miR-9 — the most MLL-specific microRNA in AML, encoded at three separate loci nobody has disambiguated
also called miR-9 · hsa-miR-9-5p · miR-9* · hsa-miR-9-3p · MIR9-1 · MIR9-2 · MIR9-3
miR-9 is a small RNA that is switched on more specifically by KMT2A (MLL) fusions than almost any other microRNA, and blocking it kills KMT2A-rearranged leukemia cells in the laboratory. The complication is that the identical mature RNA is made from three different places in the genome, and nobody has worked out which one the fusion actually turns on.
The same thing, for a scientist
The mature miR-9 sequence is encoded at three unlinked human loci: MIR9-1 (ENSG00000207933; miRBase MI0000466; GRCh38 chr1:156,420,341-156,420,429; 1q22), MIR9-2 (ENSG00000284447; MI0000467; chr5:88,666,853-88,666,939; 5q14.3) and MIR9-3 (ENSG00000284329; MI0000468; chr15:89,368,017-89,368,106; 15q26.1) [E5]. The KMT2A-r evidence is real and unusually strong for a non-coding element. Chen P et al. found miR-9 to be **the most specifically up-regulated miRNA in MLL-rearranged AML** relative to both normal controls and non-MLL-rearranged AML in a genome-wide screen; showed it is a direct target of MLL fusion proteins; and showed that an antagomiR against endogenous miR-9 inhibits growth/viability and promotes apoptosis in human MLL-rearranged AML cells, that an anti-miRNA sponge inhibits MLL-fusion-induced immortalization/transformation of mouse bone-marrow progenitors, and that forced miR-9 promotes MLL-fusion leukemogenesis in vivo [E1]. This was independently corroborated in a **647-patient** primary human AML cohort where miR-9/9* was up-regulated in most cases with **the highest expression in MLL-related leukemias harbouring 11q23 abnormalities and the lowest in t(8;21) and biallelic CEBPA**, and where the mechanism ran through post-transcriptional repression of ERG [E2]. The opposing result is real but narrower than it looks: in 90 de-novo paediatric AMLs, forced miR-9 acted as a tumour suppressor inducing monocytic differentiation — but the authors state explicitly that these functions were **restricted to t(8;21)** and **were not evident in AML blasts from patients with MLL rearrangements** [E3]. So the literature is context-dependent rather than contradictory, and the KMT2A-r reading is the well-supported one. Every functional experiment is nonetheless **myeloid AML**; there is no KMT2A::AFF1 B-ALL experiment, and PubMed returns zero records for miR-9 with KMT2A rearrangement by those index terms [E6].
The case that miR-9 is a genuine KMT2A-fusion dependency1 deeper
Chen P et al. is, on the strength of its design, the most complete functional dossier for any non-coding element in this disease 1. It contains all four of the things a dependency claim needs:
- Specificity. Genome-wide miRNA profiling put miR-9 as the single most specifically up-regulated miRNA in MLL-rearranged AML versus both normal control and non-MLL-rearranged AML. Not 'one of several' — the most specific.
- Direct regulation. miR-9 is a direct target of MLL fusion proteins and is significantly up-regulated by them in human and mouse haematopoietic stem/progenitor cells.
- Loss of function. An antagomiR against endogenous miR-9 significantly inhibited cell growth and viability and promoted apoptosis in human MLL-rearranged AML cells. An anti-miRNA sponge significantly inhibited MLL-fusion-induced immortalization/transformation of normal mouse bone-marrow progenitors in vitro.
- Gain of function in vivo. Forced miR-9 significantly promoted MLL-fusion-mediated leukemogenesis in vivo.
The independent expression confirmation is the strongest number in this node: across 647 primary human AMLs, miR-9/9* was up-regulated in most cases, with expression varying by subtype and the highest levels in MLL-related leukemias with 11q23 abnormalities 2. A 647-patient cohort putting 11q23 at the top of the distribution is a far more durable finding than any single mechanism paper.
The ERG convergence — and why it should make the lab uneasy
Two independent routes in this graph now converge on the same target. miR-9/9* blocks neutrophil differentiation by post-transcriptionally down-regulating ERG, and upregulating ERG in 32D cells rescued the miR-9/9*-induced differentiation block 2. Separately, miR-196a/b — also elevated in KMT2A-r leukemia — bind and repress the ERG 3'UTR (see noncoding-mir-196b).
