HOTAIRM1 lncRNA — the myeloid-differentiation transcript at the 3' end of HOXA, with a KMT2A-r evidence gap
also called HOTAIRM1 · HOXA transcript antisense RNA myeloid-specific 1 · HOXAAS · NCRNA00061
HOTAIRM1 is a non-coding RNA made from the opposite strand at the far end of the HOXA gene cluster, and it helps blood cells mature into granulocytes. It is often assumed to matter in KMT2A-rearranged leukemia because it sits in the HOXA cluster, but the direct evidence for that specific link does not exist yet.
The same thing, for a scientist
HOTAIRM1 (ENSG00000233429; GRCh38 chr7:27,095,647-27,100,265, + strand) is a myeloid-specific lncRNA transcribed antisense at the HOXA1/HOXA2 boundary from the same CpG island that carries the HOXA1 transcription start site [E1]. It is induced during retinoic-acid-driven granulocytic differentiation via the RA receptor; knockdown blunts RA-induced HOXA1 and HOXA4 and selectively attenuates induction of the myeloid differentiation genes CD11b and CD18, without affecting the more distal HOXA genes [E1]. Mechanistically, distinct HOTAIRM1 isoforms preferentially associate with either UTX/MLL or PRC2 complexes to tune the activating/silencing balance at the HOXA bivalent domain, and HOTAIRM1 drives physical dissociation of chromatin loops at the cluster-proximal end, delaying UTX recruitment and central HOXA transcription — with the important caveat that in a different cell type HOTAIRM1 activated proximal HOXA genes with no chromatin-conformation change at all [E2]. In AML it is a prognostic marker: across 241 AML patients high HOTAIRM1 independently predicted shorter overall survival in the intermediate-risk group [E3]. **The honest bottom line is that there is no KMT2A-rearranged-specific HOTAIRM1 experiment in the literature.** A PubMed search for HOTAIRM1 with leukemia returns 23 records and none of them isolates a KMT2A-r cohort or model; the KMT2A-r association is inferred from HOXA-locus co-regulation, not shown.
What HOTAIRM1 does in myeloid cells1 deeper
The founding paper is precise, and the precision is worth preserving because reviews tend to inflate it.
Zhang X et al. identified an intergenic transcript between HOXA1 and HOXA2 with myeloid-specific expression, up-regulated during granulocytic differentiation, transcribed antisense to the HOXA genes and originating from the same CpG island that embeds the HOXA1 start site. It has no long ORF and does not associate with polyribosomes, so it is a genuine non-coding RNA 1.
The functional result was selective, not global: knockdown quantitatively blunted RA-induced expression of HOXA1 and HOXA4, and selectively attenuated induction of CD11b (ITGAM) and CD18 (ITGB2) transcripts, but did not noticeably affect the more distal HOXA genes 1. That last clause is the one most often dropped in citation. HOTAIRM1 is a proximal-HOXA regulator. HOXA9 and HOXA10 — the genes that actually matter for KMT2A-fusion leukemia — are at the distal end and were not affected.
Upstream, HOTAIRM1 is a direct PU.1 target: PU.1 constitutively binds the HOTAIRM1 regulatory region, two PU.1 motifs around +1100 bp downstream of the TSS mediate transactivation, and the low HOTAIRM1 in APL is attributable to reduced PU.1 rather than to direct repression by PML-RARA 4. So HOTAIRM1 expression tracks myeloid transcription-factor identity.
The chromatin-architecture mechanism, and its own internal contradiction
Wang XQD and Dostie J showed that distinct HOTAIRM1 variants preferentially associate with either UTX/MLL or PRC2 complexes, modulating activating and silencing marks at the HOXA bivalent domain, and that HOTAIRM1 contributes to physical dissociation of chromatin loops at the cluster-proximal end, which delays UTX recruitment and transcription of central HOXA genes 2.
