PPIE (CyP33) — the off-switch the fusion throws away, and the reason PHD3 reinsertion kills transformation

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.

• An agent that inhibited PPIE would remove a brake on the wild-type KMT2A allele, which is the opposite of the desired direction in KMT2A-rearranged disease.
• PPIE is depleted in 59% of cell lines and is a spliceosome-associated cyclophilin; a broad-spectrum cyclophilin ligand would have wide off-tissue consequences for no fusion-specific gain.

1. Does supplying CyP33 in trans do anything to a PHD3-less fusion? Every reinsertion experiment restores the domain in cis, which is not a druggable manoeuvre; the trans experiment appears not to have been done.
2. Is the PHD3 requirement for abrogation specifically CyP33-dependent? A PHD3 reinsertion carrying point mutations that kill CyP33 binding but preserve H3K4me3 reading would separate the two functions.
3. Does the RNA competition described in vitro operate in cells, and if so does the transcriptome of a KMT2A-rearranged blast contain enough AU-rich RNA to keep CyP33 off the wild-type allele's PHD3?
4. PHD3 is intact on the retained wild-type KMT2A allele in every KMT2A-rearranged patient. Is that allele CyP33-repressed at the loci the fusion activates, and does that matter?
5. The PHD3 re-insertion result exists for MLL-ENL only. Does inserting PHD3 into an MLL-AF4/KMT2A::AFF1 construct also abolish immortalisation, or does the AFF1 transactivation domain override the CyP33-HDAC brake? This is the one experiment that would tell the graph whether the CyP33 mechanism is partner-general.
6. PPIE is unsearchable by its own symbol because 'PPIE' is the standard abbreviation for Patient and Public Involvement and Engagement, with 342 records of which one intersects a leukaemia term. How many other entries in this graph rest on a symbol whose PubMed literature is dominated by an unrelated acronym, and has anyone measured that systematically rather than one gene at a time?

1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   MLL PHD3 binds H3K4me3 and the CYP33 RRM domain at distinct, mutually inhibitory sites, making PHD3 a switch between activation and repression; CYP33 downregulates HOXC8, HOXA9, CDKN1B and MYC in a proline-isomerase-dependent manner with reduced H3K4me3 and H3 acetylation; PHD3 is lost in all MLL translocation products and reinsertion of PHD3 into MLL fusion proteins abrogates their transforming activity.

   The PHD3 domain of MLL acts as a CYP33-regulated switch between MLL-mediated activation and repression. · Park S, Osmers U, Raman G, Schwantes RH, Diaz MO, Bushweller JH · Biochemistry 2010
   PMID 20677832DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. Structural and binding work is in vitro on isolated domains; the gene-expression effects are in cell lines. The statement about PHD3 reinsertion abrogating transformation is reported by these authors as established background rather than generated in this paper, so the primary transformation experiments are one citation removed.

   ▸Evidence that argues against this (2)

   • The 'mutually inhibitory switch' is directly contradicted by a contemporaneous structural paper from a different group. Wang et al. report that 'H3K4me3/2 and CyP33-RRM target different surfaces of MLL1-PHD3 and can bind simultaneously to form a ternary complex'. Park 2010 and Hom 2010 read the two sites as competitive; Wang 2010 reads them as compatible. This entry states the competitive model as fact in its summary_expert, its section body and its data.contact_on_kmt2a block, and should be qualified: the geometry of the PHD3 surfaces is agreed, the exclusivity is not. Read: abstract in full, not full text. Wang Z et al. 2010
   • The entry frames PHD3 loss as removal of a brake - 'a lost contact with a demonstrated phenotype attached to its restoration'. Milne et al. give the opposite reading of the same domain: the PHD fingers are part of an ESSENTIAL MLL1 recruitment domain, together with the CXXC domain, controlled by direct PAF1C interaction and H3K4Me2/3 binding, and MLL1 fusion proteins, lacking the PHD fingers, 'require prebinding of a wild-type MLL1 complex and CXXC domain recognition of DNA for stable HoxA9 association'. On that account reinserting PHD3 into a fusion may abrogate transformation because it restores an activating recruitment module out of its normal regulatory context, not only because it restores a CyP33 docking site - which is precisely the separation-of-function this entry lists as an open question, and it makes the CyP33-specific interpretation weaker than stated. Read: abstract in full, not full text. Milne TA et al. 2010
2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   The Cyp33 RRM domain binds MLL PHD3 with ~2 uM affinity by isothermal titration calorimetry; the catalytic cyclophilin module does not; mutations in the PHD3-binding site or deletions in the RRM beta2-beta3 loop reduce or abolish binding; the RNA-binding pocket partially overlaps the PHD3 site; and full-length Cyp33 reduces expression of MLL targets MEIS1 and HOXA9.

