Nestin+ mesenchymal stem cell — the niche cell that MLL-AF9 leukaemia wakes up and then loses

Nestin is an intermediate filament, originally a neural stem-cell marker. Méndez-Ferrer and colleagues found that a Nestin-GFP reporter labels a rare perivascular mesenchymal population in marrow that carries all the colony-forming fibroblast activity, self-renews in serial transplantation as 'mesenspheres', and sits physically beside both HSCs and sympathetic nerve fibres. Depleting them in vivo drops marrow HSC content quickly 1.

The unresolved problem: conditional deletion of Scf from Nestin-cre cells had no effect on HSCs, while deletion from LEPR+ perivascular cells did 4. The usual explanation is that Nestin-GFP and Nestin-cre label different cells — the reporter labelling a periarteriolar-enriched subset and the cre recombining more broadly or not at all in the relevant population. That explanation is plausible and unproven. This entry therefore treats 'Nestin+ MSC' as a reporter-defined population with demonstrated depletion phenotypes, not as a molecularly defined cell type, and flags anywhere a conclusion depends on which reading is correct.

In MLL-AF9 AML the Nestin+ compartment does not simply shrink. It is activated out of quiescence, and the mesenchymal compartment expands towards an osteoblast-primed phenotype while the NG2+ periarteriolar cells that maintain HSC quiescence are lost 2. The healthy niche and the leukaemic niche are therefore different niches, not the same niche in different amounts — which is the point the authors make explicitly, proposing 'separate mesenchymal niche activities for malignant and healthy hematopoietic stem cells'.

A practical corollary the graph should not lose: in leukaemic marrow, the HSC-supporting compartment is depleted while the leukaemia-supporting compartment is expanded. That is a candidate explanation for why the normal haematopoiesis of a child with KMT2A-rearranged leukaemia recovers poorly even at low blast burden, and it is measurable in a marrow biopsy.

The functional experiment matters more than the descriptive one. Forte and colleagues report that in MLL-AF9 AML, nestin+ BMSC numbers are not reduced while bulk stroma is, and that what changes is their function: they begin supporting AML cells at the expense of non-mutated HSCs. Depleting nestin+ cells delays leukaemogenesis and, in chimeric mice, selectively decreases AML cells and not normal ones; the mechanism is increased oxidative phosphorylation, TCA cycle activity and glutathione-mediated antioxidant defence transferred from stroma to blast 3.

Hold that beside the Ebf2+ result: deleting Ebf2+ MSPCs accelerated AML [see mechanism-leukaemia-remodels-its-own-niche E3], while deleting nestin+ BMSCs delayed it. Two mesenchymal depletions in the same driver, opposite directions. This is the single most important caution in this beat: 'the stroma' is not one thing and does not have one sign. Any proposal to target the mesenchymal niche in KMT2A-rearranged disease has to name which subset, and the two subsets that have been tested disagree.

The metabolic theme is consistent across three independent studies — Prx1+ CXCL12 controlling LSC redox balance (cell-type-car-cell), nestin+ BMSCs supplying OXPHOS and glutathione 3, and marrow leptin driving complex-I activity (cell-type-bone-marrow-adipocyte). The mesenchymal niche's contribution to leukaemic survival in this genotype is repeatedly metabolic rather than a simple anti-apoptotic signal.