The same paper reports that in a different cell type HOTAIRM1 caused overall proximal HOXA gene activation without any chromatin-conformation change 2. The authors report this honestly; it means the architectural mechanism is not a general property of the transcript. For the purposes of this graph: any claim that HOTAIRM1 organises HOXA topology in a B-lymphoid KMT2A::AFF1 blast is unsupported, because the cell types studied were myeloid/RA-induced and the result was already shown to be cell-type-contingent.
Note also the direct intersection with the project's core biology: HOTAIRM1 binds UTX/MLL complexes. The bivalent-domain node mechanism-bivalent-primed-chromatin-fusion-targets describes the chromatin state; HOTAIRM1 is one of the RNA-level inputs that tunes it.
The AML clinical data — real, quantified, and not about KMT2A-r2 deeper
Díaz-Beyá M et al. measured HOTAIRM1 in 241 AML patients and analysed prognosis in 215 intermediate-risk AML (IR-AML) patients 3:
- Lowest expression in acute promyelocytic leukemia (P < 0.001); highest in t(6;9) AML (P = 0.005).
- In 215 IR-AML: high HOTAIRM1 independently associated with shorter overall survival (OR 2.04, P = 0.001), shorter leukemia-free survival (OR 2.56, P < 0.001) and higher cumulative incidence of relapse (OR 1.67, P = 0.046).
- Retained independent prognostic value within the favourable molecular subgroup (OR 3.43, P = 0.009).
- Overexpressed in NPM1-mutated AML (P < 0.001) and prognostic within that group (OR 2.21, P = 0.01).
- Associated with a specific 33-microRNA signature that includes miR-196b (P < 0.001); combining HOTAIRM1 and miR-196b stratified 5-year OS at 24% vs 42% vs 70% (P = 0.004).
What is missing from that list is any KMT2A-r stratum. The subgroups reported are APL, t(6;9) and NPM1-mutant. KMT2A-rearranged AML is not an intermediate-risk genotype in most schemes and would largely have been excluded from the 215-patient prognostic analysis. The miR-196b co-signature is the only indirect hook to KMT2A-r biology, because miR-196b is a direct KMT2A-fusion target (see noncoding-mir-196b).
Stating the gap plainly
A PubMed query for HOTAIRM1 AND leukemia returned 23 records on 2026-09-18; a query adding MLL and KMT2A returned 0 5. Reading the 23 titles, the leukemia literature on this transcript is: APL/NB4 granulocytic differentiation (several papers), PML-RARA degradation via autophagy, PU.1 regulation, AML prognosis (Díaz-Beyá), glucocorticoid resistance, plus glioma and solid-tumour work.
There is no published experiment that knocks down HOTAIRM1 in a KMT2A-rearranged cell line or model. This is a genuine, recordable gap, and it is the kind the PI asked about: the HOXA-locus association creates an intuition of relevance that the experimental record does not support. Confidence for any KMT2A-r claim about HOTAIRM1 is weak and should be labelled as such wherever it appears downstream.
The one lead that would make this interesting for infant ALL
HOTAIRM1 was overexpressed in an acquired glucocorticoid-resistant leukemia cell model, where it binds the transcriptional-inhibitory region of ARHGAP18, represses it, and thereby raises RHOA/ROCK1 signalling and blocks apoptosis; ROCK1 inhibition restored glucocorticoid responsiveness, and AML1/RUNX1 directly activates the HOTAIRM1 promoter 6.
Why this matters here: prednisone poor response is a defining feature of KMT2A-r infant ALL, and RUNX1 is one of the core transcription factors in the KMT2A-AFF1 gene regulatory network identified by Harman et al. So there is a coherent, testable chain — KMT2A::AFF1 → RUNX1 → HOTAIRM1 → ARHGAP18 → RHOA/ROCK1 → steroid resistance — in which every link is published but no link has been tested in a KMT2A-r cell. That is a hypothesis, not a finding, and it is flagged as such. The cell model in 6 is not KMT2A-rearranged.
Intervention read
HOTAIRM1 is a poor primary target and a reasonable biomarker.
