IKK/NF-kB retains MLL on chromatin — BOTH the fusion and wild-type MLL, at two promoters, in mouse myeloid AML, and it is occupancy loss rather than protein loss
also called NF-kB-dependent MLL retention · IKK inhibition evicts MLL-AF10 · Kuo 2013 retention result
Blocking the IKK kinases makes the leukaemia fusion protein fall off the two genes that matter most, even though the fusion protein itself is still present in the cell in normal amounts. The same treatment also knocks normal MLL off those genes and strips away the chemical marks both of them leave behind.
The same thing, for a scientist
**What was measured.** Mouse MLL-AF10 AML cells carrying an HA-tagged fusion were treated with IKK inhibitors and chromatin-immunoprecipitated at the **Hoxa9 and Meis1 promoters** (Ey-globin and H4 as controls). Five things fell together: Rela occupancy, **HA-tagged MLL-AF10 occupancy**, H3K79me2, **wild-type MLL occupancy** (anti-MLL C-terminus, an epitope the fusion does not contain), and H3K4me3 [E1]. **The species question is answered, and the answer is both.** The two MLL species were separated by antibody epitope — HA for the fusion, C-terminal MLL for the wild-type protein — and both lost occupancy. **This graph must not inherit the LEDGF shape here.** LEDGF is required for retention of wild-type MLL but *not* MLL fusions; NF-kB is required for both. **Retention means occupancy, not stability.** The marked reduction of promoter-associated MLL-AF10 occurred **despite stable cellular MLL-AF10 protein levels** in treated cells [E1]. So an IKK inhibitor is a chromatin-eviction agent, not a degrader. **And it is not a tether.** Physical association of RELA or p50 with MLL fusion proteins was **not detected** by IP-western in AML cells — reported as data not shown — so the authors explicitly conclude that NF-kB is not a component of the fusion complex and does not directly tether it [E1]. The proposed model is that NF-kB initiates the promoter as a prerequisite for MLL recruitment or retention. **Independent support for the wild-type half** comes from a non-leukaemic system: MLL1 translocation to chromatin at TNFa- and LPS-stimulated NF-kB target genes is NF-kB-dependent, while NF-kB activation itself is unaffected by MLL1 loss [E2].
Exactly what the ChIP panel contained, and what it did not
Antibodies: HA (ab9110) for the tagged fusion, Rela (ab7970), Mll C-term (Millipore 05-765) for wild-type MLL, H3K4me3 (ab8580), H3K79me2 (ab3594), IgG control (ab46540). Loci assayed by qPCR: Hoxa9, Meis1, Ey-globin, H4 1.
What this does not contain: no genome-wide ChIP-seq, no locus outside two MLL target promoters, no time course, no human cell, no lymphoid cell, no KMT2A::AFF1. A two-locus qPCR ChIP is a real measurement and a narrow one. Anything written here as 'MLL is evicted from chromatin' should read 'MLL-AF10 and wild-type Mll occupancy at Hoxa9 and Meis1 fell'.
Occupancy loss, not protein loss — and why that distinction is the whole point
Figure S8 of the paper shows stable cellular MLL-AF10 protein in inhibitor-treated cells while promoter-associated MLL-AF10 drops 1. If the fusion had simply been degraded, the result would say nothing about chromatin and would collide with the proteasome literature, where accumulating MLL-AF4 in pro-B leukaemias is itself tumour-suppressive (mechanism-proteasome-inhibition-splits-by-lineage-in-mll-leukaemia). The measured claim is narrower and more interesting: the protein is there and it is not on the gene.
The mark loss follows the occupancy loss, and the direction is assumed rather than shown
H3K79me2 falls with MLL-AF10 occupancy (the fusion recruits DOT1L) and H3K4me3 falls with wild-type MLL occupancy (MLL is the H3K4 methyltransferase) 1. The paper reads the mark loss as consequent on the occupancy loss. Nothing in the experiment excludes the reverse or a parallel route — IKK inhibition could alter the marks and the marks could release the proteins. A time-resolved experiment would separate them and none was done.
