start / graph / ana-q187-aff1-is-the-only-aff-scaffold-with-no-named-mendelian-syndrome

AnalysismoderateQ187-the-sec-family-and-yeats-selectivity

Three of the four AFF scaffolds have a named human Mendelian syndrome and the index fusion's partner has none — a 0/4 that is a fact about AFF1 and not about how hard anyone has looked

also called AFF family gene-phenotype census · AFF1 has no germline syndrome

Of the four related scaffold proteins, three cause a named inherited syndrome when they go wrong — FRAXE, KINSSHIP and CHOPS. The fourth is AFF1, the one fused to KMT2A in infant leukaemia, and it causes no named inherited disease at all.

The same thing, for a scientist

Queried live on 2026-09-24 with `elink dbfrom=gene db=medgen` over the four human AFF genes (NCBI Gene 4299, 2334, 3899, 27125), MedGen returns **one gene-phenotype concept each for AFF2 (FRAXE), AFF3 (KINSSHIP syndrome) and AFF4 (CHOPS syndrome), and zero for AFF1** [E1]. Three of four is the positive control: the instrument works and returns a concept whenever one exists, so the AFF1 zero is a property of AFF1. The three syndromes are well sourced — AFF3's KINSSHIP arises from de novo missense variants in a nine-amino-acid degron or deletions of it, in sixteen newly described plus two previously reported individuals [E2]; AFF4's CHOPS from germline gain-of-function missense variants, functionally linking the SEC to cohesin [E3]; AFF2/FMR2 is silenced by CCG-repeat expansion in FRAXE intellectual disability [E4]. **AFF1 is not phenotype-free in model organisms** — Aff1-deficient mice have altered lymphoid development [E5] and a dominant *Af4* point mutation causes cerebellar ataxia, Purkinje cell loss and cataracts in the *robotic* mouse [E6] — so the asymmetry is specifically about a **named human germline disorder**, and the most likely explanation, which this entry cannot distinguish from the others, is that the relevant human AFF1 lesion is somatic and is called leukaemia.

Go deeper

The census

gene	NCBI Gene	MedGen gene-phenotype links	concept
AFF1	4299	0	—
AFF2	2334	1	FRAXE
AFF3	3899	1	KINSSHIP syndrome
AFF4	27125	1	Cognitive impairment – coarse facies – heart defects – obesity – pulmonary involvement – short stature – skeletal dysplasia syndrome (CHOPS)

Why the zero is readable. Rule 67: never trust a zero from a query you have not shown can return a one. This query returns a one three times out of four, through the identical code path and the identical link name, so the instrument is demonstrably capable of the answer it did not give for AFF1.

What the zero bounds, precisely. It bounds a curated gene-to-phenotype relationship in MedGen. It does not bound the existence of an AFF1 germline variant report, a case series, or a phenotype in a model organism — and two of those three exist. A PubMed cross-check is consistent: AFF1[tiab] AND germline[tiab] returns 1 record, while AF4[tiab] AND germline[tiab] returns 10, and inspection of the latter is dominated by t(4;11) breakpoint and translocation work rather than germline syndromes. That second figure is reported unchecked and should not be quoted as a count of anything.

The three syndromes are not the same lesion class, and that is the interesting part

gene	syndrome	lesion	direction
AFF2	FRAXE intellectual disability	CCG-repeat expansion silencing the gene	loss of expression 4
AFF3	KINSSHIP	de novo missense in the nine-residue degron, or deletion of it	gain / stabilisation 2
AFF4	CHOPS	germline missense, described by the authors as gain-of-function	gain 3
AFF1	none named	—	—
Aff1 (mouse)	altered lymphoid development	gene deficiency	loss 5
Af4 (mouse, robotic)	cerebellar ataxia, Purkinje loss, cataracts	dominant point mutation reducing Siah binding	gain / stabilisation 6

Two of the three human syndromes, and the one mouse dominant, are stabilisation lesions on a family whose members share a SIAH-dependent degron. That convergence is developed in mechanism-q187-the-aff-degron-is-the-family-dosage-switch-and-three-paralogue-phenotypes-are-stabilisations; it is the substantive answer to the question of whether paralogue disease biology tells us anything about what AFF1 contributes to the fusion.

