A UBH-UBX module amplifies p97/VCP’s unfolding power to facilitate protein extraction and degradation

Huo, Xin-Yu; Liu, Di; Zou, Rong; Li, Zhao-Peng; Li, Yunxia; Pan, Lifeng; Zhang, Yaoyang; Zhang, Zai-Rong

doi:10.1038/s41467-025-65166-4

Download PDF

Article
Open access
Published: 19 November 2025

A UBH-UBX module amplifies p97/VCP’s unfolding power to facilitate protein extraction and degradation

Nature Communications volume 16, Article number: 10162 (2025) Cite this article

5822 Accesses
8 Citations
4 Altmetric
Metrics details

Subjects

Abstract

The p97-UFD1L-NPLOC4 ATPase unfolds numerous proteins for proteasomal degradation, but whether it suffices to pull proteins out of lipid bilayer remains unclear. Here, we identify a conserved ubiquitin-binding helix (UBH) in many UBX-containing p97 adapters, including FAF2, across yeast, plants, and metazoans. The UBH-UBX substantially facilitates the engagement of ubiquitinated substrates with p97-UFD1L-NPLOC4, and enhances p97 motor’s working ATPase and unfolding activities by approximately twofold. Using purified p97-UFD1L-NPLOC4-FAF2^UBH-UBX, we reconstitute membrane protein extraction from the ER and mitochondria, establishing p97-UFD1L-NPLOC4-FAF2 (p97-UNF) as a power-enhanced unfoldase. Deleting UBH or disrupting UBH-ubiquitin interaction impairs substrate targeting, reduces p97-UNF’s working ATPase and unfolding activities, and abolishes membrane protein extraction and degradation. We propose that UBH-UBX module amplifies p97’s mechanical output power, enabling the removal of challenging substrates from large assemblies and ensuring rapid responses to protein misfolding or regulatory signals in diverse physiological processes.

Cooperative assembly of p97 complexes involved in replication termination

Article Open access 03 November 2022

The UBX domain in UBXD1 organizes ubiquitin binding at the C-terminus of the VCP/p97 AAA-ATPase

Article Open access 05 June 2023

Highly specific intracellular ubiquitination of a small molecule

Article Open access 21 August 2025

Introduction

Membrane and secretory proteins constitute nearly a third of human proteome and are mainly synthesized at the endoplasmic reticulum (ER). When these proteins become misfolded or obsolete, they are destructed by 26S proteasome through the ER-associated degradation (ERAD) pathway^1,2,3, which maintains ER homeostasis and function, and regulates numerous pathological and physiological processes^4,5,6,7. Analogously, the mitochondria-associated degradation (MAD) disposes outer mitochondrial membrane (OMM) proteins^8,9,10,11,12, regulating mitochondrial homeostasis^11,13 and mass¹⁴, and apoptosis¹⁵. Degradation of ERAD or MAD substrates requires their extraction from the ER or mitochondria, a process driven by the p97/VCP-UFD1L-NPLOC4 unfoldase complex (or Cdc48-Ufd1-Npl4 in yeast)^1,2,16, following their ubiquitination by specific E3 ubiquitin ligases. The mammalian UFD1L-NPLOC4 (UN) or yeast Ufd1-Npl4 heterodimer preferentially captures K48-linked polyubiquitin chains^17,18, with NPLOC4/Npl4 unfolding an initiator ubiquitin¹⁹. This action triggers p97/Cdc48-driven, ATP hydrolysis-coupled unfolding of the remaining ubiquitin chain and the attached substrate protein^19,20,21,22, facilitating disassembly of multimeric complexes or extraction of proteins from lipid bilayers for proteasomal degradation.

Protein extraction from lipid bilayer poses unique challenges. For instance, membrane proteins are embedded within the lipid bilayer, with their hydrophobic transmembrane domains (TMDs) forming strong, stable interactions with surrounding phospholipids²³. Consequently, extraction requires not only the disruption of intermolecular chemical bonds but also the severing of extensive TMD-phospholipid interactions, imposing a high energy barrier to unfolding, as evidenced by single-molecule²⁴ and ensemble²⁵ studies. Some E3 ligase complexes have been proposed to function as retrotranslocons that locally thin the lipid bilayer, reducing the energy barrier for ERAD substrates to transit from the ER lumen to the cytosol^26,27. However, according to the law of conservation of energy, this localized membrane thinning does not decrease the total energetic cost required for p97-UN to extract membrane-integrated TMDs. If the mechanical power output of p97-UN is insufficient, the extraction process may become inefficient and prolonged, increasing the risk of protein aggregation and disruption of essential cellular pathways. Both p97-UN and Cdc48-UN have been shown to unfold soluble proteins^21,28,29, however, only the yeast Cdc48-UN has been demonstrated to extract membrane proteins in vitro^30,31,32. Given the substantial differences between yeast and mammalian UN complexes^17,28,33, it remains unclear whether the ATPase and unfolding activities of mammalian p97-UN are sufficient to efficiently extract integral membrane proteins.

The two-dimensional membrane surface imposes rotational and spatial limitations³⁴ on protein extraction, potentially limiting p97-UN’s ability to effectively engage ubiquitin chains located near the lipid bilayer. To overcome these limitations, membrane-anchored adapter proteins at the ER^35,36,37,38, mitochondria^10,15, lipid droplet³⁹, chloroplast and peroxisome^40,41,42 recruit p97/Cdc48 to membrane surfaces, enhancing its local concentration⁴³. However, whether recruitment alone is sufficient for ubiquitin chain recognition by the UN heterodimer or if additional ubiquitin-binding factors are required remains unclear. A hypothetical membrane-associated p97 adapter that can enhance ubiquitin chain engagement with p97-UN could improve the efficiency and timeliness of protein extraction from organelles.

FAF2, also known as UBXD8, is a p97 adapter protein with a UBX domain for p97 binding and a hydrophobic membrane anchor (MA) that localizes it to multiple organelles including the ER^39,44,45, OMM¹⁵, peroxisome^40,41,42, and lipid droplet^39,46. FAF2 and its orthologs (e.g., yeast Ubx2 and plant PUX10) can recruit p97/Cdc48 to membranes to facilitate the degradation of membrane proteins and the disassembly of stress granules⁴⁵, modulating various physiological activities^{10,35,36,40,47,48,49}. However, cell lacking FAF2 retains significant levels of membrane-associated p97¹⁵, consistent with findings that ER-decorated ubiquitin chain³⁰ and other adapters, such as UBXD2, UBXD6, VIMP, TMUB1, gp78, Hrd1, RHBDL4, and Derlin1/Derlin2, can also recruit p97 to ER membrane^37,38,43. This suggests that FAF2 may have functions beyond p97 recruitment, but its precise roles and mechanisms of action are poorly understood. A recent in vitro study has shown that FAF2’s coiled-coil and UBX domains can lower the ubiquitin chain length requirement for p97-mediated unfolding of MCM7 subunit of chromatin-associated CMG complex²⁸, indicating that FAF2 is required for unfolding proteins bearing short but not long ubiquitin chains; however, whether this mechanism applies to protein extraction from the ER and mitochondria—where FAF2 plays a physiologically relevant role—and the underlying molecular basis remain unknown.

In this work, we identify a ubiquitin binding helix (UBH) in many UBX-containing p97 adapters, including FAF2, FAF1 and their orthologs across eukaryotes. The UBH-UBX not only facilitates engagement of ubiquitinated substrates with p97-UN, but also accelerates p97 motor’s ATP hydrolysis activity and substrate processing by about twofold, thereby enhancing its mechanical power output and allowing FAF2 to enable the extraction from organellar surface. Notably, even substrates decorated with very long ubiquitin chains strictly require FAF2’s UBH-UBX for efficient extraction, consistent with the challenging task of extraction from membrane. The twofold-enhanced unfolding power of p97-UN-UBH-UBX thus provides a critical mechanism for efficiently clearing challenging proteins from large biological assemblies, highlighting its significance in cellular protein homeostasis.

Results

Identification of FAF2 required for HIV Vpu-Mediated CD4 Degradation

We hypothesized a potential factor facilitating association of p97-UN with ubiquitin chain at the membrane surface and sought to identify it by purifying ubiquitinated ERAD substrates. We exploited the HIV-1-encoded protein Vpu, which mediates the ERAD of the newly synthesized CD4 in the host ER (Supplementary Fig. 1A)⁵⁰. This system provided a useful model, as ubiquitin-conjugated CD4 can be efficiently purified and resolved by SDS-PAGE (Supplementary Fig. 1B, C).

Mass spectrometry analysis for the purified ubiquitinated CD4 (Supplementary Fig. 1D, E) identified known factors involved in CD4 degradation, such as the E3 ligase SCF^βTrCP ⁵¹ and p97-UN⁵², as well as the p97 adapter FAF2 (Supplementary Fig. 1E). Analysis of samples during purification confirmed that FAF2 interacted with ubiquitinated CD4 (Supplementary Fig. 1F). This was further supported by result showing that FAF2 interacted with CD4 in the presence of CB5083 and MG132, but not the E1 inhibitor TAK243 (Supplementary Fig. 1G). Tandem immunoprecipitation for CD4 and p97 revealed the presence of FAF2 in the eluted fraction, suggesting that CD4, FAF2, and p97 form a ternary complex (Supplementary Fig. 1H). FAF2 knockout (FAF2^KO) severely disrupted CD4 degradation, with the steady-state level increasing to ~10 times (Supplementary Fig. 2A, B). These results suggest that FAF2 is a previously unrecognized factor facilitating Vpu-mediated CD4 degradation.

Consistent with its role in ERAD^53,54,55, FAF2 depletion stabilized many other misfolded ERAD substrates (Supplementary Fig. 2C), including TCRα (Supplementary Fig. 2D, E), mTAP2⁵⁶ (Supplementary Fig. 2F, G), ABCG2-F208S⁵⁷ (Supplementary Fig. 2H, I), and α1-antitrypsin null Hong Kong variant (NHK, Supplementary Fig. 2J, K). Interestingly, while knocking down FAF2 stabilized CD4, depletion of other ER-localized p97 adapters, including UBXD2, UBXD6, and VIMP, did not (Supplementary Fig. 2L). These results indicate that, among many proteins capable of recruiting p97 to the membrane, FAF2 plays a specialized role in facilitating ERAD.

Identification of a UBX-adjacent α-helical domain (HD) in FAF2 required for protein extraction and degradation from the ER

In FAF2^KO cells, level of ubiquitinated CD4 was enhanced (Supplementary Fig. 3A, lane 9 vs lane 7), suggesting FAF2 is not required for ubiquitination of ERAD substrates. This prompted us to explore FAF2’s role in protein extraction from the ER using ERAD inhibitors (Fig. 1A). Impeding proteasomal degradation with MG132 led to appearance of a CD4 species that migrated slightly faster on SDS-PAGE (Fig. 1B, arrowhead) and localized predominantly in the cytosol (Fig. 1C, arrowheads). This species represents deglycosylated CD4, generated after extraction into the cytosol and subsequent deglycosylation by cytosolic NGLY1 (Fig. 1A)^56,57. Co-treatment with CB5083 and MG132 abolished this species, confirming that its appearance depends on p97 ATPase activity (Fig. 1B, lane 4). In FAF2 knockout cells, inhibition of proteasome did not yield the extracted CD4 species (Fig. 1B, lane 8). Re-expression of FAF2 in the FAF2^KO cells restored the appearance of the extracted CD4 (Fig. 1C). These results indicate that FAF2 specifically promotes p97-mediated substrate extraction from the ER into the cytosol. Extraction of ABCG2-F208S and TCRα were similarly impaired in FAF2^KO cells (Fig. 1D, Supplementary Fig. 3B, C), indicating that FAF2 is generally required for p97-mediated substrate extraction during ERAD.