Against targeting it: its documented function is pro-differentiation in myeloid cells 1, 4, so knocking it down in a leukemia whose problem is a differentiation block is mechanistically backwards. High expression predicts worse outcome in IR-AML 3, but that is a correlation in a differentiation-state-confounded axis, and no causal leukemia-maintenance experiment exists.
The only coherent intervention hypothesis runs the other way: if the glucocorticoid-resistance mechanism 6 holds in KMT2A-r ALL, then ROCK1 inhibition (a small molecule, already clinically available in other indications) would be the intervention and HOTAIRM1 would be the biomarker that selects patients. That is a cheap, testable proposition and it does not require drugging an RNA at all.
| element symbol | HOTAIRM1 |
|---|---|
| class | lncRNA |
| genome build | GRCh38 |
| locus | chr7:27,095,647-27,100,265 |
| chromosome | 7 |
| start | 27095647 |
| end | 27100265 |
| strand | + |
| cytoband | 7p15.2 |
| ensembl gene ID | ENSG00000233429 |
| ensembl canonical transcript | ENST00000434063 |
| hgnc ID | HGNC:37117 |
| position in cluster | 3' (proximal) end of the HOXA cluster, antisense/divergent at the HOXA1-HOXA2 interval, sharing the HOXA1 CpG island |
| direction of change in kmt2a r | unknown |
| kmt2a r specific evidence | none published; PubMed 'HOTAIRM1 AND leukemia AND MLL AND KMT2A' = 0 records on 2026-09-18 |
protein partners
[ "UTX/KDM6A-MLL complex (isoform-dependent)", "PRC2 (isoform-dependent)", "AML1/RUNX1" ]
upstream regulators
[ "PU.1/SPI1", "retinoic acid receptor", "AML1/RUNX1" ]
downstream targets reported
[ "HOXA1", "HOXA4", "CD11b/ITGAM", "CD18/ITGB2", "ARHGAP18" ]
cohorts with direction evidence
[
{
"cohort": "CETLAM AML, n=241 (prognostic subset n=215 intermediate-risk)",
"finding": "lowest in APL (P<0.001), highest in t(6;9) (P=0.005), overexpressed in NPM1-mutant (P<0.001); high expression → shorter OS (OR 2.04, P=0.001). No KMT2A-r stratum reported.",
"pmid": "26436590",
"n": 241,
"title": "The lincRNA HOTAIRM1, located in the HOXA genomic region, is expressed in acute myeloid leukemia, impacts prognosis in patients in the intermediate-risk cytogenetic category, and is associated with a distinctive microRNA signature.",
"authors": "Díaz-Beyá M, Brunet S, Nomdedéu J, et al.",
"venue": "Oncotarget",
"year": 2015,
"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": "weak / not recommended as a direct target",
"best_modality": "none as a direct target. If the resistance axis in [E6] replicates in KMT2A-r cells, the actionable node is ROCK1 (small molecule) with HOTAIRM1 as a selection biomarker.",
"modalities_ruled_out": [
"small molecule against the RNA (no pocket)",
"ASO knockdown as a therapeutic (documented function is pro-differentiation, so knockdown is mechanistically backwards in a differentiation-blocked leukemia)"
],
"biomarker_value": "plausible prognostic biomarker in intermediate-risk AML per [E3]; unvalidated in KMT2A-r or in ALL",
"first_experiment": "measure HOTAIRM1 in KMT2A::AFF1 versus non-KMT2A-r infant/paediatric B-ALL RNA-seq (TARGET, Interfant) — this requires no new samples and would close the single largest gap in this node"
}- Polarity risk: HOTAIRM1's established function is pro-differentiation. Any therapy that lowers it could deepen the differentiation block that is the disease's core pathology.
- Essential-function toxicity: HOTAIRM1 is required for normal RA-driven granulocytic maturation, so suppressing it risks neutrophil maturation arrest in a patient population already neutropenic.