So two KMT2A-fusion-associated microRNAs both repress ERG, and in at least one case that repression is sufficient to explain a differentiation block. That is a coherent, non-coding route to the central phenotype of this disease, running entirely outside the protein-coding fusion-target list.
It should make the lab uneasy for a specific reason: the differentiation block in KMT2A-r leukemia may be partly maintained by microRNA-mediated repression of ERG, which a menin inhibitor does not obviously reverse. If miR-9 and miR-196b remain elevated under revumenib while HOXA9 falls, the differentiation block could persist through a mechanism the drug does not touch. Nobody has measured this.
The three-locus problem — the real gap, and it is an experimental one1 deeper
This is the point the literature has simply not addressed, and it undermines the precision of every published claim.
The identical mature miR-9-5p sequence is produced from MIR9-1 on 1q22, MIR9-2 on 5q14.3 and MIR9-3 on 15q26.1 5. Every quantification in 1, 2 and 3 measures the mature species, so:
- 'miR-9 is up-regulated in MLL-rearranged AML' does not say which locus is transcribed.
- 'miR-9 is a direct target of MLL fusion proteins' 1 necessarily refers to fusion occupancy at one or more specific loci, but the mature-miRNA assay cannot attribute the output to the occupied locus.
- An antagomiR knocks down all three simultaneously, so the loss-of-function experiments cannot localise the dependency either.
The missing experiment is trivial by modern standards: look at the existing KMT2A-AFF1 and KMT2A-MLLT3 ChIP-seq tracks (Kerry 2017; Harman 2021; Godfrey 2021) at chr1:156.42 Mb, chr5:88.67 Mb and chr15:89.37 Mb, and measure pri-miR-9-1/-2/-3 by locus-specific qPCR. No new samples are needed.
A specific, checkable hypothesis: MIR9-2 sits next to MEF2C
MIR9-2 is at GRCh38 chr5:88,666,853-88,666,939 (minus strand). MEF2C is at chr5:88,717,117-88,904,257 (minus strand). The 3' end of MEF2C is therefore about 50 kb from MIR9-2 5.
MEF2C is a documented KMT2A-fusion target locus and a driver of a distinct high-risk ALL subtype. If the fusion opens the MEF2C domain, MIR9-2 is close enough to plausibly fall inside the same regulatory neighbourhood, which would give a concrete, locus-specific answer to 'which miR-9 gene does the fusion turn on' — and would tie the most MLL-specific microRNA to one of the fusion's best-established target loci.
This is our own inference from coordinate arithmetic, not a published claim, and the distance calculation is from gene-level annotation. It is offered as a hypothesis with a cheap test: check whether MIR9-2 lies within the same TAD as MEF2C in KMT2A-r Hi-C data, and whether locus-specific pri-miR-9-2 tracks MEF2C expression across KMT2A-r samples. It could be wrong for an obvious reason — 50 kb is comfortably far enough to sit across a boundary.
Where the opposing evidence really points
The spec asked whether the strong evidence sits in a different leukemia context. The honest answer, after reading the abstracts, is no for KMT2A-r, but the opposing paper is frequently mis-summarised.
Emmrich S et al. profiled miRNAs in 90 cytogenetically characterised de-novo paediatric AML cases and identified four miRNAs specifically down-regulated in MLL-rearranged, t(8;21) or inv(16) AML. Forced miR-9 reduced leukemic growth and induced monocytic differentiation of t(8;21) AML cell lines in vitro and in vivo, acting with let-7 to repress the LIN28B/HMGA2 axis. The paper's own conclusion is that miR-9 is a tumour suppressor 'in a stringent cell context-dependent manner', and it states that the tumour-suppressive functions were restricted to t(8;21) blasts and were not evident in AML blasts from patients with MLL rearrangements 3.
That is consistent with 2, where miR-9 was lowest in t(8;21) and highest in 11q23/MLL. The two datasets agree on the expression distribution. So the field's position is: miR-9 is high and oncogenic in KMT2A-r AML, low and tumour-suppressive in t(8;21) AML. Anyone citing Emmrich as evidence against a miR-9 role in KMT2A-r leukemia has read the title and not the result.
The genuinely limiting caveat is different and applies to all three papers: it is all AML. Not one experiment addresses KMT2A::AFF1 B-lymphoid leukemia, and infant KMT2A::AFF1 ALL is the project's centre of gravity. A PubMed search for miR-9 with KMT2A rearrangement returns zero records 6.
