The perivascular and sinusoidal niche — where the functional evidence actually sits, and where the marrow is genuinely hypoxic
also called vascular niche · perisinusoidal niche · perivascular niche
Most blood stem cells are not at the bone surface at all: they sit wrapped against the wall of the marrow's wide, slow-flowing drainage vessels, next to support cells that feed them. This is also the part of the marrow with the least oxygen, and it is the compartment where the strongest experimental evidence for a stem-cell niche now lies.
The same thing, for a scientist
The perivascular niche is the compartment in which the two indispensable HSC maintenance factors are made. Conditional deletion of Scf depletes HSCs when the gene is removed from **endothelial cells** or from **LEPR-positive perivascular stromal cells**, and removes most HSCs when removed from both, while deletion from haematopoietic cells, osteoblasts or Nestin-cre cells has no effect [E1]; CXCL12-abundant reticular cells, which most HSCs contact, surround sinusoidal endothelium [E2]. Within the vasculature the compartment splits: **quiescent** HSCs associate with NG2+-ensheathed arterioles, and activating the cell cycle **redistributes** HSCs to LEPR+ perisinusoidal sites, so perisinusoidal residence tracks the cycling state [E3]. It is also the genuinely hypoxic compartment — direct pO2 measurement puts the marrow minimum, about 9.9 mmHg (1.3 percent), in deep peri-sinusoidal regions [E4]. In leukaemia, the vascular wall is an active participant rather than a boundary: endothelial small extracellular vesicles carrying ANGPTL2 accelerate MLL-AF9 leukaemia through LILRB2, and blocking their secretion from endothelium - but not from perivascular cells, megakaryocytes or spleen stroma - delays leukaemia while sparing normal haematopoiesis [E5].
Why this compartment displaced the endosteum
The endosteal niche was proposed first and the perivascular niche won on method. Calvi and Zhang showed in 2003 that increasing osteoblasts increases HSCs — a sufficiency result. Ding and Morrison asked the necessity question with conditional deletion of the specific factor in the specific cell, and the answer came back endothelium and LEPR+ perivascular cells, with osteoblasts and Nestin-cre cells dispensable 1.
Sufficiency and necessity are different claims and the field conflated them for a decade. This graph should not: concept-endosteal-niche records what the bone surface does for leukaemia, which is a live question, without reinstating it as the normal HSC niche.
A second, subtler point from 3: perisinusoidal residence is a property of cycling HSCs. Quiescent HSCs are at arterioles; activate the cycle pharmacologically or genetically and they move to sinusoids. So this compartment is not the dormancy compartment, which matters directly for a graph that invokes dormancy to explain residual disease.
The oxygen result, which is the opposite of the textbook
Every review says the stem-cell niche is hypoxic and most imply that means the endosteum. Direct measurement says the low-oxygen compartment is peri-sinusoidal, at about 9.9 mmHg, with the endosteum better oxygenated because small arteries perfuse it 4.
That matters for this disease specifically, because the one hypoxia experiment performed in KMT2A-rearranged AML found that hypoxia sensitises cells to venetoclax plus the menin inhibitor MI-503, with the combination acting partly by repressing HIF-1A and its target HDAC9 6. Put those together and the prediction is uncomfortable for the niche-protection story and encouraging for the therapy: the most hypoxic part of the marrow should be where a venetoclax plus menin-inhibitor combination works best, not where cells hide from it. That prediction is untested and is carried as a hypothesis in hyp-niche-remodelling-precedes-relapse-kmt2a.