- Biomarker failure mode: the prognostic effect in [E3] was measured in intermediate-risk AML, a stratum that largely excludes KMT2A-r. Carrying the OR across to KMT2A-r patients would be an unforced error.
- Isoform confounding: if a clinical assay quantifies total HOTAIRM1 while the functionally decisive variable is isoform ratio, the biomarker will appear irreproducible across cohorts for reasons that have nothing to do with the biology.
- Is HOTAIRM1 differentially expressed in KMT2A-rearranged versus non-rearranged B-ALL at all? Nobody has published the comparison, and it is answerable today from TARGET/Interfant RNA-seq without generating a single new sample.
- Does the KMT2A::AFF1-RUNX1 axis identified by Harman et al. actually drive HOTAIRM1, as the AML1-dependence in [E6] would predict? This is a single ChIP-seq track lookup in existing data.
- Does HOTAIRM1 contribute to the prednisone poor response that defines KMT2A-r infant ALL, via ARHGAP18/RHOA/ROCK1? If yes, a ROCK1 inhibitor becomes a repurposing candidate.
- HOTAIRM1 is a proximal-HOXA regulator and KMT2A-r leukemia is a distal-HOXA (HOXA9/HOXA10) disease. Are these two halves of the cluster genuinely uncoupled in KMT2A-r cells, or does the fusion collapse the distinction?
- Which HOTAIRM1 isoform predominates in KMT2A-r cells? Because UTX/MLL versus PRC2 association is isoform-dependent [E2], total-transcript quantification may be actively misleading.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
HOTAIRM1 is a myeloid-specific non-coding transcript between HOXA1 and HOXA2, transcribed antisense to the HOXA genes from the same CpG island as the HOXA1 start site, induced during retinoic-acid granulocytic differentiation via the RA receptor; knockdown blunted RA-induced HOXA1 and HOXA4 and selectively attenuated CD11b and CD18 induction but did not noticeably affect the more distal HOXA genes.
A myelopoiesis-associated regulatory intergenic noncoding RNA transcript within the human HOXA cluster · Zhang X, Lian Z, Padden C, Gerstein MB, Rozowsky J, Snyder M, Gingeras TR, Kapranov P, Weissman SM, Newburger PE · Blood 113(11):2526-34 2009What this evidence cannot show
NB4 promyelocytic leukemia cells plus normal human haematopoietic cells, all in a retinoic-acid granulocytic-differentiation paradigm. Knockdown was siRNA/shRNA-based with modest quantitative effects ('quantitatively blunted', 'selectively attenuated'), not abolition. No B-lymphoid cells and no KMT2A-rearranged cells. The explicit negative result on distal HOXA genes is the most relevant finding for this project and is frequently omitted in citation.
Evidence that argues against this (1)
- A later study found HOTAIRM1's effect on proximal HOXA genes occurred without chromatin-conformation change in one cell type but with loop dissociation in another, so the mechanism reported in any single system should not be generalised. Wang XQD et al. 2017
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Distinct HOTAIRM1 variants preferentially associate with either UTX/MLL or PRC2 complexes to modulate activating and silencing marks at the HOXA bivalent domain; HOTAIRM1 contributes to physical dissociation of chromatin loops at the cluster-proximal end, delaying UTX recruitment and central HOXA transcription — but in a different cell type HOTAIRM1 activated proximal HOXA genes with no chromatin-conformation change.
Reciprocal regulation of chromatin state and architecture by HOTAIRM1 contributes to temporal collinear HOXA gene activation · Wang XQD, Dostie J · Nucleic Acids Research 45(3):1091-1104 2017What this evidence cannot show
Two-author study, RNAi knockdown, 5C/chromatin-conformation methods with the usual resolution limits. The cell-type discordance is reported by the authors themselves and materially limits generalisation. Isoform-specific complex association was shown by RNA immunoprecipitation, which is prone to indirect association. No KMT2A-fusion context.
