FBXL19 - a CpG-island reader that delivers CDK-Mediator and RNF20 to developmental promoters, and whose experimental localisation is mostly cytoplasm
also called FBXL19 · F-box/LRR-repeat protein 19 · FBXL19-AS1 (the antisense RNA, a different entity)
FBXL19 reads unmethylated CG islands and brings in two machines the KMT2A fusion also depends on: the Mediator complex that primes genes for switching on, and the enzyme that tags histone H2B. It is also, awkwardly, mostly found in the cytoplasm when anyone has measured it.
The same thing, for a scientist
FBXL19 (Q6PCT2, 694 aa) carries a **CXXC-type zinc finger at 32-78** immediately followed by a PHD finger at 85-151, with an F-box at 418-464 [E1] - the same tandem CXXC-PHD-plus-F-box architecture as KDM2A and KDM2B, without the JmjC domain. Its curated DOMAIN comment states the CXXC binds DNA containing unmethylated CpG [E1][E2]. Two independent studies give it CpG-island jobs that matter to this project: **FBXL19 is a CpG-island-binding protein that associates with CDK-Mediator and recruits it to CpG-island promoters of non-transcribed developmental genes, priming them for activation during lineage commitment**, with recognition of CpG islands by FBXL19 essential for mouse development [E3]; and **Fbxl19 promotes H2B K120 mono-ubiquitination at CpG-island promoters by interacting with Rnf20**, with genome-wide occupancy preferentially at CpG-island-containing promoters [E4]. Both are mouse ES-cell systems. **The awkward fact this entry insists on: UniProt's experimental subcellular annotation for FBXL19 is Cytoplasm three times over (ECO:0000269) and Nucleus once**, and its best-characterised molecular function is as an SCF substrate-recognition subunit degrading RHOA, RAC1 and RAC3 - a cytoplasmic job [E1]. A protein predominantly measured in the cytoplasm is a weak candidate competitor for a chromatin address, and this entry states that rather than letting the CpG-island literature imply otherwise. A second, smaller trap sits in the same place: UniProt carries a CAUTION that **it is uncertain whether Met-1 or Met-21 is the initiator** [E1] - a 20-residue ambiguity immediately N-terminal to a CXXC domain that starts at residue 32. In DepMap it is nothing (difference -0.016, rank 7,031 of 17,728) [E5].
This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.
I have never read one of these pictures before
A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.
Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.
Anything in yellow sticks is a bound small molecule, which is usually the drug.
Why this one is relevant to the index disease at all
Two of FBXL19's reported deliveries are machines this graph already treats as load-bearing for the KMT2A fusion:
- CDK-Mediator. FBXL19 recruits it to CpG-island promoters of non-transcribed developmental genes and primes them for activation on lineage commitment 3. This graph holds
ana-mediator-is-pan-essential-except-med13landgene-med12; the CpG-island route to Mediator is not recorded anywhere in it. - RNF20 and H2Bub1. Fbxl19 interacts with Rnf20 and its recruitment to CpG islands is required for H2B mono-ubiquitination at those promoters 4. This graph holds an entire H2Bub arm -
ana-depmap-h2bub-machinery-kmt2a-selectivity,gene-rnf20,mechanism-h2bub-dot1l-trans-histone-crosstalk,hyp-h2bub-direction-is-unknown-in-kmt2a- and none of those entries knows that a CXXC reader is one route by which RNF20 reaches CpG islands.
So FBXL19 is a candidate for a specific and testable role: a CXXC-family competitor that, if it occupies a fusion target island, would be delivering H2Bub1 and CDK-Mediator to the same place the fusion needs them. That is co-operation, not competition - and it is the reading the null in hyp-the-cxxc-address-is-a-contested-resource says to prefer until someone measures occupancy.
The localisation problem, stated in full
UniProt's subcellular-location block for FBXL19 lists Cytoplasm with three separate ECO:0000269 experimental supports and Nucleus with one 1. Its most fully characterised function - SCF-mediated ubiquitination and proteasomal degradation of RHOA in an ERK2-dependent manner, and of RAC1 and RAC3 in TGFB1-induced responses - is cytoplasmic 1.
