Search

High grade serous ovarian cancer (HGSOC) comprises 75-80% of all ovarian cancer cases and is characterised by p53 mutation and genetic heterogeneity. There are two major molecular categories: homologous recombination repair (HR) pathway proficient and deficient. Most cases are treated with a combination of surgery and chemotherapy. Patients with HR deficient cancers respond well initially, but the majority experience relapse after first line chemotherapy due to drug resistance. As p53 is mutated in virtually all HGSOC cases, these cancers rely heavily on the G2/M checkpoint for DNA damage repair and survival. The G2/M checkpoint is controlled by the phosphorylation status of Cdk1. Wee1 kinase responds to DNA damage by phosphorylating Cdk1 and causing G2 arrest, giving time for DNA repair before entry into mitosis. We and others have hypothesised that combining a DNA damaging agent with Wee1 kinase inhibition would be particularly effective in HGSOC. AZD1775 is a small molecule inhibitor of Wee1 kinase that forces cells into mitosis without the repair of damaged DNA. The overall aim of the thesis was to characterise mechanisms of resistance to AZD1775 in cultured HGSOC cells and identify ways to overcome resistance that could be ready to deploy as the inhibitor moved into the clinic. It was hypothesised that resistance to AZD1775 could occur in a number of different ways through changes in key DNA damage response and cell cycle control pathways. To address the additional hypothesis that mechanisms of resistance to AZD1775 could be different in HR-deficient and proficient HGSOC, cell lines from each category were used to generate resistant clones for mechanistic studies. Given the increased use of the PARP inhibitor olaparib as a therapy for HGSOC, olaparib resistant clones were also isolated from the same HR-deficient and proficient HGSOC cell lines with the supplementary aim of investigating the effectiveness of AZD1775 therapy in olaparib resistant HGSOC. I first investigated the effect of AZD1775 as a single agent and in combination with the DNA damaging agent cisplatin and olaparib in a panel of twelve HGSOC cell lines in cell growth and viability assays. AZD1775 enhanced the sensitivity to cisplatin and olaparib in three cell lines from the panel, two of which were HR pathway deficient. To study mechanisms of resistance to Wee1 inhibition, two cell lines, ES-2 and OVCAR8, were chosen representing HR proficient and deficient HGSOC to isolate clones resistant to AZD1775, AZD1775 in combination with cisplatin and olaparib alone. Compared to the parental lines, AZD1775 resistant clones retained normal cell cycle profiles in the presence of AZD1775. No evidence was found for mutation in the coding exons of the Wee1 kinase gene in resistant clones. Western blot analysis revealed reduced Cdk1 expression in many of the resistant clones, while others showed increased Wee1 kinase expression. RNA sequencing based comparison of AZD1775 resistant clones with the parental ES-2 cell line showed an upregulation of TGFβ signalling that fed into the cell cycle control pathway via multiple cell cycle inhibitors. An alternative potential mechanism of resistance in some ES-2 and OVCAR8 clones involved upregulation of the TNF and NF-κB signalling pathway leading to evasion of apoptosis. An attempt to overcome AZD1775 resistance by blocking TGFβ or TNFNF- κB signalling pathways was successful for one resistant ES-2 clone using a TGFβR1 inhibitor. Western blotting of olaparib resistant clones isolated from ES-2 and OVCAR8 showed known changes responsible for olaparib resistance, including reduced expression of 53BP1 and down regulation of Poly(ADP-ribose)glycohydrolase. All but one of the olaparib resistant clones tested remained sensitive to AZD1775. In conclusion, mechanisms of resistance to two important inhibitors for the improved treatment of HGSOC have been identified and ways to overcome resistance in the clinic have been suggested.

