Quantitative Proteomics Unveils Comprehensive Tissue-Specific VCP Interaction Networks in Mice

Abstract

Valosin-containing protein (VCP), a conserved AAA ATPase hexamer, participates in multiple biological processes including ERAD, ubiquitin-dependent degradation by extracting misfolded proteins for proteasomal degradation. Although its interactions with cofactors are well-characterized, and its dysregulation is implicated in multisystem proteinopathy, amyotrophic lateral sclerosis, and cancer, the tissue-specific VCP interactomes underlying its functional versatility remain elusive. Here, we generated HA-N-tagged VCP knock-in mice via CRISPR/Cas9 strategy and performed affinity purification coupled with data-independent acquisition (DIA) mass spectrometry to systematically profile VCP interactors across eight mouse tissues, yielding a high-confidence dataset. We identified 923 robust VCP-binding partners, including established interactors (UBX2B, UFD1, proteasomal subunits) and novel candidates implicated in energy metabolism (TCA cycle, oxidative phosphorylation) and protein quality control (proteasome, ERAD). Notably, we validated the interaction of VCP to two hepatic candidate proteins, DAXX and PRKAG2 (AMPK γ2 regulatory subunit), using HepG2 cells. This study establishes the first in vivo atlas of the VCP interaction network, providing mechanistic insights into its tissue-specific roles and highlighting potential therapeutic avenues for VCP-related disorders.

Background & Summary

Valosin-containing protein (VCP), also known as P97, is an essential AAA + (ATPases associated with various cellular activities) ATPase in eukaryotes. First identified as a ring-shaped hexameric ATPase in 1990¹, VCP is evolutionarily conserved, ubiquitously and abundantly expressed across tissues, and intricately linked to multiple cellular processes². It regulates a series of cellular functions, including protein homeostasis³, autophagy⁴ DNA replication⁵, immune signal transduction⁶, cell cycle progression, and apoptosis⁷. As one of the most abundant cytosolic proteins in mammals⁸, VCP exhibits high-level expression in diverse tissues such as the liver, heart, muscle, kidney, lung, brain, spleen and testis.

VCP consists of three structural domains: an N-terminal substrate-binding domain and two ATPase domains, D1 and D2. Assembled as a large ring-shaped hexamer, VCP executes its intracellular functions through the coordinated actions of these domains. The D1 and D2 domains each form a hexameric ATPase ring⁹, connected by a flexible short polypeptide (D1-D2 linker), while another linker (N-D1 linker) connects the D1 domain to the N-terminus¹⁰. The C-terminus of the D2 domain features an unstructured tail (~40 residues) that regulates ATPase activity⁹. Structurally, the N-terminal domain binds diverse cofactors and adaptor proteins. The D1 domain stabilizes the hexameric assembly, serving as the primary oligomerization module. The D2 domain is essential for full ATPase activity and biological function. The highly flexible C-terminal domain undergoes tyrosine phosphorylation, which modulates membrane fusion and nuclear translocation¹¹. Mechanistically, VCP interacts with cofactors such as UFD1 and NPL4¹². This complex recognizes ubiquitinated substrate proteins, threading them through the central pore formed by the D2 ATPase ring. ATP hydrolysis-driven conformational changes unfold the substrate, which is then transferred to the proteasome for degradation, ensuring clearance of misfolded proteins¹³.

