Oncofusion

start / graph / gene-clec12a

Genestrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q48-surface-antigens

CLEC12A (CD371, CLL-1, MICL) - the leukaemic-stem-cell antigen whose therapeutic rationale rests on an absence, and which menin inhibition appears to induce

also called CLEC12A · CD371 · CLL-1 · CLL1 · MICL · DCAL-2 · DCAL2 · KLRL1 · hKLRL1 · C-type lectin domain family 12 member A · C-type lectin-like molecule-1 · Myeloid inhibitory C-type lectin-like receptor · Q5QGZ9

CLEC12A is a protein on the surface of myeloid leukaemia cells that is reported to be missing from the normal blood stem cells you must not destroy, which is why several drugs are being built to attack it. A 2026 preprint reports that menin inhibitors - the drug class at the centre of this project - switch CLEC12A on, which would make the two treatments work better together.

The same thing, for a scientist

CLEC12A (UniProt **Q5QGZ9**, CL12A_HUMAN, **265 aa**, single reviewed human record, primary gene name CLEC12A, CD antigen **CD371**) is a type II transmembrane inhibitory C-type lectin-like receptor on myeloid cells. It is the single largest omission in this graph's immunophenotype layer: measured on 2026-09-20 by regex over all 1,465 node files, **CLEC12A appears in 4 entries and CD371 in 7, and neither had an entry of its own**. **The therapeutic rationale is an absence claim, and it is real but narrower than it is usually quoted.** Bakker 2004 reported CLL-1 absent on uncommitted CD34+/CD38- and CD34+/CD33- stem cells while present on CD34+/CD38+ progenitor subsets, and present in 92% (68/74) of AMLs [E2]. van Rhenen 2007 reported CLL-1 **completely absent** on CD34+CD38- cells in normal (n=11) and regenerating (n=6) marrow while present on the CD34+CD38- compartment in 77/89 AML patients [E3]. Willier 2021 found CLEC12A absent from CD34+CD38-CD45RA-CD90+ HSCs in paediatric AML donors [E4]. **But the normal-marrow denominators are small (n=11 and n=6), and the claim is about a flow gate, not about every cell that can reconstitute.** Yanagisawa 2020, testing six putative LSC targets across genetically defined AML, states the honest version: with the exception of CD47, there was **at most low expression** on normal HSCs, and CD123 and CLL-1 showed the greatest LSC-versus-HSC difference [E5]. *At most low* is not *absent*. Haubner 2023 built an entire cooperative-CAR design around ADGRE2-low/**CLEC12A-negative** normal HSPCs and still measured off-target HSPC toxicity comparable to a CD19-CAR [E6]. **The absence that matters clinically is not the one on HSCs.** CLEC12A is highly expressed on mature granulocytes and monocytes. In the first-in-human CLL-1 CAR-T study (n=10 adults, R/R AML) **all patients developed severe pancytopenia and two died of severe infection from chronic agranulocytosis** [E7]. The stem-cell-sparing argument does not spare neutrophils. **And CLEC12A monotherapy has now failed a clinical test on precisely this point.** KITE-222, an autologous CLL-1-directed CAR T, produced no clinically meaningful responses across three dose levels in 12 R/R AML patients; in expanding patients CLL-1-positive blasts were nearly depleted while **CLL-1-negative blasts persisted and total blasts did not fall** [E14]. The 54.5% figure was a prediction and this is its read-out. **The menin axis.** A 2026 preprint reports that menin inhibition induces robust and uniform CLEC12A expression in *KMT2A*-r and *NPM1*-mutated AML and sensitises them to CLEC12A-directed CAR cells [E8]. This is the first claim in this graph that a menin inhibitor creates an immunotherapy target rather than only killing cells, and it is **not yet peer reviewed** - see `mechanism-menin-inhibition-induces-clec12a`. **Lineage caution for the index disease.** CLEC12A/CD371 is a myeloid antigen. Bakker found it absent in 12 of 13 ALL cases [E2] and Zhao found it absent in ALL blasts (n=5) [E9]; but the iBFM prospective study found CD371 positive in **160 of 1,812 (8.8%)** paediatric BCP-ALL, associated with transient myelomonocytic switch in up to 65.4% of day-15 samples [E10]. **Both are true**: CLEC12A is rare in B-ALL and not zero, and the B-ALL cases that carry it are exactly the lineage-unstable ones.

Go deeper
Does the 'absent from normal HSC' rationale hold up? (read this before designing anything)

Verdict: it holds up as a flow-cytometric statement on small normal-marrow cohorts, it is routinely over-quoted as 'absent from normal haematopoiesis', and the toxicity that has actually been observed comes from a different compartment entirely.

sourcewhat was actually measurednormal-marrow n
Bakker 2004 2CLL-1 absent on uncommitted CD34+/CD38- and CD34+/CD33- cells; present on subsets of CD34+/CD38+ and CD34+/CD33+ progenitors; not expressed in any non-haematopoietic tissuenot stated in abstract
van Rhenen 2007 3'completely absent' on CD34+CD38- cells in normal and regenerating marrow11 normal, 6 regenerating
Willier 2021 4CLEC12A and CD33/CLEC12A double-positive expression absent in CD34+CD38-CD45RA-CD90+ HSCs and in non-haematopoietic tissue13 age-matched healthy donors
Yanagisawa 2020 5'at most low expression' of CLL-1 on normal HSCs; CD123 and CLL-1 give the greatest LSC-vs-HSC differencenot stated in abstract
Haubner 2023 6normal HSPCs are ADGRE2-low/CLEC12A-negative; yet off-target HSPC toxicity of the cooperative CAR was similar to a CD19-CARquantified in R/R patient samples and normal HSPCs

Three things a designer should carry forward.

  1. The claim is a gate claim. Every one of these measurements defines the HSC by a surface-marker gate and then asks whether CLEC12A is in it. If a reconstituting cell sits outside that gate, the measurement cannot see it. concept-flow-cytometry-immunophenotype-kmt2a-r covers the general form of this limitation.
  2. Progenitors are not spared even in the founding paper. Bakker's own result has CLL-1 on CD34+/CD38+ and CD34+/CD33+ progenitor subsets 2. Stem-cell sparing and progenitor sparing are different promises.
  3. The observed clinical toxicity is granulocytic, not stem-cell. Jin 2022 is explicit: CLL-1 is highly expressed on normal granulocytes, all ten patients had severe pancytopenia, two died of infection in chronic agranulocytosis, and the authors propose bridging transplant as the rescue 7. A target can be perfectly HSC-sparing and still produce lethal neutropenia, because the mature compartment is where the antigen is.
What the protein actually does (and why that is a bite-back, not a footnote)

CLEC12A/MICL is not an inert tag. It is an inhibitory C-type lectin-like receptor with a cytoplasmic ITIM. Malamud 2024 (Nature) showed MICL directly recognises DNA in neutrophil extracellular traps, and that loss or inhibition of MICL function leads to uncontrolled NET formation through the ROS-PAD4 pathway and an auto-inflammatory feedback loop; in rheumatoid arthritis, lupus and severe COVID-19, inhibitory anti-MICL autoantibodies are found in patients 11.