One further mechanistic link is worth recording: in THP-1 cells (KMT2A::MLLT3), the pro-differentiative miR-424-503 polycistron is under repression by the MLL-MLLT3 fusion, and miR-424/503 in turn down-regulate 'anti-differentiative' miR-9 by targeting a site in its primary transcript 4. If that holds, the fusion raises miR-9 partly indirectly, by removing a repressor of it — which would explain why fusion occupancy at the miR-9 loci has been hard to pin down.
Intervention read
Of the nine elements in this beat, miR-9 has the strongest published intervention data — and the most alarming toxicity profile for this particular patient population.
For it:
- An antagomiR against endogenous miR-9 already inhibits growth and induces apoptosis in human MLL-rearranged AML cells 1. That is a completed proof-of-concept loss-of-function experiment with the correct sign, which no other element in this node has.
- LNA antimiR chemistry is clinically mature.
- The sign is unambiguous (unlike miR-196b): lowering miR-9 should relieve the ERG-mediated differentiation block 2 and reduce viability 1. Both point the same way.
Against it, and this is decisive for infant disease:
- miR-9 is one of the most important microRNAs in mammalian neurogenesis and brain development. The dominant miR-9 literature is neural, not haematological. KMT2A::AFF1 leukemia presents in infants, often under 6 months, whose central nervous system is actively developing and whose disease frequently involves the CNS, so CNS-penetrant therapy is often deliberate. A systemically delivered antimiR-9 in a 4-month-old is a neurodevelopmental risk that cannot be dismissed as theoretical.
- The three-locus problem means an antimiR cannot be locus-selective, so there is no way to spare the neural pool while hitting the haematopoietic one with seed-directed chemistry.
- No in vivo antagomiR efficacy experiment (as opposed to sponge/transformation assays) has been published.
Verdict: the best-validated non-coding target in KMT2A-r leukemia and simultaneously the one with the most serious on-target developmental-toxicity objection. Grade: moderate tractability, high risk. The first step is not chemistry — it is to establish which locus the fusion drives and whether miR-9 matters in B-lymphoid KMT2A::AFF1 at all.
| element symbol | miR-9 (MIR9-1 / MIR9-2 / MIR9-3) |
|---|---|
| class | miRNA |
| genome build | GRCh38 |
| multi locus | true |
| locus driving expression in kmt2a r | unknown — this is the central experimental gap; all published assays measure the mature species, which is identical across the three loci |
| direction of change in kmt2a r | up |
| b lymphoid kmt2a aff1 evidence | none published |
mature species
[ "hsa-miR-9-5p", "hsa-miR-9-3p (miR-9*)" ]
loci — the numbers behind the chart
[
{
"gene": "MIR9-1",
"locus": "chr1:156,420,341-156,420,429",
"chromosome": "1",
"start": 156420341,
"end": 156420429,
"strand": "-",
"cytoband": "1q22",
"ensembl_gene_id": "ENSG00000207933",
"mirbase_accession": "MI0000466",
"rnacentral_id": "URS000075C54E",
"hgnc_id": "HGNC:31641",
"entrez_id": "407046"
},
{
"gene": "MIR9-2",
"locus": "chr5:88,666,853-88,666,939",
"chromosome": "5",
"start": 88666853,
"end": 88666939,
"strand": "-",
"cytoband": "5q14.3",
"ensembl_gene_id": "ENSG00000284447",
"mirbase_accession": "MI0000467",
"rnacentral_id": "URS00003FA6CD",
"hgnc_id": "HGNC:31642",
"entrez_id": "407047",
"note": "approximately 50 kb from the 3' end of MEF2C (chr5:88,717,117-88,904,257), a documented KMT2A-fusion target locus — our own coordinate inference, untested"
},
{
"gene": "MIR9-3",
"locus": "chr15:89,368,017-89,368,106",
"chromosome": "15",
"start": 89368017,
"end": 89368106,
"strand": "+",
"cytoband": "15q26.1",
"ensembl_gene_id": "ENSG00000284329",
"mirbase_accession": "MI0000468",
"rnacentral_id": "URS00005D97FE",
"hgnc_id": "HGNC:31646",
"entrez_id": "407051"
}
]validated or supported targets
[
{
"gene": "ERG",
"direction": "repressed",
"pmid": "26174629",
"note": "ERG upregulation rescues the miR-9/9*-induced neutrophil differentiation block",
"title": "Aberrant expression of miR-9/9* in myeloid progenitors inhibits neutrophil differentiation by post-transcriptional regulation of ERG.",
"authors": "Nowek K, Sun SM, Bullinger L, et al.",
"venue": "Leukemia",
"year": 2016,
"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."
