IKBKB (IKK-beta) — the kinase whose knockdown tracks MLL-leukaemia growth most tightly, and whose inhibitors have failed clinically for a reason
also called IKBKB · IKK-beta · IKKB · IKK2 · I-kappa-B kinase beta · O14920
IKK-beta is the kinase that starts the classical inflammation switch by tagging the brake protein for destruction. In MLL-rearranged leukaemia cells it is needed for growth, and blocking the kinase complex knocks the leukaemia fusion protein off the genes it controls — but drugs against this kinase have repeatedly failed in people because healthy immune cells need it too.
The same thing, for a scientist
IKBKB (UniProt O14920, 756 aa, primary gene name read back) is the catalytic subunit of the canonical IKK complex: it phosphorylates IkB-alpha, triggering its degradation and liberating RELA:p50 for nuclear translocation. It carries the most direct leukaemia evidence of the three IKK subunits. In the functional genomic screen that opened this field, IKKa, IKKb and IKKg were all among 33 phosphoregulators whose shRNA knockdown impaired growth of MLL-oncogene-transformed myeloid progenitors by 50% or more; on follow-up, **growth and clonogenicity suppression correlated with Ikkb knockdown level, with Ikka and Ikkg contributing 'to a lesser extent'** [E1]. The effect was selective: MLL-transduced and MLL leukaemia cells were affected, E2A-PBX1-transduced cells and normal haematopoietic progenitors were not [E1]. **The therapeutic history is the counterweight.** A 2026 paper states that aberrant IKKb-NF-kB activation occurs in roughly 40% of AML and that pharmacological inhibition of IKKb **'has been limited by serious toxicities, including neutrophilia'**, and builds a celastrol-based cereblon PROTAC (compound A9) that degrades IKKb and NR4A1 instead, suppressing leukaemia in a **KMT2A::MLLT3** mouse model *without* neutrophilia [E2]. That is the first KMT2A-specific in vivo demonstration that this node can be hit without the class toxicity — and it is a degrader of two proteins, not an inhibitor of one. **What is not established:** no clinical IKKb inhibitor exists, no study has combined one with a menin inhibitor, and every experiment here is myeloid.
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.
What the screen actually measured
A lentiviral shRNA library of 321 hairpins against 211 phosphoregulators was applied gene-by-gene to myeloid progenitors transformed by MLL oncogenes. Thirty-three candidates gave 50% or greater growth impairment; IKKa, IKKb and IKKg were among them, alongside PLK1, PPP4C and IRAK3 1. This is a growth screen in mouse myeloid progenitors — not a chromatin screen, and not a lymphoid one. The chromatin result came later and from pharmacology, not from these hairpins.
The chromatin result is complex-level, not IKKb-level
The occupancy experiments that make this beat matter used IKK inhibitors III, IV and VII (EMD 401480, 401481, 401486) — small molecules directed at the IKK complex, not at IKKb alone 1. So 'IKKb is required for MLL fusion retention' is naming and not measurement: the measurement is 'IKK activity is required'. The genetic arm of the paper reached only growth and clonogenicity, never occupancy. This is the same shape Rule 63 logs for CK2, where pan-catalytic-subunit chemistry made a subunit attribution look like a result. See protein-ikk-complex and mechanism-ikk-nfkb-retains-mll-on-chromatin.
Why the toxicity is not a generic caveat here
The published clinical failure mode for this target class is neutrophilia and systemic inflammation 2. The index disease of this graph is infant B-ALL, in patients whose adaptive immunity is still being built and who are already profoundly myelosuppressed by induction chemotherapy. A drug whose on-target class effect is dysregulated granulopoiesis is a poor fit for that population; risk-ikk-inhibition-essential-immune-function holds the argument.
| uniprot | O14920 |
|---|---|
| length AA | 756 |
| alphafold ID | AF-O14920-F1 |
| resolution note | gene_exact:IKBKB AND organism_id:9606 AND reviewed:true returned 1 reviewed row; gene_primary read back as IKBKB; length 756 aa asserted so verify-accessions.py can catch a swap. |
| pathway arm | canonical |
essentiality
{}depmap kmt2a selectivity
{
"chronos_mean_kmt2a_rearranged": -0.0082,
"chronos_mean_other_leukaemia": 0.0243,
"difference": -0.0326,
"pan_essentiality_all_lines": -0.076,
"selectivity_rank": 5068,
"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."
}- IKKb is required by essentially every immune cell. The published clinical failure mode for this target class is neutrophilia and systemic inflammation, and the index patients are infants building an immune system under myelosuppressive chemotherapy.
- Blocking the whole IKK complex also blocks IKKa, the kinase of the non-canonical arm — and the non-canonical arm SUPPRESSES MLL-AF9 leukaemia. Net sign is not predictable from either paper alone.
- The only KMT2A-specific in vivo demonstration is a dual degrader built on celastrol, a promiscuous electrophile. Attributing its effect to IKKb is not supported by the reported experiments.
- NF-kB is the principal stress-survival pathway of normal haematopoietic stem cells. The therapeutic window claimed here rests on constitutive-activation observations in primitive AML cells, not on a measured window in KMT2A-rearranged disease.
- Suppressing NF-kB removes a brake on infection in a population whose commonest cause of treatment-related death is infection.
- Is IKKb required for MLL-fusion chromatin occupancy, or only for growth? No paper has put an Ikkb knockdown and an MLL-fusion ChIP in the same figure; the occupancy result is entirely pharmacological and complex-level.
