Abstract
The transcription factor hypoxia-inducible factor-1 (HIF-1) upregulates the expression of hundreds of genes involved in the adaptation of tumours to the hypoxic microenvironment. HIF-1 levels and transcriptional activity are determined by hydroxylation of proline and asparagine residues on the regulatory alpha subunit. These signals result in HIF-1α degradation and prevention of the formation of an active transcription complex. The proline and asparagine hydroxylases that carry out these signalling reactions belong to the family of iron-containing 2-oxoglutarate-dependent dioxygenases, and depend on oxygen, iron (Fe), and the metabolic intermediate 2-oxoglutarate for activity. They also require ascorbate as a cofactor and enzyme activity has been shown to be compromised when ascorbate is limiting. Adequate supplementation of cells with ascorbate has been shown to reduce HIF-1 activation in vitro, and we hypothesise that a similar activity could affect tumour growth in vivo. Therefore, it was the aim of this thesis to investigate if increased ascorbate uptake into the tumour would be associated with a reduction in HIF-1 activity, slowing tumour progression. This hypothesis was tested in Gulo-/- mice, a model of the human ascorbate deficiency condition, in which we can manipulate the amount of vitamin C (ascorbate) delivered. Three syngeneic mouse tumour cell lines were used to implant subcutaneous tumours in Gulo-/- mice. The current study has monitored tumour growth and take rates, measured tumour ascorbate content and analysed the harvested tissues for HIF-1 activity and related gene expression. Tumour morphology was also investigated. Several different approaches were taken to investigate the effect of ascorbate on tumour growth, including manipulating dietary ascorbate intake, modelling the effect of high dose vitamin C (HDVC) administration on pre-established tumours, changing HDVC dosage schedules and combining HDVC with chemotherapy agents.
In the initial study, mice were supplemented with dietary ascorbate that resulted in optimal, sub-optimal or low tissue levels, and B16-F10 (melanoma), LL/2 (Lewis lung carcinoma) or CMT-93 (colon cancer) cells were implanted subcutaneously. Tumours formed with the B16-F10 and LL/2 cells, and these were found to be dramatically affected by ascorbate availability, with optimal tissue levels, resulting in a measurable delay in tumour initiation and growth rates. These effects were found to correlate closely with tumour ascorbate content and HIF-1 activation, with higher ascorbate levels being associated with lower HIF-1 activity. Using the LL/2 model, the effect of administering pharmacological ascorbate to mice with established tumours was investigated, and showed that raising the tumour ascorbate content decreased the level of HIF-1α and also slowed tumour growth.
Taken together, the results presented in this study have demonstrated that increased tumour ascorbate was associated with reduced tumour growth. Restoration of tumour ascorbate to 0.6 μmol/g either by optimal dietary supplementation (3300 mg/L; per OS) or by daily intra-peritoneal high dose ascorbate (1 g/kg), was significantly associated with increased time for LL/2 tumours to reach 4x volume. Levels of HIF-1 and its downstream targets CA-IX and VEGF, were also reduced as tumour ascorbate increased. In addition, the pharmacokinetic study presents novel findings, demonstrating that after high dose ascorbate administration, ascorbate is preferentially retained in the tumour, compared to the transient elevation in normal tissue.
Tumour ascorbate levels remained elevated in the presence of daily intervention, and an important aspect of this study was to investigate the effect of increased ascorbate on the tumour response to chemotherapy (cisplatin (2 mg/kg) and paclitaxel (4 mg/kg)). The rationale for this was that HIF-1 activation is thought to confer a survival advantage to the tumour cells and that decreasing this with ascorbate may result in a synergistic anti-tumour effect with the chemotherapy. The combined effect of HDVC and chemotherapy resulted in a similar anti-tumour outcome to using either HDVC or chemotherapy alone. In addition, the current study has found that pharmacological ascorbate does not induce DNA damage or caspase-mediated cell death, and have provided the first in vivo evidence for this effect. Taken together, these results suggest that pharmacological ascorbate does not behave as a pro-oxidant in vivo. Finally, this thesis has presented preliminary in vitro findings, which support the hypothesis that ascorbate is mediating its actions through the maintenance of HIF-hydroxylase function. This was shown by transfection experiments, where mutated HIF-1α was stabilised under normoxia and was no longer sensitive to ascorbate availability.
Therefore, the data from the current study indicates that restoration of optimal intracellular ascorbate, either by continuous dietary supplementation or daily pharmacological administration can be associated with a decline in tumour aggression by dampening HIF-1 levels and its transcriptional activity. In addition, we have provided evidence of the complexity surrounding ascorbate when used in combination with other chemotherapy agents, with regard to a potential mechanism of action via DNA damage and inhibition of the HIF-1 response. This data may have implications for the management of cancer, and future human clinical trials will determine whether human tumour levels of ascorbate can be similarly manipulated.