The cellular response to DNA double-strand breaks (DSB) is principally the result of two competing repair pathways. The Tumour Protein P53 Binding Protein 1 (TP53BP1) gene product plays a central role in DSB repair pathway choice, promoting non-homologous end-joining (NHEJ) and counteracting homologous recombination (HR). As with all DNA damage response (DDR) proteins, cell-cycle regulation of 53BP1 is critical to its function. The inappropriate activity of 53BP1 in mitosis, during which most of the DDR is inactivated to preserve genome stability, has been shown to have adverse effects on cell division that are characteristic of cancers. We explored the consequences of replication stress, associated with oncogene activity, and the impact of aberrantly active 53BP1 in mitosis. This work reveals interference with normal mitotic mechanisms for detecting and resolving stalled replication intermediates; most evidently manifest as reduced mitotic DNA synthesis (MiDAS), and increased accumulation of acentric micronuclei and G1 nuclear bodies. We propose that the functional sequestration of TOPBP1 may be responsible for these chromosomal mis-segregation defects. The mechanism by which 53BP1 (together with RIF1 and MAD2L2) promote NHEJ remains elusive. Genome instability in HR-defective cancers, principally those harbouring a mutation in BRCA1, can be rescued by the concomitant loss of either 53BP1, RIF1 or MAD2L2. This phenomenon has relevance for the development of PARP (Poly(ADP-ribose) polymerase) inhibitor resistance, now being used in BRCA1-deficient breast, ovarian and prostate cancers based on the principle of 'synthetic lethality'. As part of a search for other resistance mechanisms, we identified the shieldin complex (SHLD1/C20orf196 and SHLD2/FAM35A) that when lost suppressed PARP inhibitor sensitivity in BRCA1-deficient cancer cells. We demonstrate that SHLD1/2 represent the terminal end-effectors of the 53BP1-RIF1-MAD2L2 axis, supporting NHEJ activity. We identified three OB-fold domains in SHLD2/FAM35A, that bind single-stranded DNA and are necessary for NHEJ. Loss of SHLD1 or SHLD2 (SHLD1/2) caused a hyper-resection phenotype, and which, in the absence of BRCA1 was associated with rescuing RAD51 loading. Downregulation in SHLD1/2 expression was also associated with intrinsic and acquired resistance across a panel of BRCA1-deficient patient-derived xenograft models. Promisingly, this mechanism may impart an acquired vulnerability to cisplatin or radiotherapy. Thus, we have characterised a critical component of the 53BP1 axis which is deterministic in repair pathway choice and carries implications for the treatment of cancers.

Rak jajnika charakteryzuje się najwyższą śmiertelnością. Stosowane leki zwiększają przeżywalność pacjentek w IV, najwyższym stadium, jednakże 5-letni wskaźnik przeżycia ciągle pozostaje na bardzo niskim poziomie. Inhibitor polimerazy poli (ADP-rybozy) prowadzi do akumulacji pęknięć jednoniciowych. Sprzyja to powstawaniu pęknięć dwuniciowych, a w przypadku nowotworów z mutacją BRCA1/2 niemożnością ich efektywnej naprawy, skutkując śmiercią komórki na drodze tzw. syntetycznej letalności - zjawiska, w którym wyłączenie dwóch lub więcej szlaków naprawy DNA silniej indukuje śmierć komórkową niż wyłączenie każdego ze szlaków pojedynczo. Pomimo to oporność na inhibitory PARP jest zjawiskiem często występującym. Obecnie bada się efektywność połączenia olaparibu z lekami przeciwnowotworowymi, które zakłócają szlak naprawy DNA poprzez rekombinację homologiczną, wykorzystując inhibitory blokujące kinazę ATR lub CHK1. Celem niniejszej rozprawy doktorskiej była ocena cytotoksycznego działania inhibitorów kinazy ATR (AZD6738, ATRi, ceralasertib) i CHK1 (MK8776, CHK1i) w skojarzeniu z inhibitorem PARP (AZD2281, PARPi, olaparib) w trzech liniach komórkowych raka jajnika o zróżnicowanym profilu genetycznym – SKOV-3 – model BRCAWT oporny na cisplatynę – konwencjonalną terapię stosowaną w leczeniu raka jajnika, OV-90 – model raka jajnika niskozróżnicowanego BRCAWT, z mutacją TP53 oraz PEO-1 – model raka jajnika niskozróżnicowanego, BRCA2MUT, predysponowany do leczenia PARPi. Eksperymenty, stanowiące podstawę tej rozprawy doktorskiej potwierdziły, że skojarzone działanie PARPi+ATRi lub PARPi+CHK1i prowadzi do nasilonej śmierci komórek raka jajnika na drodze syntetycznej letalności, a także, że badane kombinacje hamują, zależnie od czasu, proliferację komórek raka jajnika i działają niezależnie od mutacji BRCA. Wyniki te sugerują, że koncepcja stosowania terapii skojarzonej PARPi+ATRi oraz PARP+CHK1i może przynosić potencjalne korzyści terapeutyczne w leczeniu raka jajnika.