That is a direct read-out of what an antagonistic anti-CLEC12A binder does to a normal neutrophil: humans who make functional-blocking antibodies to this receptor have exacerbated NET-driven pathology. A CAR or bispecific that engages CLEC12A is, on any neutrophil it does not kill, an agonist or antagonist of an autoregulatory brake. No CLEC12A-directed clinical programme in the sources read for this entry reports having measured NET formation. See risk-clec12a-directed-therapy-myeloid-ablation.

Identifier traps specific to CLEC12A (three of them, all live)

Trap 1 - CLL1 is a UniProt gene synonym of two different human proteins, and the wrong one sorts first. Fetched 2026-09-20:

query gene:CLL1 AND organism_id:9606 AND reviewed:true
Q9Y6Z7  primary COLEC10  synonyms CLL1   Collectin-10 (Collectin liver protein 1) (CL-L1)   277 aa   <- FIRST
Q5QGZ9  primary CLEC12A  synonyms CLL1 DCAL2 KLRL1 MICL                                   265 aa   <- wanted

This is rule 62's exact shape, and length is a poor defence: 277 against 265, twelve residues apart. Only gene_primary separates them. CL-L1 and CLL-1 differ by one hyphen and name different proteins. 12

Trap 2 - CLEC12B is 276 aa, eleven residues from CLEC12A's 265. A neighbouring family member whose length is inside the noise band of every length check this project runs. Assert Q5QGZ9 + 265, and read the primary protein name back; CLEC12B's is C-type lectin domain family 12 member B (Macrophage antigen H), which does not contain 'CLL' or 'MICL'. 12

Trap 3 - HGNC and UniProt disagree about the alias set. HGNC:31713 lists alias symbols CLL-1, MICL, CD371, DCAL-2 with hyphens, and an HGNC alias lookup on the bare string CLL1 returns nothing at all, while UniProt's returns two genes. An alias resolver that queries only one of the two databases gets either a false negative or an ambiguous pair. 13

Retrieval consequence. On PubMed, 2026-09-20: CLEC12A returns 156 records; adding the alias set (CLEC12A OR CLL-1 OR CD371 OR MICL) is necessary to reach the older CLL-1 literature, because the 2004-2010 founding papers say CLL-1 in their titles and not CLEC12A.

How each source in this entry was disambiguated (rule 66 requires this to be written down). Every primary source cited here for CLL-1 biology spells the name out in its title or abstract as C-type lectin-like molecule-1 239 or as CLL-1 with a myeloid-leukaemia context 714. COLEC10's protein name is Collectin-10 / Collectin liver protein 1 / CL-L1, a liver-expressed collectin, and no source cited in this entry uses the bare unhyphenated string CLL1 as its only identifier. So all five are resolved to CLEC12A on stated-full-name grounds, not on context alone, and none was left unresolved. A source that said only CLL1 would be ambiguous on its face and would have to be recorded unresolved.

Three-way, not two-way. The collision is worse than a pair: CLL1 -> COLEC10 (277 aa) and CLEC12A (265 aa), and the family paralogue CLEC12B is 276 aa. All three lie within 12 residues of each other, which is tighter than the SMC1A/SMC3 pair (16 residues) that brief rule 66 calls the limit of the length discriminator. For this gene length_aa is close to worthless as a swap detector and only gene_primary separates the records.

Cell type, cell state, localisation, epigenetics
  • Cell type. Myeloid: granulocytes, monocytes, dendritic cells, subsets of CD34+CD38+ progenitors; on leukaemic blasts in ~92% of AML and on the CD34+CD38- LSC compartment in 77/89 AML patients 23. Rare (8.8%) and lineage-flagging in paediatric BCP-ALL 10.
  • Cell state. Expression is inducible, which is the whole point of the menin result 8, and it is stable between diagnosis and relapse in AML (n=9 pairs) 3. In BCP-ALL, CD371 positivity marks cells capable of a transient myelomonocytic shift during induction 10.
  • Localisation. Plasma membrane, type II single-pass transmembrane, 265 residues, one C-type lectin domain, cytoplasmic ITIM 12. Internalises on antibody binding in the CLL-1 antibody programme 9.
  • Epigenetics. The menin-inhibition preprint's claim is precisely an epigenetic one: the induction is described as epigenetic priming by a Menin-KMT2A-complex inhibitor 8. Whether CLEC12A is a direct KMT2A-fusion target locus is not established by anything read for this entry - see the open questions.
The epigenetic column: not applicable as an activity, unresolved as a locus

This entry already discusses epigenetics, in one bullet inside the combined Cell type, cell state, localisation, epigenetics section, and it does so correctly. This amendment adds two things that bullet does not have: the domain-architecture basis for calling the activity question not-applicable, and a bounded search so the absence is a measurement rather than a silence. It also states, in the corpus's own field convention, the distinction the bullet gestures at.

Not applicable as an activity, and the reason is structural. Fetched live from UniProt on 2026-09-21, Q5QGZ9 is 265 residues: cytoplasmic 1-43 with an ITIM motif at 5-10, a type II signal-anchor transmembrane helix at 44-64, extracellular 65-265, and a single C-type lectin domain at 140-249, with a Pfam complement of exactly PF00059 (Lectin_C) E16. An inhibitory surface lectin receptor with its only globular domain outside the cell has no route to a nucleosome, and this entry separately records no nuclear pool. That is a category statement, not a gap, and it is the same call this project made for HLA-B and FUT9 on the compartment question.

Bounded, so the reader can judge it. CLEC12A[tiab] AND histone[tiab] returns 0. CLEC12A[tiab] AND epigenetic[tiab] returns 1, an arsenic-exposure blood methylation-array study. CLEC12A[tiab] AND methylation[tiab] returns 3, all off-topic - a COPD methylation array, the same arsenic study, and a microglial-niche paper. Bounded by titles and abstracts only.

And the part that is emphatically not not-applicable. The CLEC12A locus is the subject of a live therapeutic claim: a 2026 preprint reports that menin inhibition induces robust and uniform CLEC12A expression in KMT2A-rearranged and NPM1-mutated AML and describes this as epigenetic priming for CLEC12A-directed CAR therapy. This entry already grades that source weak and preprint, and already asks the right question in its open list - is CLEC12A a direct KMT2A-fusion target locus, or is the induction downstream of differentiation? For the epigenetic dimension that question is the entry. CLEC12A is a myeloid-maturation antigen; a drug that releases a differentiation block would be expected to raise it whether or not the fusion ever sat on its promoter. Calling the induction epigenetic priming presupposes the first answer, and no fusion-occupancy, menin-occupancy or histone-mark measurement at the CLEC12A locus was found for this entry.