| compartment | perivascular / perisinusoidal |
|---|---|
| cell cycle association | cycling HSCs redistribute here from NG2+ periarteriolar sites; this is not the dormancy compartment |
| literature size | the perivascular/sinusoidal x KMT2A PubMed intersection is 6 records, of which 3 are marrow-niche papers and 3 concern KMT2A sarcomas, busulfan pharmacokinetics and a skin presentation |
cellular constituents
[ "sinusoidal endothelial cell", "CAR / LEPR+ perivascular stromal cell", "megakaryocyte (perisinusoidal)", "adipocyte" ]
required factors
[ "SCF (KITL) from endothelium and LEPR+ cells", "CXCL12 from CAR/LEPR+ cells" ]
oxygen
{
"value": 9.9,
"unit": "mmHg (approx 1.3 percent O2)",
"note": "the lowest measured pO2 in mouse marrow, in deep peri-sinusoidal regions",
"source_evidence": "E4"
}leukaemic relevance
{
"endothelial_evs": "ANGPTL2-containing sEVs from Cdh5+/Tie2+ endothelium accelerate MLL-AF9 AML via LILRB2, with no effect on normal haematopoiesis when blocked",
"hypoxia_direction": "hypoxia SENSITISES KMT2A-rearranged AML to venetoclax plus MI-503, so the most hypoxic marrow compartment is predicted to be where that combination works best, not where cells hide",
"lymphoid_caveat": "compartment-specific Cxcl12 deletion places HSCs in a perivascular niche and EARLY LYMPHOID PROGENITORS in an endosteal niche dependent on osteoblast-derived CXCL12; infant KMT2A-rearranged B-ALL is a disease of early B-lymphoid progenitors, so the perivascular compartment may be the wrong one for it"
}atlas refs
[
{
"label": "Endothelial and perivascular SCF maintain HSCs",
"pmid": "22281595",
"title": "Endothelial and perivascular cells maintain haematopoietic stem cells.",
"authors": "Ding L, Saunders TL, Enikolopov G, et al.",
"venue": "Nature",
"year": 2012,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
},
{
"label": "Deep peri-sinusoidal marrow is the lowest-oxygen compartment",
"pmid": "24590072",
"title": "Direct measurement of local oxygen concentration in the bone marrow of live animals.",
"authors": "Spencer JA, Ferraro F, Roussakis E, et al.",
"venue": "Nature",
"year": 2014,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
}
]- Perivascular is where the functional evidence is, which means every effective niche-directed agent will act on the compartment that maintains the patient's own stem cells through the same two factors.
- The compartment protects this leukaemia through at least two independent routes - endothelial ANGPTL2 vesicles and mesenchymal CXCL12-mediated redox buffering - so single-pathway blockade is likely to be compensated.
- If perisinusoidal residence is a property of cycling rather than dormant cells, a drug that mobilises cells from this compartment would be mobilising the fraction that chemotherapy already kills best, and leaving the dormant fraction where it is.
- If deep peri-sinusoidal marrow is the hypoxic compartment and hypoxia sensitises KMT2A-rearranged AML to venetoclax plus a menin inhibitor, does that combination clear the perisinusoidal compartment preferentially? Spatially resolved response measurement in a xenograft would answer it.
- HSCs and early lymphoid progenitors occupy distinct niches. Which one does an infant KMT2A::AFF1 pro-B blast occupy, given that the fusion's target cell is a fetal pre-pro-B cell rather than an HSC?
- Does the redox-buffering protection described for mesenchymal and adipocyte compartments operate against menin inhibitors, or only against cytarabine? Nobody has tested a menin inhibitor in a stromal co-culture with a redox readout.
- Is LILRB2 blockade active in KMT2A-rearranged disease? It is the tractable end of the only normal-haematopoiesis-sparing endothelial mechanism found in this beat.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
SCF is expressed primarily by perivascular cells throughout bone marrow; HSCs are depleted when Scf is conditionally deleted from endothelial cells or from leptin-receptor-expressing perivascular stromal cells, and most HSCs are lost when deleted from both, whereas deletion from haematopoietic cells, osteoblasts, Nestin-cre or Nestin-creER cells has no effect on HSC frequency or function.
