Oncofusion

start / graph / localization-nuclear-envelope-proximity-kmt2a-amino-terminus

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q28-subnuclear

MLL was reported at regions attached to the nuclear envelope in 2000, which is indirect evidence at the lamina that this graph currently records as absent

also called MLL nuclear envelope attachment · MLL3AT localisation · Caslini amino-terminal targeting

A 2000 paper looked at where the front end of MLL goes inside the nucleus and described it sitting in speckles that included patches stuck to the nuclear envelope and the rim of the nucleolus. That is a small but real observation at the nuclear edge, and this knowledge graph currently says no such observation exists.

The same thing, for a scientist

Caslini and colleagues conditionally expressed a 50 kDa amino-terminal fragment of MLL spanning the AT-hook motifs (MLL3AT) in U937 and HeLa, and examined it by immunofluorescence, confocal laser scanning microscopy and **immunoelectron microscopy**. MLL3AT, like endogenous MLL, localised in the nucleoplasm in a punctate distribution 'including **regions attached to the nuclear envelope** and the periphery of the nucleolus'; both MLL3AT and endogenous MLL were present in interphase nuclear matrices and co-localised with topoisomerase II on mitotic chromosomal scaffolds; and nucleoplasmic and nucleolar localisation was observed **even for conditionally expressed MLL-AF9 and MLL-AF4 chimeric proteins**, which the authors take as a common target conferred by the MLL amino terminus [E1]. This paper is cited five times elsewhere in this graph - for AT-hook targeting, nuclear matrix, nucleolus and fusion architecture - but **not** in the lamina entry, which records the evidence grade for KMT2A at the nuclear lamina as 'none, direct or indirect'. The envelope-attachment clause is indirect evidence and the entry should be corrected. Two qualifications are essential and are the reason this is an entry rather than a one-line fix: the paper says *nuclear envelope*, not *lamina*, and envelope proximity in a fixed image is not lamina association; and the system is conditional overexpression of a fragment, so the attribution status is positive-but-overexpressed like everything else in this class.

Go deeper
What the paper supports and what it does not

Supports. That MLL-associated signal was seen at the nuclear periphery, by a method (immunoelectron microscopy) that resolves the envelope far better than light microscopy does. That is a stronger technique for this particular question than anything else in this literature.

Does not support. Any statement about lamina-associated domains, about LMNA or LMNB1 contact, or about the fusion being enriched at the periphery. 'Regions attached to the nuclear envelope' inside a punctate nucleoplasmic description is a minor component of a distribution, not a compartment assignment, and the paper makes no quantitative claim about it.

The right correction is therefore not 'KMT2A is at the lamina' but 'the evidence grade is weak-indirect rather than none', with the sentence quoted and the system named.

The part that matters more: the chimeras were included

The clause that has been under-used is that nucleoplasmic and nucleolar localisation was observed even for conditionally expressed MLL-AF9 and MLL-AF4 chimeric proteins 1. That makes this one of a very small number of papers that image a KMT2A fusion at all, and the only one that images two different fusions in the same system alongside the wild-type protein and an isolated N-terminal fragment.

Its attribution status is the good kind for this class: the chimeras are conditionally expressed, so the species producing the signal is known. Its weakness is the usual one: conditional overexpression, in U937 and HeLa, of proteins whose physiological level is low.

The design conclusion the authors draw - a common target conferred by the amino terminus of MLL to many if not all chimeric MLL proteins - is the same conclusion the 1997 MLL-AF6 mapping reached from the opposite direction 2, and it is the conclusion this beat keeps arriving at: compartment is a retained-KMT2A property, and the partner does not choose it.