VCP orchestrates diverse cellular processes central to proteostasis, including endoplasmic reticulum-associated degradation (ERAD), mitochondria associated degradation (MAD), ribosome quality control (RQC), chromatin remodeling and autophagy. In addition, VCP is closely related to membrane fusion as well as vesicular transport^10,14–16. Owing to the indispensable role of VCP in orchestrating cellular protein homeostasis, genetic mutations in its chromatin are capable of compromising its normal function in cellular protein quality control, giving rise to endoplasmic reticulum stress, mitochondrial damage, and cell death, thereby culminating in the onset of multi-system degenerative diseases¹⁷. Multi-system degenerative disorders induced by VCP-related mutations predominantly encompass inclusion body myopathy (IBM), frontotemporal dementia (FTD), Paget’s disease of bone (PDB), and amyotrophic lateral sclerosis (ALS)^18,19. Beyond this, emerging investigations have unveiled that VCP assumes a pivotal role in a plethora of cardiovascular diseases, such as myocardial ischemia-reperfusion injury, cardiac hypertrophy, and heart failure^20,21. VCP modulates the proliferation and migration of cancer cells through interactions with a myriad of cofactors, contributing to the pathogenesis and progression of diverse malignancies²². It has been documented that VCP is overexpressed in multiple cancer types, particularly hepatocellular carcinoma²³, and exhibits an association with prognosis, positioning it as a potential cancer biomarker. Concurrently, a multitude of studies have reported VCP and its cofactors as novel therapeutic targets for cancer²⁴.

However, current investigations into VCP-binding proteins and their molecular mechanisms in disease pathogenesis have been predominantly confined to the ubiquitin-proteasome system (UPS)²⁵, creating a prominent gap in the comprehensive characterization of the global VCP interactome, particularly regarding the exploration of its associations with pathways beyond protein degradation. Additionally, therapeutic strategies for VCP-associated disorders have largely centered on the development of direct VCP inhibitors and proteasome inhibitors²⁶. Direct VCP inhibitors often exhibit cross-inhibitory activity against other AAA + ATPases, thereby predisposing to off-target effects²⁴, whereas proteasome inhibitors are plagued by the emergence of therapeutic resistance²⁷. To date, only CB-5083 has shown clinical responsiveness, with its off-target effects impeding translational advancement; most other VCP inhibitors have failed to enter clinical studies due to pleiotropic impacts on AAA + ATPases and ill-defined mechanisms²⁵. In recent years, several emerging approaches designed to enhance therapeutic specificity have been developed, which primarily target VCP-interacting proteins, including E3 ligases, de-ubiquitinating enzymes (DUBs), and the SPRTN protease. This provides a novel direction with substantial therapeutic potential for diseases associated with VCP dysregulation. Nevertheless, exploration of strategies targeting tissue-specific VCP interactors or dysregulated VCP-interactor axes remains remarkably limited, with no achievable therapeutic efficacy in vivo.

Therefore, this study aims to systematically map the distribution of VCP-binding proteins across multiple mouse tissues, generate a spatiotemporal VCP interaction network, and identify tissue-specific and conserved binding partners. Using HA-tagged VCP mice, we collected eight major tissues including liver, heart, muscle, kidney, lung, brain, spleen and testis. Through affinity purification (AP) coupled with DIA-MS, we enriched and identified 923 candidate VCP-binding proteins in these tissues. Enrichment analysis revealed these interactors participate in energy metabolism and protein homeostasis, with significant organ-specific differences in potential partners that liver and kidney interactomes enriched in metabolic pathways, while muscle interactomes favored adrenergic signaling. These findings may elucidate VCP-driven molecular mechanisms underlying multiple diseases, provide novel therapeutic targets, and facilitate discovery of alternative therapeutic approaches beyond current core VCP-targeting strategies.

Methods and Materials

Cell culture

HepG2 (QuiCell-H086,) and HEK293T (QuiCell-2342) cells were cultured in high-glucose DMEM medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37 °C with 5% CO₂. HepG2 cell confluency reached ~90%, subculture was performed. For co-immunoprecipitation (Co-IP) assays, cells from three culture dishes were harvested and lysed using IP lysis buffer. The protein concentration of the resulting lysate was determined by BCA assay (Thermo Fisher, 23228). Subsequently, the protein solution was divided into two aliquots to perform Co-IP experiments to validate protein-protein interactions. 293 T cells are used for immunofluorescence.