**Fig. 1: Identification of an α-helical domain in FAF2 necessary for protein extraction and degradation from the ER.**

FAF2 contains several functional domains: a ubiquitin-association domain (UBA), a membrane anchor (MA) that retains FAF2 onto lipid droplet, ER, mitochondria or peroxisome, a ubiquitin-associated surface (UAS) domain, a 75-aa α-helical domain (HD, previously designated as coiled-coil), and a ubiquitin regulatory X (UBX) domain that binds to p97 (Fig. 1E). To explore their roles in ERAD, we made domain deletion mutants and tested their ability to rescue CD4 extraction and degradation in FAF2^KO cells. Expression of wild-type FAF2 or FAF2 lacking the UBA or UAS domains reduced CD4 levels (Fig. 1F, lanes 5, 6 and 8, and 1 G). By contrast, deletion of the UBX or HD domains failed to reduce CD4 levels (Fig. 1F, arrowheads), implying that they are necessary for FAF2’s function. We then analyzed CD4 extraction under proteasome inhibition. While p97-extracted CD4 species was observed in the presence of wild-type FAF2 or mutants lacking UBA or UAS domains, it was absent when FAF2 mutants lacking UBX or HD domains were expressed (Fig. 1H, I, and Supplementary Fig. 3D). These findings demonstrate that both the HD and UBX domains are required for p97-mediated substrate extraction from the ER.

The HD functions as a ubiquitin-binding helix (UBH) conserved in UBX-containing p97 adapters across eukaryotes

While the UBX domain is well-characterized for its role in recruiting p97⁵⁸, the HD’s function remains poorly understood. The HD of FAF2 and its paralog FAF1 has been previously classified as a coiled-coil domain^28,37,38, potentially involved in oligomerization. Using size-exclusion chromatography coupled with multi-angle light scattering (SEC-MALS), we demonstrated that the purified HD (Fig. 2A) existed as a monomer in solution (Supplementary Fig. 4A), suggesting that this long α-helix might interact with other molecules under physiological condition.

**Fig. 2: The HD serves as a UBH conserved in UBX-containing p97 cofactors across eukaryotes.**

Given dispensable role of FAF2’s UBA and prior findings that single α-helices can bind ubiquitin^59,60, we tested whether FAF2’s HD directly associates with K48-linked ubiquitin chains. To assess this, we purified HD with an N-terminal 2xStrepII tag and immobilized it on Strep-Tactin sepharose beads (Fig. 2A). Pulldown assays showed that the HD bound unanchored K48-linked ubiquitin chains, specifically di-ubiquitin and longer chains, but not control beads (Fig. 2B). Binding was also confirmed for SUMO-tagged HD (Supplementary Fig. 4B). Notably, the HD exhibited minimal interaction with K48-linked chains synthesized using I44A ubiquitin, suggesting dependence on the L8-I44-V70 hydrophobic patch of ubiquitin. Quantification revealed that HD bound tri-ubiquitin or longer chains more effectively than di-ubiquitin (Fig. 2C). Together, these results lead us to designate FAF2’s HD as a ubiquitin binding helix (UBH).

While UAS did not bind to K48-linked ubiquitin chains, the UBA did so more strongly than UBH (Supplementary Fig. 4C), suggesting UBH’s position next to UBX might be crucial for its essential function in protein extraction. To determine UBH’s evolutionary conservation, we performed sequence alignments of the HDs from FAF2, FAF1, their orthologs, and other selected UBX-containing proteins across eukaryotes. This analysis revealed a highly conserved N-terminal region within the HD (residues 289-307 in human FAF2) (Fig. 2D). To functionally test this conservation, we purified His-SUMO-tagged (Fig. 2E) or 2xStrepII-tagged HDs (Supplementary Fig. 4D) from diverse species, including yeast, plant, insect, and worm, and evaluated their ability to interact with K48-linked ubiquitin chains using pulldown experiments. All tested HDs bound ubiquitin chains, albeit with differences in binding affinity, indicating that they are functional UBHs. Interestingly, human FAF1’s UBH showed a sightly weak binding with ubiquitin chains, while UBHs from fruit fly and yeast displayed higher affinity than those from human, worm, and plant. Despite difference in UBH’s chain binding-affinity, FAF1’s UBH, engineered with a hydrophobic membrane anchor (MA), rescued CD4 extraction in FAF2^KO cells in a UBA- and UBL-independent manner (Fig. 2F–H), suggesting a shared ubiquitin recognition mechanism across cellular contexts.

UBH-ubiquitin interaction is essential for the extraction and degradation of diverse ER membrane proteins in cells

To identify the ubiquitin-binding sites (UBSs) in the UBH, we used AlphaFold to predict its structure in complex with three ubiquitin molecules. The model revealed two potential UBSs: UBS1 (L289) and UBS2 (Y297/L301), both interacting with the L8/I44/V70 hydrophobic patch (Fig. 3A). Supporting this, mild detergent washes partially removed ubiquitin chains from immobilized UBH (Supplementary Fig. 4E), confirming hydrophobic interactions. We validated these UBSs through mutagenesis. Strep-Tactin pulldown experiments demonstrated that mutations in UBS1 (L289N) and UBS2 (Y297H/L301N) each reduced the binding to ubiquitin chains, while combining these mutations (L289N/Y297H/L301N) nearly abolished the interaction (Fig. 3B, C). Furthermore, UBH bound to GFP modified with tri-ubiquitin or longer chains, and mutations in UBS2 (Supplementary Fig. 4F) or both UBSs (Fig. 3D) prominently reduced this binding. These results reflect differences between free and substrate-anchored ubiquitin chains, although they are both K48 linked. Together, these results establish UBS1 and UBS2 as critical for ubiquitin recognition.

**Fig. 3: UBH-ubiquitin binding is necessary for extraction and degradation of diverse membrane proteins from the ER.**

To assess the functional relevance of these UBSs, we expressed the UBH mutants in FAF2^KO cells and monitored CD4 extraction and degradation (Fig. 3E). Cells expressing UBH mutants exhibited impaired CD4 extraction (Fig. 3F–I, red arrowheads) and degradation (Fig. 3F, H, blue arrowheads), with the double UBS mutant (L289N/Y297H/L301N) exhibiting stronger defects compared to single UBS mutants. Similar results were observed for other ERAD substrates, TCRα and ABCG2-F208S (Fig. 3J–M), indicating that UBS1 and UBS2 are essential for ubiquitin recognition and protein extraction from the ER.

The UBH substantially enhances engagement of ubiquitin conjugates with p97-UN

To elucidate how UBH-ubiquitin interaction contribute to FAF2-facilitated protein extraction and degradation, we examined the interplay between FAF2 and the p97-UN complex. Size-exclusion chromatography confirmed that FAF2 co-fractionated with p97-UN as a stable p97-UNF complex (Supplementary Fig. 5A). Consistently, GST pulldown assay revealed that FAF2’s UBX alone is sufficient to bind purified p97-UN in vitro (Fig. 4A), in agreement with previous findings for FAF1’s UBX⁵⁸. Co-immunoprecipitation using recombinant proteins further showed that FAF2 marginally bound to p97, but addition of UN greatly enhanced the interaction (Fig. 4B). However, FAF2 UBH mutant (Y297H/L301N) bound to p97 even in the absence of UN, suggesting that the mutation stabilizes the interaction. Notably, the UBH mutant Y297H/L301N recovered similar amount of p97-UN complex as wild-type FAF2, suggesting that the UBH mutation impairs ubiquitin binding without disrupting p97-UN recruitment.

**Fig. 4: UBH-ubiquitin association enhances targeting of ubiquitinated proteins to p97-UFD1L-NPLOC4.**

We then tested whether FAF2 promotes the recruitment of ubiquitin-conjugated proteins to p97-UN. In vitro, super-folder GFP (sfGFP) with a N-terminal arginine was ubiquitinated by the N-end rule E3 ligase scUbr1²². The ubiquitinated sfGFP was incubated with immobilized p97 alone, or with p97 complexed with UN, FAF2 lacking the UBA and MA domains (F^ΔUBAΔM), or both (Fig. 4C). Consistent with earlier findings^22,29, p97-UN bound ubiquitinated sfGFP, with stronger binding observed for substrates with longer ubiquitin chains (Fig. 4D, lane 5). Importantly, the addition of FAF2 (F^ΔUBAΔM) significantly enhanced p97-UN’s association with ubiquitinated sfGFP bearing three or more ubiquitin molecules, including very long polyubiquitin chains (Fig. 4D, lane 6), demonstrating that FAF2 greatly facilitates the targeting of ubiquitin-modified substrates to p97-UN. In contrast, the UBH mutations Y289H/L301N severely disrupted this function of FAF2, leading to diminished recruitment of ubiquitin-conjugated sfGFP to p97-UN (Fig. 4E). Of note, in this assay, the amount of ubiquitin conjugates bound to p97 complexed with F^ΔUBAΔM was below the detection limit, likely due to weak interaction between p97 and FAF2 when UN is absent⁵⁸ (Fig. 4A, B). Additionally, UBH alone (without UBX) did not enhance engagement of ubiquitin conjugates with p97-UN (Supplementary Fig. 5B), suggesting necessary role of UBX.

To validate FAF2’s role in substrate targeting in cells, we measured the interaction between p97 and CD4 by co-immunoprecipitation in wild-type and FAF2^KO cells. Without FAF2, the amount of p97 co-precipitated with CD4 was evidently reduced (Fig. 4F, lane 2 vs lane 3, and 4 G), underscoring FAF2’s importance in linking CD4 to p97. Re-expression of wild-type FAF2 in knockout cells restored CD4 binding to p97, whereas FAF2 variants lacking the UBX or HD domains failed to do so (Fig. 4F, lanes 4, 5, and 6, and 4 G). These results demonstrate that both UBX and UBH of FAF2 are necessary for p97’s engagement with ERAD substrates. Collectively, our findings highlight the critical role of UBH in enhancing the efficiency of substrate targeting to p97 both in vitro and in cells.

The conserved UBH-UBX modules stimulate p97 motor’s working ATPase activity and accelerates substrate processing

To determine whether FAF2 enhances the unfolding power of p97, we first examined its impact on ATPase activity. Consistent with earlier findings^22,29,61, the ATP hydrolysis rate of p97-UN increased approximately threefold in the presence of a polyubiquitin chain (Fig. 5A, B), indicating that substrate processing stimulates p97-UN’s ATPase activity, representing its working ATPase activity⁶¹. We then demonstrated that, when ATP concentration was fixed at 1 mM, FAF2^ΔUBA/ΔM (lacking UBA and MA) boosted the working ATPase activity of p97-UN complex from ~1 to ~2.6 ATP molecules per second (Fig. 5A, red and cyan solid lines, 5B), while the UBH point mutant Y297H/L301N of FAF2^ΔUBA/ΔM failed to do so (Fig. 5A, red and purple solid lines, and 5B). Similar results were also observed for truncated FAF2 and FAF1 lacking UBA, MA, UAS and 2xUBL for FAF1, namely UBH-UBX domain, which accelerated ATP consumption rate to roughly twofold, but the UBH mutants did not. (Fig. 5C, D). In addition to wild-type p97, the UBH-UBX module also enhanced ATPase activity of IBMPFD-associated p97 mutants L198W and A232E when complexed with UN (Supplementary Figs. 6A, B), increasing their catalytic rate to 3-4 ATP per second.

**Fig. 5: Conserved UBH-UBX modules enhance p97-UN’s working ATPase activity and accelerate substrate unfolding in vitro.**

We next examined the steady-state ATPase kinetics of p97-UN in the presence or absence of the UBH-UBX modules from FAF1 or FAF2 (Fig. 5E). For p97-UN alone, the ATPase activity increased sigmoidally between 0-3 mM ATP, with a Hill coefficient of ~2.8 (Fig. 5F), indicating a positive cooperativity in ATP binding. At ATP concentrations above 3 mM, however, substrate inhibition was observed. Addition of UBH-UBX eliminated this substrate inhibition, increased the apparent V_max/[E]₀ from 2.7 s^-1 to 4.3 s^-1, reduced the Hill coefficient to ~1.5. These findings indicate that the UBH-UBX modules of FAF1 and FAF2 enhances p97’s motor function by increasing catalytic capacity, decreasing ATP cooperativity, and relieving ATP inhibition.