Note the KINSSHIP paper's own conclusion on redundancy, because the field is split: from the partial phenotypic overlap of the AFF3 and AFF4 syndromes and from transcriptome analyses, its authors conclude the ALF factors are not redundant and each contributes uniquely to development 2. Melko and colleagues, working on FRAXE, conclude the opposite — that AFF-family redundancy explains FRAXE's mild phenotype 4. Both are cited in this beat and they disagree.

Four readings of the AFF1 zero, and what would separate them

The human AFF1 lesion is somatic and is called leukaemia. The most economical reading. AFF1 is the commonest KMT2A partner in infant ALL; a gene whose defining human pathology is a somatic translocation would not accumulate a germline gene-phenotype concept.
Germline AFF1 loss is embryonic-lethal or too severe to ascertain. Would be distinguished by mouse data, and the mouse says otherwise — Aff1-deficient mice are viable with a lymphoid phenotype 5.
Germline AFF1 variation is tolerated. Would be distinguished by a gnomAD constraint lookup, which this beat did NOT run and which is one request. If AFF1's LOEUF is loose, reading 3 gains; if it is tight and there is still no syndrome, reading 1 or 2 gains.
Nobody has looked. Weakest reading, because exome-sequencing cohorts for intellectual disability are exactly what turned up KINSSHIP and CHOPS, and they were sequencing AFF1 at the same time as AFF3 and AFF4.

The gnomAD lookup in reading 3 is the cheapest discriminator available and is left as an open question rather than guessed at.

Facts and figures

status	run
method	NCBI E-utilities: esearch db=gene to resolve each symbol to a gene id, elink dbfrom=gene db=medgen for gene-phenotype concepts, esummary db=medgen to read each concept title; then primary-literature verification of each named syndrome by efetch abstract
result summary	AFF1 0 concepts, AFF2 1 (FRAXE), AFF3 1 (KINSSHIP), AFF4 1 (CHOPS). Instrument returns a positive 3 of 4 times, so the zero is readable. AFF1 is not phenotype-free in mouse: loss gives altered lymphoid development, a dominant Af4 point mutation gives the robotic ataxia phenotype.
date	2026-09-24
absence class	measured-not-curated
absence class basis	The AFF1 zero is a statement about one curated database's gene-phenotype table. Mouse phenotypes for Aff1 exist and are published (PMIDs 10887138, 12629167, 15459319); they are simply not human Mendelian gene-disease relationships and would never appear in this query. OMIM, ClinGen and gnomAD were NOT queried and any of them could change the reading.

gene phenotype census

inputs

[
  "NCBI Gene 4299, 2334, 3899, 27125",
  "NCBI MedGen",
  "PMIDs 33961779, 25730767, 21330300, 10887138, 12629167, 15459319 (abstracts)"
]

gene phenotype census — the numbers behind the chart

{
  "AFF1": {
    "gene_id": "4299",
    "medgen_links": 0,
    "syndrome": null
  },
  "AFF2": {
    "gene_id": "2334",
    "medgen_links": 1,
    "syndrome": "FRAXE",
    "lesion": "CCG-repeat expansion silencing the gene (loss of expression)"
  },
  "AFF3": {
    "gene_id": "3899",
    "medgen_links": 1,
    "syndrome": "KINSSHIP syndrome",
    "lesion": "de novo missense in a nine-residue degron, or deletion of it (stabilisation)"
  },
  "AFF4": {
    "gene_id": "27125",
    "medgen_links": 1,
    "syndrome": "CHOPS syndrome",
    "lesion": "germline missense described as gain-of-function"
  }
}

measurements

[
  {
    "quantity": "aff_scaffolds_with_a_curated_gene_phenotype_concept",
    "subject": "AFF1, AFF2, AFF3, AFF4 | NCBI MedGen gene-phenotype links via elink dbfrom=gene db=medgen",
    "value": 3,
    "units": "genes of 4",
    "n": 4,
    "universe": "the four human AFF-family genes, NCBI Gene ids 4299, 2334, 3899, 27125, queried live 2026-09-24; one concept returned per positive gene",
    "null": null,
    "note": "The missing one is AFF1, the index fusion's 3' partner. 3 of 4 is also the positive control for the query - the instrument returns a concept whenever one exists, so the AFF1 zero is not an instrument failure. Bounds MedGen curation only."
  }
]

