Experiment: block the beta-2 adrenergic receptor in a KMT2A-rearranged model — the leukaemogenic nerve signal runs through a receptor an old, cheap, infant-licensed drug already hits
also called propranolol in KMT2A-rearranged leukaemia · beta-2 blockade niche experiment
Leukaemia damages the nerves that control the bone marrow, and the harmful signal travels through one specific receptor — beta-2. Drugs that block beta receptors are among the oldest and cheapest in medicine and one of them is routinely given to babies. Nobody has tested whether blocking that receptor slows this leukaemia.
The same thing, for a scientist
In MLL-AF9 AML, adrenergic signalling promoting leukaemogenesis is transduced by the **beta-2, and explicitly not the beta-3**, adrenergic receptor on stromal cells of leukaemic marrow, in a setting where the leukaemia itself disrupts sympathetic nerves and Nestin+ niche-cell quiescence [E1]. Normal circadian niche physiology, by contrast, runs through **beta-3** [E2] - so the malignant and physiological circuits are pharmacologically separable, which is unusual and useful. The experiment: propranolol (non-selective beta-blocker, licensed in infants for haemangioma, decades of paediatric safety data) and a beta-2-selective antagonist such as ICI-118,551, given to KMT2A-rearranged models with and without standard chemotherapy and with and without a menin inhibitor. Primary readouts: leukaemia burden, survival, marrow sympathetic fibre density by tyrosine hydroxylase immunohistochemistry, and NG2+ periarteriolar pericyte number. The sympathetic/adrenergic intersection with the entire KMT2A literature is **two** PubMed records, both of which are cited in this beat, so this is an almost completely unexplored axis in a disease with a licensed, paediatric-dosed drug sitting on it.
Design and the reason for each arm
Arms. (1) vehicle; (2) propranolol; (3) ICI-118,551 (beta-2 selective), to separate the beta-2 contribution from non-selective beta-blockade; (4) standard chemotherapy; (5) chemotherapy + propranolol; (6) menin inhibitor; (7) menin inhibitor + propranolol. Arms 6 and 7 matter because the practical question is not whether a beta-blocker has single-agent activity - it almost certainly does not - but whether it adds to the drug class this lab cares about.
Models. Syngeneic MLL-AF9, because that is where the receptor assignment was made and the stroma must be host stroma for a stromal-receptor mechanism to operate. Then a KMT2A::AFF1 infant B-ALL PDX, accepting that the murine stroma in an immunodeficient host is a limitation for a mechanism that acts on stromal beta-2 receptors - a limitation that should be stated in any resulting paper rather than glossed.
Readouts. Survival and burden; tyrosine hydroxylase immunohistochemistry for marrow nerve fibre density, because the mechanism predicts preservation of fibres; NG2+ periarteriolar pericyte number, because their loss is the documented downstream consequence; and Nestin+ cell quiescence by Ki-67.
Negative control that makes the design interpretable. A beta-3 agonist (mirabegron) arm. If beta-2 blockade helps and beta-3 agonism does nothing, the receptor assignment holds in this model. If both help, the effect is general adrenergic modulation and the beta-2 specificity claim is not doing work.
Why this is cheap and why it has not been done
Propranolol costs almost nothing, has a paediatric formulation, and is given to infants for months at a time for haemangioma. Beta-blockade has been explored in other solid tumours with mixed results. The reason it has not been tried here is almost certainly that the mechanism is buried in a 2014 niche paper whose primary claim is about sympathetic neuropathy rather than about therapy, and that the entire sympathetic/adrenergic literature in this disease is two records.
The honest framing for the hand-back: this is a low-cost, high-uncertainty experiment on an axis with a single supporting paper, not a mature therapeutic proposal. Its value is that the downside is small, the drug is available, and the receptor assignment is specific enough to be wrong in a detectable way.
| design | seven-arm in vivo study in syngeneic MLL-AF9 and then a KMT2A::AFF1 infant B-ALL PDX: vehicle; propranolol; ICI-118,551 (beta-2 selective); chemotherapy; chemotherapy + propranolol; menin inhibitor; menin inhibitor + propranolol, with a mirabegron (beta-3 agonist) arm as the receptor-specificity control |
|---|---|
| readout | survival and leukaemia burden; marrow sympathetic nerve fibre density by tyrosine hydroxylase immunohistochemistry; NG2+ periarteriolar pericyte number; Nestin+ cell quiescence by Ki-67 |
| controls | mirabegron beta-3 arm to test receptor specificity; vehicle; all lines verified in Cellosaurus by accession; any non-KMT2A comparator screened for pathway-equivalent lesions (PICALM::MLLT10, NUP98 fusions, NPM1c, other HOXA/MEIS1 drivers) |
| cost tier | low |
| timeline | 3-4 months |
| drug availability | propranolol is licensed in infants for infantile haemangioma with a paediatric oral formulation and decades of dosing experience; ICI-118,551 is a research compound |
| literature size | the sympathetic/adrenergic x KMT2A PubMed intersection is 2 records, both cited here; this axis is essentially unexplored in this disease |
| honest framing | low cost, high uncertainty, single supporting paper; the value is that the downside is small and the receptor assignment is specific enough to be falsified |
tests hypotheses
[ "hyp-niche-remodelling-precedes-relapse-kmt2a" ]
- Beta-2 blockade risks bronchospasm in a neutropenic infant with frequent respiratory infection, and beta-blockade generally blunts the compensatory tachycardia a septic or anaemic child relies on.
