Search

Oesophageal cancer is the sixth leading cause of cancer death in the world and the eight most prevalent diagnosed cancer. Despite the improvements in diagnosis and treatment for oesophageal cancer, it remains one of the deadliest cancers worldwide, and its prognosis remains poor. Protein tyrosine kinases play an essential role in the development of different types of cancer, several of which have been found overexpressed on oesophageal cancer. Applying a combination of ligand-based design and phenotypic screening iteratively, novel potent kinase inhibitors that target oesophageal cancer were developed. This novel drug discovery approach, developed at the Innovative Therapeutics lab, served to accelerate the discovery of anticancer lead compounds with optimal drug-like properties. Firstly, an efficient synthetic route was developed to prepare pyrazolo[3,4-d]pyrimidine derivatives with novel C3 and N1 substituents. Applying this synthetic route, several compounds were produced, followed by phenotypic screens against oesophageal cancer cell lines. This process (design-synthesis-screening) was iteratively applied to generate structure-activity-relationships which informed further series of compound design and future optimisation. A total of 76 novel pyrazolo[3,4-d]pyrimidine derivatives were synthesised and tested in different oesophageal cancer cells. From the design-synthesis-screening rounds, several active compounds were identified and 3 lead compounds with low-micromolar potencies were discovered. Broader biological evaluation of the best compounds included screening in an extended cell panel, cell cycle assay and kinome profiling. The data from these studies identified lead compounds with potent antiproliferative activity, against a wide range of oesophageal cancer cell lines, and can induce cell cycle arrest and apoptosis. The kinome profiling identified Aurora A and PIK3CA/PIK3R kinases as the potential targets of the best lead compound. This finding is consistent with the data from cell screenings and cell cycle assay, AuroraA and PIK3CA kinases are among the pivotal regulatory proteins on distinct stages of the cell cycle and have been linked to the malignancy development of oesophagel cancer. This thesis represents a promising starting point for the development of novel inhibitors of oesophageal cancer, justifying further research efforts to advance the development of the lead compound.

In 2020, cancer was the leading cause of early death in over 57 countries, accounting for approximately 10 million deaths worldwide. Brain cancers remain one example of greatest unmet need, with on average 14% survival 10 years post-diagnosis. Glioblastoma multiforme (GBM) is the most common and aggressive type of brain cancer in adults, with an average prognosis of 12 months. Current treatments only extended the average prognosis by 2 months. GBM is characterised by a complex set of mutations, with significant disruption to kinase pathways, specifically RTK signalling. Drug design in the past 30 years has predominately been target-based, focussing on the development of highly selective molecules that act against specific biological targets. Parallel to this, the rate of attrition - the number of compounds that fail in the clinics because of poor physicochemical properties - has greatly increased. For complex cancers, a targeted approach is not sufficient to discover new therapies. Phenotypic drug discovery, where compounds are designed based on defined pharmacological endpoints, without prior knowledge of the molecular target, was used to develop new compounds for the improved treatment of glioblastoma. Libraries of anticancer compounds based on the pyrazolo[3,4-d]pyrimidine scaffold have been designed for the treatment of glioblastoma. Compounds were synthesised in a convergent manner, and data from phenotypic screens against glioma cell-lines were used to build up structure-activity-relationships, facilitating further rounds of design and optimisation. Over 230 novel molecules were designed, synthesised, and screened, identifying several lead molecules with low-micromolar potencies. Further evaluation of lead molecules in cancer models and target deconvolution studies identified lead compounds that demonstrated preferential activity against the mesenchymal sub-type of glioblastoma. Whilst the target of the best lead compounds remains unknown, two selective inhibitors of CSF-1R kinase were identified through kinase screening.