The contrast that makes the gap concrete. A sibling surface antigen in this graph has exactly the evidence CLEC12A lacks: CD133/PROM1 is driven by an intragenic H3K79me2/me3 enhancer element with direct MLL-AF4 binding and is PRC2-repressed when off, which is why this project's exp-surface-marker-panel-under-menin-and-dot1l-inhibition uses CD133 as its discriminating readout and gets a two-drug truth table from it. Running CLEC12A in the secondary panel of that same experiment - it already lists CD371 - would settle the question at almost no marginal cost, and the answer decides whether "menin inhibition primes CLEC12A" is a chromatin claim or a differentiation claim.

Facts and figures
uniprotQ5QGZ9
length AA265
gene primaryCLEC12A
hgnc IDHGNC:31713
cd designationCD371
resolution notegene_exact:CLEC12A AND organism_id:9606 AND reviewed:true returned exactly one row on 2026-09-20; its primary gene name equals CLEC12A and its length is 265. Cross-checked against CLEC12B (Q2HXU8, 276 aa) and against the CLL1 alias collision with COLEC10 (Q9Y6Z7, 277 aa).
expression normalgranulocytes, monocytes, dendritic cells, subsets of CD34+CD38+ and CD34+CD33+ progenitors; reported absent on CD34+CD38- HSC gates in normal (n=11) and regenerating (n=6) marrow
physiological ligandDNA in neutrophil extracellular traps (MICL is an inhibitory ITIM receptor)
gene symbolCLEC12A
gene symbol notepromoted from gene_primary 2026-09-20: verify-accessions.py reads data.gene_symbol, so a gene_primary-only entry was being length-checked or id-inferred rather than symbol-asserted
subnuclear localizationnot applicable - CLEC12A is a single-pass type II plasma membrane receptor with no reported nuclear pool
shown for fusionNO - every localisation experiment located for CLEC12A is in normal human or mouse myeloid cells. No study has imaged or fractionated CLEC12A in a KMT2A-rearranged blast. Diagnostic and CAR-target flow cytometry in AML does read the surface compartment, but that is an expression measurement made in this compartment rather than a localisation study of it.
epigenetic readoutNOT APPLICABLE AS AN ACTIVITY - AND NOT AN ABSENCE AS A LOCUS. CLEC12A writes, erases and reads no histone or DNA modification, and the domain architecture is the reason rather than a failure to find one: UniProt Q5QGZ9, fetched live 2026-09-21, is 265 aa with cytoplasmic 1-43 carrying an ITIM motif at 5-10, a type II signal-anchor transmembrane helix at 44-64, extracellular 65-265, ONE C-type lectin domain at 140-249, and a Pfam complement of exactly PF00059 (Lectin_C). There is no SET, JmjC, HAT, HDAC, bromo, chromo, Tudor, PHD, CXXC, YTH or KH module, and no nuclear pool is reported, so there is no mechanism by which this protein could modify or read chromatin. BOUNDED: CLEC12A[tiab] AND histone[tiab] returns 0; CLEC12A[tiab] AND epigenetic[tiab] returns 1 and it is an arsenic-exposure blood methylation-array study; CLEC12A[tiab] AND methylation[tiab] returns 3 and all three are off-topic (COPD methylation array, the same arsenic study, a microglial-niche paper). The instrument returns ones on the same shape elsewhere in this tranche. WHAT IS NOT NOT-APPLICABLE is the CLEC12A LOCUS as a target of epigenetic therapy, and this entry already states it carefully under its own 'Epigenetics' bullet: a 2026 preprint reports menin inhibition inducing robust and uniform CLEC12A expression and calls it EPIGENETIC PRIMING for CAR-based targeting. CAUSE OR CONSEQUENCE IS THE WHOLE QUESTION AND IT IS UNRESOLVED: nothing read for this entry establishes KMT2A-fusion or menin occupancy at the CLEC12A locus, so the induction is equally well explained as a downstream consequence of menin-inhibitor-induced DIFFERENTIATION - CLEC12A is a myeloid-maturation antigen, and a drug that releases a differentiation block would raise it without touching its chromatin. The discriminating measurement is a ChIP or CUT&RUN for KMT2A-fusion, menin and H3K79me2 at the CLEC12A locus, and it has not been done. The source is a PREPRINT, not peer reviewed. CONTRAST WITH A SIBLING ANTIGEN: CD133/PROM1 IS a direct case - an intragenic H3K79me2/me3 enhancer element with direct MLL-AF4 binding, PRC2-repressed when off - which is exactly the evidence CLEC12A lacks.
paralogue length trap
CLEC12A265.0CLEC12A: 265CLEC12B276.0CLEC12B: 276COLEC10_via_CLL1_alias277.0COLEC10_via_CLL1_alias: 277
uniprot gene synonyms
[
  "CLL1",
  "DCAL2",
  "KLRL1",
  "MICL"
]
hgnc alias symbols
[
  "CLL-1",
  "MICL",
  "CD371",
  "DCAL-2"
]
paralogue length trap — the numbers behind the chart
{
  "CLEC12A": 265,
  "CLEC12B": 276,
  "COLEC10_via_CLL1_alias": 277
}
expression leukaemia
{
  "aml_blasts_pct": 92,
  "aml_blasts_pct_second_series": 86.5,
  "aml_lsc_fraction_pct_van_rhenen": 87,
  "aml_lsc_fraction_pct_zhao": 54.5,
  "bcp_all_pct_ibfm": 8.8,
  "all_blasts": "absent in 12/13 (Bakker) and 0/5 (Zhao)"
}
modalities in development
[
  "CAR T",
  "bispecific T-cell engager",
  "trispecific NK engager",
  "naked and chimeric monoclonal antibody",
  "CLEC12A chimeric costimulatory receptor in a gated ADGRE2 CAR"
]
clinical outcomes
{
  "KITE-222_phase1": {
    "n": 12,
    "population": "R/R AML adults",
    "efficacy": "no clinically meaningful responses; CLL-1-negative blasts persisted",
    "pmid": "41941266",
    "title": "Phase 1 Study of KITE-222, an Autologous CLL-1-Directed CAR T-cell Therapy in Patients with Relapsed/Refractory Acute Myeloid Leukemia.",
    "authors": "Daver N, Blachly JS, Ghobadi A, et al.",
    "venue": "Clin Cancer Res",
    "year": 2026,
    "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."
  },
  "CLL-1_CAR_T_first_in_human": {
    "n": 10,
    "population": "R/R AML adults",
    "efficacy": "CR/CRi 70%",
    "safety": "severe pancytopenia in all; 2 deaths from infection in chronic agranulocytosis",
    "pmid": "35799191",
    "title": "First-in-human phase I study of CLL-1 CAR-T cells in adults with relapsed/refractory acute myeloid leukemia.",
    "authors": "Jin X, Zhang M, Sun R, et al.",
    "venue": "J Hematol Oncol",
    "year": 2022,
    "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."
  }
}
disambiguation of sources
{
  "method": "stated full protein name in title/abstract",
  "resolved_to_CLEC12A": [
    "15548716",
    "17609428",
    "19648166",
    "35799191",
    "41941266"
  ],
  "unresolved": [],
  "note": "No cited source identifies the protein by the bare string CLL1 alone; such a source would be recorded unresolved."
}
localization
{
  "subcellular": "plasma membrane, type II single-pass (Q5QGZ9, 265 aa), with the C-type lectin-like domain extracellular and the ITIM cytoplasmic. Surface expression is not static: MICL/CLEC12A surface levels are DOWN-REGULATED on myeloid cells under inflammatory and activation conditions, both in vitro and in an in vivo model of acute inflammation, and the protein is variably N-glycosylated between cell types.",
  "subnuclear": "not applicable - a type II membrane receptor; no nuclear or subnuclear pool has been reported in any source read for this beat",
  "how_shown": "a MICL-specific monoclonal antibody used on primary human myeloid cells, with surface expression followed through in vitro activation and an in vivo acute-inflammation model (PMID 16838277); lateral position within the membrane addressed separately by raft fractionation in mouse (PMID 37339805)",
  "systems": "primary human granulocytes, monocytes, macrophages and dendritic cells; the raft work is murine Clec12a",
  "node": "localization-plasma-membrane-leukaemic-surface-antigens"
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.0755,
  "chronos_mean_other_leukaemia": -0.0227,
  "difference": -0.0528,
  "pan_essentiality_all_lines": -0.0386,
  "selectivity_rank": 3258,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
epigenetic bounded absence
{
  "claim": "No histone- or DNA-modifying activity, and no epigenetic regulation of the CLEC12A locus, is reported in the title/abstract index.",
  "date_measured": "2026-09-21",
  "tool": "NCBI esearch db=pubmed (eutils), querytranslation read back from each response",
  "queries": [
    {
      "query": "CLEC12A[tiab] AND histone[tiab]",
      "querytranslation": "\"CLEC12A\"[Title/Abstract] AND \"histone\"[Title/Abstract]",
      "count": 0,
      "errorlist_note": "warninglist outputmessages: ['No items found.']; no phrasesnotfound, so both terms are indexed"
    },
    {
      "query": "CLEC12A[tiab] AND epigenetic[tiab]",
      "querytranslation": "\"CLEC12A\"[Title/Abstract] AND \"epigenetic\"[Title/Abstract]",
      "count": 1,
      "title_checked": "PMID 34516978, arsenic exposure and human blood DNA methylation/hydroxymethylation in Bangladesh and Spain - OFF TOPIC; 0 on topic"
    },
    {
