DTX4

DTX4 is a RING-domain E3 ubiquitin ligase of the DELTEX family that functions as a critical negative regulator of type I interferon signaling and a modulator of Notch pathway-dependent cell fate decisions. Its best-characterized activity is K48-linked polyubiquitination of TBK1, which it performs within an NLRP4 signalosome following DYRK2-mediated phosphorylation of TBK1 at Ser527 and USP38-dependent removal of competing K33-linked ubiquitin chains at Lys670, leading to proteasomal degradation of TBK1 and suppression of IFN-β production (PMID:22388039, PMID:26407194, PMID:27692986). DTX4 also ubiquitylates Notch1 at the cell surface to drive bilateral endocytosis and ADAM10-mediated cleavage in endocytic compartments (PMID:28611181), and its C-terminal DTC domain functions as a conserved ADP-ribose-binding module that recruits poly-ADP-ribosylated proteins for ubiquitination (PMID:32937373). Multiple viruses exploit DTX4-mediated TBK1 degradation to evade innate immunity—Zika virus promotes centrosomal DTX4 accumulation via NS3-CEP63 disruption (PMID:35793002), EBV lytic reactivation upregulates DTX4 translation through enhanced m6A methylation of its mRNA (PMID:36918845), and HBV induces DTX4 via p-STAT3 to degrade the antiviral factor APOBEC3B (PMID:39346550).

1. 2012 High

   Establishing that DTX4 is the effector E3 ligase in a previously uncharacterized NLRP4-dependent mechanism for shutting down type I IFN signaling answered how innate immune activation is restrained at the level of TBK1 turnover.

   Evidence Co-IP, K48-linked ubiquitination assays, and knockdown in human cells showing TBK1 degradation and IFN suppression

   PMID:22388039

   Open questions at the time

   • Upstream signal triggering NLRP4–DTX4 recruitment was unknown
   • Whether DTX4 targets TBK1 at specific subcellular sites was not addressed
   • Relative contribution of DTX4 versus other E3 ligases (e.g. TRIP) was unresolved
2. 2015 High

   Identification of DYRK2 as the kinase that phosphorylates TBK1 at Ser527 to prime DTX4 recruitment resolved how the NLRP4–DTX4 ubiquitination complex is temporally gated.

   Evidence In vitro kinase assay, mutagenesis of TBK1 Ser527, and Co-IP showing phosphorylation-dependent NLRP4/DTX4 binding

   PMID:26407194

   Open questions at the time

   • Signals activating DYRK2 in the context of infection were undefined
   • Whether other kinases can substitute for DYRK2 was untested
3. 2016 High

   Demonstrating that USP38 must first remove K33-linked ubiquitin from TBK1 Lys670 before DTX4 can attach K48-linked chains at the same site established a chain-editing mechanism governing TBK1 fate.

   Evidence Linkage-specific ubiquitination assays, in vitro deubiquitination, and Usp38-knockout mice

   PMID:27692986

   Open questions at the time

   • Identity of the E3 ligase attaching the initial K33 chains was not determined
   • Whether chain editing occurs at all DTX4 substrate sites was unknown
4. 2017 High

   Showing that DTX4 ubiquitylates Notch1 to drive bilateral endocytosis and redirect ADAM10-mediated S2 cleavage from the cell surface to endosomes expanded DTX4 function beyond innate immunity into Notch signaling mechanics.

   Evidence Live-cell imaging, Co-IP, ubiquitination assays, and dynamin inhibition in signal-sending/receiving cell pairs

   PMID:28611181

   Open questions at the time

   • The specific ubiquitin chain type on Notch1 was not defined
   • Whether DTX4-mediated Notch regulation is tissue-restricted was unclear
5. 2017 Medium

   DTX4 knockdown blocking adipogenic differentiation and elevating Wnt signaling in preadipocytes revealed a developmental role for DTX4 in lineage commitment beyond immune regulation.

   Evidence Stable shRNA knockdown in 3T3-L1 cells with lipid staining, gene expression, and cell cycle readouts

   PMID:28842252

   Open questions at the time

   • Direct ubiquitination substrate driving the adipogenic phenotype was not identified
   • In vivo adipogenesis data were lacking
   • Whether the effect is Notch- or TBK1-dependent was not tested
6. 2020 High

   Discovery that TRAF3IP3 promotes DTX4-dependent K48-linked ubiquitination of TBK1 at a second lysine (Lys372) in myeloid cells demonstrated cell-type-specific adaptors feeding into the same DTX4 degradation pathway.

