X CellMaskTM Green plasma membrane stain (Life Technologies) in PBS and imaged using a Zeiss LSM700 confocal microscope (Carl Zeiss AG; Oberkochen, Germany) with a FITC filter (Ex 522 nm, Em 535 nm).

277 278 279 280 281 282 283 284 285 286 287

MCTS Growth Inhibition Studies MCTS of OVCAR8, OV-90 and HEYA8 cell lines were grown as described above. Initial cell seeding density was chosen such that MCTS reached a diameter of about 500 µm after 7 days. Prior to combination studies, MCTS were treated with DOX (2 µM, 1 µM and 0.2 µM) for 72 h to determine a drug concentration sufficient to inhibit MCTS growth while still allowing MCTS to remain structurally intact, such that assessment of different molar combinations of DOX with OLP would be possible. MCTS were incubated for 72 h with single drug or drug combinations at selected molar ratios (n = 6 per treatment group for all studies). Following the 72 h treatment period, MCTS were washed with media. Every other day thereafter, 50 % of media was replaced prior to imaging for MCTS volume determination, as described above.

288 289 290 291 292

Statistical Analysis Results are reported as mean of at least 3 independently conducted experiments ± standard deviation (SD). Differences between treatments were compared using one-way ANOVA’s F-test followed by post-hoc analysis using Bonferroni correction, with significance assigned at p ≤ 0.05.

9 ACS Paragon Plus Environment

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

293

Results

294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328

Properties of the HGSOC Cell Panel Based on three large-scale cell line characterization studies 35, 36, 47, nine ovarian cancer cell lines were selected to represent EOC, particularly the HGSOC subtype (S3 Table). Additional criteria were BRCA1 or BRCA2 mutation status and availability of a corresponding BRCA-proficient cell line. BRCA was chosen as the most relevant genomic mutation because up until August 2017, OLP was only approved for treatment of patients with BRCA mutations. Although not considered to be of serous histology, SKOV3 was employed as an additional, well-characterized, BRCA-proficient cell line. Cell line authentication was performed on the basis of STR analysis (S1 Table). UWB1.289 is derived from a tumor of papillary serous histology, the most common form of OC 48. This cell line’s BRCA1 wild-type allele is absent as the gene is mutated within exon 11. UWB1.289 was transfected with a BRCA1 construct to restore BRCA1 function, creating the UWB1.289+BRCA1 cell line 48. PEO1 and PEO4 were derived from the ascites of the same patient before and after development of resistance to platinum-based chemotherapy, respectively 49. Subsequent analysis revealed that PEO1 had a BRCA2 loss-of-function mutation, while PEO4 showed a secondary mutation that caused functional restoration of BRCA2 protein, causing PEO4 to be more resistant to platinum or PARP inhibitors than PEO1 50. COV362 is another cell line with BRCA1 loss-of-function mutation which was originally described as being derived from a tumor of endometrioid subtype 51, with later analysis qualifying it as HGSOC 35, 36. OVCAR8 was derived from a platinum-refractory patient who showed disease progression even after receiving high doses of platinum 52. BRCA1 promoter methylation was later identified in this cell line 53. This epigenetic modification results in lack of BRCA1 mRNA expression, and has been clinically reported in 8.1 % 54 to 13.3 % 55, 56 of EOCs and in 14.8 % of HGSOC cases 57. OV-90, derived from previously untreated ovarian malignant ascites, forms tumors of serous histology and expresses functional BRCA1 and BRCA2 58. HEYA8 was derived from a mouse xenograft tumor after three passages of the patient-derived cells in immunocompromised mice 59, 60. The original xenografted tumor was described under the name HX-62 as a serous ovarian cystadenocarcinoma derived from peritoneal metastasis 53, 61. SKOV3 was established in 1973 and is one of the most widely used ovarian cancer cell lines based on publication records 36, 62. SKOV3 has been characterized as being of either serous or clear cell histotype 35 . While HEYA8 expresses wild-type p5363, all other cell lines employed here have been reported to have a p53 mutation, a common trait of HGSOC 35, 36, 64. BRCA status was confirmed by sequencing regions of BRCA1 and/or BRCA2 gene variants of interest for UWB1.289, UWB1.289+BRCA1, PEO1, PEO4 and COV362 cells (S4 Table).

