Abstract
Notch is an essential cell surface receptor whose proper functioning is critical during development in metazoans. Notch can be modified with two unique forms of O-linked glycosylation: O-fucose and O-glucose. Protein O-fucosyltransferase1 (Pofut1, OFUT1 in flies) and Protein O-glucosyltransferase (Poglut, Rumi in flies) are the glycosyltransferases that add O-fucose and O-glucose monosaccharides to Notch, respectively. Loss of Pofut1/Ofut1 or Poglut/rumi in mice or flies phenocopies Notch-null phenotypes, suggesting both forms of O-glycosylation are essential for Notch function. To gain insights into how these essential modifications impart functionality to the receptor, it is vital to identify sites of O-glycosylation, and structures of glycans at each of these sites. The extracellular domain (ECD) of mouse Notch1 contains 36-tandem epidermal growth factor-like (EGF) repeats, 16 of which contain the consensus sequence for O-glucosylation (C<super>1</super>XSXPC<super>2</super>). Using nanoLC-ESI-MS/MS to determine occupancy at these 16 sites, I found most predicted sites were efficiently modified, though efficiency of modifying O-glucose sites was cell type-dependent, and some sites were under-glucosylated. O-Glucose was elongated to Xyl-Α1,3-Xyl-Α1,3-Glc at all sites. In addition, a novel non-traditional consensus site was found to be efficiently O-glucosylated at EGF 9, leading to a revision of the consensus sequence for O-glucosylation to allow Alanine N-terminal to Cysteine 2: C<super>1</super>X<underline>S</underline>X(P/A)C<super>2</super>. The ECD of Drosophila Notch (dN) also contains 36 EGF repeats, the majority of which contain consensus sequences for O-fucosylation (C<super>2</super>XXXX(<underline>S/T</underline>)C<super>3</super>) and/or O-glucosylation (C1X<underline>S</underline>X(P/A)C2). O-Fucose moieties can be further extended to a disaccharide by Fringe, a Β3N-acetylglucosaminyltransferase and known modulator of Notch. O-Glucose moieties can be elongated by a xylosyltransferase to a disaccharide. I have mapped both O-fucose and O-glucose glycans on the ECD of dN over-expressed in S2 cells co-transfected with or without Fringe using nanoLC-ESI-MS/MS. I have also compared relative levels of O-fucosylation and Fringe elongation, as well as O-glucosylation and subsequent xylosylation, at specific sites using Multiple Reaction Monitoring (MRM), a semi-quantitative method that allows relative quantitation of glycoforms at a specific site. My results suggest the predicted glycosylation sites are very efficiently modified with the O-glycan monosaccharides and that extent of elongation of either O-linked modification is substoichiometric and site-specific.
