MLF2

MLF2 is a ubiquitously expressed chaperone-like protein that operates at the nuclear envelope and within chromatin to regulate protein and nucleoporin homeostasis, and that is co-opted in multiple cancers (PMID:8661158, PMID:32342107, PMID:40447637). At the nuclear envelope, MLF2 is a luminal component of aberrant nuclear envelope blebs that form upon Torsin ATPase manipulation; bleb biogenesis requires the transmembrane nucleoporin POM121 and occurs as cells reassemble nuclear pore complexes after mitosis, marking blebs as defective NPC intermediates (PMID:32342107). Within these blebs MLF2 suppresses ectopic FG-nucleoporin accumulation, modulates the selective properties and size of FG-nucleoporin condensates in vitro, and tunes a sequestered HSP40-HSP70 chaperone network that drives proteotoxicity in DYT1 dystonia (PMID:36302970). Beyond the envelope, MLF2 acts as a chaperone promoting SWI/SNF (BAF) complex assembly and chromatin binding, such that its acute depletion reduces chromatin accessibility at SWI/SNF-dependent sites (PMID:40447637). MLF2 protein levels are set by competing ubiquitination and deubiquitination: the CRL4-DCAF8 E3 ligase and STUB1 (ubiquitinating Lys119) drive proteasomal degradation, while USP11 and USP21 stabilize the protein, and PIM3-mediated phosphorylation at Ser65 promotes stabilization through USP21 (PMID:32703400, PMID:41090348). In cancer, MLF2 destabilizes p53 by blocking USP7-mediated deubiquitination to drive colorectal carcinogenesis (PMID:37438558), promotes osteosarcoma metastasis by disrupting the BiP-IRE1α interaction to activate the IRE1α/XBP1-S-MMP9 axis (PMID:41090348), and requires Ser24 phosphorylation for its oncogenic activity in CML (PMID:31831854). Consistent with its chaperone-like role in protein aggregation, MLF2 is sequestered into poly-GA inclusions in C9orf72 ALS/FTD and preferentially engages mutant polyQ huntingtin, where it increases aggregate mobility and reduces aggregate-induced apoptosis (PMID:28409281, PMID:27840155).

1. 1996 High

   Established MLF2 as a gene and protein in its own right, a ubiquitously expressed homolog of MLF1, defining the molecular entity before any function was known.

   Evidence cDNA cloning, FISH chromosomal mapping, and Southern blotting

   PMID:8661158

   Open questions at the time

   • No biochemical activity or interaction partner defined
   • No subcellular localization established
   • Functional role entirely unknown at this stage
2. 2014 Medium

   First functional implication: MLF2 supports breast cancer growth, metastasis, and cancer stem cell self-renewal, linking it to nitric oxide synthase signaling and hypoxia.

   Evidence shRNA/siRNA knockdown in patient-derived xenografts with RNA deep sequencing

   PMID:24876273

   Open questions at the time

   • No direct molecular mechanism connecting MLF2 to the NOS pathway
   • No physical partners identified
   • Phenotype correlative rather than mechanistic
3. 2017 Medium

   Implicated MLF2 in neurodegenerative aggregate biology by showing it is sequestered into pathological protein inclusions, hinting at a chaperone-like or aggregate-engaging activity.

   Evidence Co-IP, FRAP and viability assays with polyQ huntingtin in Neuro2A cells; immunofluorescence co-localization with poly-GA inclusions in mouse and ALS/FTD patient tissue

   PMID:27840155 PMID:28409281

   Open questions at the time

   • No reconstituted binding to demonstrate direct interaction with aggregating species
   • Mechanism of preferential mutant-HTT engagement undefined
   • Sequestration into poly-GA shown only by co-localization
4. 2020 High

   Defined MLF2 as a luminal nuclear-envelope-bleb component dependent on POM121, recasting blebs as aberrant NPC biogenesis intermediates and giving MLF2 a concrete nuclear-envelope role.

   Evidence Proteomics, live-cell imaging with an MLF2-based platform, immunofluorescence, and POM121 genetic depletion

   PMID:32342107

   Open questions at the time

   • Molecular activity of MLF2 within the bleb not yet defined
   • How MLF2 is recruited to the luminal space unknown
   • Functional consequence for mature NPCs not directly tied to MLF2
5. 2020 Medium

   Showed MLF2 abundance is post-translationally controlled by opposing ubiquitin machinery, identifying CRL4-DCAF8 as a destabilizer and USP11 as a stabilizer.