We then assessed substrate unfolding using an mEOS3.2-based p97 unfoldase activity assay^29,61. We purified mEOS3.2 fused to linear di-ubiquitin at the N-terminus and modified it with very long K48-linked ubiquitin chains using gp78RING-Ube2G2 chimera²⁹. Highly polyubiquitinated mEOS3.2 substrates were isolated by size-exclusion chromatography and incubated with p97-UN. This resulted in green fluorescent loss at a rate of 1.0 ± 0.1 arbitrary unit (AU) (Figs. 5G, H, and Supplementary Fig. 6C), consistent with mEOS3.2 unfolding. Adding FAF2^ΔM significantly increased the unfolding rate to 2.1 ± 0.1 AU, a 2-folded enhancement. Deletion of the UBA domain, or both the UBA and UAS domains (Fig. 5G, H), did not affect FAF2’s ability to accelerate unfolding, indicating that these domains are not essential for this activity. In contrast, the FAF2’s UBH mutant (L289N/Y297H/L301N) reduced the unfolding rate to 0.9 ± 0.1 AU, underscoring the role of FAF2’s UBH and its ubiquitin binding function in substrate processing (Fig. 5I, J). Similar results were also observed for FAF2’s UBH-UBX which accelerated the unfolding of the UV-irradiated, backbone cleaved mEOS3.2 (red) by approximately 2-fold, while their UBS mutants did not (Supplementary Fig. 6D, E). Altogether, we conclude that the UBH-UBX enhances p97-mediated unfolding by strengthening the coupling of p97-UN to ubiquitinated substrates, thereby stimulating energy consumption to amplify the mechanical power output of the p97 motor.

UBH-UBX enables p97-UNF-mediated protein extraction from microsome membrane in vitro

To evaluate whether the p97-UN complex can extract proteins from lipid bilayer and whether FAF2 enhances this process, we reconstituted p97-mediated extraction of ubiquitinated CD4 in vitro (Fig. 6A). We used a truncated CD4 construct (mCD4), which retains a single D4 domain in the ER lumen and has a reduced molecular weight⁵⁶. The mCD4 was expressed in FAF2^KO cells to prevent its degradation. Microsomes containing mCD4 were prepared, ubiquitinated using recombinant SCF^βTrCP, and pelleted to remove excess regents. The resuspended microsomes were then incubated with ATP, and either p97-UN alone, or p97-UN supplemented with recombinant FAF2^ΔM or its UBH mutant Y297H/L301N (Fig. 6B). Chaperonin GroEL mutant D87K was also employed to capture unfolded substrate and ubiquitin molecules, thereby preventing their aggregation, as previously described^22,29. Extracted mCD4 in the supernatant was immunoprecipitated and analyzed by western blot.

**Fig. 6: UBH-UBX enables membrane protein extraction from the ER by p97-UNF in vitro.**

Surprisingly, p97-UN alone failed to extract ubiquitinated mCD4 into the supernatant (Fig. 6C, lane 4), even those decorated with very long ubiquitin chains, contrasting with previous findings showing that p97-UN alone can unfold long ubiquitin chain modified soluble proteins, such as GFP/mEOS3.2²⁹ (Fig. 5) and the MCM7 subunit of CMG helicase complex^28,62. This result suggests that, in the absence of FAF2, p97-UN lacks sufficient power output to unfold and extract membrane proteins from the lipid bilayers, regardless of the length of ubiquitin chain. Presumably, compared with unfolding proteins in the cytosol or nucleus (e.g., GFP or MCM7), extracting a substrate from the lipid bilayer is more challenging.

We deleted FAF2’s hairpin-like MA (FAF2^ΔM) for purification and reconstitution of protein extraction from microsomal membrane. First, full-length FAF2 containing the MA domain could not be expressed in E. coli. Second, dissecting FAF2’s function beyond membrane recruitment of p97 requires removal of the MA. Notably, this approach remains physiologically relevant for several reasons. First, FAF2^ΔM/ΔUBA can assemble with p97-UN and the polyubiquitin-modified sfGFP (Fig. 4D, E) and enhancing p97’s unfolding power (Fig. 5). Therefore, deletion of MA seems not impair the ability of p97-UN-FAF2 to engage the polyubiquitinated proteins. Second, treatment with the E1 inhibitor TAK243, which blocks ubiquitination, abolished FAF2’s interaction with p97 in cells (Supplementary Fig. 6F), demonstrating that polyubiquitin chains are the primary factor driving the assembly of p97-UN with FAF2 in vivo. This mechanism mirrors the polyubiquitin-dependent assembly of p97-UN with UbxN7 in cell nucleus⁶³. These findings suggest that FAF2’s MA primarily functions to increase its local concentration at the membrane, preventing its diffusion into the cytosol. Consistent with this, altering the MA sequence does not impair FAF2’s function in protein extraction (Supplementary Fig. 6G), and its role could be compensated by the relatively high concentration of FAF2^ΔM used in our in vitro reconstitution assays. Third, p97 can still be recruited to the ER membrane in FAF2^KO cells, likely through polyubiquitin-decorated substrates or other adaptors in the ER¹⁵. This aligns with findings from Rapoport and colleagues, who showed that Cdc48 recruitment to ubiquitinated Hrd1 primarily depends on the ubiquitin chain on Hrd1, with Ubx2 stabilizing but not essential for this interaction³⁰. Finally, OMM-localized FAF2 has been shown to support ERAD in trans¹⁵, indicating that functional FAF2 does not need to be directly anchored to the ER membrane to facilitate ER substrate extraction. Taken together, these considerations justify the use of FAF2^ΔM in our reconstitution assays.

Strikingly, the addition of FAF2^ΔM (Fig. 6C, D) enabled robust extraction of mCD4, with extracted mCD4 containing short ubiquitin chain of up to three molecules detected in the supernatant (Fig. 6D, arrowhead). In contrast, the UBH mutant Y297H/L301N, which compromises ubiquitin binding, severely impaired FAF2’s function in promoting mCD4 extraction, indicating that UBH-ubiquitin interactions are essential for protein extraction from membrane. Notably, the extracted mCD4 retained N-linked oligosaccharides, which were sensitive to Endo H treatment (Supplementary Fig. 6H), confirming that the N-glycans can be retrotranslocated from the microsomal lumen into the supernatant in our system. Furthermore, purified UBH-UBX domain of FAF2 with p97-UN can also facilitate extraction from membrane of ubiquitinated mCD4, albeit with lower activity (Fig. 6E), consistent with its ability in powering up the ATPase and unfolding activities of p97 (Fig. 5C–J). In the absence of FAF2, IBMPFD-associated p97-A232E mildly increased CD4 extraction in our reconstitution assay (Supplementary Fig. 6I, long exposure, compare lane 5 with lane 3). Addition of FAF2^ΔM markedly enhanced CD4 extraction by p97-A232E or L198W (Supplementary Fig. 6I, J, lane 6 vs lane 5), underscoring FAF2’s essential role in targeting ubiquitinated substrate to p97. Notably, p97-L198W-UNF showed a slight increase in extraction efficiency for ubiquitinated mCD4 compared with p97-WT-UNF (Supplementary Fig. 6J, lane 6 vs lane 4 and 6 K). Together, these findings demonstrate that FAF2 facilitate substrate targeting to p97 and enhances the extraction power of p97-UN, with its UBH domain playing a critical role in this process.

UBH-UBX modulates pexophagy and enables protein extraction from the mitochondrial outer membrane

Beyond the ER, FAF2 also localizes to the peroxisome and outer mitochondrial membrane, where it inhibits basal pexophagy to maintain peroxisome pool^40,41,42 and promotes MAD¹⁵, respectively. However, the mechanisms underlying these functions remain incompletely understood.

To monitor pexophagy, we employed a tandem RFP-GFP reporter bearing a C-terminal SKL sequence for peroxisomal targeting (Fig. 7A). Flow cytometry revealed that, under basal conditions, 3.9 ± 0.2% of cells underwent pexophagy (Fig. 7B, C). FAF2 knockout increased this proportion to 15.5 ± 0.5%. Re-expression of wild-type FAF2 reduced pexophagy to 3.2 ± 0.2%, whereas the FAF2 UBH mutant L289N/Y297H/L301N, which lacks the ubiquitin-binding ability, only partially rescued the phenotype (to 7.0 ± 0.4%). Consistently, wild-type FAF2 fully restored the peroxisomal ABCD3 protein levels in FAF2^KO cells, while FAF2 UBH mutant L289N/Y297H/L301N failed to do so (Fig. 7D, E). Taken together, these findings demonstrate that UBH-ubiquitin binding is critical for suppressing basal autophagy and maintaining peroxisomal abundance.

**Fig. 7: UBH-ubiquitin interaction regulates pexophagy and outer mitochondria membrane associated degradation.**

To test if FAF2’s UBH facilitates MAD, we first examined the turnover of MCL1, a MAD substrate functioning in apoptosis¹⁵. In FAF2^KO cells, expression of FAF2 UBH mutant lacking ubiquitin binding activity delayed MCL1 degradation, extending its half-life from 30 min to over 90 min. Thus, the UBH-ubiquitin interaction promotes efficient MCL1 degradation from OMM (Fig. 7F, G).

We next assessed the degradation of Mitofusin1 (Mfn1), an OMM protein targeted for degradation during PINK1-Parkin-mediated mitophagy. In FAF2^KO cells, re-expression of wild-type FAF2 or a UBA-depleted variant reduced Mfn1 levels by 50%, whereas the FAF2 UBH mutant failed to do so (Fig. 7H, I). These results indicate that FAF2 is required for Mfn1 degradation, and the UBH-ubiquitin binding plays a critical role in this process.

To determine whether FAF2 promotes substrate extraction from mitochondria, we reconstituted this process in vitro (Supplementary Fig. 7A). Mitochondria harboring N-terminal FLAG-tagged PINK1⁶⁴ were isolated and subjected to in vitro ubiquitination with E1, UbcH7, Parkin, and ubiquitin, leading to robust modification of Mfn1 and other OMM proteins such as Mfn2 (Supplementary Fig. 7B). When ubiquitinated mitochondria were incubated with p97-UN alone, low or negligible release of Mfn1 and Mfn2 into the supernatant was detected (Supplementary Fig. S7C–E, lane 8). In contrast, addition FAF2^ΔM enabled prominent extraction of ubiquitinated Mfn1 and Mfn2 into supernatant (Supplementary Fig. 7C–E, lane 10). Deletion of FAF2’s UBA domain did not impair this activity (Supplementary Fig. 7F, lane 8 vs 10), consistent with our in vivo observation that the UBA is dispensable for substrate extraction and degradation from the ER (Fig. 1H). In contrast, the FAF2 UBH mutant Y297H/L301N, which disrupts ubiquitin binding, failed to promote substrate extraction (Supplementary Fig. 7F, lane 12 vs 10), underscoring the essential role of UBH-ubiquitin interactions in membrane protein extraction. Although the overall extraction efficiency was modest, this may reflect Parkin’s assembly of non-K48 ubiquitin linkages (e.g., K63, K6, and K11) on OMM proteins⁶⁵. These results demonstrate that FAF2 directly promotes p97-mediated segregation of MAD substrates from the OMM. The fact that soluble FAF2^ΔM reconstituted this activity suggests that FAF2’s cytosolic UBH-UBX module, in complex with p97-UN, is sufficient to engage polyubiquitin chains on membrane substrates and drive their extraction. This aligns with previous reports that ER-localized FAF2 can facilitate MAD in trans¹⁵.

Discussion

By reconstitution assays and mutational analyses in both cellular and in vitro systems, we demonstrate that the UBH-UBX enhances engagement of ubiquitinated substrates with the p97-UN, thereby stimulating p97 motor’s ATPase activity^22,29. This increase in ATP hydrolysis is transduced into greater mechanical unfolding power, enabling the efficient extraction of ERAD and MAD substrates from lipid bilayers (Fig. 7J). Our findings define the power-enhanced unfoldase complex, p97-UNF, and provide a mechanistic explanation for how protein substrates are efficiently removed from lipid bilayer. Given that FAF2, FAF1, and their orthologs all carry the UBH-UBX module (Supplementary Fig. 7G) and participate in diverse cellular processes across multiple organelles, our results suggest a general mechanism underlying their roles in physiological activities such as cholesterol biosynthesis^5,44, mitochondrial¹⁰ and peroxisomal^40,41,42 homeostasis, apoptosis¹⁵, lipid droplet regulation^48,49,66,67, and chloroplast homeostasis⁴⁷.