   Molecular mechanism of MLL PHD3 and RNA recognition by the Cyp33 RRM domain. · Hom RA, Chang PY, Roy S, Musselman CA, Glass KC, Selezneva AI, Gozani O, Ismagilov RF, Cleary ML, Kutateladze TG · Journal of Molecular Biology 2010
   PMID 20460131DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. 1.9 A crystal structure is of the Cyp33 RRM alone, not of the RRM-PHD3 complex, so the interface is mapped by NMR perturbation and mutagenesis rather than solved. Independent of E1's group, which matters because the two agree.

   ▸Evidence that argues against this (1)

   • Hom et al.'s finding that 'the catalytic cyclophilin module does not bind' PHD3 is reported in this entry as evidence that the RRM is the whole story. Wang et al. show the PPIase domain has a distinct, necessary role: it catalyses proline isomerisation within the PHD3-Bromo linker, disrupting the PHD3-Bromo interface and thereby FACILITATING the RRM-PHD3 interaction, with the MLL1-CyP33 interaction required for repression of HOXA9 and HOXC8 in vivo. So CyP33 engages KMT2A bivalently - a binding surface (RRM to PHD3) and a catalytic action on an adjacent linker - and a construct carrying only PHD3, such as the reinsertion experiments this entry relies on, need not reconstitute the regulation. This is also the clearest multivalent-contact case in the beat, though both engagement points lie past the breakpoint, so it does not rescue CyP33 into the fusion interactome. Read: abstract in full, not full text. Wang Z et al. 2010
3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   AU-rich RNA binds Cyp33 through an extended interface that overlaps, but is not identical to, the MLL1-PHD3 binding surface, with aptamer affinities comparable to the PHD3 interaction, so RNA can competitively disrupt the gene-repressive Cyp33-MLL1 complex in vitro.

   Cyp33 binds AU-rich RNA motifs via an extended interface that competitively disrupts the gene repressive Cyp33-MLL1 interaction in vitro. · Lloyd NR, Wuttke DS · PLoS ONE 2021
   PMID 33606679DOI
   ▸What this evidence cannot show

   Abstract read in full; not full text. Entirely in vitro (SELEX, NMR, binding assays); the title itself says 'in vitro'. No demonstration that RNA competition operates on chromatin in cells.

   ▸Evidence that argues against this (1)

   • SELEX-derived aptamers are selected in vitro to bind tightly and are not ligands the cell contains. An overlapping interface shows the surface is RNA-accessible, not that an endogenous RNA regulates it, and AU-rich elements are abundant enough that overlap is weak evidence of a dedicated mechanism. Cell Reports 2014
4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

   PPIE has DepMap mean gene effect -0.5516 across 1,208 lines with 59.4% of lines depleted, and a KMT2A-rearranged minus other-leukemia difference of -0.027 (11 vs 43 lines), against -0.866 for MEN1 in the same lines.

   DepMap CRISPR (Chronos) gene effect via the Open Targets Platform GraphQL API, with KMT2A-rearrangement status from Cellosaurus; per-gene recomputation for this beat · Broad Institute DepMap; Cellosaurus; Open Targets Platform; analysis run by this project · Database and derived analysis 2026
   ▸What this evidence cannot show

   DepMap release version not stated by the API. Eleven KMT2A-rearranged lines is small. A gene whose relevant role is a brake would not be expected to score as a dependency in a knockout screen at all, so this null is uninformative about the biology and informative only about targetability.