The non-receptor tyrosine kinase SRC plays an important and context-dependent role in a variety of cell signalling pathways that regulate cell motility, survival, division and growth factor signalling. SRC activity has been closely linked to several hallmarks of cancer (cell proliferation, metastasis, angiogenesis) and treatment resistance. Multiple studies found SRC to be aberrantly activated and/or overexpressed in solid tumours, such as breast, colorectal and ovarian cancer, which often correlated with malignancy and poor prognosis. Dual SRC/ABL kinase inhibitors (e.g. dasatinib, bosutinib) are already in clinical use for the treatment of chronic myeloid leukaemia. However, attempts to repurpose them for their SRC inhibition alone have so far failed in clinical trials. Several studies suggest that ABL kinase can act as a tumour suppressor in solid cancers and its inhibition has been linked to cardiotoxic events in patients. This complicates the use of SRC/ABL inhibitors in combination with other anti-cancer agents with cardiac side effects, such as chemotherapeutics, certain monoclonal antibodies and other kinase inhibitors; yet resistance to several of these drugs was found to involve the activation of SRC signalling pathways. Consequently, there is a clear need for more selective SRC inhibitors which have the potential to exhibit greater efficacy and fewer side effects than the ones currently available. The novel small molecule SRC inhibitor eCF506 has recently been discovered by the Innovative Therapeutics Lab at the University of Edinburgh 1. It exhibits nanomolar potency, thousand-fold selectivity for SRC over ABL kinase and superior drug-like properties compared to the current gold-standard SRC inhibitor dasatinib. This thesis continues the preclinical development of eCF506 by (i) investigating its potency against different cancers in in vitro and in vivo models, (ii) assessing its safety and pharmacokinetic properties and (iii) identifying a unique binding mode to SRC kinase.

Chemotherapeutics are potent molecules capable of systematically treating cancer. As healthy tissues contain features also inherent to cancer cells, treatment often results in unwanted sideeffect. As chemotherapeutic side-effect produces significant harm and often limits optimal drug dosing, new strategies must be developed to improve treatment selectivity. A prodrug strategy provides one option to improve the selectivity of an established chemotherapeutic. By modifying a pharmaceutically active drug, interaction with biology may be functionally masked. Subsequent ‘un-masking’ the prodrug exclusively at the intended treatment site may direct treatment only to where the anticancer effect is required. This thesis progresses the novel approach of bioorthogonal organometallic (BOOM) prodrug activation. A metal catalyst and masked chemotherapeutic constitute reaction partners to provide a new strategy for intratumoural prodrug activation. Whereby the prodrug and metal catalyst are independently non-cytotoxic, in combination the prodrug undergoes catalytic activation to deliver an anticancer affect. By positioning the metal catalyst within a tumour (i.e. by microsurgery), an administered masked prodrug sensitive to catalyst-mediated activation could allow for ‘targeted’ chemotherapy localised to the tumour site. The design, synthesis and study of new BOOM prodrug candidates are reported herein. Novel protecting groups are developed to enhance drug masking to biology and subsequent catalyst-mediated activation. Prodrug screening studies are carried out in cancer cell culture models, with zebrafish and in ex vivo rodent model tumour explants. The catalyst, a palladium (Pd0) functionalised bead system, is optimised for enhanced activation, drug release and in vivo implantation. The potentially infinite generation of active chemotherapeutics exclusively in tumour would increase the efficacy of treatment whilst reducing harmful effect on healthy tissue.

