MLLT10/AF10: twenty observed phosphosites, twelve PIKK motifs, and not one kinase-site relation in the literature - a bounded absence
also called AF10 phosphorylation absence · MLLT10 no kinase · MLLT10 bounded absence
Phosphates have been seen on the AF10 protein in twenty places, but nobody has ever shown which enzyme puts any of them there or what any of them does. We report the exact queries and counts, because 'nobody looked' and 'someone looked and found nothing' are different things.
The same thing, for a scientist
No published kinase-substrate relation exists for MLLT10 at any residue. This is a counted absence over a stated symbol set, not an inference from silence, and it sits beside a substantial positive: the PeptideAtlas/PTMeXchange re-analysis reports 20 phosphosites on P55197 and UniProt curates 7, and a scan of the canonical sequence finds 12 S/T-Q motifs - the PIKK consensus that is phosphorylated on all three of the other partners in this beat. So AF10 has abundant substrate opportunity and zero mechanistic characterisation. **This is an absence of testing, not a negative result**: no paper reports looking for AF10 phosphorylation and failing to find it. The one kinase relation that does exist for this partner runs the other way - AF10 fusions recruit and stabilise JAK1, with no site on AF10 mapped.
The queries, run 2026-09-19
| query | records |
|---|---|
(MLLT10[tiab] OR AF10[tiab] OR "AF-10"[tiab] OR ZFP-1[tiab]) AND (phosphorylation OR phosphorylated OR phosphorylates OR kinase OR phosphosite OR phosphoproteomic OR phosphoproteomics)[tiab] | 25 |
(MLLT10[tiab] OR AF10[tiab]) AND (phosphosite[tiab] OR "phosphorylation site"[tiab]) | 1 |
(MLLT10[tiab] OR AF10[tiab]) AND (serine OR threonine OR Ser OR Thr)[tiab] | 6 |
(MLLT10[tiab] OR MLLT1[tiab] OR MLLT3[tiab] OR ELL[tiab] OR ELL2[tiab] OR AF9[tiab] OR AF10[tiab]) AND (CK2 OR CSNK2 OR "casein kinase 2")[tiab] | 1 |
Of the 25 records in the broadest query, title inspection finds: one JAK/STAT paper on AF10 fusions that maps no site on AF10; one DDX3X-MLLT10 case report about JAK3 mutations; one DOT1L/BCL-2 combination paper using PI3K/AKT as a downstream readout; one GWAS of chronic dizziness naming the MLLT10 locus; and the remainder are 'AF-10' as the name of an unrelated IL-6-responsive B-cell line, C. elegans ZFP-1/AF10 RNAi and insulin-signalling papers, and veterinary and cardiology records where AF10 is an abbreviation for something else entirely. No record maps a phosphorylation site on the AF10 protein.
The six-record serine/threonine query is the sharpest bound and its hits are all incidental: a kinase-interacting-stathmin substrate paper about CATS/FAM64A, an HP1a peptide-selection paper, and four 1990s immunology papers using the AF-10 cell line.
Alias precision, stated because this project has been burned by it
AF10 is a word-like token and over-retrieves. It matches the AF-10 B-cell line, the C. elegans gene ZFP-1/AF10 and, in cardiology and veterinary records, unrelated abbreviations. The raw counts are AF10[tiab] 280, "AF-10"[tiab] 184, MLLT10[tiab] 178, so the union is substantially larger than the real literature and every hit had to be title-checked. The counted absence above is therefore a checked number, not a raw one.
The retired- and alternative-symbol rule was applied: MLLT10, AF10, AF-10 and ZFP-1 were all searched. ZFP-1 is the C. elegans orthologue symbol and was included precisely because this project has previously under-counted by a factor of ten by keying to current human symbols only. It adds no phosphorylation records.
What could still be missed. A complex-level paper that frames its findings at the level of DOT1L, the DOT1L complex or 'AF10 complex' without naming the subunit would be unreachable from any of these queries - the same trap that produced a wrong absence for MED23 in a previous cycle. And a phosphosite reported only in a supplementary table, with the protein named nowhere in title or abstract, is invisible to a [tiab] query by construction; the KMT2A-rearranged AML patient table this graph holds is exactly such a source, and it does report two MLLT10 sites.