},
{
"gene": "LIN28B / HMGA2 axis",
"direction": "repressed",
"pmid": "24270738",
"note": "in cooperation with let-7; demonstrated in t(8;21), not in MLL-rearranged cells",
"title": "miR-9 is a tumor suppressor in pediatric AML with t(8;21).",
"authors": "Emmrich S, Katsman-Kuipers JE, Henke K, et al.",
"venue": "Leukemia",
"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."
}
]cohorts with direction evidence — the numbers behind the chart
[
{
"cohort": "647 primary human AML (adult; HOVON/Ulm)",
"finding": "miR-9 and miR-9* up-regulated in most cases; HIGHEST in MLL-related leukemias with 11q23 abnormalities; LOWEST in t(8;21) and biallelic CEBPA",
"pmid": "26174629",
"n": 647,
"title": "Aberrant expression of miR-9/9* in myeloid progenitors inhibits neutrophil differentiation by post-transcriptional regulation of ERG.",
"authors": "Nowek K, Sun SM, Bullinger L, et al.",
"venue": "Leukemia",
"year": 2016,
"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."
},
{
"cohort": "genome-wide miRNA screen, MLL-rearranged vs non-MLL-rearranged AML vs normal",
"finding": "miR-9 was the single most specifically up-regulated miRNA in MLL-rearranged AML",
"pmid": "23798388",
"n": "not stated in abstract",
"title": "miR-9 is an essential oncogenic microRNA specifically overexpressed in mixed lineage leukemia-rearranged leukemia.",
"authors": "Chen P, Price C, Li Z, et al.",
"venue": "Proc Natl Acad Sci U S A",
"year": 2013,
"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."
},
{
"cohort": "90 de-novo paediatric AML (DCOG/AML-BFM/international)",
"finding": "four miRNAs downregulated in MLL-rearranged, t(8;21) or inv(16); miR-9 tumour-suppressive function restricted to t(8;21) and absent in MLL-rearranged blasts",
"pmid": "24270738",
"n": 90,
"title": "miR-9 is a tumor suppressor in pediatric AML with t(8;21).",
"authors": "Emmrich S, Katsman-Kuipers JE, Henke K, et al.",
"venue": "Leukemia",
"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."
}
]intervention assessment
{
"tractability": "moderate — the best-validated non-coding target in KMT2A-r leukemia, but with a serious on-target developmental-toxicity objection specific to infants",
"best_modality": "LNA antimiR / antagomiR (loss-of-function proof of concept already published in human MLL-rearranged AML cells)",
"sign_of_effect": "unambiguous and favourable: lowering miR-9 reduces viability and promotes apoptosis [E1] and should relieve ERG-mediated differentiation block [E2]",
"modalities_ruled_out": [
"small molecule (no precedent for selective miRNA biogenesis inhibition at this locus)",
"locus-selective knockdown by seed-directed chemistry (impossible — three loci, identical mature sequence)"
],
"critical_toxicity_concern": "miR-9 is a central regulator of mammalian neurogenesis and brain development. KMT2A::AFF1 leukemia presents in infants under 12 months with frequent CNS involvement and deliberately CNS-directed therapy, so systemic antimiR-9 in a developing brain is a first-order neurodevelopmental risk, not a theoretical one.",
"published_in_vivo_efficacy": "no — the in vivo experiment in [E1] is gain-of-function; no in vivo antagomiR treatment result exists",
"first_experiment": "locus-specific pri-miR-9-1/-2/-3 qPCR plus inspection of existing KMT2A-AFF1/KMT2A-MLLT3 ChIP-seq at chr1:156.42 Mb, chr5:88.67 Mb and chr15:89.37 Mb, and measurement of mature miR-9 in KMT2A::AFF1 B-ALL lines and PDX. No new samples required."
}- Developmental toxicity, and it is the serious one: miR-9 is essential for neurogenesis, the patients are infants, and the therapy would be systemic. This is on-target, unavoidable with seed-directed chemistry, and specific to the population the lab cares most about.
