Oncofusion

start / graph / ptm-pml-rara-k160-sumoylation-the-retained-site-that-became-a-drug

Mechanismstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.q165-the-retained-acetylation-site

PML K160 SUMOylation - the retained-portion PTM that became an approved therapy, and the benchmark every other retained site should be measured against

also called PML-RARA K160 · arsenic RNF4 SUMO · retained PTM benchmark

In one leukaemia fusion protein, a chemical tag on a spot the fusion keeps is the exact thing arsenic trioxide exploits to destroy it. Mutate that one lysine and arsenic stops working. That is what a retained modification site looks like when it has been taken all the way.

The same thing, for a scientist

PML K160 (UniProt P29590, B1 box) is SUMOylated, is retained in both the bcr1/L and bcr3/S forms of PML::RARA (breakpoints at PML aa 552 and aa 394 respectively), and is the residue through which arsenic trioxide acts: arsenic enhances SUMO conjugation, the SUMO-targeted ubiquitin ligase RNF4 then builds K48 polyubiquitin and the fusion is degraded. A K160R PML::RARA is not degraded by arsenic. This is the completed version of the argument this beat is making about KMT2A K1130 - a regulated modification in retained sequence, demonstrated ON the fusion, mechanistically linked to an approved drug - and the contrast is the point.

Go deeper
The chain, end to end, and what was shown on the fusion itself
stepshowncitation
three major SUMO acceptor sites on PML: K65, K160, K490yes, on wild-type PML1
K160R in PML/RARα abolishes arsenic-induced degradationyes, on the fusion, in retrovirally transduced NIH3T32
arsenic → SUMO → RNF4 → K48 polyUb → proteasomeyes, and differentiation is impaired by a non-degradable PML-RARα SUMO mutant and by dominant-negative RNF43
PIAS1 is the SUMO E3 for this event, on the fusionyes4

Residue check, run in this beat and not taken on report: P29590 is 882 aa; position 160 is K; context 155-170 is HQWFLKHEARPLAELR. K65 and K490 are also lysines 5.

🟠 Two corrections to the textbook version, both from the primary literature

K65 is disputed. Lallemand-Breitenbach 2001 states K65 is not in a SUMO consensus and that K65R did not change PML's conjugation pattern before or after arsenic 2. The 1998 triple-site paper's K65 call is not independently supported.

K490 is conditionally retained, not retained. PML::RARA breaks at PML aa 394 (bcr3 / S-form) or PML aa 552 (bcr1 / L-form). K65 and K160 are in both. K490 is in the L-form only and is LOST in the S-form, which is roughly a third of adult APL 6. So even in the best-worked example in oncology, one of three sites is breakpoint-variable - exactly the VARIABLE class this beat had to invent for KMT2A K1364.

And the CK2 phosphorylation site on PML is LOST. PML S565 (canonical numbering; = 'Ser517' in the short-isoform numbering of the original Cell paper, a +48 offset that was resolved and checked - P29590 position 565 is S, and the SIM VVVI sits at 556-559) lies in exon 7a, downstream of both breakpoints. Every PML::RARA lacks the SIM and the CK2 sites. A fusion can lose a regulated PTM as decisively as it keeps one 6.

Why this is the benchmark and not just an analogy

Line up what is established for PML K160 against what is established for KMT2A K1130:

PML K160KMT2A K1130
site mapped✅✅
retained in the fusion✅✅
modification demonstrated ON the fusion protein✅❌ never attempted
point mutant tested for a fusion phenotype✅ K160R❌ K→R exists but only in a fragment, for antibody validation
enzyme identified✅ PIAS1 / RNF4❌ writer unassigned at the residue
therapeutic exploitation✅ arsenic trioxide, curative in APL❌

Five of six rows are empty on the KMT2A side. That is the honest measure of how early this is, and it is also a specification: the PML::RARA literature names exactly which experiments convert a retained site into a target.