Introduction: Chromosomal translocations are common in cancer. In many cancers such as prostate cancer, leukemia and Ewing sarcoma, chromosomal translocations are the main driver of malignancy. Ewing sarcoma is a cancer diagnosed mostly in children and adolescents that has very grim outcomes for patients with metastasis and recurrent disease. Malignancy in Ewing sarcoma is due to EWS-FLI1, an aberrant transcription factor that is the result of a chromosomal translocation. EWS-FLI1 is the main driver of oncogenesis in Ewing sarcoma and has been the target of many drugs developed to treat the disease. Mithramycin (MTM) was identified as a potent inhibitor of EWS-FLI1, but despite its pre-clinical success there were serious toxicities associated with its use including hepatotoxicity. Adverse effects associated with MTM treatment resulted in treatment regimens that were not enough to decrease tumor size. Because of dose limiting toxicities there has been a collective effort to develop analogues of MTM that are less toxic but more effective against inhibiting EWS-FLI1 activity. Here I present an assessment of the unique molecular biologic activity of novel MTM analogues. Preliminary pharmacokinetic studies have shown that in mice MTM has a shorter plasma half-life of about five hours [1]. MTM analogues MTMSA-TRP and MTMSA-TRP-TRP have extended plasma half-lives of twelve hours and over eighteen hours, respectively. Due to their longer plasma half-lives we hypothesized that MTM analogues have less affinity for organic anion transporting polypeptides (OATP) because of their physiochemical properties. Additionally, DNA binding assays suggested that MTM analogues bind DNA and stabilize the EWS-FLI1 transcription factor reducing transcription of EWS-FLI1 target genes [2]. We hypothesize that MTM analogues bind DNA displacing and/or stabilizing EWS-ETS transcription factors. These analogues bind DNA resulting in DNA damage, we hypothesize that combining MTM analogues that cause DNA damage with PARP1 inhibitors will result in a synergistic response in cell lines expressing EWS-FLI1. Methods: Based on preliminary cytotoxicity and pharmacokinetic studies, the ionization constant (pKa) of MTM and analogues was measured. Based on the fluorescence of MTM core aromatic rings at increasing pH, pKa values were estimated. RKO cells stably transfected with pIRES/OATP1B1 or pIRES/OATP1B3 expression vectors were used to complete competitive inhibition studies. Measuring MTM and analogue competition into OATP vs. 8- (2- [fluoresceinyl]aminoethylthio) adenosine- 3', 5'- cyclic monophosphate (8-FcA), a fluorescent substrate of OATP transporters. OATP expressing cells were also subjected to cytotoxicity studies to evaluate uptake of MTM analogues over a 72-hour assay. FVB/N Oatp1a/1b knockout (KO) and FVB/N wildtype mice were used to complete pharmacokinetic studies evaluating differences in uptake of MTM and MTMSA-TRP-TRP after pre-treatment with the OATP inhibitor rifampicin for 1 hour. Time-resolved fluorescent energy transfer (TR-FRET) assays used CAL635 labeled DNA oligomers specific for ERG-his tagged and Sp1-his tagged peptides. Europium-labeled (EU) anti-HIS antibodies were combined with the peptide to evaluate shifts in FRET, a response to treatment with MTM or analogue. Cellular thermal shift assays (CETSA) were used to determine shifts in thermal stability of EWS-FLI1, Sp1 and RNA polymerase II (RNAPII). Protein expression studies evaluating expression of MTM analogue activity against RNAPII phosphorylated Serine 2 (p-Ser2) on its c-terminal domain (CTD). mRNA expression of DNA damage repair genes CDK12, BRCA1, FANC1 and FANCD2 in cell lines expressing EWS-FLI1 (TC-32 and A673) vs. cell lines that do not express EWS-FLI1 (PC3 and TC-32shEWS-FLI1#6) was also evaluated. Protein expression of the dsDNA damage biomarker γ-H2AX was also evaluated in TC-32 and PC3 cell lines. TC-32 and PC3 cells were also used to evaluate synergy between MTM analogues and the PARP1 inhibitor olaparib. Results from cytotoxicity studies were used to calculate the combination index (CI) values of each drug combination. Based on CI values from in vitro studies a xenograft study in female athymic nu/nu mice evaluating tumor regression after treatment with olaparib and MTMSA-TRP was conducted. Results: Evaluation of the molecular function of MTM analogues was largely based on differences in their structure and pKa. pKa estimation studies revealed that analogues with large hydrophobic conjugations to the C3 side chain of MTM, resulted in a higher pKa than the parent compound MTM. This result also correlated with OATP mediated uptake into RKO-OATP1B1 and RKO-OATP1B3 cells in competition assays and in cell viability assays. Pharmacokinetic studies in OATP1B1 humanized mice demonstrated an increase in exposure of MTM after treatment with rifampicin. TR-FRET and CETSA assays demonstrated pointed evidence of displacement and stabilization of EWS-FLI1 and ERG transcription factor DNA binding domains (DBD). In TR-FRET assays MTM analogues