What exists instead, and what it is not
Two observations stand where a mechanism would be:
- Twenty observed phosphosites on P55197 in the atlas re-analysis, seven curated by UniProt, none with a kinase 1.
- Two sites in patient material - canonical S658 and S689 after this beat's isoform correction - with no kinase, no stoichiometry and no function 2.
- Twelve S/T-Q motifs at 73, 223, 390, 406, 428, 625, 636, 927, 980, 1017, 1028 and 1050, of which the last several are retained in the fusion 1. Given that ATM or DNA-PKcs phosphorylates ENL, AF9 and ELL at S/T-Q sites, AF10 being untested at any of twelve such motifs is the most conspicuous hole this beat found.
None of that is a kinase-site relation and none of it should be written up as one.
| status | run |
|---|---|
| gene | gene-mllt10 |
| uniprot | P55197 |
| kinase site relations found | 0 |
| observed phosphosites atlas | 20 |
| observed phosphosites uniprot | 7 |
| observed in patients | 2 |
| alias precision note | AF10 is word-like and over-retrieves; every hit was title-checked and the reported number is the checked one |
| absence class | never-measured |
| what would overturn this | A single in vitro kinase assay or phosphoproteomic study reporting a kinase for any AF10 residue; a complex-level paper naming DOT1L or the AF10 complex rather than the subunit, which none of these queries can reach; or a supplementary table reporting an AF10 site with a kinase, which a [tiab] query cannot see. |
| result summary | Zero kinase-site relations for MLLT10 across four stated queries with four aliases and full title checking, against 20 observed sites and 12 PIKK motifs. |
| gene symbol | MLLT10 |
| absence class previous freetext | absence of testing, not a negative result - no study reports searching for AF10 phosphorylation and failing |
| absence class basis | The twenty phosphosites were observed, but no study assigns a kinase to any of them, which requires an experiment nobody has run. |
| absence class assigned | 2026-09-21 (cycle 47, king) |
st q motifs
[ 73, 223, 390, 406, 428, 625, 636, 927, 980, 1017, 1028, 1050 ]
queries — the numbers behind the chart
[
{
"query": "(MLLT10[tiab] OR AF10[tiab] OR \"AF-10\"[tiab] OR ZFP-1[tiab]) AND (phosphorylation OR phosphorylated OR phosphorylates OR kinase OR phosphosite OR phosphoproteomic OR phosphoproteomics)[tiab]",
"records": 25,
"on_topic_after_title_check": 0
},
{
"query": "(MLLT10[tiab] OR AF10[tiab]) AND (phosphosite[tiab] OR \"phosphorylation site\"[tiab])",
"records": 1,
"on_topic_after_title_check": 0
},
{
"query": "(MLLT10[tiab] OR AF10[tiab]) AND (serine OR threonine OR Ser OR Thr)[tiab]",
"records": 6,
"on_topic_after_title_check": 0
},
{
"query": "BROKEN AS STORED, flagged 2026-09-22: the word 'partners' is a PLACEHOLDER that leaked out of the query template and was never substituted, so this string cannot be re-run as written and its record count of 1 is not attributable to any real query. The intended form was almost certainly (MLLT10 OR AF10)[tiab] AND (CK2 OR CSNK2 OR casein kinase 2)[tiab]. NOT re-run this cycle; the surrounding queries in this list are unaffected.",
"records": 1,
"on_topic_after_title_check": 0
}
]aliases searched
[ "MLLT10", "AF10", "AF-10", "ZFP-1" ]
- The practical risk of this absence is the opposite of most: because JAK inhibitors are approved and the JAK/STAT result is strong, a programme could treat MLLT10 as a solved kinase problem and never ask the phosphorylation question at all. A kinase bound to a fusion and a kinase acting on a fusion are different objects and only the first is established here.