      "query": "CLEC12A[tiab] AND methylation[tiab]",
      "querytranslation": "\"CLEC12A\"[Title/Abstract] AND \"methylation\"[Title/Abstract]",
      "count": 3,
      "title_checked": "PMID 36655999 (COPD differentially methylated genes), 34516978 (the same arsenic study), 30451869 (competitive repopulation of an empty microglial niche) - ALL OFF TOPIC; 0 on topic"
    },
    {
      "query": "CLL-1[tiab] AND epigenetic[tiab]",
      "querytranslation": "\"CLL-1\"[Title/Abstract] AND \"epigenetic\"[Title/Abstract]",
      "count": 3,
      "role": "ALIAS ARM, run because rule 69 part 4 and this entry's own resolution_note establish that CD371/CLL1/MICL/KLRL1/DCAL2 are the names most of this literature uses; the count is small and CLL-1 is a short, word-like alias that also abbreviates chronic lymphocytic leukaemia, so these were not counted as on-topic without reading"
    }
  ],
  "aliases_searched": [
    "CLEC12A",
    "CLL-1"
  ],
  "aliases_not_searched": [
    "CD371",
    "MICL",
    "KLRL1",
    "DCAL2",
    "CLL1"
  ],
  "absence_class": "never-measured",
  "absence_class_basis": "Bounded by titles and abstracts only; supporting figures not searched (rule 80), and the alias list is incomplete - CD371, MICL, KLRL1, DCAL2 and unhyphenated CLL1 were not run, which per rule 70 is the real bound on this absence. It is recorded as never-measured for the ACTIVITY question, where the domain architecture makes it a category statement rather than a gap. For the LOCUS question the class is different and better: measured-not-reported, because the 2026 menin-inhibition preprint generated the expression data and the occupancy measurement that would interpret it was simply not made.",
  "what_would_close_it": "A ChIP or CUT&RUN for KMT2A-fusion protein, menin and H3K79me2 at the CLEC12A locus in a KMT2A-rearranged cell, with and without a menin inhibitor. This graph's exp-surface-marker-panel-under-menin-and-dot1l-inhibition already lists CD371 in its secondary flow panel, so the expression arm is costed; only the occupancy arm is missing."
}
How this could bite us back
  • **This has now failed in the clinic on exactly the predicted mechanism.** KITE-222, a CLL-1-directed CAR T, lacked clinically meaningful responses across three dose levels in 12 R/R AML patients; where CAR T cells did expand, CLL-1-positive blasts were nearly eliminated and **CLL-1-negative blasts persisted with no fall in total blasts** [E14]. Antigen-negative escape is not a theoretical risk for this target - it is the observed phase 1 outcome, and it is the single strongest argument for the induction strategy [E8] or for gated co-targeting [E6] over CLEC12A monotherapy.
  • **Neutrophil ablation, observed and fatal.** CLEC12A is highly expressed on normal granulocytes. In the only first-in-human CLL-1 CAR-T study read here, all 10 patients had severe pancytopenia and 2 died of infection during chronic agranulocytosis [E7]. HSC sparing does not buy neutrophil sparing, and the authors' own proposed mitigation is a transplant.
  • **The stem-cell absence is measured inside a gate, on small normal cohorts.** n=11 normal and n=6 regenerating marrows [E3]; a reconstituting cell outside the CD34+CD38- gate is invisible to that measurement. Yanagisawa's 'at most low expression' [E5] is the defensible version and it is not zero.
  • **Antigen-negative escape is already the expected failure mode.** Only 54.5% of AML CD34+/CD38- fractions were CLL-1 positive in one series [E9], and the leading preclinical design uses CLEC12A only as a gated costimulatory input precisely to avoid single-antigen escape [E6].
  • **Essential-function toxicity in the receptor's day job.** MICL is an inhibitory brake on neutrophil activation and NET formation; humans with function-blocking anti-MICL autoantibodies have exacerbated NET-driven pathology in RA, lupus and severe COVID-19 [E11]. An engaging binder is an uncontrolled experiment on that brake.
  • **The menin combination's benefit is the induction, so resistance to the induction defeats it.** If the CLEC12A induction reported in the 2026 preprint [E8] is a differentiation-dependent effect, then any clone that escapes menin-inhibitor-driven differentiation - the documented MEN1 mutant resistance route this graph already tracks - is simultaneously CLEC12A-low and CAR-invisible. The two therapies could share a single resistance mechanism rather than covering for each other.
  • **Off-tumour myeloid neoplasia.** CLL-1 is on 10-60% of CD34+ cells in CML and MDS [E2], so the target is not AML-restricted within myeloid disease.
  • **Delivery and lineage mismatch for the index disease.** For KMT2A-rearranged infant B-ALL there is no CLEC12A expression measurement at all in the sources read here, and the one paediatric study that did stratify by KMT2A nominated CD33/FLT3 instead [E4].
  • Antigen down-modulation as a resistance route that is not a mutation: CLEC12A surface levels fall under inflammatory and activation conditions on normal myeloid cells (PMID 16838277). If the same regulation operates on blasts, cytokine release during CAR or bispecific therapy could lower the very antigen density the therapy depends on, producing escape with no genetic lesion to find.
  • Raft position is an unmeasured variable in every CLEC12A-directed programme here. Clec12a is expelled from lipid rafts on stimulation (PMID 37339805, murine), so epitope accessibility and signalling context can change while flow-cytometric positivity does not.
  • If CLEC12A induction under a menin inhibitor is a differentiation effect rather than a chromatin effect at the CLEC12A locus, then the priming is expected to be transient and to track the differentiation response - so it would fade in exactly the patients whose disease stops differentiating, which are the ones who relapse. A CAR-T schedule built on sustained menin-inhibitor-induced antigen would then fail latest and worst. The preprint reports uniform induction; nobody has reported its durability.
Still open
  1. Does menin-inhibitor-induced CLEC12A convert the CLL-1-negative blast fraction that defeated KITE-222 [E14] into a targetable one? That is the precise question the 2026 preprint [E8] implies and does not, at abstract level, appear to test in a KITE-222-like heterogeneous R/R population.
  2. Is *CLEC12A* a direct KMT2A-fusion target locus, or is its induction under menin inhibition a downstream consequence of differentiation? Nothing read for this entry establishes fusion occupancy at the CLEC12A locus, and the distinction decides whether the menin/CAR combination is a targeted synergy or a generic differentiation effect that any differentiating agent would reproduce.
  3. What is the CLEC12A status of KMT2A-rearranged infant B-ALL specifically? Every measurement in this entry is AML or unselected BCP-ALL. The iBFM study did not report the KMT2A-r fraction of its CD371-positive group, and that single number would decide whether this target has any B-lineage relevance at all.
  4. Does a CLEC12A-engaging binder dysregulate NET formation in the surviving neutrophils of a treated patient? MICL loss-of-function drives uncontrolled NET release [E11] and no CLEC12A clinical programme read here reports measuring it.
  5. Why do van Rhenen (87% of LSC fractions positive) and Zhao (54.5%) disagree by more than thirty points on the same compartment? If the lower figure is right, roughly half of patients have a CLEC12A-negative LSC reservoir at baseline and monotherapy is futile in them.
  6. Is menin-inhibitor-induced CLEC12A a chromatin event at the CLEC12A locus or a consequence of differentiation? This entry already asks the fusion-occupancy question; the epigenetic framing sharpens it into a discriminating experiment. CD133/PROM1 is the positive control in this graph - an intragenic H3K79me2/me3 enhancer with direct MLL-AF4 binding, PRC2-repressed when off - and CLEC12A has none of that evidence. If CLEC12A behaves like CD133 the induction is a programme effect and a DOT1L inhibitor should reproduce it; if it behaves like a maturation antigen, only agents that release the differentiation block will work and the phrase 'epigenetic priming' is misleading.
  7. Would pinometostat induce CLEC12A? This is the cheapest discriminator available and it costs one arm in an experiment this graph has already designed. A menin inhibitor and a DOT1L inhibitor interrupt the fusion programme by different routes, so concordance implicates the programme and discordance implicates differentiation or an off-programme menin function.
Every source behind this entry (16)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    CLEC12A is named in 4 node files and CD371 in 7 of this graph's 1,465 node files, and before this entry neither had a node of its own.