Facts and figures
fusionPML::RARA
sitePML K160
uniprotP29590
length AA882
partner side5'
modificationSUMOylation
writerPIAS1 / Ubc9
reader effectorRNF4 (SUMO-targeted ubiquitin ligase)
fusion statusRETAINED in both bcr1 and bcr3 forms
shown for fusionPML::RARA - K160R abolishes arsenic-induced degradation of the fusion protein itself
therapeutic linkarsenic trioxide; the site is the mechanism of an approved, curative therapy in APL
residue verified in this beattrue
breakpoints pml AA
{
  "bcr3_short": 394,
  "bcr1_long": 552
}
other sites
{
  "K65": "RETAINED but disputed as a genuine SUMO acceptor",
  "K490": "VARIABLE - retained in bcr1/L only",
  "S565": "LOST in all forms - CK2 site in exon 7a, downstream of both breakpoints"
}
measurements
[
  {
    "quantity": "residue position of the SUMO acceptor in PML",
    "subject": "P29590 K160",
    "value": 160,
    "units": "residue",
    "n": 1,
    "universe": "UniProt P29590 canonical, 882 aa",
    "null": null,
    "note": "context HQWFLKHEARPLAELR at 155-170; retained under both the 394 and 552 breakpoints"
  },
  {
    "quantity": "isoform numbering offset between the Cell 2006 CK2 site and canonical PML",
    "subject": "PML 'Ser517' vs canonical S565",
    "value": 48,
    "units": "residues",
    "n": 1,
    "universe": "P29590 canonical vs isoforms -11/-12/-13",
    "null": null,
    "note": "resolved by locating the SIM motif VVVISSS at 508 in the short isoform and 556 in canonical; 517+48=565 and P29590[565]=S"
  }
]
How this could bite us back
  • The PML::RARA story is the single most seductive precedent in fusion oncology and it is almost certainly not generalisable. PML is a self-assembling nuclear-body scaffold whose degradation is its regulated fate; KMT2A is a chromatin enzyme scaffold whose abundance the source paper reports is NOT circadian. Reasoning from one to the other assumes a shared logic that the biology does not obviously supply.
  • Arsenic works in APL partly because APL is uniquely dependent on one fusion protein's presence. A KMT2A-rearranged blast carries a wild-type KMT2A allele and a partner's wild-type allele; degrading the fusion is a different proposition when the sequence being targeted is shared.
Still open
  1. What are the primary sources for the PML aa 394 and aa 552 breakpoints? Every retained/lost call for PML in this entry rests on one review, which is a weaker footing than this graph's KMT2A calls and should be upgraded.
  2. Is any KMT2A fusion degraded through a SUMO-targeted ubiquitin ligase? PML::RARA's whole therapeutic mechanism is degradation triggered at a retained PTM. This graph records no equivalent for KMT2A fusions, and the question of whether one exists has not been asked here.
Every source behind this entry (6)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    PML is SUMOylated at three major sites - K65 (RING), K160 (B1 box) and K490 (NLS) - and a triple K-to-R mutant cannot be sumoylated and mislocalises from nuclear bodies to the nucleoplasm.

    Identification of three major sentrinization sites in PML · Kamitani T, Kito K, Nguyen HP, Wada H, Fukuda-Kamitani T, Yeh ET · Journal of Biological Chemistry 273:26675-82 1998
    PMID 9756909DOI
    What this evidence cannot show

    Wild-type PML only, not the fusion. Abstract and title verified via Europe PMC in this beat; the full text was not read here and the site assignments are taken from the reporting worker's read plus the later literature.

    Evidence that argues against this (1)
    • The K65 assignment from this paper is contradicted by Lallemand-Breitenbach 2001 [E2], which reports K65 is not in a SUMO consensus and that K65R does not alter the conjugation pattern. One of the three 'major' sites is therefore doubtful.
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Mutation of K160 in PML/RARα abolishes arsenic-induced degradation of the fusion protein, and K160R PML/RARα fails to relocate to nuclear bodies on arsenic.

    Role of promyelocytic leukemia (PML) sumolation in nuclear body formation, 11S proteasome recruitment, and As2O3-induced PML or PML/RARA degradation · Lallemand-Breitenbach V, Zhu J, Puvion F, Koken M, Honoré N, Doubeikovsky A, Duprez E, Pandolfi PP, Puvion E, Freemont P, de Thé H · Journal of Experimental Medicine 193:1361-71 2001
    PMID 11413191DOI
    What this evidence cannot show

    The K160R-on-fusion result is a figure-3d legend statement in retrovirally transduced NIH3T3 cells - an overexpression system in a non-haematopoietic line, not primary APL blasts. This beat read the title and metadata directly; the figure legend was read by the reporting worker agent and is quoted at second hand.

