What Determines Fringe Specificity for a Notch EGF?

Abstract

There are four transmembrane Notch (N) signaling receptors in humans involved in developmental, cell differentiation, and cell regulatory events, and dysregulation of N signaling leads to pathologies such as developmental disorders or cancer. The extracellular domain (ECD) of the N receptors are composed of 29 to 36 tandem Epidermal Growth Factor‐like repeats (EGFs) and N signaling occurs when EGFs of the ECD interact with a membrane tethered N ligand on an adjacent cell. These EGFs are heavily modified with O‐glycans that directly modulate interaction with, and activation by, the N ligands. PROTEIN O‐FUCOSYLTRANSFERASE 1 is an ER localized enzyme that adds an O‐linked fucose to Serines or Threonines within a properly folded EGF containing the C2‐X‐X‐X‐X‐ S/T‐C3 consensus sequence. The O‐fucose affects binding affinity for ligand, as well as N signaling levels. In the Golgi, O‐fucosylated EGFs are exposed to a family of β3‐N‐acetylglucosaminyltransferases named Fringes which can elongate the O‐fucose to a GlcNAc‐β1‐3Fuc‐O disaccharide. Mirroring effects seen after addition of fucose, there are additional effects on ligand binding affinity and signaling after Fringe elongation. There are three Fringe enzymes termed LUNATIC (Lfng), MANIC (Mfng), and RADICAL FRINGE (Rfng). We have mapped the extent of Fringe elongation across N1 and N2 EGFs and we find that generally speaking, Lfng modifies 50% or more of fucosylated EGFs, Mfng modifies a sub‐set of the Lfng sites, and Rfng modifies an even smaller subset of the Lfng sites. Additionally, despite high sequence identity over the entirety of the tandem EGFs in the ECD, the pattern of Fringe‐modified EGFs is different between N1 and N2. Prior work suggested that while primary amino acid sequence of a particular EGF can determine the level of elongation by Fringe enzymes, it may not be sufficient to explain all specificity differences. Our mapping data has shown that some corresponding EGFs in N1 and N2, such as EGF5 and EGF12, show different levels of Fringe modification, despite high sequence identity. These data suggested that the position of the EGF in the linear sequence of tandem EGFs could be another determinant of Fringe enzyme specificity. To test the degree to which linear position in a N receptor or EGF amino acid sequence determines specificity, we swapped corresponding EGFs between N1 and N2 and determined whether the propensity for elongation lies within a given EGF sequence. We now have evidence that EGF amino acid sequence is the primary driver of Fringe elongation potential. We are now testing whether in vitro specificity correlates with in vivo specificity for a given EGF in well‐established Fringe enzyme assays. We will take what we learn from this work, and test whether we can eliminate Fringe elongation at specific EGFs without eliminating the O‐fucose glycan. This will allow a highly refined version of current experiments testing the importance of Fringe elongation at individual EGFs, where glycosylation is completely eliminated by mutating the glycan modified amino acid.Support or Funding InformationThis work is supported by NIH grant GM061126.