Against that, the two CpG-island papers are chromatin-immunoprecipitation studies in mouse ES cells with genome-wide occupancy maps 34, which is direct evidence of nuclear chromatin binding in that system.
The honest synthesis is that FBXL19 is a dual-compartment protein whose nuclear pool has been characterised in mouse ES cells and whose cytoplasmic pool dominates the human experimental record. For this beat the operational consequence is that FBXL19 should not be counted as a competitor for the fusion's address without a measurement of its nuclear fraction in a haematopoietic cell, which does not exist. Together with CXXC4, that removes two of twelve family members from the competition question on current evidence.
Names
FBXL19[tiab] returns 79 records and the symbol is not obviously contaminated, but two hazards are worth recording. First, FBXL19-AS1 is a separate, well-published lncRNA and a keyword search on FBXL19 returns its papers; they are about a different molecule. Second, this protein sits in a paralogue set with FBXL10 (KDM2B) and FBXL11 (KDM2A), whose symbols differ by one digit and whose domain architectures are nearly identical - the exact shape of the SMARCA2/SMARCA4 trap this project has logged. Proposed as a registry row; see the memo.
| uniprot | Q6PCT2 |
|---|---|
| gene symbol | FBXL19 |
| protein name | F-box/LRR-repeat protein 19 |
| length AA | 694 |
| alphafold ID | AF-Q6PCT2-F1 |
| accession resolution | UniProt REST gene_exact:FBXL19 AND organism_id:9606 AND reviewed:true, run 2026-09-21; one row; primary gene name read back as FBXL19 and length 694 from the same fetched record. |
| alias hazard | FBXL19-AS1 is a separate, well-published lncRNA whose papers are returned by an FBXL19 keyword search. The paralogue symbols FBXL10 (KDM2B) and FBXL11 (KDM2A) differ by one digit and share the CXXC-PHD-F-box architecture. |
| numbering hazard | UniProt CAUTION: uncertain whether Met-1 or Met-21 is the initiator. The CXXC domain is annotated at 32-78 on the Met-1 numbering; a construct built on the Met-21 start shifts every residue index by 20. |
| function | CpG-island reader that recruits CDK-Mediator and RNF20 to CpG-island promoters; separately an SCF substrate-recognition subunit degrading RHOA, RAC1 and RAC3 |
| classification | CXXC-family reader with the KDM2A/KDM2B architecture minus the JmjC domain; a candidate CO-OPERATOR rather than competitor, because both its chromatin deliveries are machines the fusion also uses |
| shown for fusion | NO - nobody has looked in any fusion |
| cell type evidence | mouse embryonic stem cells and mouse development (both chromatin studies); human cell lines (the cytoplasmic SCF work). No haematopoietic and no KMT2A-rearranged material. |
| epigenetic readout | H2B K120 mono-ubiquitination at CpG-island promoters; CDK-Mediator occupancy |
| subnuclear localization | CpG-island promoters genome-wide in mouse ES cells by ChIP (PMIDs 29809150, 28453857). The human experimental record is predominantly CYTOPLASMIC - three ECO:0000269 cytoplasmic supports to one nuclear - so the nuclear fraction in a haematopoietic cell is unmeasured. Never imaged in any KMT2A-fusion cell. |
| cell type | Mouse embryonic stem cells and mouse development (both chromatin studies); human somatic cell lines (the cytoplasmic SCF work). No haematopoietic and no KMT2A-rearranged material. |
domains — the numbers behind the chart
[
{
"name": "CXXC (CpG-binding zinc finger)",
"start": 32,
"end": 78,
"pfam": "PF02008"
},
{
"name": "PHD (plant homeodomain) zinc finger",
"start": 85,
"end": 151,
"pfam": "PF16866"
},
{
"name": "F-box",
"start": 418,
"end": 464,
"pfam": "PF12937"
}
]localization
{
"subcellular": "cytoplasm (three ECO:0000269 supports) and nucleus (one ECO:0000269) - the experimental record is predominantly CYTOPLASMIC, which is in tension with the CpG-island literature",
"subnuclear": "CpG-island promoters genome-wide in mouse ES cells by ChIP; no human subnuclear data",
"localization_shown_for_fusion": "NO"
}essentiality
{
"depmap_scan": "analyses/depmap-kmt2a-selective-scan-full",
"difference_kmt2a_r_minus_other": -0.0161,
"mean_kmt2a_rearranged": -0.1468,
"mean_other_leukemia": -0.1307,
"pan_essentiality_all_lines": -0.0953,
"n_all_lines": 1208,
"rank_of_17728": 7031,
"percentile": 39.66,
"release_note": "Open Targets dataVersion 26.06, read 2026-09-21"
}- If FBXL19 delivers H2Bub1 and CDK-Mediator to CpG islands, a molecule that blocks CXXC-DNA binding across the family would remove two activities the fusion needs at the same time as removing the fusion's own address - a confound that would make any such agent's mechanism uninterpretable.
