Oncofusion

start / graph / cell-type-lmpp

Cell typemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q4-cells

Lymphoid-primed multipotent progenitor (LMPP) — the bi-potent cell that explains how one fusion makes two lineages

also called LMPP · lympho-myeloid primed progenitor · early lymphoid progenitor · ELP · GMLP · human multilymphoid progenitor (MLP)

The LMPP is a blood progenitor that has already started down the lymphoid road but has not thrown away its myeloid options. If KMT2A::AFF1 lands here, the resulting leukaemia can present as B-cell, as myeloid, or as both.

The same thing, for a scientist

The LMPP sits immediately downstream of the multipotent progenitor and retains both lymphoid and myeloid potential while having lost megakaryocyte-erythroid potential. Its human counterpart, the multilymphoid progenitor, is a Thy-1-negative-to-low CD45RA-positive population inside the CD34-positive CD38-negative compartment that clonally generates all lymphoid lineages plus monocytes, macrophages and dendritic cells. In mouse Mll-AF4 models, the fetal liver LMPP is the compartment reported to supply the prerequisites for t(4;11) leukaemia initiation, and an LMPP population is the cell that propagates a mixed B-cell-precursor/myeloid Mll-AF4 leukaemia. This makes the LMPP the mechanistic bridge between cell of origin and the lineage plasticity that later manifests as MPAL and as lineage switch under CD19-directed therapy.

Go deeper
What defines it, in mouse and in human1 deeper

Mouse. Cell Ontology term CL:0000936 carries LMPP, lymphoid-primed multipotent progenitor, ELP and GMLP as exact synonyms and defines the cell as lineage-negative, Kit-positive, Sca-1-positive, FLT3-positive, CD34-positive, CD150-negative, glycophorin-A-negative, with a transcription factor state that is E2A-positive and Ikaros-positive but EBF-negative, Pax5-negative and PU.1-negative 4. FLT3 gain and CD150/SLAM loss are the two informative markers: the cell has left the stem-cell niche identity but has not yet committed its B-lineage transcription factor network.

Human. The functional equivalent was mapped clonally as the multilymphoid progenitor: a Thy-1-negative-to-low, CD45RA-positive population within the CD34-positive CD38-negative stem-cell compartment. It gave rise to all lymphoid cell types and to monocytes, macrophages and dendritic cells, which established that in human, as in mouse, myeloid potential is retained inside early lymphoid specification rather than being segregated away at the first branch 1.

That last sentence is the whole reason this node matters to the project. A cell that is physiologically lympho-myeloid does not need to 'transdifferentiate' to become myeloid. It only needs to stop being pushed toward B.

The transcription-factor state, not the surface markers, is the mechanism

The CL definition of the LMPP is EBF-negative and Pax5-negative 4. EBF1 and PAX5 are the commitment factors that make a B cell a B cell and enforce exclusion of myeloid programs. A progenitor that is lymphoid-primed but not yet EBF1/PAX5-locked is exactly a cell in which a chromatin-level oncogene can hold both programs accessible. This is the cell-type-level statement of what the cell-state-primed-bivalent-chromatin node describes at the chromatin level.

The direct evidence tying LMPP to KMT2A::AFF1

Initiation. In the Mll-AF4 mouse system, the fetal liver LMPP is reported to provide the prerequisites for the initiation of t(4;11) infant leukaemia 2. The same group had shown that fetal liver Mll-AF4-positive Lineage-negative Sca1-positive Kit-positive (LSK) cells are the responsive pre-leukaemic compartment, proliferating and upregulating a subset of MLL-AF4 signature genes and Toll-like receptor pathway members on exposure to poly(I:C) or LPS, while a single maternal immune activation was not sufficient to produce leukaemia 5.