We identify the highly conserved ubiquitin-binding residues in UBH and show that UBH-UBX module enhances both the working ATPase activity and unfolding ability of p97-UN. Our research suggests that improved substrate engagement directly enhances p97 motor’s mechanical power. In addition, we reveal the function and mechanism of UBH in the physiological context of ER, peroxisome and mitochondria, where FAF2 localizes. Fujisawa et al. elegantly demonstrated that, in vitro, FAF2 or FAF1’s HD/UBH promotes p97-mediated extraction of MCM7 when modified with short ubiquitin chains, whereas longer chains allow unfolding by p97-UN without cofactors²⁸. However, in vivo, UBXN7/UBXD7 and FAF1, rather than FAF2, appear to be the primary cofactors mediating CMG helicase disassembly^28,62,63. In contrast, our results reveal that FAF2, through its UBH, is indispensable for substrate extraction from various organelles, regardless of ubiquitin chain length, provided that the chain contains three or more ubiquitin units. These findings highlight the context-dependent specialization of p97 cofactors: FAF2 plays a central, non-redundant role in organellar protein quality control, whereas FAF1 and UBXN7 are more critical in nuclear replisome disassembly. Our results also suggest that short ubiquitin chains may not sufficiently stimulate p97-UN’s ATPase and unfolding activities to disassemble the CMG helicase. Future studies should examine whether UBXN7^28,62,63 also enhances p97-UN’s working ATPase and unfolding activities.

Why is FAF2 required for the substrate extraction from membrane-enclosed organelles, even when substrates carry long ubiquitin chains? First, membrane protein substrates present a unique energetic challenge: their lipid-intercalating transmembrane domains form strong hydrophobic interactions with the membrane core, creating a high barrier for unfolding and extraction. Although p97-UN’s ATPase activity increases approximately threefold upon polyubiquitin engagement^22,29,61 (Fig. 5A), this activity-sufficient for unfolding long ubiquitin chain-modified proteins in the cytosol or nucleus-is inadequate for extracting protein substrates from lipid bilayers. The UBH-UBX further enhances the interaction between ubiquitinated substrates and p97-UN, thereby accelerating ATP hydrolysis to generate either higher mechanical force or output power, ultimately promoting efficient protein unfolding and extraction from membrane.

Second, FAF2 is also required for soluble ERAD-L substrates such as NHK. This likely reflects the unique topology of ERAD, where soluble substrates are retrotranslocated by p97 through a thinned lipid bilayer region formed by the Hrd1p-Der1p/Hrd1-Derlin1 complex^26,27,68. Polyubiquitin chain localized near membrane surface or within other large macromolecular assemblies are presumably less accessible to p97-UN than those freely exposed in the cytosol or nucleoplasm. This explains why p97-UN alone can unfold heavily ubiquitinated GFP/mEOS3.2²⁹ (Fig. 5G–J) and MCM7^28,62, yet fails to extract ER-anchored proteins, even those carrying very long ubiquitin chains (Fig. 6C–E, Supplementary Fig. 6H–J). By providing additional ubiquitin binding sites, FAF2’s UBH domain compensates for this limitation, increasing substrate accessibility and enhancing the ATPase and unfolding activities. Therefore, our findings suggest that enhancing substrate targeting, akin to induced proximity, can elevate the power output of the p97 motor. Interestingly, yeast Cdc48-UN alone can extract membrane proteins from proteoliposome^31,32, likely reflecting functional differences between yeast and mammalian UN heterodimers^17,28,33. Whether yeast orthologs such as Ubx2, Ubx3 and UCP10 (Supplementary Fig. 7G) further potentiate Cdc48-mediated membrane protein extraction and degradation in a UBH-dependent manner remains an intriguing question^{31,32,35,36,69}.

Among the five human p97 adaptor proteins that contain both UBA and UBX domains, FAF1 and FAF2 are distinguished by an ~11-nm UBX-adjacent helical domain (UBH)^43,70. This domain is conserved across yeasts (Ubx2, Ubx3 and UCP10), plants (PUX8, PUX10, and PUX13 in Arabidopsis) and other metazoans (Supplementary Fig. 7G). While both UBA^{40,47,48,49,71} and UBH domains bind ubiquitin, our results show that only UBH is indispensable for protein extraction in both cellular and in vitro settings. This finding clarifies prior uncertainties regarding the role of UBA-UBX proteins such as FAF2, FAF1 and PUX10. Although their UBA domains have been proposed to link ubiquitinated substrates to p97/Cdc48, they appear dispensable for substrate degradation. Mutational analyses of UBH in homologs like PUX10, Ubx2, Ubx3, UCP10, and FAF1 could reveal whether defects in UBH impair specific physiological processes, such as lipid droplet turnover, mitochondrial quality control, and chloroplast homeostasis.

We propose that the UFD1L-NPLOC4-FAF2/FAF1 may assemble into a specialized architecture on the cis-side of p97 hexamer, where multiple ubiquitin-binding sites in these co-factors cooperate to capture and stabilize ubiquitin chains, thereby facilitating their unfolding. Alternatively, the UBH may serve to prime polyubiquitin chains for downstream processing by NPLOC4 or p97. For instance, UBH may disrupt the compact configuration of K48-linked chains and bias them toward a more ‘open’ state, potentially favoring unfolding of initiator ubiquitin by NPLOC4 or translocation through the p97 pore. However, proximity to UN and p97-ensured by UBX-mediated binding to p97-appears necessary for this conformational change to translate into efficient substrate recognition and processing. Future cryo-EM studies of p97-UNF bound to polyubiquitinated substrates will be critical to elucidate the mechanism.

We found that UBH binds ubiquitin chains with surprisingly low affinity, as only a small fraction of chains were recovered even with excess UBH. This weak binding alone cannot explain the strong enhancement of ubiquitin targeting to p97-UN (Fig. 4D, E), suggesting additional mechanisms. Although the predicted structure of UBH bound to three ubiquitin molecules provides clue, how UBH engages tri-ubiquitin or longer chains within the p97-UNF complex remains unresolved. In addition, while the reconstitution assay successfully recapitulates the motor function of substrate extraction, it does not address the conductance function within the microsome membrane. CD4 extraction appears independent of Hrd1 and gp78 ubiquitin ligase complexes⁵⁶, raising the possibility of a distinct retrotranslocon. Future studies should aim to identify the putative retrotranslocon and reconstitute the complete extraction process with purified components in proteoliposomes.

Methods

Cell culture and cell lines

HEK293T (ATCC, Cat#CRL-3216), HeLa (ATCC, Cat#CCL-2) and HEK293T-Flp-In cells (Invitrogen, Cat#R75007) were maintained in high-glucose Dulbecco’s Modified Eagle’s medium (DMEM, Thermo Fisher Scientific) supplemented with 10% FBS (GIBCO, Thermo Fisher Scientific) and 1x Penicillin-Streptomycin-Glutamine (GIBCO, Thermo Fisher Scientific). All cells were incubated at 37 °C with 5% CO₂. Expi293F and Sf9 cell lines were kindly provided by Dr. Yu Cao and Dr. Yixiao Zhang, respectively.

The HEK293T-Flp-In cells stably expressing Vpu and CD4 were generated by co-transfecting pOG44 (Invitrogen) and pCDNA5/FRT/TO-Vpu-P2A-CD4. Positive clones were selected using 50 μg/ml hygromycin.

The targeting sequence for FAF2 (5’-tgaggagcgggatctaaccc-3’) was cloned into the LentiCRISPR v2 plasmid (Addgene, Cat#52961). The plasmid was co-transfected into HEK293T cells with the second-generation lentiviral packaging plasmid psPAX2 (Addgene, Cat#52961) and envelope-expressing plasmid pMD2.G (Addgene, Cat#12259). After 48 h, puromycin (1 μg/ml, Thermo Fisher Scientific) was added for positive selection. The selected cells were cultured for 3 days and plated into 96-well plates at a density of 0.4 cells per well to establish monoclonal populations.

Constructs, antibodies and other materials

Vpu- and CD4-derived constructs were modified from published plasmids^50,72. Expression vectors of the SCF complex and UbcH3 have been described⁷³. Constructs encoding ABCG2, TCRα, mTAP2 and ubiquitin have been described^57,74. The coding sequences of p97, FAF2 and FAF1 were amplified from cDNA of HEK293T by RT-PCR and cloned into pCDNA5/FRT/TO-based vector, with FLAG or Myc tags at the N- or C-terminus, as indicated. Truncated constructs of FAF2 (ΔUBA, ΔM, ΔUAS, ΔUBX, ΔHD) were created by PCR-based cloning. The Vpu-P2A-CD4 fusion construct was generated using PCR-based cloning strategies and subcloned into the pCDNA5/FRT/TO vector. The plasmid encoding RFP-GFP-SKL was a gift from Dr. Min Zhuang (ShanghaiTech University). The EGFP coding sequence was subcloned into the pCDNA5/FRT/TO vector. All constructs were verified by Sanger sequencing.

For bacterial expression, truncated FAF2 and HDs were cloned into pET-28a with a N-terminal 14xHis-SUMO tag or 14xHis-2xStrepII tag. The sequence of p97 was cloned into pET-28a with a C-terminal 6xHis or Myc-6xHis tag, while UFD1L and NPLOC4 coding regions were cloned into pET-26a and pET-30a vectors, respectively. The S. cerevisiae gene UBC2 was cloned into pET-28a with an N-terminal 6xHis tag. To generate plasmid encoding Arg-sfGFP, the R-degron sequence²² (synthesized by Sangon Biotech) was cloned into the pET-28a vector with a N-terminal 14xHis-SUMO tag, followed by the insertion of sfGFP at the C-terminus. The sequence of GroEL was amplified from the E. coli BL21 and cloned into pET-28a.The plasmid encoding gp78RING-Ube2g2 chimera²⁹, Ub^G76V-Ub^G76V-mEOS3.2⁷⁵, and S. cerevisiae Ubr1⁷⁶ have been previously described.

Rabbit polyclonal anti-GFP⁷⁷, anti-BAG6⁷⁸, anti-AspTag⁷⁴ and anti-Vpu⁵⁶ antibodies were used as described. Other commercial antibodies and reagents used in this study are listed as in Supplementary Fig. 8. The majority of primary antibodies were used at a 1:1000 dilution, unless otherwise specified by the manufacturer.

Short-interfering RNA (siRNA) for FAF2, UbxD2, UbxD6 and VIMP were bought from GenePharma. The sequences of siRNA were listed below: siFAF2 #1 (5’-GGGACATTGTTTCATTTAT-3’), siFAF2 #2 (5’-CGCCTAGAAAGGGAAGAAA-3’), siUbxD2 #1 (5’-GCAAGTTGGGAAGATGATA-3’), siUbxD2 #2 (5’-TAACTCTCTGCTAGGTCCTTGCTTA-3’), siUbxD6 #1 (5’-CCTCATGGCTTAACTCATT-3’), siUbxD6 #2 (5’-TGGCACAAATAAAGGCTTCACTTTCA-3’), siVIMP #1 (5’-CCACCTATGGCTGGTACAT-3’), siVIMP #2 (5’-GATGAACCCTTAACCCTCGATTCAA-3’).

Cell transfection and RNA interference

Cells were transfected with plasmids using the Lipofectamine 2000 (Thermo Fisher Scientific, Cat#11668-019) according to the manufacturer’s instructions. Typically, 50-100 ng of plasmid was transfected into cells in individual well of a 24-well plate to mimic the expression level of endogenous proteins. To assess transfection efficiency, 20 ng of a plasmid encoding EGFP was co-transfected.

For short-interfering RNA (siRNA)-mediated gene knockdown, cells at ~20% confluency in 24-well plate were transfected with siRNA using Lipofectamine RNAiMAX (Thermo Fisher Scientific, Cat#13778150) according to manufacturer’s instructions. After 24 h, cells were transfected again with the same amount of siRNA to enhance knockdown efficiency. The cells were then cultured for an additional 48–72 h before analysis by immunoblotting.

Protein degradation measurement

The protein turnover was measured using cycloheximide (CHX) chase assay, as previously described⁷⁴. Briefly, cells were seeded in 24-well plate and transfected with plasmids. After 24 h, cells were treated with 100 μg/mL CHX for various time points. Following treatment, cells were harvested, lysed with 1.5x SDS-PAGE sample buffer, and boiled before analysis by immunoblotting.