329 330 331 332 333 334 335

The doubling time (T2) of all cell lines was evaluated, since previous studies on a large panel of ovarian cancer cell lines have shown that doubling time significantly correlates with in vitro sensitivity to DOX and other drugs, such as taxanes and platinum derivatives 35 . The cell line with the highest proliferation rate was HEYA8 (16 h), while COV362 had the lowest rate (99 h). T2 appeared to be unrelated to BRCA mutation status. For instance, the BRCA1-restored UWB1.289+BRCA1 cell line displayed a T2 that was 30 % shorter than that of UWB1.289. Conversely, T2 for the BRCA2-restored cell line PEO4 was 31 %

10 ACS Paragon Plus Environment

Page 10 of 35

Page 11 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

336 337

faster than for PEO1, which are cells derived from the same patient at different stages of the disease 49.

338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355

Cell Monolayer Viability Studies of DOX and OLP treatment All cell lines were treated with DOX or OLP over 72 h for determination of IC50 values. In general, the DNA damaging agent DOX was observed to be approximately three orders of magnitude more cytotoxic than OLP. Consideration of the IC50 values with respect to BRCA status revealed that BRCA1 or 2 deficiencies are not sufficient to predict sensitivity to OLP, or DOX (Fig. 1). For this analysis, cell lines were categorized as “BRCA proficient” and “BRCA deficient” according to S4 Table, and for the purpose of this analysis, OVCAR8 (light orange) was classified as “BRCA deficient”. As shown in Fig. 1, cell lines that are BRCA deficient showed a very narrow distribution of IC50 values for both drugs, whereas BRCA-proficient cells showed a more broad distribution in IC50 values, especially for OLP. It may be that some of the BRCA-proficient cell lines harbor other deficiencies in HRR that make them more sensitive to DOX or OLP. Additionally, for HEYA8, the relatively low doubling time of 16 h may be contributing to its very high sensitivity to DOX as well as OLP. When the IC50 values for each group were considered together, comparison between the two groups found no statistically significant difference. However, the BRCA proficient PEO4 (purple) and UWB1.289+BRCA1 (dark blue) were indeed two to three times more resistant to OLP than their BRCA-deficient counterparts PEO1 and UWB1.289, respectively.

356 357 358 359 360 361

Fig. 1: Single drug IC50 values. IC50 values for cell lines and and HEYA8 cells), grouped as BRCA deficient or proficient, DOX (left) or OLP (right). When the IC50 values for BRCA methylated) cell lines are considered together and compared proficient cell lines, no significant difference is identified (n=3).

362 363 364 365 366

DOX-OLP Combination Studies Following the assessment of single drug cytotoxicity, DOX and OLP combination treatments at 10 different molar ratios were evaluated over a 72 h period. For each ratio, three independent experiments were performed on different days, and the molar drug concentrations required to kill 50 % (Fa = 0.5) and 75 % (Fa = 0.75) of cells were 11 ACS Paragon Plus Environment

MCTS (OVCAR8, OV90 following treatment with deficient (i.e. mutated or to values for the BRCA

Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

367 368 369 370 371 372 373 374 375 376 377 378 379 380 381

Page 12 of 35

determined. The median-effect algorithm based on the widely used method established by Chou and Talalay was employed to calculate the CI for each ratio at both effect levels (i.e. Fa = 0.5 and Fa = 0.75) as outlined in the methods section 65. The CI equation was used to generate CI values, which categorize the effect of the drug-drug combination at specific ratios as synergistic, additive or antagonistic, as described in the methods section. A synergistic effect implies that the two drugs are more effective together than what would be expected from adding the effects of each drug when used separately 43, 44. In the case of OLP+DOX combinations, it is important to evaluate combination effects at different effect levels (i.e. Fa) since dose-response curves for both DOX and OLP are not linear. Consequently, the cytotoxicity of these drugs changes incrementally with dose, and the relative cytotoxicity contribution of each drug will differ with varying Fa 43, 66. For cancer chemotherapy, higher effect levels of Fa = 0.8, for instance, are commonly used and more relevant for treatment, where complete cancer cell eradication is the goal 44, 67. The CI results are presented in a “heat map” for each ratio of DOX:OLP at both Fa levels (Fig. 2, A, B). 382

383 384 385 386 387 388

Fig. 2: Combination Index values for DOX and OLP at different molar ratios. Summary of CI values for the combinations of DOX and OLP following 72h incubation in nine ovarian cancer cells lines at the indicated molar ratios required for Fa = 0.5 (A) and Fa = 0.75 (B). Each CI value was calculated and a heat map was generated as outlined in Materials and Methods on the basis of three independent IC50 experiments performed on different days for DOX, OLP and each combination ratio (n=3).