In recent years, there has been a major movement in the pharmaceutical industry towards the development of molecules that selectivity inhibit a previously-validated specific target. This is referred to as target-based drug discovery. It was hoped that adopting this approach would usher in a new golden age of drug discovery. However, this has not been the case, with issues arising such as the target’s mechanism of action being poorly understood, with it not playing the expected role in the disease progression, or feedback resistance mechanisms causing the target to lose its role in the disease. In contrast to this, in the past 20 years it has been argued that developing drugs in a target-agnostic way and screening them against an expressed phenotype i.e. phenotypic drug discovery, has been more successful, despite fewer programs being run in the manner. The AXL kinase is a receptor tyrosine kinase (RTK) and a member of the TAM family, along with MER and TYRO3. AXL has long been associated with numerous types of cancer. Having been first discovered in 1991 in acute myeloid leukaemia (AML), it has gone on to be more associated with advanced solid tumours such as brain, breast, and lung, with the trend being that increased AXL correlates with a poorer prognosis for the patient. Upon the activation of AXL by the vitamin K ligand GAS6, a series of downstream pathways are activated that go on to encourage cell survival, proliferation, and migration. In addition to this, AXL has been shown to be involved in crosstalk with other kinase pathways, resulting in AXL expression being associated with chemoresistance and survival mechanisms. Despite the promising outlook for AXL inhibitors, to date only one selective AXL inhibitor, BGB324 (formally R428) has entered clinical trials, with selective AXL inhibitors being difficult to develop due to a lack of a crystal structure or a reliable homology model. To address the aforementioned issues that target-based approaches can suffer from, and due to AXL lacking a crystal structure, the work in this thesis utilised a pragmatic drug design method that started from ligands/existing scaffolds known to inhibit the target from the literature (publications, clinical trials and patents). A series of small libraries were prepared and then tested against a selected phenotype e.g. cell viability, in at least two cell types: one that expressed the target (e.g. AXL) and one that did not. Hits were optimised for potency against the desired phenotype. The compounds then went through target deconvolution (kinase screening) to confirm the target of the inhibitors. Employing this approach, we initially synthesised two small libraries of potential AXL inhibitors. The potency of these compounds was tested using cell-based phenotypic assays, by evaluating cell viability in both native and chemo-resistant breast cancer cells. These libraries were optimised through focused combinatorial synthesis and phenotypic screening, to yield a small collection of antiproliferative hits. These hits were then profiled against a panel of twelve select kinases. The first library, while giving some important structural information, did not inhibit the kinases screened in a meaningful manner. However, the second library gave several potent compounds, inhibiting AXL, FLT3, and RET, with one compound being selective for AXL. The leads from this series were optimised further, through SAR studies, gaining important structural information in order to improve potency and selectivity of the compounds. The flexibility of the phenotypic cell-based approach allowed the pursuit of FLT3 inhibitors, resulting in the synthesis of one of the most potent FLT3 inhibitors synthesised to date.

I hypothesise that cancer cells with high aldehyde dehydrogenase (ALDHhigh) activity present a new therapeutic target and will be selectively sensitive to 5-nitrofuran pro-drugs. Cancers are heterogeneous and contain subpopulations of ALDHhigh cells with tumour initiating potential. ALDH enzymes metabolize toxic aldehydes, and are highly expressed in somatic and cancer stem cells (CSCs), although their function in CSCs is not fully understood. In a small molecule screen coupled with target ID, Zhou et al. (2012) recently discovered that clinically active 5-nitrofurans (5-NFNs) are substrates of ALDH2. 5-NFNs are a class of pro-drug widely used to treat bacterial and parasitic infections, where their relative specificity is driven by nitroreductases, but little is known about the enzymes that bio-activate 5-NFNs in humans. Recent clinical cancer research has found that the 5-NFN, nifurtimox, has anti-cancer properties and it is currently in Phase 2 clinical trials for neuroblastoma and medulloblastoma (ClinicalTrials.gov Identifier: NCT00601003), however the mechanism underlying this anti-cancer activity is unknown. In melanoma and other cancers, ALDH1A1 and ALDH1A3 are highly expressed in CSCs. I demonstrate the anti-cancer activity of 5-NFNs in cancer cell lines, where they express high sensitivity to 5-NFNs in cell viability assays (A375 melanoma cells EC50 = 867nM). To test if ALDH1 enzymes are substrates of 5-NFNs, I performed in vitro activity assays by monitoring NADH production (λ = 340nm). I found that the clinically available 5-NFNs, nifuroxazide and nifurtimox, in addition to our own newly synthesised 5-NFNs, are competitive substrates for human ALDH1A3 activity in vitro (P < 0.05). Notably, nifuroxazide is not a substrate for ALDH2, suggesting that nifuroxazide may show selectivity toward ALDH1. Enzymatic assays with purified human ALDH2, demonstrate that ALDH2 requires NAD+ for bio-activation of 5-NFNs. Consistent with these assays, I found that 5-NFNs are competitive substrates for ALDH activity in melanoma cells by Aldefluor™, with 5-NFNs displaying a prolonged competitive inhibition of ALDH activity compared with the known inhibitor, DEAB. Importantly, no-nitro control compounds show no activity toward ALDH enzymes in vitro or in culture. Kinetic living-cell imaging (IncuCyte ZOOM®) reveals that a subpopulation of ALDH1A3 siRNA transfected A375 cells are protected from 5-NFN toxicity (P > 0.05) and cell death (DRAQ7™: P < 0.0001), demonstrating a functional role for ALDH1A3 in mediating 5-NFN activity in cancer cells. In contrast, A375 cells overexpressing ALDH1A3 by cDNA transient transfection were hypersensitive to 5-NFNs (P < 0.001), determined by Muse™ cell viability. Computational docking studies reveal that 5-NFNs have the potential to fit within the interior of the ALDH enzymatic cavity and interact with the catalytic cysteine, thereby offering a potential mechanism for 5-NFN bio-activation. Finally, in collaboration, we show a unique interaction between 5-NFNs and ALDH using mass spectrometry and have initiated protein crystallography trials. My work demonstrates a novel and biologically relevant 5-NFN-ALDH interaction in cancer cells. I propose 5-NFNs have the potential to target ALDHhigh CSCs within a tumour and advance the repurposing of clinical 5-NFN pro-drug antibiotics as anti-cancer therapeutics.