    Q48 surface-antigen beat: regex census of data/graph/nodes/*/*.json · Q48-surface-antigens · Oncofusion knowledge graph (internal measurement) 2026
    What this evidence cannot show

    Measured 2026-09-20 by a Python regex sweep over all node JSON files (patterns `CLEC12A`, `CD371`). It counts files containing the string anywhere, including in citation titles, so it slightly over-counts substantive discussion. The dispatch brief quoted the same two numbers (4 and 7); this run reproduces them independently.

    Evidence that argues against this (1)
    • A string count is not a measure of how well covered a topic is. Seven mentions of CD371 could in principle be seven thorough treatments inside other entries. Inspection shows they are not - they are passing mentions in flow-panel lists and in one CD33 entry's calibration evidence - but the count alone cannot establish that. Oncofusion knowledge graph 2026
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    CLL-1 is a transmembrane glycoprotein restricted to the haematopoietic lineage, absent on uncommitted CD34+/CD38- and CD34+/CD33- stem cells but present on subsets of CD34+/CD38+ and CD34+/CD33+ progenitors, not expressed in any other tissue, and expressed in 92% (68 of 74) of primary AMLs; it complemented CD33 in that 8 of 12 CD33-negative AMLs were CLL-1 positive, and it was absent in 12 of 13 cases of acute lymphoblastic leukaemia.

    C-type lectin-like molecule-1: a novel myeloid cell surface marker associated with acute myeloid leukemia. · Bakker AB, van den Oudenrijn S, Bakker AQ, Feller N, van Meijer M, Bia JA, et al. · Cancer Res 2004
    PMID 15548716DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. This is the founding descriptive paper and the expression survey is antibody-based flow cytometry with a single newly raised human anti-CLL-1 antibody, so epitope-dependent under-detection cannot be excluded. The normal-marrow denominator is not given in the abstract. 'Absent on uncommitted CD34+/CD38- cells' and 'present on CD34+/CD38+ progenitors' appear in the same sentence - the paper does not claim progenitor sparing, and the therapeutic literature frequently quotes only the first half.

    Evidence that argues against this (1)
    • The same paper reports CLL-1 on CD34+ cells in 10-60% of chronic myelogenous leukaemia and myelodysplastic syndrome cases, so 'AML-restricted' is wrong as a statement about myeloid neoplasia generally - a CLEC12A agent given to a patient with an undiagnosed MDS component is not on a clean target. Xiu Y et al. 2018
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    CLL-1 is expressed on the CD34+CD38- stem-cell compartment in 77 of 89 AML patients, CD34+CLL-1+ cells engraft NOD/SCID mice with outgrowth to CLL-1+ blasts, expression did not differ between diagnosis and relapse in 9 paired cases, and CLL-1 expression is completely absent on CD34+CD38- cells in normal (n=11) and regenerating (n=6) bone marrow.