- Compensation across loci: three genomic sources of identical mature miR-9 means transcriptional silencing of one locus is buffered by the other two, so only post-transcriptional (antimiR) approaches can work — and those hit the neural pool too.
- Redundancy at the target: ERG is repressed by both miR-9 and miR-196a/b, so removing miR-9 alone may not restore ERG enough to break the differentiation block.
- Essential-function toxicity in blood: miR-9/9* blocks neutrophil differentiation in normal murine progenitors, meaning it is part of normal myeloid differentiation control; perturbing it perturbs normal granulopoiesis.
- Resistance: since the fusion may drive miR-9 indirectly by repressing miR-424/503, a cell could restore miR-9 by silencing that repressor arm — a route to antimiR resistance with no mutational signature to detect it.
- Which of the three miR-9 loci does the KMT2A fusion actually drive? No published work distinguishes them, and every dependency and expression claim in the field is locus-agnostic as a result. Answerable from existing ChIP-seq plus locus-specific qPCR.
- Does MIR9-2 share a TAD with MEF2C in KMT2A-r cells? If so, the most MLL-specific microRNA in AML would be physically coupled to one of the fusion's best-established target loci, which would be a substantial finding. Our 50 kb distance estimate makes this worth checking and easily wrong.
- Is miR-9 elevated and required in KMT2A::AFF1 B-lymphoid leukemia? Every functional experiment in the literature is myeloid AML, and the project's centre of gravity is infant B-ALL.
- Do miR-9 and miR-196b remain elevated under menin inhibition while HOXA9 falls? If they do, ERG repression and the differentiation block could persist through a route revumenib does not touch — a candidate mechanism of incomplete response that nobody has measured.
- Can miR-9 be antagonised in the haematopoietic compartment without exposing the infant brain? Without a locus- or tissue-selective approach this is the question that decides whether the target is usable at all.
- Is the reported repression of the miR-9 primary transcript by miR-424/503 real? A miRNA targeting a pri-miRNA is an unusual claim from a single cell line and would, if true, mean the fusion raises miR-9 indirectly by removing a repressor.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In a genome-wide miRNA expression assay miR-9 was the most specifically up-regulated miRNA in MLL-rearranged AML compared with both normal control and non-MLL-rearranged AML; miR-9 is a direct target of MLL fusion proteins and is up-regulated by them in human and mouse haematopoietic stem/progenitor cells; depletion of endogenous miR-9 by antagomiR significantly inhibited growth and viability and promoted apoptosis in human MLL-rearranged AML cells; an anti-miRNA sponge inhibited and forced miR-9 promoted MLL-fusion-induced immortalization/transformation of mouse bone-marrow progenitors; and forced miR-9 significantly promoted MLL-fusion-mediated leukemogenesis in vivo.
miR-9 is an essential oncogenic microRNA specifically overexpressed in mixed lineage leukemia-rearranged leukemia · Chen P, Price C, Li Z, Li Y, Cao D, Wiley A, He C, Gurbuxani S, Kunjamma RB, Huang H, Jiang X, Arnovitz S, Xu M, Hong GM, Elkahloun AG, Neilly MB, Wunderlich M, Larson RA, Le Beau MM, Mulloy JC, Liu PP, Rowley JD, Chen J · Proceedings of the National Academy of Sciences USA 110(28):11511-6 2013What this evidence cannot show
Entirely myeloid AML: human MLL-rearranged AML cell lines and mouse bone-marrow progenitors. No KMT2A::AFF1 B-lymphoid experiment. 'Direct target of MLL fusion proteins' cannot be attributed to one of the three miR-9 loci by a mature-miRNA assay, and the abstract does not state which locus was examined by ChIP. The in vivo experiment is gain-of-function (forced miR-9 accelerates leukemogenesis), not loss-of-function; there is no in vivo antagomiR efficacy result. Cohort sizes for the profiling are not given in the abstract. The putative target list is inferred from inverse expression correlation, not validated.
