Development of in vitro and in vivo models to underpin advances in human radiotherapy

Abstract

Radiotherapy (RT) is commonly used for the local control of many cancer types. Unfortunately, not all patients will achieve a therapeutic benefit, and some will develop loco-regional recurrences and/or metastatic disease. The hypoxic nature of the tumour microenvironment and the development of radioresistant cancer cells can contribute to these treatment failures. Understanding the mechanisms involved in acquired radioresistance and the development of techniques to identify and target hypoxic tumour areas has the potential to improve RT response rates. The first aim of this project was to investigate the development of acquired radioresistance and identify radiation-induced secreted biomarkers which could be used as indicators of a radiation response or radiosensitivity. Human radioresistant (RR) breast cancer cell lines were developed from parental MCF-7, ZR-751 and MDA-MB-231 cells. Parental and RR cells underwent genotypic, phenotypic and functional characterisation. RR cells exhibited enhanced migration and invasion, with evidence of epithelial-to-mesenchymal transition. MCF-7 RR and ZR-751 RR cell lines exhibited significant phenotypic changes, including loss of ERα and PgR expression and increased EGFR expression, which were associated with the down-regulation of ER signalling genes and up-regulation of genes associated with PI3K, MAPK and WNT pathway activation. A change in subtype classification from luminal A to HER2-overexpressing (MCF-7 RR) and normal-like (ZR-751 RR) subtypes was also observed, consistent with radiation and endocrine therapy resistance and a more aggressive phenotype. To identify biomarkers secreted in response to radiation, human and canine breast and ovine lung cancer cell lines were radiated. Secretome samples were analysed by liquid chromatography-mass spectrometry. Using results from the MCF- 7 cell line, 33 radiation-induced secreted biomarkers were identified which had higher (up to 12-fold) secretion levels compared to untreated controls. Based on secretion profiles and functional analysis 9 candidate biomarkers were selected (YBX3, TK1, SEC24C, EIF3G, EIF4EBP2, NAP1L4, VPS29, GNPNAT1 and DKK1) of which the first 4 underwent in-lab validation. To identify biomarkers related to radiosensitivity transcriptomic analysis identified higher expression of genes encoding 7 of the candidate biomarkers in the MCF-7 cell line compared to its radioresistant derivative. WB analysis identified increased levels of the 4 biomarkers in the conditioned media of parental cells 24 h post-radiation which was not seen in the RR cell lines. These biomarkers, which had differential gene expression and secretion profiles between parental and RR cell lines, may be useful for both predicting and monitoring a tumour’s response to RT. A further aim was to investigate the biocompatibility and functionality of an implantable electrochemical sensor, developed within the Engineering and Physical Sciences Research Council funded IMPACT project. This sensor was designed to measure tissue O2 tension (ptO2) within a tumour, enabling the identification and monitoring of radioresistant hypoxic tumour areas. This study developed a novel in vivo tumour xenograft model to evaluate the potential of 6 materials (silicon dioxide, silicon nitride, Parylene-C, Nafion, biocompatible EPOTEK epoxy resin and platinum) used in the construction of the sensor, to trigger a foreign body response (FBR) when implanted into a solid tumour. Following implantation none of the materials affected tumour growth and all mice remained healthy. Immunohistochemistry performed on the tumour showed no significant changes in necrosis, hypoxic cell number, proliferation, apoptosis, immune cell infiltration or collagen deposition around the implant site. The absence of a FBR supports their use in the construction of implantable medical devices. In vivo validation of the O2 sensor to provide real-time measurements on intra-tumoural ptO2 was performed using a novel large animal ovine model. To achieve this aim, we developed a novel computed tomography (CT) guided transthoracic percutaneous implantation technique for the delivery of sensors into naturally occurring ovine pulmonary adenocarcinoma (OPA) tumours. This model successfully integrated techniques such as ultrasound, general anaesthesia, CT and surgery into the OPA model, all of which are techniques commonly used in the treatment of human lung cancer patients. This methodology resulted in the accurate implantation of sensors into OPA tumours with minimal complications and demonstrated the sensor’s ability to detect changes in intra-tumoural ptO2 following manipulation of the inspired fractional O2 concentration (FiO2). To investigate other possible clinical applications, sensors were validated for measuring intestinal ptO2 using a novel rat model. These experiments assessed the potential of the sensor to monitor intestinal perfusion following an intestinal resection and anastomosis. The sensor was placed onto the serosal surface of the small intestine of anaesthetised rats that were subsequently exposed to ischaemic, hypoxaemic and haemorrhagic insults. Decreases in intestinal ptO2 were observed following superior mesenteric artery occlusion and reductions in FiO2; these changes were reversible after reinstating blood flow or increasing FiO2. These results provided evidence that the sensors could detect changes in intestinal perfusion which could be utilised in a clinical setting to monitor peri-anastomotic intestinal ptO2. Overall this PhD project has conducted both in vitro and in vivo work aimed at the investigation of mechanisms of radioresistance, identifying secreted biomarkers of radiosensitivity and validating the ability of an implantable sensor to measure real-time intra-tumoural and visceral surface O2 tension. Identification of factors contributing to poor RT responses, such as radioresistance development and hypoxic tumour areas could provide a means by which RT could become personalised. Patients identified as having radioresistant tumours or those not responding to RT based on radiation-induced secreted biomarkers, could be given higher dose of radiation or radiosensitising agents to improve patient outcomes.