   Evidence Reciprocal Co-IP, proteasome inhibitor assays, and overexpression/knockdown

   PMID:32703400

   Open questions at the time

   • Ubiquitination site not mapped in this study
   • No physiological trigger for degradation defined
   • Single-lab finding
6. 2022 High

   Assigned a biochemical function to MLF2 at the envelope: suppression of FG-nucleoporin condensate accumulation and modulation of a bleb-sequestered HSP40-HSP70 chaperone network underlying DYT1 dystonia proteotoxicity.

   Evidence In vitro condensate reconstitution, live-cell imaging, Co-IP, and TorsinA genetic manipulation

   PMID:36302970

   Open questions at the time

   • Direct binding determinants for FG-nucleoporins not mapped
   • Which HSP40/HSP70 members are engaged and how is incomplete
   • Stoichiometry and structural basis of condensate modulation unknown
7. 2019 Medium

   Identified phosphorylation at Ser24 as functionally required for MLF2 oncogenic activity in CML, linking a post-translational mark to transforming function.

   Evidence Phos-tag SDS-PAGE, Ser24 mutagenesis, colony formation, and mouse xenograft survival

   PMID:31831854

   Open questions at the time

   • Kinase responsible for Ser24 phosphorylation not identified
   • Downstream effectors of phospho-MLF2 in CML undefined
   • Single-lab study
8. 2023 Medium

   Established a tumor-promoting mechanism: MLF2 destabilizes p53 by competing with USP7, blocking p53 deubiquitination to drive colorectal carcinogenesis.

   Evidence Co-IP with p53 and USP7, ubiquitination assay, loss-of-function, and in vivo tumor models

   PMID:37438558

   Open questions at the time

   • Structural basis of MLF2-USP7 competition unresolved
   • Whether this axis operates in other tumor types untested here
   • Single-lab finding
9. 2025 High

   Integrated MLF2 regulation and a metastasis mechanism: PIM3-Ser65 phosphorylation stabilizes MLF2 via USP21 while STUB1 ubiquitinates Lys119 for degradation, and MLF2 disrupts BiP-IRE1α to activate the IRE1α/XBP1-S-MMP9 axis driving osteosarcoma metastasis.

   Evidence In vivo CRISPR activation screen, interactome, Co-IP, kinase and ubiquitination assays, site-directed mutagenesis, and in vivo metastasis models

   PMID:41090348

   Open questions at the time

   • How phosphorylation at Ser65 mechanistically favors USP21 over STUB1 unclear
   • Direct binding interface with BiP/IRE1α not structurally defined
   • Whether the same regulatory logic applies outside osteosarcoma untested
10. 2025 High

    Revealed a chromatin function: MLF2 acts as a chaperone promoting SWI/SNF (BAF) complex assembly and chromatin engagement, with acute loss reducing accessibility at SWI/SNF-dependent sites.

    Evidence CRISPR KO screen with epigenome editing, degron-based rapid depletion, ATAC-seq, and ChIP-seq

    PMID:40447637

    Open questions at the time

    • Which BAF subunits MLF2 directly chaperones not specified
    • Whether nuclear-envelope and chromatin roles are mechanistically linked unknown
    • Structural basis of chaperone activity undefined
11. 2025 Medium

    Extended MLF2 condensate biology by using it as a genome-wide screening reporter, linking nuclear envelope and nucleoplasmic condensate accumulation to ZNF335 and RNF26.

    Evidence High-content imaging genome-wide CRISPR/Cas9 screen with MLF2 condensate reporter and machine-learning phenotyping (preprint)

    PMID:bio_10.1101_2025.06.07.658469

    Open questions at the time

    • Not yet peer-reviewed
    • Mechanistic role of ZNF335/RNF26 relative to MLF2 not established
    • MLF2's own activity in these condensates not directly tested here
12. 2025 Low

    Suggested a cardiac role through interaction with αB-crystallin and an anti-hypertrophic effect on cardiomyocyte gene expression.

    Evidence Pulldown/Co-IP for CryAB, immunofluorescence in mouse DRM models, cardiomyocyte overexpression

    PMID:41590846

    Open questions at the time

    • Single Co-IP/pulldown without reciprocal or structural validation
    • No pathway placement for the anti-hypertrophic effect
    • Causality versus correlation in heart failure models unresolved

• Whether MLF2's distinct activities — nuclear-envelope condensate modulation, SWI/SNF chaperoning, and oncogenic ubiquitin/IRE1α signaling — reflect a single unifying chaperone-like biochemical mechanism remains unresolved.
• No structure of MLF2 or its complexes
• Core biochemical activity not defined
• Tissue-specific selection among partners unexplained