   ▸Evidence that argues against this (1)

   • The entry reads PPIE's flat DepMap profile (difference -0.027) as showing it 'behaves like a housekeeping spliceosome-associated cyclophilin', i.e. that the null is informative about targetability. A knockout-screen null is weaker than that. Valerio et al. show MOF/KAT8 - also a chromatin enzyme with no KMT2A selectivity in this graph's own scan - is nonetheless required for MLL-AF9 leukemia maintenance in vivo and pharmacologically tractable with MG149 in human KMT2A-rearranged lines. Read from the full text: the requirement runs through global H4K16ac and DNA damage rather than through Hoxa9/Meis1, and non-KMT2A lines are sensitive too. The general lesson for this entry is that a flat CRISPR difference constrains selectivity, not requirement. Read: FULL TEXT via PMC5501293. Valerio DG et al. 2017
5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

   CyP33/PPIE has no intrinsic subnuclear address: expressed alone in HeLa cells it is uniformly distributed through the nucleus and excluded from nucleoli in 100% of cells, and it acquires a nuclear-speckle distribution only when co-expressed with an MLL fragment containing the AT hooks, the nuclear targeting sequences and PHD fingers 1-3, relocating onto MLL speckles in about 30% of double-positive cells.

   Protein interactions of the MLL PHD fingers modulate MLL target gene regulation in human cells. · Fair K, Anderson M, Bulanova E, Mi H, Tropschug M, Diaz MO · Mol Cell Biol 2001
   PMID 11313484DOI
   ▸What this evidence cannot show

   Transient overexpression of two epitope-tagged constructs in HeLa, one of them a 694-residue fragment of a 3,969-residue protein. No endogenous staining, no haematopoietic cell, no fusion. The 30% figure is a fraction of double-positive cells and the paper gives no quantification of intensity. Full text was read through the PMC HTML page because the Europe PMC fullTextXML endpoint returned HTTP 500 for PMC100280.

   ▸Evidence that argues against this (1)

   • Even the interaction that produces the speckle recruitment is competed away by RNA: Cyp33 binds AU-rich RNA motifs through an extended interface that competitively disrupts the Cyp33-MLL1 interaction in vitro. In a cell containing RNA - which is every cell - the fraction of CyP33 available to be recruited onto MLL speckles is not the fraction measured in a transfection, and the direction of the error is toward less recruitment. Lloyd NR et al. 2021
6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   In murine haematopoietic stem/progenitor cells, re-inserting KMT2A PHD finger 3 into MLL-ENL restores recruitment of Cyp33 and subsequently HDAC1 to the fusion protein, produces Cyp33-dependent downregulation of HOXC8, and BLOCKS the haematopoietic stem cell immortalisation potential of the fusion in serial replating colony assays; inserting PHD fingers 1 and 2 alone has no such effect. Loss of the PHD fingers is therefore necessary for MLL-ENL-driven immortalisation.

   Loss of MLL PHD finger 3 is necessary for MLL-ENL-induced hematopoietic stem cell immortalization. · Chen J, Santillan DA, Koonce M, Wei W, Luo R, Thirman MJ, Zeleznik-Le NJ, Diaz MO · Cancer Res 2008
   PMID 18676843DOI
   ▸What this evidence cannot show

   Murine haematopoietic progenitors with retrovirally expressed MLL-ENL constructs, i.e. KMT2A::MLLT1 and not KMT2A::AFF1, and a serial replating assay rather than a transplantation or patient readout. The experiment is a gain-of-domain construct, so it shows what a PHD3-containing fusion cannot do, not what endogenous PPIE does in a normal cell. Read: abstract only.

   ▸Evidence that argues against this (1)

   • This is a partner-specific result carrying a general-sounding conclusion, which is the failure mode the cycle-36 title-drift audit was written about. It was obtained in MLL-ENL; MLL-AF4/KMT2A::AFF1 has a different C-terminal partner with a different transactivation architecture, and no equivalent PHD3 re-insertion experiment has been published in that fusion. The graph's structural reading - that the PHD cluster is deleted in every fusion, so the CyP33 contact cannot occur on any of them - is what generalises; the immortalisation-blocking result is not shown to. Oncofusion knowledge graph, analyses/depmap-kmt2a-selective-scan-full; DepMap Chronos via Open Targets; 11 KMT2A-rearranged versus 43 other leukaemia lines, 17,728 genes 2026

• ← related toKMT2A exon 11 / intron 11 breakpoints (BCR subregion C): the infant, HOXA-low, topoisomerase-II-associated break

  The CYP33/PPIE binding site on KMT2A is PHD3 (exons 14-16), not PHD1 (exon 11). This matters for the breakpoint atlas because the reciprocal-PHD1 prognostic mechanism assumes otherwise.