Microsome preparation

Crude microsomes were isolated from cultured cells as described⁵⁰. Briefly, cells expressing Vpu and CD4 (or mCD4) were washed with cold PBS and collected by pipetting in cold PBS. After centrifugation at 1000 × g for 10 min, cell pellet was resuspended with hypotonic buffer (10 mM HEPES, pH 7.4, 250 mM Sucrose, 2 mM MgCl₂, 1x protease inhibitor cocktail) at 4x the cell volume. The cell suspension was incubated on ice for 5 min and homogenized with ~15 passages through a 25-gauge needle. The homogenate was centrifuged at 3800 × g for 20 min to remove unbroken cells and nuclei. The supernatant was centrifuged again to remove nuclei thoroughly. The post-nuclear supernatant was further centrifuged at 75,000 × g for 1 h at 4 °C using a Beckman TLA100.3 rotor. The pellet (microsomes) was resuspended in homogenization buffer (10 mM HEPES, pH 7.4, 0.25 M sucrose, 1 mM MgCl₂ and 0.5 mM DTT), flash-frozen in liquid nitrogen, and stored at -80 °C.

TUBE pulldown for ubiquitination analysis

TUBE pulldowns were performed as described⁵⁷. Briefly, cells grown in a 6-well plate were treated with inhibitors when indicated and lysed in 1 mL (per well) IP buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, and 1% Triton X-100) supplemented with 1x protease inhibitors cocktail, 2 mM N-ethylmaleimide (NEM) and 10 mM iodoacetamide (IAM). The lysate was incubated on ice for 20 minutes and then centrifuged at 20,000 × g for 15 min to remove nuclear and insoluble debris. The cleared supernatant was incubated with agarose-conjugated tandem ubiquitin binding entity (TUBE) on an end-over-end rotator at 4 °C for 2 h. After three washes, bound proteins were eluted with 50 μL of 1.5x SDS-PAGE sample buffer and analyzed by immunoblotting.

Native co-immunoprecipitations

Cells from individual wells of 6-well plates were lysed in 1 mL co-IP buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.5% Triton X-100) containing protease inhibitors cocktail and incubated on ice for 20 min. Cell debris was removed by centrifugation at 20,000 x g for 10 min. The supernatant was incubated with antibody-conjugated agarose beads or Strep-Tactin Sepharose on an end-over-end rotator at 4 °C. After 3 washes with co-IP buffer, bound proteins were eluted with 50 μL of 1.5x SDS-PAGE sample buffer and analyzed by immunoblotting.

Purification of ubiquitinated CD4 from ER-derived microsome

Ubiquitinated CD4 was purified using a tandem-affinity purification approach. HEK293T cells seeded in three 15-cm dishes were transfected with plasmids encoding CD4-FLAG and 2xStrepⅡ-Ub (cells transfected empty vector were used as control). After 24 h, cells were treated with 20 μM CB5083 for 3 h, washed twice, and harvested in cold PBS. The cell pellet was resuspended with 2 mL homogenization buffer (30 mM HEPES pH 7.4, 150 mM NaCl, 2 mM MgCl₂ and 1x protease inhibitor cocktail). Cells were disrupted using a dounce homogenizer (KONTES Glass Co.) with 100 strokes. The total cell lysate was centrifuged twice at 3700 × g for 20 min at 4 °C to remove unbroken cells and nuclei. The post-nuclear supernatant was further centrifuged at 100,000 × g (Beckman TLA-100.3 rotor) at 4 °C for 1 h. The resulting pellet was resuspended in salt buffer (20 mM HEPES pH7.4, 150 mM NaCl, 2 mM MgCl₂) and lysed by incubation with 1% DDM for 30 min at 4 °C. Insoluble materials were removed by centrifugation at 20,000 × g for 15 min. The supernatant was diluted with an equal volume of salt buffer (20 mM HEPES pH7.4, 150 mM NaCl, 2 mM MgCl₂), reducing the final concentration of DDM to 0.5%.

The diluted lysate was incubated with 25 μL packed anti-FLAG affinity agarose for 2 h at 4 °C with an end-over-end rotation. The flow-through was collected for analysis, and the beads were washed extensively with salt buffer containing 0.2% DDM. Bound proteins were eluted with 1 mL of 0.5 mg/mL 3xFLAG peptides dissolved in salt buffer with 0.2% DDM. The FLAG-eluted fraction was further purified by incubation with 25 μL Strep-Tacin Sepharose for 2 h at 4 °C with an end-over-end rotation. The flow-through was collected and bound materials was eluted with 200 μL salt buffer containing 0.9 mM biotin and 1% Triton X-100.

For experiments shown in Supplementary Fig. 1H, HEK293T cells cultured in 15-cm dishes were transfected with plasmid encoding CD4-FLAG and p97-2xStrepⅡ. Upon reaching 90% confluency, cells were treated with 20 μM CB5083 for 3 h. CD4 and p97 were purified following the same procedure described above.

Sample preparation for mass spectrometry analysis

Proteins from the final eluates in Supplementary Fig. 1D were precipitated on ice by adding four volumes of cold acetone. The resulting pellets were collected by centrifugation at 14,000 g for 15 min and redissolved in buffer containing 8 M urea and 100 mM Tris-HCl (pH 8.5). Disulfide bonds were reduced with 5 mM Tris(2-carboxyethyl) phosphine (TCEP) for 20 min, followed by alkylation of cysteine residues with 10 mM iodoacetamide (IAA) for 15 min at room temperature in the dark. The urea concentration was then diluted to 2 M using 100 mM Tris-HCl (pH 8.5) supplemented with 1 mM CaCl₂. Proteins were digested overnight at 37 °C with trypsin (Cat#V5111, Promega) using an enzyme to substrate ratio of 1:50 (w/w). Tryptic digests were acidified with 1% formic acid (FA), centrifuged (18,000 g, 20 min), desalted with C18 StageTips, dried, and stored at -80 °C.

LC-MS/MS analysis

The peptide mixture was analyzed using an on-line EASY-nL-LC 1000 coupled with an Orbitrap Q-Exactive HF mass spectrometer. The sample was loaded directly onto a 15 cm home-made capillary column (C18-AQ, 1.9 μm resin, 100 μm I.D.). A gradient was developed using mobile phase A (0.1% FA, 2% ACN, and 98% H₂O) and B (0.1% FA, 2% H₂O, and 98% ACN), with a 60-min gradient set as 8% B at 4 min, 22% B at 45 min, 30% B at 53 min, 92% B at 57 min and 95% B at 60 min. A static flow rate of 300 nL/ min was used. Mass spectra were acquired in a data-dependent mode with one full scan in the Orbitrap (m/z: 350-1,500; resolution: 60,000; AGC target value: 3,000,000; maximal injection time: 20 ms; exclusion duration: 30 s), followed by MS2 scans in the linear trap (resolution: 15,000; AGC target value: 50,000; maximal injection time: 25 ms; isolation window: 1.6 m/z).

LC-MS/MS data processing

The MS/MS raw spectra were processed using MaxQuant software (version 1.6.0.1). The MS data were searched against the SwissProt human protein database (downloaded in October 2018, 20,409 entries) and the built-in contaminant protein list. Trypsin was set as the enzyme, and the maximum missed cleavage was set to 2. The precursor mass tolerance and the fragment mass tolerance were set to 20 ppm and 0.1 Da, respectively. The cysteine carbamidomethyl (delta mass = 57.02) served as a static modification, while the methionine oxidation (delta mass = 15.99) as variable modifications. The false discovery rates at the peptide spectral match level and the protein level were both controlled below 1%. Only unique peptides and razor peptides were used for quantification, and the minimum ratio count for protein identification was 1. After filtering out the contaminants, a total of 396 proteins were identified. Differentially expressed proteins were defined as those with a fold change greater than 2.

Fractionation of cytosolic and membrane proteins

HEK293T cells were semi-permeabilized by 0.015% digitonin (CalBioChem, Cat#300410) as described⁵⁷. Briefly, cells grown in a 24-well plate were washed once with cold PBS and incubated on ice for 12 minutes in a buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1x protease inhibitor cocktail and 0.015% digitonin. The released cytosolic fraction was collected and centrifuged at 20,000 x g for 10 minutes. The membrane fraction was extracted using co-IP buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.5% Triton X-100). The cytosolic and membrane protein fractions were reserved for downstream analysis.

Protein expression and purification from bacteria

Expression and purification of Ube2D2⁷⁹ and gp78(RING)-Ube2g2 chimera²⁹ were performed as previously described. The Cul1/Rbx1 and βTrCP1/Skp1 were expressed, purified and assembled according to established protocols^79,80.

Non-tagged GroEL-D87K, 6xHis-tagged Ube1, 6xHis-tagged Ube2D2, 6x His-tagged ubiquitin, 6xHis-tagged UbcH7, 12xHis-tagged scUlp1, 6xHis-MBP-tagged TUBE, 14xHis-StrepII-tagged HDs, 14xHis-SUMO^eu1-tagged HDs, 14xHis-SMT3-tagged Arg-sfGFP, Ub^G76V-Ub^G76V-mEos3.2, and 14xHis-SMT3-tagged Parkin were expressed in E. coli BL21 (DE3). The p97 with C-terminal Myc-6xHis, 14xHis-tagged FAF2^ΔM, and 14xHis-tagged FAF2^ΔM or FAF1 with additional mutations or truncations were expressed in E. coli Rosetta 2 (DE3). Protein expression was induced with 0.5 mM IPTG at 16 °C for 18 h.

Bacterial pellets were resuspended in bacterial lysis buffer (50 mM Tris-HCl, pH 7.5, 300 mM NaCl, 1 mM β-mercaptoethanol, 5% glycerol, and 1 mM PMSF), and lysed using a French Press (JN-02C, JNBIO). Cell lysates from 1 L of E. coli culture was clarified by centrifugation at 14,000 × g for 30 min in a JA-25.50 Rotor (Beckman) and purified using 1 ml Ni-NTA beads (Smart-Lifesciences). Proteins were eluted with wash buffer (same as bacterial lysis buffer without PMSF) containing 300-500 mM imidazole and dialyzed against storage buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5% glycerol). For proteins with a 14xHis-SMT3 (SUMO) tag, the tag was cleaved using scUlp1 (SUMO protease) at 4 °C for 1 h and removed via a HiTrap Q HP column. The 6xHis-tag in ubiquitin was cleaved by 6xHis-tagged TEV protease and removed by Ni-NTA beads to purify non-tagged ubiquitin.

To purify the UFD1L-NPLOC4 complex, bacterial pellets mixed at a 1:3 ratio of 6xHis-UFD1L to NPLOC4 were processed as described above.

GST-tagged UBX were expressed in E. coli Rosetta 2 (DE3) from pGEX-6p1 vector. Cell lysate was incubated with Glutathione Beads (Smart-Lifesciences) at 4 °C for 2 h. The resin was washed three times with wash buffer and eluted twice with 50 mM reduced glutathione at RT for 10 min. Elution fractions were pooled and subsequently passed through a PD-10 desalting column equilibrated with storage buffer to remove glutathione.

scUbr1 was expressed in Saccharomyces cerevisiae BY4741 grown in YPD medium. The protein was purified by anti-FLAG affinity chromatography and eluted with wash buffer supplemented with 3xFLAG peptides. Eluates were loaded onto a Superdex 200 column to remove excess 3xFLAG peptides. The pooled fractions were snap-frozen in liquid nitrogen, and stored at -80 °C.