How this could bite us back

A 3-of-4 census is a tidy result and tidy results in this project have repeatedly been artefacts of the instrument. This one rests on a single database, its zero has four competing explanations, and the three cheapest cross-checks (OMIM, ClinGen, gnomAD) were all skipped. Quote it as 'MedGen names a syndrome for three of the four AFF genes and not for AFF1', never as 'AFF1 causes no human germline disease'.
If reading 3 is right and germline AFF1 variation is genuinely tolerated, that is mildly reassuring for any therapeutic that lowers AFF1 function and mildly damaging for the argument in `mechanism-q187-the-aff-degron-is-the-family-dosage-switch-and-three-paralogue-phenotypes-are-stabilisations`, which treats the family as tightly dosage-controlled. The gnomAD lookup could move the beat's conclusions in either direction and was not done.
The disagreement between the two redundancy conclusions in this entry's own evidence (E2 non-redundant, E4 redundant) is not resolved anywhere in this beat, and both readings are used elsewhere in it. Anyone building on the target-set analysis should know the developmental-genetics literature has not settled whether these scaffolds substitute for one another.

Still open

What is AFF1's gnomAD constraint (pLI, LOEUF)? One request, and it is the cheapest discriminator between 'germline AFF1 variation is tolerated' and 'the human AFF1 lesion is somatic'. This beat did not run it.
Does OMIM or ClinGen name an AFF1 gene-disease relationship that MedGen does not? A single-database census is the weakest form of this claim and two cross-checks are one request each.
Is there an unascertained human AFF1 stabilisation syndrome? The robotic mouse says a dominant Af4 degron-region mutation causes cerebellar degeneration; the equivalent human variant class is exactly what found KINSSHIP, and a targeted query of AFF1 de novo variants in intellectual-disability cohorts has not been run here.
Are the AFF3 and AFF4 syndromes phenocopied by the KMT2A::AFF1 fusion in any tissue? The fusion is somatic and haematopoietic, so the question is whether any non-haematopoietic feature of KMT2A-rearranged disease overlaps KINSSHIP or CHOPS. Nobody has asked.

Every source behind this entry (6)

