LFNG

LFNG (Lunatic Fringe) is a Golgi-resident fucose-specific β1,3-N-acetylglucosaminyltransferase that modifies O-fucose residues on EGF-like repeats of Notch receptors and Notch ligands (DLL1, DLL3), thereby modulating Notch signaling output in both signal-receiving and signal-sending cells (PMID:19061953, PMID:35429490). In the presomitic mesoderm, LFNG expression oscillates as part of the segmentation clock, where its dosage tunes the clock period, its N-terminal processing and secretion are required for normal clock function, and post-transcriptional regulation via miR-125a-5p-mediated transcript turnover contributes to oscillatory dynamics (PMID:24560643, PMID:26811377, PMID:23484856). Beyond somitogenesis, LFNG acts downstream of androgen receptor signaling in Sertoli cells to modulate Notch-dependent GDNF production for spermatogonial stem cell maintenance, and marks a pancreatic centroacinar cell population where it is required for Notch3-dependent tumor initiation upon oncogenic Kras activation (PMID:39407201, PMID:39548190). Loss-of-function mutations in the LFNG catalytic domain, including variants p.D201N and p.R360C, abolish glycosyltransferase activity and cause spondylocostal dysostosis type 3 in humans (PMID:30531807, PMID:38565611).

1. 2008 High

   Establishing that LFNG encodes a fucose-specific β1,3-N-acetylglucosaminyltransferase that glycosylates Notch receptors resolved the molecular basis by which Fringe proteins modulate Notch signaling and explained the somitogenesis defects observed in Lfng mutant mice and humans.

   Evidence Genetic studies in mouse and human combined with biochemical characterization of enzymatic specificity

   PMID:19061953 PMID:19899223

   Open questions at the time

   • Whether LFNG modifies Notch ligands in addition to receptors was unknown
   • How LFNG protein processing and turnover contribute to clock function was unresolved
   • No structural model of LFNG-substrate interaction existed
2. 2013 High

   Demonstrating that miR-125a-5p targets the Lfng 3′ UTR to destabilize its transcript in vivo revealed a post-transcriptional mechanism contributing to the oscillatory expression pattern required by the segmentation clock.

   Evidence In vivo miRNA-target blocking in chick embryos with 3′ UTR reporter assays and clock gene expression analysis

   PMID:23484856

   Open questions at the time

   • Whether miR-125a-5p is essential or redundant in mammalian somitogenesis was untested
   • Other post-transcriptional regulators of Lfng oscillation were not characterized
3. 2014 Medium

   An allelic series of Lfng hypomorphic mice showed that the amplitude of oscillatory LFNG expression quantitatively modulates segmentation clock period, establishing LFNG dosage as a tunable parameter of the clock rather than a simple on/off switch.

   Evidence Quantitative clock period measurements in Lfng hypomorphic allelic series in mouse

   PMID:24560643

   Open questions at the time

   • The mechanism by which LFNG dosage translates to clock period change was not molecularly resolved
   • Differential effects on anterior versus posterior body axis were descriptive
4. 2016 High

   A knock-in replacing LFNG's signal sequence with a Golgi-retention motif showed that LFNG secretion and normal N-terminal processing are required for clock function, and that intracellular accumulation of LFNG creates dominant gain-of-function phenotypes including Hes7 stabilization.

   Evidence Lfng^RLFNG knock-in mouse with in vivo clock gene expression analysis

   PMID:26811377

   Open questions at the time

   • Whether secreted LFNG has extracellular signaling functions or secretion simply controls intracellular levels was unclear
   • The molecular basis by which retained LFNG stabilizes Hes7 was not defined
5. 2017 Medium

   Mutation of miR-125 binding sites in the mouse Lfng 3′ UTR confirmed their role in transcript turnover in the presomitic mesoderm, but germline loss of miR-125a-5p did not disrupt somitogenesis, indicating compensatory mechanisms exist in mammals.

   Evidence 3′ UTR reporter transgenes and germline miR-125a-5p knockout in mouse

   PMID:28710840

   Open questions at the time

   • Identity of compensatory miRNAs or RNA-binding proteins was not determined
   • Partial discordance with chick data leaves the essential post-transcriptional regulators in mammals unresolved
6. 2018 Medium

   Identification of the p.D201N variant in the DxD catalytic motif as enzymatically dead, causing spondylocostal dysostosis type 3, directly linked LFNG glycosyltransferase activity to human congenital vertebral segmentation defects.

   Evidence In vitro glycosyltransferase enzyme assay on active-site mutant; compound heterozygous patient sequencing

   PMID:30531807

   Open questions at the time

   • Only a single patient family was reported
   • No structural basis for why D201N abolishes activity was provided
7. 2018 Medium

   Demonstration that TGFBR2 signaling upregulates LFNG to increase GlcNAc incorporation into Notch1 in colorectal cancer cells revealed an upstream regulatory input linking TGF-β pathway to Notch glycosylation status.

   Evidence TGFBR2-reconstituted HCT116 cells with radiolabeled GlcNAc incorporation and Notch1 immunoprecipitation

   PMID:27156840

   Open questions at the time

   • Whether this TGF-β–LFNG–Notch axis operates in non-cancer contexts was not tested
   • Functional consequences for Notch signaling output were not directly measured
8. 2022 High

   Mass spectrometry and genetic epistasis showed that LFNG glycosylates DLL1 and DLL3 ligands in signal-sending cells, and that DLL3 loss is epistatic to LFNG loss in the segmentation clock, establishing that LFNG functions on both receptor and ligand sides of Notch signaling.

   Evidence Mass spectrometry of glycosylated EGF repeats, Dll3/Lfng double-mutant mouse, cell-based Notch signaling assays

   PMID:35429490

   Open questions at the time

   • Which specific EGF repeats on DLL1/DLL3 are functionally critical targets of LFNG glycosylation was not resolved
   • Whether LFNG modification of ligands matters outside the segmentation clock context was untested
9. 2024 Medium

   ChIP-seq demonstrated direct androgen receptor binding at the Lfng promoter in Sertoli cells, connecting androgen signaling to LFNG-dependent Notch modification and downstream GDNF production for spermatogonial stem cell maintenance — the first link between hormone-regulated LFNG transcription and germ cell niche function.

   Evidence ChIP-seq, RNA-seq, enzalutamide androgen blockade in Sertoli cells

   PMID:39407201

   Open questions at the time

   • Whether LFNG's glycosyltransferase activity is directly required for GDNF regulation was not formally tested
   • In vivo validation with Sertoli-specific Lfng deletion was not performed
10. 2024 Medium

    Lineage-tracing and conditional deletion revealed that Lfng marks pancreatic centroacinar cells and is required for Notch3-dependent tumor initiation upon Kras/p53 activation, extending LFNG's role from development to cancer initiation.

    Evidence Lineage-specific Cre-mediated Lfng deletion in mouse PDAC models with Kras/p53 mutations

    PMID:39548190

    Open questions at the time

    • Whether LFNG's enzymatic activity on Notch3 is the mechanistic basis for tumor initiation was not directly shown
    • Whether this role is unique to centroacinar cells or generalizable to other pancreatic cell types was not addressed

• A structural understanding of how LFNG recognizes and glycosylates specific EGF repeats on Notch receptors versus ligands, and the mechanism by which LFNG's intracellular trafficking and secretion regulate its clock function, remain unresolved.
• No crystal or cryo-EM structure of LFNG in complex with a substrate EGF repeat exists
• The functional role of secreted LFNG (extracellular versus disposal) is unknown
• The identity of compensatory post-transcriptional regulators in the mammalian segmentation clock is undefined