henhaobio
Sign in

bioRxivCell BiologyPreprint

Structural Mechanisms of DNAJC13 Dimeric Assembly and InsP6 binding in Recycling Endosome Regulation

Tao FuiD, Chan LeeiD, Frances V HundleyiD, João A. PauloiD, J. Wade HarperiD

PDFDOICodeData
In-depth full-text digest available · Jump to digest ↓

Abstract

The balance of plasma membrane protein degradation and recycling during endocytosis is regulated, in part, by the large J-domain-containing protein DNAJC13/RME-8 and WASH complexes, which function together with Retromer to support cargo trafficking into recycling endosomes. Despite extensive cellular and biochemical studies, the structural basis and proteomic landscape of DNAJC13 function remain elusive. Here, we find that DNAJC13 forms an unexpected antiparallel homodimeric architecture involving distinctive symmetrical interactions between composite IWN1 and -solenoid ARM2 domains in each protomer. Additionally, two PH-like domains of unknown function (PHL2 and PHL3), adjacent to the PI(3)P-binding PHL1 domain, form an unanticipated composite, positively charged pocket occupied by InsP6, as visualized structurally and verified by mass spectrometry. Proteomic profiling of DNAJC13-associated endosomes revealed enrichment of recycling endosomal components and WASH complexes. Mutations disrupting the dimer interface disable recruitment of WASH complexes to endosomes and result in elongated endosomal tubulation. Disruption of InsP6 binding impairs DNAJC13 binding to PI(3)P-containing vesicles in vitro and Transferrin-positive endosomes in cells. We further demonstrate that DNAJC13 dimerization and InsP6 binding promote melanin production during melanosome maturation, a process known to require recycling endosomes. This work provides a structural and mechanistic framework for understanding DNAJC13 function in recycling endosome control.

In-depth summary

Claude Sonnet 4.6 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (29 Jul 2026) · CC BY 4.0

Cryo-EM at 2.4 Å shows DNAJC13 as an antiparallel arch-shaped homodimer with a buried InsP6 pocket; dimerization controls WASH recruitment, tubule fission, and melanin production.

DNAJC13 supports recycling endosome function with Retromer and WASH, but prior work relied on co-immunoprecipitation of truncations and co-localization imaging, leaving its structural assembly and endosome proteome unknown. This work supplies a cryo-EM structure and an Endo-IP proteomic handle.

Key results

  1. ShownDNAJC13(1-1486) forms a symmetric arch-shaped antiparallel homodimer held by an electrostatic Site 1 interface and a hydrophobic Site 2 interface. Site 1 mutations abolish dimerization, while the Site 2 mutant stays dimeric.

    Consensus map 2.4 Å and four focused maps 2.5-2.7 Å. Mass photometry of 1-2191 gave a 504 kDa dimer against a 251 kDa theoretical monomer. Both Site 1 mutants showed only monomer and monomer plus 70 kDa peaks, s2m kept dimer peaks, and ALFA co-IP failed for both Site 1 mutants. · Cryo-EM structure of a DNAJC13 Dimer; DNAJC13 dimerization via two discrete interfaces; Figure legends

  2. ShownPHL2 and PHL3 form a composite positively charged pocket occupied by InsP6, and an eight-residue alanine mutant (8A) lacks bound InsP6 by two orthogonal readouts.

    Cryo-EM density matched InsP6 in the PHL2/PHL3 pocket. LC-MS detected the [M-2H]2- ion at 328.9229 m/z in WT but not 8A across three truncations. A g-CPD cleavage biosensor (three replicate assays) responded to WT but not 8A samples. · Composite InsP6 Binding Site Formed by PHL2-PHL3

  3. SuggestedDisrupting InsP6 binding reduces DNAJC13 recruitment to PI(3)P-containing GUVs in vitro and weakens Transferrin-endosome colocalization in cells. The 8A mutant was also less abundant, which confounds the cell result.

    At 80 nM on GUVs with 5% PI(3)P, WT accumulated over time while 8A stayed largely diffuse. E264A, expressed near WT levels, lowered the Pearson coefficient with Transferrin at all time points (5-25 min). · InsP6 Binding Supports DNAJC13 Association with…

  4. ShownDNAJC13 Endo-IP isolates a TGN and recycling-endosome-biased population enriched for WASH subunits relative to EEA1 Endo-IP. Dimerization-defective mutants lose WASH subunits but keep Retromer and most cargo.