E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
NCBI MedGen, queried live on 2026-09-24 via `elink dbfrom=gene db=medgen` for human AFF1 (Gene 4299), AFF2 (2334), AFF3 (3899) and AFF4 (27125), returns zero gene-phenotype concepts for AFF1 and exactly one each for AFF2 (FRAXE), AFF3 (KINSSHIP syndrome) and AFF4 (the CHOPS syndrome concept spelled out as cognitive impairment - coarse facies - heart defects - obesity - pulmonary involvement - short stature - skeletal dysplasia syndrome).
Oncofusion KG live query, 2026-09-24: NCBI E-utilities elink dbfrom=gene db=medgen over the four human AFF genes, with esummary of each returned concept · Q187-the-sec-family-and-yeats-selectivity · this project 2026
What this evidence cannot show
This bounds a CURATED gene-to-phenotype relationship in one database, not the literature and not clinical reality. MedGen curation lags publication, and a 2026 report of an AFF1 germline syndrome would not necessarily have a concept yet. The gene ids were resolved by `esearch db=gene term=<SYM>[sym] AND human[orgn]` and each returned exactly one id, so no symbol collision arose - but the lookup was not cross-checked against HGNC, and the collision registry's `AF4 / AFF4` row warns that AFF4's protein name contains the string AF4 while AF4 is AFF1.
Evidence that argues against this (1)
- A single-database census is exactly the shape this project has been burned by: an instrument that returns three positives looks healthy while its fourth answer is a curation artefact rather than a biological fact. The independent cross-checks available and NOT run here are OMIM's own gene-phenotype table, ClinGen's gene-disease validity classifications and a gnomAD constraint lookup. Any one of them could overturn the reading of this zero, and none costs more than a request. Voisin N et al. 2021
E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
De novo missense variants in the degron of AFF3 — a nine amino acid sequence important for its binding to ubiquitin ligase — or de novo deletions of that region cause an autosomal dominant disorder named KINSSHIP syndrome in sixteen newly identified plus two previously reported individuals, partially overlapping AFF4-associated CHOPS syndrome; knock-in mice modelling a microdeletion showed mesomelic limb deformities and the commonest missense variant caused early lethality, AFF3 overexpression in zebrafish produced body axis anomalies supporting a pathological effect of increased AFF3, and the authors conclude from partial phenotypic overlap and prior transcriptome analyses that the ALF transcription factors are NOT redundant.
Variants in the degron of AFF3 are associated with intellectual disability, mesomelic dysplasia, horseshoe kidney, and epileptic encephalopathy. · Voisin N, Schnur RE, Douzgou S, Hiatt SM, Rustad CF, Brown NJ, Earl DL, Keren B, Levchenko O, Geuer S, Verheyen S, Johnson D, Zarate YA, Hančárová M, Amor DJ, Bebin EM, Blatterer J, Brusco A, Cappuccio G, Charrow J, Chatron N, Cooper GM, Courtin T, Dadali E, Delafontaine J, Del Giudice E, et al. · American Journal of Human Genetics 2021
PMID 33961779 DOI
What this evidence cannot show
Abstract only. The abstract does not state the degron's residue coordinates, so this entry records 'a nine amino acid sequence' and no numbers - a residue range typed from inference would be exactly the class of error this project logs. The zebrafish overexpression result is described by the authors themselves as providing only 'some support'. The author list is truncated at 25 names as fetched; the full list is longer.
Evidence that argues against this (1)
- Homozygous Aff3 knockout mice, per this paper's own abstract, display skeletal anomalies, kidney defects, brain malformations and neurological anomalies - a loss-of-function phenotype that overlaps the human gain-of-function syndrome. When loss and stabilisation of the same scaffold produce overlapping phenotypes, 'the disease is caused by stabilisation' is a weaker inference than the degron localisation suggests, and the dosage model in the sibling mechanism entry has to accommodate it. Voisin N et al. 2021
E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Exome sequencing identified germline missense mutations in AFF4, a core component of the super elongation complex, in three unrelated probands with a syndrome named CHOPS that phenotypically overlaps Cornelia de Lange syndrome; the paper is titled as reporting gain-of-function mutations, and transcriptome and ChIP-seq analyses showed similar alterations of genome-wide AFF4, cohesin and RNAP2 binding in CdLS and CHOPS, with direct molecular interaction of SEC, cohesin and RNAP2 demonstrated.
Germline gain-of-function mutations in AFF4 cause a developmental syndrome functionally linking the super elongation complex and cohesin. · Izumi K, Nakato R, Zhang Z, Edmondson AC, Noon S, Dulik MC, Rajagopalan R, Venditti CP, Gripp K, Samanich J, Zackai EH, Deardorff MA, Clark D, Allen JL, Dorsett D, Misulovin Z, Komata M, Bando M, Kaur M, Katou Y, Shirahige K, Krantz ID · Nature Genetics 2015
PMID 25730767 DOI
What this evidence cannot show
Three probands. 'Gain-of-function' is asserted in the title; the abstract does not state the mechanism as stabilisation or give residue positions, so this entry does not claim the CHOPS variants are degron variants - that is stated only for AFF3, where the title says so. Abstract only.
Evidence that argues against this (1)
- CHOPS is framed by its discoverers as a cohesinopathy-adjacent disorder, with the mechanism routed through altered genome-wide binding of AFF4, cohesin and RNAP2 rather than through AFF4 protein dosage as such. If the pathogenic axis is the cohesin interaction rather than scaffold abundance, CHOPS is a weaker member of the 'stabilisation of an AFF scaffold' pattern than its gain-of-function label implies, and the pattern is then two instances plus a mouse, not three. Voisin N et al. 2021