with amino acid side chains displaced ERG binding more potently than they did SP1, and CETSA assays demonstrated a physical interaction between EWS-FLI1 and MTM analogues. Sp1 CETSA experiments showed minimal shifts in thermal stability between the different analogues, suggesting that there is no physical interaction between analogue and the Sp1 transcription factor. RNAPII CETSA experiments showed that MTM and MTMSA analogues were more likely to stabilize this protein vs. MTMox analogues. Protein expression of RNAPII CTD p-Ser2 also demonstrated a concentration and EWS-FLI1 dependent decrease in expression. mRNA expression of DNA damage repair proteins CDK12, BRCA1, FANC1 and FANCD2 was assessed by qRT-PCR, indicating that MTM and analogues also induce a decrease in their expression in an EWS-FLI1 and concentration dependent manner. These interactions were the rationale to investigate the possibility of MTM analogues being DNA damaging agents and transcriptional inhibitors. MTM analogues were shown to induce DNA damage in TC-32 cells but not in PC3 cells, demonstrating a concentration and EWS-FLI1 dependent increase in γH2AX expression. Cell viability studies combining MTM, analogues and olaparib proved synergy specific to cells expressing EWS-FLI1. Xenograft synergy studies also showed that we can lower the dose of MTMSA-TRP combined with olaparib and gain a more sustained reduction in tumor size. Conclusion: These studies in whole highlight the importance of developing and understanding the underlying mechanism of these novel analogues. Physiochemical properties of the analogues compared to MTM directly contribute to their exposure. And binding properties correlate with specificity and cytotoxicity in ETS translocation expressing cell lines. MTMSA-TRP has also shown promise in xenograft studies however, in these studies we combined it with olaparib to take advantage of the synthetic lethality imparted on cell lines that express EWS-FLI1.

Response to polyadenosine diphosphate ribose polymerase (PARP) inhibitors in platinum-resistant ovarian cancer and in the absence of BRCA mutations is very low. Combining PARP inhibitors with other agents might overcome this lack of activity. Here we describe the rationale and design of GEICO1601-ROLANDO (resistant ovarian cancer treated with olaparib and pegylated liposomal doxorubin; NCT03161132). ROLANDO is a Phase II single-arm multicenter trial in which patients are treated with a combination of olaparib and pegylated liposomal doxorubicin (PLD) in platinum-resistant epithelial ovarian, primary peritoneal, or Fallopian tube cancer regardless of the BRCA mutation status. The primary end point is progression-free survival at 6 months. Other secondary end points are response rate, disease control rate, quality of life and overall survival. Lay abstract: PARP inhibitors as a single agent have shown very modest activity in platinum-resistant ovarian cancer in a BRCA nonselected population. The GEICO1601-ROLANDO trial is a protocol designed with the aim of assessing efficacy and safety of the combination of olaparib and pegylated liposomal doxorubin followed by olaparib maintenance in this setting.

Small Cell Lung Cancer (SCLC) is a highly aggressive neuroendocrine tumor with a 5% survival rate over 5 years. Though SCLC comprises 13% of all cases of lung cancer the median survival time of 14.5 months has seen little improvement over the last four decades. Standard treatment relies on DNA damaging agents such as Cisplatin/Etoposide (EP) which induce a high response rate of 60-70%. Despite this initial response, nearly all patients will relapse rendering first-line therapies ineffective. Furthermore, SCLC has been shown to develop chemotherapy cross-resistance in which resistance to first-line chemotherapies will confer resistance to additional DNA damaging agents thereby reducing treatment efficacy and duration of response. Cross-Resistance constitutes a major clinical issue whose underlying mechanisms remain a mystery. The modest improvements in SCLC patient outcomes over the decades may be partially explained by the existing systems of study. Current methodologies of SCLC study rely on cell lines, patient samples, and Genetically Engineered Mouse Models which have little functional correlation to clinical outcomes. While few sources have proposed Patient Derived Xenograft (PDX) systems as an improved alternative, significant data remains sparse. Without a robust model system which accurately recapitulates patient outcomes, molecular pathways driving resistance cannot be uncovered. Here we present the generation of 34 SCLC PDX models which maintain both genomic and functional fidelity. Furthermore, treatment of a 30-model subset with first-line chemotherapy EP and a novel chemotherapy Olaparib/Temozolomide (OT) allowed for functional and molecular comparison between groups. Our findings demonstrate incomplete independent resistance mechanisms between EP and OT treatment with a small overlap of 31 genes involved in glycolysis and xenobiotic metabolism.