    TMTpro 16-plex, four biological replicates: 441 shared proteins (FC > 1.5, q < 0.01), 69 DNAJC13-specific (all WASH subunits, HSP70s, AP-4), 159 EEA1-specific. s1m1 and s1m2 Endo-IPs essentially lacked WASH subunits, unlike WT and s2m. · Proteomic landscape of endosomes using DNAJC13…; DNAJC13 dimerization promotes WASH complex…

  5. ShownMonomeric DNAJC13 fails to normalize the elongated endosomal tubules of knockout cells and, like the 8A mutant, fails to rescue melanin production in MNT-1 cells.

    SUM159 DNAJC13-KO at 4 min: tubule Feret 1.8 μm (WT rescue) vs 2.45 μm (s1m2); puncta 0.98 vs 1.1 μm. In MNT-1 DNAJC13-KO cells, WT re-expression restored melanin while s1m2 and 8A did not. · DNAJC13 Dimerization Controls Endosomal Tubular…; DNAJC13 dimerization and PHL2/3-domain integrity…

Methods, data and reproducibility

Single-particle cryo-EM of DNAJC13(1-2191) with AlphaFold-based composite modeling; InsP6 detected by LC-MS and a cleavage biosensor; GUV and Transferrin assays of structure-guided mutants; TMTpro Endo-IP proteomics in knockout and rescue cells.

  • Cryo-EM dataset of FLAG-DNAJC13(1-2191) purified from Expi293F cells; model built for residues 1-1486 · human · n = 21786 movies; 599,851 particles in final reconstruction · PDB 37FE; EMD-78135 (composite), EMD-78131 (consensus)
  • TMTpro Endo-IP proteomics of DNAJC13- and EEA1-associated endosomes, dimer mutants, and WASHC2A-KO cells in HEK293T · human · n = 4 biological replicates per condition; TMTpro 16-plex · PXD080914, PXD080921
  • LC-MS detection of InsP6 on WT vs 8A DNAJC13 purified from Expi293F cells · human · n = 3 DNAJC13 truncations
  • Transferrin uptake, tubulation and colocalization imaging in SUM159 DNAJC13-KO and WASHC2A-KO cells · human
  • Melanin quantification in MNT-1 melanoma DNAJC13-KO and rescue lines · human
  • Integrated full-length model and AlphaFold3 predictions of DNAJC13 dimers and lipid complexes · human · PDB-IHM 5-FDMP; ModelArchive ma-z9nt8, ma-sgzb0, ma-germ0, ma-vp0iy
Methods & tools
cryoSPARC v4.4.1, Coot v.0.9.8, Phenix v.1.19.2, AlphaFold2-Multimer v2.3.2, AlphaFold3, FoldSeek, ChimeraX v.1.10.1, Refeyn mass photometry, Orbitrap Exploris 240 LC-MS, FragPipe 23.1, Comet search engine, MSstatsTMT
Software & environment
CryoSPARC v4.4.1, Coot v.0.9.8, Phenix v.1.19.2, ChimeraX v.1.10.1, FragPipe 23.1, AlphaFold2-Multimer v2.3.2; UniProt Swiss-Prot release 2024-01 used as the human reference proteome.
Assay conditions & compute
AlphaFold predictions were run on the O2 cluster (Harvard Medical School) or the AlphaFold3 web server; GPU types and runtimes are not stated.
Reproducibility gaps
Structure and map datasets are to be released at publication; cell lines are provided upon request; additional raw data are available from the lead contact on request.
Limitations (5)
  • The C-terminal IWN3-ARM5 region and extreme C-terminus lack experimental structure (flexible), so the full-length composite relies on AlphaFold predictions and how these regions engage partners is unclear. · stated by the authors
  • How HSP70 engages DNAJC13 is unresolved: the J-domain in the composite structure would clash with the other protomer, and HSP70 may act on folding rather than J-domain function. · stated by the authors
  • It is unresolved whether InsP6 acts purely as a structural cofactor or as a regulated signal, and the surfaces engaging sorting nexins, Retromer and WASH are undefined. · stated by the authors
  • The 8A mutant (eight substitutions) was expressed at reduced levels in cells, so its weaker endosome targeting cannot be cleanly attributed to loss of InsP6 binding alone. · noted in this summary
  • Functional assays rely on overexpressed variants in knockout lines, and quantified blots are the mean of 3 technical repeats, not independent biological replicates. · noted in this summary

Read this if you study endosomal recycling, WASH or Retromer biology, or want cryo-EM and Endo-IP proteomics evidence linking DNAJC13 dimerization and InsP6 binding to trafficking.

Citations
1

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISAdoff '25Norris '22Cui '26Laidlaw '21Rosendale '19Dostal '22
77% semantic matchCell Biology
Hayden Adoff, Brandon Novy, Emily Holland+1 more·Molecular Biology of the Cell·Jul 30, 2025· 3 citations