Facts and figures
attributionpositive but conditionally overexpressed
correction owedlocalization-nuclear-lamina-lads-kmt2a-r records evidence grade 'none, direct or indirect'; this paper's envelope-attachment clause is indirect evidence
compartments reported
[
  "nucleoplasm (punctate)",
  "regions attached to the nuclear envelope",
  "periphery of the nucleolus",
  "interphase nuclear matrix",
  "mitotic chromosomal scaffold"
]
methods
[
  "immunofluorescence",
  "confocal laser scanning microscopy",
  "immunoelectron microscopy"
]
systems
[
  "U937",
  "HeLa"
]
constructs
[
  "MLL3AT (50 kDa N-terminal fragment)",
  "MLL-AF9",
  "MLL-AF4",
  "endogenous MLL"
]
How this could bite us back
  • If the fusion has a peripheral pool, a nuclear-export or envelope-disrupting agent would act on wild-type KMT2A identically, since the targeting module is shared.
  • Any claim that the fusion is excluded from silenced peripheral chromatin would have to survive this observation, and it is currently made without reference to it.
Still open
  1. Does the envelope-attached MLL signal co-localise with lamin B1, or is it at nuclear pores, which the immunoelectron images could in principle distinguish?
  2. Is the fraction of MLL or of a KMT2A fusion at the periphery different from the genome-average expectation, or is it simply what a chromatin-binding protein looks like at the edge of a nucleus?
  3. Do the original immunoelectron micrographs allow the envelope-attached fraction to be quantified retrospectively?
Every source behind this entry (3)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Conditionally expressed MLL3AT, a 50 kDa amino-terminal MLL fragment spanning the AT-hook motifs, localised in U937 and HeLa - by immunofluorescence, confocal microscopy and immunoelectron microscopy - in the nucleoplasm in a punctate pattern including regions attached to the nuclear envelope and the periphery of the nucleolus, was present with endogenous MLL in interphase nuclear matrices and co-localised with topoisomerase II on mitotic chromosomal scaffolds, and nucleoplasmic and nucleolar localisation was observed even for conditionally expressed MLL-AF9 and MLL-AF4 chimeric proteins.

    PMID 11069025DOI
    What this evidence cannot show

    Abstract-level for the sentence quoted; the Europe PMC core record returns no pmcid for this PMID. Conditional overexpression of a fragment and of chimeras in U937 and HeLa, neither of which carries the translocation. 'Nuclear envelope' is not 'nuclear lamina', and the envelope observation is a minor component of a punctate nucleoplasmic distribution with no quantification. **Retrieval correction by the coordinator, 2026-09-20.** This beat logged PMID 11069025 as having no PMCID and being readable at abstract level only. It has **PMCID PMC7543881**. `fullTextXML` returns HTTP 500, but the **PMC HTML page returns 63,706 characters of stripped text**. The paper is readable and was read; the full text supplies the image-level detail *"the presence of label at the edge of the nuclear envelope"* and confirms that the word *lamin* appears nowhere in it.

    Evidence that argues against this (1)
    • The same punctate nucleoplasmic pattern is seen for wild-type MLL and for the fusions in this system, so the shared pattern cannot discriminate them, and in cells that actually carry a translocation the same pattern appears with and without the rearrangement. Butler LH et al. 1997
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The region of the MLL N-terminus responsible for nuclear localisation of MLL-AF6 was mapped to a region containing the AT-hook motifs, by immunofluorescence and cell fractionation - the same amino-terminal module implicated in the 2000 targeting study.

    PMID 9349501DOI
    What this evidence cannot show

    Abstract only; no PMCID. A region assignment rather than a signal identification, and nuclear entry is not distinguished from nuclear retention.

    Evidence that argues against this (1)
    • The one direct test of MLL's AT-hooks found they read DNA structure rather than AT-richness, so the module implicated in both papers may be acting as a DNA-binding element whose apparent targeting is retention. Yano T et al. 1997
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Chimeric ALL-1 proteins expressed in transfected cells show a punctate distribution assignable to two short trithorax-conserved sequences within the N-terminal ~1,100 residues, independently placing the targeting information in retained KMT2A sequence.

    Nuclear punctate distribution of ALL-1 is conferred by distinct elements at the N terminus of the protein · Yano T, et al. · Proceedings of the National Academy of Sciences of the United States of America 1997
    PMID 9207083DOI
    What this evidence cannot show

    Transfected overexpression system; deletion mapping in constructs. Two independent groups converging on the amino terminus is the strength here, not either result alone.

    Evidence that argues against this (1)
    • If the punctate pattern is conferred by retained N-terminal sequence shared with wild-type KMT2A, it cannot be a fusion phenotype, and the many papers reporting it for fusions are reporting a wild-type property. Caslini C et al. 2000
Connected to