VCP^N-HA mouse strain generation and validation

The N-HA-tagged VCP (VCP^N-HA) mouse strain used in this study was developed by the Genome Tagging Project (GTP)²⁸ at the Center for Excellence in Molecular Cell Science, Chinese Academy of Sciences, Shanghai. The GTP identification number for this strain is 20000948. The strain was generated via CRISPR/Cas9-mediated in situ tagging of the Vcp gene in male haploid embryonic stem cells (ESCs), with an HA tag inserted at the N-terminus of VCP. The tagged haploid ESCs were then introduced into oocytes by microinjection to establish the transgenic line. All tagged mice used in experiments were genotyped and validated by Western blotting (WB) to confirm the genotype and expression of the HA-tagged VCP.

All animal experimental protocols used in this study were approved by the Animal Protection and Utilization Committee of the Institute of Animal Models, Nanjing University, China. Experiments were conducted in strict accordance with the approved guidelines. Tagged mice were housed under specific pathogen-free (SPF) conditions with a 12-hour light/dark cycle (lights turn on at 8:00 am and turn off at 8:00 pm), controlled temperature and humidity, and provided with standard rodent chow and ad libitum access to water.

Hepatic perfusion and multi-organ collection in mice

Eight-week-old male mice were used for tissue collection. Prior to tissue harvesting, hepatic saline perfusion was performed to eliminate blood interference. Mice were anesthetized with isoflurane gas, placed supine on an operating table, and the abdominal cavity was incised in a U-shape. Abdominal viscera were retracted to the left to expose the inferior vena cava and hepatic portal vein. Pre-warmed (37 °C) saline (approximately 40 mL total volume) was perfused via the inferior vena cava until the liver swelled and lightened in color, at which point the hepatic portal vein was incised for blood drainage. Perfusion ceased when the liver turned tan. Thoracic and cranial cavities were then opened to harvest the liver, heart, muscle, kidney, lung, brain, spleen and testis. The extracted tissue was placed in tissue cryopreservation tubes, rapidly frozen using liquid nitrogen, and then transferred to a −80 °C freezer for storage.

Genotyping

Mouse tissue samples were incubated in 200 μL of 1.5 mol/L NaCl and 0.5 mol/L NaOH, then heated at 95 °C for 30 min on a thermal platform. The reaction was terminated by adding 20 μL of 5 mol/L NaCl and 0.5 mol/L Tris-HCl (pH 7.0), yielding DNA fragments. Using this DNA as a template, the HA-tagged VCP fragment was amplified by PCR with the following parameters in Table 1.

Table 1.

Primer list.

Primer	Sequence
HA-Forward	ATCCGTATGATGTGCCGGATTAT
HA- Reverse	AACAGGTCTCCCACAGTGAAAAA
WT- Forward	TCCTGCCTCTCATCTTCCCA
WT-Reverse	GGCTCCAGAGGCCATGGC

The amplified PCR products were resolved by 1% agarose gel electrophoresis at constant voltage of 130 V for 30 minutes, followed by visualization.

Protein sample preparation and affinity purification

The eight harvested mouse tissues including liver, heart, muscle, kidney, lung, brain, spleen and testis were lysed in IP lysis buffer (25 mM Tris-HCl, pH 7.5; 150 mM NaCl; 2 mM EDTA; 1% Triton X-100; 5% glycerol) supplemented with protease inhibitor cocktail (MCE, HY-K0011) and PMSF. For each 50 mg of mice tissue, 1 mL of lysis buffer and three steel beads (Servicebio, G0103-200G) were added, followed by grinding three times in homogenizer (60 Hz, 60 s). Lysates were incubated on ice for 30 min, then centrifuged at 12,000 rpm for 15 min at 4 °C, and the supernatants were collected. Protein concentrations were determined by BCA assay and adjusted to 1 μg/μL. For each 2 mg of protein, 2.5 μg of anti-HA (Santa Cruz, sc-7392) or anti-IgG antibody (Santa Cruz, sc-2025) was added, and samples were rotated overnight at 4 °C. The next day, 50 μL of Protein A/G PLUS-Agarose beads (Santa Cruz, sc-2003), pre-washed three times with IP buffer, were added and incubated for 4 h at 4 °C. After centrifugation (3,000 × g, 3 min, 4 °C), the supernatants were collected as flow-through fractions. Beads were washed three times with IP buffer, then 20 μL beads were mixed with 20 μL of 2 × loading buffer and denatured by heating at 95 °C for 10 min. The remaining 30 μL beads were used for on beads digestion. Samples were subjected to Western blotting and mass spectrometry analysis.