Propagation, and the lineage fork. The most informative experiment is a two-arm one. In an entirely murine Mll-AF4 in vivo model, two MLL-AF4 target microRNAs drove the pre-leukaemic-to-leukaemic transition, and they produced different diseases from the same fusion:

co-driverdiseasepropagating population
miR-130b overexpressionmixed B-cell-precursor / myeloid leukaemiaLMPP population
miR-128a overexpressionpro-B ALLexpanded Il7r-positive Kit-positive blast population

Both recapitulated human features including CNS infiltration and activation of an MLL-AF4 expression signature 3. So the lineage output of KMT2A::AFF1 is set not by the fusion but by which progenitor compartment ends up carrying the leukaemia-propagating activity. That is a testable, mechanistic account of lineage plasticity rather than a descriptive one.

Where these cells physically sit

Cell Ontology locates the mouse early lymphoid progenitor in bone marrow 4. For infant KMT2A-rearranged disease the relevant compartment is the fetal liver LMPP 2, and by the second and third trimester the equivalent progenitors are in fetal bone marrow, where B-progenitors already make up a large fraction of the stem/progenitor pool. Human multilymphoid progenitors were mapped from both neonatal cord blood and adult bone marrow 1, so the compartment itself persists after birth; what does not persist is its fetal transcriptional program.

Facts and figures
ontologyCL:0000936
locationFetal liver in the relevant infant-leukaemia window; bone marrow after birth; human multilymphoid progenitors isolated from both cord blood and adult bone marrow
relevance to kmt2aBi-potent progenitor identity provides a cell-intrinsic explanation for MPAL, for lineage switch under CD19-directed therapy, and for the mixed/pro-B immunophenotype of infant KMT2A-r disease
markers
[
  "Lineage-",
  "CD34+",
  "FLT3/CD135+",
  "CD150/SLAMF1-",
  "Kit+ (mouse)",
  "Sca-1+ (mouse)",
  "Thy-1/CD90 neg-to-low (human MLP)",
  "CD45RA+ (human MLP)",
  "CD38- (human MLP)"
]
transcription factor state
[
  "E2A+",
  "Ikaros+",
  "EBF1-",
  "PAX5-",
  "PU.1-"
]
lineage potential
[
  "B",
  "T",
  "NK",
  "dendritic cell",
  "monocyte/macrophage"
]
lineage potential lost
[
  "megakaryocyte-erythroid (by definition of the LMPP stage)"
]
atlas refs
[
  {
    "label": "Clonal mapping of seven human progenitor classes, cord blood and adult bone marrow",
    "pmid": "20543838",
    "title": "Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development.",
    "authors": "Doulatov S, Notta F, Eppert K, et al.",
    "venue": "Nat Immunol",
    "year": 2010,
    "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": "Mll-AF4 murine model, LMPP-propagated mixed BCP/myeloid leukaemia",
    "pmid": "34111240",
    "title": "miR-130b and miR-128a are essential lineage-specific codrivers of t(4;11) MLL-AF4 acute leukemia.",
    "authors": "Malouf C, Antunes ETB, O'Dwyer M, et al.",
    "venue": "Blood",
    "year": 2021,
    "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."
  }
]
How this could bite us back
  • A bi-potent cell of origin means antigen-directed therapy is structurally fragile: removing the B-lineage cells leaves a cell that was never obliged to be a B cell.
  • If menin dependency differs between the LMPP-like and pro-B-like compartments of the same patient, menin inhibition could debulk the lymphoid disease while enriching a myeloid-capable reservoir.
  • Targeting FLT3, which the LMPP definition includes as a positive marker, hits a normal early progenitor compartment as well as the leukaemia.
Still open
  1. Is the human multilymphoid progenitor, rather than the CD10-negative PreProB progenitor, the cell that actually carries the translocation in infant t(4;11)? Single-cell genotype-plus-phenotype on diagnostic marrow could separate these.
  2. Does an LMPP-like subpopulation persist inside established human KMT2A::AFF1 B-ALL, and is it the reservoir that produces myeloid relapse after CD19-directed therapy?
  3. Are LMPP-propagated and pro-B-propagated KMT2A::AFF1 leukaemias differentially sensitive to menin inhibition? If the LMPP reservoir is less menin-dependent, single-agent menin inhibition selects for the lineage-switch-capable cell.
  4. What is the human equivalent of the miR-130b versus miR-128a fork, and is either microRNA measurable at diagnosis as a lineage-risk biomarker?
Every source behind this entry (5)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Human multilymphoid progenitors, a Thy-1-negative-to-low CD45RA-positive population within the CD34-positive CD38-negative stem-cell compartment of neonatal cord blood and adult bone marrow, clonally give rise to all lymphoid cell types as well as monocytes, macrophages and dendritic cells, showing that human haematopoiesis does not segregate myeloid from lymphoid at the first branch.

    Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development · Doulatov S, Notta F, Eppert K, Nguyen LT, Ohashi PS, Dick JE · Nature Immunology 2010
    PMID 20543838DOI
    What this evidence cannot show

    Clonal potential assays report what a cell CAN do in permissive culture and xenograft conditions, which is an upper bound on what it does in vivo. The gate is a surface-marker approximation to a continuum that later single-cell work has shown is not cleanly discrete.

    Evidence that argues against this (2)
    • The multilymphoid progenitor gate is not a homogeneous cell. Clonal functional and single-cell transcriptional analysis of cord-blood MLPs, LMPPs and GMPs found all three heterogeneous, with most individual progenitors having only a single lineage potential and bi- or multi-lineage clones rare, supporting a continuum rather than discrete lympho-myeloid populations. A population-level demonstration that a gate contains cells making lymphoid and myeloid output does not show that single cells in it are bipotent. Read: abstract only. Karamitros D et al. 2018
    • The hierarchy this population sits in is not conserved across ontogeny. Single-cell fate mapping of CD34+ cells showed fetal liver rich in oligopotent progenitors with intermingled myeloid, erythroid and megakaryocytic fates while adult bone marrow contains almost none, collapsing to a two-tier multipotent/unipotent structure. A progenitor definition validated in cord blood and adult marrow cannot be assumed to name the same cell in the fetal liver where KMT2A::AFF1 infant leukaemia is thought to start. Read: abstract only. Notta F et al. 2016
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    The fetal liver lymphoid-primed multipotent progenitor provides the prerequisites for the initiation of t(4;11) infant leukaemia.

    PMID 29903765DOI
    What this evidence cannot show

    Published as a Letter, and no abstract is indexed in PubMed or PMC, so the claim here rests on the title and the journal record rather than on a full reading of the data. Confidence is therefore capped at moderate. It is a mouse Mll-AF4 model, and mouse Mll-AF4 models are known not to reproduce human t(4;11) infant ALL faithfully.

    Evidence that argues against this (2)
    • The same fusion introduced into different starting populations can give the same leukaemia. HSCs and committed granulocyte/macrophage progenitors transduced with one MLL fusion were both immortalised and produced AML with similar latency and an identical immunophenotype and expression profile. If starting population is not recoverable from the resulting disease, an LMPP 'prerequisite' shown by transduction and transplantation is not evidence that the human leukaemia starts there. Read: abstract only. Cozzio A et al. 2003
    • Transduction-and-transplant readouts in this system are prone to scoring engraftment advantage as transformation. MLL-AF4 in human cord-blood CD34+ HSPCs enhanced repopulation and clonogenicity but did not initiate leukaemia; adding activated KRAS still did not, producing extramedullary and CNS-homing phenotypes instead. What survives transduction and engrafts is not necessarily what transforms. Read: abstract only. Montes R et al. 2011
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In an entirely murine Mll-AF4 in vivo model, miR-130b overexpression produced a mixed B-cell-precursor/myeloid leukaemia propagated by the LMPP population whereas miR-128a overexpression produced a pro-B ALL maintained by an expanded Il7r-positive Kit-positive blast population; both recapitulated human features including CNS infiltration and an MLL-AF4 expression signature.

    miR-130b and miR-128a are essential lineage-specific codrivers of t(4;11) MLL-AF4 acute leukemia · Malouf C, Antunes ETB, O'Dwyer M, Jakobczyk H, Sahm F, Landua SL, Anderson RA, Soufi A, Halsey C, Ottersbach K · Blood 2021
    PMID 34111240DOI
    What this evidence cannot show

    Overexpression of a microRNA is a strong, non-physiological perturbation, so the lineage outcome may reflect the dose of the co-driver rather than a choice the endogenous disease makes. All in vivo work is murine Mll-AF4. The mapping from a mouse LMPP gate to the human multilymphoid progenitor is by analogy.