Purification of p97 from mammalian cells

For the reconstitution assays in Figs. 5, 6, Supplementary Fig. 6 and 7, p97 purified from mammalian cells was utilized due to its superior performance compared to bacterially expressed p97⁶¹. The p97 with a 2xStrepII tag at the C-terminal was expressed in Expi293F and purified using Strep-Tactin Sepharose beads. Expi293F were transfected with plasmids encoding p97-2xStrepII, p97^L198W-2xStrepII, or p97^A232E-2xStrepII using EZ-trans (Life-iLab, Cat#AC04L082) at the density of approximately 2 × 10⁶ mL^-1 and cultured at 37 °C. After 12–16 h of transfection, the cells were treated with 10 mM sodium butyrate and cultured for an additional 48 h at 30 °C. Cells were collected by centrifugation at 1000 × g for 10 min and washed with ice-cold PBS. The cell pellet was resuspended in lysis buffer (30 mM HEPES, pH7.5, 150 mM KCl, 10 mM MgCl₂, 5% glycerol) plus 1x protease inhibitor cocktail, 1 mM DTT and 1 mM ATP. The cells were lysed by ultra-sonication on ice and centrifugated at 20,000 × g for 15 min to remove unbroken cells and debris. The clarified lysate was incubated with Strep-Tactin Sepharose for 2 h with gentle rotation. The beads were washed 3 times with lysis buffer supplement with 0.1 mM ATP. The bound p97 protein was eluted for three times with lysis buffer containing 1 mM ATP and 2 mM Biotin. The eluted fractions were concentrated using a 4 mL Amicon Ultra Centrifugal Filter (Merck, UFC8100) and loaded onto a 24 mL Superdex 200 column pre-equilibrated with lysis buffer containing 1 mM ATP. Fractions containing p97 were pooled and re-concentrated for further use.

Analytic size exclusion chromatography and MALS

Recombinant p97, UFD1L-NPLOC4, and FAF2^ΔM mixed in a molar ratio of 3:1:1 was loaded onto a Superose 6 Increase 10/300 GL column (GE Healthcare) pre-equilibrated with running buffer containing 30 mM HEPES (pH 7.5), 150 mM NaCl, and 5 mM MgCl₂. The eluted samples were analyzed by SDS-PAGE to assess complex formation and protein composition.

2xStrepII-tagged HD in buffer (30 mM HEPES, pH 7.5, 150 mM NaCl) was injected onto a Superdex 75 size exclusion column connected in-line with a DAWN HELEOS-II MALS detector and an Optilab T-rEX differential refractive index detector at room temperature.

AlphaFold Prediction

Structural prediction presented in Fig. 3A was generated in AlphaFolder Server (available at https://alphafoldserver.com/). For this analysis, three individual ubiquitin sequences and UBH sequence were inputted into the server, and the output file was visualized in ChimeraX.

Synthesis of ubiquitin chains and analysis of their binding with α-helical domains

Free ubiquitin chains were synthesized in a reaction mixture containing 0.5 μM Ube1, 10 μM gp78(RING)-Ube2g2 chimera, 20 μM His-tagged wild-type or I44A ubiquitin (or non-tagged ubiquitin in Fig. 2E and Supplementary Fig. 4B), and 5 mM ATP in the ubiquitination buffer (30 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM DTT, 5 mM MgCl₂). The reaction was carried out at 25 °C overnight. During the first 4.5 h, 5 μM ubiquitin was added incrementally every 1.5 h. For experiments shown in Figs. 2B, E, 3B, Supplementary Fig. 4B-E, His-tagged or no-tagged ubiquitin was added immediately before downstream analysis to generate a mixture of both mono-ubiquitin and ubiquitin chains of two or more molecules. The mixture was centrifuged at 20,000 × g at 4 °C for 10 min to remove insoluble proteins.

To assess the interaction between ubiquitin and α-helical domains (HD), recombinant 2xStrepII-tagged or 14xHis-SUMO-tagged HDs from various species were immobilized on Strep-Tactin Sepharose or TALON resin, respectively. The synthesized ubiquitin chains were diluted with DDM washing buffer (50 mM Tris, pH 7.5, 150 mM NaCl and 0.3% DDM) and incubated with immobilized HDs at 4 °C for 2 h. After extensive washing with DDM washing buffer to eliminate non-specific interactions, the bound materials were eluted using 2x SDS-PAGE sample buffer and analyzed by immunoblotting.

Non-tagged ubiquitin chains used for ATPase assays were further purified using a HiTrap^TM Heparin HP column according to established protocols⁸¹.

Interaction between p97-UN and FAF2

To examine interactions between p97-UN and either UBX domain, FAF2^ΔUBAΔM, or its mutant FAF2^{ΔUBAΔM/Y297H/L301N}, GST-tagged UBX, FLAG-tagged FAF2^ΔUBAΔM (or FAF2^{ΔUBAΔM/Y297H/L301N}) were incubated with either p97 alone or the p97-UN complex at 4 °C for 20 min and diluted with Triton X-100 washing buffer (50 mM Tris, pH 7.5, 150 mM NaCl, and 0.1% Triton X-100) before incubation with glutathione Sepharose (Smart-Lifesciences) or anti-FLAG agarose at 4 °C for 2 h. The beads were washed three times and bound proteins were eluted with 1.5x SDS-PAGE sample buffer before analysis by immunoblotting and Coomassie blue staining.

Ubiquitination of Arg-sfGFP and analysis of its binding to p97-UN-FAF2

Ubiquitination of Arg-sfGFP was conducted by incubating 0.5 μM Ube1, 0.5 μM Ubc2, 20 nM scUbr1, 1 μM Arg-sfGFP, 5 μM His-Ubiquitin, and 5 mM ATP in a reaction buffer (30 mM Tris-HCl, pH 7.5, 100 mM NaCl, 1 mM DTT, 5 mM MgCl₂) for 12 minutes at 30 °C. The samples were centrifuged at 20,000 × g for 10 minutes to remove insoluble proteins.

The p97 alone or p97 complexed with UFD1L-NPLOC4, FAF2 lacking UBA and MA domains (FAF2^ΔUBAΔM), or both was immobilized on anti-Myc agarose beads (via Myc-tagged p97). Typically, purified Myc-tagged p97, UFD1L-NPLOC4, and FAF2^ΔUBAΔM were mixed at a molar ration of 6(monomer):3:4 at 4 °C for 15 min before incubation with anti-Myc beads for 2 h in cold room. The beads with bound proteins were incubated with ubiquitin chains at 4 °C for 1 h with shaking. The beads were then washed three times with Triton X-100 washing buffer (50 mM Tris, pH 7.5, 150 mM NaCl, and 0.1% Triton X-100) and bound proteins were eluted with 2x SDS-PAGE sample buffer for analysis by immunoblotting.

ATPase assays

Briefly, proteins were diluted in the assay buffer (30 mM HEPES, pH 7.4, 60 mM KCl, 10 mM MgCl₂ and 0.6 mM TCEP), preincubated at room temperature for 10 min, and then cooled on ice for 5 min. The reaction was initiated by adding a mixture containing 1 mM ATP, 1 mM NADH, 7.5 mM phosphoenolpyruvate, 16 units/mL pyruvate kinase and 22 units/mL lactic dehydrogenase. A₃₄₀ was monitored every 20 s for the following 1.5 h. The final concentrations of proteins in each reaction were 100 nM p97 hexamer (50 nM in Fig. 5C), 130 nM UN and 200 nM FAF2^ΔM (or truncations of FAF1 or FAF2).

The ATPase velocities were calculated using the extinction coefficient of NADH 6220 M^-1cm^-1.

Steady-state ATPase kinetics of p97-UN and p97-UNF

The Hill coefficient (\({n}_{{{{\rm{Hill}}}}}\)) for p97 ATPase was determined from ATP concentration-dependent reaction rates. The fractional saturation Y was calculated as \(Y=\frac{V}{{V}_{\max }}\), where \({V}_{\max }\) is the max observed rate. Hill coefficient and \({[{ATP}]}_{0.5}\) values were obtained by fitting data to the Hill equation \(Y=\frac{{[{{{\rm{ATP}}}}]}^{{n}_{{{{\rm{Hill}}}}}}}{({[{{{\rm{ATP}}}}]}_{0.5}^{{n}_{{{{\rm{Hill}}}}}}+{[{{{\rm{ATP}}}}]}^{{n}_{{{{\rm{Hill}}}}}})}\) using nonlinear regression analysis in GraphPad Prism (Fig. 5E, F).

Polyubiquitylated substrate synthesis and unfolding assays

Ub^G76V-Ub^G76V-mEOS3.2 was photo-converted to Ex510/Em590 by irradiation with a UV-lamp for 2 h at 4 °C, poly-ubiquitinated by gp78RING-Ube2G2 chimera, and isolated by size-exclusion chromatography as previously reported²⁹. The non-irradiated mEOS3.2 was poly-ubiquitinated using the same method. Typically, the unfolding of highly ubiquitinated mEOS3.2 was performed in a reaction containing 430 nM p97 hexamer, 560 nM UN, 830 nM FAF2, 1 μM GroEL-D87K, 2x energy regenerating system [20x ERS: 20 mM ATP, 20 mM GTP, 0.8 mg/ml creatine kinase, and 200 mM creatine phosphate] in the assay buffer. Two types of control reactions were used: those lacking ERS (no ATP) or those containing Ub^G76V-Ub^G76V-mEOS3.2 that was not further polyubiquitination by the E2-E3 chimera (no ubiquitin). The unfolding of irradiated or non-irradiated mEOS3.2 was measured every 15 s at 37 °C for 32 min. Data were normalized and the unfolding rate was determined from the time window indicated in figure legends.

Reconstitution of mCD4 extraction from microsome

Reconstitution of mCD4 extraction from microsomes was performed using crude microsomes isolated from FAF2 knockout cells expressing Vpu and FLAG-tagged mCD4 or its mutant mCD4-K453. Microsomes were prepared as described earlier, and ubiquitination was performed using one-half of the total microsome isolated from a single 10-cm dish in a reaction mixture containing 0.4 μM Ube1, 0.2 μM Ube2D2, 5 μM Cdc34, 0.27 μM SCF^βTrCP, 22 μM His-Ub, 2 mM ATP and 1 mM DTT, in 30 mM HEPES, pH 7.5, 150 mM NaCl, and 2 mM MgCl₂. The reaction proceeded at 25 °C for 12 minutes, followed by centrifugation at 100,000 × g for 45 min over a sucrose cushion (0.5 M sucrose, 30 mM HEPES, pH 7.5, 75 mM KCl and 2 mM MgCl₂). The microsome pellet was resuspended with 250 μL of extraction buffer (30 mM HEPES, pH 7.4, 60 mM KCl, 10 mM MgCl₂ and 0.6 mM TCEP) for the following extraction assay. As a control, the microsome pellet was directly lyzed in RIPA buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 1% Triton X-100 and 0.5% sodium deoxycholate) to evaluate total ubiquitinated mCD4.

The extraction of mCD4 was performed in a 75 μL reaction containing 500 nM p97 hexamer, 1 μM UFD1L-NPLOC4 complex, 2 μM FAF2^ΔM or its HD mutant (Y297H/L301N), 2 μM GroEL-D87K, 2 mM ATP in the extraction buffer. After incubation at 37 °C for 40 min, the mixture was centrifuged at 100,000 × g for 45 min to separate supernatant and pellet fractions. The supernatant was diluted with RIPA buffer before incubation with anti-FLAG agarose for 2 h at 4 °C. The recovered FLAG-tagged mCD4 was analyzed by immunoblotting. An aliquot of the total extraction reaction was analyzed by Coomassie staining and immunoblotting to assess protein levels.

FACS-based analysis

HeLa cells stably expressing RFP-GFP-SKL were washed twice with PBS and collected with FACS buffer (2% BSA in PBS). The cell suspension was filtered through 35 μm nylon meshes and analyzed using the BD LSRFortessa. The data were collected with BD FACSDiva and analyzed using FlowJo v10.8.

Mitochondria preparation

Mitochondria were prepared from HEK293T cells expressing 3xFLAG-tagged PINK1, which were grown to approximately 90% confluency in a 10-cm dish. The cells were washed with cold PBS, collected by centrifugation at 800 × g for 5 minutes at 4 °C, and resuspended in 1 mL of cold NHS buffer (50 mM HEPES, pH 7.4, 150 mM NaCl, 250 mM sucrose) supplemented with 1x protease inhibitor cocktail. The resuspended cells were homogenized by 20 passages through a 25-gauge needle and the resulting homogenate was clarified by centrifugation at 1000 × g for 5 minutes at 4 °C. This step was repeated twice to remove nuclei and unbroken cells. The crude mitochondria were isolated by centrifugation of the post-nuclear supernatant at 13,000 × g for 10 min. For further purification, the crude mitochondria pellet was washed once with homogenization buffer (50 mM HEPES, pH 7.4, 1 mM EDTA and 250 mM sucrose), resuspended, and centrifuged again at 13,000 x g for 10 min over a sucrose cushion containing 50 mM HEPES, pH 7.4, 1 mM EDTA and 500 mM sucrose. The enriched mitochondria were used for downstream applications.