   Evidence Co-IP, ubiquitination assay, Traf3ip3-KO primary myeloid cells, and myeloid-specific deletion mouse model

   PMID:32366851

   Open questions at the time

   • How Lys372 and Lys670 ubiquitination are coordinated or independently regulated was unresolved
   • Whether TRAF3IP3 and NLRP4 act in the same or separate complexes was unknown
7. 2020 High

   Structural and biochemical demonstration that the DTC domain of DTX4 binds ADP-ribose and recruits PAR-modified proteins for ubiquitination provided a molecular logic for substrate recognition beyond protein–protein interaction.

   Evidence Label-free mass spectrometry, structural analysis, biochemical binding assays, and domain mutagenesis

   PMID:32937373

   Open questions at the time

   • Whether PAR-binding is required for TBK1 or Notch1 targeting specifically was not tested
   • Structural resolution of the DTX4 DTC domain itself was not reported
8. 2021 Medium

   Identification of MAP4K1 as another adaptor that recruits DTX4 to TBK1 and IKKε broadened the repertoire of upstream scaffolds converging on DTX4-mediated innate immune suppression.

   Evidence Yeast two-hybrid, Co-IP, ubiquitination assay, DTX4 knockdown rescue, and IFN-β reporter

   PMID:34908452

   Open questions at the time

   • Redundancy among NLRP4, TRAF3IP3, and MAP4K1 in physiological settings was not dissected
   • Single-lab finding without independent replication
9. 2022 Medium

   Showing that ZIKV NS3 disrupts CEP63, enabling centrosomal DTX4 accumulation and local TBK1 degradation, established a spatially resolved viral immune evasion mechanism exploiting DTX4.

   Evidence Confocal microscopy, Co-IP, protein stability assays, and ZIKV infection with CEP63 knockdown

   PMID:35793002

   Open questions at the time

   • Whether centrosomal TBK1 pool has distinct signaling roles was not determined
   • Generalizability to other flaviviruses was untested
10. 2023 Medium

    Multiple findings in 2023 collectively showed that diverse viruses (EBV, HBV) and cancers exploit DTX4 upregulation—via m6A-enhanced translation or STAT3 signaling—to degrade antiviral factors (TBK1, APOBEC3B) or promote tumor growth through SCD regulation.

    Evidence MeRIP-seq and polysome profiling for EBV [PMID:36918845]; Co-IP, ubiquitination, and HBV-infected liver chimeric mice for HBV [PMID:39346550]; xenograft models and SCD inhibitor rescue for thyroid cancer [PMID:37612343]

    PMID:36918845 PMID:37612343 PMID:39346550

    Open questions at the time

    • Whether DTX4 directly ubiquitinates SCD or acts indirectly in thyroid cancer was not established
    • APOBEC3B ubiquitination site was not mapped
    • Each study from a single lab
11. 2023 Medium

    DTX4 overexpression in the postnatal mouse retina reduced AII amacrine cell number, linking DTX4-Notch regulation to quantitative control of retinal neuron fate.

    Evidence In vivo retinal overexpression, quantitative cell counting, and QTL analysis

    PMID:36816119

    Open questions at the time

    • Whether the effect is Notch-dependent was assumed but not formally tested
    • Loss-of-function data in the retina were absent
12. 2024 High

    Demonstrating that DTX4, unlike DTX3L, cannot ubiquitylate nucleic acids via ester bonds clarified the functional boundaries within the DELTEX family despite their shared DTC domain architecture.

    Evidence In vitro ubiquitylation assays with purified DELTEX family members comparing nucleic acid substrates

    PMID:39242775

    Open questions at the time

    • Whether DTX4 has unique substrates not shared with DTX3L was not explored
    • The structural basis for differential nucleic acid ubiquitylation activity was not resolved

• A unified model explaining how DTX4 partitions between its TBK1-degradation, Notch-regulatory, and PAR-dependent substrate recognition functions—and whether these represent independent or interconnected pathways—remains unestablished.
• No structural model of full-length DTX4 exists
• Relative importance of PAR-binding versus adaptor-mediated substrate recruitment in vivo is unknown
• Comprehensive identification of DTX4 substrates beyond TBK1, Notch1, and APOBEC3B has not been performed