    Evidence that argues against this (1)
    • The founding property of the murine LMPP — loss of megakaryocyte/erythroid potential on acquiring Flk2/Flt3 — was contested directly. Quantitative in vivo assays at population and clonal level showed Flk2+ early progenitors retain MegE potential, and the authors concluded that Flk2 expression is not at the expense of full multipotency. 'Propagated by the LMPP population' therefore names a sort gate whose defining lineage restriction is disputed. Read: abstract only. Forsberg EC et al. 2006
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Cell Ontology term CL:0000936 (early lymphoid progenitor) lists LMPP, lymphoid-primed multipotent progenitor, ELP and GMLP as synonyms and defines the cell as bone-marrow-resident, lineage-negative, Kit-positive, Sca-1-positive, FLT3-positive, CD34-positive, CD150-negative and glycophorin-A-negative, with an E2A-positive Ikaros-positive but EBF-negative Pax5-negative PU.1-negative transcription factor state, giving rise to B, T, NK and dendritic cells.

    Cell Ontology term CL:0000936 early lymphoid progenitor, retrieved via the EMBL-EBI Ontology Lookup Service OLS4 API · EMBL-EBI Ontology Lookup Service; Cell Ontology consortium · Database 2026
    What this evidence cannot show

    The ontology record itself states that the markers are associated with mouse cells, so the marker panel should not be applied directly to human flow cytometry. Ontology definitions lag the single-cell literature and treat a continuum as a discrete class.

    Evidence that argues against this (2)
    • The ontology's marker definition encodes a model that single-cell data do not support. Integrating index sorting, transcriptomes and function showed human HSCs acquiring lineage biases gradually along multiple directions without passing through discrete hierarchically organised progenitor populations — a continuum of low-primed HSPCs — so a Lin-Kit+Sca1+Flt3+CD34+ compartment is a region of that continuum, not a cell type. Read: abstract only. Velten L et al. 2017
    • The Cell Ontology definition quoted here is a mouse definition (Kit, Sca-1), and the human hierarchy it is being used to describe differs across ontogeny: oligopotent intermediates are abundant in fetal liver and largely absent in adult marrow. Mapping mouse LMPP onto human fetal LMPP is an assumption the ontology does not license. Read: abstract only. Notta F et al. 2016
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Fetal liver Mll-AF4-positive Lineage-negative Sca1-positive Kit-positive cells proliferate and upregulate a subset of MLL-AF4 signature genes and Toll-like receptor pathway genes on direct exposure to poly(I:C) or LPS, but a single maternal immune activation with poly(I:C) does not initiate leukaemia; ageing Mll-AF4-positive mice instead lose their B-lymphoid bias and show reduced stem and multipotent progenitor frequencies.

    PMID 31381950DOI
    What this evidence cannot show

    A single dose of poly(I:C) is a weak proxy for the repeated real-world infectious exposures that the delayed-infection hypothesis invokes, so the negative in vivo result is not evidence against infection-driven progression in general. Mouse pre-leukaemia model.

    Evidence that argues against this (1)
    • A single maternal immune activation not producing leukaemia is a negative from one exposure protocol at one gestational time. Proliferation and partial signature upregulation on direct poly(I:C) or LPS exposure is measured on sorted cells in culture, where TLR agonists activate almost any progenitor, so the specificity to the fusion signature requires a matched non-fusion control the entry does not cite. Oncofusion knowledge graph (internal analysis) 2026
Connected to