EIF3H

EIF3H is a non-core subunit of the eukaryotic translation initiation factor 3 (eIF3) complex that selectively promotes translation of mRNAs bearing upstream open reading frames (uORFs) and, through its interaction with METTL3, facilitates mRNA looping and ribosome recycling to enhance translation of m6A-modified transcripts (PMID:15548739, PMID:30232453, PMID:23716667). As a JAMM/MPN-family deubiquitylase, EIF3H removes K48-linked polyubiquitin chains from multiple oncoproteins—including YAP, Snail, HAX1, OGT, CCND1, and β-catenin—thereby stabilizing them against proteasomal degradation and activating downstream signaling cascades such as MAPK/ERK and Wnt/β-catenin (PMID:32269044, PMID:32867821, PMID:38514606, PMID:37559097). Within the eIF3 holocomplex, EIF3H associates with 40S ribosomal subunits and is dispensable for bulk translation but is required for transcript-specific translational control, including mRNAs critical for vertebrate lens and neural development (PMID:19061185, PMID:23716667). Conditional deletion of Eif3h suppresses colorectal tumorigenesis in mouse models, and a colorectal cancer risk variant at 8q23.3 modulates EIF3H expression through long-range promoter interaction (PMID:38514606, PMID:20862326).

1. 2004 High

   The first evidence that eIF3h is not required for general translation but selectively promotes reinitiation on uORF-containing mRNAs established its role as a transcript-specific translational regulator rather than a core initiation factor.

   Evidence Polysome fractionation and transient expression assays in Arabidopsis eif3h mutants showing selective loss of uORF-containing mRNA translation

   PMID:15548739

   Open questions at the time

   • Mechanism of reinitiation promotion not resolved at the molecular level
   • Unclear whether the COP9 signalosome interaction is functionally relevant to translation
2. 2006 Medium

   Gain- and loss-of-function experiments demonstrated that EIF3H overexpression is sufficient to drive oncogenic transformation, linking its translational activity to cell proliferation and cancer.

   Evidence Tet-Off overexpression in NIH 3T3 cells showing anchorage-independent growth; siRNA knockdown inhibiting breast and prostate cancer cell growth

   PMID:16652384

   Open questions at the time

   • Mechanism was not distinguished between translation-dependent and deubiquitylase-dependent functions
   • Single lab study
3. 2007 High

   Identification of measles virus N protein as a direct EIF3H-binding partner that inhibits translation established EIF3H as a viral target for host translational shutdown.

   Evidence Yeast two-hybrid, mammalian Co-IP, and in vitro translation inhibition with recombinant GST-N protein

   PMID:17686838

   Open questions at the time

   • Relevance to natural MV infection pathogenesis not demonstrated
   • Whether other eIF3 subunits are co-targeted was not resolved
4. 2008 High

   Biochemical purification of the fission yeast eIF3 complex showed that eIF3h is a non-core subunit dispensable for holocomplex integrity and general translation, but required for meiosis/sporulation, confirming its auxiliary regulatory role across eukaryotes.

   Evidence TAP-tag purification, polysome profiling, genetic deletion and human ortholog complementation in S. pombe

   PMID:19061185

   Open questions at the time

   • Mechanism of meiotic defect not elucidated
   • Whether distinct eIF3 subcomplexes have separable functions was not resolved
5. 2010 Medium

   A colorectal cancer risk SNP at 8q23.3 was shown to physically contact the EIF3H promoter via long-range chromatin interaction, directly linking genetic variation in EIF3H expression to cancer susceptibility.

   Evidence Chromosome conformation capture (3C) and reporter assays in CRC cell lines

   PMID:20862326

   Open questions at the time

   • Causal variant not fine-mapped to single nucleotide resolution
   • Effect on endogenous EIF3H expression not quantified in patient tissue
6. 2013 High

   Genome-wide polysome profiling in zebrafish embryos revealed that eIF3h controls translation of a specific mRNA cohort required for lens and neural development, extending its transcript-selective function to vertebrate organogenesis.