- Is AF10 phosphorylated after DNA damage at any of its twelve S/T-Q motifs? Three of the four partners in this beat are PIKK substrates and AF10 has never been asked.
- What kinase produces canonical S689, which sits 15 residues before the DOT1L-binding OM-LZ and is seen in patient material?
- Is AF10 a JAK1 substrate? JAK1 is physically held against the fusion and nobody has run the kinase assay.
- Would a complex-level search strategy - querying 'DOT1L complex' or 'DotCom' as phrases rather than subunit symbols - reach a paper these queries cannot?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
The PeptideAtlas/PTMeXchange re-analysis reports 20 phosphosites on MLLT10 (P55197) and UniProt curates 7, none with an assigned kinase; the canonical sequence contains 12 S/T-Q PIKK-consensus motifs, at residues 73, 223, 390, 406, 428, 625, 636, 927, 980, 1017, 1028 and 1050.
What this evidence cannot show
Motif presence is not evidence of phosphorylation and this project's standing rule is that a short linear motif match means almost nothing on its own. Twelve S/T-Q motifs in 1068 residues is close to the number expected by chance for a protein of that composition, so the motif count is included as a description of substrate opportunity, not as a prediction.
Evidence that argues against this (1)
- The comparator that disciplines this: AF9 has exactly ONE S/T-Q motif in 568 residues and that one is the mapped functional site, whereas AF10 has twelve. A protein with many motifs gives a PIKK many chances and gives an analyst many false leads; the motif argument is strong for AF9 for precisely the reason it is weak for AF10. UniProt Consortium 2026
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The only MLLT10 phosphosites reported in KMT2A-rearranged patient material are two sites from a 74-patient AML phosphoproteomic study, which assigns no kinase to any site in its table.
Integrative phosphoproteomics defines two biologically distinct groups of KMT2A rearranged acute myeloid leukaemia with different drug response phenotypes. · Casado P, et al. · Signal Transduct Target Ther 2023What this evidence cannot show
Reached through this graph's ana-phosphosites-observed-in-kmt2a-r-aml-patients entry rather than by re-parsing the supplementary file in this session. The values are presence/absence row labels with no per-patient counts, and their residue numbering required an isoform correction, documented in ana-mllt10-patient-phosphosite-isoform-resolution. The cohort is AML only.
Evidence that argues against this (1)
- That table's own limitation cuts both ways: KMT2A S136 has direct functional evidence and appears nowhere in it for peptide-coverage reasons, so the two MLLT10 sites are a floor on what is present in patients, not a census, and a site's absence from the table is not evidence against it. EMBL-EBI 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The nearest thing to a kinase relation for this partner runs in the opposite direction: AF10 fusion proteins directly recruit and stabilise JAK1, activating a JAK/STAT inflammatory cascade, without any phosphorylation site on AF10 being mapped.
A JAK/STAT-mediated inflammatory signaling cascade drives oncogenesis in AF10-rearranged AML. · Chen BR, et al. · Blood 2021What this evidence cannot show
Full text read 2026-09-19. The authors attribute increased JAK1 phosphorylation in fusion-on cells to higher total JAK1 rather than to increased activity, and they do not test whether AF10 is a JAK1 substrate.
Evidence that argues against this (1)
- A published clinical failure limits how far the JAK axis can be taken in this partner: a DDX3X-MLLT10 T-lymphoblastic leukaemia with activating JAK3 mutations did not achieve an objective response to tofacitinib. Wong J et al. 2020
zero kinase-site relations over four queries
- related to →AF10 fusions recruit and stabilise JAK1 - a kinase bound to the fusion with no phosphosite on the fusion
the one kinase relation runs the other way
- related to →KMT2A::MLLT10 (MLL-AF10), t(10;11)(p12;q23)
no phospho-regulation known for this fusion
same title-checking discipline applied to AF10
MLLT10 is the untested partner the hypothesis predicts about
counted absence 3
SURVIVES: 23 records, 0 on topic, and the entry's own AF10 over-retrieval warning is confirmed
confirms that entry's own alias_precision_note: AF10 over-retrieves, and the interference includes a B cell line of the same name