- FBXL19 is also an SCF substrate-recognition subunit for RHOA and RAC GTPases. Interfering with it has a cytoskeletal and migratory arm that has nothing to do with chromatin, and in a paediatric patient that is an unexamined liability.
- This graph's H2Bub arm does not know that a CXXC reader delivers RNF20 to CpG islands. Does FBXL19 contribute the H2Bub1 at KMT2A-fusion target islands, and would depleting it phenocopy RNF20 loss there?
- What fraction of FBXL19 is nuclear in a haematopoietic cell? The human experimental record is three cytoplasmic supports to one nuclear, and without that number FBXL19 cannot be counted as a competitor for the fusion's address.
- Do the CDK-Mediator and RNF20 mechanisms operate at the same islands or different ones? Two papers, two mechanisms, no cross-citation of a shared model.
- Is the Met-1/Met-21 ambiguity resolved anywhere? It determines whether the CXXC domain starts at residue 32 or residue 12.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
FBXL19 (Q6PCT2, 694 aa) carries a CXXC-type zinc finger at 32-78, a PHD finger at 85-151 and an F-box at 418-464; its curated DOMAIN comment states the CXXC binds DNA containing unmethylated CpG; its experimental subcellular annotation is Cytoplasm three times (ECO:0000269) and Nucleus once; and a CAUTION records that it is uncertain whether Met-1 or Met-21 is the initiator.
UniProtKB Q6PCT2 (FXL19_HUMAN), F-box/LRR-repeat protein 19 · UniProt Consortium · UniProtKB 2026What this evidence cannot show
Fetched 2026-09-21; accession resolved by gene_exact:FBXL19 AND organism_id:9606 AND reviewed:true, one row, primary gene name read back as FBXL19 and length 694 from the same record. Three isoforms are annotated. The initiator-methionine ambiguity means the residue numbering of the CXXC domain could be offset by 20 - which is the same class of error as the ASH2L 94-residue offset this project has documented, and any structural work on this domain must be checked against the construct used.
Evidence that argues against this (1)
- Two ChIP studies map FBXL19 across CpG-island promoters genome-wide in mouse ES cells, which is stronger evidence of a nuclear chromatin function than the count of UniProt localisation supports suggests. Dimitrova E et al. 2018
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
FBXL19's CXXC domain was measured alongside every other human CXXC domain in the systematic family comparison of DNA-binding selectivity.
DNA Sequence Recognition of Human CXXC Domains and Their Structural Determinants. · Xu C, Liu K, et al. · Structure 2018What this evidence cannot show
Isolated domain, short duplexes, in vitro; FBXL19 carries a tandem CXXC-PHD module, which that study treats as a separate structural class, so the isolated-CXXC measurement may not describe the tandem's behaviour.