- Preserving the leukaemia-associated adrenergic signal cuts both ways: beta-3 signalling normally mobilises HSCs, and interfering with adrenergic tone could alter normal stem-cell trafficking during count recovery.
- The axis rests on a single 2014 paper; a negative result would be uninformative about the mechanism and a positive result would need independent replication before anyone acted on it.
- If chemotherapy denervates the marrow anyway, a beta-blocker given alongside chemotherapy may have no substrate to act on.
- Does a PDX in an immunodeficient mouse with murine stroma support a mechanism that acts on stromal beta-2 receptors, or does the model defeat the experiment?
- Is beta-2 blockade antagonistic to a menin inhibitor? A differentiating agent and a niche-preserving agent could plausibly interfere, and arm 7 is where that would show.
- Does propranolol at infant haemangioma doses achieve marrow exposures relevant to stromal beta-2 receptors?
- If the mirabegron arm also helps, what does that leave of the beta-2 specificity claim, and would it point instead at general adrenergic modulation of the mesenchymal compartment?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In an MLL-AF9 mouse AML model, adrenergic signalling promoting leukaemogenesis is transduced by the beta-2, but not the beta-3, adrenergic receptor expressed on stromal cells of leukaemic bone marrow, in a setting where AML development disrupts sympathetic nervous system nerves and the quiescence of Nestin-positive niche cells and expands osteoblast-primed mesenchymal progenitors at the expense of HSC-maintaining NG2-positive periarteriolar cells.
Acute myelogenous leukemia-induced sympathetic neuropathy promotes malignancy in an altered hematopoietic stem cell niche · Hanoun M, Zhang D, Mizoguchi T, Pinho S, Pierce H, Kunisaki Y, Lacombe J, Armstrong SA, et al. · Cell Stem Cell 2014What this evidence cannot show
Mouse, myeloid, single paper; the receptor assignment rests on subtype pharmacology and genetics whose selectivity is imperfect. No lymphoid, KMT2A::AFF1 or infant arm, and no therapeutic beta-blockade arm - this experiment proposes what the source paper did not do. Read: abstract only.
Evidence that argues against this (1)
- Marrow sympathetic neuropathy is not specific to MLL-AF9: it is present in JAK2V617F MPN, NRAS-G12D pre-leukaemic myelopoiesis and p53-deletion lymphoid disease, and in human MPN and AML marrow, and is also induced by chemotherapy. If denervation is a general consequence of marrow disease and of treatment, preserving adrenergic signalling with a receptor antagonist may be treating a bystander lesion, and a beta-blockade arm could show nothing while the mechanism remains real but non-limiting. Read: abstract only. Bernal A et al. 2025
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Normal circadian control of the niche runs through the beta-3 adrenergic receptor: circulating HSCs oscillate in antiphase with marrow Cxcl12, and the oscillation is driven by clock genes acting through circadian noradrenaline secretion from marrow sympathetic nerves transmitted to stromal cells by the beta-3 adrenergic receptor, decreasing nuclear Sp1 and rapidly downregulating Cxcl12.
Haematopoietic stem cell release is regulated by circadian oscillations · Mendez-Ferrer S, Lucas D, Battista M, Frenette PS · Nature 2008What this evidence cannot show
Mouse, nocturnal, normal physiology. Cited here to establish that the physiological and malignant adrenergic circuits use different receptor subtypes, which is what makes selective beta-2 blockade a coherent proposal; beta-subtype pharmacological selectivity is relative, not absolute, and propranolol is non-selective. Read: abstract only.
Evidence that argues against this (1)
- Nestin+ MSCs, the cells physically paired with adrenergic fibres, have HSC-maintenance genes that are selectively downregulated by beta-3 adrenoreceptor activation - so beta-3 also modulates the mesenchymal niche compartment that MLL-AF9 leukaemia disturbs. The two receptor circuits converge on overlapping stromal populations, and a clean functional separation between beta-2 'malignant' and beta-3 'physiological' signalling may not survive testing. Read: abstract only. Mendez-Ferrer S et al. 2010
tests whether preserving adrenergic niche signalling alters leukaemia progression
targets the beta-2 receptor that transduces the leukaemogenic adrenergic signal
readouts include marrow sympathetic fibre density by tyrosine hydroxylase immunohistochemistry