In vitro ubiquitination and extraction of Mfn1/Mfn2 from isolated mitochondria

The ubiquitination of Mfn1 and Mfn2 was performed in a 120 μL reaction mixture containing 0.2 μM Ube1, 0.4 μM UbcH7, 2 μM Parkin, 5 mM ATP, 10 μM ubiquitin (wild-type or K63R mutant), along with mitochondria from a single 10-cm dish. The mitochondrial assay buffer comprised 30 mM HEPES, pH 7.4, 50 mM NaCl, 5 mM MgCl₂, 0.25 mM DTT. The reactions were incubated at 26 °C for 15 min, after which the ubiquitinated products were used for extraction assays. Of note, the use of K63R ubiquitin enhanced the generation of K48-linked ubiquitin chains, as Parkin is capable of catalyzing ubiquitin chains with various lysine linkages; however, wild-type ubiquitin yields similar results, albeit with a slightly lower extraction efficiency.

The extraction assays were performed in a 40 μL reaction containing 500 nM p97 hexamer, 1 μM UFD1L-NPLOC4 complex, 2 μM FAF2^ΔM (or its variants, FAF2^ΔUBAΔM and FAF2^{ΔUBAΔM/Y297H/L301N}), 2 μM GroEL-D87K, 2 mM ATP in extraction buffer consisting of 30 mM HEPES, pH 7.4, 60 mM KCl, 10 mM MgCl₂ and 0.6 mM TCEP. The reaction was incubated at 37 °C for 30 min, followed by centrifugation at 13,000 × g for 10 minutes. The resulting mitochondria pellet (P) and supernatant (S) fractions were analyzed by immunoblotting.

Quantification and statistics

Western blot band intensities were quantified by densitometry using ImageJ. Student t-test was applied for two-group comparisons, and ANOVA was performed for multiple-group analyses using GraphPad Prism10.1 or Microsoft excel. Error bars represent the ± standard error of mean ( ± SEM).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data are available within the main article and Supplementary Information files. All experiments were independently repeated at least three times or more with similar results. Original western blot source data underlying the figures are provided in the Source Data file. Additional data or materials related to this paper are available upon request from the corresponding author. Source data are provided with this paper.

References

Christianson, J. C. & Carvalho, P. Order through destruction: how ER-associated protein degradation contributes to organelle homeostasis. EMBO J. 41, e109845 (2022).
Article CAS PubMed PubMed Central Google Scholar
Christianson, J. C., Jarosch, E. & Sommer, T. Mechanisms of substrate processing during ER-associated protein degradation. Nat. Rev. Mol. Cell Biol. 24, 777–796 (2023).
Krshnan, L., van de Weijer, M. L. & Carvalho, P. Endoplasmic Reticulum-Associated Protein Degradation. Cold Spring Harb. Perspect Biol. 14, a041247 (2022).
van den Boomen, D. J. H., Volkmar, N. & Lehner, P. J. Ubiquitin-mediated regulation of sterol homeostasis. Curr. Opin. Cell Biol. 65, 103–111 (2020).
Article PubMed Google Scholar
Schumacher, M. M. & DeBose-Boyd, R. A. Posttranslational regulation of HMG CoA reductase, the rate-limiting enzyme in synthesis of cholesterol. Annu Rev. Biochem 90, 659–679 (2021).
Article CAS PubMed Google Scholar
Wu, S. A., Li, Z. J. & Qi, L. Endoplasmic reticulum (ER) protein degradation by ER-associated degradation and ER-phagy. Trends Cell Biol. 35, 576–591 (2025).
Needham, P. G., Guerriero, C. J. & Brodsky, J. L. Chaperoning endoplasmic reticulum-associated degradation (ERAD) and protein conformational diseases. Cold Spring Harb Perspect Biol 11, a033928 (2019).
Karbowski, M. & Youle, R. J. Regulating mitochondrial outer membrane proteins by ubiquitination and proteasomal degradation. Curr. Opin. Cell Biol. 23, 476–482 (2011).
Article CAS PubMed PubMed Central Google Scholar
Zheng, J., Li, L. & Jiang, H. Molecular pathways of mitochondrial outer membrane protein degradation. Biochem Soc. Trans. 47, 1437–1447 (2019).
Article CAS PubMed Google Scholar
Martensson, C. U. et al. Mitochondrial protein translocation-associated degradation. Nature 569, 679–683 (2019).
Article ADS CAS PubMed Google Scholar
Cao, Y. et al. A mitochondrial SCF-FBXL4 ubiquitin E3 ligase complex degrades BNIP3 and NIX to restrain mitophagy and prevent mitochondrial disease. EMBO J. 42, e113033 (2023).
Article CAS PubMed PubMed Central Google Scholar
Pickles, S., Vigie, P. & Youle, R. J. Mitophagy and quality control mechanisms in mitochondrial. Maint. Curr. Biol. 28, R170–R185 (2018).
Article CAS Google Scholar
Tanaka, A. et al. Proteasome and p97 mediate mitophagy and degradation of mitofusins induced by Parkin. J. Cell Biol. 191, 1367–1380 (2010).
Article CAS PubMed PubMed Central Google Scholar
Sun, Y. et al. A mitophagy sensor PPTC7 controls BNIP3 and NIX degradation to regulate mitochondrial mass. Mol. Cell 84, 327–344.e9 (2024).
Article CAS PubMed Google Scholar
Zheng, J., Cao, Y., Yang, J. & Jiang, H. UBXD8 mediates mitochondria-associated degradation to restrain apoptosis and mitophagy. EMBO Rep. 23, e54859 (2022).
Article CAS PubMed PubMed Central Google Scholar
van den Boom, J. & Meyer, H. VCP/p97-mediated unfolding as a principle in protein homeostasis and signaling. Mol. Cell 69, 182–194 (2018).
Article PubMed Google Scholar
Tsuchiya, H. et al. In vivo ubiquitin linkage-type analysis reveals that the Cdc48-Rad23/Dsk2 axis contributes to K48-linked chain specificity of the proteasome. Mol. Cell 66, 488–502 e7 (2017).
Article CAS PubMed Google Scholar
Ye, Y., Meyer, H. H. & Rapoport, T. A. Function of the p97-Ufd1-Npl4 complex in retrotranslocation from the ER to the cytosol: dual recognition of nonubiquitinated polypeptide segments and polyubiquitin chains. J. Cell Biol. 162, 71–84 (2003).
Article CAS PubMed PubMed Central Google Scholar
Twomey, E. C. et al. Substrate processing by the Cdc48 ATPase complex is initiated by ubiquitin unfolding. Science 365, eaax1033 (2019).
Ji, Z. et al. Translocation of polyubiquitinated protein substrates by the hexameric Cdc48 ATPase. Mol. Cell 82, 570–584.e8 (2022).
Article CAS PubMed Google Scholar
Williams, C., Dong, K. C., Arkinson, C. & Martin, A. The Ufd1 cofactor determines the linkage specificity of polyubiquitin chain engagement by the AAA+ ATPase Cdc48. Mol. Cell 83, 759–769 (2023).
Bodnar, N. O. & Rapoport, T. A. Molecular mechanism of substrate processing by the Cdc48 ATPase complex. Cell 169, 722–735.e9 (2017).
Article CAS PubMed PubMed Central Google Scholar
Marinko, J. T. et al. Folding and misfolding of human membrane proteins in health and disease: from single molecules to cellular proteostasis. Chem. Rev. 119, 5537–5606 (2019).
Article CAS PubMed PubMed Central Google Scholar
Min, D., Jefferson, R. E., Bowie, J. U. & Yoon, T. Y. Mapping the energy landscape for second-stage folding of a single membrane protein. Nat. Chem. Biol. 11, 981–987 (2015).
Article CAS PubMed PubMed Central Google Scholar
Chang, Y. C. & Bowie, J. U. Measuring membrane protein stability under native conditions. Proc. Natl. Acad. Sci. USA. 111, 219–224 (2014).
Article ADS CAS PubMed Google Scholar
Wu, X. & Rapoport, T. A. Translocation of Proteins through a Distorted Lipid Bilayer. Trends Cell Biol. 31, 473–484 (2021).
Article CAS PubMed PubMed Central Google Scholar
Wu, X. et al. Structural basis of ER-associated protein degradation mediated by the Hrd1 ubiquitin ligase complex. Science 368, eaaz2449 (2020).
Fujisawa, R., Polo Rivera, C. & Labib, K. P. M. Multiple UBX proteins reduce the ubiquitin threshold of the mammalian p97-UFD1-NPL4 unfoldase. Elife 11, e76763 (2022).
Blythe, E. E., Olson, K. C., Chau, V. & Deshaies, R. J. Ubiquitin- and ATP-dependent unfoldase activity of P97/VCP*NPLOC4*UFD1L is enhanced by a mutation that causes multisystem proteinopathy. Proc. Natl Acad. Sci. USA 114, E4380–E4388 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Stein, A., Ruggiano, A., Carvalho, P. & Rapoport, T. A. Key steps in ERAD of luminal ER proteins reconstituted with purified components. Cell 158, 1375–1388 (2014).
Article CAS PubMed PubMed Central Google Scholar
Schmidt, C. C., Vasic, V. & Stein, A. Doa10 is a membrane protein retrotranslocase in ER-associated protein degradation. Elife 9, e56945 (2020).
Natarajan, N., Foresti, O., Wendrich, K., Stein, A. & Carvalho, P. Quality control of protein complex assembly by a transmembrane recognition factor. Mol. Cell 77, 108–119.e9 (2020).
Article CAS PubMed PubMed Central Google Scholar
Meyer, H. H., Wang, Y. & Warren, G. Direct binding of ubiquitin conjugates by the mammalian p97 adaptor complexes, p47 and Ufd1-Npl4. EMBO J. 21, 5645–5652 (2002).
Article CAS PubMed PubMed Central Google Scholar
Lowe, M., Kalacheva, M., Boersma, A. J. & Kedrov, A. The more the merrier: effects of macromolecular crowding on the structure and dynamics of biological membranes. FEBS J. 287, 5039–5067 (2020).
Article PubMed Google Scholar
Neuber, O., Jarosch, E., Volkwein, C., Walter, J. & Sommer, T. Ubx2 links the Cdc48 complex to ER-associated protein degradation. Nat. Cell Biol. 7, 993–998 (2005).
Article CAS PubMed Google Scholar
Schuberth, C. & Buchberger, A. Membrane-bound Ubx2 recruits Cdc48 to ubiquitin ligases and their substrates to ensure efficient ER-associated protein degradation. Nat. Cell Biol. 7, 999–1006 (2005).
Article CAS PubMed Google Scholar
Hanzelmann, P. & Schindelin, H. The interplay of cofactor interactions and post-translational modifications in the regulation of the AAA+ ATPase p97. Front Mol. Biosci. 4, 21 (2017).
Article PubMed PubMed Central Google Scholar
Stach, L. & Freemont, P. S. The AAA+ ATPase p97, a cellular multitool. Biochem J. 474, 2953–2976 (2017).
Article CAS PubMed Google Scholar
Olzmann, J. A., Richter, C. M. & Kopito, R. R. Spatial regulation of UBXD8 and p97/VCP controls ATGL-mediated lipid droplet turnover. Proc. Natl Acad. Sci. USA 110, 1345–1350 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Koyano, F. et al. AAA+ ATPase chaperone p97/VCP(FAF2) governs basal pexophagy. Nat. Commun. 15, 9347 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Kim, C. et al. FAF2 is a bifunctional regulator of peroxisomal homeostasis and saturated lipid responses. Sci. Adv. 11, eadu9104 (2025).
Article ADS CAS PubMed PubMed Central Google Scholar
Montes, I. D. et al. The p97 ATPase and its adaptor UBXD8 maintain peroxisome pools by preventing pexophagy. J. Cell Biol. 224, e202409024 (2025).
Braxton, J. R. & Southworth, D. R. Structural insights of the p97/VCP AAA+ ATPase: How adapter interactions coordinate diverse cellular functionality. J. Biol. Chem. 299, 105182 (2023).
Article CAS PubMed PubMed Central Google Scholar
Loregger, A. et al. Haploid mammalian genetic screen identifies UBXD8 as a Key determinant of HMGCR degradation and cholesterol biosynthesis. Arterioscler Thromb. Vasc. Biol. 37, 2064–2074 (2017).
Article CAS PubMed PubMed Central Google Scholar
Gwon, Y. et al. Ubiquitination of G3BP1 mediates stress granule disassembly in a context-specific manner. Science 372, eabf6548 (2021).
Article CAS PubMed PubMed Central Google Scholar
Schrul, B. & Kopito, R. R. Peroxin-dependent targeting of a lipid-droplet-destined membrane protein to ER subdomains. Nat. Cell Biol. 18, 740–751 (2016).
Article CAS PubMed PubMed Central Google Scholar
Li, N. & Jarvis, R. P. Recruitment of Cdc48 to chloroplasts by a UBX-domain protein in chloroplast-associated protein degradation. Nat. Plants 10, 1400–1417 (2024).
Article CAS PubMed PubMed Central Google Scholar
Deruyffelaere, C. et al. PUX10 Is a CDC48A adaptor protein that regulates the extraction of ubiquitinated oleosins from seed lipid droplets in Arabidopsis. Plant Cell 30, 2116–2136 (2018).
Article CAS PubMed PubMed Central Google Scholar
Kretzschmar, F. K. et al. PUX10 is a lipid droplet-localized scaffold protein that interacts with CELL DIVISION CYCLE48 and is involved in the degradation of lipid droplet proteins. Plant Cell 30, 2137–2160 (2018).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Z. R., Bonifacino, J. S. & Hegde, R. S. Deubiquitinases sharpen substrate discrimination during membrane protein degradation from the ER. Cell 154, 609–622 (2013).
Article CAS PubMed PubMed Central Google Scholar
Margottin, F. et al. A novel human WD protein, h-beta TrCp, that interacts with HIV-1 Vpu connects CD4 to the ER degradation pathway through an F-box motif. Mol. Cell 1, 565–574 (1998).
Article CAS PubMed Google Scholar
Magadan, J. G. et al. Multilayered mechanism of CD4 downregulation by HIV-1 Vpu involving distinct ER retention and ERAD targeting steps. PLoS Pathog. 6, e1000869 (2010).
Article PubMed PubMed Central Google Scholar
Mueller, B., Klemm, E. J., Spooner, E., Claessen, J. H. & Ploegh, H. L. SEL1L nucleates a protein complex required for dislocation of misfolded glycoproteins. Proc. Natl Acad. Sci. USA 105, 12325–12330 (2008).
Article ADS CAS PubMed PubMed Central Google Scholar
Christianson, J. C. et al. Defining human ERAD networks through an integrative mapping strategy. Nat. Cell Biol. 14, 93–105 (2011).
Article PubMed PubMed Central Google Scholar
Zhang, T., Xu, Y., Liu, Y. & Ye, Y. gp78 functions downstream of Hrd1 to promote degradation of misfolded proteins of the endoplasmic reticulum. Mol. Biol. Cell 26, 4438–4450 (2015).
Article CAS PubMed PubMed Central Google Scholar
Shi, J. et al. A technique for delineating the unfolding requirements for substrate entry into retrotranslocons during endoplasmic reticulum-associated degradation. J. Biol. Chem. 294, 20084–20096 (2019).
Article CAS PubMed PubMed Central Google Scholar
Hu, X. et al. RNF126-mediated reubiquitination is required for proteasomal degradation of p97-extracted membrane proteins. Mol. Cell 79, 320–331 e9 (2020).
Article CAS PubMed Google Scholar
Hanzelmann, P., Buchberger, A. & Schindelin, H. Hierarchical binding of cofactors to the AAA ATPase p97. Structure 19, 833–843 (2011).
Article PubMed Google Scholar
Dikic, I., Wakatsuki, S. & Walters, K. J. Ubiquitin-binding domains - from structures to functions. Nat. Rev. Mol. Cell Biol. 10, 659–671 (2009).
Article CAS PubMed PubMed Central Google Scholar
Husnjak, K. & Dikic, I. Ubiquitin-binding proteins: decoders of ubiquitin-mediated cellular functions. Annu Rev. Biochem 81, 291–322 (2012).
Article CAS PubMed Google Scholar
Blythe, E. E., Gates, S. N., Deshaies, R. J. & Martin, A. Multisystem Proteinopathy Mutations in VCP/p97 Increase NPLOC4.UFD1L Binding and Substrate Processing. Structure 27, 1820–1829.e4 (2019).
Article CAS PubMed PubMed Central Google Scholar
Tarcan, Z., Poovathumkadavil, D., Skagia, A. & Gambus, A. The p97 segregase cofactor Ubxn7 facilitates replisome disassembly during S-phase. J. Biol. Chem. 298, 102234 (2022).
Article CAS PubMed PubMed Central Google Scholar
Kochenova, O. V., Mukkavalli, S., Raman, M. & Walter, J. C. Cooperative assembly of p97 complexes involved in replication termination. Nat. Commun. 13, 6591 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Okatsu, K., Kimura, M., Oka, T., Tanaka, K. & Matsuda, N. Unconventional PINK1 localization to the outer membrane of depolarized mitochondria drives Parkin recruitment. J. Cell Sci. 128, 964–978 (2015).
CAS PubMed PubMed Central Google Scholar
Ordureau, A. et al. Quantitative proteomics reveal a feedforward mechanism for mitochondrial PARKIN translocation and ubiquitin chain synthesis. Mol. Cell 56, 360–375 (2014).
Article CAS PubMed PubMed Central Google Scholar
Wang, C. W. & Lee, S. C. The ubiquitin-like (UBX)-domain-containing protein Ubx2/Ubxd8 regulates lipid droplet homeostasis. J. Cell Sci. 125, 2930–2939 (2012).
CAS PubMed Google Scholar
Suzuki, M. et al. Derlin-1 and UBXD8 are engaged in dislocation and degradation of lipidated ApoB-100 at lipid droplets. Mol. Biol. Cell 23, 800–810 (2012).
Article CAS PubMed PubMed Central Google Scholar
Nejatfard, A. et al. Derlin rhomboid pseudoproteases employ substrate engagement and lipid distortion to enable the retrotranslocation of ERAD membrane substrates. Cell Rep. 38, 110578 (2022).
Article CAS PubMed Google Scholar
Hartmann-Petersen, R. et al. The Ubx2 and Ubx3 cofactors direct Cdc48 activity to proteolytic and nonproteolytic ubiquitin-dependent processes. Curr. Biol. 14, 824–828 (2004).
Article CAS PubMed Google Scholar
Schuberth, C. & Buchberger, A. UBX domain proteins: major regulators of the AAA ATPase Cdc48/p97. Cell Mol. Life Sci. 65, 2360–2371 (2008).
Article CAS PubMed PubMed Central Google Scholar
Waltho, A. et al. K48- and K63-linked ubiquitin chain interactome reveals branch- and length-specific ubiquitin interactors. Life Sci. Alliance 7, e20240274 (2024).
Magadan, J. G. & Bonifacino, J. S. Transmembrane domain determinants of CD4 Downregulation by HIV-1 Vpu. J. Virol. 86, 757–772 (2012).
Article CAS PubMed PubMed Central Google Scholar
Pierce, N. W., Kleiger, G., Shan, S. -o & Deshaies, R. J. Detection of sequential polyubiquitylation on a millisecond timescale. Nature 462, 615–619 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Wang, L. et al. TMUB1 is an endoplasmic reticulum-resident escortase that promotes the p97-mediated extraction of membrane proteins for degradation. Mol Cell 82, 3453–3467 (2022).
Blythe, E. E., Olson, K. C., Chau, V. & Deshaies, R. J. Ubiquitin- and ATP-dependent unfoldase activity of P97/VCP•NPLOC4•UFD1L is enhanced by a mutation that causes multisystem proteinopathy. Proceedings of the National Academy of Sciences 114, E4380–E4388 (2017).
Pan, M. et al. Structural insights into Ubr1-mediated N-degron polyubiquitination. Nature 600, 334–338 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Stefanovic, S. & Hegde, R. S. Identification of a targeting factor for posttranslational membrane protein insertion into the ER. Cell 128, 1147–1159 (2007).
Article CAS PubMed Google Scholar
Mariappan, M. et al. A ribosome-associating factor chaperones tail-anchored membrane proteins. Nature 466, 1120–1124 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Saha, A. & Deshaies, R. J. Multimodal activation of the ubiquitin ligase SCF by Nedd8 conjugation. Mol. Cell 32, 21–31 (2008).
Article CAS PubMed PubMed Central Google Scholar
Li, T., Pavletich, N. P., Schulman, B. A. & Zheng, N. High-level expression and purification of recombinant SCF ubiquitin ligases. Methods Enzymol. 398, 125–142 (2005).
Article CAS PubMed Google Scholar
Li, C. et al. Biosynthesis of long polyubiquitin chains in high yield and purity. Anal. Biochem 664, 115044 (2023).
Article CAS PubMed Google Scholar

