ID1 - the only OTHER named factor whose requirement for a KMT2A fusion depends on cell of origin: needed in fetal-liver-derived MLL-AF9 leukaemia, actively restraining in bone-marrow-derived
also called inhibitor of DNA binding 1 · Inhibitor of differentiation 1 · bHLHb24 · ID1_HUMAN
ID1 is a small protein that stops other proteins from telling a cell to specialise. In mice, removing it makes MLL-AF9 leukaemia from fetal liver cells develop more slowly - but makes the same leukaemia from adult bone marrow develop faster. In human patients, ID1 is high in the infant, fetal-origin form of the disease and low in the later-onset form.
The same thing, for a scientist
ID1 (HGNC:5360, 20q11.21, ENSG00000125968, UniProt P41134, 155 aa) is a helix-loop-helix protein lacking a basic DNA-binding domain; it heterodimerises with bHLH transcription factors and blocks their DNA binding, acting as a differentiation inhibitor [E1]. Man et al. expressed MLL-AF9 in fetal liver (FL) or bone marrow (BM) cells from wild-type, *Id1*-/-, *p21*-/- or *Id1*-/-*p21*-/- mice and transplanted them: **loss of Id1 significantly prolonged survival in recipients of FL cells but ACCELERATED leukaemogenesis in recipients of BM cells**, an effect rescued in both models by deleting *Cdkn1a* (p21), identifying p21 as the critical Id1 target. In human material, **ID1 is elevated in fetal-origin (infant) MLL-fusion-protein AML and reduced in postnatal MLL-FP AML** [E2]. **This is the second named molecular instance of a cell-of-origin-conditional requirement in a KMT2A fusion, and the first found outside AFF1.** It is not, strictly, a *fetal-restricted* factor in the LIN28B sense - ID1 is expressed in adults too - but its *requirement* is ontogeny-conditional, and the sign reverses, which is the same structural phenomenon.
What makes this the right comparator for LIN28B, and where the analogy breaks
| LIN28B / KMT2A::AFF1 | ID1 / MLL-AF9 (KMT2A::MLLT3) | |
|---|---|---|
| species of the functional result | human | mouse |
| lesion | endogenous CRISPR translocation | retroviral MLL-AF9 |
| cell of origin compared | fetal liver only (no adult arm reported in the abstract) | fetal liver AND bone marrow, side by side, same experiment |
| direction in fetal cells | required for transformation | required - Id1 loss prolongs survival |
| direction in adult cells | not tested | reversed - Id1 loss accelerates leukaemogenesis |
| mechanism | transcript stabilisation (B-lymphoid, epigenetic, cell-cycle, anti-apoptotic genes) | via p21/CDKN1A - deleting p21 rescues Id1 loss in both models |
| human patient data | LIN28B expressed in infant ALL | ID1 high in fetal-origin infant MLL-FP AML, low in postnatal MLL-FP AML |
| factor itself fetal-restricted? | yes - LIN28B is a fetal gene that falls postnatally | no - ID1 is expressed in adults; it is the REQUIREMENT that is ontogeny-conditional, not the factor |
The last row is the important one and must not be blurred. LIN28B is a fetal-restricted factor. ID1 is an ontogeny-conditional dependency. Those are different kinds of object and they suggest different therapeutic logics: a fetal-restricted factor is absent from the adult patient and therefore intrinsically selective; an ontogeny-conditional dependency is present everywhere and selective only because the disease's origin differs.
And the ID1 study did the comparison LIN28B's did not. It ran fetal and adult arms in the same experiment. That is exactly the design exp-kmt2a-mllt3-versus-aff1-in-lin28b-null-human-fetal-liver calls for on the LIN28B side, done a decade earlier in mouse with a different factor.
The bounded absence around ID1 in the index fusion
Live PubMed esearch, 2026-09-23, one [tiab] tag per term:
| query | translation | count |
|---|---|---|
ID1[tiab] AND "MLL-AF9"[tiab] | "ID1"[Title/Abstract] AND "MLL-AF9"[Title/Abstract] | 2 |
ID1[tiab] AND "MLL-AF4"[tiab] | "ID1"[Title/Abstract] AND "MLL-AF4"[Title/Abstract] | 0 |
ID1[tiab] AND KMT2A[tiab] | "ID1"[Title/Abstract] AND "KMT2A"[Title/Abstract] | 0 |
ID1 has never been studied in KMT2A::AFF1 or under the modern KMT2A gene name at all. The positive control (MLL-AF9) returns 2, so the instrument works. Caveat per rule 72: ID1 is a short token and the query was not run with Id1, inhibitor of DNA binding 1 or ID-1; per rule 80, the bound is titles and abstracts only. data/symbol-collisions.json has no ID1 row, checked with word-boundary matching (rule 88) - though a two-character-plus-digit symbol is exactly the kind that deserves one, and a future beat searching ID1 in free text should expect hits on identifier fields and figure labels.