Evidence that argues against this (1)
- In paediatric AML, forced miR-9 acts as a tumour suppressor that induces monocytic differentiation and reduces leukemic growth — but the authors state this is restricted to t(8;21) and is NOT evident in MLL-rearranged blasts, so the polarity conflict is context-dependent rather than a direct contradiction of the KMT2A-r claim. Emmrich S et al. 2014
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Across a cohort of 647 primary human AMLs, miR-9 and miR-9* were significantly up-regulated in most cases with expression varying by subtype — highest in MLL-related leukemias harbouring 11q23 abnormalities and lowest in AML with t(8;21) and biallelic CEBPA mutations; enforced miR-9/9* expression blocked normal neutrophil development in 32D cells and in primary murine lineage-negative bone-marrow cells, ERG was the only common predicted target of AMLs with high miR-9/9* and of 32D-miR-9/9* cells, and upregulation of ERG rescued the miR-9/9*-induced block in neutrophil differentiation.
Aberrant expression of miR-9/9* in myeloid progenitors inhibits neutrophil differentiation by post-transcriptional regulation of ERG · Nowek K, Sun SM, Bullinger L, Bindels EMJ, Exalto C, Dijkstra MK, van Lom K, Döhner H, Erkeland SJ, Löwenberg B, Jongen-Lavrencic M · Leukemia 30(1):229-37 2016What this evidence cannot show
Adult AML (HOVON/Ulm cohorts); '11q23 abnormalities' is a cytogenetic category that does not resolve KMT2A fusion partner and includes non-KMT2A 11q23 events. The functional work is in murine 32D cells and lineage-negative marrow, i.e. a normal-differentiation model rather than a leukemia-maintenance model. The mature-miRNA assay cannot distinguish the three miR-9 loci. ERG was identified as 'the only common potential target' by expression profiling, so the ERG mechanism is supported by rescue but not by direct 3'UTR mutagenesis in this paper. No paediatric or B-lymphoid data.
Evidence that argues against this (1)
- miR-196a/b, also elevated in KMT2A-rearranged leukemia, independently repress ERG via its 3'UTR, so ERG repression is not specific to miR-9 and the relative contribution of each microRNA to the differentiation block is unresolved. Coskun E et al. 2011
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In a miRNA expression profiling of 90 cytogenetically characterised de-novo paediatric AML cases, four miRNAs were specifically downregulated in MLL-rearranged, t(8;21) or inv(16) AML; forced miR-9 expression reduced leukemic growth and induced monocytic differentiation of t(8;21) AML cell lines in vitro and in vivo through cooperation with let-7 to repress the LIN28B/HMGA2 axis, but these tumour-suppressive functions were specifically restricted to t(8;21) cells and primary blasts and were NOT evident in AML blasts from patients with MLL rearrangements.
miR-9 is a tumor suppressor in pediatric AML with t(8;21) · Emmrich S, Katsman-Kuipers JE, Henke K, Khatib ME, Jammal R, Engeland F, Dasci F, Zwaan CM, den Boer ML, Verboon L, Stary J, Baruchel A, de Haas V, Danen-van Oorschot AA, Fornerod M, Pieters R, Reinhardt D, Klusmann JH, van den Heuvel-Eibrink MM · Leukemia 28(5):1022-32 2014What this evidence cannot show
This paper is routinely mis-cited as evidence that miR-9 is a tumour suppressor in MLL-rearranged leukemia. It is not: the authors state the effects were restricted to t(8;21) and absent in MLL-rearranged blasts. The abstract does not specify which of the four downregulated miRNAs belongs to which cytogenetic group, so it cannot be used to claim miR-9 is low in MLL-rearranged paediatric AML. n=90 across multiple cytogenetic groups means individual subgroups are small. Paediatric AML only, not ALL.
Evidence that argues against this (1)
- The much larger 647-patient adult AML cohort found miR-9 HIGHEST in 11q23/MLL-related leukemias and LOWEST in t(8;21), which is concordant rather than discordant with a t(8;21)-restricted tumour-suppressor role. Nowek K et al. 2016
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In THP-1 cells, which carry the MLL-MLLT3 (KMT2A::MLLT3) fusion, the pro-differentiative polycistronic precursor mir-424-503 is under repression by the MLL-MLLT3 leukemogenic fusion, and miR-424 and miR-503 down-regulate the anti-differentiative miR-9 by targeting a site in its primary transcript.