389

12 ACS Paragon Plus Environment

Page 13 of 35 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

390 391 392 393 394 395 396 397 398 399 400 401 402 403 404

Overall, the combination of OLP and DOX resulted in synergistic effects when the molar concentrations of each drug differed from each other to a greater extent, such as DOX:OLP at 1:100, 1:50, 50:1 and 100:1. Taking this into consideration, we can extrapolate from the data, that there needs to be a minimum difference in the molar ratios of these 2 drugs in order to observe synergism across the cell lines (based on the average CI values). As it is not feasible to determine the CI values for every possible permutation of molar ratios of DOX:OLP, we have identified from the data presented here, that the magnitude of the molar ratio should be greater than 1:10 or 10:1. We have observed that once these ratios have been achieved, there was very little difference, with respect to CI values between 1:100 and 1:50 (CI values of 0.81 vs. 0.83 respectively) and 50:1 and 100:1 (0.78 vs. 0.74 respectively). More equimolar ratios of 1:2, 1:5 and 1:10 as well as the inverse resulted in an additive effect. For Fa = 0.75, a trend towards more synergy was generally observed in comparison to Fa= 0.5, with the exception of the 50:1 and 100:1 ratios (Fig. 2, B). Notably, OLP concentrations required when in combination with DOX were 10 to 1000-fold below the IC50 values of OLP as a single agent.

405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422

Neither BRCA status nor sensitivity to DOX or OLP single-drug treatment appeared to be predictive of DOX-OLP synergism. For example, although the BRCA1-deficient cell line UWB1.289 was more sensitive to single drug treatments by a factor of three to four in comparison to UWB1.289+BRCA1, the CI values for UWB1.289 were lower than for UWB1.289+BRCA1 at certain ratios (1:100, 1:50, 10:1, 50:1 and 100:1) but not others (1:10, 1:5, 1:2, 2:1, 5:1). CI values for the BRCA2-deficient cell line PEO1 were generally lower than for its BRCA2-proficient counterpart PEO4, except for the 1:50 and 100:1 ratios. Furthermore, while HEYA8 and PEO1 are both quite sensitive to DOX, most DOX:OLP ratios resulted in synergism in the former cell line and an additive effect or antagonism in the latter. Conversely, DOX-OLP combinations for UWB1.289+BRCA1, the most DOX-resistant cell line, were found to be very effective, especially at Fa = 0.75. As was the case with DOX, single-drug OLP sensitivity did not correlate with combination outcomes. While PEO4 showed the strongest resistance to single-drug OLP as well as overall antagonistic effects in response to DOX-OLP combinations, OVCAR8 was highly sensitive to single-drug OLP but the DOX-OLP combination did not yield particularly synergistic effects. In summary, the data suggests that synergism of the DOX:OLP combination is drug-ratio-dependent and more pronounced with higher Fa, but cannot be predicted from BRCA status or sensitivity to single-drug treatments.

423 424 425 426 427 428 429 430 431 432

In order to ascertain the mechanism of action of the DOX and OLP combination, the level of DSBs induced by the combination was assessed in two cell lines with varying BRCA status (i.e. UWB1.289 and UWB1.289+BRCA) using the γH2AX assay (Fig. 3). For the OLP group, the levels of γH2AX foci were not significantly different from the levels observed in the untreated controls for both of the cell lines (p>0.05). According to the CI values (Fig. 2), the 50:1 ratio was synergistic in UWB1.289 and UWB1.289+BRCA cells. This data was supported by the γH2AX assay, where 50:1 resulted in significantly more (p0.05) between the 50:1 and 1:2 DOX:OLP ratios. Again, this data is supported by the CI values, where both displayed strong synergy in this cell line (CI values of 0.36 and 0.70 respectively). For the UWB cell line, the 50:1 ratio had a significantly higher amount of DSBs relative to 1:2, which also displayed significantly lower levels of γH2AX foci relative to DOX alone (p