Western blotting

A 10% polyacrylamide gel was used for electrophoresis, and samples were divided into input, IP, and FT groups with loading volumes of 10 μL (input/FT groups) and 5 μL (IP group). Electrophoresis was performed at 80 V for 30 min, then adjusted to 120 V after the samples entered the separating gel until the bromophenol blue front reached the bottom of the gel. A 0.22 μm PVDF membrane was activated with methanol, followed by wet transfer at 300 mA constant current for 90 min under ice bath conditions. The membrane was blocked with 5% skim milk (prepared with TBST) at room temperature for 1 h, then incubated with primary antibodies (anti-HA 1:500, anti-VCP 1:2000, anti-β-actin 1:10000) at 4° C overnight. After 3 washes with TBST, the membrane was incubated with HRP-labeled secondary antibody (1:10000) at room temperature, then washed again. ECL luminescent solution was added for 1–2 min dark reaction, and finally exposed with a developer.

Coomassie blue staining

After electrophoresis, the gel was immersed in 0.25% Coomassie Brilliant Blue R-250 staining solution (composed of 45% methanol, 10% acetic acid, and deionized water) and incubated at 4 °C overnight. The gel was destained 3–4 times with destaining solution (same composition as staining solution without dye), with each destaining step to ensure clear background and distinct protein bands. The gel was rinsed with deionized water and then scanned for imaging.

Immunofluorescence

HEK293T cells were seeded on sterile glass coverslips in 12-well plates and cultured overnight. Cells were then transiently transfected with plasmids encoding wild-type VCP or HA-tagged VCP using a standard transfection reagent. At 24 h post-transfection, the cells were washed with PBS, fixed with 4% paraformaldehyde for 15 min at room temperature, and permeabilized with 0.5% Triton X-100 in PBS for 10 min. After blocking with normal goat serum for 30 min at room temperature, the cells were incubated with rabbit anti-HA antibody (Proteintech, 51064-2-AP, 1:200) for 2 h at room temperature. Following three washes with PBS, the cells were incubated with Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:200) for 1 h at room temperature in the dark. After three additional washes with PBS, the cells were then incubated with mouse anti-VCP antibody (Proteintech, 60316-1-Ig, 1:200) overnight at 4 °C. The next day, the cells were washed three times with PBS and incubated with Alexa Fluor 594-conjugated goat anti-mouse IgG (1:200) for 1 h at room temperature in the dark. After three final washes with PBS, the coverslips were mounted onto glass slides using a mounting medium containing DAPI to stain the nuclei. Fluorescence images were acquired using a Zeiss LSM900 laser scanning confocal microscope.