| hgnc ID | HGNC:5360 |
|---|---|
| symbol | ID1 |
| ensembl gene ID | ENSG00000125968 |
| uniprot | P41134 |
| length AA | 155 |
| chromosomal location | 20q11.21 |
| identifier provenance | HGNC REST fetch/symbol/ID1 (numFound 1) and UniProtKB P41134 fetched live 2026-09-23; PMID 26944543 fetched from PubMed the same day and the title read back. data/symbol-collisions.json has no ID1 row (word-boundary check). |
| localization evidence quality | POOR. Neither location call carries ECO:0000269. Do not quote ID1's localisation as measured. |
| mechanism | Dominant-negative HLH protein: heterodimerises with bHLH transcription factors and blocks their DNA binding, inhibiting differentiation. In the MLL-AF9 context the critical target is CDKN1A/p21 - deleting p21 rescues Id1 loss in both fetal and adult models. |
| is the factor itself fetal restricted | false |
| what is ontogeny conditional | the REQUIREMENT, not the expression - unlike LIN28B, where the factor itself is fetal-restricted |
| cell type of the observation | mouse fetal liver cells and mouse bone marrow cells, transplanted; plus human MLL-FP leukaemia clinical samples |
| shown for fusion | KMT2A::MLLT3 (MLL-AF9), mouse, retroviral. ZERO title/abstract records for ID1 with MLL-AF4 or with KMT2A under its modern symbol - it has never been examined in the index fusion. |
| measured in kmt2a aff1 | false |
| epigenetic role | none direct; it is a transcription-factor sequestrant, not a chromatin enzyme |
prev and alias symbols
[ "bHLHb24", "dJ857M17.1.2" ]
subcellular location uniprot
[ "Nucleus (UNCODED - no evidence)", "Cytoplasm (ECO:0000250 - sequence-similarity inference, NOT an experiment)" ]
ontogeny conditional requirement
{
"fetal_liver_derived_mll_af9": "Id1 loss PROLONGS survival - Id1 is required/pro-leukaemic",
"bone_marrow_derived_mll_af9": "Id1 loss ACCELERATES leukaemogenesis - Id1 is restraining",
"rescued_by": "Cdkn1a (p21) deletion, in both models",
"human_correlate": "ID1 elevated in fetal-origin (infant) MLL-FP leukaemia, reduced in postnatal MLL-FP AML",
"pmid": "26944543"
}- ID1 is not fetal-restricted, so an ID1-directed agent has no intrinsic developmental selectivity - unlike a LIN28B agent, it would act on adult tissue too.
- Its requirement reverses with cell of origin. In an adult or older child with KMT2A-rearranged AML, inhibiting ID1 would be expected to ACCELERATE disease on this data, which makes patient age or disease origin a safety-critical stratifier rather than an efficacy one.
- ID1 acts through p21, a core tumour suppressor and cell-cycle brake with roles in every proliferating tissue. Any intervention on this axis is an intervention on the p21 setpoint.
- The entire functional dataset is mouse retroviral MLL-AF9. Importing it into human infant KMT2A::AFF1 crosses species, lesion type, partner and lineage at once - the same importation problem this project documents repeatedly.
- Is ID1 required for KMT2A::AFF1 transformation in human fetal liver CD34+ cells? Zero papers address it, the reagents exist, and it is the direct import of this result into the index fusion.
- Is ID1 expression higher in infant KMT2A::AFF1 ALL than in childhood KMT2A::AFF1 ALL? The human arm of Man et al. compared infant with postnatal MLL-FP AML; this graph already holds that fetal-specific expression separates infant from childhood ALL, so the same split is predicted and is testable on existing cohorts.
- Do ID1 and LIN28B converge? Both routes end in cell-cycle control - ID1 through p21, LIN28B through stabilisation of cell-cycle transcripts - and nobody has asked whether they are the same axis approached from two directions.
- Why does the p21 rescue work in BOTH directions? If deleting p21 rescues Id1 loss in a model where Id1 loss helps and in a model where it hurts, p21 is downstream of a sign that is set somewhere else, and that somewhere else is the interesting part.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
ID1 is HGNC:5360 at 20q11.21, ENSG00000125968, UniProt P41134, 155 aa; it is a transcriptional regulator lacking a basic DNA-binding domain that negatively regulates basic helix-loop-helix transcription factors by forming heterodimers and inhibiting their DNA binding and transcriptional activity, implicated in growth, senescence, differentiation, apoptosis, angiogenesis and neoplastic transformation.