Tumour necrosis factor receptor associated factors (TRAFs) play a key role in signal transduction in mammalian cells. Several members of the TRAF family have been identified but only TRAF2 and TRAF6 are implicated in the regulation of both osteoclastic bone resorption and breast cancer. Here I studied the role of TRAF2 and TRAF6 in breast cancer induced osteoclastogenesis and osteolysis. I observed that TRAF2, but not TRAF6, is highly expressed in a highly metastatic bone-tropic clone of the human MDA-MB-231-BT (MDA-231-BT) breast cancer cells when compared to parental MDA-MB-231 (MDA-231) cells. Targeted knockdown of TRAF2, but not TRAF6, in both parental MDA-231 and bone-tropic MDA-231-BT breast cancer cells by siRNAs markedly reduced cell migration and significantly reduced the ability of these cells and their conditioned medium to induce osteoclast formation in RANKL stimulated bone marrow cultures. Encouraged by these data, I generated stable parental MDA-231 and bone-tropic MDA-231-BT breast cancer cell lines overexpressing TRAF2 using a retroviral approach. Then, I went on to show that overexpression of TRAF2 in parental MDA-231 cell line significantly stimulated directed cell migration and 3D invasion in vitro. Bone-tropic MDA-231-BT breast cancer cells over expressing TRAF2 or their conditioned medium were significantly effective in enhancing RANKL induced osteoclast formation in vitro. Mechanistic studies in parental MDA-231 and bone-tropic MDA-231-BT breast cancer cells revealed that over-expression of TRAF2 enhanced cell migration and osteoclastogenesis via a mechanism that involves the activation of the breast cancer oncogene IKKepsilon (IKKε) coupled with significant increase in levels of Vascular endothelial growth factor (VEGF). Ex vivo studies in human MDA-231-mouse calvaria organ co-cultures showed that conditioned medium obtained from MDA- 231 cells enhanced calvarial osteolysis. In vivo studies showed that overexpression of TRAF2 in the human breast cancer cells MDA-231 enhanced tumour incidence and tumour volume after orthotopic injection and exacerbated osteolysis after supracalvarial injection of conditioned medium from these cells. In conclusion, our studies showed that the TRAF2/IKK/VEGF axis in breast cancer cells regulates breast cancer cell motility in vitro, osteoclastogenesis in vitro and osteolysis ex vivo and in vivo. However, the role of TRAF2 in bone metastasis associated with breast cancer will require further in vivo investigation.