Filter-aided sample preparation (FASP) digestion

100 μg protein solution was placed in a 1.5 mL Eppendorf tube. For reduction and alkylation, 2 μL reducing reagent (50 mM TCEP) was added per 100 μg protein, followed by incubation at 55 °C for 1 h in a water bath. Then 1 μL cysteine-blocking reagent (50 mM MMTS) was added; the mixture was vortexed and incubated at room temperature for 30 min in the dark. The protein solution was transferred to a 10 K ultrafiltration tube (Sartorious, VN01H02) and centrifuged at 9,900 rpm at 4 °C until the solution collected at the bottom. Next, 100 μL 8 M urea was added, centrifuged under the same conditions, and the filtrate was discarded; this step was repeated three times. Subsequently, 100 μL 0.5 M TEAB was added, centrifuged as above, and the filtrate was discarded; this step was repeated at least six times until all solution was at the tube bottom. The ultrafiltration tube was transferred to a new collection tube, and 60 μL 0.5 M TEAB plus 25 μL trypsin (0.2 μg/μL) was added. The tube was sealed with parafilm and incubated at 37 °C overnight. The next day, the ultrafiltrate was collected by centrifugation. An additional 40 μL 0.5 M TEAB was added to the tube, centrifuged, and the eluates were combined. The solution was lyophilized, stored at −80 °C, and reconstituted with loading buffer prior to mass spectrometric analysis.

DIA-MS and database searching

Mass spectrometry analysis was performed on an ACQUITY UPLC M-Class system (Waters) coupled to a ZenoTOF 7600 mass spectrometer (SCIEX, USA). 0.1 μg/μL peptide samples were injected (4 μL) onto an ACQUITY BEH C18 column (1.7 μm, 2.1 × 150 mm, 37 °C, VDAP-4025PBS-631-100, CNW). Peptide separation used buffer A (0.1% formic acid (FA) in ultrapure water) and buffer B (0.1% FA, 2% H₂O in acetonitrile (ACN)) with a gradient: 3%–80% B over 46 min, 80% B wash (2 min), 3% B re-equilibration (7 min); flow rate 5 μL/min. MS operated in positive-ion SWATH mode (85 variable windows, m/z 400–900). ESI parameters: spray voltage 5,000 V, source temperature 200 °C, desolvation temperature 350 °C, gas flows 20/60/35 psi (nebulizing gas/desolvation gas/cone gas); Zeno Trap enabled. MS¹: m/z 350–1,500, ≥ 45,000 resolution, 100 ms; MS²: m/z 200-1,500, ≥ 42,000 resolution, 13 ms, dynamic collision energy (15–60 eV). Raw data (.wiff) were processed via DIA-NN (UniProt database, version 1.8.1). Quantification based on MS² XIC (median normalization); FDR < 1% (peptides/proteins).

Bioinformatics analysis

All the missing values in the dataset were imputed with the minimum value 0.1. For the screening of candidate binding proteins, a two-tailed Student’s t-test was employed with significance defined as P < 0.05 (two-tailed Student’s t-test) and fold change (FC) > 1.2. For the overall analysis, the protein abundance of HA pull-down samples was subtracted by that of IgG control samples to perform background subtraction, and these processed data were used for subsequent bioinformatics analysis. Quantitative data were visualized using GraphPad Prism 8.0.1 and RStudio. Principal component analysis (PCA) was plotted using the RStudio program. Pathway enrichment analysis of differentially expressed proteins was performed via STRING database, where candidate proteins meeting criteria were uploaded with Mus musculus specified. Enrichment significance for Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways was determined by hypergeometric test (FDR < 0.05), with results visualized in RStudio. All methods are reported in accordance with ARRIVE guidelines.

Data Records

To ensure research transparency and reproducibility, all datasets generated/analyzed in this study are accessible: raw DIA-MS data of VCP-interacting proteins from eight murine tissues (liver, heart, muscle, kidney, lung, brain, spleen and testis) have been deposited in the iProX^29,30 (https://www.iprox.org/) with the identifier IPX0014478000. Simultaneously, iProX uploaded this project to ProteomeXchange and applied for a PXD number: PXD071501³¹. For raw files acquired by the ZenoTOF 7600, each individual sample comprises three files: wiff, wiff2, wiffscan and timeseries.data. Output files generated by the DIA-NN software are uploaded concurrently. DIA-NN produced eight tsv output files, with each file named after the specific tissue it represented. This work ultimately summarized candidate binding proteins identified across eight tissues. Additionally, we have compiled the peptide information corresponding to the identified proteins generated by DIA-NN across the eight tissues. Both of these files can be found in the dataset we uploaded to ProteinXchange.