   Evidence Morpholino knockdown with genome-wide polysome profiling and UTR dissection in zebrafish

   PMID:23716667

   Open questions at the time

   • cis-regulatory elements beyond UTRs not identified
   • Whether the deubiquitylase activity contributes to developmental phenotypes was not tested
7. 2018 High

   Discovery that METTL3 directly binds eIF3h to promote mRNA looping between 3′ m6A sites and 5′ cap-binding complexes revealed the molecular mechanism by which m6A enhances translation through ribosome recycling, and established the METTL3-eIF3h axis as a driver of oncogenic translation.

   Evidence Co-IP, electron microscopy of polyribosomes, tethering assays, and reporter translation in human cells

   PMID:30232453

   Open questions at the time

   • Structural basis of METTL3-eIF3h interface not determined at atomic resolution
   • Whether this mechanism operates on all m6A-containing mRNAs or a subset is unknown
8. 2020 High

   Characterization of EIF3H as a JAMM-family deubiquitylase that removes K48-linked ubiquitin from YAP and Snail established a second biochemical activity independent of its translation initiation role, directly linking it to EMT and tumor invasion.

   Evidence Structure-based modeling with catalytic triad mutagenesis (D90/D91/Q121) and ubiquitination assays for YAP; reciprocal Co-IP/MS, CHX chase, and in vivo models for Snail

   PMID:32269044 PMID:32867821

   Open questions at the time

   • No crystal structure of EIF3H DUB domain solved
   • Whether DUB activity occurs within or outside the eIF3 holocomplex is unknown
   • Catalytic mechanism details (zinc coordination, processivity) not biochemically characterized with purified enzyme
9. 2023 Medium

   Extension of the DUB substrate repertoire to OGT and CCND1 showed that EIF3H stabilizes diverse oncoproteins across cancer types, broadening its role as a general oncoprotein stabilizer acting through K48-linked deubiquitylation.

   Evidence Co-IP with domain mapping (JAB/MPN–GT domain interaction for OGT); knockdown with CCND1 degradation assays in HCC and cholangiocarcinoma cells

   PMID:36350008 PMID:37559097

   Open questions at the time

   • In vitro reconstitution with purified components not performed
   • Substrate selectivity determinants not defined
   • Whether CCND1 is a direct or indirect target needs clarification
10. 2024 High

    Conditional Eif3h knockout in a mouse colorectal cancer model suppressed tumorigenesis and revealed that EIF3H deubiquitylates HAX1 to potentiate RAF1-MEK1-ERK signaling, while Wnt/β-catenin signaling induces EIF3H expression, establishing a feedforward oncogenic loop.

    Evidence Conditional knockout mouse (AOM/DSS model), Co-IP, ubiquitination assays, orthotopic and PDX models

    PMID:38514606

    Open questions at the time

    • Whether EIF3H's translation and DUB functions are independently required for tumorigenesis was not dissected
    • Tissue specificity of the Wnt-EIF3H feedforward loop not determined
11. 2025 Medium

    Phosphorylation of METTL3 at Ser2 was identified as a regulatory switch that disrupts the METTL3-eIF3H interaction, showing that the translational activity of the axis is dynamically regulated and targetable for chemosensitization.

    Evidence High-throughput base editor screen in gastric cancer cells with Co-IP validation of interaction disruption

    PMID:41385641

    Open questions at the time

    • Kinase responsible for METTL3 Ser2 phosphorylation not identified
    • In vivo validation of chemosensitization limited
12. 2025 Medium

    EIF3H was reported to deubiquitylate β-catenin, directly activating Wnt signaling in anaplastic thyroid cancer, with EIF3H expression itself regulated by m6A/IGF2BP2, closing a regulatory circuit between m6A RNA modification and Wnt pathway activation.

    Evidence Co-IP, ubiquitination assay, m6A reader identification in ATC cell lines

    PMID:39984062

    Open questions at the time

    • Single study without in vivo validation
    • Whether β-catenin deubiquitylation is direct (purified enzyme) not shown

• Key unresolved questions include: (1) whether the translation initiation and deubiquitylase activities of EIF3H are structurally and functionally separable within the eIF3 holocomplex; (2) what determines substrate selectivity among the growing list of DUB substrates; and (3) whether a high-resolution structure of the EIF3H MPN/JAMM domain can be obtained to enable pharmacological targeting.
• No atomic-resolution structure of EIF3H
• No separation-of-function mutant distinguishing translation from DUB activity
• No unbiased substrate screen (e.g., TUBE/diGly proteomics) performed