MS2133 - the first molecule to degrade DOT1L rather than inhibit it
also called MS2133 · compound 13 (Yim 2025) · first-in-class DOT1L PROTAC
MS2133 is the first reported molecule that hands DOT1L to the cell's protein-shredder instead of blocking its active site. It is a laboratory tool rather than a medicine, but it is the reagent that lets anyone ask whether removing DOT1L entirely does more than switching its chemistry off.
The same thing, for a scientist
MS2133 (compound 13) is the first-in-class DOT1L PROTAC degrader, reported 2025 [E1]. It induces concentration- and time-dependent DOT1L degradation without changing DOT1L mRNA; degradation requires binding to both DOT1L and the E3 ligase **VHL** and proceeds through the ubiquitin-proteasome system; it is selective for DOT1L over other methyltransferases and inhibits growth of MLL-rearranged leukaemia cells while reported as non-toxic to normal cells. The authors' framing is careful and this entry adopts it: **a chemical biology tool first, a potential therapeutic second**. Its value to this project is not as a drug candidate but as the reagent that makes the central question of this beat experimentally addressable - whether the pinometostat result reflects a refuted target or an unreachable function. No DC50, no in vivo data and no quantitative normal-cell toxicity appear in the abstract, so this entry is deliberately thinner than `drug-dot1l808`, which reports both.
Why a tool compound gets its own entry
Because the graph's granularity rule is one entry per thing, and because the two published DOT1L degrader series are genuinely different objects: MS2133 is a VHL-recruiting first-in-class report with no potency figure or animal data in its abstract 1, while DOT1L705/DOT1L808 are a later series with stated DC50 values, menin-inhibitor cross-resistance data and an in vivo regression 2. Collapsing them into one 'DOT1L degraders' entry would lose the fact that the field's first degrader was a tool and the second was a candidate.
The VHL dependence is worth recording explicitly. It means MS2133's activity is contingent on VHL expression in the cell being tested, which varies across lineages and is lost in some tumours - so a negative result with MS2133 in a given model is not by itself evidence that DOT1L degradation does not work there.
| target | gene-dot1l |
|---|---|
| modality | PROTAC degrader |
| e3 ligase | VHL |
| status | preclinical chemical-biology tool, first reported 2025; no trial registered |
| dc50 | not stated in the abstract |
| in vivo | none reported |
| role in this graph | reagent that makes the refuted-target versus wrong-modality question testable |
mechanism confirmed by
[ "VHL-binding requirement", "proteasome dependence", "unchanged DOT1L mRNA" ]
trials
[]
key ae
[ "none reported" ]
resistance
[ "VHL loss or downregulation" ]
- VHL-dependence means the compound reports nothing in VHL-low cells, so an apparently resistant model may simply be a model where the degrader cannot work.
- Degrading DOT1L removes the catalytic function that adult marrow needs, so the modality that best addresses efficacy is the one most likely to import a myelosuppression problem the catalytic inhibitor did not have.
- What is MS2133's DC50 and Dmax, and does it degrade DOT1L in primary KMT2A-rearranged blasts rather than in cell lines?
- Does MS2133 phenocopy DOT1L knockout or DOT1L catalytic inactivation? That comparison, run in the same cells, is the cleanest available test of how much of the leukaemic dependency is catalysis.
- Which normal cells were tested for the no-toxicity claim, and would a primary haematopoietic progenitor colony assay give the same answer?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Compound 13 (MS2133) is the first-in-class DOT1L PROTAC degrader; it potently induces DOT1L degradation in a concentration- and time-dependent manner without affecting DOT1L mRNA expression, requires binding to the E3 ligase VHL and to DOT1L, acts through the ubiquitin-proteasome system, is selective for DOT1L over other methyltransferases, and inhibits growth of MLL-rearranged leukaemia cells while having no toxicity on normal cells. The stated rationale is that DOT1L has biological functions independent of its methyltransferase activity.
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. No DC50, Dmax, in vivo data or pharmacokinetics in the abstract, and 'no toxicity on normal cells' is unquantified and unspecified as to which normal cells. Single report, not independently replicated. VHL-dependence makes activity contingent on VHL expression in the test system. The authors present it as a chemical-biology tool as much as a therapeutic.
Evidence that argues against this (1)
- A non-toxic-to-normal-cells claim is hard to reconcile with the genetics: acute Dot1l inactivation in adult mouse marrow depletes myeloid progenitors within 7 days and then long-term HSCs, and an enzymatically inactive DOT1L fails to rescue sustained haematopoiesis. Either the degradation achieved in normal cells was partial, or the in vitro normal-cell assay was not a haematopoietic progenitor assay. Grigsby SM et al. 2021
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A later DOT1L PROTAC series reports DC50 values of 0.33 micromolar (DOT1L705) and 5 nanomolar (DOT1L808), MLL-rearrangement-dependent viability effects, retained activity in menin-inhibitor-resistant cells, and complete tumour regression in an orthotopic leukaemia model.
Development of DOT1L-Targeted Protein Degraders for Treating MLL-r Leukemia. · Quan S, Unno K, Gupta DG, et al. · J Med Chem 2026What this evidence cannot show
Abstract read from the eutils record; full text not read. Cited here only to establish that a second, more advanced degrader series exists and that MS2133 is therefore not the state of the art despite being first. The two series have not been compared head to head.
Evidence that argues against this (1)
- That two independent groups converged on degradation is encouraging for the modality and says nothing about whether the modality solves the clinical problem. The clinical problem with pinometostat was efficacy at a dose that was never escalated to its maximum, and no degrader paper addresses dose-limiting toxicity in humans. Stein EM et al. 2018
first-in-class VHL-recruiting DOT1L PROTAC
degradation as the modality for non-enzymatic DOT1L function
- ← related toHow much sugar the cell has decides how much DOT1L there is - a metabolic dial on the leukaemia's mark
DOT1L has a native E3 (UBE3C) that the VHL-recruiting PROTAC does not use