    Evidence that argues against this (1)
    • NIH3T3 is a mouse fibroblast line. A degradation result there does not by itself establish the mechanism in a human promyelocyte, and the clinical validation of the arsenic mechanism came separately and later.
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Arsenic degrades PML and PML-RARα through SUMO-triggered, RNF4-dependent ubiquitin-mediated proteolysis, and differentiation is impaired in cells transformed by a non-degradable PML-RARα SUMOylation mutant.

    Arsenic degrades PML or PML-RARalpha through a SUMO-triggered RNF4/ubiquitin-mediated pathway · Lallemand-Breitenbach V, Jeanne M, Benhenda S, Nasr R, Lei M, Peres L, Zhou J, Zhu J, Raught B, de Thé H · Nature Cell Biology 10:547-55 2008
    PMID 18408733DOI
    What this evidence cannot show

    Abstract-level for this beat. The co-discovery paper (Tatham et al., Nat Cell Biol 2008) was deliberately NOT cited because it was not retrieved, so the RNF4 result rests here on one of the two papers that reported it.

    Evidence that argues against this (1)
    • Read at abstract level only. The abstract does not say whether the impaired-differentiation result used a non-degradable PML-RARα or a non-degradable PML, and this entry's summary asserts the fusion - a distinction that matters and that was not checked in the text. The co-discovery paper (Tatham et al., Nat Cell Biol 2008) was deliberately not retrieved, so a finding reported simultaneously by two groups is carried here on one of them.
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    PIAS1-mediated SUMOylation of PML-RARA is essential for arsenic-trioxide-induced degradation of the fusion.

    The SUMO E3-ligase PIAS1 regulates the tumor suppressor PML and its oncogenic counterpart PML-RARA · Rabellino A, Carter B, Konstantinidou G, Wu SY, Rimessi A, Byers LA, Heymach JV, Girard L, Chiang CM, Teruya-Feldstein J, Scaglioni PP · Cancer Research 72:2275-84 2012
    PMID 22406621DOI
    What this evidence cannot show

    Abstract-level. A single group's assignment of the E3; SUMO E3 ligases are redundant and PIAS-family assignments have been revised elsewhere.

    Evidence that argues against this (1)
    • SUMO E3 ligases are substantially redundant and single-group E3 assignments are revised often, so 'PIAS1 is the E3' should be read as 'PIAS1 is sufficient in this system'. There is also a non-independence problem inside this entry: the Scaglioni group produced both this PIAS1 result and the CK2/PML-Ser517 result cited in the body, so two of the entry's PML claims share a laboratory. Abstract-level only.
  5. E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt P29590 (PML) is 882 aa; positions 65, 160 and 490 are lysines and position 565 is a serine, with the SIM motif VVVI at 556-559.

    What this evidence cannot show

    Residue identity verified in this beat against the live UniProt FASTA on 2026-09-24, independently of the worker agent that supplied the reference. Canonical isoform P29590 only. PML has many isoforms with different C-termini and the literature numbering is inconsistent between them, which is exactly the trap the S517/S565 offset represents.

    Evidence that argues against this (1)
    • PML has multiple major isoforms with different C-termini and the literature numbers inconsistently between them - which is exactly the trap this entry documents in resolving 'Ser517' to canonical Ser565. That reconciliation is an INFERENCE from the position of the VVVISSS motif, not a mapping stated by any source, and if the Cell 2006 paper used a different short isoform than P29590-11/-12/-13 the offset would differ and the S565 identification would be wrong. The K65/K160/K490 positions do not depend on it.
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review

    PML::RARA breakpoints lie at PML aa 394 (bcr3 / short form) and aa 552 (bcr1 / long form), so K65 and K160 are retained in both while K490 is retained only in the long form; all PML-RARα fusions lack the PML SIM and the exon 7a CK2 phosphorylation sites.

    Differential Roles of PML Isoforms · Nisole S, Maroui MA, Mascle XH, Aubry M, Chelbi-Alix MK · Frontiers in Oncology 3:125 2013
    PMID 23734343DOI
    What this evidence cannot show

    THIS IS A REVIEW and it is the only source here for the breakpoint coordinates. No primary breakpoint-mapping paper was retrieved in this beat, so every retained/lost call for PML residues rests on a secondary source. That is weaker than the equivalent KMT2A calls in this beat, which rest on Ensembl exon coordinates.

    Evidence that argues against this (1)
    • If the aa 394 / aa 552 coordinates are wrong or isoform-dependent, the K490 conditional-retention claim changes. PML isoform numbering is notoriously variable and the review does not state which isoform its coordinates refer to.
Connected to