The brake pad is actively thinned: autophagy degrades HEXIM1 in MLL-fusion and NPM1c AML cells
also called HEXIM1 autophagic degradation · NPM1 and HEXIM1 proteolysis
In AML cells carrying an MLL fusion or a mutant NPM1, the cell's own recycling machinery chews up HEXIM1, the protein that holds the elongation enzyme in check. Blocking that recycling brings HEXIM1 back.
The same thing, for a scientist
Huang and colleagues found autophagy activated in acute leukaemia cells expressing mutant NPM1 (NPMc+) or MLL-fusion proteins, and that this activation degrades **NPM1 and HEXIM1 - two negative regulators of BET pathway activation**. Inhibiting autophagy pharmacologically or by ATG5 knockdown raises both proteins, and the BET inhibitors JQ1 and I-BET-151 also inhibit autophagy and raise NPM1 and HEXIM1 [E1]. The conclusion drawn is that HEXIM1 and NPM1 degradation contributes to BET pathway activation in these AML subtypes. Read alongside the independent finding that HEXIM1 knockdown in MOLM-13 raises c-Myc and growth and blunts JQ1 [E2], and the human genetics showing that dismantling the reservoir causes developmental disease rather than cancer [E3], this is the most direct published evidence that the reservoir is **actively depleted** in MLL-fusion leukaemia rather than merely bypassed. It is also, as of this beat, the only such evidence, and it does not resolve the fusion partner or the lineage.
What this does and does not establish
Does: that HEXIM1 protein is turned over by autophagy in AML cells carrying MLL fusions or NPM1c, and that blocking autophagy restores it 1.
Does not: identify which MLL fusion partner, show the effect in B-lineage or infant disease, quantify how much HEXIM1 is lost, or show that the loss changes the free-versus-sequestered P-TEFb partition. No 7SK measurement appears in the paper. The causal chain from HEXIM1 loss to BET-pathway activation is inferred from co-varying protein levels, and the confounder is explicit in the paper's own data: JQ1 both inhibits autophagy and raises HEXIM1, so the drug and the proposed mechanism move the same marker 1.
Note also the direction relative to the rest of this beat. BET inhibitors release P-TEFb acutely and induce HEXIM1 transcriptionally, and here they additionally raise HEXIM1 by blocking its degradation. Three mechanisms, two of which raise HEXIM1 and one of which frees P-TEFb, all triggered by the same compound. Any attempt to attribute a BET inhibitor's leukaemia activity to one of them needs a separation-of-function experiment that has not been done.
| mechanism class | protein turnover |
|---|---|
| target protein | HEXIM1 (and NPM1) |
| route | autophagy; blocked by pharmacological inhibitors, ATG5 knockdown, and by JQ1/I-BET-151 |
genotypes reported
[ "NPM1c+ AML", "MLL-fusion AML (partner not specified)" ]
not tested
[ "KMT2A::AFF1", "B-lineage", "infant disease", "7SK abundance", "free versus sequestered P-TEFb" ]
related entries
[ "gene-hexim1", "gene-brd4", "protein-7sk-snrnp" ]
- Autophagy inhibition to preserve HEXIM1 would be a global intervention in a pathway that normal haematopoietic stem cells depend on heavily.
- If a BET inhibitor's activity partly depends on raising HEXIM1, then any resistance mechanism that lowers HEXIM1 - including the autophagy activation described here - is a built-in escape route.
- Is HEXIM1 protein lower in KMT2A::AFF1 blasts than in matched non-KMT2A leukaemia blasts? A blot on primary material would settle it and none exists.
- Does blocking autophagy raise HEXIM1 enough to change the free P-TEFb pool, or only enough to change a blot?
- Which of the three routes by which a BET inhibitor moves this axis - acute P-TEFb release, HEXIM1 transcriptional induction, blocked HEXIM1 degradation - carries its anti-leukaemia activity?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Autophagy is activated in acute leukaemia cells expressing mutant NPM1 or MLL-fusion proteins and results in degradation of NPM1 and HEXIM1, two negative regulators of BET pathway activation; pharmacological autophagy inhibition or ATG5 knockdown increases both proteins, and JQ1 and I-BET-151 also inhibit autophagy and increase NPM1 and HEXIM1.
Autophagy mediates proteolysis of NPM1 and HEXIM1 and sensitivity to BET inhibition in AML cells. · Huang M, Garcia JS, Thomas D, Zhu L, Nguyen LX, Chan SM, Majeti R, Medeiros BC, Mitchell BS · Oncotarget 2016What this evidence cannot show
AML cell lines and primary AML cells. The MLL-fusion arm is not resolved to a partner and no B-lineage, infant or KMT2A::AFF1 cell appears. No 7SK or P-TEFb partition measurement. Published in Oncotarget.
Evidence that argues against this (1)
- The proposed mechanism and the drug used to probe it move the same marker in the same direction, so the experiment cannot separate 'BET inhibition works by raising HEXIM1' from 'BET inhibition happens to raise HEXIM1'. Huang M et al. 2016
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In MOLM-13 (KMT2A::MLLT3) AML cells, HEXIM1 knockdown raises c-Myc and in vitro growth and reduces JQ1-induced differentiation, apoptosis and in vivo survival benefit, while ectopic HEXIM1 attenuates growth.
HEXIM1 induction is mechanistically involved in mediating anti-AML activity of BET protein bromodomain antagonist. · Devaraj SG, Fiskus W, Shah B, Qi J, Sun B, Iyer SP, Sharma S, Bradner JE, Bhalla KN · Leukemia 2016What this evidence cannot show
One KMT2A-rearranged line plus one primary AML sample; read from the PMC full text because the PubMed record has no abstract. HEXIM1 knockdown alone did not change in vivo growth.
Evidence that argues against this (1)
- BRD4's ET domain can activate transcription independently of P-TEFb through JMJD6 or NSD3, which the authors invoke to explain why HEXIM1 knockdown does not abolish the BET response; the HEXIM1 arm is one route among several. Devaraj SG et al. 2016
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Disassembling the 7SK snRNP in a human - through a de novo MEPCE nonsense variant - produces a neurodevelopmental disorder with enhanced P-TEFb-dependent transcription, not a leukaemia.
de novo MEPCE nonsense variant associated with a neurodevelopmental disorder causes disintegration of 7SK snRNP and enhanced RNA polymerase II activation. · Schneeberger PE, Bierhals T, Neu A, Hempel M, Kutsche K · Sci Rep 2019What this evidence cannot show
A single patient, assayed in fibroblasts. A rare syndrome is not powered to exclude a modest cancer risk.
Evidence that argues against this (1)
- The same is true of biallelic LARP7 loss (Alazami syndrome), so two independent human genotypes point the same way; but both are constitutional, lifelong and germline, which is a different perturbation from an acquired change in a blast. Schneeberger PE et al. 2019
autophagic degradation lowers HEXIM1 protein in MLL-fusion and NPM1c AML
BET inhibitors block the degradation and raise HEXIM1, confounding attribution
a mechanism by which the brake could thin without the fusion touching the reservoir