Chemotherapy is widely used to treat various forms of cancer. However, some chemotherapeutic drugs, due to their antineoplastic properties, also act upon healthy cells which normally replicate rapidly causing a plethora of undesirable side effects. One rising and promising therapeutic strategy is the development of prodrugs. Prodrugs are derivatives of the pharmaceutically active drugs but require an enzymatic or biochemical transformation within a certain biological space in order for it to become activated and capable of exerting the desired pharmacological effect. As a novel prodrug approach, this thesis describes the pioneering use of a bioorthogonal organometallic (BOOM) activation strategy to develop spatially-controlled anticancer treatments. Bioorthogonal reactions are selective chemical processes between two abiotic reagents in a biological system that do not interfere with the system’s biotic components. In BOOM reactions, one of the reagents is a metal catalyst, which if immobilized, could in principle allow for the local transformation of a continuous flow of a bioorthogonal chemo-substrate indefinitely. To exploit the benefits of this paradigm in anticancer therapy, this thesis reports the design, synthesis and screening of a set of prodrugs masked with bioorthogonal protecting groups sensitive to activation by a catalysts-based “activating device”. Specifically, it describes the synthesis of palladium (Pd0) functionalized resins (the activating device) capable of activating cytotoxic pyrimidine analogue prodrugs masked with Pd0-labile protecting groups. Both the Pd0 functionalized resins and the BOOM-activated prodrugs are independently non-cytotoxic. However, once in combination together, the Pd0 is capable of mediating the removal of the masking groups in situ and rendering the drugs in their cytotoxic state with comparable antiproliferative properties to the unmodified parental drugs in vitro. The Pd0 resins also display biocompatibility and local catalytic activity inside zebrafish embryos. This approach is intended to generate a more targeted therapeutic treatment regime while minimizing harm to normal healthy tissues through the local generation of prodrugs which are not dependent on intrinsic biological activators but by an external activating device, thus reducing the systemic presence of the drug.

Traditionally, drug discovery programs have focused on prioritising compounds by their affinity to a specific target in isolation, which was hypothesised to be the cause of a particular disease. Through chemical inhibition, the disease could, thus, be prevented or at the very least, controlled. These hypotheses require significant validation before drug screening can begin which relates to lengthy and expensive programs. Furthermore, drug screening against a single target in isolation is not a realistic model of cellular behaviour and is not appropriately tailored to more complex diseases such as cancer. Phenotypic drug discovery, on the other hand, bypasses any involvement of known targets, instead focusing on the desired outcome – the phenotype. In this way, drugs are biased by their potency on the phenotype and not against any particular targets. The molecular mechanism of action need not be known at all, however, it can be useful to later reveal the target(s) involved by various deconvolution methods. This thesis describes a cooperative ligand based phenotypic drug discovery approach, undertaken in order to develop more suitable small molecule drugs for cancer treatment. For this purpose, the promiscuous pyrazolopyrimidine inhibitor PP1 was chosen as a starting model compound. Modification of PP1 on the N1 position allowed a series of water solubilising groups to be incorporated into the pyrazolopyrimidine scaffold which created an initial 12-membered library. Testing against MCF7 breast cancer cells and looking at phenotypic end points such as cell proliferation, cell mobility and cell cycle, generated early target-agnostic structure/anti-proliferative activity relationships. These early results, along with compounds published in recent literature, were used to generate further libraries. Profiling lead compounds against a selection of 18 kinases known to be targeted by PP1, showed the compounds were inhibiting either SRC family or mTOR kinases which enabled the creation of two, structure specific, groups of inhibitors. Further lead optimisation led to the rapid discovery of preclinical candidates with excellent drug-like properties and potencies in both cellular assays and against their respective targets. Compounds also showed improved selectivity profiles compared to PP1 and commonly known inhibitors of SRC and mTOR kinases. Reported, herein, is the discovery of the first sub-nanomolar SRC inhibitor which does not inhibit the kinase ABL and shows excellent properties suitable for further preclinical development.