Endothelial and perivascular cells maintain haematopoietic stem cells · Ding L, Saunders TL, Enikolopov G, Morrison SJ · Nature 2012What this evidence cannot show
Mouse conditional genetics; cre-driver specificity is the field's central methodological weakness and the negative Nestin-cre result is disputed on that basis. SCF is one factor among several. Read: abstract only.
Evidence that argues against this (1)
- Nestin-GFP-defined perivascular MSCs are required for marrow HSC content and progenitor homing when depleted in vivo, so a cell can be dispensable as an SCF source and still be required as a population. The functional partition of the perivascular compartment is unresolved. Read: abstract only. Mendez-Ferrer S et al. 2010
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Most HSCs are found in contact with CXCL12-abundant reticular (CAR) cells, which surround sinusoidal endothelial cells or are located near the endosteum, and induced deletion of CXCR4 in adult mice severely reduces HSC numbers and increases sensitivity to myelotoxic injury.
Maintenance of the hematopoietic stem cell pool by CXCL12-CXCR4 chemokine signaling in bone marrow stromal cell niches · Sugiyama T, Kohara H, Noda M, Nagasawa T · Immunity 2006What this evidence cannot show
Mouse; CAR identity rests on a CXCL12 reporter. The HSC-contact claim is histological. Read: abstract only.
Evidence that argues against this (1)
- In MLL-AF9 leukaemia CXCR4 is essential to the leukaemic cell in a CXCL12-INDEPENDENT manner: Cxcr4 deletion eradicated leukaemia without impairing marrow homing, and CXCL12 was dispensable for leukaemia development in recipients. The retention logic that governs normal HSCs does not transfer intact to the leukaemic cell. Read: abstract only. Ramakrishnan R et al. 2020
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Quiescent HSCs associate specifically with NG2+-pericyte-ensheathed arterioles rather than with sinusoids, and pharmacological or genetic activation of the HSC cell cycle alters the distribution of HSCs from NG2+ periarteriolar niches to LEPR+ perisinusoidal niches.
Arteriolar niches maintain haematopoietic stem cell quiescence · Kunisaki Y, Bruns I, Scheiermann C, Ahmed J, Pinho S, Zhang D, et al. · Nature 2013What this evidence cannot show
Mouse whole-mount imaging with computational association analysis; the redistribution is a population-level shift, not tracking of individual cells. Read: abstract only.
Evidence that argues against this (1)
- The arteriolar-versus-sinusoidal axis does not describe where lymphoid cells sit, and for a B-lymphoid disease that is decisive. Cxcl12-DsRed knock-in and compartment-specific Cxcl12 deletion showed that deleting Cxcl12 from ENDOTHELIUM depleted HSCs but not lymphoid progenitors, deleting it from PERIVASCULAR stromal cells depleted HSCs and certain restricted progenitors and mobilized them, and deleting it from OSTEOBLASTS depleted certain early lymphoid progenitors but not HSCs or myeloerythroid progenitors and did not mobilize them. The authors' conclusion is explicit: HSCs occupy a perivascular niche and early lymphoid progenitors occupy an ENDOSTEAL niche. Infant KMT2A-rearranged B-ALL is a disease of early B-lymphoid progenitors, so the perivascular compartment may be the wrong one for it. Read: abstract only. Ding L et al. 2013
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Direct in vivo two-photon phosphorescence lifetime measurement in live mice places the lowest bone-marrow oxygen tension, about 9.9 mmHg (1.3 percent), in deep peri-sinusoidal regions, with overall marrow pO2 below 32 mmHg despite very high vascular density, and the endosteal region less hypoxic because it is perfused by small arteries; these values change markedly after radiation and chemotherapy.
Direct measurement of local oxygen concentration in the bone marrow of live animals · Spencer JA, Ferraro F, Roussakis E, Klein A, Wu J, Runnels JM, et al. · Nature 2014What this evidence cannot show
Mouse calvarium; normal, not leukaemic, marrow. Local extracellular pO2 is not the cell's hypoxic transcriptional state. Read: abstract only.