- Does an IKKb-selective degrader evict the fusion from chromatin the way the pan-IKK inhibitors did, or does degrading one subunit leave the complex partly functional?
- Is there any IKKb dependency in KMT2A-rearranged B-ALL? Every experiment in this entry is myeloid.
- Do IKK inhibition and menin inhibition cause the same occupancy loss or additive loss? Nobody has run the pair.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
A 321-shRNA screen over 211 phosphoregulators in MLL-oncogene-transformed myeloid progenitors identified 33 candidates causing 50% or greater growth impairment, among them IKKa, IKKb and IKKg; genetic reduction of Ikka, Ikkb or Ikkg decreased growth and clonogenicity in MLL-oncogene-transduced and MLL AML cells but not in E2A-PBX1-transduced cells or normal haematopoietic progenitors, with suppression correlating with Ikkb knockdown level and Ikka and Ikkg contributing to a lesser extent. The pharmacological arm used IKK inhibitors III, IV and VII, which act on the complex rather than on a single subunit.
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
Mouse myeloid progenitors and mouse/human AML lines throughout. No lymphoid MLL model and no KMT2A::AFF1 B-ALL experiment. The subunit-level claim rests on shRNA growth assays; the occupancy claim rests on complex-level chemistry. Read in full this session via efetch db=pmc (PMC3816582), 61,617 characters after tag-stripping.
Evidence that argues against this (1)
- The same pathway read the other way: stabilising NIK activates the non-canonical arm, represses the canonical arm, and SUPPRESSES MLL-AF9 AML. A pan-IKK inhibitor also inhibits IKKa, the non-canonical kinase, so it removes a protective signal at the same time as a permissive one. Xiu Y et al. 2018
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Aberrant activation of the IKKb-NF-kB pathway occurs in approximately 40% of AML cases; pharmacological inhibition of IKKb has been limited by serious toxicities including neutrophilia. A celastrol-based PROTAC, compound A9, degrades IKKb and NR4A1 through cereblon, kills multiple AML lines and primary AML samples, and suppresses leukaemia progression in a KMT2A::MLLT3 AML mouse model without inducing neutrophilia.
PROTAC-mediated targeting of IKKβ and NR4A1 for AML therapy. · Maharjan CK, Liu Y, ... Zhang W · Oncogene 2026What this evidence cannot show
Abstract read; full text not retrieved. A9 is a dual degrader of IKKb AND NR4A1, so the in vivo effect cannot be attributed to IKKb degradation alone, and the authors call the two roles non-redundant. The 40% figure is an association statement in an abstract, not a number checked against a cohort here. Celastrol is a promiscuous electrophile, which is why it binds both targets; off-target proteome effects are not excluded by the abstract. The bioRxiv version (PMID 41040298, doi 10.1101/2025.09.24.678355) puts the toxicity point more bluntly: 'nearly all IKKβ inhibitors have failed prior clinical trials due to their serious side effects such as neutrophilia and systematic inflammation'.
Evidence that argues against this (1)
- NR4A1 has previously been described as a tumour suppressor in AML, and this work reverses that assignment for (pro)monocytic AML. A dual degrader whose second target has a disputed direction of effect is a weak instrument for attributing anything to IKKb. Maharjan CK et al. 2025
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt reviewed human record O14920 (IKKB_HUMAN) has primary gene name IKBKB and length 756 aa; the query gene_exact:IKBKB AND organism_id:9606 AND reviewed:true returned exactly one row and its gene_primary field was read back as IKBKB.
UniProtKB entry O14920 (IKKB_HUMAN), Inhibitor of nuclear factor kappa-B kinase subunit beta · UniProt Consortium · UniProt 2026What this evidence cannot show
Resolved 2026-09-20 with gene_primary requested and every returned row read. One row, primary match. No paralogue ambiguity for this symbol, unlike TRAF2.
Evidence that argues against this (1)
- A reviewed database record fixes the canonical sequence and primary symbol on the date it was read, and nothing more. It does not establish that the protein is expressed in KMT2A-rearranged blasts, which isoform is present, or that the subunit acts in this disease. Accession and length are asserted so that a swap fails verify-accessions.py - a defence this graph needs because IKK-family and NF-kB subunit names collide heavily in the older literature - not because the record carries biological weight. Xiu Y et al. 2018
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Counted absence. (KMT2A OR MLL OR "mixed-lineage leukemia") AND (IKBKB OR "IKK-beta"[tiab] OR "IKKbeta"[tiab] OR "IKK beta"[tiab]) returns 5 PubMed records. Title-checking leaves three on topic: the two versions of the IKKb/NR4A1 PROTAC work and the 2013 Cancer Cell paper. There is no IKKb-focused study in KMT2A-rearranged B-ALL of any kind.
PubMed E-utilities esearch, db=pubmed, run 2026-09-20 · NCBI · PubMed 2026What this evidence cannot show
querytranslation was read: PubMed expanded KMT2A to the supplementary-concept term and IKBKB to 'i kappa b kinase', so recall is wider than the literal string and IKK2 is covered by that expansion. A count is only as good as its symbol set.
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
catalytic beta subunit of the canonical IKK complex
Ikkb shRNA impairs MLL-transformed growth; the occupancy arm is complex-level, not subunit-level
KMT2A::MLLT3 mouse AML suppressed by an IKKb/NR4A1 dual degrader without neutrophilia
class failure in prior clinical trials on neutrophilia and systemic inflammation