Evidence that argues against this (1)
- The paper contains its own counter-result: HOTAIRM1 activated proximal HOXA genes without chromatin conformation change in a second cell type, meaning the architecture mechanism is not required for the transcriptional output. Wang XQD et al. 2017
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In 241 AML patients HOTAIRM1 expression was lowest in APL (P<0.001) and highest in t(6;9) AML (P=0.005); in 215 intermediate-risk AML patients high HOTAIRM1 independently predicted shorter overall survival (OR 2.04, P=0.001), shorter leukemia-free survival (OR 2.56, P<0.001) and higher cumulative incidence of relapse (OR 1.67, P=0.046), was overexpressed in NPM1-mutated AML (P<0.001), and correlated with a 33-microRNA signature including miR-196b, with combined HOTAIRM1/miR-196b stratifying 5-year OS at 24% vs 42% vs 70% (P=0.004).
The lincRNA HOTAIRM1, located in the HOXA genomic region, is expressed in acute myeloid leukemia, impacts prognosis in patients in the intermediate-risk cytogenetic category, and is associated with a distinctive microRNA signature · Díaz-Beyá M, Brunet S, Nomdedéu J, Pratcorona M, Cordeiro A, Gallardo D, Escoda L, Tormo M, Heras I, Ribera JM, Duarte R, Queipo de Llano MP, Bargay J, Sampol A, Nomdedeu M, Risueño RM, Hoyos M, Sierra J, Monzo M, Navarro A, Esteve J · Oncotarget 6(31):31613-27 2015What this evidence cannot show
Retrospective single-cooperative-group (CETLAM) cohort; prognostic analysis restricted to 215 intermediate-risk patients, a stratum that largely excludes KMT2A-rearranged AML, so these ORs cannot be transferred to KMT2A-r disease. Published in Oncotarget, a journal whose peer review has been criticised and which was delisted from some indexes. Odds ratios are reported rather than hazard ratios for time-to-event endpoints. No validation cohort. Expression measured at diagnosis only.
Evidence that argues against this (1)
- The direction of the prognostic effect is hard to reconcile with the transcript's pro-differentiation function: HOTAIRM1 knockdown blocks granulocytic differentiation markers, yet HIGH expression predicts worse survival. This suggests the prognostic signal is confounded by differentiation state or lineage rather than reflecting a leukemia-promoting activity. Zhang X et al. 2009
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
PU.1 constitutively binds the HOTAIRM1 regulatory region and directly activates it during granulocytic differentiation via two PU.1 motifs located around +1100 bp downstream of the transcriptional start site; the low HOTAIRM1 expression in APL cells is attributable to reduced PU.1 expression rather than to direct repression by PML-RARA.
PU.1 controls the expression of long noncoding RNA HOTAIRM1 during granulocytic differentiation · Wei S, Zhao M, Wang X, Li Y, Wang K · Journal of Hematology & Oncology 9(1):44 2016What this evidence cannot show
Cell-line ChIP and reporter assays (NB4/APL model). Reporter transactivation is a weak proxy for endogenous regulation. Does not address whether HOTAIRM1 is PU.1-dependent in B-lymphoid or KMT2A-rearranged cells, where PU.1 levels and the myeloid programme differ substantially.
Evidence that argues against this (1)
- This is granulocytic differentiation and APL, not KMT2A-rearranged leukaemia. The mechanism proposed — that low HOTAIRM1 in APL follows from reduced PU.1 rather than from PML-RARA repression — is specific to a fusion this beat is not about, and PU.1 is itself one of the genes this project has shown to behave as a myeloid-lineage artefact rather than a KMT2A-selective dependency. Oncofusion knowledge graph (internal analysis) 2026
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
As of 2026-09-18 no published study isolates HOTAIRM1 function or expression in a KMT2A-rearranged cohort or model: a PubMed query for HOTAIRM1 with leukemia returns 23 records, and a query additionally requiring MLL and KMT2A returns zero.