    The novel AML stem cell associated antigen CLL-1 aids in discrimination between normal and leukemic stem cells. · van Rhenen A, van Dongen GA, Kelder A, Rombouts EJ, Feller N, Moshaver B, et al. · Blood 2007
    PMID 17609428DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. **This is the single load-bearing citation for the whole CLEC12A therapeutic rationale and its normal-marrow denominators are n=11 and n=6.** 'Completely absent' on eleven donors bounds the frequency of a CLL-1-positive normal CD34+CD38- population only loosely - with 11 donors an event occurring in, say, 10% of donors is missed about 31% of the time. The regenerating-marrow control (n=6) is the more important of the two for a patient recovering from induction, and it is the smaller. NOD/SCID engraftment underestimates human HSC activity relative to NSG.

    Evidence that argues against this (1)
    • The same paper reports that in remission marrow, BOTH CLL-1-negative normal and CLL-1-positive malignant CD34+CD38- cells were present. So in the clinical state where a CLEC12A agent would most plausibly be used as consolidation, the normal and malignant stem compartments coexist and are separated only by this one antigen. Xiu Y et al. 2018
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 36 paediatric AML bone marrow samples versus 13 age-matched healthy donors, CD33 and CLEC12A were the most upregulated immunotargets by RNA-seq and showed the highest surface expression by flow cytometry; CLEC12A and CD33/CLEC12A double-positive expression was absent in CD34+CD38-CD45RA-CD90+ haematopoietic stem cells and in non-haematopoietic tissue, while CD33 and FLT3 are expressed on HSCs. KMT2A-mutated infant AML clustered separately by RNA-seq and overexpressed FLT3, for which the authors propose CD33/FLT3 co-targeting as the subgroup-specific option.

    CLEC12A and CD33 coexpression as a preferential target for pediatric AML combinatorial immunotherapy. · Willier S, Rothamel P, Hastreiter M, Wilhelm J, Stenger D, Blaeschke F, et al. · Blood 2021
    PMID 33094319DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. **The KMT2A-relevant conclusion of this paper is NOT a CLEC12A conclusion**: the authors name CLEC12A and CD33 as the generic paediatric AML pair and name CD33/FLT3 as the KMT2A-mutated infant AML pair. Citing this paper as support for CLEC12A in KMT2A-rearranged disease would misread it. n=36 AML and n=13 donors is small for a subgroup claim, and the abstract does not give the KMT2A-mutated subgroup's size.

    Evidence that argues against this (1)
    • This paper is the strongest available argument AGAINST prioritising CLEC12A specifically in KMT2A-rearranged infant AML, because the same analysis that nominates CLEC12A generically nominates FLT3 for that subgroup instead. The 2026 menin preprint [E8] argues the opposite by a different route - induction rather than baseline expression - and the two are compatible only if menin inhibition changes the baseline the 2021 paper measured. Rausch J et al. 2026
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Across genetically defined AML subtypes, six putative LSC targets (CD25, CD26, CD47, CD96, CD123 and CLL-1) were all expressed across heterogeneous LSC phenotypes, and with the exception of CD47 there was at most low expression of these targets on normal haematopoietic stem cells; CD123 and CLL-1 demonstrated the greatest expression differences between putative LSCs and normal HSCs.

    Expression of putative leukemia stem cell targets in genetically-defined acute myeloid leukemia subtypes. · Yanagisawa B, Perkins B, Karantanos T, Levis M, Ghiaur G, Smith BD, et al. · Leuk Res 2020
    PMID 33220589DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. Cohort size and the per-subtype breakdown are not in the abstract, and the abstract does not state whether any KMT2A-rearranged cases were included. **The wording that matters is 'at most low expression', not 'absent'** - this is a quantitative-window claim, not an all-or-none one, and it is the more defensible form.

    Evidence that argues against this (1)
    • The paper's own therapeutic conclusion goes to CD123, not CLL-1: CD123 monoclonal antibodies were cytotoxic in vitro to putative LSCs from all AML subtypes with limited to no toxicity against normal HSCs and progenitors. A beat writing CLEC12A up as the best LSC target has to explain why this head-to-head comparison chose the other one. Xiu Y et al. 2018
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Quantification of target expression in relapsed/refractory AML patient samples and normal HSPCs revealed a therapeutic window for gated co-targeting of ADGRE2 and CLEC12A, engaging ADGRE2-positive/CLEC12A-positive leukaemic stem cells over ADGRE2-low/CLEC12A-negative normal HSPCs; the combination prevented antigen escape in xenografts, and off-target HSPC toxicity was similar to that of a CD19-CAR and could be mitigated by reducing CAR T cell-derived interferon-gamma.

    Cooperative CAR targeting to selectively eliminate AML and minimize escape. · Haubner S, Mansilla-Soto J, Nataraj S, Kogel F, Chang Q, de Stanchina E, et al. · Cancer Cell 2023
    PMID 37802054DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. Xenograft and humanised-mouse data, not clinical. The design deliberately uses CLEC12A as a **costimulatory** receptor rather than the primary CAR, which is itself an admission that CLEC12A alone was not considered safe enough to drive killing. The AML here is not KMT2A-stratified in the abstract.

    Evidence that argues against this (1)
    • 'Off-target HSPC toxicity similar to a CD19-CAR' is the sentence that undercuts the whole HSC-sparing framing: a CD19-CAR's HSPC toxicity is supposed to be near zero, but the comparison is reported as a floor the design reached rather than as zero, and the mitigation offered is cytokine reduction rather than target selectivity. Xiu Y et al. 2018
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the first-in-human phase I study of CLL-1 CAR-T cells in 10 adults with relapsed/refractory AML, all patients developed cytokine release syndrome (6 of 10 high grade), severe pancytopenia occurred in all patients, two patients died of severe infection due to chronic agranulocytosis, and the CR/CRi rate was 70% (7 of 10); the authors note that CLL-1 is also highly expressed on normal granulocytes and propose bridging haematopoietic stem cell transplantation to rescue long-term agranulocytosis due to off-target toxicity.

    First-in-human phase I study of CLL-1 CAR-T cells in adults with relapsed/refractory acute myeloid leukemia. · Jin X, Zhang M, Sun R, Lyu H, Xiao X, Zhang X, et al. · J Hematol Oncol 2022
    PMID 35799191DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. n=10, single arm, adults, median follow-up 173 days - far too small and too short for a safety conclusion, and adult R/R AML patients arrive with prior marrow damage that confounds attribution of pancytopenia to the CAR. **No paediatric or KMT2A-rearranged data.** The two deaths are the most decision-relevant numbers in this entry and they come from a ten-patient study.

