AbstractModification of proteins on serine and threonine residues with O-linked N-acetyl-β-Dglucosamine (O-GlcNAc) was first observed on the surface of lymphocytes in 1984. It was then found that over 1,000 nucleocytoplasmic proteins are modified by O-GlcNAc (O-GlcNAcylation). Interestingly, O-GlcNAc homeostasis is regulated by a single pair of enzymes: O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA). O-GlcNAcylation has been found to modulate enzyme catalytic activity, protein stability, protein–protein interaction and protein subcellular localisation, and these could lead to regulation of cell proliferation and differentiation. Mutations in OGT that resulted in compromised OGT stability or catalytic activity, have been identified as causing a new syndromic form of intellectual disability classified as OGT-linked Congenital Disorder of Glycosylation (OGT-CDG). O-GlcNAcylation has been shown to provide beneficial effects on cells in coping with environmental stresses, such as hypoxia and glucose deprivation, and on the development of many cancers where O-GlcNAc levels are often up-regulated.
The action of OGT relies on the steady supply of UDP-GlcNAc, the end product of the hexosamine biosynthetic pathway, synthesized by UDP-N-acetylhexosamine pyrophosphorylase (UAP1). Therefore, I considered whether UAP1 gene mutations would phenocopy OGT-CDG mutations. A de novo heterozygous mutation of UAP1 (NM_001324114:c.G685A:p.A229T, DDD entry: https://decipher.sanger.ac.uk/search/ddd-research-variants/results?q=uap1), which encodes a missense variant UAP1A229T, was reported in a patient with intellectual disability. The aim of the second chapter of this thesis is to investigate the pathogenic potential of the UAP1 A229T mutation. I include the following findings in Chapter 2. First, clinical bioinformatics tools predict that the mutation is likely to be pathogenic, whilst biochemical analysis with recombinant AGX1A229T reveals that the A229T mutation causes a reduction of protein thermal stability. Compared to wild type AGX1, AGX1A229T has lower activity in producing UDP-GlcNAc. X-ray crystallographic structural characterization demonstrates that the A229T mutation lies proximal to the active site. The mutation induces local structural shift which weakens the hydrogen bond network connecting the N-terminal and catalytic domain, leading to conformational changes of the N-terminal domain that explains changes in catalytic activity. Together, these in vitro data suggest that the UAP1A229T missense mutation could contribute to the patient phenotype.
UDP-N-acetylhexosamine pyrophosphorylase-like protein 1 (UAP1L1) is a paralogue of UAP1, sharing 59% sequence identity with UAP1. Recently, UAP1L1 was reported as a tumour promotor of both liver and gastric cancer. UAP1L1 has been found to interact with the catalytic domain of OGT in human carcinoma HepG2 cells and with CDK6 in gastric cancer cells. Genetic knockdown of UAP1L1 in HepG2 cells inhibited O-GlcNAcylation, cell proliferation and tumourigenesis. Recombinant UAP1L1 was found to not catalyse the formation of UDP-GlcNAc, and the mechanism by which UAP1L1 regulates O-GlcNAcylation was left unresolved. The aim of the third chapter of this thesis is to investigate the function of UAP1L1 using tools ranging from bioinformatics, biochemistry and structural and molecular biology. I show that UAP1L1 and its active site are conserved in vertebrates. Recombinant UAP1L1 forms multimers in solution, and monomeric UAP1L1 is stabilized by UTP. UAP1L1 multimerization is majorly maintained through the N-terminal domain. UAP1L1 monomer or low molecular weight oligomer, which is the expressed form of UAP1L1 11 in HepG2 cells, is enzymatically active in synthesizing UDP-GlcNAc in vitro. Importantly, the crystal structure of a UAP1L1-UAP1 chimera sheds lights on the catalytic basis of UAP1L1 and reveals a disulfide in the active site regulating UAP1L1 thermal stability. Neither knockout nor overexpression of UAP1L1 modulates OGlcNAcylation or cell proliferation in mouse embryonic stem cells, suggesting that UAP1L1 is nonessential for signaling/metabolic regulation in these cells. This knowledge could inspire the development of chemical tools for dissecting the function of UAP1L1 in in vivo models such as HepG2 cells and gastric cancer cells.
Genetic knockdown of UAP1L1 inhibits cell proliferation and tumour growth in vivo. However, due to ablation of whole UAP1L1 protein, it is unknown whether these effects are resulted by the loss of catalytic or scaffolding activity of UAP1L1, and this issue could be resolved with chemical tools that specifically inhibit the catalytic activity of UAP1L1. The aim of the last chapter of this thesis is to discover chemical tools that specifically inhibit UAP1L1 for dissecting the function of UAP1L1 in in vivo models. Auranofin was identified as a selective inhibitor against human UAP1L1 over UAP1 and fungal UAP1 homologues from a high-throughput screening campaign. Mechanistically, auranofin inhibits UAP1L1 through selectively and covalently modifying a UAP1L1-specific cysteine C231, which suggests that UAP1L1 C231 is a site that could be used as a handle for the future development of specific UAP1L1 inhibitors.
|Date of Award
|China Scholarship Council
|Daan van Aalten (Supervisor)