How this joins the rest of the graph
This is the layer above everything this graph holds about elongation and chromatin. The fusion's own retention machinery — menin/LEDGF at the N-terminus, DOT1L and the SEC at the C-terminal partner — is downstream of an occupancy that an unrelated kinase cascade controls. Two consequences follow. First, an IKK inhibitor is a menin-independent route to the same endpoint a menin inhibitor is aiming at, which is hyp-ikk-inhibition-is-a-menin-independent-chromatin-eviction-route. Second, because menin independently represses RELA (mechanism-menin-represses-nfkb-transactivation), a menin inhibitor may push this axis in the wrong direction, which is hyp-menin-inhibition-liberates-rela-and-reinforces-the-fusion-programme.
| mechanism class | signalling-to-chromatin |
|---|---|
| readout | ChIP-qPCR occupancy, with total fusion protein by western as the stability control |
| perturbation | IKK inhibitors III/IV/VII (EMD 401480/401481/401486) |
| system | mouse MLL-AF10 AML cells |
| is tether | false |
| tether evidence | RELA/p50 co-IP with MLL fusion proteins not detected in AML cells (reported as data not shown) |
species measured
[ "MLL-AF10 fusion (HA epitope)", "wild-type MLL (C-terminal epitope)" ]
loci measured
[ "Hoxa9 promoter", "Meis1 promoter" ]
control loci
[ "Ey-globin promoter", "H4 promoter" ]
marks measured
[ "H3K79me2", "H3K4me3" ]
- A chromatin-eviction agent that works through an immune kinase cascade evicts the fusion in every cell that signals through that cascade, which is all of them. The selectivity claimed rests on growth assays in normal progenitors, not on occupancy measurements in normal cells.
- If eviction is transient and signalling-dependent, the fusion returns the moment drug levels fall. Continuous inhibition of IKK is the toxic regimen, and intermittent inhibition may achieve nothing durable.
- Wild-type MLL is evicted too. Losing H3K4me3 at MLL targets in normal haematopoietic cells is the mechanism by which this could damage normal blood production rather than only leukaemia.
- Cells with constitutively active downstream NF-kB, or with HOXA9/MEIS1 driven independently of the fusion, are predicted to escape - and the paper shows exactly that, since Hoxa9/Meis1-transformed progenitors are relatively resistant to IKK inhibition.
- Does IKK inhibition evict KMT2A::AFF1 from chromatin in a B-lymphoid cell? Nothing in the literature answers this, and the LEDGF precedent in this graph shows that fusion and wild-type MLL can behave oppositely on exactly this question.
- Is the eviction genome-wide or confined to promoters carrying kB sites? Two loci cannot distinguish 'NF-kB licenses MLL binding at its own targets' from 'IKK inhibition destabilises MLL chromatin binding generally'.
- Does the mark loss precede, follow, or accompany the occupancy loss? No time course exists.
- If NF-kB does not tether the fusion, what does the initiation event actually change - nucleosome occupancy, a DNA modification, or a cofactor's availability? The paper offers a chromatin-context model and a BRD4-sharing model and tests neither.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In mouse MLL-AF10 AML cells, ChIP-qPCR at the Hoxa9 and Meis1 promoters showed Rela occupancy peaking near consensus NF-kB sites with a profile similar to HA-tagged MLL-AF10; IKK inhibitor treatment reduced Rela occupancy, substantially reduced promoter occupancy of MLL-AF10 despite stable cellular MLL-AF10 protein levels, reduced H3K79 dimethylation, and also substantially reduced promoter occupancy of wild-type MLL (anti-MLL C-terminus) and H3K4 trimethylation at the same genes. Physical association of RELA or p50 with MLL fusion proteins was not detected by immunoprecipitation-western in AML cells (reported as data not shown), leading the authors to conclude that NF-kB is not a component of the MLL fusion complex and does not directly tether the oncoprotein to chromatin.
Epigenetic roles of MLL oncoproteins are dependent on NF-κB. · Kuo HP, Wang Z, Lee DF, Iwasaki M, Duque-Afonso J, Wong SHK, Lin CH, Figueroa ME, Su J, Lemischka IR, Cleary ML · Cancer Cell 2013What this evidence cannot show
One fusion (MLL-AF10), one species of cell (mouse AML), two loci (Hoxa9, Meis1), qPCR ChIP rather than ChIP-seq, pharmacological IKK inhibition rather than genetic. The co-IP negative that makes the mechanism indirect is reported as 'data not shown', so the key structural claim of the model is the least documented thing in the paper. Full text read this session via efetch db=pmc (PMC3816582), 61,617 characters.