Evidence that argues against this (1)
- The MLL-AF9 metabolic-imaging literature assigns the hypoxic niche to the endosteum instead, and localises leukaemia-initiating cell homing and symmetric division there. The two spatial assignments are incompatible as written. Read: abstract only. Hao X et al. 2019
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Conditional Vps33b knockout in seven cell-type-specific cre lines showed that blocking small extracellular vesicle secretion from Cdh5+ or Tie2+ endothelial cells, but not from Lepr+ perivascular cells, Osx+ osteoprogenitors, Pf4+ megakaryocytes or Tcf21+ spleen stromal cells, markedly delayed MLL-AF9-induced AML progression with no effect on normal haematopoiesis; the endothelial vesicles carry high ANGPTL2 which accelerates leukaemia by binding LILRB2, and ANGPTL2-sEVs are also required for maintenance of primary human AML cells.
ANGPTL2-containing small extracellular vesicles from vascular endothelial cells accelerate leukemia progression · Huang D, Sun G, Hao X, He X, Zheng Z, Chen C, Yu Z, Xie L, et al. · Journal of Clinical Investigation 2021What this evidence cannot show
Mouse MLL-AF9; VPS33B has trafficking roles beyond sEV secretion, so the perturbation is not sEV-specific. Single primary study with no replication in this genotype. Read: abstract only.
Evidence that argues against this (1)
- The same perivascular compartment supplies protection through a completely different route in a different study of the same driver, where CXCL12 from Prx1+ mesenchymal cells maintains MLL::AF9 LSCs via redox and energy balance and the defect caused by Cxcl12 deletion is attenuated by antioxidant therapy. If the compartment protects through multiple redundant routes, blocking one may achieve little. Read: abstract only. Vinado AC et al. 2022
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In KMT2A-rearranged AML, hypoxia induces HDAC9 expression; venetoclax plus a menin inhibitor inhibits the hypoxia pathway, especially HIF-1A, and its target HDAC9, HDAC9 is indispensable for KMT2A-rearranged AML proliferation, and hypoxia induction SENSITISED KMT2A-rearranged AML to venetoclax plus MI-503-mediated proliferation inhibition and apoptosis induction.
Repressing HIF-1alpha-induced HDAC9 contributes to the synergistic effect of venetoclax and MENIN inhibitor in KMT2Ar AML · Ling Q, Zhou Y, Qian Y, Qian J, Zhang Y, Wang J, Zhu Y, Zhou Y, et al. · Biomarker Research 2023What this evidence cannot show
PubMed classifies the record as a Letter. The work is cell lines, primary ex vivo cells and a MOLM-13 xenograft; hypoxia is induced in culture rather than measured in a marrow compartment, so the connection between the culture condition and the peri-sinusoidal region is an inference this entry makes, not one the paper makes. Myeloid, adult-derived lines. Read: abstract only.
Evidence that argues against this (1)
- Three independent niche compartments protect KMT2A-rearranged leukaemia by buffering oxidative stress - Prx1+ mesenchymal CXCL12, nestin+ BMSC glutathione and OXPHOS transfer, and adipocyte leptin driving complex I with a compensatory antioxidant programme. A redox-adapted state that is protective against cytarabine would not obviously be sensitising to venetoclax plus a menin inhibitor, so the two literatures make opposite predictions about the same metabolic state and neither has tested the other's drug. Read: abstract only. Forte D et al. 2020
vascular compartment; where the conditional-deletion evidence for HSC maintenance sits
the vessel wall itself
perivascular stromal constituent
deep peri-sinusoidal marrow is the measured minimum, about 9.9 mmHg
Scf is primarily expressed by perivascular cells throughout marrow
LEPR+ perivascular stroma is the required SCF source compartment
vascular-niche chemoresistance receptor