Oncofusion

start / graph / localization-pml-nuclear-bodies-pmlrara

Mechanismestablished confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.Q4b-localization

PML nuclear bodies and PML::RARA — the best-worked case of a fusion destroying a nuclear compartment, and what it actually predicts for KMT2A

also called PML oncogenic domains · PODs · ND10 · PML NB disruption · microspeckled pattern APL

In one leukemia, APL, the fusion protein shatters a normal nuclear structure called the PML body into hundreds of tiny specks, and the drug that cures the disease puts those structures back together. It is the clearest example in cancer of a fusion working by putting things in the wrong place — but the twist is that reassembling the bodies turned out not to be the cure; destroying the fusion protein was.

The same thing, for a scientist

In 1994 two groups reported back-to-back in Cell that PML normally resides in a discrete subnuclear organelle (PML oncogenic domains / PODs, now PML nuclear bodies), that in APL blasts PML::RARA disrupts this organelle into a microparticulate pattern, and that retinoic acid triggers reorganisation into normal-appearing bodies [E1][E2]. Weis et al. added that the hybrid drags RXR and other nuclear antigens into aberrant chromatin-bound structures — a dominant-negative mislocalisation of partners — and that ATRA relocalises all of them [E2]. This is the canonical precedent for fusion-driven mislocalisation and a differentiation-therapy success. But the field's own later verdict is the part that matters for this graph: the curative mechanism of both ATRA and arsenic is now understood as targeted catabolism of PML::RARA — arsenic binding, oxidation, SUMOylation on PML nuclear bodies and RNF4-mediated ubiquitination — not restoration of nuclear-body architecture as such [E3]. Arsenic at pharmacological concentrations in fact inhibits PML-body recycling and sequesters PML and PML::RARA in cytoplasmic assemblies, the opposite of rebuilding the compartment, while still curing the disease [E4]. The precedent therefore predicts something specific and deflationary for KMT2A::AFF1: a mislocalisation phenotype is diagnostically powerful and mechanistically suggestive, but the therapeutic lever ended up being degradation of the oncoprotein.

Go deeper
What was seen, in what material, by what method1 deeper

Dyck et al. used immunohistochemistry and immunofluorescence on APL cells and on cells expressing PML-RARa, and described PML as part of a novel macromolecular organelle containing at least three other nuclear proteins. In APL cells the POD is disrupted into a microparticulate pattern as a consequence of PML-RAR expression; RA treatment triggers reorganisation to normal-appearing PODs. They proposed PML-RAR is a dominant-negative oncoprotein that inhibits POD assembly 1.

Weis et al., in the adjacent paper, localised PML to nuclear bodies recognised by patient autoimmune sera, using immunofluorescence and electron microscopy. In APL cells the PML-RARa hybrid shows abnormal localisation and directs RXR and other nuclear antigens into aberrant structures that are tightly bound to chromatin; RA induces complete relocalisation of each protein 2.

Method note that matters for this beat: both 1994 papers worked with APL-derived cells stained for endogenous protein, in part with human autoimmune sera, not with overexpressed GFP constructs. That is a higher evidence grade than anything that exists for KMT2A fusions, where every published image is of an exogenously expressed construct.

Why 'the fusion breaks the body' is the easy half of the story

The 1994 model — PML::RARA is oncogenic because it prevents POD assembly — is elegant and was enormously productive. It is also not what the cure turned out to run on. By 2011 the physiopathological consensus was that RA and As2O3 both work by triggering degradation of PML::RARA, with arsenic binding to PML, oxidation, SUMOylation on PML nuclear bodies and RNF4-mediated ubiquitination as the steps 3. Here the nuclear body is the venue for the degradation reaction rather than the thing whose repair is therapeutic.

Lang et al. sharpen this. At pharmacologically relevant arsenic concentrations (0.02-0.5 uM) ATO inhibits formation of progeny PML bodies and stabilises cytoplasmic precursor compartments (CyPNs), sequestering PML and PML::RARA in the cytoplasm after cell division; PML and PML::RARA carrying mutations found in arsenic-resistant APL patients are impeded in their ability to be sequestered in CyPNs 4. So the curative drug, at curative doses, does not rebuild the compartment. It moves the protein out of the nucleus and destroys it.

What the precedent predicts for KMT2A::AFF1 — three concrete transfers1 deeper

1. Expect mislocalisation of partners, not only of the fusion. The most reproducible PML finding was not where PML::RARA went but that it dragged RXR and other nuclear antigens with it into aberrant chromatin-bound structures 2. The KMT2A analogue is untested and cheap: does KMT2A::AFF1 pull menin, LEDGF/PSIP1, AFF4, ENL or DOT1L out of their normal distribution in a t(4;11) blast? A multiplex immunofluorescence panel on fixed patient marrow would answer it, and unlike almost everything else in this beat it can be done on archival material.

