The synthesis and biological evaluation of glycosphingolipids for improved cancer immunotherapy

The immune system plays a crucial role in providing the first line of defence against invading pathogens such as bacteria, viruses and parasites. It is activated when immune cells known as dendritic cells (DCs) detect specific molecules that are foreign to the host, and present them to T cells. This in turn causes the activation of T cells, which marks the start of an immune response leading to the clearance of the invader. The pathogen-derived molecules recognised by the immune cells are typically peptides and their role as activators of the immune system is well established. While T cells were originally thought to only recognise peptide antigens, it is now evident that T cells are also able to recognise nonpeptide antigens. Recognition of non-peptide antigens confers protection against pathogens that have cell surfaces that are highly functionalised with carbohydrate moieties, such as glycolipids, glycopeptides and polysaccharides. Specifically, glycosphingolipids (GSLs) can activate invariant Natural Killer T (iNKT) cells via their T cell receptor (TCR) when presented by the CD1d molecule found on the surface of DCs. α-Galactosylceramide (α-GalCer 1, Figure 1), a synthetic analogue of a GSL extracted from the marine sponge Agelas mauritianus, was discovered to be a potent stimulator of iNKT cells when presented by CD1d. α-GalCer is currently being used in clinical trials as an adjuvant to boost the activation of immune cells during cancer immunotherapy. Although the molecular interaction of α-GalCer with CD1d and iNKT cells is well established, it is not fully understood how the glycolipid interacts with different subsets of DCs. Greater understanding of the fate of the glycolipid during cancer immunotherapy will provide crucial information on how the current therapy can be improved. In this thesis, the design and synthesis of two fluorescent α-GalCer probes, dansyl-α-GalCer (2, Figure 1) and BODIPY-α-GalCer (3, Figure 1) is reported. Dansyl-α-GalCer was able to activate DCs and iNKT cells in a similar fashion to the parent glycolipid α-GalCer. Its activity was CD1d-dependent and DCs that have taken up α-GalCer in vitro can be detected by flow cytometry. Unfortunately, the fluorescence of dansyl-α-GalCer was too weak to be detected by fluorescent microscopy due to photobleaching of the dye. Accordingly, another α-GalCer probe bearing a brighter fluorescent group, BODIPY, was synthesised. The α-GalCer probes were made via two synthetic strategies and the benefits and shortcomings of each synthetic route are discussed. Isoglobotrihexosylceramide (4, iGb3, Figure 2) is another GSL known to activate iNKT cells. Like α-GalCer, it is presented by DCs in the context of a CD1d molecule. iGb3 contains a sphingosine lipid backbone β-linked to a trisaccharide head group, which is in contrast to the α-linked phytosphingosine lipid found on α-GalCer. Despite the structural difference, iGb3 can stimulate iNKT cells, though to a lesser extent than α-GalCer. The intriguing activity of iGb3 provides a platform to further investigate the molecular interactions between CD1d, glycolipid and TCR of iNKT cell. The crystal structure of iGb3 in complex with mouse CD1d and TCR of mouse iNKT cell show compelling evidence that the terminal galactose moiety is crucial for the observed activity and this is attributed to the hydrogen bond between the 6´´´-OH and Thr159 on the CD1d. To unambiguously determine the importance of the hydrogen bond conferred by 6´´´-OH, 6´´´-deoxy-iGb3-sphingosine (5, Figure 2) was synthesised. 6´´´-deoxy-iGb3-sphinganine 6 was also synthesised to study the role of the double bond on the sphingosine backbone. A novel synthetic route for the synthesis of iGb3 analogues was established. This allowed for the expedient synthesis of 6´´´-deoxy-iGb3 derivatives that will subsequently be crystallised with CD1d and TCR of iNKT cell, to provide further insight into the structural requirements of β-linked GSLs. Studies have also revealed that the length and saturation of the N-acyl chain of GSLs greatly influences their activity. It is speculated that varying the length of the acyl chain affects the processing and loading of the glycolipid onto CD1d and also TCR binding affinity. To this end, the syntheses of a series of acyl chain analogues of iGb3, including the shorter chain homologue C12:0 7 (Figure 3) and the unsaturated C20:2 derivative 8 are reported. A divergent synthetic route was employed, whereby a common intermediate from the synthesis of 6´´´-deoxyiGb3 was used. This allowed for efficient syntheses of the acyl chain analogues that will facilitate a greater understanding of the structure-activity relationships. Taken together, the GSLs synthesised provide crucial insight into how they modulate the immune system and will guide future optimisation of cancer immunotherapy regimes.