Technical Validation

To investigate the VCP binding protein networks in different mouse tissues, we collaborated with the GTP in Shanghai to generate Vcp^N-HA transgenic mice using CRISPR/Cas9 technology²⁸ (Fig. 1a). The expression of the HA tag at both the gene and protein levels was validated through genotyping and Western blotting. Gene-level analysis confirmed high expression of both the HA tag and Vcp (Fig. 1b). Eight major tissues including liver, heart, kidney, lung, brain, spleen, muscle and testis, were collected for protein extraction. Western blotting revealed specific expression of HA-tagged VCP in all tested tissues, demonstrating VCP^N-HA expression. Coomassie brilliant blue staining of total protein lysates served as a loading control for tissue samples (Fig. 1c). To investigate whether the HA tag affects VCP expression and subcellular localization, we performed western blotting and immunofluorescence assays. Western blot analysis showed that VCP expression levels were comparable between wild-type (WT) mice and Vcp^N-HA tagged mice (Fig. 1d). Immunofluorescence staining further revealed that the overexpressed HA-tagged VCP exhibited a subcellular localization pattern identical to that of untagged VCP (Fig. 1e).

Fig. 1.

Construction of VCP^N-HA mice and expression validation. (a) Strategy for constructing VCP^N-HA mice with CRISPR/Cas9 technology. (b) Phenotypic validation of the HA tag and VCP gene via PCR. (c) Expression validation of HA tagged VCP in eight tissues of mice by WB using both HA and VCP antibodies. (d) Verification of VCP expression in wild-type and HA-tagged VCP mice. (e) Immunofluorescence analysis of VCP localization in 293 T cells overexpressing wild-type and HA-tagged VCP.

To systematically characterize VCP-binding proteins across eight mouse tissues, we established an integrated workflow combining affinity purification (AP) with DIA-MS (Fig. 2a). Western blotting validated efficient enrichment of HA-tagged VCP in all tissues (Fig. 2b). The enriched samples were processed using the FASP digestion method to obtain peptide fragments, and resulting peptides were analyzed by DIA-MS. All the above experiments, including tissue collection, protein extraction, enrichment of binding proteins, and FASP digestion, involved four biological replicates. For the mass spectrometry identification step, each biological sample was run in triplicate to generate three technical replicates, and finally 12 replicates were obtained for both the IgG group and the HA group. Differential protein screening (Fold Change > 1.2, p < 0.05) revealed tissue-specific VCP interactomes: liver (1,015 total proteins, 200 upregulated proteins), heart (769 total proteins, 114 upregulated proteins), muscle (1,090 total proteins, 57 upregulated proteins), kidney (791 total proteins, 185 upregulated proteins), lung (1,009 total proteins, 138 upregulated proteins), brain (1,566 total proteins, 184 upregulated proteins), spleen (1,308 total proteins, 121 upregulated proteins), and testis (828 total proteins, 187 upregulated proteins) (Fig. 2c). To evaluate the reproducibility of the 12 replicates in both the IgG and HA groups across each tissue type, we conducted quality control analyses including correlation heatmaps (Figure S1) and principal component analysis (PCA, Figure S2). To visualize the protein expression profiles across the eight tissue types, volcano plots were generated (Figure S3). VCP was found to be among the proteins with the highest enrichment levels. To provide a quantitative summary, we present a table (Table 2) reporting the absolute protein abundance and the fold change of VCP. The t-SEN analysis showed partial clustering of tissue samples, indicating both shared and distinct proteomic profiles (Fig. 2d), while heatmap visualized significant inter-tissue expression disparities (Fig. 2e).

Fig. 2.