2. Do not expect compartment repair to be the therapeutic endpoint. If menin inhibition is the KMT2A field's ATRA, the PML history predicts that the readout that matters is fusion-protein clearance, not restoration of a normal speckle pattern. That is an argument for putting degrader-modality work ahead of compartment-normalisation assays.

3. Expect the mislocalisation phenotype to be diagnostically useful before it is mechanistically decisive. The microspeckled PML pattern became a rapid diagnostic test for APL long before anyone agreed what it meant. A validated antibody giving a distinctive subnuclear pattern for KMT2A fusions would have real clinical value even if the mechanism stayed unresolved.

And one transfer that fails

PML::RARA disrupts a compartment that already existed and had a function: the PML body is a bona fide organelle with a SUMO-dependent assembly pathway and roles in senescence, apoptosis and antiviral defence. Nothing in the KMT2A literature identifies an equivalent pre-existing compartment that KMT2A occupies and that the fusion destroys. Wild-type KMT2A's punctate distribution is not a named organelle with an independent functional literature. The honest statement is that the PML precedent shows what strong evidence for fusion-driven mislocalisation looks like, and by that standard the KMT2A case is currently very weak — a reason to do the experiment, not a reason to assume the answer.

Facts and figures
compartmentPML nuclear body (POD / ND10)
fusion studiedfusion-pml-rara
evidence class for that fusiondirect imaging in APL-derived cells with endogenous-protein staining, two independent groups back-to-back
evidence class for kmt2anone — no KMT2A study of comparable design exists
therapeutic lever that workedoncoprotein degradation (ATRA + As2O3), not compartment restoration
marker proteins
[
  "PML",
  "SP100",
  "SUMO1",
  "DAXX"
]
methods used
[
  "immunohistochemistry",
  "indirect immunofluorescence with patient autoimmune sera",
  "electron microscopy"
]
How this could bite us back
  • Reading the PML precedent as 'fix the compartment and you cure the disease' would misdirect effort: the compartment was not repaired by the curative drug, the oncoprotein was destroyed. A KMT2A programme built on compartment normalisation as the primary endpoint inherits a hypothesis the source field abandoned.
  • Arsenic-resistant APL arises through PML mutations that block the sequestration/degradation step, so resistance targeted the localisation-dependent mechanism itself. A KMT2A therapy that depended on relocalising the fusion should expect the same class of escape mutation at the relocalisation interface.
  • Differentiation therapy in APL carries differentiation syndrome; the same toxicity is already seen with menin inhibitors, so a successful KMT2A differentiation therapy inherits the APL safety problem alongside its precedent.
Still open
  1. Does KMT2A::AFF1 mislocalise its partners — menin, LEDGF/PSIP1, AFF4, ENL, DOT1L — the way PML::RARA mislocalises RXR? A five-colour immunofluorescence panel on archival t(4;11) marrow is the closest available analogue of the 1994 experiment and has never been run.
  2. Is there any pre-existing nuclear compartment that wild-type KMT2A occupies and that the fusion measurably disrupts? If not, the PML analogy is structurally inapplicable and should be retired from grant text rather than repeated.
  3. Would a subnuclear staining pattern for KMT2A fusions have diagnostic value even without a mechanism, as the microspeckled PML pattern did for a decade before its meaning was settled?
  4. Does menin inhibition change the subcellular partition of the fusion the way arsenic moves PML::RARA to the cytoplasm, or does it leave the protein nuclear and merely unbound?
Every source behind this entry (4)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    PML is part of a novel macromolecular nuclear organelle (PML oncogenic domain, POD) containing at least three other nuclear proteins; in APL cells the POD is disrupted into a microparticulate pattern as a consequence of PML-RAR expression, and retinoic acid treatment triggers reorganisation of PML into normal-appearing PODs.

    A novel macromolecular structure is a target of the promyelocyte-retinoic acid receptor oncoprotein · Dyck JA, Maul GG, Miller WH, Chen JD, Kakizuka A, Evans RM · Cell 1994
    PMID 8293467DOI
    What this evidence cannot show

    1994 immunohistochemistry and immunofluorescence at the resolution then available; 'microparticulate' is a qualitative description without particle counting or super-resolution. The dominant-negative POD-assembly model proposed here is not the mechanism that ultimately explained the cure.

    Evidence that argues against this (1)
    • The field's later consensus is that RA and arsenic cure APL by triggering degradation of PML::RARA, with nuclear bodies serving as the site of the SUMO/RNF4 reaction rather than the therapeutic endpoint. Lallemand-Breitenbach V et al. 2011
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    PML localises to a discrete subnuclear compartment corresponding to nuclear bodies recognised by patient autoimmune sera; in APL cells the PML-RAR alpha hybrid shows abnormal localisation and directs RXR and other nuclear antigens into aberrant structures tightly bound to chromatin, and retinoic acid induces complete relocalisation of each of these proteins.