E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
AFF2/FMR2 is silenced in FRAXE intellectual disability; all four AFF family members (AFF1/AF4, AFF2/FMR2, AFF3/LAF4, AFF4/AF5q31) localise to nuclear speckles and bind G-quadruplex RNA, and enlarged nuclear speckles are present in FRAXE fibroblasts, with the authors proposing functional redundancy among AFF members as an explanation for the relatively mild FRAXE phenotype.
Functional characterization of the AFF (AF4/FMR2) family of RNA-binding proteins: insights into the molecular pathology of FRAXE intellectual disability. · Melko M, Douguet D, Bensaid M, Zongaro S, Verheggen C, Gecz J, Bardoni B · Human Molecular Genetics 2011
PMID 21330300 DOI
What this evidence cannot show
This paper is a functional characterisation, not the founding FRAXE genetic report, which this beat did not identify or fetch - the first PMID reached for from memory turned out to be an unrelated bipolar-disorder linkage study and was discarded. The FRAXE-AFF2 assignment is therefore carried here by this paper's statement plus the independent MedGen link in E1, not by the primary discovery paper. The speckle and G-quadruplex work is largely overexpression-based.
Evidence that argues against this (1)
- This paper's redundancy conclusion directly contradicts the KINSSHIP paper's non-redundancy conclusion, and the disagreement is load-bearing for this whole beat: if the AFF scaffolds are functionally redundant, the distinct target sets reported for SEC, SEC-L2 and SEC-L3 are harder to interpret and a drug reaching all three is less consequential than the non-redundancy reading implies. Neither paper measured the other's endpoint, so this is an unresolved disagreement rather than a refutation either way. Voisin N et al. 2021
E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Mice deficient for the mAF4 (Aff1) proto-oncogene are viable and show altered lymphoid development, so AFF1 is not phenotype-free on loss of function in a model organism despite having no named human germline syndrome.
Altered lymphoid development in mice deficient for the mAF4 proto-oncogene. · Isnard P, Coré N, Naquet P, Djabali M · Blood 2000
PMID 10887138
What this evidence cannot show
Title and journal verified by esummary 2026-09-24; PubMed returns NO DOI for this record, so the citation carries a PMID only. The abstract itself was not read by this beat - the claim here is what the title asserts plus the viability implied by a developmental phenotype being described, and the severity, penetrance and lineage detail are not established here. Mouse, 2000.
Evidence that argues against this (1)
- A mouse lymphoid phenotype on Aff1 loss and no named human syndrome are not in tension and should not be presented as a puzzle: mouse knockouts routinely show phenotypes that produce no ascertainable human Mendelian disorder, because the human equivalent may be subclinical, embryonic-lethal or simply never sequenced into a cohort. This item therefore weakens reading 2 in the four-readings section only slightly, and the entry should not lean on it. Isaacs AM et al. 2003
E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
A dominant point mutation in mouse Af4 (Aff1) causes the robotic mouse phenotype of cerebellar ataxia, Purkinje cell loss and cataracts, so AFF1 gain-of-function has a documented neurological phenotype in a mammal even though no human AFF1 germline syndrome is named.
A mutation in Af4 is predicted to cause cerebellar ataxia and cataracts in the robotic mouse. · Isaacs AM, Oliver PL, Jones EL, Jeans A, Potter A, Hovik BH, Nolan PM, Vizor L, Glenister P, Simon AK, Gray IC, Spurr NK, Brown SD, Hunter AJ, Davies KE · The Journal of Neuroscience 2003
PMID 12629167 DOI
What this evidence cannot show
The 2003 title says 'is predicted to cause' - the causal attribution was strengthened by the follow-up (PMID 15459319), which established the Siah mechanism. Mouse, ENU mutagenesis, one allele. This entry cites it for the existence of an AFF1 gain-of-function mammalian phenotype and not for any human claim.
Evidence that argues against this (1)
- The robotic phenotype is cerebellar and ocular, with nothing haematopoietic reported, while the Aff1 loss-of-function mouse is lymphoid with nothing neurological reported. A single scaffold whose gain and loss hit two entirely different organ systems is hard to fold into any simple dosage model, and it means neither mouse is a good predictor of what the AFF1 moiety contributes inside a leukaemic fusion. Isnard P et al. 2000

Connected to

related to →AFF1 / AF4 — the 3' half of the fusion, and a target with almost no structural footing
AFF1 is the only AFF scaffold with no MedGen gene-phenotype concept
related to →AFF2 (FMR2) — the SEC-L2 scaffold, and the AFF-family member with no dependency anywhere
AFF2 -> FRAXE, a loss-of-expression lesion
related to →AFF3 (LAF4) — the AFF-family scaffold of SEC-L3, and a KMT2A fusion partner in its own right
AFF3 -> KINSSHIP syndrome, de novo variants in the degron
related to →AFF4 — the textbook SEC scaffold that KMT2A::AFF1 cells need LESS than other leukemias
AFF4 -> CHOPS syndrome, germline gain-of-function
supports →What the paralogues' disease biology says about AFF1: the AFF family shares one SIAH degron — and the index fusion DISCARDS it, because the whole degron lies in the AFF1 leader that the junction at residue 347 throws away
the lesion-direction column that the degron mechanism is built from