Pipeline and quality control of affinity purification-mass spectrometry. (a) Schematic workflow of AP-MS. (b) WB validation of AP enrichment efficiency. (c) Number of proteins identified and significantly up-regulated proteins (P < 0.05, Fold Change > 1.2) in each of the eight tissues. (d) t-SNE analysis of protein expression profiles across eight mouse tissues. (e) Heatmap shows proteins identified in each of the eight tissues (after deduction of the background).

Table 2.

Expression levels and statistical difference analysis of VCP between IgG and HA groups across 8 tissues.

Tissue	Group	Mean ± SD	Fold Change (HA/IgG)	P value	Padjust_fdr
Liver	IgG	49.94 ± 10.33	226.69	6.43E-07	1.45E-05
	HA	11321.86 ± 5668.19
Heart	IgG	90.04 ± 113.17	112.25	9.5E-07	1.22E-05
	HA	10107.13 ± 5166.79
Muscle	IgG	95.16 ± 23.99	47.78	2.25E-07	1.06E-05
	HA	4546.79 ± 2092.92
Kidney	IgG	35.85 ± 15.07	201.84	1.3E-11	5.7E-10
	HA	7235.98 ± 1961.57
Lung	IgG	32.33 ± 13.96	292.56	5.22E-06	7.22E-05
	HA	9459.6 ± 5469.74
Brain	IgG	34.5 ± 20.64	186.18	9.4E-05	0.002066
	HA	6422.4925 ± 4647.39
Spleen	IgG	48.68 ± 4.09	88.50	4.12E-06	0.000131
	HA	4307.84 ± 2429.86
Testis	IgG	37.58 ± 7.77	240.10	3.49E-08	7.04E-07
	HA	9023.05 ± 3770.33

We next summarized the upregulated proteins from all eight tissues, yielding 923 candidate VCP-binding proteins. The candidate binding proteins we identified include a great number of previously reported VCP adapters (such as UBX2B, NSF1C, UFD1, UBC) and proteasome subunits (such as PSMD1, PSMD2, PSMD3). We performed comprehensive characterization of all candidate proteins, heatmap visualized tissue-specific expression patterns of these candidates (Fig. 3a), Pearson correlation heatmap was constructed using normalized protein abundance data to assess inter-tissue similarity of VCP interactomes (Fig. 3b). For functional characterization, we performed bioinformatics analyses including GO enrichment (Fig. 3c) and KEGG pathway analysis (Fig. 3d). The functions of the potential binding proteins we identified encompass known VCP functions, such as protein quality control (proteasome, ERAD), which supports the reliability of our data.

Fig. 3.

Summary of DIA quantitative proteomics results of VCP interaction proteins. (a) Heatmap showing the expression of the candidate VCP binding proteins in eight tissues. (b) Correlation heatmap showing the tissue correlation of candidate VCP binding proteins. (c) Top 10 of the three annotations of GO for all candidate VCP binding proteins are also shown. (d) Top 30 KEGG analysis of all candidate VCP binding proteins.

Additionally, based on the correlation heatmap, we selected the shared proteins between liver and testis for bioinformatics analysis. The GO analysis results (Fig. 4a) showed that these shared proteins are associated with biological processes such as the ubiquitin-dependent ERAD pathway and regulation of protein catabolic process. Moreover, these proteins were found to be components of the VCP-NSFL1C complex and to exert molecular functions including unfolded protein binding. The KEGG analysis (Fig. 4b) results indicated that these proteins are involved in pathways such as protein processing in the endoplasmic reticulum and the TCA cycle. These findings are consistent with the core functions of VCP in protein unfolding and ERAD.

Fig. 4.

Analysis of VCP interacting proteins shared between liver and testis. (a) GO analysis of VCP interacting proteins shared between liver and testis. (b) KEGG analysis of VCP interacting proteins shared between liver and testis.