    Retinoic acid regulates aberrant nuclear localization of PML-RAR alpha in acute promyelocytic leukemia cells · Weis K, Rambaud S, Lavau C, Jansen J, Carvalho T, Carmo-Fonseca M, Lamond A, Dejean A · Cell 1994
    PMID 8293468DOI
    What this evidence cannot show

    Immunofluorescence and electron microscopy on APL-derived cells; 'tightly bound to chromatin' rests on extraction conditions of the era. The claim that partner mislocalisation is causal for transformation is an inference, not a demonstrated separation of function.

    Evidence that argues against this (1)
    • The dominant-negative reading has a genetic test, and it points at apoptosis rather than at mislocalisation as the load-bearing consequence. Wang et al. showed that Pml-null mice and cells are protected from the lethal effects of ionising radiation and anti-Fas antibody, that Pml is required for Fas-, TNF-, ceramide- and type I/II interferon-induced programmed cell death and for caspase 1 and caspase 3 activation, and - the point for this item - that the PML-RAR alpha fusion renders haematopoietic progenitors resistant to Fas-, TNF- and IFN-induced apoptosis with a lack of caspase 3 activation, 'thus acting as a Pml dominant-negative product'. The functional dominant-negative readout is therefore an apoptotic one measured genetically; the microparticulate immunofluorescence pattern is a correlate of it, and the entry should not treat the redistribution of RXR and other antigens as the demonstrated mechanism. Wang ZG et al. 1998
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.review

    The curative activity of retinoic acid and arsenic trioxide in APL is attributed to targeted degradation of the PML/RARA oncoprotein, with arsenic binding, oxidation, SUMOylation on PML nuclear bodies and RNF4-mediated ubiquitination as the molecular steps.

    Curing APL through PML/RARA degradation by As2O3 · Lallemand-Breitenbach V, Zhu J, Chen Z, de The H · Trends in Molecular Medicine 2011
    PMID 22056243DOI
    What this evidence cannot show

    A review, cited as a map of the field's settled interpretation rather than as primary data, and written by the group that advanced the degradation model, so not a neutral survey. The cure rates it quotes (~70% arsenic alone, ~90% combination) come from trials summarised there, not reported there.

    Evidence that argues against this (1)
    • Degrading the fusion and restoring wild-type PML function are not separable by the evidence the degradation model rests on, and the knockout shows why that matters. Wang et al. established that Pml itself is required for Fas-, TNF-, ceramide- and interferon-induced apoptosis and for caspase 1 and caspase 3 activation, with Pml-null animals protected from lethal irradiation, and that PML-RARA confers the same apoptotic resistance on haematopoietic progenitors as a dominant negative. Catabolism of PML/RARA therefore simultaneously removes an oncoprotein and releases an apoptotic effector from dominant-negative inhibition. A review that attributes the cure to targeted degradation - written by the group advancing that model - does not distinguish those, and the distinction is the one a KMT2A programme would need before copying the strategy. Wang ZG et al. 1998
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    At pharmacologically relevant concentrations (0.02-0.5 uM) arsenic trioxide inhibits the formation of progeny PML bodies and stabilises cytoplasmic precursor compartments (CyPNs), sequestering PML and PML/RARA in the cytoplasm after cell division, and PML and PML/RARA variants carrying mutations from arsenic-resistant APL patients are impeded in CyPN sequestration.

    PMID 22692509DOI
    What this evidence cannot show

    Immunofluorescence in cultured cells including tumour cell lines; the causal link between CyPN sequestration and clinical cure is correlative, resting on the behaviour of resistance-associated mutants rather than on a designed separation-of-function allele.

    Evidence that argues against this (1)
    • Cytoplasmic sequestration is one of several mechanisms arsenic is credited with, and the genetics of PML says the functional endpoint to measure is apoptotic competence, not location. Wang et al. showed Pml is essential for multiple apoptotic pathways - Fas, TNF, ceramide, type I and II interferons - with caspase 1 and caspase 3 activation lost in its absence, Pml-null mice and cells protected from lethal irradiation and anti-Fas antibody, and PML-RARA acting as a Pml dominant-negative in haematopoietic progenitors. Sequestering PML and PML/RARA into cytoplasmic precursor compartments removes both from the nucleus, so the same manipulation that disables the oncoprotein also disables the tumour suppressor. That resistance-associated mutants are impeded in CyPN sequestration is consistent with the model and equally consistent with those mutants being generally impaired in PML body dynamics. Wang ZG et al. 1998
Connected to