Intracellular O-GlcNAc glycosylation

Several cytoplasmic and nuclear proteins carry O-linked GlcNAc on specific serine and threonine residues, thereby defining another class of glycosylation sometimes referred to as “O-GlcNAc-ylation”. Gerald Hart discovered O-GlcNAc-ylation by demonstrating that presence of glycoproteins in the nucleus and cytosol of lymphocytes. While O-GlcNAc-ylation remained little known for about 20 years, the ubiquity and importance of this type of glycosylation have been clearly established.

FIG: INTRACELLULAR O-GLCNAC

O-GlcNAc is transferred to and removed from proteins dynamically, in a way similar to protein phosphorylation/dephosphorylation. Although hundreds of proteins are O-GlcNAc-ylated, a single O-GlcNAc-transferase (OGT) enzyme mediates the transfer of O-GlcNAc. The OGT gene is localized on the X-chromosome; the deletion of the OGT gene is lethal in mammals and even cultured cells require an intact OGT gene to survive. The OGT gene yields three isoforms by alternative splicing: the nuclear and cytoplasmic OGT (ncOGT) of 116 kDa, the mitochondrial OGT (mOGT) of 103 kDa and the short OGT (sOGT) of 70 kDa. The three isoforms contain the same catalytic domain but a different number of tetratrico peptide repeats (TPR) at their N-terminus. These TRPs are involved in the recognition of the multiple substrates of OGT. The crystal structure of the complete human OGT has been recently described. The two lobes of the catalytic domain (N-Cat and C-Cat) have Rossmann folds typical of GT-B glycosyltransferases. The intervening domain between the two lobes of the catalytic domain is only found in metazoans, but no catalytic contribution could be deduced for this domain. The co-crystalization of the enzyme with UDP and a peptide acceptor suggests that OGT first binds UDP-GlcNAc and then the peptide substrate.

FIG: STRUCTURE OF OGT

The removal of O-GlcNAc from proteins is also mediated by a single O-GlcNAcase (OGA) enzyme. The human OGA gene maps to chromosome 10. Two isoforms of the OGA protein are found in cells, a long form of 103 kDa and a short form of 76 kDa, which lacks the acetyltransferase-like domain at the C-terminus. The short form is mainly found in the nucleus. Both isoforms contain a hyaluronidase domain at their N-terminus and both isoforms are active of O-GlcNAcase. The disruption of the mouse OGA gene leads to postnatal lethality following embryonic developmental delay. The impact of OGA activity was also be studied in cell culture using inhibitors like O-(2-acetamido-2-deoxy-D-glucopyranosylidene) amino N-phenyl carbamate (PUGNAc). Inhibition of OGA in vitro leads to over O-GlcNAc-ylation and to multiple defects ranging from deficient proliferation, decreased protein translation to increased apoptosis. Surprisingly, the deletions of the OGT and OGA genes in the nematode Caenorhabditis elegans are not lethal but impair the energy metabolism of the worm.

FIG: STRUCTURE OF OGA

The extent of O-GlcNAc on proteins is proportional to intracellular levels of UDP-GlcNAc. The formation of UDP-GlcNAc itself is regulated by the rate-limiting glutamine fructose-6-phosphate amidotransferase (GFAT) enzyme, which is fuelled by nutrients like Glc, glutamine, acetyl-CoA, uridine and ATP. Accordingly, the levels of UDP-GlcNAc, which average 5 mM, reflect the energy status of cells. High levels of UDP-GlcNAc do not only increase O-GlcNAc-ylation but also the branching of N-glycans as discussed in the chapter dedicated to Golgi glycosylation.

FIG: O-GLCNAC AND ENERGY SUPPLY

The presence of O-GlcNAc can be surveyed through various proteomic approaches. Such studies rely on the detection of O-GlcNAc by mass-spectrometry. A new fragmentation technique called electron-transfer dissociation (ETD) is turning out as a major advantage in mapping the O-GlcNAc-ylated sites since it allows for the first time the generation of ions keeping the glycan attached to the peptide backbone. Besides ETD, O-GlcNAc-ylated sites can be tagged using various techniques. For example, cells or animals can be supplemented with azido-GlcNAc (GlcNAz), which in metabolized into UDP-GlcNAz. After transfer to proteins, GlcNAz can be modified by Staudiger ligation. Alternatively, O-GlcNAc-ylated sites can be mapped using b-elimination and Michael addition reaction (BEMAD), although this tagging method does not discriminate between O-GlcNAc-ylated and phosphorylated sites. O-GlcNAc can also be modified by addition of a reactive Gal residue, e.g. keto-Gal, using an engineered form of the β1-4 Gal-transferase-1 enzyme. Since O-GlcNAc-ylation is a dynamic process, it is difficult to generate an exhaustive catalogue of modified proteins. The majority of O-GlcNAc-ylated proteins are involved in gene regulation and encompass polymerases, transcription and translation factors. Such proteins include the nuclear pore complex, c-Myc, p53, eNOS, pyruvate kinase, UDP-Glc pyrophosphorylase, E-cadherin, myosin and tubulin. The addition of O-GlcNAc affects the activity of the target proteins, such as by decreasing enzymatic activity or increasing the DNA binding of transcription factors. In several proteins, O-GlcNAc-ylation and phosphorylation compete for the same acceptor sites, meaning that the attribution of specific functions to one of the other type of protein modification may be difficult to establish.

As mentioned above, O-GlcNAc-ylation is regulated by the intracellular concentration of UDP-GlcNAc, which is itself affected by the energy metabolism. Accordingly, altered O-GlcNAc-lyation leads deregulation of the insulin and leptin pathways and thereby to diabetes and hyperlipidemia. Transgenic mouse models featuring increased GFAT or OGT activity outlined the importance of the UDP-GlcNAc pathway in metabolically active tissues. Increased GFAT activity leads to higher UDP-GlcNAc and O-GlcNAc-ylation. Altogether, these mouse models demonstrate that increased O-GlcNAc-ylation desensitizes insulin response, which causes hyperglycemia and diabetes. O-GlcNAc-ylation is often considered to act as general nutrient signal, which fine-tunes various cellular processes, such as cell cycle progression, protein translation or signaling cascades based on nutrient availability.