Evidence that argues against this (1)
- The tandem CXXC-PHD arrangement is shared with KDM2A and KDM2B, whose PHD fingers contribute to chromatin binding, so a CXXC-only affinity understates the module. Blackledge NP et al. 2010
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
FBXL19 is a CpG-island-binding protein in mouse embryonic stem cells that associates with the CDK-Mediator complex and recruits it to CpG-island-associated promoters of non-transcribed developmental genes, priming them for activation during lineage commitment, and recognition of CpG islands by FBXL19 is essential for mouse development.
FBXL19 recruits CDK-Mediator to CpG islands of developmental genes priming them for activation during lineage commitment. · Dimitrova E, Kondo T, et al. · Elife 2018What this evidence cannot show
Mouse embryonic stem cells and mouse development. The 'priming' claim rests on expression changes on differentiation rather than on a direct measurement of pre-initiation-complex assembly. Nothing haematopoietic and nothing malignant.
Evidence that argues against this (1)
- A second study attributes FBXL19's CpG-island function to RNF20 recruitment and H2B mono-ubiquitination rather than to CDK-Mediator; the two mechanisms have not been tested against each other. Lee BK et al. 2017
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Fbxl19 promotes H2B K120 mono-ubiquitination at CpG-island-containing promoters by interacting with Rnf20, occupies CpG-island promoters preferentially genome-wide, and its chromosomal binding is required for H2Bub1 at its targets and for proper ES-cell differentiation.
Fbxl19 recruitment to CpG islands is required for Rnf20-mediated H2B mono-ubiquitination. · Lee BK, Lee J, et al. · Nucleic Acids Res 2017What this evidence cannot show
Mouse embryonic stem cells; up- and down-regulation of Fbxl19 rather than acute degradation, so indirect effects are not excluded. H2Bub1 is measured globally and by ChIP at targets. No leukaemia system.
Evidence that argues against this (1)
- The same protein's other reported nuclear job is recruiting CDK-Mediator, and the two studies do not cite a shared mechanism, so which activity dominates at a given island is unknown. Dimitrova E et al. 2018
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
FBXL19 shows no dependency in KMT2A-rearranged cell lines: difference -0.0161, rank 7,031 of 17,728, pan-essentiality -0.0953 across 1,208 lines.
Genome-wide KMT2A-rearranged selectivity scan, 17,728 genes (DepMap CRISPR Chronos via the Open Targets Platform GraphQL API) · Oncofusion knowledge-graph project · analyses/depmap-kmt2a-selective-scan-full/results.json, generated 2026-09-19; Open Targets dataVersion year 26 month 06, re-read from the API on 2026-09-21 2026What this evidence cannot show
The scan compares 11 Cellosaurus-annotated KMT2A-rearranged lines against 43 other leukaemia lines; none of the 11 is infant-derived and the arm is myeloid-enriched, so a difference mixes fusion status with lineage. Both the difference and the pan-essentiality across all 1,208 lines are quoted, because a gene required everywhere is not a selective dependency. Open Targets dataVersion 26.06; DepMap's own release identifier is not returned by that endpoint.
Evidence that argues against this (1)
- FBXL19's characterised nuclear role is in priming non-transcribed developmental genes for later activation, a function that a proliferation-based CRISPR screen in established lines is poorly designed to detect. Dimitrova E et al. 2018
Binds unmethylated CpG islands and delivers CDK-Mediator and RNF20 there - the same address class, with cargo the fusion also needs.
One of the eleven other human CXXC domains a ligand for this interface must discriminate against; tandem CXXC-PHD architecture like KDM2A and KDM2B.
Fbxl19 interacts with Rnf20 and its CpG-island recruitment is required for H2B K120 mono-ubiquitination at those promoters - a CXXC route into this graph's H2Bub arm that the arm did not know about.
A CpG-island-targeted route to H2Bub1 that is not in that analysis's gene set.
One of the twelve human CXXC proteins.
Shares the tandem CXXC-PHD plus F-box architecture, without the JmjC domain; FBXL19, FBXL11 (KDM2A) and FBXL10 (KDM2B) are one symbol series.