Evidence that argues against this (1)
- The activity is not simply proportional to canonical NF-kB output. Stabilising NIK represses canonical signalling and suppresses MLL-AF9 AML, but overexpressing RelA fails to rescue that suppression - so canonical output and leukaemia maintenance are not a single dial, and 'less RELA equals less fusion on chromatin equals less leukaemia' is too simple a chain. Xiu Y et al. 2018
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
MLL1 selectively regulates activation of NF-kB target genes stimulated by TNFa and LPS: MLL1 deficiency reduces expression of a group of NF-kB downstream genes while activation of NF-kB itself is unaffected, the MLL1 complex associates with NF-kB, and ChIP shows that translocation of MLL1 to chromatin is dependent on NF-kB. The authors conclude MLL1 is recruited to its target genes by activated NF-kB.
MLL1, a H3K4 methyltransferase, regulates the TNFα-stimulated activation of genes downstream of NF-κB. · Wang X, Zhu K, ... Wu M · Journal of Cell Science 2012What this evidence cannot show
Abstract read; full text not retrieved. This is wild-type MLL1 at inflammatory target genes in a non-leukaemic system, not a fusion and not a leukaemia cell. It is cited here only for the direction of the wild-type half of the mechanism, which is the half it actually tests. It is the paper Kuo et al. cite as 'Wang et al, 2012' for the same point.
Evidence that argues against this (1)
- This is WILD-TYPE MLL1, and the finding is that MLL1 is required for NF-kB TARGET gene expression while NF-kB activation itself is unaffected - the arrow runs from MLL1 to the target genes. Using it to support NF-kB retaining MLL on chromatin runs the inference backwards from what was measured. Kuo HP et al. 2013
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Bounded absence. No published experiment measures MLL-fusion chromatin occupancy under IKK or NF-kB inhibition in a lymphoid or KMT2A::AFF1 system. A PubMed search for (KMT2A OR MLL) AND "infant"[tiab] AND ("NF-kappaB"[tiab] OR "NF-kB"[tiab] OR NFKB[tiab]) returns 0 records.
PubMed E-utilities esearch, db=pubmed, run 2026-09-20 · NCBI · PubMed 2026What this evidence cannot show
querytranslation was read and contained no [UID] artefact; the zero is a real zero for that symbol set. 'Infant' in title/abstract is a narrow proxy for the disease - a paper on KMT2A::AFF1 B-ALL that does not use the word would be missed - so this bounds the absence of an explicitly infant-framed NF-kB paper, not of all lymphoid work.
Evidence that argues against this (1)
- A zero at the symbol level is not a zero at the pathway level, and on this beat that is demonstrable rather than hypothetical: PMID 29320732 studies the non-canonical arm directly in MLL-AF9 leukaemia and is unreachable from any of these subunit symbol queries, because it names the kinase and the pathway rather than the subunit. The count is honest about what was searched and should not be read as evidence that the biology is unstudied - only that it is not indexed under this symbol. Xiu Y et al. 2018
the beat that found the mechanism on IKK subunits
IKK inhibitors III/IV/VII are the perturbation that produced the eviction result
Ikkb shRNA impairs MLL-transformed growth; the occupancy arm is complex-level, not subunit-level
- regulates →KMT2A / MLL1 — the gene that breaks
IKK inhibition strips wild-type MLL and H3K4me3 from Hoxa9/Meis1
- regulates →KMT2A::MLLT10 (MLL-AF10), t(10;11)(p12;q23)
the fusion actually measured: HA-tagged MLL-AF10 occupancy falls with stable total protein
one of the two promoters at which occupancy and marks were measured
the other measured promoter; RELA occupies near consensus kB sites
RELA occupancy at Hoxa9/Meis1 falls with IKK inhibition alongside MLL occupancy
H3K79me2 loss follows fusion eviction, consistent with loss of recruited DOT1L
- analogous to →PSIP1 / LEDGF p75 — the tether with the best therapeutic index in the whole interactome
contrast: LEDGF retains wild-type MLL but not the fusion; IKK/NF-kB retains both
bounds the retention result to myeloid systems; untested in the index lymphoid disease
the hypothesis is built on the eviction result and on its co-IP negative