    Evidence that argues against this (1)
    • A 70% CR/CRi rate in R/R AML is a strong efficacy signal and the counter-argument to reading this study as a safety failure: the question is not whether CLEC12A-directed therapy causes agranulocytosis - it plainly does - but whether that is acceptable when bridged to transplant, which is the same bargain already accepted for several AML regimens. Xiu Y et al. 2018
  8. E8weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.preprint

    A 2026 preprint reports that menin inhibition primes KMT2A-rearranged and NPM1-mutated AML for CAR-based targeting by inducing robust and uniform expression of CLEC12A, that menin inhibitors did not impair T or NK cell viability, phenotype or effector function, and that second-generation CLEC12A-directed CAR T cells combined with menin inhibition outperformed either monotherapy across in vitro systems and xenograft models.

    Menin-Inhibition Sensitizes Acute Myeloid Leukemia to CLEC12A-Directed CAR Cell Therapy · Rausch J, Wendel P, Dzama-Karels M, Steiner M, Gierschek F, Fetsch V, et al. · Preprint, not peer reviewed (Europe PMC PPR1223455) 2026
    DOI
    What this evidence cannot show

    **Preprint, not peer reviewed.** Abstract read in full from the Europe PMC core record (PPR1223455) on 2026-09-20; the DOI 10.64898/2026.02.15.703376 was copied from that same fetched record. Europe PMC reports `isOpenAccess: N` and no PMID or PMCID, so the full text was not retrieved through routes 1-4 and only the abstract and significance statement were read. Every quantitative claim - how much induction, in how many models, over what time - is therefore unread. Confidence is `weak` because the claim is unreviewed and read only at abstract level, not because the work looks poor.

    Evidence that argues against this (1)
    • Willier 2021, peer reviewed, nominates CD33/FLT3 rather than CLEC12A for KMT2A-mutated infant AML on baseline expression grounds. If menin-inhibitor-induced CLEC12A is real it reconciles the two, but until the preprint is reviewed the peer-reviewed literature points the other way for this specific subgroup. Willier S et al. 2021
  9. E9moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    C-type lectin-like molecule-1 was expressed in 86.5% (45 of 52) of AML cases and in 54.5% (12 of 22) of AML CD34+/CD38- stem cells but not in acute lymphoblastic leukaemia blasts (n=5); anti-CLL-1 monoclonal antibodies mediated complement-dependent and antibody-dependent cellular cytotoxicity against AML lines and against freshly isolated AML blasts in 15 of 16 cases, and were efficiently internalised on binding.

    Targeting C-type lectin-like molecule-1 for antibody-mediated immunotherapy in acute myeloid leukemia. · Zhao X, Singh S, Pardoux C, Zhao J, Hsi ED, Abo A, et al. · Haematologica 2010
    PMID 19648166DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. **The 54.5% figure is the important one and it contradicts the usual framing**: only about half of AML CD34+/CD38- stem-cell fractions were CLL-1 positive in this series, against 77/89 (87%) in van Rhenen 2007 [E3]. Different antibodies, different gates, different eras. The ALL denominator is n=5, which bounds nothing.

    Evidence that argues against this (1)
    • If only 54.5% of AML LSC fractions are CLEC12A positive, then a CLEC12A-directed agent leaves a CLEC12A-negative LSC fraction behind in roughly half of patients - which is the antigen-escape argument that motivated the cooperative ADGRE2/CLEC12A design [E6] and is the main reason CLEC12A monotherapy is not the plan anywhere. Haubner S et al. 2023
  10. E10established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Among 1,812 children with newly diagnosed BCP-ALL in trial AIEOP-BFM ALL 2009, 160 (8.8%) were CD371-positive at diagnosis; this correlated with older age, lower ETV6::RUNX1 frequency, immunophenotypic immaturity and strong CD34 and CD45 expression; CD371-positive cases showed a transient myelomonocytic switch in up to 65.4% of samples at day 15 and slow early response, yet 5-year event-free survival was 88.3%; and 27 of 28 (96.4%) DUX4-fusion-positive cases in the validation cohort were CD371-positive.

    CD371-positive pediatric B-cell acute lymphoblastic leukemia: propensity to lineage switch and slow early response to treatment. · Buldini B, Varotto E, Maurer-Granofszky M, Gaipa G, Schumich A, Bruggemann M, et al. · Blood 2024
    PMID 38215390DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read; the citation identifiers were re-fetched from PubMed for this entry rather than copied from the sibling entry `gene-cd33` that also cites this paper. **The abstract does not report the KMT2A-rearranged fraction of the CD371-positive group**, so this cannot be used to state how often KMT2A-r B-ALL is CD371-positive. Infants are largely absent from AIEOP-BFM ALL 2009 by protocol, so this says little about infant KMT2A-r ALL specifically. The strong association is with DUX4 fusions, not with KMT2A.

    Evidence that argues against this (1)
    • The 88.3% 5-year EFS is the counter to reading CD371 positivity in B-ALL as a danger signal: a marker can flag a large, reproducible, transient myeloid phenotypic shift and carry essentially no prognostic weight under MRD-directed therapy. It is also the counter to assuming that CD371 in B-ALL identifies the same biology as CLEC12A in AML. Xiu Y et al. 2018
  11. E11established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    MICL (CLEC12A) directly recognises DNA in neutrophil extracellular traps, and loss or inhibition of MICL functionality leads to uncontrolled NET formation through the ROS-PAD4 pathway and an auto-inflammatory feedback loop; inhibitory anti-MICL autoantibodies are found in patients with rheumatoid arthritis, lupus and severe COVID-19, while dysregulated NET release is protective during systemic Aspergillus fumigatus infection.

    Recognition and control of neutrophil extracellular trap formation by MICL. · Malamud M, Whitehead L, McIntosh A, Colella F, Roelofs AJ, Kusakabe T, et al. · Nature 2024
    PMID 39143217DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. Mouse models plus human patient autoantibody data; the human arm is an association between autoantibody presence and disease, not a controlled blockade experiment. Nothing in this paper concerns leukaemia or CAR therapy - the inference to CLEC12A-directed therapeutics is mine, and it is an inference.

    Evidence that argues against this (1)
    • The Aspergillus result cuts the other way: dysregulated NET release was protective during systemic fungal infection, so blocking MICL is not uniformly harmful and in a neutropenic leukaemia patient at fungal risk the sign of the effect is genuinely unclear. Xiu Y et al. 2018
  12. E12established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    The reviewed human UniProt record for CLEC12A is Q5QGZ9 (CL12A_HUMAN), primary gene name CLEC12A, 265 aa, gene synonyms CLL1 DCAL2 KLRL1 MICL, CD antigen CD371; the neighbouring paralogue CLEC12B is Q2HXU8, 276 aa; and a `gene:CLL1 AND organism_id:9606 AND reviewed:true` query returns COLEC10 (Q9Y6Z7, Collectin-10 / CL-L1, 277 aa) ahead of CLEC12A.

    UniProtKB reviewed entry Q5QGZ9 (CL12A_HUMAN) · UniProt Consortium · UniProtKB 2026
    What this evidence cannot show

    Fetched live from the UniProt REST API on 2026-09-20 with fields accession, gene_primary, gene_synonym, protein_name, length; every row was read and only the row whose primary gene name equalled the queried symbol was kept, per rule 62. A database record states nomenclature, not biology: the 265 aa length is the canonical isoform and says nothing about the surface form's glycosylation or about isoform usage in leukaemic cells.