Download references

Acknowledgements

We thank the staff members of the Mass Spectrometry System at the National Facility for Protein Science in Shanghai (NFPS), China for providing technical support and assistance in data collection. We also thank Min Zhuang (ShanghaiTech University) for providing plasmids and Edward B. Lee (University of Pennsylvania) for valuable discussions. This work was supported by National Natural Science Foundation of China (Grant 32570911 and 32270824) and the Shanghai Basic Research Pioneer Project.

Author information

These authors contributed equally: Xin-Yu Huo, Di Liu, Rong Zou.

Authors and Affiliations

Interdisciplinary Research Center on Biology and Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, 100 Haike Road, Bldg. 13, Rm. 340, Pudong New District, Shanghai, China
Xin-Yu Huo, Di Liu, Rong Zou, Zhao-Peng Li, Yunxia Li, Yaoyang Zhang & Zai-Rong Zhang
University of Chinese Academy of Sciences, 19 Yuquan Road, Shijingshan District, Beijing, China
Xin-Yu Huo, Di Liu, Rong Zou & Zhao-Peng Li
State Key Laboratory of Chemical Biology, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai, China
Lifeng Pan

Authors

Xin-Yu Huo
View author publications
Search author on:PubMed Google Scholar
Di Liu
View author publications
Search author on:PubMed Google Scholar
Rong Zou
View author publications
Search author on:PubMed Google Scholar
Zhao-Peng Li
View author publications
Search author on:PubMed Google Scholar
Yunxia Li
View author publications
Search author on:PubMed Google Scholar
Lifeng Pan
View author publications
Search author on:PubMed Google Scholar
Yaoyang Zhang
View author publications
Search author on:PubMed Google Scholar
Zai-Rong Zhang
View author publications
Search author on:PubMed Google Scholar

Contributions

D.L. discovered and characterized the interaction between the HD domain and ubiquitin chains (Figs. 2B, C, 3B–D, 5A–J, 6E, Supplementary Figs. 4C, 4E, F, 5B, 6A–C). R.Z. conducted analyses of HD conservation and experiments on the extraction of OMM proteins (Figs. 2E, 7F–I, andSupplementary Fig. 4B, D, 7A–F). Z.-P.L. performed the experiment in Fig. 2H and the SEC analysis of the p97-UNF complex (Supplementary Fig. 5A). X.-Y.H. performed all the other experiments. Y.L., L.P., and Y.Z. provided technical assistance. All authors analyzed the data. Z.-R.Z. supervised the research and wrote the manuscript with input from all authors.

Corresponding author

Correspondence to Zai-Rong Zhang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Yihong Ye, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (download PDF )

Reporting Summary (download PDF )

Transparent Peer Review file (download PDF )

Source data

Source Data (download ZIP )

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Huo, XY., Liu, D., Zou, R. et al. A UBH-UBX module amplifies p97/VCP’s unfolding power to facilitate protein extraction and degradation. Nat Commun 16, 10162 (2025). https://doi.org/10.1038/s41467-025-65166-4

Download citation

Received: 03 May 2025
Accepted: 07 October 2025
Published: 19 November 2025
Version of record: 19 November 2025
DOI: https://doi.org/10.1038/s41467-025-65166-4