Induction of microRNAs, mir-155, mir-222, mir-424 and mir-503, promotes monocytic differentiation through combinatorial regulation · Forrest ARR, Kanamori-Katayama M, Tomaru Y, Lassmann T, Ninomiya N, Takahashi Y, de Hoon MJL, Kubosaki A, Kaiho A, Suzuki M, Yasuda J, Kawai J, Hayashizaki Y, Hume DA, Suzuki H · Leukemia 24(2):460-6 2010What this evidence cannot show
Single cell line (THP-1) in a phorbol-ester differentiation paradigm, part of the FANTOM4 consortium effort rather than a dedicated mechanism study. The claim that miR-424/503 target a site in the miR-9 primary transcript is unusual (a miRNA targeting a pri-miRNA) and has not been independently replicated. The 'repression by MLL-MLLT3' conclusion is inferred within a single line. The relevance to KMT2A::AFF1 B-ALL is unestablished.
Evidence that argues against this (1)
- THP-1 is KMT2A::MLLT3 and monocytic, and the beat's index disease is KMT2A::AFF1 B-lymphoid. The entry's own next item records that no indexed study examines miR-9 in KMT2A::AFF1 at all, so this mechanism is being carried across both a partner and a lineage boundary on a single cell line. Oncofusion knowledge graph (internal analysis) 2026
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The mature miR-9 sequence is encoded at three unlinked loci: MIR9-1 (ENSG00000207933, miRBase MI0000466, RNAcentral URS000075C54E, HGNC:31641, GRCh38 chr1:156,420,341-156,420,429, minus strand, 1q22); MIR9-2 (ENSG00000284447, MI0000467, URS00003FA6CD, HGNC:31642, chr5:88,666,853-88,666,939, minus strand, 5q14.3); and MIR9-3 (ENSG00000284329, MI0000468, URS00005D97FE, HGNC:31646, chr15:89,368,017-89,368,106, plus strand, 15q26.1). MIR9-2 lies approximately 50 kb from the 3' end of MEF2C (chr5:88,717,117-88,904,257, minus strand).
HGNC REST API records for MIR9-1, MIR9-2, MIR9-3 and Ensembl REST API gene lookups (MIR9-1/2/3, MEF2C), assembly GRCh38 · HUGO Gene Nomenclature Committee; Ensembl / EMBL-EBI · HGNC / Ensembl 2026What this evidence cannot show
The ~50 kb MIR9-2-to-MEF2C distance is computed by us from gene-level Ensembl coordinates and is a linear genomic distance only; it carries no implication about shared TAD membership, which is the question that actually matters and which has not been checked in KMT2A-r cells. Coordinates are for the precursor hairpins. The three loci produce identical mature miR-9-5p, which is the source of the attribution problem described in this node.
Evidence that argues against this (1)
- Three unlinked loci encoding an identical mature sequence is precisely why miR-9 functional work is hard to interpret: no expression measurement of the mature microRNA attributes it to a locus, knocking out one locus leaves two, and the three are differently regulated. Claims about miR-9 regulation by a fusion cannot be assigned to a target the fusion could plausibly reach. Ensembl, EMBL-EBI 2026
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
As of 2026-09-18 a PubMed search for miR-9 with MLL and leukemia returns 4 records, and a search for miR-9 with KMT2A rearrangement and leukemia returns zero; no indexed study examines miR-9 in KMT2A::AFF1 or in B-lymphoid KMT2A-rearranged leukemia.
PubMed database queries 'miR-9 AND MLL AND leukemia' (4 records) and 'miR-9 AND KMT2A AND rearranged AND leukemia' (0 records), executed 2026-09-18 · National Library of Medicine, PubMed · PubMed 2026What this evidence cannot show
A symbol-plus-keyword search is a blunt instrument: miR-9 data could be reported in a supplementary miRNA table of an infant-ALL paper without appearing in the title or abstract, and older papers use 'MLL' rather than 'KMT2A'. The 4-record count for 'miR-9 AND MLL' is implausibly small relative to the field and reflects PubMed's term translation rather than the true literature size, so this evidence item should be read as indicative of an ALL/B-lineage gap, not as a census.
Evidence that argues against this (1)
- Zero indexed records is an absence measured through a search interface. miR-9 papers routinely name the locus (MIR9-1/2/3) rather than the mature microRNA, "MLL" and "KMT2A" partition the literature by era, and the project's own rule is that alias searches fail in both directions. This is a reason to look harder, not an established gap. NCBI 2026
- regulates →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
the most KMT2A-specific microRNA in myeloid leukemia
miR-9 is among the microRNAs reported silenced by CpG hypermethylation in t(4;11)
miR-9 cooperates with let-7 to repress the LIN28B/HMGA2 axis in t(8;21) AML - and does NOT do so in MLL-rearranged blasts