Year

2017

Degree Name

Doctor of Philosophy

Department

School of Chemistry

Abstract

The presence of up to 30-40% more sialic acid (or hypersialylation) on the tumour cell surface compared to normal cells, along with a marked up-regulation of sialyltransferase (ST) activity is a well-established hallmark of cancer. Due to the critical role of these glycan-building enzymes in tumour growth and progression, ST inhibition has emerged as a potential new antimetastatic strategy for a range of cancers including pancreatic and ovarian cancer. Human STs are divided into four subfamilies or groups based on their linkages (ST3Gal, ST6Gal, ST6GalNAc and ST8Sia) with each subtype controlling the synthesis of specific sialylated structures each with a unique biological role. This has important implications for inhibitor development, as STs also play significant roles in immune responses, inflammation, viral infection and neurological disorders. Thus, to advance to the clinic it is essential to develop subtype selective ST inhibitors.

A wide range of ST inhibitors have been reported including inhibitors from design, from Nature and from high-throughput screening. In the absence of crystal structures of the human STs, the early inhibitors were designed and their activity optimised by using empirical observations for almost two decades. However, the recent publication of crystal structures of mammalian forms of the major enzyme subtypes ST3Gal, ST6Gal and ST8Sia, has furthered our understanding of the differences between the subfamilies and provided the opportunity to use computational tools for the design and optimisation of inhibitors.

All ST subtypes use CMP-Neu5Ac as the natural donor, and the most potent ST inhibitors todate are all transition-state analogue inhibitors that essentially mimic the donor structure. As part of our study we published a major review focussing on cytidine-based inhibitors comprised of three key fragments: a nucleoside component and a sialic acid mimic linked by a phosphate group. Despite the nanomolar potency of the reported phosphate-linked inhibitors, little is known about their selectivity towards the different ST subtypes or in vivo activity. Furthermore, their synthesis is often challenging and utilises expensive and low yielding methods. To facilitate an in-depth biological study of ST inhibitors, the design of novel compounds must be directed by the newly emerging structure activity relationships to create libraries of inhibitors via highly efficient and inexpensive synthetic routes.

Therefore, the current project was aimed at the convergent synthesis of triazole-linked inhibitors, which would allow the versatile ligation of nucleoside derivatives with a variety of functionalised sialic acid mimics. The choice of the building blocks for this strategy was guided by computational methods using the recently released ST crystal structures. A series of 19 α- hydroxyphosphonae esters were synthesised for the preparation of carbamate-based inhibitors and also examined for their own inherent anti-inflammatory activity. A series of 10 nucleoside derivatives and 15 alkynes fragments were also prepared, leading to 27 novel triazolonucleoside products. The preliminary testing of our library of triazole-based derivatives towards ST8Sia-II showed promising inhibition up to 98% when tested at a concentration of 100 μM. The incorporation of a fluorine atom on the nucleoside or the sialic acid mimic as well as heterocyclic structures provided the greatest inhibitory activity. Finally, the absence of toxicity of our candidates at concentrations of up to 300 μM allows for further cell-based evaluation of the effects of our compounds on tumour cell adhesion, migration and invasiveness.

FoR codes (2008)

030503 Organic Chemical Synthesis, 030401 Biologically Active Molecules, 030405 Molecular Medicine, 030402 Biomolecular Modelling and Design

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.