How much sugar the cell has decides how much DOT1L there is - a metabolic dial on the leukaemia's mark
also called DOT1L O-GlcNAcylation · DOT1L S1511 O-GlcNAc · UBE3C degrades DOT1L · hexosamine pathway and H3K79
DOT1L carries a sugar tag at one spot near its tail, and that tag is added when glucose is plentiful. The tag stops a disposal enzyme from finding DOT1L, so well-fed leukaemia cells keep more DOT1L, make more of the H3K79 mark, and grow faster - and starving that pathway makes them more sensitive to the DOT1L drug.
The same thing, for a scientist
DOT1L protein abundance is set by extracellular glucose through the **hexosamine biosynthetic pathway (HBP)** [E1]. DOT1L is O-GlcNAcylated at the evolutionarily conserved residue **S1511** in its C terminus. **UBE3C** is identified as a DOT1L E3 ubiquitin ligase that promotes DOT1L degradation, and its interaction with DOT1L is **susceptible to O-GlcNAcylation** - so the sugar modification blocks the ligase and stabilises the enzyme. Consequently HBP flux raises H3K79 methylation and the expression of *HOXA9* and *MEIS1*, promoting proliferation in MLL-fusion leukaemia, and **inhibiting HBP or OGT increases cellular sensitivity to a DOT1L inhibitor**. Three things make this more than a metabolic curiosity for this project. It is a **combination rationale** with an orthogonal mechanism - lowering the enzyme's abundance while inhibiting what remains. It identifies an **endogenous degradation route** for DOT1L, which is directly relevant to the PROTAC programme, since a native ligase already exists and the cell already uses ubiquitin-proteasome turnover of DOT1L as a regulated step. And it means **H3K79 methylation is metabolically conditioned**, so the same leukaemia may carry different amounts of the mark in the glucose-rich marrow and in blood, which is a caveat on every pharmacodynamic measurement made in one compartment.
Why UBE3C is the most interesting name in this paper
A PROTAC works by borrowing an E3 ligase that has no natural reason to touch the target. The published DOT1L PROTACs recruit VHL 2. But DOT1L already has a native E3 ligase, UBE3C, whose engagement is switched by a metabolic signal 1.
That suggests a different chemical strategy from the standard one: rather than building a bivalent molecule to a borrowed ligase, remove the block on the native route - inhibit OGT, or otherwise strip the S1511 O-GlcNAc - and let UBE3C do the degradation. The published data already contain the phenotype this predicts: inhibiting HBP or OGT increases sensitivity to a DOT1L inhibitor 1.
The obvious objection is selectivity. OGT is a single enzyme responsible for O-GlcNAcylation of thousands of proteins, so an OGT inhibitor is about as non-selective an intervention as exists in this space, and any anti-leukaemic effect would have many possible explanations. That is why the mechanism is recorded here as a rationale worth testing with a UBE3C-directed or S1511-directed reagent, and not as a therapeutic proposal.
The pharmacodynamic caveat it creates
If DOT1L abundance tracks glucose availability through HBP, then the amount of H3K79 methylation in a leukaemic cell is a function of its metabolic environment. A blast in the glucose-rich marrow and the same blast in peripheral blood, or in a culture dish at 25 mM glucose, are not guaranteed to carry the same mark level.
This matters because the pinometostat pharmacodynamic endpoint was H3K79 methylation reduction measured in patient material 3. A percentage reduction computed against a baseline that is itself metabolically variable is a noisier quantity than it appears, and the direction of the bias is unknown. It is one more reason the beat's proposed measurement uses calibrated absolute occupancy rather than fold-change.
It also gives a mechanistic reading to a clinical observation this graph does not otherwise explain: variability in pharmacodynamic response between patients on the same dose.
| mechanism class | post-translational regulation of enzyme abundance |
|---|---|
| protein | gene-dot1l |
| modification | O-GlcNAcylation at S1511 (C-terminal, evolutionarily conserved) |
| upstream pathway | hexosamine biosynthetic pathway, driven by extracellular glucose; writer is OGT |
| native e3 ligase | UBE3C |
| logic | O-GlcNAc at S1511 blocks the UBE3C-DOT1L interaction, so high HBP flux stabilises DOT1L, raising H3K79 methylation and HOXA9/MEIS1 expression |
| pharmacological consequence | inhibiting HBP or OGT sensitises MLL-fusion leukaemia cells to a DOT1L inhibitor |
| relevance to degraders | DOT1L has a native ubiquitin-proteasome degradation route that a PROTAC does not use; the published degraders recruit VHL instead |
| pharmacodynamic caveat | H3K79 methylation level is metabolically conditioned, so fold-change endpoints have a variable baseline |
| replication status | single report; no independent confirmation of S1511 or UBE3C found in this beat's reading |
- OGT is responsible for O-GlcNAcylation of thousands of proteins, so any HBP or OGT intervention has an enormous off-target surface and an anti-leukaemic effect would be very hard to attribute to DOT1L.
