Design Synthesis and Evaluation of Sialyltransferase Inhibitors as Radiosensitising Agents

Proteins and lipids are ubiquitously decorated with sugar chains known as glycans. Sialic acid is the sugar often found at the terminal position of cell membrane glycans, where it plays an integral role in inter- and intracellular communications and physiological function. Cancer cells often hijack these pathways, overexpressing sialic acid to promote immune evasion, invasiveness and treatment resistance. Reducing sialylation through the use of sialyltransferase inhibitors has been repeatedly demonstrated to reduce deadly metastatic spread and increase a tumour’s responsiveness to treatment with chemotherapy. However, there remains an ongoing need to improve the design of sialyltransferase inhibitors, particularly in terms of their potency, selectivity and bioavailability. Furthermore, sialyltransferase inhibitors have not been thoroughly explored as radiosensitising agents,

This research project first explored the rational design of triazole-linked sialyltransferase inhibitors. These compounds were more analogous to the sialyltransferase substrate, CMP-Neu5Ac, in order to increase their potency relative to the current best-in-class sialyltransferase inhibitors. Two triazole-linked ST inhibitors were synthesised and screened. The 9ʹ-N-triazole was prepared over 11 steps and was measured to have a K_i of 7 ± 3 μM against ST6GAL1, which was a four-fold increase in potency relative to the analogous ethertriazole. The 6ʹ-N-triazole was prepared over five steps and was measured to give 54% inhibition of ST6GAL1 at 10 μM, which is comparable to the ether-triazole lead compound.

Computationally aided drug design was used to investigate the design of ST6GAL1-selective inhibitors. Relative binding free energy perturbations were used to screen a library of nucleoside derivatives targeting the cytidine 4N-pocket. The synthesis and screening of selected derivatives showed these simulations to be particularly powerful when using the ST6GAL1 crystal structure, giving a coefficient of determination (R²) of 0.93. Using the AlphaFold2 model of ST3GAL1 was less accurate, with an R² value of 0.12. These experiments led to the discovery of the 4N-hydroxy and 4N-benzyl derivatives, which showed increased potency and ST6GAL1 selectivity, relative to the uridine lead compound.

Investigations into improving the bioavailability of these inhibitors examined the design of neutral phosphonate bioisosteres, nanoparticle formulations and prodrug derivatives. A neutral, benzoboroxole-functionalised analogue was synthesised over just three steps. The benzoboroxole exhibited an excellent ADME profile and was found to have comparable activity to the phosphonate, giving 41% inhibition of ST6GAL1 at 10 μM. A selection of these inhibitors were then screened in a novel radiosensitisation assay at the Australian Nuclear Science and Technology Organisation (ANSTO) radiobiology laboratories through a collaboration between the University of Wollongong (UOW) and ANSTO. The radiosensitisation effect was observed to be highly variable, ranging from a radioprotective effect of up to 28% in primary cultured mouse astrocytes to a radiosensitisation effect of up to 17% in the A375 melanoma cell line.

This research demonstrates a step-wise improvement to the design of sialyltransferase inhibitors in terms of their potency, selectivity and bioavailability. Moreover, these inhibitors have been used to demonstrate, for the first time, a variable radiosensitisation effect across several different types of tumour and healthy cells. It is hoped that this work will form the foundation for the next generation of sialyltransferase inhibitors that can be used to one day improve the treatment outcomes for cancer patients.