What this evidence cannot show
Fetched live 2026-09-23; HGNC numFound was 1. Note the subcellular annotation is weak: 'Nucleus' carries NO evidence code and 'Cytoplasm' carries only ECO:0000250 (sequence-similarity inference), not ECO:0000269. For a transcriptional regulator whose whole mechanism is nuclear heterodimerisation, an uncoded nuclear call is a poor record, and this graph should not quote ID1's localisation as measured.
Evidence that argues against this (1)
- UniProt's localisation record for ID1 is unusually weak - 'Nucleus' is uncoded and 'Cytoplasm' carries only ECO:0000250 sequence-similarity inference - so the database entry that supplies this gene's identity supplies almost nothing about where it acts. For a protein whose entire mechanism is nuclear heterodimerisation, that is a gap, not a detail. Blood 2016
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MLL-AF9 was expressed in fetal liver or bone marrow cells from wild-type, Id1-/-, p21-/- or Id1-/-p21-/- mice and transplanted into syngeneic recipients; although all groups developed AML, loss of Id1 significantly prolonged median survival of mice receiving fetal liver cells but accelerated leukaemogenesis in recipients of bone marrow cells, and deletion of Cdkn1a (p21) rescued the effect of Id1 loss in both models; analysis of clinical samples showed ID1 expression elevated in fetal-origin (infant) MLL-fusion-protein leukaemia and reduced in postnatal MLL-FP AML.
Differential role of Id1 in MLL-AF9-driven leukemia based on cell of origin. · Man N, Sun XJ, Tan Y, García-Cao M, Liu F, Cheng G, Hatlen M, Xu H, Shah R, Chastain N, Liu N, Huang G, Zhou Y, Sheng M, Song J, Yang FC, Benezra R, Nimer SD, Wang L · Blood 2016What this evidence cannot show
Abstract read; full text not opened, so effect sizes, n per arm and the clinical-sample cohort size are unknown. MOUSE for the functional arms, with RETROVIRAL MLL-AF9 rather than an endogenous translocation, and MYELOID disease. The fetal-liver transplant model is explicitly described by the authors as a MIMIC of human fetal-origin infant MLL leukaemia - it is an analogy, not the thing. The human expression data are correlative and the abstract gives no cohort size. Most importantly, 'fetal-origin (infant)' versus 'postnatal' in the human analysis is an inference about origin from clinical category, not a measured origin.
Evidence that argues against this (1)
- The ontogeny-conditional direction in the KMT2A field is not consistent: for LIN28B in mouse myeloid MLL-ENL and MLL-AF9, the fetal state is PROTECTIVE and the factor is tumour-suppressive, the opposite of Id1's fetal-liver requirement. Two factors, same disease family, opposite ontogeny logic. Leukemia 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Live PubMed esearch on 2026-09-23 returns 2 records for `ID1[tiab] AND "MLL-AF9"[tiab]`, 0 for `ID1[tiab] AND "MLL-AF4"[tiab]` and 0 for `ID1[tiab] AND KMT2A[tiab]`, each with a non-empty querytranslation.
What this evidence cannot show
Rule 72: only the string `ID1` was tried; `Id1`, `ID-1` and `inhibitor of DNA binding 1` were not, and PubMed's [tiab] matching is case-insensitive so `Id1` is covered but the hyphenated and spelled-out forms are not. Rule 80: titles and abstracts only. The `MLL-AF9` positive control at 2 is thin - it bounds the absence but does not demonstrate a well-covered field.
Evidence that argues against this (1)
- These zeros were produced by the same method that generated a false headline elsewhere in this beat, where adding the legacy symbol MLL turned a zero into three records. ID1 was not queried against MLL, MLL-AF4 or t(4;11) here, so the 'never studied in the index fusion' claim is narrower than it sounds and should be re-run before use. Oncotarget 2016
- depends on →KMT2A::MLLT3 (MLL-AF9), t(9;11)(p21.3;q23.3)
Id1 loss prolongs survival in fetal-liver-derived MLL-AF9, accelerates it in bone-marrow-derived
- expressed in →Fetal liver haematopoietic stem and progenitor cell — the leading candidate cell of origin for t(4;11) infant ALL
the compartment where the Id1 requirement is pro-leukaemic
the only other named factor with an ontogeny-conditional role in a KMT2A fusion; ID1 is a conditional DEPENDENCY, LIN28B a fetal-restricted FACTOR
ID1 high in fetal-origin infant MLL-FP leukaemia, low in postnatal
ID1 gives the fetal state a PERMISSIVE role in a MYELOID KMT2A fusion, which the lineage hypothesis does not predict
human MLL-FP AML: ID1 high in infant/fetal-origin, low in postnatal
the partial analogue - an ontogeny-conditional dependency, not a fetal-restricted factor