- A metabolically conditioned target abundance means the effective dose of any DOT1L-directed agent varies with the tumour's microenvironment, and the glucose-rich marrow niche is the compartment where abundance - and therefore required dose - would be highest.
- Sensitising leukaemia cells by starving a central metabolic pathway will sensitise normal proliferating tissue by the same mechanism, and HBP flux is not leukaemia-specific.
- Has the S1511 O-GlcNAc site or the UBE3C assignment been independently replicated? Both are single-report findings on which a combination rationale would rest.
- Would a UBE3C-recruiting DOT1L PROTAC outperform a VHL-recruiting one, given that UBE3C is the native ligase and is already competent for this substrate?
- Does H3K79me2 in KMT2A-rearranged blasts differ measurably between marrow and peripheral blood, as the metabolic-conditioning model predicts? This is checkable in existing paired patient samples.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
DOT1L protein stability is regulated by extracellular glucose through the hexosamine biosynthetic pathway; DOT1L is O-GlcNAcylated at the evolutionarily conserved S1511 in its C terminus; UBE3C is a DOT1L E3 ubiquitin ligase promoting DOT1L degradation whose interaction with DOT1L is susceptible to O-GlcNAcylation; HBP flux enhances H3K79 methylation and expression of DOT1L target genes including HOXA9 and MEIS1, promoting proliferation in MLL-fusion leukaemia; and inhibiting HBP or O-GlcNAc transferase increases cellular sensitivity to a DOT1L inhibitor.
DOT1L O-GlcNAcylation promotes its protein stability and MLL-fusion leukemia cell proliferation. · Song T, Zou Q, Yan Y, et al. · Cell Rep 2021What this evidence cannot show
Abstract read from the eutils record; full text not read. Cell-line work; no in vivo or primary-sample data are described in the abstract, and the MLL fusions used are not named. OGT and HBP inhibitors are broadly non-selective, so 'increases sensitivity to a DOT1L inhibitor' has many possible routes besides DOT1L abundance. The sensitisation is not quantified in the abstract. Single report; this beat's reading found no independent replication of the S1511 site or the UBE3C assignment.
Evidence that argues against this (1)
- If DOT1L's leukaemic contribution includes catalysis-independent scaffolding, then lowering DOT1L abundance and inhibiting DOT1L catalysis are not merely additive on the same axis - they are different perturbations, and the observed sensitisation could reflect loss of the scaffold rather than loss of enzyme. The abstract's framing assumes a single axis. Cao K et al. 2020
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The first-in-class DOT1L PROTAC degrader requires binding to the E3 ligase VHL and degrades DOT1L through the ubiquitin-proteasome system without affecting DOT1L mRNA.
Discovery of the first-in-class DOT1L PROTAC degrader. · Yim H, Sun R, Xu Z, et al. · Eur J Med Chem 2025What this evidence cannot show
Abstract read from the eutils record; full text not read. Cited here to contrast a borrowed E3 ligase with the native one; the paper makes no reference to UBE3C or to O-GlcNAcylation, and the comparison between the two routes is this graph's inference.
Evidence that argues against this (1)
- Borrowing VHL rather than using the native ligase may be the correct design choice: a native ligase whose engagement is metabolically switched would make degradation depend on the tumour's glucose environment, which is exactly the variability a drug wants to avoid. Song T et al. 2021
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The pinometostat adult phase 1 used reduction of H3K79 methylation in patient material as its pharmacodynamic endpoint.
The DOT1L inhibitor pinometostat reduces H3K79 methylation and has modest clinical activity in adult acute leukemia. · Stein EM, Garcia-Manero G, Rizzieri DA, et al. · Blood 2018What this evidence cannot show
Abstract read from the eutils record; full text not read; the tissue, assay and methylation state used are not stated there. The argument that a metabolically variable baseline adds noise to a fold-change endpoint is an inference of this graph, not a reported finding of the trial.
Evidence that argues against this (1)
- A within-patient paired design, comparing post-treatment to pre-treatment marrow in the same person, controls for much of the metabolic variability this entry raises, so the practical size of the problem may be small. It would matter more for cross-patient comparison than for the pharmacodynamic endpoint as reported. Richter WF et al. 2021
O-GlcNAc at S1511 blocks UBE3C and stabilises the protein
hexosamine flux raises H3K79 methylation via DOT1L abundance
DOT1L has a native E3 (UBE3C) that the VHL-recruiting PROTAC does not use
OGT is the enzyme in this mechanism
bounded failure; the one topic-exact GEO hit belongs to a different paper