To further characterize tissue-specific VCP interactomes, we performed pathway enrichment analysis on upregulated proteins from eight tissues. As a key AAA + ATPase, VCP is indispensable in three core metabolic organs: liver (regulating metabolic homeostasis), kidney (mediating protein processing), and muscle (maintaining energy metabolism and myofiber homeostasis). Therefore, we selected these three tissues covering core metabolic processes, aiming to representatively reveal the metabolism-related functions of VCP and its pathological significance. Enrichment results for liver, kidney, and muscle metabolic organs are shown in Fig. 5 (bioinformatics data for other tissues in Figures S4–5). Liver GO (Fig. 5a) and KEGG analysis (Fig. 5b), kidney GO (Fig. 5c) and KEGG analysis (Fig. 5d), muscle GO (Fig. 5e) and KEGG analysis (Fig. 5f) were significantly enriched in known VCP-associated biological processes and pathways, such as protein catabolism, ERAD, and energy metabolism, which supported the biological relevance of the dataset.

Fig. 5.

Tissue-specific VCP interacting proteins with distinct functions identified in mouse liver, kidney, and muscle. (a) Top 10 of the three annotations in the GO analysis of VCP binding proteins in liver. (b) Top 30 of the KEGG annotations of VCP binding proteins in liver. (c) Top 10 of three annotations in GO analysis of VCP binding proteins in kidney. (d) KEGG annotations of VCP binding proteins in kidney. (e) Top 10 of the three annotations in the GO analysis of VCP binding proteins in muscle. (f) KEGG annotations of VCP binding proteins in muscle.

We selected the HCC cell line HepG2 for validating candidate binding proteins to improve the reliability of our data. Among the liver-enriched candidate proteins, we selected DAXX and PRKAG2 for subsequent binding validation experiments. DAXX is a multifunctional protein that is highly expressed in various cancer types and closely associated with cancer initiation, progression, and therapeutic resistance³². PRKAG2, the γ regulatory subunit of the AMPK complex, may affect the activity of the AMPK holoenzyme through its binding to VCP. As a core regulator of systemic energy metabolism, AMPK is involved in the pathogenesis of multiple metabolic diseases and cancers³³; meanwhile, targeting AMPK activation has emerged as a promising therapeutic strategy for type 2 diabetes, obesity, and cancer in recent years³⁴. Western blotting results confirmed the binding of VCP to both proteins (Fig. 6). Collectively, these findings indicate that our study provides a reliable database for developing novel therapeutic targets against VCP-binding proteins.

Fig. 6.

Validation of two liver VCP interactors DAXX and PRKAG2. Western blotting was used to verify that VCP binds to DAXX and PRKAG2.

In summary, this study innovatively integrated the CRISPR/Cas9-mediated HA-tagged transgenic mouse model with DIA based quantitative proteomics technology, successfully constructing a spatiotemporal profile of VCP interactomes across eight key tissues in mice. This not only yielded a VCP-binding protein dataset with broad coverage and high reliability, but also provided a novel perspective for in-depth deciphering the tissue-specific functional diversity of VCP, while laying a solid experimental foundation for subsequent VCP-related mechanistic research and translational medicine exploration.

Author contributions

Bing Bai, Lei Fang, Jingzi Zhang and Nannan Wang designed this project together. Nannan Wang was responsible for collecting mouse tissue samples and processing mass spectrometry samples. Yining Li was responsible for mouse maintenance and genotype identification. Na Li was responsible for preliminary analysis of mass spectrometry data. Linru Shen, Chengzhi Wang, Huiming Zhu, Ziyue Zhou and Yibing Ding assisted Nannan Wang in all experimental procedures. All authors participated in the writing and review of the manuscript.

Funding

This work was supported by grants from National Natural Science Foundation of China (82172354, 82472333 to Bing Bai; 32071256, 32371336 to Lei Fang).

Code availability

No new code was used in this study.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41597-026-07626-0.