    Evidence that argues against this (1)
    • A UniProt canonical length is opposable on scope exactly as the brief says of other database facts: it is the reviewed canonical sequence, and an entry asserting 265 aa is asserting which record was used, not that leukaemic cells express a 265-residue polypeptide. UniProt Consortium 2026
  13. E13established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    HGNC record HGNC:31713 gives CLEC12A as the approved symbol with alias symbols CLL-1, MICL, CD371 and DCAL-2, no previous symbols, and UniProt id Q5QGZ9; an HGNC alias_symbol lookup on the unhyphenated string CLL1 returns no records at all.

    HGNC record HGNC:31713 (CLEC12A) · HUGO Gene Nomenclature Committee · HGNC 2026
    What this evidence cannot show

    Fetched live from the HGNC REST API on 2026-09-20 (`/fetch/symbol/CLEC12A` and `/search/alias_symbol/CLL1`). The negative half of this claim is a property of HGNC's indexing of hyphenation on one day, not a stable fact; it should be re-run before being relied on. It is recorded because it is the mechanism by which a one-database alias resolver silently disagrees with a two-database one.

    Evidence that argues against this (1)
    • HGNC's own free-text search on CD371 does return CLEC12A, so the database is not blind to the CD number - the failure is specific to the unhyphenated CLL1 string in the alias field. An implementer should not conclude from this that HGNC is unusable for alias resolution. HUGO Gene Nomenclature Committee 2026
  14. E14strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the phase 1 dose-escalation study of KITE-222, an autologous CLL-1-directed CAR T-cell therapy, 12 patients with relapsed/refractory AML were treated across three dose levels; safety was acceptable (one DLT of prolonged grade 4 cytopenias, no grade >=3 CRS), but clinically meaningful responses were lacking at every dose despite detectable CAR T-cell expansion, and in two of five cohort-3 patients with clear expansion the CLL-1-positive marrow blasts were nearly completely depleted while CLL-1-negative blasts persisted and total blasts did not fall.

    Phase 1 Study of KITE-222, an Autologous CLL-1-Directed CAR T-cell Therapy in Patients with Relapsed/Refractory Acute Myeloid Leukemia. · Daver N, Blachly JS, Ghobadi A, Advani A, Muffly L, Garciaz S, et al. · Clin Cancer Res 2026
    PMID 41941266DOI
    What this evidence cannot show

    Abstract read in full from raw NCBI efetch XML on 2026-09-20; full text not read. n=12, dose escalation, adults with R/R AML - a negative efficacy result from a phase 1 is not proof that the target is wrong, and the authors themselves attribute the failure to CLL-1 heterogeneity and poor in vivo expansion rather than to the target being invalid. No paediatric and no KMT2A-rearranged patients are described in the abstract. **This trial did not combine with a menin inhibitor**, so it does not test the 2026 preprint's proposal [E8].

    Evidence that argues against this (1)
    • The trial's own framing states CLL-1 is 'predominantly expressed on myeloid cells but absent on normal hematopoietic stem cells and other tissues' - i.e. the sponsor's own rationale restates the HSC-absence claim, and the observed safety profile (no grade >=3 CRS, one cytopenia DLT) is consistent with it. The failure here was efficacy, not the selectivity premise. Xiu Y et al. 2018
  15. E15strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Human MICL/CLEC12A is a surface receptor of granulocytes, monocytes, macrophages and dendritic cells whose SURFACE level is actively regulated - it is down-regulated under inflammatory and activation conditions in vitro and during an in vivo model of acute inflammation - and it is variably N-glycosylated between cell types.

    Human MICL (CLEC12A) is differentially glycosylated and is down-regulated following cellular activation. · Marshall AS, Willment JA, Pyz E, Dennehy KM, Reid DM, Dri P, Gordon S, Wong SY, Brown GD · Eur J Immunol 2006
    PMID 16838277DOI
    What this evidence cannot show

    Normal myeloid cells only; no leukaemia. The regulated down-modulation matters for a CLEC12A-directed therapeutic because it means surface antigen density is a state variable and not a fixed property of the cell, but that has never been tested in leukaemic blasts. Differential glycosylation between cell types also means an antibody's apparent avidity can differ between normal and leukaemic myeloid cells without any change in copy number.

    Evidence that argues against this (1)
    • Surface presence is not the whole of the localisation story: murine Clec12a is EXPELLED from lipid rafts on monosodium urate stimulation, and the inhibitory signalling it delivers depends on where in the membrane it sits rather than merely on whether it is there. A flow-cytometry-positive cell can therefore be functionally different from another flow-cytometry-positive cell, and no CLEC12A-directed therapeutic programme in this graph measures raft residence. Xu Y et al. 2023
  16. E16established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    CLEC12A HAS NO CHROMATIN MODULE, WHICH IS WHY THE EPIGENETIC ACTIVITY QUESTION IS NOT APPLICABLE RATHER THAN UNANSWERED: UniProt Q5QGZ9 (primary gene CLEC12A, 265 aa, fetched 2026-09-21) annotates a cytoplasmic region 1-43 containing an ITIM motif at 5-10, a helical type II signal-anchor transmembrane segment at 44-64, an extracellular region 65-265, and a single C-type lectin domain at 140-249, with Pfam cross-references consisting of PF00059 (Lectin_C) alone and no methyltransferase, demethylase, acetyltransferase, deacetylase, bromodomain, chromodomain, Tudor, PHD, CXXC, YTH or KH domain.

    What this evidence cannot show

    A domain-architecture argument, not an experiment: absence of an annotated chromatin module makes an epigenetic activity implausible and does not formally exclude a moonlighting function, and UniProt feature tables are incomplete for poorly studied regions. The type II topology places the sole globular domain extracellularly, so any nuclear function would require an unreported cleaved or alternatively targeted species - and this entry records no nuclear pool. This item bounds what CLEC12A can do to chromatin; it says nothing about what chromatin does to CLEC12A, which is the live question and is handled separately.

    Evidence that argues against this (1)
    • The not-applicable call is about the protein's activity and must not be read as 'epigenetics is irrelevant to CLEC12A', because the thing that matters clinically about this antigen is its level, and its level is what an epigenetic drug is claimed to move. The founding immunotherapy characterisation makes the stakes explicit: CLL-1/CLEC12A is expressed in 86.5% of AML cases but in only 54.5% of AML CD34+/CD38- stem cells, and antibody-negative fractions are precisely the reservoir that defeats a surface-directed agent. A drug that raises CLEC12A on the negative fraction would be therapeutically decisive whether it does so through chromatin or through differentiation, so the mechanism question is not academic - it determines which drug class to combine. Zhao X et al. 2010
Connected to