• related to →KMT2A / MLL1 — the gene that breaks

  corrective: H3K4me3/2 and the CyP33 RRM bind different PHD3 surfaces and can occupy them simultaneously in a ternary complex (Wang 2010), so the switch is not simply mutually exclusive

• related to →PAF1 — pan-essential subunit of a complex the fusion needs, where the window is in the interface and not in the protein

  PHD3 sits inside the same essential MLL1 recruitment module that PAF1C controls, so PHD3 loss removes an activating contact as well as a repressive one

• binds →KMT2A / MLL1 — the gene that breaks

  CyP33 RRM binds KMT2A PHD3 at ~2 uM (ITC), competitively with H3K4me3

• related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)

  NOT an interactor: PHD3 is deleted from all KMT2A translocation products, and reinserting PHD3 abrogates transformation

• regulates →HOXA9 — required for survival of established MLL-rearranged leukemia, and dispensable for starting it

  CyP33 represses HOXA9 (and HOXC8, CDKN1B, MYC) via PHD3, isomerase-dependent

• regulates →MEIS1 — rate-limiting for MLL leukemia stem cells, a surviving co-dependency, and embryonic-lethal when removed

  full-length Cyp33 lowers MEIS1 expression

• related to →HCFC1 (HCF-1) — a founding member of the wild-type KMT2A complex that the fusion cannot bind

  both are wild-type-only KMT2A contacts on the discarded C-terminal fragment

• ← related toCREBBP (CBP) — binds the KMT2A transactivation domain, which no fusion keeps

  two mapped wild-type-only contacts on the discarded KMT2A fragment, one activating and one repressive

• ← bindsPHD3 and the CyP33 RRM: Kd 7-15 uM, and a 2023 ternary structure that settles the 2010 disagreement - H3K4me3 and the RRM CAN be held at once, and the real switch is RNA

  CyP33 RRM canonical RNA-binding surface

• ← supportsAn H3K4me3-reading PHD finger fused to NUP98 makes a potent leukaemia oncoprotein; the H3K4me3-reading PHD finger of KMT2A is deleted from every KMT2A fusion and reinstating it abolishes transformation. Same module class, opposite sign - and the one NUP98 fusion that carries KMT2A's PHD cassette is the weakest member of its family.

  CyP33 docking on PHD3 is the proposed determinant of the opposite sign

• related to →Nuclear speckles — the compartment everyone assumes KMT2A::AFF1 uses, on evidence that skips AFF1 itself

  CyP33 enters speckles only by recruitment onto wild-type KMT2A PHD1-3, a surface every fusion deletes

• supports →The clustered PHD fingers (KMT2A aa 1416-1640) are a chromatin-targeting module and the canonical t(4;11) junction removes them by ten residues

  the PHD cluster carries CyP33's only route to a subnuclear compartment and is lost at the breakpoint

• related to →Bone marrow haematopoietic stem cell — the postnatal comparator, and why the same fusion behaves differently here

  re-inserting the CyP33-binding KMT2A PHD3 into MLL-ENL BLOCKS haematopoietic stem cell immortalisation; the cell type in which PPIE's absence is necessary for transformation, not one in which PPIE is required

• inhibits →The stem-like self-renewal state — a limited stem-cell module grafted onto a committed cell

  the CyP33-PHD3 module suppresses fusion-driven self-renewal; every KMT2A fusion deletes the PHD cluster, so the brake is absent in KMT2A::AFF1 by construction

• ← related toLost from the KMT2A half: PHD3 and the bromodomain are one cassette, not two modules (aa 1566-1765)

  CyP33 isomerises the PHD3-bromo linker to open the cassette

• ← related toAudit, CORRECTED: the PHD add-back result was replicated in 2008 by a second group in a different fusion, and this graph already cited it

  the CyP33-dependent arm of the same single study

• ← regulatesHypothesis: the fusion is exempt from KMT2A's regulation, not free of it - the brakes still operate at the same loci, on the other allele

  CyP33 induction is the hypothesis's central perturbation

• ← testsExperiment: induce CyP33 in a KMT2A-rearranged cell and ask, allele by allele, whether the brake still works at HOXA9

  inducible CyP33 with PPIase-dead and RRM-mutant controls