PubMed database queries 'HOTAIRM1 AND leukemia' (23 records) and 'HOTAIRM1 AND leukemia AND MLL AND KMT2A' (0 records), executed 2026-09-18 · National Library of Medicine, PubMed · PubMed 2026What this evidence cannot show
A negative literature search is bounded by indexing and terminology: a paper reporting HOTAIRM1 in KMT2A-r leukemia only inside a supplementary table, or using 'MLL-AF4' without the indexed term, could be missed. Absence of a title/abstract hit is not proof of absence of data. Preprints and non-indexed venues are not covered. This evidence item supports a claim about the published record, not about the underlying biology.
Evidence that argues against this (1)
- HOTAIRM1 does appear indirectly in the KMT2A-r orbit: it correlates with a 33-miRNA signature containing miR-196b, itself a direct KMT2A-fusion target, so locus-level co-regulation with KMT2A-fusion biology is documented even though HOTAIRM1-specific KMT2A-r experiments are not. Díaz-Beyá M et al. 2015
- E6weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
HOTAIRM1 is overexpressed in an acquired glucocorticoid-resistant leukemia cell model and in patients with poor prognosis; it binds the transcriptional inhibitory region of ARHGAP18 and represses it, raising RHOA/ROCK1 signalling and blocking apoptosis, and ROCK1 inhibition restored glucocorticoid responsiveness; AML1/RUNX1 directly activates the HOTAIRM1 promoter.
The long noncoding RNA HOTAIRM1 controlled by AML1 enhances glucocorticoid resistance by activating RHOA/ROCK1 pathway through suppressing ARHGAP18 · Liang L, Gu W, Li M, Gao R, Zhang X, Guo C, Mi S · Cell Death & Disease 12(7):702 2021What this evidence cannot show
Graded weak for this graph's purposes. The model is an in-vitro-derived acquired glucocorticoid-resistant leukemia cell line, NOT a KMT2A-rearranged cell, and the patient data are correlative. ChIRP-seq binding-site calls are sensitive to probe design and background. Single-group study, no independent replication. The link to KMT2A-r infant ALL steroid resistance is an inference of ours, not a claim of the paper.
Evidence that argues against this (1)
- No paper tests this axis in KMT2A-rearranged cells, so the inference to infant-ALL steroid resistance is untested in either direction. Notably, the other leukemia literature on HOTAIRM1 casts it as pro-differentiation and therefore tumour-suppressive, which is the opposite polarity to this resistance-promoting role. Wei S et al. 2016
- E7established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
HOTAIRM1 is ENSG00000233429, biotype lncRNA, GRCh38 chr7:27,095,647-27,100,265 on the plus strand, HGNC:37117, canonical transcript ENST00000434063; it lies between HOXA1 (chr7:27,092,993-27,096,039, minus strand) and HOXA2 (chr7:27,099,978-27,102,686, minus strand) and is therefore antisense and divergent to HOXA1.
Ensembl REST API gene lookup, homo_sapiens, assembly GRCh38 · Ensembl / EMBL-EBI · Ensembl 2026What this evidence cannot show
Gene-level Havana model coordinates; HOTAIRM1 has multiple annotated isoforms whose boundaries differ, and the isoform identity matters functionally per [E2]. Annotation is revised between Ensembl releases.
Evidence that argues against this (1)
- Coordinates and biotype are an annotation, not a fact about the locus. Ensembl/GENCODE biotype calls for lncRNAs are revised between releases, locus boundaries depend on which transcript model is canonical, and "antisense and divergent to HOXA1" describes the relationship between two annotation records. Nothing here establishes that the transcript is functional rather than a by-product of an active HOXA promoter. Ensembl, EMBL-EBI 2026
HOXA-cluster lncRNA at the myeloid-differentiation end
- related to →Step 5 — The output program: HOXA9/HOXA10/MEIS1/PBX3 plus MEF2C, MYB, BCL2, FLT3 and PROM1/CD133
sits at the differentiation end of the same cluster the output program is read from